Volatile Analyses of Single-Strength Orange Juice Inoculated with Penicillium digitatum

MISSING IMAGE

Material Information

Title:
Volatile Analyses of Single-Strength Orange Juice Inoculated with Penicillium digitatum
Physical Description:
1 online resource (104 p.)
Language:
english
Creator:
Shook,Gabriel L
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Food Science and Human Nutrition
Committee Chair:
Rouseff, Russell L
Committee Members:
Goodrich, Renee M
Danyluk, Michelle D.
Roberts, Stephen M

Subjects

Subjects / Keywords:
active -- analyses -- aroma -- compounds -- contamination -- digitatum -- fungal -- juice -- orange -- penicillium -- volatile
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre:
Food Science and Human Nutrition thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Post-pasteurization contamination of orange juice does occasionally occur, resulting in appearance and aroma changes. Previous microbial research regarding citrus has focused on bacteria and fermentative yeasts. Molds were not considered a microbial problem because the majority of juice was 65 ?Brix concentrate and its low water activity did not favor mold growth. An increased potential for spoilage from mold was a result of the shift in consumer orange juice products from concentrate to single strength, higher water activity juice. Penicillium digitatum is responsible for up to 90% of the citrus fruit loss experienced during storage and readily grows in juice. This study examined the changes in volatiles, and their impact on aroma, produced by P. digitatum in single strength orange juice. The primary objective was to determine qualitative and/or quantitative differentiating volatiles in inoculated and uninoculated (control) samples. A secondary objective was to correlate differentiating volatiles with aroma active volatiles. These objectives were accomplished by sampling inoculated and control juices using headspace, solid phase microextraction and analyzing the volatile profiles with gas chromatography ? mass spectrometry, ? sulfur, and ? olfactometry. Total cloud destruction and an 11% reduction in ?Brix was observed in inoculated juices at four days. Methanol, ethyl propanoate, ethyl 2-methylbutanoate, 3-methylbutanol, (Z)-3-hexenol, and 1-octen-3-ol were qualitative differentiating volatiles found only in inoculated samples. Neryl acetate was a qualitative differentiating volatile observed only in control samples. Other volatiles, including sulfur volatiles, were found to be quantitative differentiating volatiles, either greater than or less than that in the control. Neryl acetate was a differentiating volatile that also had aroma activity. The net effect of P. digitatum growth in orange juice produced an overall decrease in total volatiles as well as total aroma active volatiles.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Gabriel L Shook.
Thesis:
Thesis (M.S.)--University of Florida, 2011.
Local:
Adviser: Rouseff, Russell L.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043442:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 VOLATILE ANALYSES OF SINGLESTRENGTH ORANGE JUICE INOCULATED WITH PENICILLIUM DIGITATUM By GABRIEL LOUIS SHOOK A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Gabriel Louis Shook

PAGE 3

3 To my fiance, for without you this would have been an interesting paper

PAGE 4

4 ACKNOWLEDGMENTS The first person I must acknowledge is Dr. Russell Rous eff, for without him I would not be where I am today. His knowledge and expertise guided me as I developed knowledge in flavor chemistry. His trust and devotion allowed me to push myself and learn through experience, always knowing he would be there as a safety net. Dr. Rouseff also understood the value of internships, especially in the current ec onomic situation, granting me a leave of absence to experience industry through my summer internship with Kraft Foods. This internship led to my first job offer, securing employment immediately upon graduation. The tools he bestowed upon me go beyond those necessary to understand flavor chemistry, but also include life lessons, which helped shape me as a person. My gratitude must also be extended to each of my committee me mbers: Dr. Jan Narciso, Dr. Ren e Goodrich Schneider, Dr. Steve Roberts and Dr. Michelle Danyluk. Dr. Narciso offered extensive use of her lab, equipment, and expertise in fungal microbiology. This allowed me to successfully complete the mic robial portion of this study while furthering my knowledge in food microbiology. My knowledge of food science grew immensely with each class I took with Dr. Goodrich. Her class on food toxicology sparked my interest in toxicology as a whole, which led me to Dr. Roberts. His classes in general toxicology and toxic substances established a strong foundation for my further interest in the field. Dr. Danyluk was also always present for support and offered the use of her lab during the government shutdown sc are. Without these members compassion and understanding, I would never have successfully completed my graduation requirements.

PAGE 5

5 I owe sincere gratitude to the lab technician in Dr. Rouseffs lab, Jack Smoot, and Dr. Narcisos lab, Chris Ference, who wer e behind the scenes making sure the show ran smoothly. They were always there to offer a helping hand and make sure I had the supplies I needed. If an instrument ever went offline, I could call on them for the help to get my experiment back up and runnin g. Both were even on call during my long nights and weekends if an issue ever arose in the lab. My sniffers, Dr. Xiaofen Du, Song Mei Cau, and June Rouseff, always had a smile on their faces when they were called upon to sit and sniff at the GC O for 30 minutes. I knew the task was not the most enjoyable and each had their own work to accomplish, so taking 30 minutes out of their day was a lot to ask. Without them, however, I would not have been able to successfully complete an objective of this study. To my lab mates, Whitney Johnson and Kyle Edwards, it has been quite a ride, and I appreciate you lending a helping hand and providing a listening ear. We formed a bond through our experiences that no one else will understand. I hope to keep in touch as the years progress and we each go our separate ways. Finally, none of this would have been possible without my family and friends. My mom and dad have been there through thick and thin, always supporting my endeavors without question. Zach and Jon, yo u are the fuel to my competitive fire and wi thout your amazing achievements I may not have been as inspired to set the bar so high for myself. To all my friends, I thank you for being there to lend your support and advice. This especially includes Marianne Fatica, who has been there from the beginning of my journey through graduate school, lending advice and support from her own experiences through this crazy ride of ups and downs. Lastly, I must acknowledge my fianc e

PAGE 6

6 Rachel, for without you I do not k now where I would be today. I do know this paper would not be as well written without your amazing editing abilities. We have had quite a journey so far and I cannot wait to travel down the road of life with you by my side.

PAGE 7

7 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 9 LIST OF FIGURES ........................................................................................................ 10 LIST OF ABBREVIATIONS ........................................................................................... 12 ABSTRACT ................................................................................................................... 13 CHAPTER 1 INTRODUCTION .................................................................................................... 15 2 LITERATURE REVIEW .......................................................................................... 19 Microbial Spoilage of Foods .................................................................................... 19 Penicillium digitatum ............................................................................................... 21 Overview .......................................................................................................... 21 Growth Conditions ............................................................................................ 21 Specificity to Orange Juice ............................................................................... 22 Association with Orange Juice ......................................................................... 23 Contamination .................................................................................................. 24 Known Effects in Citrus .................................................................................... 24 Enzyme production .................................................................................... 24 Cloud destruction, pH, and Brix changes ................................................. 25 Volatile metabolites and biotransformation ................................................ 26 Metabolomics and Chemosystematics .................................................................... 28 Overview .......................................................................................................... 28 Previous Applications ....................................................................................... 28 Volatile Analysis ...................................................................................................... 30 Overview .......................................................................................................... 30 H eadspace Solid Phase Microextraction .......................................................... 31 Application to Orange Juice .................................................................................... 32 3 MATERIALS AND METHODS ................................................................................ 34 Penicillium digitatum Preparation ............................................................................ 34 Sample Preparation ................................................................................................ 34 Sampling ................................................................................................................. 35 Brix and pH Analysis ............................................................................................. 35 Volatile Analysis ...................................................................................................... 36 Headspace Solid Phase Microextraction .......................................................... 36 Gas Chromatography Mass Spectrometry .................................................... 36

PAGE 8

8 GC MS general program ......................................................................... 37 GC MS limonene and SIR program ........................................................ 37 GC MS peak identification ....................................................................... 37 Gas Chromatography Sulfur .......................................................................... 38 Gas Chromatography Olfactometry ............................................................... 38 Aroma Impact Compound Identification ........................................................... 39 Data Treatment ................................................................................................ 39 4 PHYSICAL AND CHEMICAL CHANGES ............................................................... 43 Cloud Destruction ................................................................................................... 43 Change in pH .......................................................................................................... 44 Change in Brix ....................................................................................................... 44 5 CHANGES IN ORANGE JUICE VOLATILES DUE TO PENICILLIUM DIGITATUM ............................................................................................................ 48 Gas Chromatography Mass Spectrometry Volatiles ............................................ 48 Chromatographic Differences ........................................................................... 48 Volatiles of G reatest Difference ........................................................................ 49 Volatile Changes During Incubation ................................................................. 50 Qualitative differences: Volatiles present in only inoculated or control samples .................................................................................................. 50 Quantitative differences: Increasing volatiles ............................................. 51 Quantitative differences: Decreasing volatiles ........................................... 52 Quantitative differences: Intermediate volatiles .......................................... 53 Gas Chromatography Sulfur Volatiles .................................................................. 53 6 MECHANISMS OF VOLATILE FORMATION ......................................................... 75 Volatile Formation as a Result of the Metabolic Pathway ....................................... 75 Volatile Formati on as a Result of Enzymes ............................................................ 76 Volatile Formation as a Result of Bioconversions ................................................... 76 7 CHANGES IN ORANGE JUICE AROMA IMPACT VOLATILES DUE TO PENICILLIUM DIGITATUM ..................................................................................... 83 8 CONCLUSIONS AND FURTHER WORK ............................................................... 94 APPENDIX A VOLATILE TABLES ................................................................................................ 96 B AROMA IMPACT COMPOUNDS ........................................................................... 98 LIST OF REFERENCES ............................................................................................... 99 BIOGRAPHICAL SKETCH .......................................................................................... 104

PAGE 9

9 LIST OF TABLES Table page 4 1 pH and Brix percent differenc e fr om control ...................................................... 47 5 1 Differentiating volatiles ....................................................................................... 56 5 2 Qualitative differentiating volatiles ...................................................................... 65 5 3 Quantitative different iating volatiles that increased ............................................. 65 5 4 Quantitative differentiating volatiles that decreased ........................................... 66 5 5 Quantitative intermediate volatiles ...................................................................... 67 5 6 Differentiating sulfur volatiles .............................................................................. 68 7 1 Differentiating volatiles which impact aroma with aroma descriptors .................. 86 7 2 Differentiating volatiles which impact arom a with percent difference .................. 86 A 1 All peaks detected using GC MS ..................................................................... 96 A 2 All peaks detected using GC S ........................................................................ 97 B 1 All aroma impact compounds detected using GC O ........................................ 98

PAGE 10

10 LIST OF FIGURES Figure page 2 1 Fungal metabolic pathway s ................................................................................ 33 3 1 HS SPME volatile sampling apparatus setup ..................................................... 41 3 2 GC O instrument .............................................................................................. 42 4 1 Image of c ontrol and inoculated sample at incubation day four ......................... 46 5 1 MS chromatograms at incubation day one ......................................................... 58 5 2 MS chromatograms at incubation day four ......................................................... 59 5 3 Top 20 differentiating volatiles at incubation day one ......................................... 60 5 4 Top 20 differentiating volatiles at incubation day two ......................................... 61 5 5 Top 20 differentiating volatiles at incubation day three ....................................... 62 5 6 Top 20 differentiating volatiles incubation day four ............................................. 63 5 7 Overall trend in volatiles ..................................................................................... 64 5 8 Sulfur chromatograms at incubation day four ..................................................... 69 5 9 Overall trend in sulfur volatiles ............................................................................ 70 5 10 Major differentiating sulfur volatile s at incubation day one ................................. 71 5 11 Major differentiating sulfur volatile s at incubation day two .................................. 72 5 12 Major differentiating sulfur volatile s at incubation day three ............................... 73 5 13 Major differentiating sulfur volatile s at incubation day four ................................. 74 6 1 Esterification of acetic acid and ethanol to form ethyl acetate ............................ 78 6 2 Formation of ethanol from glucose ..................................................................... 78 6 3 Relationship between Brix and ethanol ............................................................. 79 6 4 Relationship between ethyl hexanoate and ethanol ........................................... 80 6 5 De esterification of ethyl hexanoate to form hexanoic acid and ethanol ............. 81 6 6 Bioconversion of limonene ................................................................................. 81

