 

.4 v.11 ‘
arty)
.5

ca

.. ..
.inn

.r
SW“. ‘

and A».
mummmﬂuumﬁz
. gm.

4

m

‘2‘

a ..
14 . kﬁa. n
. a

u‘ 5;»:

‘. l. 1.;

7

i

.3...

.. .
1...... I
;. ...

.1 u. u...

 

e. 3!. as,
‘91)... h .53. r:
; PHI... taut]?
n: I.
u.

91“: an “134.71.... .
a .ﬁ, .ﬁnmmzmv. v

.34 .n;

‘1...

Z.- ‘A ”B E EITNERSITY
MICHIGAN STAT
2N8 EAST LANSING, MICH 48824-1048

93330593

This is to certify that the
dissertation entitled

TOMATO AROMA: INFLUENCE OF FATTY ACID SUBSTRATE AND
HYDROPEROXIDE LYASE ON VOLATILE PROFILE AND SENSORY
QUALITY

presented by
Mauricio Alejandro Caﬁoles

has been accepted towards fulﬁllment
of the requirements for the

PhD degree in Horticulture

 

MSU is an Afﬁnnative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/Date0ue.p65-p.15

 

 

TOMATO AROMA: INFLUENCE OF FATTY ACID SUBSTRATE AND
HYDROPEROXIDE LYASE ON VOLATILE PROFILE AND SENSORY QUALITY
BY

Mauricio Alejandro Caﬁoles

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Horticulture

2004

 

.lll“ I].

ABSTRACT

TOMATO AROMA: INFLUENCE OF FATTY ACID SUBSTRATE AND
HYDROPEROXIDE LYASE ON VOLATILE PROFILE AND SENSORY QUALITY

BY

Mauricio Alejandro Caﬁoles

The aldehydes cis-3-hexenal, hexanal, and trans-Z-hexenal, the alcohols 1-
hexanol, and cis-3-hexenol, and the ketone 1-penten-3-one are produced as a
consequence of lipid degradation following tissue disruption and are among the
most important volatile compounds in tomato aroma. The biosynthesis of cis-3-
hexenal and other volatiles noted involves the action of a sequence of enzymes
including lipase, lipoxygenase (LOX), hydroperoxide Iyase (HPL), isomerase, and
alcohol dehydrogenase (ADH) on glycerolipids containing the fatty acids, linoleic
acid (18:2) and Iinolenic acid (18:3), via the LOX pathway. 18:2 serves as a
precursor to hexanal formation; 18:3 serves as precursor to unsaturated C6-
aldehydes. In the current work, the formation and sensory perception of volatile
compounds was studied in tomato plant lines where (1)3 fatty acid desaturase
(0)3 FAD) and HPL activity was genetically altered. Reduced trienoic and
increased dienoic fatty acid content in leaves and fruits decreased biosynthesis of
cis-3-hexenal and trans-Z-hexenal, and increased hexanal formation. There was
a linear relation between 18:2 content in leaves and production of hexanal, while
the relation between 18:3 and unsaturated C6-aldehyde production better ﬁt a

logarithmic expression. The data suggest that at least one component of the LOX

pathway has a high speciﬁcity for 18:3, resulting in preferential synthesis of
18:3-derived C6-volatiles. This speciﬁcity is likely related to HPL preferential
activity for 18:3. In fruits, C6-volatile production increased with the onset of
ripening, probably related to an enhancement in LOX activity. LeHPL co-
suppression dramatically reduced the production of lipid-derived C6-volatiles in
leaves, but in fruits, only unsaturated C6-volatile production was affected,
suggesting LeHPL-independent formation of hexanal occurs in fruits, but not in
leaves. Increased production of 5-carbon volatiles is proposed as an alternative
way to metabolize 13-hydroperoxy Iinolenic acid in plants with reduced LeHPL
activity. Changes in the volatile proﬁle of leaves and fruits of tomato lines in
which LeFad7 or LeHPL activity is reduced markedly are readily detected by non-
trained sensory panels. Reduction of the unsaturated C6-aldehydes ci5-3-hexenal
and trans-Z-hexenal was detrimental for tomato sensory quality relative to wild
type tissue. An increase in hexanal, without an increase in other LOX pathway-
derived volatiles was also related to a negative preference. The studies
demonstrate that a marked reduction in the activity of one of the most critical
steps in the LOX pathway can markedly impact sensory perception. Efforts to
improve total volatile formation may require the modiﬁcation of LOX pathway at
several steps simultaneously, including precursor formation, and LOX and HPL

activities.

To my parents, Mario Cafioles and Rosa Salvo, and
my wife Marlene Ayala

ACKNOWLEDGMENTS

I would like to thank my advisor Dr. Randy Beaudry for all the support
given through all these years at MSU. I would like to thank the members of my
committee Dr. Robert Herner, Dr. Gregg Howe, and Dr. Dean Della Penna for
their important contribution to this research.

I am especially grateful to Dr. Howe and his laboratory members Sastry
Jayanty, Chuanyou Li and Tony Schilmiller, for letting me use their plant and
genetic material, as well as their constant help in the development of some of
the methods and protocols used in this research. In addition, I would like to
thank Dr. John Ohlrogge and Gustavo Bonaventure for their help in lipid analysis,
and Dr. Janice Harte for her help in the sensory evaluations.

Special thanks to Marisol Soto who has been working hard in the lab and
greenhouse making the tomato plant produce the fruit of this research.

I gratefully acknowledge ﬁnancial support from MIDEPLAN, Gobierno de
Chile, through the scholarship “Beca Presidente de la Republica".

I would like to thank all those “international" friends who where there
along my Ph.D. program, and made easier being far away from family and
friends back home: David Mota and Karen Cichy, Elzette Van Rooyen and Conrad
Schutte, Marcel Lenz, Marcus and Donna Duck, Sonali Padhye, Jose Cisneros,
Costanza Zavalloni and Daniele Trebi, Dario Stefanelli, Mohamed Tawﬁk, Randy

Vos, Gianni Sorrenti, Adriana Nikoloudi, and Rafael Auras. Special thanks to Dr.

Ron Perry and his wife Anne for make us part of their family traditions, and Dr.
Ken Poff for his friendship and support in those moments when we need it the
most.

Finally my eternal gratitude and love to my parents Rosa Salvo and Mario
Cafioles and my sister Catalina Caﬁoles, and to my wife Marlene Ayala for being

there and made possible this goal that we started together 5 years ago.

vi

TABLE OF CONTENT

LIST OF TABLES ............................................................................................. ix
LIST OF FIGURES .......................................................................................... xii
CHAPTER I
INTRODUCTION .................................................................................................. 1
REFERENCES ........................................................................................... 5
CHAPTER II
LITERATURE REVIEW ........................................................................................... 6
LIPOXYGENASE (LOX) PATHWAY .................................................................. 9
VOLATILE FORMATION BY THE TOMATO LIPOXYGENASE PATHWAY ...................... 12
TOMATO LIPOXYGENASES ............................................................... 13
TOMATO HYDROPEROXIDE LYASE ..................................................... 15
FATTY ACID SUBSTRATES ......................................................................... 17
FATTY ACID DESATURASES ....................................................................... 18
REFERENCES ......................................................................................... 21
CHAPTER III
DEFICIENCY OF LINOLENIC ACID OF Lefad7 MUTANT TOMATO CHANGES THE VOLATILE
PROFILE AND SENSORY PERCEPTION OF LEAVES AND FRUITS .......................................... 30
MATERIALS AND METHODS ....................................................................... 34
PLANT MATERIAL ......................................................................... 34
FATTY ACID COMPOSITION ............................................................. 34
EVALUATION OF AROMA COMPOUNDS ................................................ 34
SENSORY EVALUATION ................................................................... 36
STATISTICAL ANALYSIS .................................................................. 37
RESULTS .............................................................................................. 37
FATTY ACID COMPOSITION ............................................................. 37
VOLATILE AROMA COMPOUNDS ........................................................ 38
SENSORY EVALUATION ................................................................... 39
DISCUSSION ......................................................................................... 40
REFERENCES ......................................................................................... 45
CHAPTER IV
GENETIC MANIPULATION OF LINOLENIC ACID CONTENT ALTERS C6-VOLATILE SYNTHESIS AND
SENSORY PERCEPTION OF TOMATO LEAVES AND FRUITS ............................................... 56
MATERIALS AND METHODS ....................................................................... 59
PLANT MATERIAL ......................................................................... 59
FATTY ACID COMPOSITION ............................................................. 60
EVALUATION OF AROMA COMPOUNDS ................................................ 60

Vii

SENSORY EVALUATION ................................................................... 61

STATISTICAL ANALYSIS .................................................................. 62
RESULTS .............................................................................................. 63
FATTY ACID COMPOSITION ............................................................. 63
C6-ALDEHYDE VOLATILES .............................................................. 64
SENSORY EVALUATION ................................................................... 67
DISCUSSION ......................................................................................... 67
REFERENCES ......................................................................................... 73

CHAPTER V
REDUCED LeHPL ACTIVITY INCREASES 1-PENTEN-3-ONE AND DIFFERENTIALLY DECREASES C6-
ALDEHYDES BIOSYNTHESIS IN TOMATO LEAVES AND FRUITS, RESULTING IN AN IMPACT ON

SENSORY PERCEPTION AND PREFERENCE ................................................................. 84

MATERIALS AND METHODS ....................................................................... 87

PLANT MATERIAL ......................................................................... 87

A grobaa‘erium tumefasciens—MEDIATED TRANSFORMATION ............... 87

ADDITION OF RECOMBINANT LeHPL TO FRUIT HOMOGENATE .................. 88

FATTY ACID COMPOSITION ............................................................. 89

EVALUATION OF AROMA COMPOUNDS ................................................ 89

SENSORY EVALUATION ................................................................... 90

STATISTICAL ANALYSIS .................................................................. 91

RESULTS .............................................................................................. 92

FATTY ACID COMPOSITION ............................................................. 92

VOLATILE ANALYSIS .................................................................. 92

SENSORY EVALUATION ................................................................... 94

DISCUSSION ......................................................................................... 95

REFERENCES ......................................................................................... 99
CHAPTER VI

CONCLUSIONS ............................................................................................... 11 1

APPENDIX .................................................................................................. 116

viii

LIST OF TABLES

CHAPTER II

Table 1. Odor descriptors, fatty acid precursor and logarithms of odor units for
C6- and CS-volatile compounds from lipid oxidation in tomato fruit. ...... 27

CHAPTER III

Table 1. Odor descriptors, fatty acid precursor and logarithms of odor units for
C6- and C5-volatiles from lipid oxidation in tomato fruit. ....................... 49

Table 2. Signiﬁcance for triangle test in leaves and fruits, and preference test

in fruits. ............................................................................................. 49
Table 3. Total C6-volatile headspace concentration produced by homogenized

immature green and red mature fruits (nL-L‘l). .................................... 50
CHAPTER IV

Table 1. Ratios between total polyunsaturated octadecanoic acids (umol-g'l)
on a fresh weight basis and total C6-volatiles (nL-L") in leaves.
Abbreviations are as follows: C18 PUFA, polyunsaturated fatty acid of
18 carbons (18:2 and 18:3); Ln, linolenic acid; La, linoleic acid; VFLn,
C6-volatiles derived from linolenic acid; and VFLa, C6-volatiles derived
from linoleic acid. ............................................................................... 76

Table 2. Relative amount of C6-volatiles derived from 18:3 at different fruit

ripening stages (% of total C6-volatiles). ............................................. 76
Table 3. Signiﬁcance for triangle test in leaves and fruits. ............................... 77
Table 4. Signiﬁcance for preference test in leaves and fruits. .......................... 77
CHAPTER V
Table 1. Signiﬁcance for triangle test in leaves and fruits. ............................. 102
Table 2. Signiﬁcance for preference test in leaves and fruits. ........................ 102

APPENDD<

Table 1. Original data for Figure 1 Chapter III. Values in Limol-g‘1 fresh weight
(standard error). .............................................................................. 116

Table 2. Original data for Figure 2 Chapter III. Values in mole percent. ......... 117

Table 3. Original data for Figure 3 Chapter III. Values in nL-L’1 (standard
error). ............................................................................................. 117

Table 4. Original data for Figure 4 Chapter III. Values in nL-L'1 (standard
error). ............................................................................................. 118

Table 5. Original data for Figure 5 Chapter III. Relative content of aldehydes. 118

Table 6. Original data for Figure 1 Chapter N. Values in pmolcg'1 fresh weight
(standard error). .............................................................................. 119

Table 7. Original data for Figure 2 Chapter IV. Values in mole percent
(standard error). .............................................................................. 119

Table 8. Original data for Figure 3 Chapter IV. Values in mole percent
(standard error). .............................................................................. 119

Table 9. Original data for Figure 4 Chapter IV. Values in nL-L'1 (standard
error). ............................................................................................. 120

Table 10. Original data for Figures 5, 6 and 7 Chapter IV. Values in nL-L’1
(standard error). .............................................................................. 120

Table 11. Original data for Figure 8 Chapter IV. Values in pmol-g'1 for FA and
nL-L‘1 for volatiles. ............................................................................ 121

Table 12. Original data for Figure 9 Chapter IV. Values in units of area at mass
108. ................................................................................................ 121

Table 13. Original data for Figure 2 Chapter V. Values in pmol-g'1 fresh weight
(standard error). .............................................................................. 122

Table 14. Original data for Figures 3 and 5 Chapter V. Values in nLoL'1
(standard error). .............................................................................. 122

Table 15. Original data for Figures 4 and 6 Chapter V. Values in nL-L‘1
(standard error). .............................................................................. 123

Table 16. Original data for Figure 7 Chapter V. Values in nL-L'1 (standard
error). ............................................................................................. 123

xi

LIST OF FIGURES

CHAPTER II

Figure 1. Lipoxygenase pathway for C6- and C9-aldehydes and alcohols
biosynthesis. Abbreviations are as follows: 18:2, linoleic acid; 18:3,
linolenic acid; 9-LOX, lipoxygenase with speciﬁcity for peroxydizing
carbon 9; 13-LOX, lipoxygenase with Speciﬁcity peroxydizing carbon 13;
9-HPOD, 9-hydroperoxylinoleic acid; 9-HPOT, 9-hydroperoxy linolenic
acid; 13-HPOD, 13- hydroperoxylinoleic acid; 13-HPOT, 13-
hydroperoxylinolenic acid; HPL, hydroperoxide Iyase; ADH, alcohol
dehydrogenase; Z, cis; E; trans. (Vick and Zimmerman, 1987; Mack et
al., 1987; Hildebrand, 1989; Siedow, 1991; Feussner and Wasternack,
2002) ............................................................................................... 28

Figure 2. Jasmonic acid biosynthesis pathway. Abbreviations are as follows:
18:3, linolenic acid; 13-HPOT, 13-hydroperoxylinolenic acid; 12,13-EOT,
12,13-epoxylinolenic acid; OPDA, 12-oxo-phytodienoic acid; OPC-8:0,
3-oxo-2-(2’(Z)-pentenyI)-cyclopentane-1-octanoic acid; 13-LOX, 13-
Iipoxygenase; AOS, allene oxide synthase; AOC, allene oxide cyclase;
OPR, OPDA reductase. (Creelman and Mullet, 1997) ............................. 29

CHAPTER III

Figure 1. Fatty acid composition of tomato tissue on a fresh weight basis. A,
leaf; B, green immature fruit; and C, red mature wild type and Lefad7
fruit. Vertical bars represent standard error (n between 4 and 6). ......... 51

Figure 2. Fatty acid composition (mole percent). A, leaf; B, green immature
fruit; and C, red mature wild type and Lefaa’7 fruit. .............................. 52

Figure 3. C6-aldehyde concentration in (A) young and (B) mature wild type
and Lefad7leaves (nL-L‘l). Vertical bars represent standard error
(n=4). 2, cis; E, trans. ........................................................................ 53

Figure 4. C6-aldehyde concentration in (A) immature green and (B) mature
red wild type and Lefad7 fruits (nL-L‘l). Vertical bars represent standard
error (n between 4 and 6). ................................................................. 54

Figure 5. Relative C6-aldehyde headspace concentration in leaves, immature
(green) and mature (red) Wild type and Lefad? tomato fruits. ............... 55

xii

CHAPTER IV

Figure 1. Leaf fatty acid composition (umol-g'l) in a fresh weight basis.
Columns within fatty acids with different letters are signicantly different
(P<0.05). Vertical bars represent standard error (n=4). ....................... 78

Figure 2. Fruit fatty acid composition (mole percent). Columns within fatty
acids with different letters are signicantly different (P<0.05). Vertical
bars represent standard error (Wild Type n=20, Lefad7n=17, R183
n=6). ................................................................................................ 78

Figure 3. Relative content of linoleic (18:2) and linolenic acid (18:3) in wild
type, Lefad7 and R183 fruits at different pooled ripening stages: (0-2)
immature green, green, and breaker, (3-6) turning, pink, light red, and
red. Vertical bars represent standard error (n between 4 and 6). .......... 79

Figure 4. C6-Aldehyde headspace concentration (nL-L'l) from leaves. Columns
within aldehyde with different letters are signicantly different (P<0.05).
Vertical bars represent standard error (n=4). 2, cis; E, trans ................. 79

Figure 5. C6-Aldehyde headspace concentration (nL'L'l) from fruits. Columns
within aldehyde with different letters are signicantly different (P<0.05).
Vertical bars represent standard error (n=4). ....................................... 80

Figure 6. Hexanal headspace concentration (nL-L‘l) in fruits at different
ripening stages: (0) immature green, (1) green, (2) breaker, (3)
turning, (4) pink, (5) light red, and (6) red. Vertical bars represent
standard error (n between 3 and 5). Asterisk (*) represent signiﬁcant
difference from wild type (o=0.05). ..................................................... 80

Figure 7. cis-3-Hexenal headspace concentration (nL-L'l) in fruits at different
ripening stages: (0) immature green, (1) green, (2) breaker, (3)
turning, (4) pink, (5) light red, and (6) red. Vertical bars represent
standard error (n between 3 and 5). Asterisk (*) represents signiﬁcant
difference from wild type (o=0.05). Double asterisk (**) represents
signiﬁcant difference from R183 (o=0.05) ............................................ 81

Figure 8. trans-Z-Hexenal headspace concentration (nL-L") in fruits at
different ripening stages: (0) immature green, (1) green, (2) breaker,
(3) turning, (4) pink, (5) light red, and (6) red. Vertical bars represent
standard error (n between 3 and 5). Asterisk (*) represents signiﬁcant
difference from wild type (o=0.05). Double asterisk (**) represents
signiﬁcant difference from R183 (o=0.05) ............................................ 81

xiii

Figure 9. Correlations between fatty acid content and C6-aldehyde
concentration at the headspace from leaves of wild type (:1), Lefad7
(A), and R183 (0) lines: (A) hexanal and 18:2; (B) cis-3-hexenal and
18:3; and (C) trans-Z-hexenal and 18:3. .............................................. 82

Figure 10. 6-Methyl-5-hepten-2-one headspace concentration in fruits at
different ripening stages: (0) immature green, (1) green, (2) breaker,
(3) turning, (4) pink, (5) light red, and (6) red, represented as area
units of the chromatogram peak calculated at mass 108. Vertical bars
represent standard error (n between 3 and 6). .................................... 83

CHAPTER V

Figure 1. Construction 355::LeHPL (A) and conﬁrmation of Agrobacterium-
mediated transformation by PCR in cultivars Castemart and Micro-Tom:
(B) Primers F1 and R1, and (C) F2 and R2. Indicated sized in panel B
correspond to 1Kb DNA ladder (Gibco Life Technology Inc.,
Rockville,MD). R, T-DNA right border; Np, NOS promoter; nptII,
neomycin phosphotransferase II; Nt, NOS terminator; 35$, cauliﬂower
mosaic virus 35$ promoter; L, T-DNA left border ................................ 103

Figure 2. Fatty acid composition (pmoI-g'l) on a fresh weight basis of tomato
leaf CV. Castlemart and CSH lines (CSH24, CSH27 and CSH37). Vertical
bars represent standard error (n=3). ................................................. 104

Figure 3. Headspace concentration of 1-penten-3-one, hexanal, ci5-3-hexenal,
and trans-Z-hexenal for Micro-Tom leaves. Vertical bars represent
standard error of mean (n=3). Asterisk represent signiﬁcant (o=0.05)
difference from wild type. 2, cis; E, trans. .......................................... 105

Figure 4. Headspace concentration of 1-penten-3-one, hexanal, cis-3-hexenal,
and trans-Z-hexenal for Castlemart leaves. Vertical bars represent
standard error of mean (n=3). Asterisk represent signiﬁcant (o=0.05)
difference from wild type. ................................................................. 106

Figure 5. Headspace concentration of 1-penten-3-one, hexanal, cis-3-hexenal,
and trans-Z-hexenal for Micro-Tom fruits. Vertical bars represent
standard error of mean (n=3). Asterisk represent signiﬁcant (o=0.05)
difference from wild type. ................................................................. 107

xiv

Figure 6. Headspace concentration of 1-penten-3-one, hexanal, ci5-3-hexenal,
and trans-Z-hexenal for Castlemart fruits. Vertical bars represent
standard error of mean (n=3). Asterisk represent signiﬁcant (o=0.05)
difference from wild type. ................................................................. 108

Figure 7. Relative production of 1-penten-3-one, hexanal, ciS-3-hexenal, and
trans-Z-hexenal in CSHS fruit homogenate after 3 minutes incubation
with water (CSH5 Control); 50 mM K2PO4 pH 7.5 and 5% glycerol
buffer (CSHS + Buffer); and recombinant LeHPL crude extract (CSHS +
HPL). Vertical bars represent standard error of mean (n=3). Asterisk
represent signiﬁcant (o=0.05) difference from CSHS Control ............... 109

Figure 8. Propose pathway for C5-volatile formation in tomato. (Source:
Degousée et al., 1995; Gardner et al., 1996). .................................... 110

CHAPTER I

INTRODUCTION

Volatiles are detected by the olfactory system to be cognitively integrated
inﬂuencing the perception or aroma and ﬂavor of food (Baldwin et al., 2000;
Jackson, 2002). The aroma of a speciﬁc produce is given by the interaction of
many compounds in the volatile mixture. This makes the improvement of aroma
a Challenging and problematic task.

The knowledge of the pathway for aroma and ﬂavor formation, as well as
its regulation, is a key factor in the control of one the most desirable qualities of
fruit. Understanding of volatiles biosynthetic pathway is a main concern for fruit
ﬂavor researchers, in contrast to their early focus on identiﬁcation and
quantiﬁcation of those compounds (T eranishi et al., 1999). This allows the
manipulation of desired compound through speciﬁc Changes of precursors,
enzymes and cultural practices (Leahy and Roderick, 1999).

For tomato, several of the most important volatiles have been
characterized, and their biosynthesis have been described. Lipid oxidation
through the lipoxygenase pathway is one of those processes. However, the
relationship between substrate availability, concentration of C6-volatiles and
sensory quality perception is not fully understood.

Tomato is one of the most important vegetables grown in United States
with more than 173,000 ha harvested in the last ﬁve seasons, equivalent to a
production of 11,500,000 tons per year (FAOST AT data, 2004). Most of national
production is consumed internally, but approximately 185,000 tons ($161 million

value) are exported. Imports amount to approximately 800,000 tons ($790

million value). Interestingly, about 600,000 tons of domestic production is
considered waste, in part because many fruit do not achieve minimum market
quality and also due to losses during postharvest life.

Tomato quality is a function of the Characteristics and attributes that are
able to satisfy the demands and expectations of the person who is making the
judgment. Tomato growers tend to emphasize ease of growth and yield while
trying to meet market demands for fruit appearance. However, shippers,
receivers, and market distributors tend to emphasize ﬁrmness for improved
shipability, slowed ripening behavior, and extended shelf life, in addition to
having adequate visual quality. Finally, consumers want tomatoes that are
visually appealing, are ﬁrm, and have good ﬂavor and high nutritive value
(Grierson and Kader, 1986).

Despite the fact that tomato quality attributes comprises size, color,
ﬁrmness, sensory quality (aroma, sugar and acid ﬂavors, and off-ﬂavors), and
nutritional value (Frenkel and Jen, 1989; Salunkhe and Desai, 1984), consumers
perceive a disconnection between this factors. During the past 25 years,
consumer dissatisfaction with fresh-market tomato ﬂavor reached an all-time
high (Jones, 1986; Hobson, 1988). It was stated by Roberts (1988), “Green,
ﬂavorless, and rock hard, the store-bought tomato is perhaps the cruelest
invention of American agriculture”. This dissatisfaction provided the stimulus for
extensive research of factors affecting tomato ﬂavor beginning in the late 1970’s

and early 1980’s (Baldwin et al., 2000 and citations therein).

The current work integrates the use of biochemical, molecular, and
sensory techniques in order to better understand the biogenesis of some of the
most important volatiles, including the most abundant compound, cis-3-hexenal,
their impact on quality, and establish some bases for future improvement of

tomato aroma and ﬂavor.

REFERENCES

Baldwin, E.A., Scott, J.W., Shewmaker, GK. and Schuch, W. 2000. Flavor trivia
and tomato aroma: biochemistry and possible mechanisms for control and
important aroma components. HortScience 35(6):1013-1022.

FAOSTAT data, 2004. Agriculture Data.

(http:[[appsfaoorgipageicollections?subset=agriculture). Last accessed
February 2004.

Frenkel, C. and Jen, J.J., 1989. Tomatoes. In: Eskin, N.A.M., eds, Quality and
preservation of vegetables. CRC Press InC., Boca Raton, Florida.

Grierson, D. and Kader, AA, 1986. Fruit ripening and quality. In: Atherton, 16.
and Rudich, J., eds, The tomato crop, a scientiﬁc basis for improvement.
Champman & Hall, London, UK.

Hobson, GE. 1988. How the tomato lost its taste. New Scientist 19:46-50.

Jackson, J.F. 2002. Molecular biology of taste and aroma receptors: implications
for taste and aroma of plants products. p.1-5. In: Jackson, J.F. and
Linskens, H.F. (eds.) Analysis of taste and aroma. Springer-Verlag, Berlin
Heidelberg New York.

Jones, RA. 1986. Breeding for improved post-harvest tomato quality: genetics
aspects. Acta Horticulturae 190:77-87.

Leahy, MM. and Roderick, RC. 1999. Fruit ﬂavor biogenesis. p. 275-285. In:
Teranishi, R., Wick, EL. and Hornstein, I. (eds.) Flavor Chemistry 30 years
of progress. Kluwer Academic/Plenum Publishers, New York.

Roberts, L. 1988. Genetic Engineers Build a Better Tomato. Science
241(4871):1290.

Salunkhe, D.K. and Desai, 8.8., 1984. Postharvest biotechnology of vegetables.
Volume 1. CRC Press Inc., Boca Raton, Florida.

Teranishi, R., Wick, EL. and Hornstein, I. 1999. Flavor chemistry 30 years of
progress, an overview. p. 1-8. [m Teranishi, R., Wick, EL. and Hornstein,
I. (eds.) Flavor chemistry 30 years of progress. Kluwer Academic/Plenum
Publishers, New York.

CHAPTER II

LITERATURE REVIEW

Aroma volatiles are perceived by the olfactory system at concentrations in
air as low as parts-per-trillion. Perception of aroma is a very complex process
involving both detection and cognitive integration (Baldwin et al., 2000). Within
the olfactory bulb, some compounds are thought to bind single receptors and
others may bind several (Jackson, 2002). Odorants may, in fact, compete for
binding sights. Binding duration differs for different compounds and negative
aroma attributes can further complicate the response. Given that the ‘aroma’ of
most consumables is a function of many volatiles, assessing the inﬂuence of a
single volatile is a challenge and developing strategies to improve aroma is,
perhaps, even more problematic.

Of the more than 400 volatiles reported for tomato fruit (Grierson and
Kader, 1986; Frenkel and Jen, 1989; Tandon et al., 2000), only 30 are present at
a concentration higher than 1 part per billion (nL-L'l). Of those, only 16 have
positive ‘log odor units’, which is deﬁned as the logarithm of ratio of the
concentration of an aroma volatile to its odor threshold (Buttery, 1993). The
remaining compounds negatively impact tomato fruit aroma. The most
important aroma compounds are cis-3-hexenal, B-ionone, hexanal, B-
damascenone, 1-penten-3-one, 2+3-methylbutanal, trans-Z-hexenal, 2-
isobutylthiazole, 1-nitro-2-phenylethane, trans-2-heptenal, phenylacetaldehyde,
6-methyl-5-hepten-2-one, cis-3-hexenol, 2-phenylethanol, 3-methylbutanol, 3-

methylbutanal, methyl salicylate, pentanal, geranylacetone, and acetone

(Baldwin et al., 2000; Tandon, et al., 2000). Fatty acids, free amino acids, and
carotenoids are the main precursors of these aroma compounds.

The aldehydes ci5-3-hexenal, hexanal, and trans-Z-hexenal, the alcohols
1-hexanol, and ciS-3-hexenol, and the ketone 1-penten-3-one are produced as a
consequence of lipid degradation (Baldwin et al., 2000). These lipid breakdown
products provide some of the most important ﬂavor notes to tomato (Table 1).
Of all the aroma compounds produced by tomato, cis-3-hexenal has the greatest
potential impact based on its odor descriptor, its production rate, and a relatively
low human olfactory threshold (Buttery, 1993; Baldwin et al., 2000). cis-3-
Hexenal has the highest positive log odor units with 3.7. The next highest log
odor units are attributed to hexanal and beta-ionone, which have roughly 1/10th
the impact of ciS-3-hexenal as measured by this scale.

Our current understanding is that the formation of the compounds listed
in Table 1 involves the action of a sequence of enzymes [lipase, lipoxygenase
(LOX), hydroperoxide Iyase (HPL), isomerase, and alcohol dehydrogenase (ADH)]
on glycerolipids containing the fatty acids, linoleic acid (18:2) and linolenic acid
(18:3) (Galliard and Matthew, 1977; Riley et al., 1996; Bate et al., 1998), in a
process known as lipoxygenase pathway (Figure 1). The lipoxygenase (LOX)
pathway is responsible for producing several important aroma compounds from
fatty acid precursors. The relative amount of these two polyunsaturated fatty
acid substrates is variable and dependent on several factors including species,

cultivated variety, and organ, but in general they are the most abundant in

tomato accounting for the 60 to 70% of total fatty acid content (Gray et al.,

1999; Wang et al., 1996).

LIPOXYGENASE (LOX) PATHWAY

The LOX pathway is one of four enzymatic systems in plants by which
fatty acids are oxidatively modiﬁed; the other three pathways being a-oxidation,
B-oxidation, and (1)-oxidation (Hildebrand, 1989; Feussner and Wasternack,
2002). Lipoxygenases are a class of dioxygenases that catalyze the addition of
molecular oxygen to polyunsaturated molecules that contain a cis,cis-1,4-
pentadiene bond system (Mack et al., 1987). As noted, the two most common
substrates for LOX in plants are the fatty acids linoleic (18:2) and linolenic acid
(18:3) (Vick and Zimmerman, 1987).

The LOX pathway has been extensively studied and described (Vick and
Zimmerman, 1987; Mack et al., 1987; Hildebrand, 1989; Siedow, 1991; Feussner
and Wasternack, 2002) and its importance in aroma compound formation
characterized (Galliard and Matthew, 1977; Grifﬁths et al., 1999a; Baldwin et al.,
2000). The LOX pathway is also recognized as a source of important regulatory
compounds such as jasmonic acid (Creelman and Mullet, 1997; Howe and
Schilmiller, 2002) and traumatin (Vick and Zimmerman, 1987).

In the ﬁrst step, free fatty acids are released from membrane lipids by
action of lipases or acyl hydrolases (Huang, 1993; Wang, 2001). The majority of

LOXS use free fatty acids as substrate, but they can also attack phospholipids

while positioned in membranes (Hildebrand, 1989; Porta and Rocha-Sosa, 2002;
Feussner and Wasternack, 2002). In the second step, linolenic and linoleic acids
are oxidized by LOX to 13- and 9-hydroperoxy acids. The relation between 9-
and 13-HPO depends on the LOX isoform that attacks the substrate (Hornung et
al., 1999). LOX speciﬁcity for dioxygen insertion at position 9 or 13 is determined
by a combination of amino acid residues in the polypeptide. Thr/His, Thr/Phe,
Ser/Phe or CVS/Phe between positions 550 and 620 are common in 13-LOX,
while Thr/Val between 575 and 580 is related to 9-LOX (Hornung et al., 1999).
Hydroperoxide Iyase (HPL) catalyzes the cleavage of hydroperoxy fatty
acids (HPOs) to aldehydes and oxoacids. Based on substrate speciﬁcity, HPLs are
divided in two groups (Feussner and Wasternack, 2002; Grechkin, 2002): 13-
HPLS, that acts speciﬁcally on 13-HPOs, and 9/13-HPLS, which accept as
substrate both 9- and 13-HPOs. Hexanal is produced from the 13-hydroperoxy
acid of linoleic acid (13-HPOD), while cis-3-hexenal is produced from the 13-
hydroperoxy acid of linolenic acid (13-HPOT). The oxoacid 12-oxo-ci5-9-
dodecenoic acid is also a product from both reactions. The action of HPL on 9-
hydroperoxy linolenic acid (9-HPOT) leads to the formation of 3-cis-6-cis-
nonadienal, which is transformed to 2-t/ans-6-cis-nonadienal, and 9-oxo-
nonanoic acid. The action of HPL on 9-hydroperoxy linoleic acid (9-HPOD) results
in 3-ci5-nonenal, 2-trans-nonenal and 9-oxo-nonanoic acid (Vick, 1993). The
physiological role of 9-carbon aldehydes resulting from 9-HPO cleavage could be

related to wounding-healing agent or pest defense, however, no role has yet

10

been proposed for 9-oxo-nonanoic acid (Vick, 1993). C9-Aldehydes and alcohols
are important component of the volatile mixture of fruits from the Cucurbitaceae
family (Palma-Harris et al., 2002; Hayata et al., 2002).

In the next step of the pathway, at least in C6-Volatile biosynthesis, ADH
reduces hexanal to hexanol, and cis-3-hexenal to cis-3-hexenol. In addition,
trans-Z-hexenal is formed from cis-3-hexenal by isomerization, which may or
may not be catalyzed by an enzyme in vivo (Suurmeijer et al., 2000).

The LOX pathway is also employed by the plant for the biosynthesis of
jasmonates (Figure 2). Jasmonates are derived from 13-HPOT subsequent to the
action of allene oxide synthase (AOS) and allene oxide cyclase (AOC). Jasmonic
acid (JA) is produced from 12-oxo-phytodienoic acid following reduction and loss
of six carbons via three cycles of the B-oxidation pathway (Vick and Zimmerman,
1986; Mueller, 1997; Ziegler et al., 1999; Miersch and Wasternack, 2000;
Schaller, 2001).

In general, LOXS seem to have three main physiological roles in plants: i)
provide substrates for other metabolic enzymes, ii) participate in formation of
compounds that directly affect plant growth and development, and iii) participate
in membrane degradation during senescence and following wounding and
infection (Mack et al., 1987; Hildebrand, 1989; Blee, 2002; Feussner and
Wasternack, 2002).

Plants have numerous genes encoding enzymes with LOX activity. The

enzymes differ in their speciﬁcity to produce 9- or 13-HPOS, are present in

11

chloroplast and cytosol, and can be soluble or associated with membranes
(Creelman and Mullet, 1997; Riley and Thompson, 1997). Because of the
variability in enzyme substrate speciﬁcity and cellular location, ascertaining the
inﬂuence of LOX on the ﬂux of material through the pathway is difﬁcult. In
addition, apparently there is evidence for some degree of feedback regulation.
For example, in an ADH-deﬁcient mutant of Arabidopsis thaliana increased levels
of trans-Z-hexenal were associated with higher levels of HPL, suggesting that

HPL activity is inﬂuenced by products of the LOX pathway (Bate et al., 1998).

VOLATILE FORMATION BY THE TOMATO LIPOXYGENASE PATHWAY

Riley and Thompson (1998) investigated whether aroma compounds
associated with the LOX pathway in tomato accumulate during ripening or
whether they are formed upon tissue disruption. Their evidence suggested that
hexanal and cis-3-hexenal are present in the fruit at very low levels before tissue
disruption, regardless of maturity stage, and that after maceration they increase
several-fold depending on the maturity stage. Boukobza et al. (2001) found that
the production of C6-aldehydes and alcohols increased rapidly after fruit tissue
disruption and that the volatiles accumulated to their maximum concentration
after 3 minutes, demonstrating the importance of LOX pathway enzymes after
disruption, as was previously suggested by Galliard and Matthew (1977).

Somewhat paradoxically, Yilmaz et al. (2001) found no strong relationship

between LOX, HPL, and ADH activities and the amount of volatiles produced.

12

They also conclude that improving volatile formation is a complex goal that may
not be achievable by modifying a single step in the LOX pathway.

In terms of concentration at the headspace, the most abundant C6-
volatiles are derived from linolenic acid (Buttery, 1993). However, the most
abundant fatty acid present in tomato fruit is linoleic acid (Wang et al., 1996).
This suggests an 18:3-speciﬁcity of the LOX pathway in tomato that could be
related to expression and activity of LOXS and HPLs.

Collectively, the lack of a clear relationship between enzyme activity and
volatile aldehyde production, and the differential relations between fatty acids
and volatiles derived from them, support the necessity of a better understanding
of each of the biosynthesis of C6-volatiles, including the availability of substrate,

to improve fruit aroma.

TOMATO LIPOXYGENASES

Five different LOX genes have been reported in tomato: TomLoxA (Ferrie
et al., 1994), Tom/.oxB(Ferrie et al., 1994) or U13681 (Kausch and Handa,
1995, 1997), TomLoxC(Heitz et al., 1997), TomLoxD(Heit2 et al., 1997), and
TomLoxE (NCBI Accession: AY008278). TomLoxA, TomLoxB and TomLoxC are
found in the fruit and undergo Changes in expression during fruit ripening.
TomLoxD mRNA has been detected in very low levels in mature green and

breaker stages (Heitz et al., 1997). TomLOXA and TomLOXB are 9-LOXs,

13

whereas TomLOXC and TomLOXD are 13-LOXS (Grifﬁths et al., 1999a, Heitz et
al., 1997; Feussner and Wasternack, 2002)

Tom/.oxA is highly expressed in mature green and breaker stages, while
TomLoxB is expressed at breaker and ripe stages, but most strongly at the ripe
stage (Ferrie et al., 1994; Grifﬁths et al., 1999b). Only TomLOXB is fruit speciﬁc
and seems to be associated primarily with degradative processes during
senescence (Ferrie et al., 1994). TomLOXC and TomLOXD are both targeted to
the Chloroplast (Heitz et al., 1997). TomLOXD expression is wound-inducible
expression in leaves and may therefore play a role in JA-mediated defense
signaling in response to herbivore and pathogen attack. TomLOXC has no wound-
related expression in leaves, but it does accumulate in fruits from breaker stage
onward (Grifﬁths et al., 1999b). It was suggested that TomLOXC might
participate in disintegration of the thylakoid membrane during the transition from
chloroplast to chromoplast (Heitz et al., 1997).

The change in expression of different LOX genes during tomato ripening
(Ferrie et al., 1994; Kausch and Handa, 1997) suggests that different isoforms
could be responsible for the production of 13-HPOs during tissue disruption,
which are required for the production of six-carbon aroma compounds such as
hexanal and cis-3-hexenal. Riley et al. (1996) found that 20% of LOX is
membrane-associated in the microsomal fraction, where most of the HPL is
located. Between 5- and 9-times more HPO is produced by microsomal LOX than

is consumed by microsomal HPL. Microsomal or membrane associated LOX

14

activity increases during tomato fruit development, with a peak during breaker
stage. This suggests that tomato aroma compounds are produced at a
membrane Site (Riley et al., 1996).

Grifﬁths et al. (1999a) used antisense technology to reduce TomLOXA and
TomLoxB mRNA levels in tomato fruit, resulting in a reduction in lipoxygenase
activity. However, ﬂavor analysis showed no signiﬁcant changes between the
wild type and transformed fruit. The importance of TomLOXC in fruit aroma
production was recently demonstrated (Grierson, 2004, personal
communication), as tomato lines silencing TomLOXC) by co-suppression and
antisense, have reduced capacity to synthesize cis-3-hexenal, trans-Z-hexenal
and hexanal.

The main products of fruit LOX activity are 9-HPOS (Galliard and Matthew,
1977; Hatanaka et al., 1992); the ratio of 9-HPOS to 13-HPOs 5 approximately
95:5. However, C9-aldehydes and alcohols are not present in the volatile mixture
of tomato, as they are in other species such as cucumber (Palma—Harris et al.,

2002) and melon (Hayata et al., 2002).

TOMATO HYDROPEROXIDE LYASE
HPL activity has been extensively studied in tomato fruit (Riley et al.,
1996; Hatanaka et al., 1992; Suurmeijer et al., 2000) and leaf (Fauconnier et al.,

1997; Matsui et al., 2001).

15

Most of the HPL activity in tomato fruit is located in the microsomal
fraction (Riley et al., 1996). Its activity is not affected by pH variations, and does
not change from mature green to ripe stage (Riley et al., 1996). Regarding HPL
activity in different fruit tissues, it was found evenly distributed from the skin to
the locular material (Hatanaka et al., 1992).

It has been demonstrated that 9-HPOS are not substrates for HPL in fruits,
only 13-HPOs from 18:2 and 18:3 can be cleaved yielding C6-aldehydes
(Hatanaka et al., 1992; Matsui et al., 2001; Suurmeijer et al., 2000). Although
the production of C9-aldehydes by incubations of 9-HPOs with crude tomato
extract has been observed (Hatanaka et al., 1992), most probably they are
formed by the action of another enzyme (Suurmeijer et al., 2000).

Leaf and fruit HPL activities have a Clear afﬁnity for 13-HPOT over 13-
HPOD (Fauconnier et al., 1997; Suurmeijer et al., 2000). The enzyme converts
13-HPOT to cis-3-hexenal 8 times faster than 13-HPOD to hexanal (Suurmeijer et
al., 2000).

A tomato gene coding for HPL was recently identiﬁed and cloned (Howe et
al., 2000). Characterization of the recombinant protein showed that it is speciﬁc
for the metabolism of 13-HPOS, with 20-fold greater afﬁnity for 13-HPOT than
13-HPOD. LeHPL expression analysis showed a slow accumulation of mRNA in
fruits, suggesting that further work on the role of LeHPL in C6-aldehydes
production, speciﬁcally ci5-3-hexenal, needs to be done by alteration of LeHPL

expression (Howe et al., 2000).

16

FATTY ACID SUBSTRATES

That the LOX pathway is the primary route of volatile C6 aldehyde
formation during fruit maceration is generally not argued, despite the fact that
has been suggested (Wang et al., 1996) that small amounts of hexanal can be
produced from autoxidation of palmitoleic acid (16:1). Substrate feeding
experiments with protein extracts or homogenates from fruit demonstrated that
the addition of 18:2 enhances the production of hexanal and 1-hexanol, and
addition of 18:3 enhances cis-3-hexenal, trans-Z-hexenal and cis-3-hexenol
formation (Galliard and Matthew, 1977; Riley and Thompson, 1998; Boukobza et
al., 2001).

Gray et al. (1999) measured the amounts of C6-saturated and
unsaturated aldehydes produced by tomato cultivars that differ in amount of
18:2 and 18:3. They found a signiﬁcantly higher hexanal to hexenal ratio for
tomato fruit in which the ratio of 18:2 to 18:3 was higher. Wang et al. (1996)
found essentially the same phenomenon using a transgenic approach to alter
LOX substrate levels. In this work, the gene for A9 desaturase from yeast was
constitutively expressed in tomato. The amounts of 16:1, 18:1, and 18:2 in the
fruit were signiﬁcantly increased as was the total amount of fatty acids. The
amount of 18:3 was slightly reduced. The volatiles hexanal and 1-hexanol were

enhanced 267% and 407%, respectively, relative to untransformed tomato.

17

FATTY ACID DESATURASES

Fatty acid desaturases (FADS) are enzymes that introduce double bonds
into fatty acyl chains. They are present in all groups of organisms (bacteria,
fungi, plants and animals), and play an important role in the maintenance of the
proper structure and function of biological membranes (Los and Murata, 1998).
In higher plants, there are three important desaturases: A9 acyl-acyl carrier
protein (ACP) desaturase, A12 desaturases (FADZ is ER-located, and FAD6 is
chloroplast-located), and A15 or (03 desaturases (FAD3 is ER-located, and FAD7
and FAD8 are Chloroplast-located) (Nishida and Murata, 1996). A9 acyl-ACP
desaturase is an acyl-ACP desaturase, while A12 desaturase and (03 desaturase
are acyl-lipid desaturases, which are speciﬁc to fatty acids esteriﬁed to
glycerolipids (Nishida and Murata, 1996).

Manipulation of the degree of fatty acid saturation has revealed an
important physiological role for fatty acids in stress resistance. Decreased levels
of trienoic fatty acids (18:3 and 16:3) was shown to increase the heat tolerance
of plants (Murakami et al., 2000), but it has also been shown that this Change
reduces salt and drought tolerance (Im et al., 2002). When plants are subjected
to chilling conditions, there is an increase in the trienoic fatty acids (Kodama et
al., 1995; Berberich et al., 1998; Horiguchi et al., 2000; Dyer et al., 2001). It is
not clear whether overexpression of (03 desaturases is able to increase chilling
tolerance by itself (Nishida and Murata, 1996). However, Kodama et al. (1995)

demonstrated an increase in chilling resistance of transgenic tobacco plants

18

overexpressing (o3 FAD. It has been shown that 033 FAD expression increases in
Capsicum annuum upon wounding (Kwon et al., 2000).

The production of 18:3 from 18:2 in plants is catalyzed by enzymes called
03 FAD. At least three ESTS for enzymes with putative w3-FAD activity have
been identiﬁed in tomato.

Howe and Ryan ( 1999) initially identiﬁed a tomato mutant impaired in its
ability to respond to wounding. The mutant line does not produce wound-
inducible proteinase inhibitor in response to wounding. The mutant, sprZ (for
Suppressed in 3SS::Prosystemin-mediated Responses) was subsequently
identiﬁed as being deﬁcient in JA biosynthesis (Li et al., 2002). Li et al. (2003)
determined that the sprZ mutant had considerably lower levels of 18:3 than wild
type and, furthermore, contained no detectable levels of hexadecatrienoic acid
(16:3) in leaves. Expression of the wild type gene SprZ was found in stem,
petiole, leaf and ﬂower, but not in immature fruit or roots. It has been recently
demonstrated that the SprZ gene of tomato, which is required for JA-mediated
defense against herbivores (Howe and Ryan, 1999; Li et al., 2002) encodes a
Chloroplastic (03 FAD (Li et al., 2003), now called LeFAD7. The linolenic acid
serves as a precursor to JA. It was concluded that the inability of the sprZ
tomato mutant to synthesize signiﬁcant quantities of trienoic fatty acids reduced
JA production sufﬁciently to prevent wound-induced defense responses.