PAGE 11

11 Figure page 6 7 terpineol, terpinolene, and 4terpineol ................................................................................ 82 7 1 Trend in aroma impact compounds .................................................................... 87 7 2 Control and inoculated aromagram at incubation day one ................................. 88 7 3 Control and inoculated aromagram at incubation day f our ................................. 89 7 4 Difference in sample aroma intensity at incubation day one ............................... 90 7 5 Difference in sample aroma intensity at incubation day two ............................... 91 7 6 Difference in sample aroma intensity at incubation day three ............................ 92 7 7 Difference in sample aroma intensity at incubation day four .............................. 93

PAGE 12

12 LIST OF ABBREVIATIONS Aw Water activity DI Deionized DVB D ivinylbenzene FID Flame ionized detector GC MS Gas chromatography mass spectrometry GC O Gas chromatography olfactory GC S Gas chromatography sulfur LRI Linear retention i ndex OJ Orange juice PDA Potato dextrose agar PDMS P olydimethylsiloxane PFPD Pulsed flame photometric detector RT Retention time SDE Steam distillation extraction SPME Solid phase microextraction USFDA United States Food and Drug Administration

PAGE 13

13 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the R equirements for the Degree of Master of Science VOLATILE ANALYS ES OF SINGLESTRENGTH ORANGE JUICE INOCULATED WITH PENICILLIUM DIGITATUM By Gabriel Louis Shook August 2011 Chair: Russell L. Rouseff Major: Food Science and Human Nutrition Post pasteurization contamination of orange juice does occasionally occur, resulting in appearance and aroma changes. Previous microbial research regar ding citrus has focused on bacteria and fermentative yeasts. Molds were not considered a microbial problem because the majority of juice was 65 Brix concentrate and its low water activity did not favor mold growth. An increased potential for spoilage fr om mold was a result of the shift in consumer orange juice products from concentrate to single strength, higher water activity juice. Penicillium digitatum is responsible for up to 90% of the citrus fruit loss experienced during storage and readily grows in juice. This study e xamined the changes in volatile s, and their impact on aroma, produced by P digitatum in single strength orange juice. The primary objective was to determine qualitative and/or quantitative differentiating volatiles in inoculated and uninoculated (control) samples. A secondary objective was to correlate differentiating volatiles with aroma active volatiles These objectives were accomplished by sampling inoculated and control juices using headspace, solid phase mi croextraction and analyzing t he volatile profiles with gas chromatography mass spectrometry, sulfur, and olfactometry. Total c loud destruction and an 11% reduction in Brix was observed in inoculated juices

PAGE 14

14 at four days Methanol, ethyl propanoate, ethyl 2methylbut anoate, 3methylbutanol, (Z) 3 hexenol, and 1octen3 ol were qualitative differentiating volatiles found only in inoculated samples. Neryl acetate was a qualitative differentiating volatile observed only in control samples Other volatiles, including su lfur volatile s, were found to be quantitative differentiating volatiles, either greater than or less than that in the control. Neryl acetate was a differentiating volatile that also had aroma activity. The net effect of P digitatum growth in orange juic e produced an overall decrease in total volatiles as well as total aroma active volatiles.

PAGE 15

15 CHAPTER 1 INTRODUCTION Despite all precautions, post pasteurization contamination of orange juice (OJ) does occasionally occur and can produce changes in the juices appearance, aroma, and taste (Huis in't Veld 1996; Larsen and Frisvad 1995a; Parish 1997) The majority of microbial research regarding citrus fruits and products has focused on the detection and prevention of spoilage caused by bacteria and fermentative yeasts such as Lactobacillus and Saccharomyces species, respectively. Of primary concern is the production of volatile metabolites that contribute to off flavors in the juice. For example, a buttery off note, caused by diacetyl, is associated with the presence of bacteria, including Lactobacillus and Leuconostoc species, and an alcoholic off note is associated with the presence of fermentative yeasts, such as Saccharomyces (Parish 1991; Parish and Higgins 1989; Wyatt and others 1995) Molds were not considered a microbial problem initially with OJ, as the majority of juice globally distributed was 65 Brix concentrate. Fungi do not easily grow in the freezing conditi ons (< 0C) or low water activity (Aw) of the concentrate. With recent technological advances in processing and storage, a shift from concentrate to singlest rength, not from concentrate juice has occurred. This shift resulted in an increased potential f or spoilage from mold, once not considered a problem (Parish 1991; Wyatt et al. 1995) Molds associated with citrus include Penicillium digitatum and Penicillium italicum Both of these molds are capable of thriving under current storage temperatures (0 4C), Aw (>0.98), and increased shelf life (35 65+ days) of single strength, not from concentrate OJ (Filtenborg and others 1996; Narciso and Parish 1997; Parish 1991; Parish and Higgins 1989; Raccach and Mellatdoust 2007; Tournas and others 2006)

PAGE 16

16 P digitatum is r esponsibl e for up to 90% of the loss of citrus fruit during storage and is considered one of the most damaging and important pathogenic fungi of harvested citrus (Ariza and others 2002; Droby and others 2008; Kavanagh and Wood 1971; Pesis and Marinansky 1990) Spores of this fungus have been shown to readily grow in post pasteurized OJ (Eckert and Ratnayake 1994) The United States Food and Drug Administration (USFDA) requires a 5log pathogen reduction during processing of OJ; however, post pasteurization contamination is of great concern due to the high amounts of P digitatum observed in the atmosphere of juice processing plants (Dav and others 1981; Narciso and Parish 1997; Parish 1991; Pelser and Eckert 1977; Rac cach and Mellatdoust 2007; Tournas et al. 2006; Wyatt and Parish 1995) A survey of single strength, not from concentrate OJ in the Washington DC area found 22% of juice was contaminated with Penicillium species, including P digitatum (Parish 1991; Tournas et al. 2006) The gable top juice cartons containing the juice are also a source of contamination. The multilayer cartons are not aseptic and many species of fungi have been isolated from the cardboard. Unfinishe d edges of the packaging can allow fungi access to the juice. The organism grows and forms a mat with spore production upon reaching the juice/air interface at the top of the carton (Narciso and Parish 1997) P digitatum produces a wide spectrum of enzymes that are stable and active at low pH and low temperature (Alaa and others 1990; Bush and Codner 1968; Bush and Codner 1970; Filtenborg et al. 1996) These enzymes are known for causing softening and destruction of the citrus peel and analogous destruction in the juice cloud (Alaa et al. 1990; Barmore and Brown 1979; Wyatt et al. 1995) Enzymes associated with P

PAGE 17

17 digitatum have been sh own to catalyz e the bioconversion of volatile s, altering the metabolic profile (Adams and others 2003; Demyttenaere and others 2001; Janssens and others 1992) For example, P 450 monooxygenases are considered the primary enzymes causing the bioconversion o terpineol (Demyttenaere et al. 2001; Duetz and others 2003) The organism also produces additional secondary metabolites through various metabolic pathways which can impact aroma and taste (Ariza et al. 2002; French 1985; Fries 1973; Huis in't Veld 1996; Larsen and Frisvad 1995a) The specific volatile metabolite profile, also known as differentiating volatiles, produced by bioconversion and production of s econdary metabolites can be used to differentiate two organisms down to the species level (Larsen and Frisvad 1994; Larsen and Frisvad 1995b) For example, in the study performed by Vikram and others (2004b), dim ethyl ether and propanal were found to be specific to Penicillium growing on Cortland apples. In contrast, methyl acetate and styrene were associated with Monilinia on the same apple (Vikram and others 2004b) Studies have also shown the ability to differentiate various Penicillia species based on a specific or series of differentiating volatiles produced as the organism grows on media, including potato dextrose agar (PDA) (Larsen and Frisvad 1995b) Ethyl acetate was found to be the volatile produced in largest amount specific to P digitatum while P italicum could be distinguished by the presence of both ethyl acetate and isopentanol on PDA (Larsen and Frisvad 1995a) Metabolomics is the study of all the metabolites produced by a specific organism, while chemosystematics is the use of this information to identify and classify the

PAGE 18

18 organism (Cevallos Cevallos and others 2009) The use of differentiating volatiles to id entify various organisms, and the disease associated with the specific organism, has been examined in potato tubers, mangos, carrots, and apples. S tudies have found that organisms, and the resulting disease they cause, can often be differentiated from one another based on differentiating volatile s (Lui and others 2005; Moalemiyan and others 2007; Moalemiyan and others 2006; Vikram and others 2006; Vikram and others 2004a; Vikram et al. 2004b) Based on this and pr evious studies that demonstrated the ability of Penicillia to produce differentiating volatiles, the presence of P digitatum should be detectable in OJ through volatile analysis (Larsen and Frisvad 1995b; Schnrer and others 1999) Since fungi that contaminate juice initially grow while submerged, the present study focused on determining the changes in volatiles produced by P digitatum submerged in single strength, not from concentrate OJ. The primary objective w as to determine qualitative and/or quantitative differentiating volatiles in inoculated and uninoculated (control) samples. A secondary objective was to correlate differentiating volatiles with aroma active volatiles.