The mutant Lefad7 contains no 16:3 and <10% of the wild type level of

18:3 in leaves. There are no differences in fatty acid composition of roots

19

between the mutant and the wild type. Fruit fatty acid composition has not yet
been analyzed. A search of the tomato EST database has identiﬁed a tentative
consensus sequence, constructed from multiple overlapping EST 5 (Quackenbush
et al., 2000), that is clearly orthologous to the well Characterized FADB gene that
performs m3 fatty acid desaturation in the endoplasmic reticulum (ER) of
Arabidopsis (Yadav et al., 1993) and other plants. The enzyme encoded by this
gene (LeFad3) could be responsible for 18:3 accumulation in roots of the mutant
Lefad7, and the residual levels of 18:3 in leaves of sprZ plants. Leaf and fruit

volatile composition of the mutant Lefad7 has not been analyzed.

The current work focused on the relationship between tomato fruit
sensory quality and volatile proﬁle, and how these traits are affected by Changes
in substrate availability by alterations of 18:2 and 18:3 content, and LOX

pathway speciﬁcity by modiﬁcation of HPL activity.

20

REFERENCES

Baldwin, E.A., Scott, J.W., Shewmaker, OK. and Schuch, w. 2000. Flavor trivia
and tomato aroma: biochemistry and possible mechanisms for control and
important aroma components. HortScience 35(6):1013-1022.

Bate, N.J., Riley, J.C.M., Thompson, J.E. and Rothstein, 5.3., 1998. Quantitative
and qualitative differences in C6-volatile production from the lipoxygenase
pathway in an alcohol dehydrogenases mutant of Arabidopsis thaliana.
Physiologia Plantarum 104:97-104.

Berberich, T., Harada, M., Sugawara, K., Kodama, H., Iba, K. and Kusano, T.
1998. Two maize genes encoding omega-3- fatty acid desaturase and
their differential expression to temperature. Plant Molecular Biology
36(2):297-306.

Blee, E. 2002. Impact of phyto-oxylipins in plant defense. Trend in Plant Science
7(7):315-321.

Boukobza, F.,Dunphy, P.J. and Taylor, A.J. 2001. Measurement of lipid
oxidation-derived volatiles in fresh tomatoes. Postharvest Biology and
Technology 23:117-131.

Buttery, R.G., 1993. Quantitative and sensory aspects of ﬂavor of tomato and
other vegetables and fruits. In: Acree, TE. and Teranishi, R., eds, Flavor
science sensible principles and techniques, Chapter 8. American Chemical
Society, Washington, DC.

Creelman, RA. and Mullet, J.E., 1997. Biosynthesis and action of jasmonates in
plants. Annual Review of Plant Physiology and Plant Molecular Biology
48:355-381.

Dyer, J.M., Chapital, D.C., Cary, J.W. and Pepperman, AB. 2001. Chilling-
sensitive, post-transcriptional regulation of a plant fatty acid desaturase
expressed in yeast. Biochemical and Biophysical Research
Communications 282(4): 1019-1025.

Fauconnier, M.L., Perez, A.G., Sanz, C. and Marller, M. 1997. Puriﬁcation and

characterization of tomato (Lycopersicon esculentum Mill.) hydroperoxide
Iyase. Journal of Agricultural and Food Chemistry 45:4232-4236.

21

Ferrie, B.J., Beaudoin, N., Burkhart, w., Bowsher, CG. and Rothstein, SJ. 1994.
The Cloning of two tomato lipoxygenase genes and their differential
expression during fruit ripening. Plant Physiology 106:109-118.

Feussner, I. And Wasternack, C. 2002. The lipoxygenase pathway. Annual
Review of Plant Biology 53:275-297.

Frenkel, C. and Jen, J.J., 1989. Tomatoes. In: Eskin, N.A.M., eds, Quality and
preservation of vegetables. CRC Press Inc., Boca Raton, Florida.

Galliard, T. and Matthew, J. 1977. Lipoxygenase-mediated cleavage of fatty acid
to carbonyl fragments in tomato fruits. Phytochemistry 16:339-343.

Gray, D.A., Prestage, S., Linforth, R.S.T. and Taylor, A.J., 1999. Fresh tomato
speciﬁc ﬂuctuations in the composition of lipoxygenases-generated C6
aldehydes. Food Chemistry 64:149-155.

Grechkin, A.N. 2002. Hydroperoxide Iyase and divinyl ether synthase.
Prostaglandins & other Lipids Mediators 68-69:457-470.

Grierson, D. and Kader, AA, 1986. Fruit ripening and quality. In: Atherton, J.G.
and Rudich, J., eds, The tomato crop, a scientiﬁc basis for improvement.
Champman & Hall, London, UK.

Grifﬁths, A., Prestage, 5., Linforth, R., Zhang, J., Taylor, A. and Grierson, D.,
1999a. Fruit-speciﬁc lipoxygenase suppression in antisense-transgenic
tomatoes. Postharvest Biology and Technology 17:163-173.

Grifﬁths, A., Barry, C., Alpuche-Solis, AG. and Grierson, D., 1999b. Ethylene and
developmental signals regulate expression of lipoxygenase genes during
tomato fruit ripening. Journal of Experimental Botany 50(335):793-798.

Hatanaka, A., Kajiwara, T., Matsui, K. and Kitamura, A. 1992. Expression of
lipoxygenase and hydroperoxide Iyase activities in tomato fruits. Journal of
Biosciences 47(5-6):369-374.

Hayata, Y., Sakamoto, T., Kozuka, H., Sakamoto, K. and Osajima, Y. 2002
Analysis of aromatic volatile compounds in 'Miyabi' melon (Cucumis melo
L.) using the Porapak Q column. Journal of the Japanese Society for
Horticultural Science 71(4):517—525.

22

Heitz, T., Bergey, DR. and Ryan, CA, 1997. A gene encoding a Chloroplast-
targeted lipoxygenase in tomato leaves is transiently induced by
wounding, systemin, and methyl jasmonate. Plant Physiology 114:1085-
1093.

Hildebrand, D.F., 1989. Lipoxygenases. Physiologia Plantarum 76:249-253.

Horigushi, G., Fuse, T., Kawakami, N., Kodama, H. and Iba, K. 2000.
Temperature-dependent translational regulation of the ER omega-3- fatty
acid desaturase gene in wheat root tips. Plant Journal 24(6):805-813.

Hornung, E., Walther, M., Kiihn, H. and Feussner I., 1999. Conversion of
cucumber linoleate 13-lipoxygenase to a 9-lipoxygenating species by Site-
directed mutagenesis. Proceedings of National Academy of Sciences of
United States of America 96:4192-4197.

Howe, G. and Ryan, C. 1999. Suppressors of systemin signaling identify genes in
the tomato wound response pathway. Genetics 153:1411-1421.

Howe, G.A., Lee, G.I., Itoh, A., Li, L. and DeRocher, A. 2000. Cytochrome P450-
dependent metabolism of oxylipins in tomato: Cloning and expression of
allene oxide synthase and fatty acid hydroperoxide Iyase. Plant Physiology
123:711-724.

Howe, G. and Schilmiller 2002. Oxylipin metabolism in response to stress.
Current Opinions in Plant Biology 5:230-236.

Huang, A.H.C., 1993. Lipases. In: In: Moore, T.S., ed, Lipid metabolism in plants.
Chapter 14. CRC Press InC., Boca Raton, Florida.

Im, Y.J., Han, O, Chung, GO and Cho, EH. 2002. Antisense expression of an
Arabidopsis omega-3- fatty acid desaturase gene reduces salt/drought
tolerance in transgenic tobacco plants. Molecules and Cells 13(2):264-271.

Jackson, J.F. 2002. Molecular biology of taste and aroma receptors: implications
for taste and aroma of plants products. p.1-5. In: Jackson, J.F. and
Linskens, H.F. (eds.) Analysis of taste and aroma. Springer-Verlag, Berlin
Heidelberg New York.

Kausch, K.D. and Handa, A.K., 1995. Molecular cloning and nucleotide sequence

of a lipoxygenase cDNA from ripening tomato fruit. Plant Physiology
107:669-670.

23

Kausch, K.D. and Handa, A.K., 1997. Molecular cloning of a ripening-speciﬁc
lipoxygenase and its expression during wild-type and mutant tomato fruit
development. Plant Physiology 113:1041-1050.

Kodama, H. Horiguchi, G., Nishiuchi, T., Nishimura, M. and Iba, K. 1995. Fatty-
acid desaturation during Chilling acclimation is one of the factors involved
in conferring low-temperature tolerance to young tobacco-leaves. Plant
Physiology 107(4): 1177-1185.

Kwon, J.H., Lee, Y.M. and An, OS, 2000. cDNA cloning of chloroplast omega-3-
fatty acid desaturase from Capsicum annuum and its expression upon
wounding. Molecules and Cells 10(5):493-497.

Li, L., Li, C., Lee, GI. and Howe, GA. 2002. Distinct roles for jasmonic acid
synthesis and action in the systemic wound response of tomato.
Proceedings of National Academy of Sciences of United States of America
99:6416-6421.

Li, C., Liu, G., Zhao, Y., Bauer, P., Ganal, M., Ling, H., Xu, C. and Howe, GA.
2003. The tomato SUPPRESSOR OF PROSYSTEMIN-MEDIATED
RESPONSESZ gene encodes a fatty acid desaturase required for systemic
induced resistance. Abstract presented at "Frontiers of Plant Cell Biology:
Signals and Pathways", the 22nd Symposium in Plant Biology Riverside
Convention Center Riverside, California (Jan 15-19, 2003).

Los, DA. and Murata, N., 1998. Structure and expression of fatty acid
desaturases. Biochemica et Biophysica Acta-Lipids and Lipid Metabolism
1394 (1):3-15.

Mack, A.J., Peterman, T.K. and Siedow, J.N., 1987. Lipoxygenase isozymes in
higher plants: biochemical properties and physiological role. Isozymes:
Current TOpics in Biological and Medical Research 13:127-154.

Matsui, K., Fukutomi, S., Wilkinson, J., Hiatt, B., Knauff, V. and Kajwara, T. 2001.
Effect of overexpression of fatty acid 9-hydroperoxide Iyase in tomatoes
(Lycopersicon esculentum Mill.). Journal of Agricultural and Food
Chemistry 49(11):5418-5424

Mueller, NJ. 1997. Enzymes involved in jasmonic acid biosynthesis. Phisiologia
Plantarum 100:653-663.

Murakami, Y., Tsuyama, M., Kobayashi, Y., Kodama, H. and Iba, K. 2000.

Trienoic fatty acids and plant tolerance to high temperature. Science
287(5452):476-479.

24

Nishida, I. and Murata, N. 1996. Chilling sensitivity in plants and cyanobacteria:
the crucial contribution of membrane lipids. Annual Review of Plant
Physiology and Plant Molecular Biology 47:541-568.

Palma-Harris, C., McFeeters, RF. and Fleming, HP 2002. Fresh cucumber ﬂavor
in refrigerated pickles: Comparison of sensory and instrumental analysis
Journal of Agricultural and Food Chemistry 50(17):4875-4877.

Porta, H. and Rocha-Sosa, M. 2002. Plant lipoxygenases. Physiological and
molecular features. Plant Physiology 130:15-21.

Quackenbush, J., Liang, F., Holt, 1., Pertea, G. and Upton, J. 2000.The TIGR
Gene Indices: reconstruction and representation of expressed gene
sequences. Nucleic Acids Research 28(1):141-145.

Riley, J.C.M., Willemot, C. and Thompson, J.E., 1996. Lipoxygenase and
hydroperoxide Iyase activities in ripening tomato fruit. Postharvest Biology
and Technology 7:97-107.

Riley, J.C.M. and Thompson, J.E., 1997. Subcellular generation and distribution
of lipid-derived volatiles in ripe tomato. Journal of Plant Physiology
150:546-551.

Riley, J.C.M. and Thompson, J.E., 1998. Ripening-induced acceleration of volatile
aldehyde generation following tissue disruption in tomato fruit. Physiologia
Plantarum 104:571-576.

Siedow, J.N., 1991. Plant lipoxygenase: structure and function. Annual Review of
Plant Physiology and Plant Molecular Biology 42:145-188.

Suurmeijer, C.N.S.P., Perez-Gilabert, M., van Unen, D.J., van der Hijden HTWM,
Veldink GA, and Vliegenthart, J.F.G. 2000. Puriﬁcation, stabilization and
characterization of tomato fatty acid hydroperoxide Iyase. Phytochemistry
53:177-185

Tandon, K.S., Baldwin, EA. and Shewfelt, R.L. 2000. Aroma perception of
individual volatile compound in fresh tomatoes (Lycopersicon esculentum,
Mill.) as affected by medium of evaluation. Postharvest Biology and
Technology 20:261-268.

Vick, BA. and Zimmerman, DC. 1986. Characterization of 12-oxo-phytodienoic

acid reductase in corn - the jasmonic acid pathway. Plant Physiology
80(1):202-205.

25

Vick, BA. and Zimmerman, DC, 1987. Oxidative system for modiﬁcation of fatty
acids: the lipoxygenase pathway. In: Stumpf, P.K., Conn, E.E., eds, Lipids:
structure and function. The biochemistry of plants: a comprehensive
treatise, Vol 9. Academic Press, Orlando, Florida.

Vick, BA, 1993. Oxygenated fatty acids of the lipoxygenase pathway. In: Moore,
T.S., ed, Lipid metabolism in plants. Chapter 5. CRC Press Inc., Boca
Raton, Florida.

Wang, C., Chin, C.K., Ho, C.T., Hwang, C.F., Polashock, J. and Martin, CE,
1996. Changes of fatty acids and fatty acid derived compounds by
expressing the yeast A-9 desaturase gene in tomato. Journal of
Agricultural Food Chemistry 44:3399-3402.

Wang, X. 2001. Plant phospholipases. Annual Review of Plant Physiology and
Plant Molecular Biology 52:211-231.

Yadav N. S., Wierzbicki A., Aegerter M., Caster C. S., Perez-Grau L., Kinney A. J.,
Hitz W. D., Booth Jr., J. R., Schweiger 8., Stecca K. L., Allen S. M.,
Blackwell M., Reiter R. 5., Carlson T. J., Russell S. H., Feldmann K. A.,
Pierce J. and Browse J. 1993. Cloning of Higher Plant [omega]-3 Fatty
Acid Desaturases. Plant Physiology 103(2):467—476.

Yilmaz, E., Tandon, K.S., Scott, J.W., Baldwin, EA. and Shewfelt, R. 2001.
Absence of a Clear relationship between lipid pathway enzymes and
volatile compounds in fresh tomatoes. J. Plant Physiology 158:1111-1116.

Ziegler, J., Wasternack, C. and Hamberg, M. 1999. On the speciﬁcity of allene
oxide cyclase. Lipids 34(10):1005-1015.

26

Table 1. Odor descriptors, fatty acid precursor and logarithms of odor units for

C6- and CS-Volatile compounds from lipid oxidation in tomato fruit.

 

 

c 0mm. d Odor descriptor1 Fatty acid precursor Lag? r
Cis-3-hexenal Tomato, citrus Linolenic Acid (18:3) 3.7
hexanal Stale, grassy, green Linoleic Acid (18:2) 2.8
trans-Z-hexenal Stale, green, vine Linolenic Acid (18:3) 1.2
1-hexanol Glue, oil Linoleic Acid (18:2) -1.9
cis-3-hexenol Green, celery Linolenic Acid (18:3) 0.3
1-penten-3-one Fresh, sweet Linolenic Acid (18:3) 2.7
2-pentenal Stale, oil Linolenic Acid (18:3) -1.0
1-penten-3-ol Pungent, green, fruity Linolenic Acid (18:3) -0.6

 

1Tandon et al., 2000; 2Buttery, 1993

27

\r/\\ A‘v/\‘\\ ) //\ ”‘\ J”: / /—\9| M
. ./ V \v \ \/ \/,/—_-\V’ _ _ \

E-Z-Nonenol Z-3-Nonenol Z-3,Z-6-Nonadienol E-2,Z-6-Nonadienol
4;“; A A .4
. i . l ADH i l i i
COOH

E-Z-Nonenal <4- 3 Z-3-Nonenal 9-oxo-Nonanoic acid Z-3,Z-6-Nonadienal C-—-§> E-2,Z-6-Nonadienal

O - M
\v/“vAv’TS/E /‘\/\/:\V9 .‘r‘l 4, ,1}; {K , .4 \/=\r——\9 /\=/\/\9
\- ’ HPL "
90H H
are \./\e»\.,A.,.COOH 9-HPOD 9-HPOT \/_:‘\/’=%$Vv/\/m
‘\ 9-LOX /4
ﬁv\/=~Vﬁ—-\,r-\.,,/\,»\V.CIX)I'I 1822 18:3 \/=\/=\/=\/\/\/\/m'i
//A
l:/ 13.-Lox XE
OOH
,z-Vx/m,1~——\/~\/~V~L.COOH 13-HPOD 13-HPOT \v,=\/ \ — vWVCCDH
/7" HPL /
Hexanal 12-oxo-Z-9-Dodecenoic acid Z-3-Hexenal ELF?) E-Z-Hexenal
T K/WCIDH F! H
\2’ ADH ‘7' “.7
I-Hexanol Z-3-Hexenol E-2-Hexenol
/-\/.-\/9H VABH My

Figure 1. Lipoxygenase pathway for C6- and C9-aldehydes and alcohols
biosynthesis. Abbreviations are as follows: 18:2, linoleic acid; 18:3, linolenic acid;
9-LOX, lipoxygenase with speciﬁcity for peroxydizing carbon 9; 13-LOX,
lipoxygenase with speciﬁcity peroxydizing carbon 13; 9-HPOD, 9-
hydroperoxylinoleic acid; 9-HPOT, 9-hydroperoxy linolenic acid; 13-HPOD, 13-
hydroperoxylinoleic acid; 13-HPOT, 13-hydroperoxylinolenic acid; HPL,
hydroperoxide Iyase; ADH, alcohol dehydrogenase; Z, cis; E; trans. (Vick and
Zimmerman, 1987; Mack et al., 1987; Hildebrand, 1989; Siedow, 1991; Feussner
and Wasternack, 2002).

28

18:3 — —
13-LOX ﬂ
OOH
13-HPOT _ \ _ oooa
AOS ﬂ
0
12,13-EOT _ \ _ goon
Aoc , ﬂ

OPDA mam“

OPR 0 ﬂ

OPDA-8:0 WCOOH
£1 °

3 Cycles
B-Oxidation ll

0
ﬂ WW
0

JASMONIC
ACID COOH

Figure 2. Jasmonic acid biosynthesis pathway. Abbreviations are as follows: 18:3,
linolenic acid; 13-HPOT, 13-hydroperoxylinolenic acid; 12,13-EOT, 12,13-
epoxylinolenic acid; OPDA, 12-oxo-phytodienoic acid; CFC-8:0, 3-oxo-2-(2’(Z)-
pentenyl)-cyclopentane-1-octanoic acid; 13-LOX, 13-lipoxygenase; AOS, allene
oxide synthase; AOC, allene oxide cyclase; OPR, OPDA reductase. (Creelman and
Mullet, 1997).

29

CHAPTER III

DEFICIENCY OF LINOLENIC ACID OF Lefad7 MUTANT TOMATO CHANGES

THE VOLATILE PROFILE AND SENSORY PERCEPTION OF LEAVES AND FRUITS

30

Perception of aroma is a very complex process involving both detection
and cognitive integration (Baldwin et al., 2000). Given that the aroma of most
consumables is a function of many volatiles, assessing the inﬂuence of a single
volatile is a challenge, and developing strategies to improve aroma is, perhaps,
even more problematic.

Of the more than 400 volatiles reported for tomato fruit (Grierson and
Kader, 1986; Frenkel and Jen, 1989; Tandon et al., 2000), only 30 are present at
a concentration higher than 1 nL-L‘l. Of those, only 16 have a positive ‘log odor
unit’, which is deﬁned as the logarithm of ratio of the concentration of an aroma
volatile to its odor threshold (Buttery, 1993). The most important aroma ‘impact
compounds’, in order of their log odor unit, are cis-3-hexenal, B-ionone, hexanal,
1-penten-3-one, 3-methylbutanal, trans-Z-hexenal, 2-isobutylthiazole, 6-methyl-
S-hepten-Z-one, Cis-3-hexenol, 2-phenylethanol, 3-methylbutanol, geranyl-
acetone, 2-pentenal, and hexanol (Baldwin et al., 2000). Fatty acids, free amino
acids, and carotenoids are the main precursors of these aroma compounds.

The lipoxygenase (LOX) pathway is responsible for producing the
aldehydes cis-3-hexenal, hexanal, and trans-Z-hexenal, and the alcohols 1-
hexanol, and cis—3-hexenol from fatty acid precursors (Baldwin et al., 2000).
These lipid breakdown products provide some of the most important ﬂavor notes
to tomato (Table 1). Of all the aroma compounds produced by tomato, cis-3-

hexenal has the greatest potential impact based on its odor descriptor, and its

31

production rate, which result in the highest positive log odor units of all volatiles
produced (Buttery, 1993; Baldwin et al., 2000).

Our current understanding is that the formation of these ﬁve compounds
involves the action of a sequence of enzymes [lipase, lipoxygenase (LOX),
hydroperoxide Iyase (HPL), isomerase, and alcohol dehydrogenase (ADH)] on
glycerolipids containing the fatty acids (FA) linoleic acid (18:2) and linolenic acid
(18:3) (Galliard and Matthew, 1977; Riley et al., 1996; Bate et al., 1998). Fatty
acid composition of tomato is variable and dependent on the cultivated variety,
but in general the most abundant are 18:3 and 18:2, which account for the 60 to
70% of the total FA content (Gray et al., 1999; Wang et al., 1996; Li et al.,
2003).

Substrate feeding experiments using protein extracts or homogenates
from fruit demonstrated that the addition of 18:2 enhances the production of
hexanal and 1-hexanol, and addition of 18:3 enhances ci5-3-hexenal, trans-2-
hexenal, and cis-3-hexenol formation (Galliard and Matthew, 1977; Riley and
Thompson, 1998; Boukobza et al., 2001). Somewhat paradoxically, no strong
relationship was found between LOX, HPL, and ADH activities and the amount of
volatiles produced, suggesting that improving volatile formation is a complex
goal that may not be achievable by modifying a single step in the LOX pathway.
(Yilmaz et al., 2001).

Gray et al. (1999) measured the amounts of saturated and unsaturated

C6-aldehydes produced by tomato cultivars that differ in amount of 18:2 and

32

18:3. They found a signiﬁcantly higher hexanal to hexenal ratio for tomato fruit
in which the ratio of 18:2 to 18:3 was higher. Wang et al. (1996) found similar
results using a transgenic approach to alter LOX substrate levels by constitutively
expression of a yeast A9 desaturase gene. The amounts of 16:1, 18:1, and 18:2
in the fruit were signiﬁcantly increased as was the total amount of FA. The
amount of 18:3 was slightly reduced relative to the wild type. The volatiles
hexanal and 1-hexanol were enhanced 267% and 407%, respectively, relative to
untransformed tomato.

Howe and Ryan (1999) identiﬁed a tomato mutant impaired in its ability to
respond to wounding. The mutant, sprZ (for Suppressed in 355::Prosystemin-
mediated Responses) was subsequently identiﬁed as being deﬁcient in jasmonic
acid (JA) biosynthesis (Li et al., 2002). Li et al. (2003) determined that the sprZ
mutant had considerably lower levels of 18:3 than wild type and, furthermore,
contained no detectable levels of hexadecatrienoic acid (16:3) in leaves. It has
been recently demonstrated that the SprZ gene of tomato encodes a
Chloroplastic (n3 FA desaturase (Li et al., 2003), now called LeFAD7.

The current work analyzes the effect of the dramatic change in the
polyunsaturated fatty acid (PUFA) composition in the sprZ mutant (renamed
Lefad?) on C6-Volatile concentration in tomato leaves and fruits at different
developmental stages and how this change in the volatile proﬁle is sensorially

perceived by a non-trained panel.