PAGE 19

19 CHAPTER 2 LITERATURE REVIEW Microbial Spoilage of Foods As food products are harvested, processed, and handled, the potential of microbial contamination by bacteria, yeasts, and molds increases. This contamination could lead to microbial spoilage of the food product rendering it unacceptabl e for consumption (Huis in't Veld 1996; Lacey 1989) Microbial spoilage of citrus fruits and products involves acid tolerant bacteria, yeasts, and molds. The primary bacterial contaminates include Lactobacillus a nd Leuconostoc species, which produce 2,3butanedione, also known as diacetyl, as a secondary metabolite. Diacetyl imparts a buttery flavor to the contaminated product and is considered an undesirable attribute. Yeast species associated with citrus include Saccharomyces and Candida spp. These are fermentative yeasts and will ferment sugars under anaerobic conditions to produce alcohols and other metabolic end products, all of which are undesirable. Various molds have been associated with citrus products including Penicillium Geotrichum Fusarium Cladosporium Trichoderma, and Tri chophyton (Parish 1991) In rare cases, pathogenic microbes such as Escherichia coli are introduced post pasteurization through contamination and lead to foodborne disease outbreaks (Parish 1997) All microorganisms have a series of pathways that allow them to adapt readily to their environment By products of such pathways are known as secondary metabolites. These secondary metabolites can alter the food products appearance, aroma, and taste (Huis in't Veld 1996; Larsen and Frisvad 1995c) Besides s afety, one of the primary concerns is the alteration of taste, which results from an increase or decrease in volatile s known as aroma impact compounds through enzymatic pathways specific to

PAGE 20

20 the individual organisms (French 1985; Fries 1973; Moalemiyan et al. 2007) Bioconversion is also one of the pathways that result in the synthesis or breakdown of aroma active compounds (Janssens et al. 1992) The majority of microbial research regarding citrus fruit and citrus products involves the detection and prevention of spoilage due to growth of lactic acid bacteria and fermentative yeasts mentioned previously (Parish and Higgins 1989; Wyatt et al. 1995) Molds were previously not an issue, as the majority of juice was concentrated to 65 Brix and frozen, creating an inhospitable environment due to low Aw and low storage temperature (Wyatt et al. 1995) With the increase in availability of single strength, not from concentrate OJ and technology that allows for greater shelf life, molds are now becoming a problem in citrus products, specifically OJ (Parish 1991; Parish and Higgins 1989) The aforementioned molds can utilize a variety of substrates such as pectin, carbohydrates, and organic acids found in OJ and have the ability to thrive in low pH (3 8) and low temperatures (0 30C) (Huis in't Veld 1996; Lacey 1989) The ability to tolerate these conditions associated with single strength, not from concentrate OJ affords them an advantage over other yeasts and bacteria (Lacey 1989) Prior studies that examined fungal growth of the two primary fungi associated with citrus P digitatum and P italicum on citrus fruit demonstrated the organisms ability to substantially soften the fruit and produce a range of secondary metabolites. No previous studies have examined volatiles produced by this organism in contaminated citrus juice, specifically OJ (Filtenborg et al 1996; Schnrer et al. 1999)

PAGE 21

21 Penicillium digitatum Overview P digitatum also known as green mold, is classified as a soft rot wound parasite that grows primarily on the surface of post harvest oranges. Green mold is considered the major cause of or ange fruit deterioration during storage and transportation and one of the most important and damaging pathogenic fungi of harvested citrus (Ariza et al. 2002; Cole and Wood 1970; Droby et al. 2008; Eckert and Ratnay ake 1994; Kavanagh and Wood 1971; Pesis and Marinansky 1990) R esponsible for 90% of citrus losses during storage, this mold has not been reported to cause progressive decay on any other fruits or vegetables (Ariz a et al. 2002; Droby et al. 2008; Eckert and Ratnayake 1994; Pelser and Eckert 1977) Post harvest decay begins when spores enter microinjuries in the flavedo of the orange peel and germinate. Once germination begins, a mycelia mat and yellow green spores develop and the fruit begins to soften (Bush and Codner 1970; Caccioni and others 1998) As the organism develops it produces enzymes and secondary metabolites that lead to the biotransformation of volatiles. This causes an increase or decrease in volatile metabolites present (Achilea and others 1985; Adams et al. 2003; Bush and Codner 1970) Growth Conditions Growth results when specific conditions and interactions of temperature, pH, Aw, and various stimulators are present (French 1985; Fries 1973) Filamentous fungi, especially P digitatum hav e been shown to survive and grow under extreme conditions, including low temperatures, pH, Aw and oxygen, not suitable for other microorganisms (Nar ciso and Parish 1997) P digitatum spores lay dormant on the surface of oranges for ext ended periods of time and grow rapidly upon peel injury

PAGE 22

22 (Droby et al. 2008) The organisms temperature growth range is 6 37C with optimum growth occurring between 20 25C; however, growth can occur at lower temperatures between 0 6C with extensive growth at 7.6C (Lacey 1989; Parish and Higgins 1989; Raccach and Mellatdoust 2007; Wyatt and Parish 1995) Studies have also shown the organisms tolerance f or acids associated with citrus. G ermination occurred 90% of the time in OJ with a pH of 3.5 with rapid germ tube development at pH 4. The organic acids in OJ buffer the juice, aid ing in the maintenance of a favorable growth pH (Filtenborg et al. 1996; Fries 1973; Pelser and Eckert 1977) These extreme temperature and pH conditions give this organism an advantage over others by decreasing competition, while components and metabolites specific to OJ facilitate and stimulate fungal infection (Ariza et al. 2002; Lacey 1989) Specificity to Orange Juice Sugars found in OJ, including fructose, glucose, sucrose, and xylose, were associated with over 2/3 of P digitatum growth, with glucose support ing 50 70% of germination (Filtenborg et al. 1996; Kavanagh and Wood 1971; Pelser and Eckert 1977) Pectin was also found to assist in the growth of the organism causing increased infection (Kavanagh and Wood 1971) Fungal growth increased with volatile and nonvolatile extracts emitted from wounded orange peel tissue. This demonstrated host selectivity. These extracts included myrcene, acetaldehyde, ethanol, ethylene, Lascorbic acid, citric acid, and Lmalic acid (Droby et al. 2008; Eckert and Ratnayake 1994; Filtenborg et al. 1996; French 1985; Kavanagh and Wood 1971; Pelser and Eckert 1977) Research performed by French (1985) and Fries (1973) ionone, and benzaldehyde were associated with the greatest stimulatory effects (French 1985; Fries 1973) One

PAGE 23

23 component alone, however, did not stimulate growth. Mixtures of var ious volatile and nonvolatile extracts at specific concentrations were needed (Droby et al. 2008; Eckert and Ratnayake 1994; Stange and others 2002) Once infection begins, the fungus possesses the ability to transform its environment making it more hospitable. It does this by inhibiting other organisms through activating or blocking enzymatic reactions, removing or introducing inhibitors, and influencing nutrient uptake. An example of this is P digitatum s abi lity to produce mycotoxins that are only harmful to bacteria (Filtenborg et al. 1996; Fries 1973) Association with Orange Juice Even under ideal extraction conditions, prepasteurized juice contains a wide range of microorganisms (Faville and Hill 1951) The USFDA requires a 5log pathogen reduction during processing, which is stated in the USFDA Juice Hazard Analysis and Critical Control Point Regulation. This pathogen reduction normally occurs during pasteuriza tion, in order to eliminate most pathogens and heat sensitive microbes (Narciso and Parish 1997; Raccach and Mellatdoust 2007; Tournas et al. 2006) Post pasteurization contamination, however, remains a concern (Parish 1991) Technological advances that allow for increased shelf life of single strength, not from concentrate OJ, such as the gable top multilayer carton, have led to a shift from 65 Brix juice concentrates to the more convenient pour and serve single strength, not from concentrate juic e. This shift made mold contamination, which was not a major concern with concentrated juice, a growing problem (Parish 1991; Wyatt et al. 1995) The storage conditions of single strength, not from concentrate juice involve nonfreezing conditions (0 4C) In addition, single strength, not from concentrate juice has a high er Aw (>0.98) then conce ntrate. These conditions allow for mold growth and spoilage

PAGE 24

24 (Parish 1991; Parish and Higgins 1989; Raccach and Mellatdoust 2007; Tournas et al. 2006) P digitatum has been i solate d from pasteurized juice. I n a recent sampling of pasteurized, single strength, not from concentrate OJ from grocery stores in the Washington DC area, 22% contained fungal contamination of Penicillium species (Parish and Higgins 1989; Tournas et al. 2006) Contamination S pores of P digitatum have been isolated from and are abundant in the atmosphere of citrus groves, packinghouses, and juice processing plants, and on food contact surfaces. Some i solated samples are tolerant of presently used fungicides including 2amino butane, sodium orthophenylphenate, thiabendazole, and benomyl, and an increasing incidence of resistance is occurring. The ubiquitous presence of spores is making post pasteurization contamination of OJ a primary concern, as neither cartons nor filling systems are typically aseptic (Dav et al. 1981; Narciso and Parish 1997; Pelser and Eckert 1977; Wyatt and Parish 1995; Wyatt et al. 1995) Gable top cartons themselves are an area of contamination. Fungi have been isolated from the cardboard layer used in their construction. Inappropriately sealed or unfinished areas of the cartons allow the organism access to the juice. Eventually, the germ tube will reach the juice/air interface at the top of the carton and a mycelia mat will form (Narciso and Parish 1997) Known Effects in Citrus Enzyme production P digitatum has the ability to produce a wide range and vast number of enzymes that are stable and active at low pH and low temperature. These enzymes drive reactions that result in the production and destruction of compounds, including volatiles

PAGE 25

25 some of which aid in the proliferation of the organism as previously discussed Even after the removal of the organism, the enzymatic reactions will continue until they are deactivated (Alaa et al. 1990; Bush and Codner 1968; Bush and Codner 1970; Filtenborg et al. 1996) Some of the enzymes associated with this organism as it grows on infected citrus fruit include pectin lyase, pectin methylesterase, pectin esterase, pectin transeliminase, olygalacturonase, exopolygal amylase, proteinases, and cytochrome P 450 enzymes (Alaa et al. 1990; Barmore and Brown 1979; Bush and Codner 1968; Cole and Wood 1970; Duetz et al. 2003; Kavanagh and Wood 1971) Cloud destruction, pH, and Brix changes Some of the enzymes produced, such as pectin lyase and pectin esterase, are 1,4 glycosidic bonds between galacturonic acid and carbohydrates This results in OJ cloud destruction (Alaa et al. 1990; Barmore and Brown 1979; Wyatt et al. 1995) The reaction of pectin esterase breaks down pectin into pectate and methanol (Bush and Codner 1968) The primary role s of these enzymes is to allow the mold greater access to the fruit once it has entered the microinjuries and breakdown compounds into simpler components for use as energy sources (Barmore and Brown 1979) These enzymes have shown acidity resistance (Barmore and Brown 1979; Patrick and Hill 1959) The process of peel softening and cloud destruction is considered a threestep process, beginning with the collapse of the cytoplasm, followed by the swelling of cell walls, and finally cell separation. When this occurs in OJ the viscosity decreases as the juice clarifies (Alaa et al. 1990; Barmore and Brown 1979) A slight

PAGE 26

26 reduction in Brix was noted after 12 days of P digitatum incubation at 10C, while the pH and t otal titratable acidity did not change (Wyatt et al. 1995) Volatile metabolites and biotransformation One hundred, ninety six volatile metabolites, consisting of monoterpenes, alcohols, esters, ketones, and alkenes, most of which are considered s econdary metabolites, have been associated with P digitatum inoculated citrus fruit Substrate composition greatly influenced the type and amount of metabolites produced (Ariza et al. 2002; Lacey 1989; Schnrer et al. 1999) Of the metabolites mentioned previously alcoho ls, including ethanol, 1octene3 ol, and methanol, are considered to be extraordinary indicators of this organism, as they are produced in large amounts through various pathways (Ariza et al. 2002; Larsen and Frisvad 1995a; Nilsson and others 1996; Pesis and Marinansky 1990; Schnrer et al. 1999) Following alcohols, e sters are the second largest group of volatile s identified and associated with this organism in whole fruit The production of alcohols and esters is the resul t of fungal metabolic pathways, seen in Figure 2 1, and enzymatically driven reactions (Ariza et al. 2002; Larsen and Frisvad 1995a; Schnrer et al. 1999) For example, an increase in ethyl acetate is considered a quantitative indicator for the presence of green mold. Ethyl a cetate formation result s from the metabolic pathway that bioconverts acetaldehyde to acetic acid by acetyl CoA. Acetic acid then undergoes an esterification reaction with ethanol to form ethyl acetate (Pesis and Marinansky 1990; Schnrer et al. 1999) Microbiologists have been aware of the ability of fungi to bioconvert compounds since the late 1960s. The mechanisms by which they accomplish these conversions are not well understood but include oxidation, reduction, dehydration, degradation, and

PAGE 27

27 C C bond formation reactions (Demyttenaere et al. 2001; Janssens et al. 1992) The majority of reactions are considered enzymatically driven, as they are not noted with simple acid catalyzation at pH 3.5, which is the pH of most OJ (Adams et al. 2003) One of the most well studied bioconversions involving P digitatum is the conversion of ( R ) (+) limonene to ( R ) (+) terpineol. ( R ) (+) limonene converts more readily then ( S ) ( ) limonene at pH 3.5. The conversion has low specificity and leads to the formation of other volatile s including trans carveol, cis carveol, trans limonene oxide, trans p menth2 en1 o terpinene, terpinolene, and menth3 en1 ol. At a temperature of 28C and pH 4.5, the bioconversion occurred within 8 12 hours, terpineol noted between 1 and 2 days of incubation. Products were not rec overed from a control, uninoculated flask at pH 3.5, confirming the reaction was enzymatically driven rat her than simply acidcatalyzed (Adams et al. 2003; Demyttenaere et al. 2001; Duetz et al. 2003; Tan and Day 1998a; Tan and Day 1998b; Tan and others 1998) The mechanism of this bioconversion remains a great debate. An early suggested mechanism included hydrolysases and hydratases; however, a newly proposed mechanism involves P 450 monooxygenases with the initial step being an epoxidation of the 89 double bond followed by reductive cleavage of the epoxide (Demyttenaere et al. 2001; Duetz et al. 2003) The majority of studies examining volatile metabolites produced by P di gitatum inv olve the whole, contaminated orange or media containing orange pumpout an industry term for OJ concentrates whose major volatiles have been removed during the concentration process The reactions and metabolites observed in infe cted oranges w ere shown to be unique and associated with

PAGE 28

28 individual species, furthering the idea that differentiating volatiles can be used as indicators of fungal growth (Larsen and Frisvad 1995a; Pesis and Marinansky 1990) Met abolomics and Chemosystematics Overview Metabolomics is the study of all metabolites in a system, while chemosystematics, also referred to as chemotaxonomy, is the classification and identification of organisms based on all the metabolites produced in a s ystem (Cevallos Cevallos et al. 2009) Volatile metabolites are metabolites in a gaseous phase or those that have a high vapor pressure allow ing for their liberation from a cell or substance into the headspace (Cai and others 2001; Hutchinson 1973) By focusing on the volatile metabolites produced by an organism, metabolomics and chemosystematics can be employed to develop a database of volatile s and methodology for examining volatile s for the use of identifying the microorganism based on specific differentiating volatiles (Cevallos Cevallos et al. 2009; Larsen and Frisvad 1994; Schnrer et al. 1999) Previous metabolomic and chemosystematic studies have shown microbes can be distinguished with a clear separation of taxa down to the species level, based on their differentiating volatile s (Larsen and Frisvad 1994; Larsen and Frisvad 1995b) Previous Applications The previously discussed concept of metabolomics for the detection of microbes has been applied to the food industry. One study examined Rus set Burbank potato tubers for the identification and discrimination of potato tuber diseases caused by Phytophthora infestans Pythium ultimum and Botrytis cinerea. Each organism produced two to three differentiating volatiles that allowed for their identification and subsequent identification of potato tuber disease (Lui et al. 2005) Other studies

PAGE 29

29 examined mangos for the abili ty to identify and discriminate stem end rot disease caused by Lasiodiplodia theobromae and anthracnose caused by Colletotrichum gloeosporioides Thirty seven volatiles were identified, with several being specific to each organism W hen common to both or ganisms, the volatile varied in its abundance (Moalemiyan et al. 2007; Moalemiyan et al. 2006) Carrots were examined for the ability to identify Botrytix cinerea, Erwinia carotovora subsp. Carotovora, Aspergillus n ig er and Fusarium avenaceum As with the previous studies, each organism produced a specific series of differentiating volatiles that allowed for its identification (Vikram et al. 2006) The concept of differentiating volatiles was also app lied to studies involving McIntosh, Cortland, and Empire apples for the identification and differentiation of Botrytix cinerea, Penicillium expansum Mucor piriformis and Monilinia spp. Each organism produced a specific set of volatiles that allowed for it to be differentiated (Vikram et al. 2004a; Vikram et al. 2004b) This concept has also been applied outside the food industry for the identification of mold in building supplies. Similar results to the previous studies were seen with the organisms tested (Wady and others 2003) The methodology developed for testing the food products involved the use of a portable GC MS. The intact samples were placed in mason jars and their headspace was sa mpled using the portable GC MS. V olatile metabolites were concentrated on a Carbopack X carbon trap prior to being desorbed and analyzed using the GC MS (Lui e t al. 2005; Moalemiyan et al. 2007; Moalemiyan et al. 2006; Vikram et al. 2006; Vikram et al. 2004a; Vikram et al. 2004b) The identification of mold in building supplies was performed using a SPME fiber to sample the headspace of a vial containing the s pecific

PAGE 30

30 building material. The fiber was then desorbed and analyzed using a GC MS (Wady et al. 2003) Volatile Analysis Overview Traditional application s of volatile evaluation invol ved a trained sniffer who would sniff storage containers for distinctive odors associated with specific diseases. This method was highly subjective, required well trained personnel, and by the time odors were at detectable concentrations, significant damage to the product had already occurred. Despite its faults, this method has been used for years in taxonomic classification of some microbes (Janssens et al. 1992; Moalemiyan et al. 2007) Newer methodology, which allows for improved and more standardized detection, involves the use of various methods to extract, concentrate, and analyze volatiles (Cai and others 2001) Some techniques currently employed to sample volatile metabolites include steam distillation extractions ( SDE), purge and trap, direct headspace sampling with a syringe, and diffusion, which includes HS SPME (Larsen and Frisvad 1994; Larsen and Frisvad 1995c; Schnrer et al. 1999) SDE lacks sensitivity for certain volatiles, is time consuming, requires laborious manipulations, and ut ilizes heat. The heat can lead to the development of artifacts and/or the breakdown of volatiles (Cai et al. 2001; Larsen and Frisvad 1995c; Schnrer et al. 1999) Purge and trap is a good alternative to SDE, but r equires a pump. Direct headspace sampling with a syringe or diffusion with a SPME fiber is considered optimal (Larsen and Frisvad 1994; Larsen and Frisvad 1995c)

PAGE 31

31 Headspace Solid Phase Microextraction HS SPME is a s olvent free extraction technique that uses a small fused silica fiber coated with a polymeric organic liquid. The fiber is introduced into the headspace above a sample. Volatiles in the headspace are adsorbed and concentrated in the coating of the fiber, which is then desorbed into analytical equipment such as a GC MS for analysis (Cai et al. 2001; Nilsson et al. 1996; Zhang and Pawliszyn 1993) The type of fiber used determines the selectivity of extraction, as various fibers have differing affinities for assorted volatile s Studies examining these fibers to determine an optimal fiber for use in the HS SPME method determined that polydimethyls iloxane/divinylbenzene (PDMS/DVB) fibers have the highest sensitivity and lowest detection limits 2003) Temperature was also shown to play an important role in the equilibration of volatile s between a liquid medium and the headspace and the absorption of volatiles to the fiber. The optimum temperature was found to be 40 50C, as this allowed for the most rapid equilibration and absorption times Agitation of the liquid medium also quickens the partitioning of volatiles between the medium and the headspace decreasing the equilibration time (Zhang and Pawliszyn 1993) As compared to SDE, SPME can identify and dif ferentiate more volatile s without using a high temperature extraction Vola tiles collected using HS SPME were comparable to those collected under similar conditions using a purge and trap methodology; however, the purge and trap technique required a pump (Nilsson et al. 1 996; Schnrer et al. 1999) HS SPME has many advantages including simplicity, solvent free nature of the extraction, short extraction time, no apparent sample hydrolysis, possibility of automation, and suitability for use in routine screening (Cai et

PAGE 32

32 Studies have shown the validity and advantage of this rapid, solvent free extraction method and its capabilities to detect volatile metabolites produced by P digitatum (Nilsson et al. 1996; Wady et al. 2003; Zhang and Pawliszyn 1993) Application to Orange Juice Filamentous fungi, specifically P digitatum are known to produce secondary volatile metabolites that can be detected using HS SPME (Filtenborg et al. 1996; Hutchinson 1973) By applying volatile metabolomics and chemosystematics to OJ and fungal contaminants, the detected volatiles can be associated with specific organisms and subsequently used to identify fungi down to the species level more rapidly than conventional microbiological methods (Cevallos Cevallos et al. 2009; Hutchinson 1973; Larsen and Frisvad 1995a; Larsen and Frisvad 1995b) Studies have shown that Penicillia species can be distinguished from each other solely on their differentiating volatiles when grown on various media, such as PDA These volatile met abolites can be detected after two days of fungal growth. This precedes the time required for the identification by morph ological differences on the PDA plates (Larsen and Frisvad 1995b; Schnrer et al. 1999) By detecting the contamination early, appropriate steps can be taken to significantly reduce losses if adequate control meas ures are implemented (Lacey 1989; Vikram et al. 2006)

PAGE 33

33 Figure 21. Fungal metabolic pathways and their resulting volatile metabolites from the initial substrate glucose Glycolysis Pyruvate Acetyl CoA TCA cycle 2 methyl 1 pr opanol Leucine 3 methyl1 butanol Fatty Acids Alkanes Alkenes Secondary Ketones C8 compounds & 1 octen3 ol Primary alcohols Esters and lactones Ethanol Mevalonate Mono & sesquiterpenes Oxaloacetate Aspartate Threonine 2 methyl 1 butanol Methionine Sulfur compounds

PAGE 34

34 CHAPTER 3 MATERIALS AND METHODS This study was undertaken to evaluate the bio chemical changes that occur in single strength, not from concentrate OJ when contaminated with P digitatum Volatile changes in juice stored at room temperature (~22 C) occurring as a result of the submerged fungal colony within the first four day s of inoculation were assessed. The experiment was repeated three times. Penicillium digitatum Preparation P digitatum (Link) is an imperfect fungus commonly associated with decaying citrus The culture for this study was obtained from the USDA/ARS, Citrus and Subtropical Products Research Laboratory, Winter Haven, FL. Spores of P digitatum were plated on PDA plates (Difco, NJ ) and incubated at room temperature until sporulation occurred, which was determined by morphological changes, such as the development of an olive green mat Mature colony plates were stored under refrigeration (5C) in plastic bags to prevent dehydration. Inoculum was prepared by placing a small amount of sterile 0.1% Tween 20 on the sporulating mat and slightly spreading the liquid to release spores. Spores were picked up from the plate surface with an inoculating loop and transferred into 9 mL of sterile 0.1% phosphate buffer and vortexed to keep the cells in suspension. Spore concentration was calculated using a haemocytometer and adjusted to 106/mL. Sample Preparation Simply Orange, pulpfree not from concentrate 100% pure pasteurized OJ (The Coca Cola Company FL) was purch ased from a local Publix grocery store The bottles of juice were frozen at 20C at the Citrus Research and Education Center, Lake Alfred,