33

MATERIALS AND METHODS
PLANT MATERIAL

Tomatoes (Lycopersicon esculentum Mill.) CV. Castlemart and the mutant
of this cultivar, Lefad7, obtained by ethyl methanesulfonate mutagenesis (EMS)
(Howe and Ryan, 1999), were grown in a greenhouse under a constant
temperature of 20°C. Young and mature leaves and immature-green and

mature-red fruits were sampled for FA and volatile analysis.

FATTY ACID COMPOSITION

Total lipids were extracted twice from 400 mg of leaf and 500 mg of fruit
tissue with 6 mL of hexane and once with 4 mL of 2:7 isopropanolzhexane.
Esteriﬁcation/transesteriﬁcation of FA was performed with methanoliczH2504
(2.5%) to generate fatty acid methyl esters (FAMES) (Conconl et al., 1996).
FAMES were analyzed by gas chromatography (GC)-ﬂame ionization detector
(FID) on a DB-23 capillary column (J&W Scientiﬁc, Folsom, CA) as described by

Bonaventure et al. (2003).

EVALUATION OF AROMA COMPOUNDS

Sample Preparation. One leaf disk (35 mg) from each experimental unit
was crushed in a 20-mL amber glass vial. After 3 minutes, 1 mL of 50% CaClz
solution was added to stop further enzymatic reactions. The vial was immediately

Closed with Mininert valve (Supelco, Bellefonte, PA), and stored at -20°C until

34

Volatile analysis. From each tomato fruit, 10 g of pericarp tissue were
homogenized in 5 mL of distilled water using a Brinkmann Homogenizer (PT
10/35, Brinkmann Instrument Co, Switzerland). After 3 minutes, 6 mL of 50%
CaClz solution was added and stored at -80°C. Before volatile analysis, 5 mL of
the homogenate were placed in a 20-mL amber glass vial and Closed with a
Mininert valve.

GC-MS Analysis. Homogenate vials were placed in a water bath held at
37°C for 30 minutes before analysis. Volatile compounds were extracted from the
headspace using a solid phase micro extraction (SPME) ﬁber (65 um PDMS-DVB,
Supelco, Bellefonte, PA) according to the method of Song et al. (1998). The ﬁber
was held in the vial for 3 minutes to allow absorption of volatile compounds. A
gas chromatograph (HP 6890 Series GC, Hewlett-Packard Co., Wilmington, DE)
was used for analyte separation, and a time-Of-ﬂight mass spectrometer
(Pegasus II, Leco, St. Joseph, MI) was used for analyte detection, identiﬁcation,
and quantiﬁcation. The GC column (SupelcoWax-10, Supelco, Bellefonte, PA) was
30 m x 0.2 mm with a 0.2 pm-thick coating. In order to quantify the headspace
concentration of hexanal, cis-3-hexenal, trans-Z-hexenal, 1-hexanol, and cis-3-
hexenol, gas standards were created using authenticated compounds according
to the method of Song et al. (1998). For leaf analysis, 0.25 pg of 2-octanone (10
pL of a 25 ng-uL'1 solution in water) was added in the vial as internal standard.

For fruit samples, 0.4 ug of 2-octanone was added to the vial.

35

SENSORY EVALUATION

For leaves, ﬁve leaf disks (approximately 170 mg) were crushed in a 20-
mL vial, closed with teﬂon-lined caps, and held at room temperature. One mL of
a 50% CaClz aqueous solution was added after 3 minutes to stop enzymatic
reactions. Vials were kept at -20°C until the sensory test was performed. For
fruits, 30 g of pericarp were blended and 20 mL of 50% CaClz solution were
added after 3 minutes. Five mL of the homogenate were transferred into 20-mL
vials and kept at -20°C until the sensory test.

A triangle test was performed as indicated by Meilgaard et al. (1999).
Panelists were presented with three vials; two vials held the same homogenate
and the third held a different homogenate. Panelists were asked to choose which
vial differed from the other two. The order in which the samples, mutant or wild
type, were presented to the panelists were arranged so that all six possible
combinations were used. There were at least ﬁve replications of each sample
combination presented to panelists.

A preference test was performed together with the triangle test for fruits
(Meilgaard et al., 1999). The panel was asked if the homogenate they Chose as
being different from the other two was preferable or not-preferable. The
preference test was designed to qualify the sensory differences detected in the
triangle test. The hypothesis tested was that more than 50% of panelists would

prefer one aroma over the other (Meilgaard et al., 1999).

36

STATISTICAL ANALYSIS

All data for FA content and headspace volatile concentration were
expressed as the mean :L- standard error of mean. Data were analyzed using
one-way ANOVA by PROC MIXED of a commercial statistical software package
(SAS version 8e, SAS Institute InC., Cary, NC). Statistical signiﬁcance of sensory

evaluation data was determined using Table T8 in Meilgaard et al. (1999).

RESULTS
FATTY ACID COMPOSITION

The total FA content of leaves did not differ between wild type and Leiad7
mutant lines, averaging 29.3 pmol-g'l. Of the eight FAS evaluated for mutant and
wild type leaves, only dienoic (16:2 and 18:2) and trienoic (16:3 and 18:3) FAS
differed (Figure 1A). Mutant leaf 16:3 content was 1% of wild type, while the
18:3 content was 16% of wild type. The content of 16:2 and 18:2 in mutant
leaves was 13- and 3.5-fold higher than in wild type leaves. In the wild type
tomato leaf, the most abundant FA was 18:3, constituting about 44% of total FA
while 18:2 constituted 18%; but in Lefad7leaves, the most abundant was 18:2
comprising about 57% of the total (Figure 2A), while 18:3 represented only 7%
of total FA. Octadecanoic acids accounted for approximately 65% of total FA.
Maturity did not affect total FA content or composition for wild type and mutant

leaves (not shown).

37

AS in leaves, fruit FA composition and content was not affected by
maturity stage (Figures 18 and C) for wild type and mutant fruits, averaging 2.7
umol-g'l. Trienoic and dienoic forms of hexadecanoic acid (16:3 and 16:2) were
absent or in very small amounts in tomato fruit, octadecanoic acids accounted
for more than 60% of the total FA content (Figures 28 and C). Mutant fruits had
a lower 18:3 content than wild type fruit. The loss of FAD7 activity caused the
mole fraction of 18:2 to increase from approximately 40% to 60%, and caused
the mole fraction of 18:3 to decrease from approximately 20% to 1-3% of the
total FA content. The impact of Lefad7 mutation on 18:3 content reduction was

greater for fruits than for leaves.

VOLATILE AROMA COMPOUNDS

C6-aldehydes accounted for 99% of C6-volatiles derived from the LOX
pathway for all the tissues and stages analyzed in wild type and Lefad7 plants.

In wild type leaves, the most abundant C6-aldehyde was cis-3-hexenal
(93%), a product of 18:3 degradation. In Lefad7 mutant leaves, the most
abundant C6-aldehyde was hexanal (70%), a product of 18:2 degradation. In
mutant leaves cis-3-hexenal accounted for only 29% of the total C6-aldehydes,
most likely derived from the residual amounts of 18:3 produced by non-
Chloroplastic FAD (FAD3). For wild type and mutant leaves, the concentration of

trans-Z-hexenal was low, amounting to 3 and 1% respectively (Figure 5). There

38

was no difference in volatile composition and concentration for young and
mature leaves (Figures 3A and B).

When fruit homogenate volatile quantity and composition was compared
between immature green and mature red stages, total C6-aldehydes for red
tomatoes were 100-fold higher than green fruit in the wild type, despite no shift
in the content or composition of FA precursors. Fruit maturity also had a marked
impact on volatile production by mutant fruits, with total C6-Volatiles of the red
fruit being 50-fold higher than green fruits (Figures 4A and B). In terms of
composition of the C6-aldehydes, wild type green fruits had more hexanal (60%
of total C6-aldehydes) than cis-3 and trans-Z-hexenals, while red fruit had
produced a greater proportion (74%) of hexenals, of which cis-3-hexenal was
the most abundant with 69% of total C6-aldehydes (Figure 5). Although the
most abundant C6-aldehyde produced by mutant fruits was hexanal for both
developmental stages, at the red stage it was possible to detect the presence of
small amounts of cis-3-hexenal (120 nL-L‘l), while at green immature stage this

aldehyde was not detected.

SENSORY EVALUATION

In the triangle test for leaves, 26 out of 32 subjects were able to pick the
correct odd sample between the three vials (Table 2). For fruits, 28 out of 36
subjects were able to differentiate the odd sample from the set of three Vials.

The results of both triangle tests demonstrated that the difference in volatiles of

39

mutant and wild type leaf and fruit tissues can be readily perceived (p-value
<0.0005).

In the preference test, the tested hypothesis was that more than 50% of
the panel prefered the tomato aroma of wild type fruit. The majority of
respondents expressed a preference for the aroma of the wild type tomato fruit

homogenate and the hypothesis was accepted (p-value <0.0005).

DISCUSSION

The mutation of the O3 FA desaturase gene (LeFad7) was previously
Characterized (Li et al., 2003) and was shown to reduce the linolenic acid content
in leaves, but did not affect root FA content. Fruit FA content was not reported.
The FA composition and content of leaves in this study was similar to those
previously published (Li et al., 2003).

The fact that the total FA content was not affected by LeFad7 mutation
contrasts with results of overexpression of a yeast-derived A9 desaturase in
tomato, in which fruit FA content was increased approximately 100% in fruits
(Wang et al.,1996), and 25% in leaves (Wang et al., 2001). One might have
expected that reduced expression of a gene critical to FA biosynthesis might
have reduced total FA accumulation. However, Lefad7 data are consistent with
observations regarding an Arabidopsis thaliana fad7 mutant (Zhuang et al.,
1996) and a fad3-2 fad7-2 fad8 triple mutant (Routaboul et al., 2000), for which

the total FA content is not affected by the lack of (.03 FA desaturase activity.

40

The reduction of 18:3 in tomato vegetative and reproductive tissue and
the accompanying dramatic decreased in the concentration of the aldehyde and
alcohol volatiles derived from its peroxidation, conﬁrm that 18:3 is the primary
precursor to cis-3-hexenal, trans-Z-hexenal, and cis—3-hexenol, as suggested by
Baldwin et al. (2000).

Considering that FA and C6-volatile composition was not signiﬁcantly
different between young and mature leaves, the data suggest that the regulation
of FA oxidation in leaves does not change developmentally. However, in fruits,
the 50- to 100-fold increase in C6-aldehyde production associated with
maturation and ripening (Table 3) with no accompanying shift in FA content,
likely reﬂects that an increase in LOX pathway activity accompanies fruit
ripening. Changes in lipase activity could also be affecting the availability of
substrate for LOX pathway. However, the only lipase expression analyses during
tomato fruit ripening have been done on phospholipase D (PLD) (Jandus et al.,
1997; Whitaker et al., 2001 ; Pinhero et al., 2003). In some cultivars PLD activity
increased during ripening process, but in others it decreased or did not change.

The ripening-related enhanced LOX pathway activity is most likely
associated with an increase in expression of lipoxygenase genes (Grifﬁths et al.,
1999b). The gene most likely to control fruit LOX activity during ripening is
TomLoxC, given that TomLoxD is not expressed in fruits (Heitz et al., 1997), and
antisense-mediated silencing of TomLoxA and TomLoxB did not result in a

Change in volatile composition (Grifﬁths et al., 1999a). The relative importance of

41

LOX activity to the other enzymes in the pathway, HPL, is further supported by
the ﬁndings that HPL activity is steady during tomato fruit ripening (Riley et al.,
1996). However, differences in the relative aldehyde composition (Figures 4 and
5) between the two stages of fruits, indicate a marked increase in the production
of 18:3 oxidation products (eg. cis-3-hexenal) without a similar increase in the
formation of hexanal from 18:2, occurred during ripening. This developmental
Change in substrate preference in the LOX pathway is most likely related to HPL,
which has a higher afﬁnity for 13-hydroperoxide of 18:3 (13-HPOT) than 13-
hydroperoxide of 18:2 (13-HPOD) (Fauconnier et al., 1997; Howe et al., 2000;
Suurmeijer et al., 2000). Our data therefore imply a Change in HPL activity in
viva, contrary to the ﬁndings of Riley et al. (1996). The large increase in 18:3-
derived C6-volatiles relative to 18:2-derived C6-volatiles is likely not related to
LOX enzymes, which have been shown to have no speciﬁcity for 18:2 and 18:3.

The large increase in C6-volatile production by ripening tomato fruit
relative to immature fruit with no change in FA content or composition suggests
that increase in volatile biosynthesis in wild type fruit during ripening is not a
function of FA substrate availability. Nevertheless, the near elimination of 0:3 FAs
as a substrate for the LOX pathway did impact the capacity for synthesis of C6-
volatiles.

The perception of differences in the volatiles proﬁle of wild type and
Lefad7 leaves and fruits by a sensory panel is the ﬁrst demonstration that

alteration in FA composition under isogenic background can alter the aroma of

42

vegetative or reproductive tissue. The LeFad7 mutation did not appear to
negatively impact LOX pathway performance, thus perceived quality changes are
attributable to a single factor, the balance between dienoic and trienoic fatty
acids. It is not certain, however, whether panelists were responding to a loss in
18:3-related volatiles or to an increase in 18:2-related volatiles. Likely, a trained
panel would be needed to make this determination.

The data for chemical and sensory impacts of the reduction in the activity
of LeFAD7 highlight an important feature of manipulating single genes in this and
other metabolic pathways. Despite the relative simplicity of the LOX pathway for
volatile formation, a single gene Change impacts multiple compounds, suggesting
that the enhancement of a single desirable compound or Class of compounds,
may not be achievable. While the minimal unit of change is a single gene, the
minimal scope of change is likely to be numerous products of the pathway
impacted.

The current work demonstrated that in viva reduction of 18:3 content in
tomato tissue reduces the biosynthesis of several of the most important volatile
compounds in tomato aroma, including the volatile with the highest log-odor
units cis-3-hexenal, but it does not reduce the capacity of C6-aldehyde and
alcohol production by the LOX pathway. The altered concentration of LOX
pathway oxidation products, mainly aldehydes, was detected by the olfactory
sense of humans, affecting the sensory quality of tomato fruits. The current

results open the possibility of improvement of tomato fruit sensory quality by

43

changing FA composition. An increase in 18:3 content and/or the rate of
oxidation of linolenic acid may result in an increase in the relative concentration
of the most important tomato volatile compound, ci5-3—hexenal. An increase on
18:3 content may also be attended by increases in non-target compounds such
as 1-penten-3-one and other CS-Volatiles from fatty acid degradation, which may

also impact fruit quality attributes and may alter plant physiology.

REFERENCES

Baldwin, E.A., Scott, J.W., Shewmaker, GK. and Schuch, W. 2000. Flavor trivia
and tomato aroma: biochemistry and possible mechanisms for control and
important aroma components. HortScience 35(6):1013-1022.

Bate, N.J., Riley, J.C.M., Thompson, J.E. and Rothstein, S.J., 1998. Quantitative
and qualitative differences in C6-volatile production from the lipoxygenase
pathway in an alcohol dehydrogenases mutant of Arabidopsis thaliana.
Physiologia Plantarum 104: 97-104.

Bonaventure, G., Salas, J.J., Pollard, MR and Ohlrogge, J.B. 2003. Disruption of
the FATB gene in Arabidopsis demonstrates an essential role of saturated
fatty acids in plant growth. The Plant Cell 15(4):1020-1033.

Boukobza, F.,Dunphy, P.J. and Taylor, A]. 2001. Measurement of lipid
oxidation-derived volatiles in fresh tomatoes. Postharvest Biology and
Technology 23:117-131.

Buttery, R.G., 1993. Quantitative and sensory aspects of ﬂavor of tomato and
other vegetables and fruits. In: Acree, TE. and Teranishi, R., eds, Flavor
science sensible principles and techniques, Chapter 8.American Chemical
Society, Washington, DC.

Conconi, A., Miquel, M., Browse, JA. and Ryan, CA. 1996. Intracellular levels of
free linolenic and linoleic acids increase in tomato leaves in response to
wounding. Plant Physiology 111(3): 797-803.

Fauconnier, M.L., Perez, A.G., Sanz, C. and Marlier, M. 1997. Puriﬁcation and
characterization of tomato (Lycopersicon esculentum Mill.) hydroperoxide
Iyase. J. Agric. Food Chem. 45:4232-4236.

Frenkel, C. and Jen, J.J., 1989. Tomatoes. In: Eskin, N.A.M., eds, Quality and
preservation of vegetables. CRC Press Inc., Boca Raton, Florida.

Galliard, T. and Matthew, J. 1977. Lipoxygenase-mediated Cleavage of fatty acid
to carbonyl fragments in tomato fruits. Phytochemistry 16: 339-343.

Gray, D.A., Prestage, S., Linforth, R.S.T. and Taylor, A.J., 1999. Fresh tomato

speciﬁc ﬂuctuations in the composition of lipoxygenases-generated C6
aldehydes. Food Chemistry 64: 149-155.

45

Grierson, D. and Kader, AA, 1986. Fruit ripening and quality. In: Atherton, J.G.
and Rudich, J., eds, The tomato crop, a scientiﬁc basis for improvement.
Champman & Hall, London, UK.

Grifﬁths, A., Prestage, 5., Linforth, R., Zhang, J., Taylor, A. and Grierson, D.,
1999a. Fruit-speciﬁc lipoxygenase suppression in antisense-transgenic
tomatoes. Postharvest Biology and Technology 17: 163-173.

Grifﬁths, A., Barry, C., Alpuche-Solis, AG. and Grierson, D., 1999b. Ethylene and
developmental signals regulate expression of lipoxygenase genes during
tomato fruit ripening. Journal of Experimental Botany 50(335):793-798.

Heitz, T., Bergey, DR. and Ryan, CA, 1997. A gene encoding a chloroplast-
targeted lipoxygenase in tomato leaves is transiently induced by
wounding, systemin, and methyl jasmonate. Plant Physiology 114: 1085-
1093.

Howe, G. and Ryan, C. 1999. Suppressors of systemin signaling identify genes in
the tomato wound response pathway. Genetics 153:1411-1421.

Howe, G.A., Lee, G.I., Itoh, A., Li, L. and DeRocher, A. 2000. Cytochrome P450-
dependent metabolism of oxylipins in tomato: Cloning and expression of
allene oxide synthase and fatty acid hydroperoxide Iyase. Plant Physiology
123: 711-724.

Jandus, J., Valentova, O., Kas, J., Daussant, J. and Thevenot, C. 1997.
Phospholipase D during tomato fruit ripening. Plant Physiology and
Biochemistry 35(2): 123-128.

Li, L., Li, C., Lee, GI. and Howe, GA. 2002. Distinct roles for jasmonic acid
synthesis and action in the systemic wound response of tomato.
Proceedings of National Academy of Sciences of United States of America
99:6416-6421.

Li, C., Liu, G., Xu, C., Lee, G.I, Bauer, P., Ling, H., Ganal, M., and Howe, GA
2003. The tomato suppressor ofprosystemin-mediated responsesz gene
encodes a fatty acid desaturase required biosynthesis of jasmonic acid
and the production of a systemic wound signal for defense gene
expression. The Plant Cell 15:1646-1661.

Meilgaard, M., Civille, G.V., and Carr, B.T. 1999. Sensory Evaluation
Techniques, 3rd Edition. CRC Press, Boca Raton, Florida.

46

Pinhero, R.G., Almquist, K.C., Novotna, Z. and Paliyath, G. 2003. Developmental
regulation of phospholipase D in tomato fruits. Plant Physiology and
Biochemistry 41(3):223-240.

Riley, J.C.M., Willemot, C. and Thompson, J.E., 1996. Lipoxygenase and
hydroperoxide Iyase activities in ripening tomato fruit. Postharvest Biology
and Technology 7: 97-107.

Riley, J.C.M. and Thompson, J.E., 1998. Ripening-induced acceleration of volatile
aldehyde generation following tissue disruption in tomato fruit. Physiologia
Plantarum 104: 571-576.

Routaboul Jean-Marc, Fischer Steven F., and Browse John. 2000. Trienoic fatty
acids are required to maintain chloroplast function at low temperatures.
Plant Physiology 124:1697-1705.

Song, J., Fan, L.H. and Beaudry, RM. 1998. Application of solid phase
microextraction and gas Chromatography time-of-ﬂight mass spectrometry
for rapid analysis of ﬂavor volatiles in tomato and strawberry fruits.
Journal of Agricultural and Food Chemistry 46(9): 3721-3726.

Suurmeijer, C.N.S.P., Perez-Gilabert, M., van Unen, D.J., van der Hijden HTWM,
Veldink GA, and Vliegenthart, J.F.G. 2000. Puriﬁcation, stabilization and
characterization of tomato fatty acid hydroperoxide Iyase. Phytochemistry
53: 177-185

Tandon, K.S., Baldwin, EA. and Shewfelt, R.L. 2000. Aroma perception of
individual volatile compound in fresh tomatoes (Lycopersicon esculentum,
Mill.) as affected by medium of evaluation. Postharvest Biology and
Technology 20: 261-268.

Wang, C., Chin, C.K., Ho, C.T., Hwang, C.F., Polashock, J. and Martin, CE,
1996. Changes of fatty acids and fatty acid derived compounds by
expressing the yeast A-9 desaturase gene in tomato. Journal of
Agricultural Food Chemistry 44: 3399-3402.

Wang, C., Xing, J., Chin, C.K., Ho, CT. and Martin, GE, 2001. Modiﬁcation of
fatty acids Change ﬂavor volatiles in tomato leaves. Phytochemistry
58:227-232.

Whitaker, B.D., Smith, D.L. and Green, KC. 2001. Cloning, characterization and

functional expression of a phospholipase D alpha cDNA from tomato fruit.
Physiologia Plantarum 112(1):87-94.

47

Yilmaz, E., Tandon, K.S., Scott, J.W., Baldwin, EA. and Shewfelt, R. 2001.
Absence of a clear relationship between lipid pathway enzymes and
Volatile compounds in fresh tomatoes. J. Plant Physiology 158:1111-1116.

Zhuang, H., Hamilton-Kemp, T.R., Andersen, RA and Hildebrand, DE 1996.
The impact of alteration of polyunsaturated fatty acid levels on C5-
aldehyde formation of Arabidopsis thaliana leaves. Plant Physiology
111:805-812.

48

Table 1. Odor descriptors, fatty acid precursor and logarithms of odor units for
C6- and C5-volatiles from lipid oxidation in tomato fruit.

 

 

colirigomtfnd Odor descriptor1 Fatty acid precursor Lag?
cis-3-Hexenal Tomato, citrus Linolenic Acid (18:3) 3.7
Hexanal Stale, grassy, green Linoleic Acid (18:2) 2.8
trans-Z-Hexenal Stale, green, vine Linolenic Acid (18:3) 1.2
1-Hexanol Glue, oil Linoleic Acid (18:2) -1.9
Cis-3-Hexenol Green, celery Linolenic Acid (18:3) 0.3
1-Penten-3-one Fresh, sweet Linolenic Acid (18:3) 2.7
2-Pentenal Stale, oil Linolenic Acid (18:3) -1.0

1-Penten-3-ol Pungent, green, fruity Linolenic Acid (18:3) -0.6

 

1Tandon et al., 2000; 2Buttery, 1993

Table 2. Signiﬁcance for triangle test in leaves and fruits, and preference test in
fruits.

 

 

 

 

 

Number of subjects
Correct Total p-Value1
Triangle test leaves 26 32 < 0.0005
Triangle test fruits 28 36 < 0.0005
Wild type better Total p-Value
Preference test fruits 23 28 < 0.0005

 

1 T8 “Critical Numbers of Correct Responses in a Triangle Test” (Meilgaard et al., 1999)

49

Table 3. Total C6-volatile headspace concentration produced by homogenized
immature green and red mature fruits (nL-L'l).

 

 

 

 

Wild Type Lefad7
Green Red R egg: e n Green Red R egg: e n
Aldehydes
Hexanal 33 1,398 42.3 142 7,064 49.7
cis-3-Hexenal 16 3,738 233.6 0 1 19 -
trans-Z-Hexenal 5 254 50.8 1 18 18.0
Total 54 5,390 99.8 143 7,201 50.3
Alcohols
1-Hexanol 0 0 n.d.1 3 3.0
cis-3-Hexenol 4 4.0 0 1 n.d.
Total 1 4 4.0 4 4.0

 

1 n.d., not detected.

50

20

 

IMHWm

FL
16 1 El Lefad7 A

-.- lzi

 

 

 

 

 

0 _j . I 4 ‘TI 4
16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3
Fatty acid

 

 

 

 

 

0.0" 'Tmr —l ILT

16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3
Fatty acid

 

 

 

 

0.0 “ T T T —L

 

16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3
Fatty acid

Figure 1. Fatty acid composition of tomato tissue on a fresh weight basis. A, leaf;
B, green immature fruit; and C, red mature wild type and Lefad7 fruit. Vertical
bars represent standard error (n between 4 and 6).

51

01
O

 

I Wild Type A
El Lefad7

3 ‘0"
¥l¥ﬁ

Mole percent
on
O

 

 

H
01
O
H
F"
H
H
0'1
N

H N
O O O
E__Ll ﬁi__' _L__

16:3 18:0 18:1 18:2 18:3
Fatty acid

 

 

Mole percent

l
l
l
l
l
l

1011 I

16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3

 

 

Fatty acid
70 7 7 ﬂ 7 _— ei-—A—-k
60 «1 C
.. 50l
C
(U
e 40 l
0) i
D.
2 30 a
‘z’
20 1
10 4
0 1 4—r_-.WEL ,_-—l *

16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3
Fatty acid

Figure 2. Fatty acid composition (mole percent). A, leaf; B, green immature fruit;
and C, red mature wild type and Lefad7 fruit.

52

10,000

 

 

)

to
o
o
o
>

I Wild Type
[I] Leiad7

I

8,000 -
7,000 ~
6,000 .
5,000 —
4,000 ~
3,000 4
2,000 4
1,000 i
0 4! r 1 -_.__4
Hexanal Z-3-Hexenal E-2-Hexenal

 

 

Headspace concentration (nL-L’1

 

 

 

 

 

 

 

16,000

 

14,000 — 3
12,000 -
10,000 I

1
8,000 I
l

 

6,000 1
I

4,000 7|
l

 

 

Headspace concentration (nL-L'l)

2,000

 

 

 

 

 

4
i
OF T

 

1 __T___——
Hexanal Z-3-Hexenal E-Z-Hexenal

 

Figure 3. C6-aldehyde concentration in (A) young and (B) mature wild type and
Lefad7 leaves (nL-L‘l). Vertical bars represent standard error (n=4). 2, cis; E,
trans.

53

A
n

1

Headspace concentration (nLoL

Headspace concentration (nL-L‘l)

200 -

 

180 T
160

140 ..
120 4‘
100 ‘
80 ‘
60

40 1
20 .

A

Hexanal

l
I Wild Type 1
El Lefad?

 

_TA -‘T_.._1 7 L,-__.._ 4

Z-3-Hexenal E-2-Hexenal

 

8,000 7
7,000
6,000 i
5,000 ‘
4,000 —i
3,000 «i

3 _i_

 

Hexanal

 

 

Z-3-Hexenal E-2-Hexenal

Figure 4. C6-aldehyde concentration in (A) immature green and (B) mature red
wild type and Lefad7fruits (nL-L‘l). Vertical bars represent standard error (n

between 4 and 6).

54

 

I Hexanal
I Z-3-Hexenal
CI E-2-Hexenal

.0 .0 P .o P 1"
U1 0‘ \1 co ‘0 O
I I l l 4

0.4 _
0.3 _

F)
N
l

  

Relative composition of C6-aldehydes

.0
H
l

   

Wild Type Lefad7 Wild Type Lefad? Wild Type Lefad7
LEAVES GREEN FRUIT RED FRUIT

P
O
l

 

 

Figure 5. Relative C6-aldehyde headspace concentration in leaves, immature
(green) and mature (red) wild type and Lefad7tomato fruits.

55

CHAPTER IV

GENETIC MANIPULATION OF LINOLENIC ACID CONTENT ALTERS C6-
VOLATILE SYNTHESIS AND SENSORY PERCEPTION OF TOMATO LEAVES AND

FRUITS

56

The lipoxygenase (LOX) pathway is responsible for producing volatile
compounds from fatty acid precursors critical to the aroma of tomato fruit (Mack
et al., 1987; Hildebrand, 1989; Feussner and Wasternack, 2002). The aldehydes
cis—3-hexenal, hexanal, and trans-Z-hexenal, and the alcohols 1-hexanol, and cis-
3-hexenol are produced as a consequence of lipid oxidation in tomato leaves and
fruits (Baldwin et al., 2000).

The formation of these compounds involves the action of a sequence of
enzymes in the LOX pathway on linoleic acid (18:2) and linolenic acid (18:3)
(Galliard and Matthew, 1977; Riley et al., 1996; Bate et al., 1998). These two
polyunsaturated fatty acids (PUFAS) are the most abundant in tomato leaves and
fruits, accounting for the 60 to 70% of the total fatty acid content (Gray et al.,
1999; Wang et al., 1996).

Substrate feeding experiments, using 18:2 and 18:3 on protein extracts or
homogenates from tomato fruit, have demonstrated that the addition of 18:2
enhances the production of hexanal and 1-hexanol, and addition of 18:3
enhances cis-3-hexenal, trans-Z-hexenal and cis-3-hexenol formation (Galliard
and Matthew, 1977; Riley and Thompson, 1998; Boukobza et al., 2001).

When fruits from different varieties of tomatoes have been analyzed, a
signiﬁcantly higher hexanal to hexenals (cis-3-hexenal and trans-2-hexenal) ratio
has been found in fruit which the ratio of 18:2 to 18:3 was higher (Gray et al.,

1999). Similarly, when fatty acid (FA) composition has been affected using

57

transgenic approaches, the ratio of C6-Volatiles has been affected in leaves and
fruits (Wang et al., 1996; Wang et al., 2001).

Analysis of the Lefad7 mutant, which lacks in active chloroplastic (03 FA
desaturase (LeFAD7), showed the importance of 18:3 in the biosynthesis of C6-
volatiles, speciﬁcally cis-3-hexenal, trans-Z-hexenal, and ci5-3-hexenol (Chapter
III). Leaves and fruits of the Lefad7 mutant produced signiﬁcantly lower
amounts of cis-3-hexenal, trans-Z-hexenal, and cis-3-hexenol, relative to wild
type, and higher levels of hexanal and 1-hexanol. It was possible to conclude
that the ratios and amounts of C6-saturated and unsaturated aldehydes and
corresponding alcohols produced by tomato leaves and fruits are in part
dependent on PUFA substrate levels, and the sensory quality of fruit was
negatively affected by LeFAD7 deﬁciency.

From analysis of the Lefaa’7 mutant, it was demonstrated that increased
levels of 18:2 in tomato leaves and fruits increases hexanal synthesis, and
reduced levels of 18:3 decreases the production of cis-3-hexenal and trans-2-
hexenal. Despite the fact that Leiad7 analysis provided important information
about the signiﬁcant role of FAS as substrate for C6-volatiles in tomato, it was
not possible assess the impact of intermediate levels of 18:2 and 18:3 on
volatiles. The availability of a tomato line with intermediate levels of PUFAs made
it possible to further investigate whether a moderated decrease in 18:3 content
would lead to a similar reduction in cis-3-hexenal and trans-Z-hexenal and a

similar shift in sensory quality.

58

The present study evaluated the relationship between FA composition,
volatile proﬁle and sensory quality between wild type, Lefad7 mutant and an
intermediate phenotype, which had a less dramatic reduction of 18:3 and
increase of 18:2, most likely due to co-Suppression of LeFad7. The impact of
ripening stage on C6-aldehyde formation derived from the LOX pathway in

tomato fruit of wild type, mutant and CO-suppressed lines was also investigated.

MATERIALS AND METHODS
PLANT MATERIAL

A line with reduced level of 18:3 and increased levels of 18:2 was
generated by Li et al. (2003), using Agmbacterium tumefasciens-mediated
transformation of Castlemart cotyledon with a construction containing genomic
Clone of LeFad7 under the control of 355 promoter of Cauliﬂower Mosaic Virus
(35S::LeFaa7). The phenotype of this line suggested that LeFad7 was co-
suppressed; it was named R183.

Tomatoes (L ycopersicon esculentum Mill.) CV. Castlemart (wild type),
mutant Lefad7, and R183 lines were grown in a greenhouse under constant day
and night temperature of 20°C. FA and volatile analyses were performed on
leaves and fruits. For fruits, different ripening stages were sampled. Ripening
stages of tomato fruits were visually determined according to color classiﬁcation

requirements (USDA, 1975): (1) green, (2) breaker, (3) turning, (4) pink, (5)

59

light red, and (6) red. In addition, the category (0) was included as immature

green fruit.

FATTY ACID COMPOSITION

Total lipids were extracted twice from 400 mg of leaf and 500 mg of fruit
tissue with 6 mL of hexane and once with 4 mL of 2:7 isopropanol:hexane.
Esteriﬁcation/transesteriﬁcation of fatty acids (FA) was performed with
methanoliczH2S04 (2.5%) to generate fatty acid methyl esters (FAMES) (Conconl
et al., 1996). FAMES were analyzed by gas Chromatography (GC)-ﬂame
ionization detector (FID) on a DB-23 capillary column (J&W Scientiﬁc, Folsom,

CA) as described by Bonaventure et al. (2003).

EVALUATION OF AROMA COMPOUNDS

Sample Preparation. One leaf disk (approximately 35 mg) from each
experimental unit was crushed in a 20-mL amber glass vial. After incubation for 3
minutes at 20°C, 1 mL of 50% C30; solution was added to stop further
enzymatic reactions. The vial was immediately closed with a Mininert valve
(Supelco, Bellefonte, PA), and stored at -20°C until volatile analysis. From each
tomato fruit, 10 g of pericarp tissue were homogenized in 5 mL of distilled water
using a Brinkmann Homogenizer (PT 10/35, Brinkmann Instrument Co,
Switzerland). After incubation for 3 minutes, 6 mL of 50% CaClz solution was

added and the mixture stored at -80°C. For volatile analysis, the homogenate

60

was thawed and 5 mL were placed in a 20-mL amber glass vial and Closed with a
Mininert valve. Homogenates were heated in a water bath held at 37°C for 30
minutes before analysis.

Volatile Analysis. Volatile compounds were extracted from the headspace
of the 20-mL vials using a solid phase micro extraction (SPME) ﬁber (65 pm
PDMS-DVB, Supelco, Bellefonte, PA) according to the method of Song et al.
(1998). The ﬁber was held in the vial for 3 minutes to allow absorption of volatile
compounds. Gas chromatograph-mass spectrometer (GC-MS) analysis was

performed as previously described (Chapter III).

SENSORY EVALUATION

For leaves, 1 g of tissue was crushed in a mill and incubated for 3 minutes
in a 20-mL glass-amber vial, and 1 mL of 50% CaClz solution was added after
incubation. Vials were kept at -20°C until sensory analysis. For fruits, 50 g of red
ripe tissue were blended with 50 mL of distilled water and 35 mL of 50% C30;
solution was added after 3 minutes. 4 mL from the homogenate were transferred
into 20-mL glass-amber vials and kept at -20°C until sensory test. Four triangle
tests comparing leaf and fruits from wild type and R183 lines, and leaf and fruits
from Lefad7 and R183 lines were performed as indicated by Meilgaard et al.
(1999) to determine if differences in volatile proﬁle were perceived by human
olfaction. As described in the previous chapter, panelists were asked to Choose

in a set of three which vial differed from the other two. All six possible set

61

combinations were used for each triangle test. A total of 39 subject were
assessed for this analysis.

A preference test was performed together with each triangle test for
leaves and fruits (Meilgaard et al., 1999). The panel was asked if the fruit
homogenate or crushed leaves they Chose as being different from the other two
was preferable or not-preferable. The preference test was conducted to qualify
the sensory differences detected in the triangle test. The hypothesis tested was
that more than 50% of panelists would prefer the aroma of one genotype over

another (Meilgaard et al., 1999).

STATISTICAL ANALYSIS

All data for FA content and headspace volatile concentration were
expressed as the mean :l: standard error of the mean. Data were analyzed using
one-way ANOVA by PROC MD(ED of a commercial statistical software package
(SAS version 8e, SAS Institute Inc., Cary, NC). Volatile data was transformed by
natural logarithm (In) to homogenize variances. Correlations between fatty acid
content and headspace concentration were performed with the statistical
functions of commercially available software (Microsoft Ofﬁce Excel 2003,
Microsoft Corporation, Redmond, WA). Statistical signiﬁcance of sensory

evaluation data was determined using Table T8 in Meilgaard et al. (1999).

62

RESULTS
FATTY ACID COMPOSITION

The wild type, Lefad? mutant, and R183 lines differed in FA composition
of their leaves (Figure 1). The mutant line had decreased levels of trienoic FA
compared to wild type leaves. Trienoic FAS from the R183 line were 20% of wild
type, but 2-fold higher than Lefad7 line. The total FA content was not affected by
LeFad7 mutation or by the likely co-suppression; the average FA content was 22
pmol-g'1 on a fresh weight basis. For leaves, the content of C18 PUFAS (18:2 and
18:3) was similar in the three lines, about 13 pmol-g'1 of fresh weight, but the
ratio of 18:3 to 18:2 was lower in mutant (0.07) and R183 lines (0.17) than wild
type (2.48) (Table 1).

The FA composition of the fruit for the three lines differed for dienoic and
trienoic fatty acids. Wild type levels of 18:3 were higher, and 16:2 and 18:2 were
lower than for mutant and R183 lines (Figure 2). Unlike leaves, no statistical
differences were detected between l.efad7 and 18:3 lines, although the average
for the R183 line for 18:3 and 18:2 were between those of wild type and mutant.
As in leaves, fruit total FA content was not affected by LeFad7 mutation or
probable CO-suppression, averaging 2.4 pmol-g'1 on a fresh weight basis. For the
three lines analyzed, the fruit FA content was approximately 10 times lower than
leaf content.

The relative content 18:2 and 18:3 in fruits at early stages of maturity

(pooled stages 0, 1 and 2) and later stages of development (pooled stages 3, 4,

63

5 and 6) did not differ for any of the tomato lines (Figure 3). Summed, 18:2 and
18:3 accounted for approximately 60% of total FAS in fruits, with 18:2 being the

most abundant for each line.

C6-ALDEHYDE VOLATILES

The aldehydes hexanal, cis-3-hexenal, and trans-Z-hexenal accounted for
99% of C6-volatiles produced by LOX pathway in leaves and fruits of all lines,
except for leaves of mutant plants where C6-aldehydes corresponded to 96% of
C6-volatiles.

Hexanal, derived from 18:2, was increased in mutant and R183 leaves,
29- and 23-fold, respectively, compared to wild type leaves (Figure 4).
Conversely, cis-3-hexenal was 60% and 2% that of the wild type in R183 and
Lefad7lines, respectively. trans-Z-Hexenal had a trend similar to that of cis-3-
hexenal; the headspace concentration was 50% and 13% of wild type in C0-
suppressed and mutant lines, respectively.

For fruits, hexanal production by the wild type was relatively stable during
the ripening process. For Lefad? and R183 lines, which had approximately 3-fold
higher levels of dienoic FA than wild type, however, hexanal level increased as
the fruit ripened, reaching a maximum of approximately 2,500 nL-L’1 at breaker
stage, 3-4 times higher than wild type fruit, and maintained a similar level

throughout ripening (Figures 5 and 6).

64

cis-3-Hexenal production by wild type fruit increased alter stage 2, the
mature green stage (Figure 7). Both mutant and R183 lines had much lower Cis-
3-hexenal production. Although Changes in the production of this aldehyde
started at the same ripening stage in the three lines, the maximum concentration
reached by R183 and mutant lines was 1/250th and 1/1,000th of the wild type,
respectively (Figures 5 and 7).

trans-Z-Hexenal, like the other C6-aldehydes, increased after the mature
green stage. Wild type fruit produced the greatest amounts, yielding a
headspace concentration close to 500 nL-L'1 at the turning stage, while R183 and
mutant lines only reached a maximum of 120 and 15 nL-L’l, respectively (Figures
5 and 8).

In leaves, the headspace concentration of total C6-aldehydes and alcohols
was reduced to 67% and 25% of the wild type for R183 and mutant lines,
respectively (Table 1). The ratio between C6-Volatiles derived from 18:3 (VFLn)
and C6-volatiles derived from 18:2 (VFLa) also decreased in both lines (Table 1,
column D). However, the magnitude of this change was not linearly related to
the Change in 18:3 and 18:2 ratio, as shown in columns B, D and F in Table 1.
An increase of 2.5-fold in the 18:3 to 18:2 ratio (anLa) between mutant and
R183 lines corresponded to a 9-fold Change in the VFLn:VFLa ratio. A 14.5-fold
increase in the anLa ratio between R183 and wild type lines was accompanied

by an increase of 40- to 50-fold in the VFLn:VFLa ratio.

65

In leaves, as 18:2 content increased from its lowest levels in wild type to
highest levels in R183 and mutant lines, hexanal production also increased
(Figure 9A). The increase in hexanal was best described by a linear relationship.
The production of cis-3-hexenal and trans-Z-hexenal also increased with
increased content of their precursor, 18:3 (Figures 9B and C). However, the
relationship appeared to be non-linear; a small increase in 18:3 content was
associated with a relatively larger increase in its oxidized products.

For fruits, the total C6-volatile production increased from immature green
to mature red fruits approximately 10-fold for wild type, and 6-fold for mutant
and R183 lines. There was a shift in the proportions of C6-volatiles derived from
18:3 in wild type and R183 lines (Table 2). Before the ripening process started,
most of the C6-Volatiles present in the headspace were derived from 18:2. From
immature to mature green stage, the proportion of C6-volatiles derived from
18:3 increased from 7% to 43% in wild type fruits, from 1% to 5% in co-
suppressed line fruits, and did not differ in mutant fruits. From mature green to
red ripe, 18:3-derived C6-volatiles increased from 43% to 93% in wild type
fruits, but did not Change in co-Suppressed and mutant line fruits. Although 18:3
content in R183 fruits was not signiﬁcantly higher than in mutant fruits, the small
difference of 18:3 available (1%) between those two lines, created a difference
of 4% in the C6-volatiles derived from 18:3.

Another odor-active aroma volatile produced by fruits, but derived from a

different pathway, 6-methyl-5-hepten-2-one, also increased during ripening for

66

all three tomato lines (Figure 10). The amount of 6-methyl-5-hepten-2-one
produced and its pattern of increase during ripening was not affected by altered

levels of 18:2 and 18:3 in Lefad7 mutant and R183 line.

SENSORY EVALUATION

Sensory evaluation showed that the differences in the leaf volatile proﬁle
between the three lines were readily detected by sensory panel (Table 3). For
fruits, the difference between mutant and R183 line was detected, but wild type
and co-suppressed lines could not be distinguished.

The aroma of the R183 line was preferred to that of the mutant, but it
was not preferred relative to wild type (Table 4). R183 fruit aroma was
determined to be preferred compared to that of the mutant. Although the
triangle test for wild type and R183 line fruits did not detect a perceptible
difference between both aromas, for the subjects who perceived a difference,

the aroma of wild type fruits was preferred to that of the R183 line fruits.

DISCUSSION

Mutation and loss of function of Chloroplastic (03 FA desaturase (LeFAD7)
in tomato allowed the comparison between volatile proﬁles of two extreme
phenotypes in terms of 18:2 and 18:3 content (Chapter III). This observation
was conﬁrmed and extended in the present study, noting the biosynthesis of

hexanal was highly increased in Lefad7line while the production of unsaturated

67

C6-aldehydes (hexenals) was very low compared to wild type for leaves and
fruits. In the current study, the perception of volatile differences between wild
type and Lefad? by sensory panels, and the preference for the wild type volatile
proﬁle over the mutant is also consistent with our previous report.

The analysis of the R183 line allowed us to further assess the relationship
between fatty acid composition, volatile production from the oxidation of 18:2
and 18:3, and the perception of these genetic and phenotypic manipulations by
sensory panel.

The increase in the relative abundance of 18:3 to 18:2 was associated
with a 10-fold greater increase in the ratio of 18:3-derived volatiles to 18:2-
derived volatiles. Thus, the increase in the relative availability of 18:3 for
oxidation was accompanied by a greater than expected increase in 18:3-derived
volatiles. This relative preference for 18:3 oxidation may be related to the
speciﬁcity of the reaction catalyzed by tomato hydroperoxide Iyase (HPL), which
is speciﬁc for 13-hydroperoxides (13-HPOS); no activity for 9-hydroperoxides (9-
HPOS) has been detected (Fauconnier et al., 1997; Matsui et al., 2001). A HPL
clone from tomato obtained by Howe et al. (2000) has a high speciﬁcity for the
13-hydroperoxide of 18:3 (13-HPOT). The relative activity of LeHPL for 13-HPOT
and 13-hydroperoxide of 18:2 (13-HPOD) were reported as 100% and 9.8%,
respectively (Fauconnier et al., 1997), reﬂecting a 10-fold greater speciﬁcity for
the formation of unsaturated C6-Volatiles. Lipoxygenase (LOX), like HPL, is

speciﬁc for the insertion of the dioxygen at position GB or O9 in

68

octadecadienoic and trienoic acids, but unlike HPL, it does not discriminate
between 18:2 or 18:3 (Feussner and Wasternack, 2002) and therefore should
have no direct inﬂuence on the ratio of unsaturated to saturated C6-aldehydes.
These considerations make HPL a reasonable candidate to explain the relatively
high sensitivity of unsaturated C6-aldehydes production to changes in 18:3
content (Table 1 and Figure 8).

Given that fruit fatty acid composition and content in fruit do not change
during ripening, the dramatic Change in the production of volatiles during
ripening likely reﬂects changes in LOX pathway activity, and most probably
related with the climacteric increase in ethylene production (Baldwin et al.,
1991). The increase of C6-volatiles during ripening could be explained by an
increase in 13-LOX activity. Riley et al. (1996) showed that total LOX activity
does not change between mature green and breaker fruits. However, this
Characterization did not discriminate between 9- and 13-LOX activity. Later,
Grifﬁths et al. (1999) demonstrated that expression of the 13-LOX, TomLOXC)
started only at breaker stage, being an ethylene dependent process. It has been
also demonstrated that formation of C6-volatiles in tomato fruit depends directly
on TomLongene expession (Grierson, 2004, personal communication).

Despite the fact that no C9-aldehydes are produced by tomato, it has
been extensively reported that most of the hydroperoxides produced by tomato
fruit LOX correspond to 9-HPOS of linoleic and linolenic acid in a ratio of 9-HPO

to 13-HPO of 95:5 (Galliard and Matthew, 1977; Hatanaka et al., 1992).

69

However, as in leaves, fruit HPL activity is highly speciﬁc for the Cleavage of 13-
HPOS, yielding C6-aldehydes and alcohols rather than 9 carbon compounds
(Suurmeijer et al., 2000; Matsui et al., 2001). In addition, HPL activity in fruits is
preferential for 13-HPOT over 13-HPOD (Suurmeijer et al., 2000), as it was
previously demonstrated in leaves (Fauconnier et al., 1997; Howe et al., 2000).
This explains how the ratio of 2:1 for 18:2 to 18:3 FAS, which serve as substrate
for C6-volatile biosynthesis, yielded a ratio of 7:93 for 18:2- to 18:3-derived C6-
volatiles in ripe wild type fruits (Table 2). Similarly, afﬁnity and kinetic analysis of
HPL explain the reduced total C6-Volatile concentration in the headspace (25% of
the wild type) for mutant and co-suppressed line fruits.