PAGE 35

35 FL. When needed, a bottle was thawed in a refrigerator at 5C and 100 mL of thawed juice was poured into sterile 250 mL Erlenmeyer flasks with screw caps The juice was re pasteurized by submerging the flasks 2/3 in a circulating water bath (model#2864, Thermo Scientific, NH), at 90C for 5 minutes. The juice reached an internal temperature of 72C. Flasks were immediately transferred to an ice water bath, removed and then left to equilibrate to ambient temperature (22C). When the juice was cooled, it was inoculated with P digitatum using 0.01 mL of inoculum per pasteurized flask. A juice sample treated as the experimental with no inoculum was used as a control. Samples were placed on an Innova 2100 platform shaker plate (New Brunswick Scientific, NJ) set to 150 RPM and allowed to incubate at ambient room temperature (22C). Sampling Samples from the control and inoculated juices were collected on days one, two, three, and four An aliquot (10 mL) of juice was aseptically transferred to a 40 mL glass vial with a silicone/PTFE septa screw cap that contained a stir bar. On days one and three, 0.25 mL samples from the control and inoculated flasks were transferred to PDA plates to confirm P digitatum Brix and pH Analysis Procedures to determine Brix and pH were adapted from Wyatt and others (Wyatt et al. 1995) On days one, two, three, and four of incubation, prior to collecting the 10 mL samples used for volatile analysis, Brix a nd pH analysis was performed. For Brix 0.25 mL of sample was placed on the lens of a PR 101 Palette digital refractometer (Atago, WA) that was previously calibrated using deion ized (DI) water. The reading was taken. S ubsequently the sample was removed and the lens was cleaned thoroughly

PAGE 36

36 with KimWipes. This process was repeated in triplicate for each sample on each day, and the average and standard error was calculated. A MiniLab, ISFET pH meter (model# IQ125, IQ Scientific, Loveland, CO) was calibrated using a standard buffer solution pH 7.00 (SB107500) and pH 4.00 (SB101500) (Fisher Scientific, NC). After calibrating, a 0.25 mL sample from the control flask was transferred to the probe to obtain the pH reading. The procedure was then repeated with a 0.25 mL sample from the inoculated flask, with adequate cleansing with DI water in between samples. Volatile Analysis Headspace Solid Phase Microextraction The 40 mL glass vial prepared earlier was 1/3 submerged in a 40C water bath containing a stir bar spinning at low speed and allowed to equilibrate for 15 minutes. DVB/Carboxen/ PDMS StableFl ex fiber (Supelco, PA) was inserted into the vial, suspended ove r the headspace and exposed for 5 minutes. The exposure time was lowered to 10 seconds for the analysis of limonene only on the GC MS. The apparatus setup can be seen in Figure 31 At the end of the exposure time, the fiber was fully retracted into the needle, removed from the bottle, and injected into the appropriate GC as seen below. Gas Chromatography Mass Spectrometry A Claurus 500 GC MS system (Perkin Elmer, MA) was used for volatile analysis and identification. The system contained a Stabi lwax polar (wax) column 60 m, 0.25 mm I PA) using helium as the carrier gas at 2 mL/min. Both the GC and MS were controlled using TurboMass software, vers ion 5.4.2 (Perkin Elmer MA).

PAGE 37

37 GC MS general program The general program was developed to examine the overall volatile spectrum while not identifying limonene, as the high concentration of limonene overloaded the detector and overshadowed smaller peaks. The GC temperature program began at 40C for 2 minutes followed by a ramp of 7 C/min up to 240C with a sampling rate of 1.5625 pts/sec for a total run time of 30.57 minutes The injector temperature was set at 220C, while the detector was set at 275C. The MS program consisted of a spectrometer scanning 25 to 300 m/z from 2.00 to 30.10 minutes with a delay at 13.25 to 13.79 minutes GC MS l imonene and SIR p rogram This program was established to examine limonene only, as the SPME exposure time had to be reduced to 10 sec onds The GC temperature program began at 40C for 2 minutes followed by a ramp of 7C/min up to 140C. At 140C, a second ramp of 20C/min began and went up to a final oven temperature of 240C. The sampling rate was 1.5625 pts/sec, and the program had a total run time of 21.29 minutes The injector temperatur e was set at 220C, while the detector was set at 275C. This GC temperature program was run with two different MS programs, each used with separate injections. The first program denoted limonene, consisted of a spectrometer scanning 25 to 300 m/z from 2.00 to 19.40 minutes while the second program denoted SIR, scanned from 3.01 to 20 minutes while only detecting masses at 121 and 136 m/z. GC MS peak i dentification Chromatograms were produced and analyzed using the TurboMass software. Peak area mi nimums were established at 500,000 for detection. Identification of peaks was first performed by examining the spectral analysis followed by confirmation using

PAGE 38

38 web database linear retention index (LRI) values as compared to observed (1%) which was calcu lated by a standard curve generated from a series of low and high alkane standards comprised of compounds with various carbon chain lengths and known retention times Gas Chromatography Sulfur Volatile sulfur metabolites were analyzed using a 7890A GC System (Agilent Technologies, CA) with a pulsed flame photometric detector (PFPD) and a Stabilwax polar (wax) column 30 m, 0.32 mm PA). The instrument data was collected by Chrom Perfect Data software, version 6.0.4 (Justice Laborator y Software, NJ). The oven temperature program began at 40C followed by a ramp of 7C/min up to 240C with a 5 minute end hold for a total run time of 34 minutes. The injector temperature was set at 200 C, and the PFPD detector was set to 250C. The chr omatograms were analyzed using Chrom Perfect LRI values were calculated based on a standard curve generated from a series of low and high alkane standards with various carbon chain lengths and known retention times. These were then compared to LRI val ues from standards with various carbon chain lengths and known retention times run on the GC S with a Stabilwax, polar (wax) column in Dr. Russell Rouseffs lab (%). Gas Chromatography Olfactometry A 6890N Network GC System (Agilent Technolog ies, CA ) in split mode with a flame ionized detector (FID) was used for the detection of aroma impact compounds The column inside the machine was a DB Wax polar 30 m, 0.32 mm column (J&W Scientific, Agilent Technologies, CA). The instrument data was collected by Chrom Perfect Data software, version 6.0.4 (Justic e Laboratory Software, NJ). The

PAGE 39

39 oven temperature program began at 40C followed by a ramp of 7C/min up to 240C with a 5 minute end hold for a total run time of 34 minutes. The injector temperature was set at 220C, and the FID detector was set to 275C. Aroma Impact Compound Identification Sniffing began approximately 2 minutes after the injection to all ow for retained volatiles to purge from the column. The nose was positioned inch from the olfactometer while breathing normally. As an odor was detected, the indicator bar of the Tigre Laboratory Interface (Justice Laboratory Soft ware, NJ) was moved in proportion to the intensity of the odor. The bar was quickly returned back to zero after indicating the intensity recorded as peak height A descriptor of the odor was recorded, along with the time of elution, by the sniffer. This data was recorded as an aromagram by the Chrom Perfect software. The set up of this instrument can be seen in Figure 3 2 The aromagrams were analyzed using Chrom Perfect and the LRI values were calculated based on a standard curve generated from a series of low and high alkane standards with various carbon chain lengths and known retention times and compared to those identified on the GC MS (%). Data Treatment The experiment was performed three times. Percent difference from control for GC MS and GC S volatiles was calculated based on averages. The averages, along with the standard error for the averages, were calculated using Microsoft Excel software F or GC O peak heights were averaged and standard errors calculated also using Microsoft Excel software. Values were analyzed and subjected to principal component analysis using The Unscrammbler X, versions 10. 0.1 (Camo Software, NJ). The principal component analysis was used as a tool to assess differences

PAGE 40

40 between days based on totals from volatiles determined to be differentiating. To determine differentiating volatiles, the specific volatile in question had to be significantly different when comparing control and inoculated samples. From consultation with a statistician, equation 3 1 was used to calculate a highly conservative percent difference (80%), based on the number of control samples (N1 = 3) and inoculated samples (N2 = 3), required to determine significant difference. This calculated value, 80% difference, was too conservative for the study and l owered to 50% difference required to state significant difference between samples on recommendation from the statistician. With the ability to state a specific volatile was significantly different from control and inoculated samples allowed for the volati le to be classified as differentiating. 3 1. Percent difference for significant difference = (1.96 x 1 + 1/N2)) x 100

PAGE 41

41 Figure 31. HS SPME volatile sampling apparatus setup

PAGE 42

42 A B Figure 32. GC O instrument with A) tube aromas elute from and B) Tigre Laboratory Interface aroma intensity indicator bar

PAGE 43

43 CHAPTER 4 PHYSICAL AND CHEMICAL CHANGES Cloud Destruction C loud destruction as a result of P digitatum contamination was noticeable after two days of incubation. In the Erlenmeyer flask c ontaining the inoculated sample, a mycelia puck, ~30 mm diameter and ~10 mm thick, began forming at incubation day two and was fully formed by day three. This mat was absent from the control sample ( Figure 41, A ). The mycelia puck was a bright orange co lor, similar to the color of the control juice. When the juice was extracted for sampling, the control sam ple was an opaque orange while the inoculated sample was clear with a yellow hue (Figure 41, B). By day three of incubation, the juice achieved max imum cloud destruction and there were no visible differences in the inoculated juice between further incubation days. This physical change was observed in previous studies performed by Alaa and others (1990), Barmore and Brown (1979), Bush and Codner (1968), and Wyatt and others (1995). V arious pectin degrading enzymes, including pectin esterases, cause this change by split ting 1,4 glycosidic bonds between galacturonic acid and carbohydrates (Alaa et al. 1990; Barmore and Brown 1979; Bush and Codner 1968; Wyatt et al. 1995) The rapid cloud destruction supports the idea that multiple enzymes were present in abundance and worked in conjunction (Alaa et al. 1990; Bush and Codner 1968; Doyle and Beuchat 2007; Filtenborg et al. 1996) In addition, the juice pH was initially 3.7, close to the optimum pH of 4.5 for some enzymes including polygalacturonase and exopolygalacturonase which assisted with the rapid cloud destruction (Barmore and Brown 1979; Patrick and Hill 1959)

PAGE 44

44 Change in pH The pH of the control and inoculated juice was monitored over the four day incubation period. The pH of juice was ~3.7. It can be seen in Tabl e 41 that the percent difference between samples was at most 2.6 1.52% on incubation day three and decreased to a percent differences of 1.8 0.90 on day 4. Other incubation days percent differences were less than 2.6 1.52%. The relative stability in the juice pH was noted in literature and is explained by the buffering capabilities of the organic acids such as citric acid (Filtenborg et al. 1996; Fries 1973; Pelser and Eckert 1977; Wyatt et al. 1995) pH stability helps to maintain an optimum environment for P digitatum and its enzymes while inhibiting competition from others (Ariza et al. 2002; Lacey 1989) Change in Brix N oted in previous studies by Wyatt and others (1995) and seen in this study was a reduction in Brix within a few days of inoculation. Wyatt found that Brix began decreasing after 12 days at 10C (Wyatt et al. 1995) This study, however, found an 11.05 1.52% reduction in Brix began after day three at 22C (Table 41) The higher incubation temperature (22C as compared with 10C) of this study can explain the earlier decrease in sugar content, as the organism was more active at the elevated incubation temperature. One would expect the sugar content to decrease starting at day one of incubation, as sugars are one of the primary energy sources entering the metabolic pathway through glycolysis seen in Figure 21 (Filtenborg et al. 1996; Kavan agh and Wood 1971; Pelser and Eckert 1977) A relationship exist s between sugar content and breakdown of pectin. The enzymes causing cloud destruction are dissolving the cellulose, a component of cell walls, into simpler components including sugars (Barmore and Brown

PAGE 45

45 1979) This production and supply of sugar could be relatively equal to the amount being used by the organism for the first three days of incubation. Starting at the fou rth day, the supply of sugar from the breakdown of cellulose becomes depleted, so the sugars used by the organism are not being replenished, explaining the observed decrease. Visual observation of cloud destruction reaches its optimum between day three and four after which no increase in cloud destruction is noted. This timeline of cloud destruction coincides with the beginning of a decrease in Brix supporting the correlation between observed cloud destruction and decrease in Brix

PAGE 46

46 A B Figure 41. Control and inoculated sample, respectively at incubation day four, A) taken while in Erlenmeyer flasks and B) after 10 mL samples were extracted.

PAGE 47

47 Table 41. pH and Brix percent difference from control with standard error for the four day incubation period Percent difference from control standard error Incubation period p H Brix Day 0 ( ) 0.90 0.90 ( ) 0.38 0.19 Day 1 0.88 0.88 ( ) 0.18 0.25 Day 2 ( ) 1.76 0.88 ( ) 0.37 0.34 Day 3 ( ) 2.63 1.52 ( ) 2.61 1.31 Day 4 ( ) 1.80 0.90 ( )11.05 1.52

PAGE 48

48 CHAPTER 5 CHANGES IN ORANGE JU ICE VOLATILES DUE TO PENICILLIUM DIGITATUM Gas Chromatography Mass Spectrometry Volatiles The HS SPME analysis of the samples analyzed using a GC MS with a polar (wax) column identified 102 volatiles ( Table A 1). Of those 102, 44 volatiles differentiated from the control at least ~50% by incubation day four. The ~50% difference from control was established to identify the compound as a differentiating volatile (Table 51 ). The exception to this rule was 4terpineol as its maximum percent difference was 32.9 4.7% on day four. It was retained as a differentiating volatile for further analysis of the bioconversion of limonene. These differentiating volatiles are either only present in the control or inoculated sample (qualitative difference) or are pres ent in both, but differ in quantity, either greater than or less than that in the control by at least ~50% (quantitative difference). These qualitative and quantitative differences between the inoculated sample and the control sample can be used to distin guish the sample contaminated with P digitatum Chromatographic Differences Comparing chromatograms of the control sample to those from inoculated samples for each day allowed for an initial analysis of both qualitative and quantitative differences. The chromatogram for day one, with the inoculated samples chromatogram above the control s chromatogram, can be seen in Figure 51. Days two and three are a progression of increasing and decreasing peaks culminating at day 4. The chromatogram for day four c an be seen in Figure 52 On first observation, it is easy to see that some peak s are either greater or lesser in the inoculated as compared to the control sample. The volatile that eluted at a retention time ( RT) of 7.06 minutes

PAGE 49

49 with an LRI of 947, identified a s ethanol, increased over the four day incubation time. Limonene, at a RT of approximately 13.67 minutes, LRI 1230, decreased in percent difference from ( ) 7.3 4.8 at incubation day 1 to ( ) 99.0 0.3 at incubation day four Two other obvious peaks include those at RT 16.16 minutes (LRI 1354) identified as hexanol, and RT 20.15 minutes (LRI 1569) identified as octanol, which both increase d initially, but decrease d on days three and four The visual analysis is an extremely rudimentary method, however, and more quantitative evaluation is required. Volatiles of Greatest Difference By examining the percent difference of inoculated volatiles, as compared to the control, volatiles of greatest percent difference for each incubation day can be determined. Figures 5 3, 54, 5 5, and 56 display the top 10 volatiles with the greatest percent increase from control and the top 10 volatiles with the greatest percent decrease from control. For incubation day one, the top two volatiles with the greatest percent increase are terpineol. For day two, however, the top two terpineol and terpinolene. These change again for day three, hexanol and terpinolene, and day four hexanol and ethanol. The top two volatile s of greatest decrease for day one are neryl acetate and hexa nal. For incubation day two, it is neryl acetate and benzaldehyde, for day three, neryl acetate and citronellyl, and day four selinene. The percent difference changes between each incubation day for each quantitative volat ile, except when the volatile decreased to undetectable limits, making it a qualitative volatile. Once below detectable limits, the volatile would only be detected in the control. Since each day is unique, it may be possible to not only identify P digit atum from differentiating volatiles but also determine how long the organism has been present based on the percent difference of various volatiles. For application of the

PAGE 50

50 concept, further studies and validation would be required under different incubation conditions. Volatile Changes During Incubation Analysis of the FID chromatograms initially identified that there were at least 102 volatile s. Of those 102 volatiles, 44 were determined to be differentiating volatiles as discussed previously The differ entiating volatiles were categorized based on whether they were differentiating from the control qualitatively or quantitatively. Volatiles with quantitative differences were then classed as whether their percent difference from control was increasing, decreasing, or intermediate. Intermediate volatiles initially increased in percent difference then decreased during days three and four C hanges from the control represent the differentiating volatiles used to detect contamination and identify P digitatum Figure 57 is a representation of the overall trend in volatiles throughout the four days of incubation. An initial increase in the amount of volatiles in the inoculated sample is noted. This increase in volatiles is possibly due to the organisms metabolic pathway (Figure 21) using sugar as an energy source, resulting in the production of volatiles. The decrease in volatiles seen in days three and four is representative of the bioconversion or degradation of volatiles due to enzymes into nonvolatile compounds or compounds with a lower affinity to the SPME fiber used, making them less detectable. Qualitative differences: Volatiles present in only inoculated or control samples Seven volatile s, consisting of alcohols and est ers, were found to be qualitative differentiating volatiles, either only in inoculated or control juices (Table 52 ). The majority of these volatile s, including methanol, ethyl propanoate, ethyl 2-

PAGE 51

5 1 methylbutanoate, 3methylbutanol, (Z) 3 hexenol, and 1octen3 ol, were only present in the inoculated sample. Neryl acetate, an ester, was the only qualitative volatile present in the control and not in inoculated juices The qualitative differences between samples represent the most promising potential target volatiles. Target volatiles are volatiles that can be used to easily identify the presence of contamination since they are either present or absent in the juice. Their presence or absence would signify contamination. For example, i f the juice was sampled and either of the volatiles only seen in the inoculated sample was detected, or in the case of neryl acetate, not detected, it would be an important indicator that P digitatum contamination has occurred. Quantitative differences: Increasing volatiles Of the 44 differentiating volat iles, a mixture of seven alcohols and esters, were found to be increasing in percent difference as compared to the control over the four day incubation period (Table 53 ). This was expected because alcohols followed by esters, are volatiles consistently associated with P digitatum discussed previously (Ariza et al. 2002; Larsen and Frisvad 1995a; Nilsson et al. 1996; Pesis and Marinansky 1990; Schnrer et al. 1999) Volatile s that increased as a result of inoc ulation can be used to detect contamination and determine the presence of P digitatum based on the increase in percent difference when there is significant difference (>50% difference). Unlike qualitative differences which are either present or absent, quantitative differences can only be determined as such when compared to an uninoculated control sample or baseline concentration of the differentiating volatiles. This comparison is required to determine the percent difference in order to classify the vo latile as increasing in amount.

PAGE 52

52 For example, ethyl acetate is considered a quantitative indicator for the presence of P digitatum (Ariza et al. 2002; Larsen and Frisvad 1995a; Schnrer et al. 1999) This study f ound that ethyl acetate does increase by 361.4 173.6% by incubation day four as the result of P digitatum contamination. The determination that ethyl acetate did increase in amount, however, was only accomplished by comparing the inoculated sample to t he control sample and calculating percent difference Without the comparison, the percent difference and overall change in amount would not have been able to be determined. Quantitative differences: Decreasing volatiles The majority of differentiating v olatiles, as compared to the control, decrease d during the four day incubation period. This helps explain the overall decrease in volatile peak area seen in table 57. Table 54 contains the 25 volatile s, consisting of aldehydes, alcohols, esters, monoterpenes, cyclic terpenes, bicyclic terpenes, sesquiterpenes, bicyclic sesquiterpenes, tricyclic sesquiterpenes, a terpinoid, and a furanoid, that decreased in percent difference compared to the control Some volatile s, such as ethyl hexanoate, decreased to undetectable levels four days after inoculation. As with the volatile s which increased as a result of inoculation, those that decreased can be used to detect contamination and determine the presence of P digitatum A value to compare the differentiati ng volatiles to is required in order to determine that a percent decrease occurred. With some compounds, like ethyl hexanoate, this comparison is not needed since the compound decreases to undetectable limits by day four. This means the quantitative ethy l hexanoate is now a qualitative volatile and its absence signifies contamination. In addition to signifying contamination, since the compound is present in the inoculated sample until day four,

PAGE 53

53 the absence of the compound signifies the organism has been present in the sample for at least four days. Quantitative differences: Intermediate volatiles Consisting of esters, alcohols, a cyclic terpene, and a bicyclic terpene, 12 intermediate volatiles were identified (Table 55 ). These volatile s, like those f ound to be increasing or decreasing, can only be determined as an intermediate when compared to a control or standard baseline in order to calculate percent difference. As seen in Table 5 5 each volatile has a specific day it reaches its maximum percent difference. Based on this information, these volatile s and their maximum percent difference could be used as indicators of how long the organism has been present. For example, if a sample was tested and hexyl acetate was determined to be at its maximum percent difference it could be assumed that the organism has been present for two days, based on this study. Application of this concept would require not only a baseline concentration for comparison, as is required to determine an increasing or decreasing volatile but it would also require specific studies performed under the storage and quality conditions of the applicable OJ. Alterations in storage conditions, such as fluctuations in temperature or pH, would alter the day a maximum percent difference is observed since substrate composition and environmental factors influence the production of metabolites (Larsen and Frisvad 1995a) Gas Chromatography Sulfur Volatiles The purpose of running samples on the GC S with a PFPD detector was to examine sulfur volatile s specifically, as they normally do not appear with analysis using a GC MS with FID detector (Acree and Arn 2004; Reineccius 2006) Since the GC S was not attached to a MS and multiple columns per sample could not be used,

PAGE 54

54 tentative identification was based on the com parison of observed LRI values with standards run on the GC S with a Stabilwax, polar (wax) column in the lab. The analysis using the GC S observed 14 volatiles (Table A 2) with 12 of those being differentiating vo latiles, four of which c ould be tentatively identified (Table 56 ). The differentiating volatiles were determined based on the same requirements as the GC MS volatiles, a percent difference from control of at least ~50%. Some of these differences in su lfur volatiles were noticeable upon initial analysis of the control and inoculated chromatogram. Figure 58 is the chromatogram from incubation day four with the inoculated sample chromatogram above the control chromatogram. It is clear that the compound at RT of 6.82 minutes (LRI 1093), tentatively identified as dimethyl disulfide, has decreased in the inoculated sample from control. As with the GC MS, the overall trend for differentiating sulfur volatiles in the inoculated sample was a decrease in amount, seen in Figure 59, as compared to the control. This decrease did not begin until incubation day four. Until day four, the amount of sulfur volatiles in the inoculated sample was similar to that of the control. When examining the differentiating sulfur volatiles in more detail, they were observed to differentiate from the control sample qualitatively and quantitatively For example, Sulfur volatile LRI 1484 was a qualitative volatile being onl y present in the control sample. Sulfur volatile LRI 738, tentatively identified as dimethyl sulfide, is observed to decrease in the inoculated sample as compared to the control classifying it as a quantitative difference. These differentiating volatiles, both qualitative and quantitative were further exam ined. Volatiles of greatest difference for each incubation day can be seen in

PAGE 55

55 figures 510, 511, 512, and 513. Sulfur volatile LRI 1789 was the volatile of greatest increas e at day one; however, this compound begins to decrease on day two and by day t hree and four this sulfur volatile wa s of greatest decrease. Dimethyl disulfide was of greatest i ncrease during incubation days two and three, identifying it as an intermediate volatile. The applications of these findings are similar to the GC MS. Ea ch day the inoculated sample contained a series of differentiating volatiles, both qualitative and quantitative, which can be used to identify P digitatum contamination along with how long the organism has been present. These differentiating volatile s u tilize the same application as those identified with the GC MS. The differentiating sulfur quantitative volatiles require a baseline for comparison in order to determine any change. Further studies are needed to accurately identify these volatile s. The addition of this knowledge further supports the potential of the concept that the differentiating volatiles can be used as indicators of P digitatum contamination.