According to Riley et al. (1996), HPL activity in tomato fruits does not
change during ripening and is not affected by pH. Studies of developmental
expression of LeHPL have not been done speciﬁcally for fruit, although Howe et
al. (2000) reported that expression at immature green tomatoes is detected at
very low levels, and undetectable in mature green or red fruits. The higher
relative production of C6-Volatiles derived from 18:2, mainly hexanal, by
immature and mature green wild type, as well the relatively constant hexanal
production during ripening can not be fully explained given our present level of
understanding. Possibilities that need to be addressed include a Change in HPL
activity or afﬁnity, a different gene expressed in fruits coding for a HPL with no
preferences between 13-HPOT and 13-HPOD, or HPL-independent formation of

hexanal in fruits.

70

It was previously demonstrated that a non-trained panel was able to
detect the difference between wild type and Lefad7 mutant line volatiles, for
leaves and fruits. It was also found that the Change in volatile proﬁle caused by
LeFad7 mutation is detrimental for the sensory quality of tomato fruit. The
sensory evaluation in the current study demonstrated that small Changes in the
saturation level of FAS can result in a sufﬁcient shift in the volatile proﬁle in
leaves to be detected by olfaction. Despite the fact that leaf aroma is not an
important horticultural trait of tomato, there was a direct relation between
unsaturated C6-aldehyde content and aroma preference.

The ability of the sensory panel to discriminate between the fruit of the
R183 and mutant lines, but not between the R183 and wild type lines suggests
that perception may not be directly linked to absolute levels of volatiles. The data
suggest that low levels of unsaturated C6-aldehydes, such as produced by the
R183 line, are sufﬁcient to prevent detection of a difference in aroma perception
and preference from wild type. However, the near complete loss of 18:3-derived
volatiles, as in the mutant line, is readily perceived by a panel, negatively
affecting sensory quality. The lack of differences in hexanal production between
R183 and mutant lines suggests that the preference of R183 over the mutant line
was not due to hexanal. It is possible the slight increase in hexenals may have
been sufﬁcient for detection. However, the fact that our sensory panel did not
detect the difference in aroma between R183 and wild type fruit does not allow

us to be conclusive about the lack of sensory differences between those two

71

lines. A greater number of subjects could be necessary to elucidate that
hypothesis and support the preference of wild type over R183 fruit aroma
obtained in this study.

The sensory data do not necessarily support Obtaining additional
improvement in aroma quality with continued increase in the relative content of
18:3 over 18:2. However, the impact of such an enhancement also needs to be
evaluated in the context of cultural factors such as harvest maturity and storage

temperature and duration.

72

REFERENCES

Baldwin, E.A., Nisperos-Carriedo, MO. and Moshonas, MG. 1991. Quantitative
analysis of ﬂavor and other volatiles and for other constituents of two
tomato varieties during ripening. Journal of American Society for
Horticultural Science 116:265-269.

Baldwin, E.A., Scott, J.W., Shewmaker, GK. and Schuch, W. 2000. Flavor trivia
and tomato aroma: biochemistry and possible mechanisms for control and
important aroma components. HortScience 35(6):1013-1022.

Bate, N.J., Riley, J.C.M., Thompson, J.E. and Rothstein, 5.1, 1998. Quantitative
and qualitative differences in C6-volatile production from the lipoxygenase
pathway in an alcohol dehydrogenases mutant of Arabidopsis thaliana.
Physiologia Plantarum 104:97-104.

Bonaventure, G., Salas, J.J., Pollard, MR and Ohlrogge, J.B. 2003. Disruption of
the FATB gene in Arabidopsis demonstrates an essential role of saturated
fatty acids in plant growth. The Plant Cell 15(4):1020-1033.

Boukobza, F.,Dunphy, P.J. and Taylor, A.J. 2001. Measurement of lipid
oxidation-derived volatiles in fresh tomatoes. Postharvest Biology and
Technology 23:117-131.

Buttery, R.G., 1993. Quantitative and sensory aspects of ﬂavor of tomato and
other vegetables and fruits. In: Acree, TE. and Teranishi, R., eds, Flavor
science sensible principles and techniques, Chapter 8. American Chemical
Society, Washington, DC.

Conconi, A., Miquel, M., Browse, J.A. and Ryan, CA. 1996. Intracellular levels of
free linolenic and linoleic acids increase in tomato leaves in response to
wounding. Plant Physiology 111(3):97-803.

Fauconnier, M.L., Perez, A.G., Sanz, C. and Marlier, M. 1997. Puriﬁcation and
characterization of tomato (Lycopersicon esculentum Mill.) hydroperoxide
Iyase. Journal of Agricultural and Food Chemistry 45:4232-4236.

Feussner, I. And Wasternack, C. 2002. The lipoxygenase pathway. Annual
Review of Plant Biology 53:275-297.

Galliard, T. and Matthew, J. 1977. Lipoxygenase-mediated cleavage of fatty acid
to carbonyl fragments in tomato fruits. Phytochemistry 16:339-343.

73

Gray, D.A., Prestage, S., Linforth, R.S.T. and Taylor, A.J., 1999. Fresh tomato
speciﬁc ﬂuctuations in the composition of lipoxygenases-generated C6
aldehydes. Food Chemistry 64:149-155.

Grifﬁths, A., Barry, C., Alpuche-Solis, AG. and Grierson, D., 1999. Ethylene and
developmental signals regulate expression of lipoxygenase genes during
tomato fruit ripening. Journal of Experimental Botany 50(335):793-798.

Hatanaka, A., Kajiwara, T., Matsui, K. and Kitamura, A. 1992. Expression of
lipoxygenase and hydroperoxide Iyase activities in tomato fruits. Journal of
Biosciences 47(5-6):369-374.

Hildebrand, D.F., 1989. Lipoxygenases. Physiologia Plantarum 76:249-253.

Howe, G.A., Lee, G.I., Itoh, A., Li, L. and DeRocher, A. 2000. Cytochrome P450-
dependent metabolism of oxylipins in tomato: Cloning and expression of
allene oxide synthase and fatty acid hydroperoxide Iyase. Plant Physiology
123:711-724.

Li, C., Liu, G., Zhao, Y., Bauer, P., Ganal, M., Ling, H., Xu, C. and Howe, GA.
2003. The tomato SUPPRESSOR OF PROSYSTEMIN-MEDIATED
RESPONSES2 gene encodes a fatty acid desaturase required for systemic
induced resistance. Abstract presented at "Frontiers of Plant Cell Biology:
Signals and Pathways", the 22nd Symposium in Plant Biology Riverside
Convention Center Riverside, California (Jan 15-19, 2003).

Mack, A.J., Peterman, T.K. and Siedow, J.N., 1987. Lipoxygenase isozymes in
higher plants: biochemical properties and physiological role. Isozymes:
Current Topics in Biological and Medical Research 13:127—154.

Matsui, K., Fukutomi, S., Wilkinson, J., Hiatt, B., Knauff, V. and Kajwara, T. 2001.
Effect of overexpression of fatty acid 9-hydroperoxide Iyase in tomatoes
(LyCOpersicon esculentum Mill.). Journal of Agricultural and Food
Chemistry 49(11):5418-5424

Meilgaard, M., Civille, G.V., and Carr, B.T. 1999. Sensory Evaluation
Techniques, 3rd Edition. CRC Press, Boca Raton, Florida.

Riley, J.C.M., Willemot, C. and Thompson, J.E., 1996. Lipoxygenase and

hydroperoxide Iyase activities in ripening tomato fruit. Postharvest Biology
and Technology 7:97—107.

74

Riley, J.C.M. and Thompson, J.E., 1998. Ripening-induced acceleration of volatile
aldehyde generation following tissue disruption in tomato fruit. Physiologia
Plantarum 104:571-576.

Song, J., Fan, L.H. and Beaudry, R.M. 1998. Application of solid phase
microextraction and gas chromatography time-of-ﬂight mass spectrometry
for rapid analysis of ﬂavor volatiles in tomato and strawberry fruits.
Journal of Agricultural and Food Chemistry 46(9): 3721-3726.

Suurmeijer, C.N.S.P., Perez-Gilabert, M., van Unen, D.J., van der Hijden HTWM,
Veldink GA, and Vliegenthart, J.F.G. 2000. Puriﬁcation, stabilization and
characterization of tomato fatty acid hydroperoxide Iyase. Phytochemistry
53:177-185.

USDA. 1975. Color Classiﬁcation requirements in tomatoes. United Fresh Fruit
and Vegetables Association, USDA Agricultural Marketing Service Fruit and
Vegetable Division. The John Henry Company, Lansing, MI.

Wang, C., Chin, C.K., Ho, C.T., Hwang, C.F., Polashock, J. and Martin, GE,
1996. Changes of fatty acids and fatty acid derived compounds by
expressing the yeast A-9 desaturase gene in tomato. Journal of
Agricultural Food Chemistry 44:3399-3402.

Wang, C., Xing, J., Chin, C.K., Ho, CT. and Martin, CE, 2001. Modiﬁcation of

fatty acids change ﬂavor volatiles in tomato leaves. Phytochemistry
58:227-232.

75

Table 1. Ratios between total polyunsaturated octadecanoic acids (pmol-g") on a
fresh weight basis and total C6-volatiles (nL-L’l) in leaves. Abbreviations are as
follows: C18 PUFA, polyunsaturated fatty acid of 18 carbons (18:2 and 18:3); Ln,

linolenic acid; La, linoleic acid; VFLn, C6-volatiles derived from linolenic acid; and
VFLa, C6-volatiles derived from linoleic acid.

 

 

 

 

 

FA C6-VOLATILES RATIOS
LINE T‘l’ES'FCAla 5:31 Total (% of WT) WIPES/1L3 (C):(A) (D):(B)
(A) (B) (C) (D) (E) (F)
Wild Type 12.2 2.48 30,684 (100) 180.13 2,515 72.6
R183 12.6 0.17 20,687 (67) 4.23 1,641 24.9
Lefad7 13.4 0.07 7,580 (24) 0.48 565 6.9

 

Table 2. Relative amount of C6-volatiles derived from 18:3 at different fruit
ripening stages (% of total C6-volatiles).

 

 

 

RIPENING STAGE
LINE Immature green (0) Mature green Ripening
(1) (2 to 6L
Wild Type 7 43 93
R183 1 5 5
Lefad7 1 1 1

 

76

Table 3. Signiﬁcance for. triangle test in leaves and fruits.

 

 

 

Number of subjects
Correct Total p-Value1
Leaves
Mutant and R183 23 39 <0.0005
Wild Type and R183 27 39 <0.0005
Fruits
Mutant and R183 28 39 <0.0005
Wild Type and R183 16 39 >0.1000

 

1 T8 “Critical Numbers of Correct Responses in a Triangle Test” (Meilgaard et al., 1999)

Table 4. Signiﬁcance for preference test in leaves and fruits.

 

 

 

Number of subjects
Better Total p-Value1
Leaves
R183 better than Mutant 11 23 <0.0500
Wild Type better than R183 18 27 <0.0005
Fruits
R183 better than Mutant 17 28 <0.0010
Wild Type better than R183 11 16 <0.0010

 

1 T8 “Critical Numbers of Correct Responses in a Triangle Test” (Meilgaard et al., 1999)

77

 

14 T a
l IMMWm

12 4' b
. El Lefad7

10 l I R183

 

 

 

 

16:0 16:1 16:2 16:3 18:0 18:1
Fatty acid

Figure 1. Leaf fatty acid composition (pmol-g'l) in a fresh weight basis. Columns

within fatty acids with different letters are signicantly different (P<0.05). Vertical
bars represent standard error (n=4).

18:2 18:3

\l
O

 

 

 

 

 

1 I Wild Type
60 “1 El Lefad7
50 - IR183
E
840 —
a a a a
g 30 ~
2
20 -
10 i a a a b a a a a a
0 l h F T * 1
16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3
Fatty acid

Figure 2. Fruit fatty acid composition (mole percent). Columns within fatty acids
with different letters are signicantly different (P<0.05). Vertical bars represent
standard error (Wild Type n=20, Lefad7 n=17, R183 n=6).

78

80,7 7 77 77, ,77 777,77 777

 

 

 

 

 

 

 

 

 

70 ‘ D182
60 . —1_ —j-— +
‘5 50 -
8
840 .
2
g 30
20
10
0 . 7 7 .7 7 7 7 7 7
0-2 3-6 0-2 3-6 0-2 3-6

Wild Type Lefad? R183

Figure 3. Relative content of linoleic (18:2) and linolenic acid (18:3) in wild type,
Lefad7 and R183 fruits at different pooled ripening stages: (0—2) immature
green, green, and breaker, (3-6) turning, pink, light red, and red. Vertical bars
represent standard error (n between 4 and 6).

 

 

 

40,000 .

n . a I Wild T

'1' 35'000 1 l3 Lefad7ype

_l

530,000 4 'R183

.6 .

:5 25,000 ,1

b l

9‘3 20,000 4

C l

8 15,000 4

§ 0 0 1

10 o

e ' l a b

8 5,000 7 a b

:T: 1 c . c

, I. _ ,___7

Hexanal Z-3-Hexenal E-2-Hexenal

Figure 4. C6-Aldehyde headspace concentration (nL-L'l) from leaves. Columns
within aldehyde with different letters are signicantly different (P<0.05). Vertical
bars represent standard error (n=4). 2, US, E, trans.

79

 

10,000 i

 

 

 

 

 

 

3,000 . 5:22.9ee

g 3,000 — IR183

.5 7,000 _

5 6,000 _

5 5,000 _

5‘; 4,000 -

8 3,000 -

g 2,000 e

3% 1,003 : C b 1 a C b
Hexanal Z-3-Hexenal E-2-Hexenal

Figure 5. C6-Aldehyde headspace concentration (nL-L'l) from fruits. Columns
within aldehyde with different letters are signicantly different (P<0.05). Vertical

bars represent standard error (n=4).

 

 

3,500 r _
-B-Wlld Type alt a:
.3." 3 000 ttefaa’7 * *
_i I
'9- R183

 

2,000 l
l
1,500 4

 

Headspace concentration (nLo

'? t\
2,500 I O ..
1 - ‘3» ¢ ‘

1,000 I" d,/é
I 2 13:7 {.2

 

 

 

500 .’/ 3 T
0 ‘T T l l T T l _1
0 1 2 3 4 5 6
Ripening stage

Figure 6. Hexanal headspace concentration (nL-L") in fruits at different ripening
stages: (0) immature green, (1) green, (2) breaker, (3) turning, (4) pink, (5)

light red, and (6) red. Vertical bars represent standard error (n
5). Asterisk (*) represents signiﬁcant difference from wild type

80

between 3 and
(o=0.05).

 

I . 100
A12,000 1 -B-Wild Type

   
 
  

 

 

.. " 90 :E
3‘ ! Lefad7 h 80 g
a 10,000 I . u.
T; 1 R183 b 70 g
.9 (D
E 8,000 1 5 5o 8
u 3
I: n
g 6,000 7 1 50 §
8 g . 40 3
cu l o
g 4,000 I r 30 i
. :1
§ alnl: i L 20 '7
1 "I I. A. ** F
0 2:3 /1ﬂ . . . . '. 0
0 1 2 3 4 5 6
Ripening stage

Figure 7. cis-3-Hexenal headspace concentration (nL-L’l) in fruits at different
ripening stages: (0) immature green, (1) green, (2) breaker, (3) turning, (4)
pink, (5) light red, and (6) red. Vertical bars represent standard error (n between
3 and 5). Asterisk (*) represents signiﬁcant difference from wild type (o=0.05).
Double asterisk (**) represents signiﬁcant difference from R183 (o=0.05).

 

 

700 I

500 1 ﬁ- Lefad?

-e- R183

~B-Wild Type

 
  
     

U1

0

O
7177

l

 

 

 

 

L

_'l

5

C

g

.9 40047

C I

§ l

8 3001 al: at:

8

g 200

U)

'0 _LL

§ 100 V 57 5

** *f f*
O " T IT‘ #4
0 1 2 3 4 5 6
Ripening stage

Figure 8. trans-Z-Hexenal headspace concentration (nL-L") in fruits at different
ripening stages: (0) immature green, (1) green, (2) breaker, (3) turning, (4)
pink, (5) light red, and (6) red. Vertical bars represent standard error (n between
3 and 5). Asterisk (*) represents signiﬁcant difference from wild type (o=0.05).
Double asterisk (**) represents signiﬁcant difference from R183 (o=0.05).

81

 

 

Hexanal (nL-L")

y = 1247.1x - 1487.3
¥=0%%

f I

2.0 3.0 4.0
Linoleic Acid (umolog'1 FW)

1.0

O

 

6.0

 

20,000 1
15,000 e
‘7’ 10,000 e
5,000 4

 

0

y = 11505Ln(x) + 16029
R2 = 0.8333

0

[3E]

DB

 

 

0.0

800

i T

0.5 1.0

I i

1.5 2.0 2.5
Linolenic Acid (umol-g'1 FW)

1

3.0

 

700 e
600 -
500 -
400 —
300 -
200 I
100 _

E-Z-Hexenal (nL-L")

y = 207.92Ln(x) + 332.83
R2 = 0.9123

 

 

0

0.0

l

1.0 1.5 2.0 2.5

Linolenic Acid (umol-g'1 FW)

0.5

T

3.0

I

3.5

 

4.0

Figure 9. Correlations between fatty acid content and C6-aldehyde concentration
at the headspace from leaves of wild type (a), Lefad7(A), and R183 (o) lines:
(A) hexanal and 18:2; (B) cis-3-hexenal and 18:3; and (C) trans-Z-hexenal and

18:3.

82

 

  
   

 

 

 

 

1 -B-Wild Type
-A-Lefad7
”+05 ” -e-R183
l
,9 1.E+05 _
C
3
:6 l
0)
E 1.E+04 e
l
1.E+03 1
1.E+02 i l A f , . I
0 1 2 3 4 S 6
Ripening stage

Figure 10. 6-Methyl-5-hepten-2-one headspace concentration in fruits at different
ripening stages: (0) immature green, (1) green, (2) breaker, (3) turning, (4)
pink, (5) light red, and (6) red, represented as area units of the chromatogram
peak calculated at mass 108. Vertical bars represent standard error (n between 3
and 6).

83

CHAPTER V

REDUCED LeHPL ACTIVITY INCREASES 1-PENTEN-3-ONE AND
DIFFERENTIALLY DECREASES C6-ALDEHYDES BIOSYNTHESIS IN TOMATO
LEAVES AND FRUITS, RESULTING IN AN IMPACT ON SENSORY PERCEPTION

AND PREFERENCE

84

The aldehydes cis-3-hexenal, hexanal, and trans-Z-hexenal, the alcohols
l-hexanol, and cis-3-hexenol, and the ketone 1-penten-3-one are produced as a
consequence of lipid degradation (Baldwin et al., 2000). These lipid breakdown
products provide some of the most important ﬂavor notes to tomato (Table 1
Chapter II). Of all the aroma compounds produced by tomato, cis-3-hexenal has
the greatest potential impact based on its odor descriptor, its production rate,
and a relatively low human olfactory threshold (Buttery, 1993; Baldwin et al.,
2000). Our current understanding is that the formation of these C6-volatile
compounds involves the action of a sequence of enzymes [lipase, lipoxygenase
(LOX), hydroperoxide Iyase (HPL), isomerase, and alcohol dehydrogenase (ADH)]
on glycerolipids containing the fatty acids (FAS), linoleic acid (18:2) and linolenic
acid (18:3) (Galliard and Matthew, 1977; Riley et al., 1996; Bate et al., 1998), in
a process known as lipoxygenase pathway. The action of LOX has been reported
to be responsible for the formation of C5-Volatiles such as 1-penten-3-one, 2-
pentenal, and 1-penten-3-ol in other species such as soybean (Degousée et al.,
1995; Gardner et al., 1996).

In terms of concentration in the headspace, the most abundant C6-
volatiles are derived from 18:3 (Buttery, 1993). However, the most abundant FA
present in tomato fruit is 18:2 (Wang et al., 1996). This suggests an 18:3-
speciﬁcity of the LOX pathway in tomato that could be related to expression and

activity of LOXS and/or HPLS.

85

HPL activity has been extensively studied in tomato fruit (Hatanaka et al.,
1992; Riley et al., 1996; Suurmeijer et al., 2000) and leaf (Fauconnier et al.,
1997; Matsui et al., 2001). HPL activity in tomato fruit co-locates with the
microsomal fraction (Riley et al., 1996). Its activity does not Change from the
mature green to ripe stage, and Changes in pH between 6 and 8 do not affect it
(Riley et al., 1996). HPL activity was found evenly distributed from the skin to
the locular material (Hatanaka et al., 1992).

Tomato fruit HPL has activity on 13-hydroperoxy linolenic acid (13-HPOT)
and 13-hydroperoxy linoleic acid (13-HPOD); 9-hydroperoxides (9-HPOs) are not
substrate for this enzyme (Hatanaka et al., 1992; Matsui et al., 2001 ; Suurmeijer
etaL,2000)

In addition to having speciﬁcity toward 13-HPOs, leaf and fruit HPL
activity has a strong afﬁnity for 13-HPOT over 13-HPOD (Fauconnier et al., 1997;
Suurmeijer et al., 2000). The enzyme converts 13-HPOT to cis-3-hexenal 8 times
faster than 13-HPOD to hexanal (Suurmeijer et al., 2000).

The tomato gene coding HPL was recently identiﬁed and Cloned (Howe et
al., 2000). Characterization of the recombinant protein showed that it is speciﬁc
for the metabolism of 13-HPOs with approximately 10-fold greater afﬁnity for 13-
HPOT than 13-HPOD.

In this study, we reduced the LeHPL activity in leaves and fruit, most likely
by co-suppression of LeHPL, and analyzed the effect of lack of LeHPL activity on

the volatile proﬁle of leaves and fruits of two different cultivars of tomato. We

86

also assessed the sensory quality difference, in terms of aroma, between these
co-suppressed lines and wild type and Lefad7 mutant lines. In addition, we
propose the formation of CS-Volatiles as an additional fate for accumulation of

13-HPOT.

MATERIALS AND METHODS
PLANT MATERIAL

Tomatoes (Lycopersicon esculentum Mill.) CV. Castlemart, were grown in a
greenhouse under constant day and night temperature of 20°C. Cultivar Micro-
Tom was grown in a growth chamber under 16 hours of light at constant
temperature of 25°C. Leaves and ripe fruits were sampled for fatty acid and

volatile analysis.

A grobacterium tumefasciens—MEDIATED TRANSFORMATION

A 1.9 Kb cDNA fragment containing LeHPL was digested from pBluescript
SK(-) (Stratagene, La Jolla, CA) with BamHI and XhoI and Cloned into BamHI
and Xhol sites of the binary vector in pB1121 (Clontech, Palo Alto, CA) under the
control of 35S promoter of Cauliﬂower Mosaic Virus. The binary vector containing
the construction 35S::I.eHPL was transformed into Agrobacterium tumefasciens
strain AGLO as described by Li et al. (2003). AgrobacteriunT-mediated
transformation of wild type cotyledons of cultivars Castlemart and Micro-Tom

was performed as described by Li and Howe (2001).