PAGE 56

56 Table 51. Differentiating volatiles Compound Observed LRI Literature LRI* Day 1 percent difference from control (% standard error) Day 4 percent difference from control (% standard error) Ethyl acetate 900 890 109.1 73.1 361.4 173.6 Methanol 909 899 0.0 0.0 100.0 0.0 Ethanol 947 947 19.6 11.2 965.5 29.2 Ethyl propanoate 974 951 0.0 0.0 100.0 0.0 Methyl butanoate 1002 990 ( ) 34.4 0.0 585.5 441.7 Ethyl butanoate 1050 1048 13.4 13.3 ( ) 47.1 29.7 Ethyl 2 methylbutanoate 1065 1062 0.0 0.0 100.0 0.0 Hex anal 1098 1099 ( ) 61.8 11.4 ( ) 95.8 2.1 3 carene 1161 1148 ( ) 8.3 4.0 ( ) 62.6 5.7 Myrcene 1171 1172 69.9 62.5 ( ) 97.2 1.7 Methylbutanol, 3 1212 1212 0.0 0.0 100.0 0.0 Limonene 1230 1205 ( ) 7.3 4.8 ( ) 99.0 0.3 phellandrene 1238 1241 3.4 2.0 ( ) 91.0 3.5 Ethyl hexanoate 1245 1246 2.8 1.0 ( ) 100.0 0.0 terpinene 1267 1260 ( ) 5.4 1.9 ( ) 100.0 0.0 Hexyl acetate 1286 1267 ( ) 4.2 3.6 ( ) 53.3 26.3 terpinolene 1300 1298 ( ) 6.6 3.5 353.6 101.2 Octanal 1310 1309 ( ) 38.8 9.0 ( ) 95.9 1.2 Hexanol 1359 1364 357.2 125.0 3078.2 829.8 (E) 2 hexenol 1383 1388 ( ) 44.1 28.0 ( ) 100.0 0.0 (Z ) 3 hexenol 1398 1399 0.0 0.0 100.0 0.0 Nonanal 1417 1416 ( ) 14.9 3.9 ( ) 90.1 1.7 *Database LRIs (Rouseff 2006)

PAGE 57

57 Table 51. Continued Volatile Observed LRI Database LRI* Day 1 percent difference from control (% standard error) Day 4 percent difference from control (% standard error) Perillene 1442 1409 15.7 1.9 ( ) 10 0.0 0.0 Ethyl octanoate 1451 1450 11.4 9.2 ( ) 86.7 1.4 Octen 3 ol, 1 1465 1452 0.0 0.0 100.0 0.0 Octyl acetate 1492 1480 10.2 1.7 ( ) 92.1 2.3 Decanal 1526 1523 ( ) 6.6 3.9 ( ) 85.1 3.7 copaen e 1531 1536 13.6 2.3 ( ) 100.0 0.0 Linalool 1558 1557 8.7 6.9 ( ) 46.9 10.3 Octanol 1569 1565 177.8 50.7 57.1 51.6 Benzaldehyde 1575 1555 ( ) 10.4 7.1 ( ) 100.0 0.0 elemene 1626 1595 13.5 1.4 ( ) 100.0 0.0 Terpineol, 4 1636 1616 ( ) 4.6 3.3 20.0 2.0 caryophyllene 1646 1641 15.6 1.3 ( ) 85.1 4.8 Citronellyl acetate 1680 1672 28.7 5.6 ( ) 100.0 0.0 selinene 1690 1711 11.8 5.5 ( ) 89.0 3.0 terpineol 1730 1724 264.6 112.5 178.8 108.1 Neryl acetate 1746 1742 ( ) 100.0 0.0 ( ) 100.0 0.0 Valencene 1770 1763 11.8 5.3 ( ) 98.4 0.5 selinene 1779 1724 7.2 1.5 ( ) 100.0 0.0 Carvone 1788 1779 ( ) 5.4 4.4 ( ) 100.0 0.0 cadinene 1798 1794 11.8 0.1 ( ) 74.4 7.2 panasinsen 1819 1840 10.1 0.3 ( ) 80.5 5.9 Cis carveol 1864 1855 10.6 0.5 ( ) 61.4 5.5 *Database LRIs (Rouseff 2006)

PAGE 58

58 Figure 51. MS chromatograms of control (lower) and inoculated (upper) OJ at incubation day one Time (min) LRI 947 LRI 1050 LRI 1171 LRI 1230 LRI 1300 LRI 1531 LRI 1569 LRI 1558

PAGE 59

59 Figure 52. MS chromatograms of control (lower) and inoculated (upper) OJ at incubation day four Time (min) LRI 947 LRI 1050 LRI 1171 LRI 1230 LRI 1300 LRI 1531 LRI 1569 LRI 1558 LRI 1354

PAGE 60

60 Figure 53. Top 20 differentiating volatiles of greatest difference from control at incubation day one

PAGE 61

61 Figure 54. Top 20 differentiating volatiles of greatest difference from control at incubation day two

PAGE 62

62 Figure 55. Top 20 differentiating volatiles of greatest difference from control at incubation day three

PAGE 63

63 Figure 56. Top 20 differentiating v olatiles of greatest difference from control at incubation day four

PAGE 64

64 Figure 57. Overall trend in volatiles

PAGE 65

65 Table 52. Qualitative differentiating volatiles Volatile Observed LRI Database LRI* Day one percent difference from control (% standard error) Day four percent difference from control (% standard error) Methanol 909 899 0.0 0.0 100.0 0.0 Ethyl propanoate 974 951 0.0 0.0 100.0 0.0 Ethyl 2 methylbutanoate 1065 1062 0.0 0.0 100.0 0.0 Methylbutanol, 3 121 2 1212 0.0 0.0 100.0 0.0 (Z) 3 hexenol 1398 1399 0.0 0.0 100.0 0.0 Octen 3 ol, 1 1465 1452 0.0 0.0 100.0 0.0 Neryl acetate 1746 1742 ( ) 100.0 0.0 ( ) 100.0 0.0 *Database LRIs (Rouseff 2006) Table 53. Quantitative differentiating volatiles that increased in percent difference from control Compound Observed LRI Database LRI* Day one percent di fference from control (% standard error) Day four percent difference from control (% standard error) Ethyl acetate 900 890 109.1 73.1 361.4 173.6 Methanol 909 899 0.0 0.0 100.0 0.0 Ethanol 947 947 19.6 11.2 965.5 29 .2 Ethyl propanoate 974 951 0.0 0.0 100.0 0.0 Methyl butanoate 1002 990 ( ) 34.4 0.0 585.5 441.7 Methylbutanol, 3 1212 1212 0.0 0.0 100.0 0.0 *Database LRIs (Rouseff 2006)

PAGE 66

66 Table 54. Quantitative differentiating volatiles that decreased in percent difference from control Volatile Observed LRI Database LRI* Day one percent differ ence from control (% standard error) Day four percent difference from control (% standard error) Hexanal 1098 1099 ( ) 61.8 11.4 ( ) 95.8 2.1 Myrcene 1171 1172 69.9 62.5 ( ) 97.2 1.7 Limonene 1230 1205 ( ) 7.3 4.8 ( ) 99.0 0.3 phellandrene 1238 1241 3.4 2.0 ( ) 91.0 3.5 Ethyl hexanoate 1245 1246 2.8 1.0 ( ) 100.0 0.0 terpinene 1267 1260 ( ) 5.4 1.9 ( ) 100.0 0.0 Octanal 1310 1309 ( ) 38.8 9.0 ( ) 95.9 1.2 (E) 2 hexenol 1383 1388 ( ) 44.1 28.0 ( ) 100.0 0.0 Nonanal 1417 1416 ( ) 14.9 3.9 ( ) 90.1 1.7 Perillene 1442 1409 15.7 1.9 ( ) 100.0 0.0 Ethyl octanoate 1451 1450 11.4 9.2 ( ) 86.7 1.4 Octyl acetate 1492 1480 10.2 1.7 ( ) 92.1 2.3 D ecanal 1526 1523 ( ) 6.6 3.9 ( ) 85.1 3.7 copaene 1531 1536 13.6 2.3 ( ) 100.0 0.0 Linalool 1558 1557 8.7 6.9 ( ) 46.9 10.3 Benzaldehyde 1575 1555 ( ) 10.4 7.1 ( ) 100.0 0.0 elemene 1626 1595 13.5 1.4 ( ) 100.0 0.0 caryophyllene 1646 1641 15.6 1.3 ( ) 85.1 4.8 selinene 1690 1711 11.8 5.5 ( ) 89.0 3.0 Valencene 1770 1763 11.8 5.3 ( ) 98.4 0.5 selinene 1779 1724 7.2 1.5 ( ) 100.0 0.0 Carvone 1788 1779 ( ) 5.4 4.4 ( ) 100.0 0.0 cadinene 1798 1794 11. 8 0.1 ( ) 74.4 7.2 panasinsen 1819 1840 10.1 0.3 ( ) 80.5 5.9 Cis carveol 1864 1855 10.6 0.5 ( ) 61.4 5.5 *Database LRIs (Rouseff 2006)

PAGE 67

67 Table 55. Quantitative intermediate volatiles Volatile Observed LRI Database LRI* Day of maximum percent difference Day one percent difference from control (% standard error) Day four percent difference from control (% standard error) Ethyl butanoate 1050 1048 3 13.4 13.3 200.6 32.3 Ethyl 2 methylbutanoate 1065 1062 3 0.0 0.0 100.0 0.0 3 carene 1161 1148 2 ( ) 8.3 4.0 114.3 60.4 Hexyl acetate 1286 1267 2 ( ) 4.2 3.6 70.2 8.4 terpinolene 1300 1298 2 ( ) 6.6 3.5 1779.0 94.4 Hexanol 1359 1364 3 357.2 125.0 4445.1 1214.8 (Z) 3 hexenol 1398 1399 3 0.0 0.0 100.0 0.0 Octen 3 ol, 1 1465 1452 2 0.0 0.0 100.0 0.0 Octanol 1569 1565 2 177.8 50.7 596.4 101.8 Terpineol, 4 1636 1616 3 ( ) 4.6 3.3 32.9 4.7 Citronellyl acetate 1680 1672 2 28.7 5.6 172.0 43.3 terpineol 1730 1724 2 264.6 112.5 11811.1 2109.6 *Database LRIs (Rouseff 2006)