87

Transgene insertion was conﬁrmed by PCR using two sets of primers: F1
(5'- CAA ATA GAG GAC CTA ACA GAA CTC GC -3') and R1 (5'- TAG CCA TGA ATC AAG AAA
GAT GTA GC -3') (Figure 1A and B), and F2 (5'- TAC ATC ‘lTl' CIT GAT TCA TGG CTA GC -
3') and R2 (5'- CAT CGC AAG ACC GGC AAC AGG -3') (Figure 1A and C). The expected
sizes of the ampliﬁed fragments were 1.4 Kb and 1.1 Kb, respectively.

Three independent fertile lines with reduced HPL activity were obtained
for Castlemart (CSH24, CSH27 and CSH37) and two for Micro-Tom (CSH3 and
CSHS). Alter tissue culture, those lines were grown and Characterized as

indicated for wild type cultivars.

ADDITION OF RECOMBINANT LeHPL TO FRUIT HOMOGENATE

Recombinant LeHPL was obtained from transformed Eco/i strain M15 with
expression construct pQE-HPL as described by Howe et al. (2000).

Red fruits (4 g) from wild type Micro-Tom and lines CSHS were blended
with 2 mL crude extract of recombinant protein (approximately 2 pg-mL‘1 total
protein) or 2 mL of distilled water (control) using a Brinkmann Homogenizer.
After 3 minutes of incubation in a capped 20-mL amber glass vial, 1.5 mL of 50%
C802 solution were added to stop further enzymatic reactions. The vial was
immediately Closed with a Mininert valve (Supelco, Bellefonte, PA).

HLP activity of the crude extract of recombinant protein was conﬁrmed
incubating 5 pg of 13S-hydroperoxy-9Z,11E,1SZ-octadecatrienoic acid (Cayman

Chemical, Ann Arbor, MI) diluted in 0.8 mL of distilled water with 0.2 mL of

88

crude extract of recombinant protein. After 3 minutes, 0.5 mL of 50% CaClz
solution were added to stop further enzymatic reactions.
Production of C5- and C6-aldehydes was analyzed for tomato

homogenates and controls by GC-MS as indicated below for volatile analysis.

FATTY ACID COMPOSITION

Esteriﬁcation/transesteriﬁcation of fatty acids (FA) was performed on
approximately 240 mg of leaf tissue with 2 mL methanoliC:HZSO4 (2.5%) to
generate fatty acid methyl esters (FAMES). FAMES were analyzed by gas
chromatography (GC)—ﬂame ionization detector (FID) on a DB-23 capillary

column (J&W Scientiﬁc, Folsom, CA) as described by Bonaventure et al. (2003).

EVALUATION OF AROMA COMPOUNDS

Sample Preparation. One leaf disk (approximately 7 mg) was crushed in a
20-mL amber glass vial. After incubation for 3 minutes, 0.5 mL of 50% CBCiz
solution were added to stop further enzymatic reactions. The vial was
immediately closed with a Mininert valve, and stored at -20°C until volatile
analysis. From each tomato fruit, 5 g of pericarp tissue were homogenized in 3
mL of distilled water using a Brinkmann Homogenizer (PT 10/35, Brinkmann
Instrument Co, Switzerland). After incubation for 3 minutes, 3 mL of 50% CBCiz
solution was added and stored at -80°C. For volatile analysis, the homogenate

was thawed and 4 mL were placed in a 20-mL amber glass vial and Closed with a

89

Mininert valve. Homogenates were heated in a water bath held at 37°C for 30
minutes before analysis.

Volatile Analysis. Volatile compounds were extracted from the headspace
from the 20-mL vial using a solid phase micro extraction (SPME) ﬁber (65 pm
PDMS—DVB, Supelco, Bellefonte, PA) according to the method of Song et al.
(1998). The ﬁber was held in the vial for 2 minutes to allow absorption of volatile
compounds. A gas chromatograph (HP 6890 Series GC, Hewlett-Packard Co.,
Wilmington, DE) was used for analyte separation, and a time-of-ﬂight mass
spectrometer (Pegasus II, Leco, St. Joseph, MI) was used for analyte detection,
identiﬁcation, and quantiﬁcation. The CG column (SupelcoWax-lO, Supelco,
Bellefonte, PA) was 30 m x 0.2 mm with a 0.2 pm-thick coating. In order to
quantify the headspace concentration of hexanal, cis-3-hexenal, trans-Z-hexenal,
1-hexanol, cis-3-hexenol, and 1-penten-3—one, gas standards were created using

authenticated compounds according to the method of Song et al. (1998).

SENSORY EVALUATION

Leaf tissue (0.5 g) from CV. Castlemart and CSH27 line was crushed in a
mill and incubated for 3 minutes in a 20 mL amber vial, and 1 mL of 50% CaClz
solution was added after incubation. For fruits, 100 g of red ripe tissue were
blended with 100 mL of distilled water and 33 g of CaClz were added after 3
minutes. 4 mL from the homogenate were transferred into 20-mL amber vials.

Vials were not refrigerated because the sensory test was performed 1 hour after

90

sample preparation. Four triangle tests comparing leaf and fruits from wild type
and reduced HPL activity (CSH) lines, and leaf and fruits from Lefad7 and LeHPL
co-suppressed lines were performed as indicated by Meilgaard et al. (1999) to
determine differences in volatile proﬁle as perceived by human olfactory sense.
As described in the previous Chapter, panelists were asked to choose in a set of
three which vial differed from the other two. All six possible set combinations
were used for each triangle test.

A preference test was performed together with each triangle test for
leaves and fruits (Meilgaard et al., 1999). The panel was asked if the
homogenate or crushed leaves they Chose as being different from the other two
was preferable or not-preferable. The preference test was conducted to qualify
the sensory differences detected in the triangle test. The hypothesis tested was
that more than 50% of panelists would prefer one aroma over the other

(Meilgaard et al., 1999).

STATISTICAL ANALYSIS

All data for fatty acid content and headspace volatile concentration were
expressed as the mean :L- standard error of mean. Data were analyzed using
one-way ANOVA by PROC MIXED of a commercial statistical software package
(SAS version 8e, SAS Institute Inc., Cary, NC). Statistical signiﬁcance of sensory

evaluation data was determined using Table T8 in Meilgaard et al. (1999).

91

RESULTS
FATTY ACID COMPOSITION

Leaf FA composition was not affected by reduction of HPL activity (Figure
2). As previously observed (Chapters III and IV) the most abundant FA was
linolenic acid (18:3), which comprised in average 43% of total FA. Total FA
content was not altered by reduction of HPL activity; the average of all tomato

lines was 21.5 pmol-g'1 on a fresh weight basis.

VOLATILE ANALYSIS

In Micro-Tom, C6-aldehydes accounted for more of 95% of total C6-
volatiles of wild type and CSH line leaves, and wild type fruits. However, for
CSH3 and CSHS fruits, C6-aldehydes were 56% and 61% of total C6-volatiles,
respectively. For Castlemart wild type leaves, more than 99% of total C6-volatiles
were C6-aldehydes; for CSH24 CSH27 and CSH37, they accounted for 98%, 88%
and 93%, respectively. In the case of fruits, they were more than 99% for all
lines.

Reduction of HPL activity resulted in a decreased production of saturated
and unsaturated C6-aldehydes by leaves in Micro-Tom and Castlemart (Figures 3
and 4). In addition, there was an increase between 4- and 12.5-fold in the
production of 1-penten-3-one along the different CSH lines.

In leaves, hexanal, derived from 18:2, comprised 9% and 8% of wild type

Micro-Tom C6-Volatiles in lines CSH3 and CSH5, respectively. Lines CSH24,

92

CSH27 and CSH37 produced 65%, 21% and 13%, respectively, of wild type
hexanal for Castlemart.

Leaf unsaturated C6-aldehydes were highly reduced in co-suppressed
lines. Relative to wild type cultivars, cis-3-hexenal concentration was reduced to
0.3% and 4% in CSH3 and CSHS, and to 10.5% in CSH24 and 0.3% in CSH27
and CSH37. Similarly, trans-Z-hexenal concentration was reduced to 2%, 16%,
17%, 4% and 3% in CSH3, CSHS, CSH24, CSH27 and CSH37, respectively.

For Micro-Tom CSH line fruits, there was a reduction of all C6-aldehydes
and an increase in 1-penten-3-one concentration in the headspace, similar to the
responses of leaves (Figure 5). However, hexanal production was much less
affected than in leaves, being on average 70% of the wild type. The unsaturated
C6-aldehydes cis-3-hexenal and trans-Z-hexenal were reduced to 1% and 14%
of the wild type, respectively. The increase of 1-penten-3-one was between 2.3-
to 3-fold greater than wild type.

In Castlemart fruits, similar results were obtained (Figure 6). However,
hexanal production was completely unaffected by the reduction of HPL activity,
which was probably decreased by LeHPL co-suppression. On the other hand, the
decrease on cis-3-hexenal was even more dramatic than in leaves, producing
only 0.01% to 0.07% of the wild type. The reduction of trans-2-hexenal was to
approximately 15% of wild type concentration. 1-Penten-3-one production was

increased 7- to 8-fold in the CSH lines.

93

Total concentration of C6-Volatiles was reduced in all CSH lines in a range
of 0.3% to 17% of wild type; the magnitude that was largely associated to the
extent of the reduction of cis-3-hexenal, the most abundant C6-Volatile in both
tissues, leaf and fruit, and in both cultivars, Micro-Tom and Castlemart. The
increase in the amount of 1-penten-3-one did not compensate for the decrease
of unsaturated C6-aldehydes. However, there was an increase of other C5-
volatiles that were not quantiﬁed in this study, caused by reduction of HPL
activity in leaves and fruits, such as 2-pentenal, 1-penten-3-ol and 2-penten-1-ol
(data not shown).

The requirement for HPL activity to produce unsaturated C6-Volatiles was
corroborated by addition of crude extract of recombinant LeHPL from Eco/i to
tomato fruit homogenate. Production of cis-3- and trans-Z-hexenal was increased

in CSHS fruit homogenate by addition of LeHPL (Figure 7).

SENSORY EVALUATION

Sensory evaluation showed that differences in leaf and fruit volatile proﬁle
between the three lines analyzed (wild type, Lefad7, and CSH27) were detected
by the non-trained panel (Table 1).

Preference test results showed that the volatile proﬁle of wild type was
considered better than CSH27 line in leaves and fruit (Table 2). CSH27 leaf
aroma was not better than Lefad7, but the fruit volatile proﬁle was found to be

better than the mutant.

94

DISCUSSION

Genetic manipulation of hydroperoxide lyase (HPL) activity has been
studied in tomato (Matsui et al., 2001) and other species from So/anaceae family
such as potato (Vancanneyt et al., 2001) and tobacco (Kessler et al., 2004).
However, this is the ﬁrst report of reduction of HPL activity in tomato and its
effect on volatile proﬁle and sensory quality of leaves and fruits.

AS expected from results obtained in previous studies, the reduction of
HPL activity, most likely by LeHPL silencing by co-suppression, produced a
dramatic reduction in C6-aldehyde biosynthesis from lipid oxidation through the
lipoxygenase pathway in leaves and fruits. When HPL was depleted in potato
(Vancanneyt et al., 2001) and tobacco (Kessler et al., 2004) by antisense, the
production of the C6-aldehydes hexanal and cis-3-hexenal was also reduced in
leaves, conﬁrming HPL activity as a major contributor to 13-HPO degradation in
leaves. Our results supported this hypothesis, however, when C6-aldehyde
production was analyzed in fruits, surprisingly the hexanal production was less
affected than in leaves for Micro-Tom and was unaffected in Castlemart. This
ﬁnding suggests that a signiﬁcant proportion of hexanal formation is LeHPL-
independent, perhaps resulting from the activity of an as yet unidentiﬁed HPL
that speciﬁcally acts on fruits and is speciﬁc for degradation of 13-HPOD. The
action of LOX is still required for hexanal formation, speciﬁcally TomLOXC,

considering that silencing of TomLOXC by anti-sense and co-suppression

95

markedly reduced both saturated and unsaturated C6-aldehydes in transgenic
Ailsa Craig tomato fruits (Grierson, 2004, personal communication).

The 13-HPOS was not measured, but the enhanced production of C5-
volatiles of all co-suppressed lines analyzed provides evidence that 13-HPOS
accumulated. 1-Penten-3-one is commonly found in the tomato fruit volatile
mixture (Baldwin et al., 2000). Potato plants with silenced HPL by antisense also
had an increase in the leaf production of the CS-volatlles ethyl vinyl ketone (1-
penten-3-one), 1-penten-3-ol, and cis-Z-pentenal (Vancanneyt et al., 2001).
Although we did not measure the concentration of 1-penten-3-ol and cis-Z-
pentenal, we observed an increase in them for fruits and leaves (data not sown).
Moretti et al. (2002) mentioned the formation of 1-penten-3-one from lipid
oxidation, suggesting 18:2 as its source, without involvement of LOX pathway.
However, Gardner et al. (1996) described the pathway for 1-penten-3-one and
2-pentenal formation in soybean having 18:3 as its precursor and being
synthesized as result of the action of LOX. Our results suggest that a similar
pathway exists in tomato for the formation of 1-penten-3-one (Figure 8), with
LOX oxidizing 18:3 to form 13-HPOT. HPL and LOX would be expected to
compete for 13-HPOT as substrate. It seems likely that HPL would tend to out-
compete LOX for 13-HPOT in wild type tissue, given that the bulk of 18:3-derived
volatile is cis-3-hexenal. According to Degousée et al. (1995), as HPL activity is
reduced, as in co-suppressed lines, LOX could form a pentadienyl radical that is

further converted to the alcohols 1-penten-3-ol and 2-penten-1-ol. Those

96

alcohols could be oxidized by alcohol dehydrogenase (ADH), yielding 1-penten-3-
one and 2-pentenal, respectively. Lefad7 mutant had a 300-fold reduction in the
amount of 1-penten-3-one produced (data not shown), conﬁrming 18:3, and not
18:2, as precursor of this CS-volatile.

We demonstrated in our previous studies that reduction of unsaturated
C6-aldehydes was detrimental for tomato fruit sensory quality (Chapters III and
IV). However, the analysis of the mutant Lefad7 and CSH lines was not able to
distinguish between the effect of cis-3- and trans-Z-hexenal reduction, hexanal
increase, or both. The CSH lines provided the opportunity to evaluate fruit aroma
for which hexanal synthesis remained constant, but unsaturated C6-aldehydes
were highly reduced, in fact, to an even greater extent than for Lefad7 (Chapter
III). The results of the preference test showed that lower levels of cis-3-hexenal
and trans-Z-hexenal were detrimental for tomato fruit aroma, as wild type aroma
was preferred to CSH27 line fruit aroma. The comparison between CSH27 and
Lefad7 fruit showed that not only the decrease of unsaturated C6-aldehydes
reduced sensory quality, but also the increase in hexanal in Lefad7fruits may
have further eroded sensory quality, given the preference for CSH27 line. It
seems that the increase of 1-penten-3-one in fruit was less inﬂuential in the
reduction of preference than the increase of hexanal. A different Situation was
encountered in leaves, where the increased levels of 1-penten-3-one on the
volatile mixture were more detrimental for the aroma perception than the

increase in hexanal of Lefad7. One of the odor descriptors of 1-penten-3-one is

97

‘pungent’. Other C5-volatiles have a similar descriptor. Interestingly in many of
the comments from preference test in leaves, it was mentioned that CSH line
was not preferred because of a pungent odor. Additional research using
descriptive sensory analysis is needed to improve our understanding of the

impact of manipulating LeHPL expression on tomato fruit.

98

REFERENCES

Baldwin, E.A., Scott, J.W., Shewmaker, GK. and Schuch, w. 2000. Flavor trivia
and tomato aroma: biochemistry and possible mechanisms for control and
important aroma components. HortScience 35(6):1013-1022.

Bate, N.J., Riley, J.C.M., Thompson, J.E. and Rothstein, S.J., 1998. Quantitative
and qualitative differences in C6-Volatile production from the lipoxygenase
pathway in an alcohol dehydrogenases mutant of Arabidopsis thaliana.
Physiologia Plantarum 104:97-104.

Bonaventure, G., Salas, J.J., Pollard, MR. and Ohlrogge, J.B. 2003. Disruption of
the FATB gene in Arabidopsis demonstrates an essential role of saturated
fatty acids in plant growth. The Plant Cell 15(4):1020-1033.

Buttery, R.G., 1993. Quantitative and sensory aspects of ﬂavor of tomato and
other vegetables and fruits. In: Acree, TE. and Teranishi, R., eds, Flavor
science sensible principles and techniques, Chapter 8. American Chemical
Society, Washington, DC.

Degousée, N., Triantaphylides, C., Starek, S., Iacazio, G., Martini, D., Bladier, C.,
Voisine, R. and Montillet, J.L. 1995. Measurement of thermally produced
volatile alkanes - an assay for plant hydroperoxy fatty-acid evaluation.
Analytical Biochemistry 224(2):524-531.

Fauconnier, M.L., Perez, A.G., Sanz, C. and Marlier, M. 1997. Puriﬁcation and
characterization of tomato (Lycopersicon esculentum Mill.) hydroperoxide
Iyase. Journal of Agricultural and Food Chemistry 45:4232-4236.

Galliard, T. and Matthew, J. 1977. Lipoxygenase-mediated Cleavage of fatty acid
to carbonyl fragments in tomato fruits. Phytochemistry 16:339-343.

Gardner, H.W., Grove, NJ. and Salch, Y.P. 1996. Enzymic pathway to ethyl vinyl
ketone and 2-pentenal in soybean preparations. Journal of Agricultural
and Food Chemistry 44(3):882-886.

Hatanaka, A., Kajiwara, T., Matsui, K. and Kitamura, A. 1992. Expression of

lipoxygenase and hydroperoxide Iyase activities in tomato fruits. Journal of
Biosciences 47(5-6):369-374.

99

Howe, G.A., Lee, G.I., Itoh, A., Li, L. and DeRocher, A. 2000. Cytochrome P450-
dependent metabolism of oxylipins in tomato: Cloning and expression of
allene oxide synthase and fatty acid hydroperoxide Iyase. Plant Physiology
123:711-724.

Kessler, A., Halitschkle, R. and Baldwin, LT. 2004. Silencing the jasmonate
cascade: induced plant defenses and insect populations. Science 305:365-
368.

Moretti, C.L., Baldwin, E.A., Sargent SA. and Huber, DJ. 2002. Internal bruising
alters aroma volatile proﬁles in tomato fruit tissues. Hortscience
37(2):378-382.

Li, L. and Howe, GA. 2001. Alternative splicing of prosystemin pre-mRNA ,
produces two isoforms that are active as signals in the wound response
pathway. Plant Molecular Biology 46(4):409-419.

Li, C., Liu, G., Zhao, Y., Bauer, P., Ganal, M., Ling, H., Xu, C. and Howe, GA.
2003. The tomato SUPPRESSOR OF PROSYSTEMIN-MEDIATED
RESPONSES2 gene encodes a fatty acid desaturase required for systemic
induced resistance. Abstract presented at "Frontiers of Plant Cell Biology:
Signals and Pathways", the 22nd Symposium in Plant Biology Riverside
Convention Center Riverside, California (Jan 15-19, 2003).

Matsui, K., Fukutomi, 5., Wilkinson, J., Hiatt, B., Knauff, V. and Kajwara, T. 2001.
Effect of overexpression of fatty acid 9-hydroperoxide Iyase in tomatoes
(Lycopersicon esculentum Mill.). Journal of Agricultural and Food
Chemistry 49(11):5418-5424

Meilgaard, M., Civille, G.V., and Carr, B.T. 1999. Sensory Evaluation
Techniques, 3rd Edition. CRC Press, Boca Raton, Florida.

Riley, J.C.M., Willemot, C. and Thompson, J.E., 1996. Lipoxygenase and
hydroperoxide Iyase activities in ripening tomato fruit. Postharvest Biology
and Technology 7:97-107.

Song, J., Fan, L.H. and Beaudry, R.M. 1998. Application of solid phase
microextraction and gas Chromatography time-of-ﬂight mass spectrometry
for rapid analysis of ﬂavor volatiles in tomato and strawberry fruits.
Journal of Agricultural and Food Chemistry 46(9): 3721-3726.

100

Suurmeijer, C.N.S.P., Perez-Gilabert, M., van Unen, D.J., van der Hijden HTWM,
Veldink GA, and Vliegenthart, J.F.G. 2000. Puriﬁcation, stabilization and
Characterization of tomato fatty acid hydroperoxide Iyase. Phytochemistry
53:177-185.

Vancanneyt, G., Sanz, C., Farmaki, T., Paneque, M., Ortego, F., Castanera, P.
and Sanchez-Serrano, J.J. 2001. Hydroperoxide Iyase depletion in
transgenic potato plants leads to an increase in aphid performance.
Proceedings of the National Academy of Sciences of the United States of
America 98(14):8139-8144.

Wang, C., Chin, C.K., Ho, C.T., Hwang, C.F., Polashock, J. and Martin, CE,
1996. Changes of fatty acids and fatty acid derived compounds by
expressing the yeast A-9 desaturase gene in tomato. Journal of
Agricultural Food Chemistry 44:3399-3402.

101

Table 1. Signiﬁcance for triangle test in leaves and fruits.

 

 

 

Number of subjects
Correct Total p-Value1
Leaves
Mutant and CSH 22 39 <0.0050
Wild Type and CSH 24 39 <0.0005
Fruits
Mutant and CSH 31 40 <0.0005
Wild Type and CSH 22 40 <0.0050

 

1 T8 “Critical Numbers of Correct Responses in a Triangle Test” (Meilgaard et al., 1999)

Table 2. Signiﬁcance for preference test in leaves and fruits.

 

 

 

Number of subjects
Better Total p-Value1
Leaves
CSH better than Mutant 9 22 >0.3000
Mutant better than CSH 13 22 <0.0500
Wild Type better than CSH 16 24 <0.0010
Fruits
CSH better than Mutant 19 31 <0.0050
Wild Type better than CSH 17 22 <0.0005

 

1 T8 “Critical Numbers of Correct Responses in a Triangle Test” (Meilgaard et al., 1999)

102

 

 

ﬂ Mm 355>—1 LeHPL a
PI»

< R1
F2> ‘ R2

1.4 Kb

1.1 Kb

 

Castlemart Micro—Tom 1 Castlemart B
CSH27 CHS37 WT CSH3 CSHS Pl Pr ‘ ' CSH24
I
- ' "' --
‘ Castlemart Micro-Tom - C
\NT CSH24 CSH27 CHS37 WT CSH3 CSHS Pl Pr
3.0 Kb
2.0 Kb
1.5 Kb
1.0 Kb
0.5 Kb

 

Figure 1. Construction 35S::LeHPL (A) and conﬁrmation of Agrobacterium-
mediated transformation by PCR in cultivars Castemalt and Micro-Tom: (B)
Primers F1 and R1, and (C) F2 and R2. Indicated sized in panel B correspond to
1Kb DNA ladder (Gibco Life Technology Inc., Rockville,MD). R, T-DNA right
border; Np, NOS promoter; nptII, neomycin phosphotransferase II; Nt, NOS
terminator; 35$, cauliﬂower mosaic virus 35S promoter; L, T—DNA left border.

103

 

1 IWild Type
10 1 ElCSH lines

l

l

 

 

8
TC) ,
a 61.
E l
:L
4
2 1
0 71 _—_—I77 1 71
16:0 16:1 16:2 16:3 18:0 18:1 18:2 18:3
Fatty acid

Figure 2. Fatty acid composition (pmol-g'l) on a fresh weight basis of tomato leaf

cv. Castlemart and CSH lines (CSH24, CSH27 and CSH37). Vertical bars represent
standard error (n=3).

104

 

 

   

 

 

 

 

 

 

 

 

 

 

1,000 1
1-Penten-3-one
8001 *
1 *
6001
1
400 I1
2001
0 l
7 l
6 1 Hexanal
5 1
e1
1.: 31
_J
5 21
.5 11 * *
E 0
J—ﬁ
5 25,000 1
E 1 Z-3-Hexenal
8 20,000 4
°’ 1
U
8 15,0001
.8 1
8 10,0001
I 1
5,0001 *
*
0
200
1 . E-2-Hexenal i
1601 e e 1
120 1
801
40 ‘ * ~
1 -
o 1 . ,
Wild Type CSH3 CSHS

Figure 3. Headspace concentration of 1-penten-3-one, hexanal, cis-3-hexenal,
and trans-Z-hexenal for Micro-Tom leaves. Vertical bars represent standard error
of mean (n=3). Asterisk represent signiﬁcant (o=0.05) difference from wild type.
2, cis; E, trans.

105

 

1,000

1-Penten-3-one
800

600

 

7 n

6- " Hexanal
5-

4_

37

2‘1 * *

11 7—77
01

16,000 —— e —
14,000 - Z-3-Hexenal

12,000 -
10,000 e

 

   

 

 

 

Headspace concentration (nL-L")

 

 

 

300 1 E-2-Hexenal

 

Wild Type CSH24 CSH27 CSH37

Figure 4. Headspace concentration of 1-penten-3-one, hexanal, cis-3-hexenal,
and trans-Z-hexenal for Castlemart leaves. Vertical bars represent standard error
of mean (n=3). Asterisk represent signiﬁcant (o=0.05) difference from wild type.

106

2,000

 

1-Penten-3-one
1,500 - *

1,000 7

500 -

 

     

 

Hexanal

 

Headspace concentration (nL~L")

 

 

 

 

 

 

Z-3-Hexenal
* =1:
E-2—Hexenal
* al:
Vlﬁld Type CSH3 CSHS

Figure 5. Headspace concentration of 1-penten-3-one, hexanal, cis-3-hexenal,
and trans-Z-hexenal for Micro-Tom fruits. Vertical bars represent standard error
of mean (n=3). Asterisk represent signiﬁcant (o=0.05) difference from wild type.

107

 

1:38: i 1-Penten-3-one

1,200 1 * * *

1,000
800 1

 

   

350 I Hexanal 1
l
l
l
l
l
l

50 "1
0 4‘ 7
20,000 1

   

#1

 

Z-3-Hexenal
15,000

10,000

Headspace concentration (nL-L")

5,000

717—4——

 

 

 

 

E-2-Hexenal

 

Wild Type CSH24 CSH27 CSH37

Figure 6. Headspace concentration of 1-penten-3-one, hexanal, cis-3-hexenal,
and trans-Z—hexenal for Castlemart fruits. Vertical bars represent standard error
of mean (n=3). Asterisk represent signiﬁcant (o=0.05) difference from Wild type.

108

 

4.5
4.0 _ ICSHS Control I

 

3-5 . ICSHS + Buffer
3.0 1
2.5
2.0 1
1.5 ~
1.0 4
0.5 —
0.0 —

 

DCSHS + HPL *

Relative Change

 

 

 

 

 

     

7|

1-Penten-3-one Hexanal Z-3-Hexenal E-2-Hexenal

 

 

 

Figure 7. Relative production of 1-penten-3-one, hexanal, cis-3-hexenal, and
trans-Z-hexenal in CSHS fruit homogenate after 3 minutes incubation with water
(CSHS Control); 50 mM K2PO4 pH 7.5 and 5% glycerol buffer (CSHS + Buffer);
and recombinant LeHPL crude extract (CSHS + HPL). Vertical bars represent
standard error of mean (n=3). Asterisk represent signiﬁcant (o=0.05) difference
from CSH5 Control.

109

1—1
13-HPOT l:> C6-Volatiles

13-oxo-cis- 9-Tridecadienoic acid

M330” + M

 

/ MOI-l
1-Penten-3-ol M 2-Penten-1-ol

HO
l—1
ADH + mm

0
{7 *1 'e
\./’
/ _O
1-Penten-3-one 2-Pentenal

O

Figure 8. Propose pathway for CS-Volatile formation in tomato. (Source:
Degousée et al., 1995; Gardner et al., 1996).

110

CHAPTER VI

CONCLUSIONS

111

In the current work, the formation of important volatile compounds in
tomato aroma mixture was studied. The main focus of my research was the
effect of fatty acid substrate availability on biosynthesis of C6-Volatile formation
through the lipoxygenase (LOX) pathway, by the analysis of plants defective in
(03 fatty acid desaturase activity. I also analyzed the last step of LOX pathway on
C6-volatile production by reduction of HPL activity, likely by co-suppression of
LeHPL gene. Finally, I evaluated if the impact of gene expression on volatile
proﬁle caused by substrate availability or HPL activity alterations were perceived
by non-trained panelists in leaves and fruits, and their effect on non-trained
panel preferences.

The main conclusions and remarks from my work are:

CHAPTER III:

. Mutation of the tomato chloroplastic 003 fatty acid desaturase (FAD) gene
(LeFad7), resulting in loss of FAD7 activity, decreased linolenic acid (18:3)
content and increased linoleic acid (18:2) content in tomato fruits. The
impact of the mutation on 18:3 content was greater for fruits than for
leaves.

. Total fatty acid (FA) content in fruits was not affected by LeFad7 mutation
or reduction of LeFAD7 activity.

. In vivo reduction of 18:3 directly impacted the biosynthesis of cis-3-
hexenal and trans-Z-hexenal, while the increase of 18:2 produced an

increase of hexanal formation. This conﬁrmed that cis-3-hexenal and

112

 

trans-Z-hexenal are derived from 18:3, and hexanal is, at least in part,
derived from 18:2.

. There was a linear relation between 18:2 content in leaves and production
of hexanal, while the relation between 18:3 and unsaturated C6-aldehyde
production was more curvilinear. This suggests that a higher speciﬁcity for
18:3-derived C6-volatiles resides in the LOX pathway.

. The preferential Cleavage of 13-hydroperoxy linolenic acid (13-HPOT) by

 

hydroperoxide Iyase (HPL) is most likely the reason for the abundance of
18:3-derived C6-volatiles relative to 18:2-derived C6-volatiles in tomato
tissue.

CHAPTER IV:

. There was a large increase in C6-volatiles accompanying the ripening
process in tomato fruits, most likely related to an increase in activity of
the LOX pathway, speciﬁcally TomLOXC.

. The small change in hexanal relative to unsaturated C6—aldehydes during
ripening, as well the relatively high concentration of this saturated C6-
aldehyde in immature stages, suggested the possibility of HPL-
independent hexanal production.

CHAPTER V:

. Reduction of LeHPL activity dramatically reduced the production of lipid

derived C6-Volatiles in leaves. In fruits, only unsaturated C6-volatile

production was markedly affected. This ﬁnding supports the possibility of

113

LeHPL-independent formation of hexanal in fruits but not in leaves, likely
the existence of a fruit-speciﬁc, novel HPL that has preferential activity
toward 13-HPO of 18:2.

. The production of 5-carbon volatile compounds, such as 1-penten-3-one,
2-pentenals and 1-penten-3-Ol, presumably by Cleavage of the 13-HPOT
by LOX likely result from accumulation of 13-HPOT caused by reduction of
LeHPL activity.

SUMMARY:

. Changes in volatile proﬁle caused by reduction in FAD7 and HPL activity
are readily detected by non-trained panels.

. These changes are detrimental to tomato aroma. The decline in aroma
preference was associated with the reduction of the unsaturated C6-
aldehydes cis-3-hexenal and trans-Z-hexenal, given that wild type tissue
was always preferred over Lefad7 mutant, R183, and CSH lines.

. Increase of hexanal seemed to also be negatively affecting fruit
preferences, considering that the CSH line, with normal levels of hexanal,
was preferred over Lefad7 mutant, with increased levels of hexanal.
However, the increase in 1-penten-3-one caused by reduction of LeHPL

activity can also be detrimental as was observed in leaves.

114

APPENDIX

115

Table 1. Original data for Figure 1 Chapter III. Values in pmol-g’1 fresh weight

(standard error).

 

 

 

 

 

 

 

 

 

 

 

 

Leaves

FA Wild Type Lefad7
16:0 6.01 (0.25) 5.81 (0.25)
16:1 1.44 (0.09) 1.37 (0.09)
16:2 0.18 (0.07) 2.37 (0.07)
16:3 2.44 (0.14) 0.02 (0.14)
18:0 0.53 (0.03) 0.47 (0.03)
18:1 0.64 (0.07) 0.93 (0.07)
18:2 4.93 (0.39) 16.90 (0.39)
18:3 12.58 (0.62) 2.01 (0.62)
Total 28.74 (1.46) 29.88 (1.46)

Green Immature Fruit

Wild Type Lefad7
16:0 0.68 (0.05) 0.83 (0.05)
16:1 0.03 (0.01) 0.04 (0.01)
16:2 0.00 (0.00) 0.01 (0.00)
16:3 0.01 (0.00) 0.00 (0.00)
18:0 0.09 (0.01) 0.09 (0.01)
18:1 0.18 (0.01) 0.15 (0.01)
18:2 1.05 (0.09) 1.94 (0.10)
18:3 0.47 (0.02) 0.02 @03)
Total 2.52 (0.16) 3.09 (0.18)

Red Mature Fruit

Wild Type Lefad7
16:0 0.80 (0.05) 0.77 (0.06)
16:1 0.04 (0.00) 0.04 (0.00)
16:2 0.00 (0.00) 0.00 (0.00)
16:3 0.04 (0.01) 0.00 (0.02)
18:0 0.10 (0.01) 0.08 (0.01)
18:1 0.13 (0.04) 0.07 (0.05)
18:2 1.03 (0.07) 1.58 (0.09)
18:3 0.54 (0.04) 0.06 (0.05)
Total 2.67 (0.17) 2.60 (0.22)

 

116

Table 2. Original data for Figure 2 Chapter III. Values in mole percent.

 

 

 

 

Leaves Green Immature Fruit Red Mature Fruit

FA Wild Type Lefad7 Wild Type Lefao7 Wild Type Lefad7
16:0 20.9 19.4 27.2 26.8 29.8 29.6
16:1 5.0 4.6 1.3 1.3 1.4 1.4
16:2 0.6 7.9 0.0 0.3 0.0 0.1
16:3 8.5 0.1 0.3 0.0 1.6 0.0
18:0 1.8 1.6 3.6 3.0 3.8 3.1
18:1 2.2 3.1 7.0 5.0 4.9 2.6
18:2 17.1 56.6 41.9 62.8 38.4 60.9
18:3 43.8 6.7 18.8 0.8 20.0 2.3
Total 100.0 100.0 100.0 100.0 100.0 100.0

 

Table 3. Original data for Figure 3 Chapter III. Values in nL-L‘1 (standard error).

 

Young Leaves

 

 

 

 

 

 

 

Volatile Wild Type Lefaa7
Hexanal 335.7 (200) 3623.8 (200)
cis-3-Hexenal 8571.3 (816) 1505.9 (816)
trans-2-

Hexenal 273.5 (32) 43.1 (32)
1-Hexanol 8.5 (11) 134.3 (11)
Cis-3-Hexenol 131.2 (17) 27.0 (17)
Total 9320.2 5334.0

Mature Leaves
Wild Type Lefad7

Hexanal 99.3 (492) 4503.4 (492)
cis-3-Hexenal 12812.1 (1762) 1915.6 (1762)
trans-Z- .

Hexenal 200.7 (45) 46.5 (45)
1-Hexanol 1.3 (3) 29.4 (3)
Cis-3-Hexenol 22.1 (2) 8.2 (2)
Total 13135.6 6503.2

 

117

 

Table 4. Original data for Figure 4 Chapter III. Values in nL-L‘1 (standard error).

 

Green Immature Fruit

 

 

 

 

 

 

 

Volatile Wild Type Lefad7
Hexanal 33.4 (37) 141.9 (37)
cis-3-Hexenal 16.5 (3) 0.0 (3)
trans-2-

Hexenal 5.4 (1) 0.9 (1)
1-Hexanol 0.0 (0) 1.0 (O)
cis-3-Hexenol 0.8 (0) 0.0 (0)
Total 56.0 143.8

Red Mature Fruit
Wild Type Lefad7

Hexanal 1397.9 (667) 7064.0 (504)
cis-3-Hexenal 3738.0 (304) 119.1 (230)
trans-2-

Hexenal 254.1 (74) 17.9 (56)
1-Hexanol 0.0 (0) 3.2 (0)
Cis-3-Hexenol 3.7 (1) 0.9 (0)
Total 5393.7 7205.1

 

Table 5. Original data for Figure 5 Chapter III. Relative content of aldehydes.

 

 

 

 

 

 

 

 

 

Leaves

Volatile Wild Type Lefad7
Hexanal 0.04 0.70
cis-3-Hexenal 0.93 0.29
trans-2-Hexenal 0.03 0.01
Green Immature Fruit

Wild Type Lefad7
Hexanal 0.60 0.99
Cis-3-Hexenal 0.30 0.00
trans-2-Hexenal 0.10 0.01

Red Mature Fruit

Wild Type Lefad7
Hexanal 0.26 0.98
Cis-3-Hexenal 0.69 0.02
trans-Z-Hexenal 0.05 0.00

 

118

Table 6. Original data for Figure 1 Chapter IV. Values in pmol-g’1 fresh weight
(standard error).

Table 7. Original data for Figure 2 Chapter IV. Values in mole percent (standard

error).

Table 8. Original data for Figure 3 Chapter IV. Values in mole percent (standard

error).

 

 

 

FA Wild Type Lefad7 R183
16:0 4.8 (0.18) 4.6 (0.14) 4.7 (0.14)
16:1 1.2 (0.08) 1.1 (0.05) 1.1 (0.08)
16:2 0.2 (0.01) 1.5 (0.03) 1.1 (0.13)
16:3 1.5 (0.06) 0.0 (0.00) 0.2 (0.04)
18:0 0.4 (0.04) 0.3 (0.03) 0.4 (0.01)
18:1 0.7 (0.06) 0.9 (0.07) 0.6 (0.08)
18:2 3.5 (0.19) 12.5 (0.08) 10.7 (0.76)
18:3 8.7 (0.19) 0.9 @21) 1.8 (0.38)
Total 21.0 (0.55) 21.8 (0.35) 20.7 (0.87)

 

 

 

FA Wild Type Lefad7 R183
16:0 30.3 (0.57) 30.3 (0.69) 31.5 (0.80)
16:1 1.1 (0.09) 0.9 (0.13) 0.7 (0.11)
16:2 0.1 (0.03) 0.4 (0.07) 0.4 (0.01)
16:3 0.7 (0.30) 0.0 (0.00) 0.0 (0.00)
18:0 3.7 (0.19) 3.2 (0.13) 3.3 (0.13)
18:1 5.2 (0.60) 3.5 (0.44) 4.0 (0.66)
18:2 39.8 (0.96) 60.0 (0.58) 57.7 (0.66)
18:3 19.1 (0.78) 1.2 (0.21) 2.4 (0.66)

 

 

 

 

 

Line Stage 18:2 18:3
Wild Type 0-2 40.718 (3.35) 19.601 (2.61)
3-6 39.318 (4.76) 18.883 (3.98)
Lefad? 0-2 61.142 (2.80) 0.977 (0.36)
3-6 59.345 (2.02) 1.3355 (1.06)
R183 0-2 56.626 (1.13) 1.5303 (0.29)
3-6 57.88 (2.02) 2.808 (1.92)

 

119

Table 9. Original data for Figure 4 Chapter IV. Values in nL-L‘1 (standard error).

 

 

Volatile Wild Type Lefad7 R183

Hexanal 168.53 (58) 4857.8 (54) 3847 (175)
Cis-3-Hexenal 29738 (4268) 2363.7 (1001) 16386 (6355)
trans-Z-Hexenal 579.32 (40) 75.478 (19) 293.7 (78)
1-Hexanol 0.8718 (1) 257.47 (26) 111.62 (26)
cis-3-Hexenol 197.32 (38) 25.292 47) 48.547 48)

 

Table 10. Original data for Figures 5, 6 and 7 Chapter IV. Values in nL-L‘1
(standard error).

 

 

 

 

 

 

 

 

 

Wild Type
Stage Hexanal cis-3-Hexenal trans-2-Hexenal
0 793.2 (181) 0.0 (0) 58.9 (33)
1 780.2 (121) 664.7 (665) 30.2 (9)
2 611.3 (142) 8655.1 (2831) 382.4 (96)
3 723.9 (63) 9469.8 (2542) 496.9 (45)
4 678.5 (128) 8107.2 (1771) 438.3 (73)
5 694.8 (121) 7914.0 (1136) 490.0 (64)
6 950.0 (169) 6875.7 (1582) 534.3 (76)
Lefad7
Hexanal cis-3-Hexenal trans-Z-Hexenal
0 355.5 (253) 0.0 (0) 1.8 (2)
1 1179.4 (485) 0.5 (0) 4.0 (4)
2 2265.8 (347) 4.4 (3) 11.3 (6)
3
4 2341.8 (277) 5.3 (1) 12.4 (4)
5 2153.7 (148) 3.4 (1) 14.9 (2)
6 2461.2 (605) 3.1 (1) 13.8 (3)
R183
Hexanal cis-3-Hexenal trans-2-Hexenal
0 428.2 (136) 0.0 (0) 0.2 (0)
1 1145.5 (550) 6.8 (7) 82.7 (80)
2 2640.2 (60) 26.7 (10) 107.7 (15)
3
4 2214.9 (206) 38.4 (17) 119.6 (18)
5 2592.5 (418) 25.3 (9) 105.5 (26)
6 2083.8 (379) 18.7 (8) 109.2 (34)

 

120

Table 11. Original data for Figure 8 Chapter IV. Values in pmol-g'1 for FA and
nL-L‘1 for volatiles.

 

 

 

 

FA Volatiles
Line 18:2 18:3 Hexanal cis-3-Hexenal trans-2-Hexenal
WT 1.62 3.42 327 24976 683
WT 1.36 3.70 100 36797 526
WT 1.35 3.40 67 20122 508
WT 1.26 3.38 179 37057 600
Lefad7 5.03 0.29 4835 2738 105
Lefad7 4.93 0.61 4777 5040 112
Lefad7 5.00 0.25 5015 1032 46
Lefad7 5.07 0.27 4804 645 38
R183 5.17 0.36 4056 3711 105
R183 4.11 1.01 3389 25831 438
R183 4.07 0.96 3768 15421 401
R183 3.80 0.63 4175 7906 231

 

Table 12. Original data for Figure 9 Chapter IV. Values in units of area at mass
108.

 

Stage Wild Type Lefad7 R183
0 11097 (9769) 671 (582) 6261 (12835)
1 3378 (5361) 5296 (6605) 92380 (78744)
2 16479 (15050) 40905 (54273)
3 121682 (78784) 270935 (172994) 217950
4
5
6

 

308441 (392981) 1545189 (2308402) 618822 (530185)
1009204 (1 1 17839) 545300 (206206) 1305090 (1050963)
3219936 (2190700) 2334312 (2951054) 2303103 (2201022)

 

121

Table 13. Original data for Figure 2 Chapter V. Values in pmol-g’1 fresh weight
(standard error).

 

 

 

FA Wild Type CSH
16:0 4.33 (0.29) 4.60 (0.06)
16:1 1.91 (0.08) 1.79 (0.08)
16:2 0.15 (0.01) 0.13 (0.02)
16:3 1.65 (0.05) 1.36 (0.14)
18:0 0.44 (0.02) 0.39 (0.01)
18:1 0.46 (0.06) 0.40 (0.08)
18:2 3.37 (0.44) 3.33 (0.30)
18:3 9.37 (0.26) 9.24 (0.09)
Total 21.68 (0.92) 21.23 (0.53

 

Table 14. Original data for Figures 3 and 5 Chapter V. Values in nL-L’1 (standard
error).

 

 

 

 

 

 

 

 

 

Leaves
Volatiles Wild Type CSH3 CSHS
1-Penten-3-one 47.0 (2.1) 512.9 (131.4) 590.0 (217.1)
Hexanal 5.8 (0.8) 0.5 (0.2) 0.5 (0.1)
cis-3-Hexenal 19545.5 (1633.0) 60.5 (22.7) 869.2 (410.5)
trans-Z-Hexenal 171.6 (2.2) 3.7 (1.8) 28.0 (6.9)
1-Hexanol 1.3 (0.7) 1.5 (0.5) 0.5 (0.2)
Cis-3-Hexenol 76.5 (7.3) 2.1 (1.5) 10.8 (2.7)
Total C6-Volatile 19800.8 (1627.8) 68.4 (23.2) 908.9 (419.1)
Fruits
Wild Type CSH3 CSHS
1-Penten-3-one 456.6 (8.7) 1086.3 (14.6) 1402.7 (7.4)
Hexanal 720.9 (21.7) 518.5 (20.7) 490.4 (10.5)
cis-3-Hexenal 4632.8 (463.4) 38.3 (13.6) 70.8 (19.8)
trans-Z-Hexenal 333.8 (16.2) 39.4 (1.5) 53.2 (1.8)
1—Hexanol 447.1 (11.2) 475.9 (4.5) 395.5 (4.9)
Cis-3-Hexenol 0.6 (0.1) 0.1 (0.0) 0.1 (0.0)
Total C6-Volatile 6135.2 (493.8) 1072.1 (17.7) 1010.0 (17.9)

 

122

Table 15. Original data for Figures 4 and 6 Chapter V. Values in nL-L'1 (standard
error).

 

 

 

 

 

 

 

 

Leaves
Volatiles Wild Type CSH24 CSH27 CSH37
1-Peneten-3-one 70.0 (13.1) 796.0 (40.3) 525.9 (93.9) 310.1 (56.0)
Hexanal 5.8 (0.7) 3.8 (1.0) 1.2 (0.4) 0.8 (0.3)
cis-3-Hexenal 12780.9 (2198.2) 1342.3 (183.6) 34.8 (10.4) 40.4 (9.8)
trans-Z-Hexenal 302.0 (21.9) 50.5 (6.6) 11.1 (1.8) 8.6 (1.9)
1-Hexanol 1.3 (0.7) 1.9 (0.1) 1.3 (0.3) 0.6 (0.2)
cis-3-Hexenol 23.6 (7.0) 27.0 (4.8) 5.0 (1.4) 2.9 (0.5)
Total C6-Volatile 13113.6 (2181.9) 1425.5 (194.1) 53.4 (12.0) 53.3 (11.21
Fruits
Wild Type CSH24 CSH27 CSH37
1-Peneten-3-one 140.8 (1.9) 1018.0 (11.1) 995.3 (8.7) 1138.1 (5.8)
Hexanal 293.2 (15.2) 316.9 (5.4) 300.8 (4.6) 326.8 (17.7)
Cis-3-Hexenal 17053.4 (115.0) 12.5 (1.2) 4.3 (1.5) 4.4 (0.6)
trans-Z-Hexenal 201.8 (5.5) 36.3 (0.2) 16.1 (0.1) 38.6 (1.3)
1-Hexanol 0.0 (0.0) 0.0 (0.0) 0.3 (0.3) 0.0 (0.0)
CiS-3-Hexenol 0.2 (0.1) 0.0 (0.0) 0.0 (0.0) 0.0 (0.0)
Total C6-Volatile 17548.7 (111.4) 365.7 (6.3) 321.5 (5.9) 369.9 (19.0)

 

Table 16. Original data for Figure 7 Chapter V. Values in nL-L‘1 (standard error).

 

 

Volatiles CSH Control CSH + Buffer CSH + HPL
1-Peneten-3-one 1.000 (0.05) 1.110 (0.07) 0.963 (0.01)
Hexanal 1.000 (0.04) 1.254 (0.06) 0.842 (0.02)
cis-3-Hexenal 1.000 (0.02) 1.111 (0.04) 2.041 (0.11)
trans-Z-Hexenal 1.000 (0.05) 0.821 (0.08) 3.832 (0.15)

 

123

 

 

   

111111111111111:11