PAGE 68

68 Table 56. Differentiating sulfur volatiles Sulfur volatile Observed LRI Database LRI* Difference from control Day one percent difference from control (% standard error) Day four percent difference from control (% standard error) Dimethyl sulfide 738 736 dec rease ( ) 6.4 13.7 ( ) 67.1 3.0 Unidentified 942 only in inoculated 0.0 0.0 100.0 0.0 Dimethyl disulfide 1093 1064 intermediate 8.8 27.7 54.0 61.1 Unidentified 1413 decrease 12.6 0.6 ( ) 100.0 0.0 Unidentified 1484 o nly in control ( ) 100.0 0.0 ( ) 100.0 0.0 Unidentified 1557 only in inoculated, increase 0.0 0.0 ( ) 100.0 0.0 Unidentified 1563 only in inoculated, intermediate 0.0 0.0 ( ) 100.0 0.0 Dithiane, 1,4 1597 1589 decreas e ( ) 23.0 13.4 ( ) 100.0 0.0 Unidentified 1670 decrease 5.5 8.0 ( ) 100.0 0.0 Unidentified 1743 only in inoculated, increase 0.0 0.0 ( ) 100.0 0.0 Phenethyl mercaptan 1779 1753 decrease 4.3 1.7 ( ) 100.0 0.0 Unidentified 1789 decrease 15.2 9.3 ( ) 100.0 0.0 LRI values from standards run on the GC S with a Stabilwax column in Dr. Russell Rouseffs lab

PAGE 69

69 Figure 58. Sulfur chromatograms of control (upper) and inoculated (lower) OJ at incubati on day four Intensity (millivolts)

PAGE 70

70 Figure 59. Overall trend in sulfur volatiles

PAGE 71

71 Figure 510. Major differentiating sulfur volatile differences from control at incubation day one

PAGE 72

72 Figure 511. Major differentiating sulfur volatile differences from control at incubat ion day two

PAGE 73

73 Figure 512. Major differentiating sulfur volatile differences from control at incubation day three

PAGE 74

74 Figure 513. Major differentiating sulfur volatile differences from control at incubation day four

PAGE 75

75 CHAPTER 6 MECHANISMS OF VOLATI LE FOR MATION The production and breakdown of volatiles mentioned above results from the biosynthesis of metabolites via fungal metabolic pathways and enzymatically driven reactions. OJ is a very complex system and it cannot be assumed that the reactions taking place are independent of one another. The reactions are most likely working in conjunction, as the resulting product of one can be the substrate or reactant in another. In addi tion, multiple reactions produce or use similar volatiles These reactions ar e spec ific to OJ contaminated with P digitatum The mechanisms are not well understood, but involve oxidation, reduction, hydrolytic, dehydration, degradation and C C bond forming reactions (Demyttenaere et al. 2001; Janssens et al. 1992) Volatile Formation as a Result of the Metabolic Pathway Figure 21 an overview of fungal metabolic pathways and the resulting volatile metabolites, has been adapted from Schnner et al (1999) This figure helps explain the f ormation of esters and alcohols seen in this study. The increase in ethyl acetate in the inoculated sample is the result of the metabolic pathway involving the conversion of glucose to pyruvate to acetic acid via acetyl CoA. Acetic acid then reacts with ethanol through the process of esteri fication to form ethyl acetate (Figure 61 ) (Pesis and Marinansky 1990; Schnrer et al. 1999) One o f the earliest pathways seen in Figure 21 is the breakdown of glucose by gl ycolysis to produce pyruvate. Once formed, pyruvate los es carbon dioxide to form acet aldehyde, which is then r educed by NADH to form ethanol (Figure 62 ). The association between the decrease in sugar content, Brix, and increase in ethanol due to fungal contamination can be seen in Figure 63 In the control samples, the Brix and ethanol remain consistent. I n the inoculated sample,

PAGE 76

76 however, the ethanol begins to increase as the Brix begins to decrease. Fungal metabolic pathways are not the sole means of producing alcohols in this system. Volatile Formation as a Result of Enzymes Methanol, a qualitative differentiating volatile only found in the inoculated samples forms through a fungal metabolic pathway, but in this system also forms as a result of an enzymatic reaction. P digitatum produces a wide range of enzymes that are active at the low pH associated with OJ (Alaa et al. 1990; Bush and Codner 1968; Bush and Codner 1970; Filtenborg et al. 1996) One of these enzymes, pectin esterase, is associated with softening of orange peel and cloud destruction. Through a deesterification reaction, this enzyme breaks down pectin into pectate and methanol (Bush and Codner 1968) This enzyme is produced by P digitatum thus is only present in contaminated OJ, explaining why methanol is a qualitative volatile. Other esterases contribute to the decrease associated with some ester volatile s present and an increase in alcohols. Ethyl hexanoate, found consistently in the control sa mple, ye t decreasing in the inoculated (Figure 64 ), is an example of such. Esterase enzymes produced by P digitatum catalyze the deesterification of ethyl hexanoate into ethanol and hexanoic acid (Figure 65 ). This reaction is another source of ethanol, secondary to that produced by the metabolic pathway mentioned previously. The breakdown of compounds from enzymatic reactions can help explain the overall decrease in differentiating volatiles, and sulfur volatiles, seen in Figures 5 7 and 59. Vola tile Formation as a Result of Bioconversions Enzymes produced by P digitatum also catalyze the bioconversion of volatiles terpineol, and as expected, this bioconversion was observed in this study. Due to the reactions low

PAGE 77

77 specificity, other volatile terpinolene and 4terpineol, were also produced (Figure 66) (Adams et al. 2003; Demyttenaere et al. 2001; Duetz et al. 2003; Tan and D ay 1998a; Tan and Day 1998b; Tan and others 1998) Figure 67 graphically terpinolene, 4terpineol, terpineol as compared to the control samples. As observed by Tan and others, the higheterpineol occurred on day 2, along with the highest amount of terpinolene (Tan and Day 1998a; Tan and Day 1998b; Tan et al. 1998) The highest amount of 4terpineol was not observed until day 3 of incubation, suggesting the terpinolene to 4terpineol. As discussed in Chapter 2, the mechanism of this bioconversion is debated, with the current proposed mechanism involving P 450 monooxygenases causing epoxidation of the 89 double bond of limonene followed by reductive cleavage of the epoxide (Demyttenaere et al. 2001; Duetz et al. 2003) All of the bioconversions and metabolic pathways are under debate, and the interactions between the different pathways and bioconversion only increase the possibilities.

PAGE 78

78 Figure 61. Esterification of acetic acid and ethanol to form ethyl acetate Figure 62. Formation of ethanol from glucose acetic acid ethanol ethyl acetat e (esterification) + glucose pyruvate acetaldehyde ethanol NADH NAD + CO 2

PAGE 79

79 Figure 63. Relationship between the decrease in Brix and the increase in ethanol

PAGE 80

80 Figure 64. Relationship between the decrease in ethyl hexanoate and the increase in ethanol

PAGE 81

81 Figure 65. De esterific ation of ethyl hexanoate to form hexanoic acid and ethanol Figure 6terpineol, terpinolene, and 4terpineol + ethyl hexanoate hexanoic acid ethanol limonene terpineol terpinolene 4 terpineol

PAGE 82

82 Figure 6terpineol, terpinolene, and 4terpineol

PAGE 83

83 CHAPTER 7 CHANGES IN ORANGE JU ICE AROMA IMPACT VOL ATILES DUE TO PENICILLIUM DIGITATUM Determination of aroma impact compounds was accomplished by the methodology sta ted in CHAPTER 3, performed by four different sniffers. In order to be considered an aroma impact compound, three out of the four sniffers had to identify it as such. Based on this requirement, 20 volatiles were found to impact aroma (Table B 1 ). These 20 volatile s were t hen compared to the GC MS data (Table 71 ) to determine only nine of th e 44 differentiating volatiles were aroma impact compounds The association between the GC MS and GC O data was aided by identifying the aroma impact compounds based on their calculated LRI values and observed aroma descriptors as compared to literature. These tentative identifications were then compared to the identification made using the GC MS. If both the GC MS and GC O had volatile s with similar LRIs (1%) and the same identification, it was concluded that the volatile observed to impact aroma on the GC O wa s the same differentiating volatile observ ed on the GC MS. Based on the observed aroma descriptors seen in Table 71, coelution probably occurred. For example at LRI 1246, ethyl hexanoate, a fruity aroma, was observed, but this volatile is not associated with the green, grassy odor also observ ed. These other odors are the result of another unidentified volatile coeluting. Trends in aroma impact compounds followed that of the GC MS volatiles when comparing the aroma intensity (peak height) difference from control (Table 72). The overall t rend was a decrease in perceived aroma intensity not only when comparing the inoculated to the control, but also within the inoculated sample over the four day incubation period (Figure 71). This decreasing trend in perceived intensity can also be

PAGE 84

84 observ ed when comparing the aromagram from day one (Figure 72 ) to the aromagram from day four (Figure 73 ). The aromas observed during day one in the control closely mirror the aromas observed in the inoculated sample. For day four, however, only two aromas w ere observed in the inoculated sample while seven were observed in the control. When looking at the difference between control and inoculated peak height over the four d ay incubation period (Figure 74, 75, 7 6, and 77 ), the decrease is also quite evide nt. The only volatile observed to increase after day three terpinolene. All other volatiles decrease in aroma intensity When the sniffers were asked to smell each sample and describe the overall odor profile, the control maintained its orange aroma throughout the 4 days. The inoculated, however, was descr ibed as decreasing in orange aroma while increasing in alcoholic aroma. The only alcohol identified as an aroma impact compound was linalool, but the aroma associated with this volatile is floral, green, and citrus (Rouseff 2006) The sniffers on the GC O did not identify alcoholic aromas, and subsequently alcohol volatile s, possibly because the concentrations of these vo latiles were below the aroma thresholds. For example, methanol, a volatile that would contribute an alcoholic odor, has a high aroma threshold at 100 ppm (Acree and Arn 2004) The lack of adequate sniffer training for the specific samples also contributed to their inability to detect aroma impact compounds eluting from the GC O (Reineccius 2006) Despite only 9 volatiles being i dentified as aroma impact compounds using this methodology, other volatile s are probably contributing to the change in aroma of the inoculated juice. Based on the relationship of aroma and flavor, it can be assumed that the inoculated juice would have a different flavor profile then the control (Reineccius

PAGE 85

85 2006) These off flavors are of primary concern when dealing with microbial spoilage because they render the juice unacceptable to the consumer and a loss to the manufacturer. The sooner the organism can be identified, the quicker control measures can be taken to limit the spoilage of the juice.

PAGE 86

86 Table 71. Differentiating volatiles which impact aroma with aroma descriptors Volatile GCO LRI GCMS LRI Database LRI* Observed aroma descriptor Literature aroma descriptor* Hexanal 1074 1098 1099 Green, grassy, fresh Fatty, green, grassy, powerfu l Myrcene 1151 1171 1172 Musty, moldy, metallic, alcoholic Musty, wet soil Unidentified 1191 Leafy, pungent, spicy, floral Ethyl hexanoate 1246 1245 1246 Fruity, fresh, green, grassy Fruity terpinolene 1296 1300 1298 Green, citrus, fruity Citru s, pine Octen 3 one, 1 1310 1315 Mushroom Mushroom like, metallic, musty Octyl acetate 1497 1492 1480 Minty, fruity, pine like, bamboo Fruity, slightly fatty Linalool 1545 1558 1557 Floral, flowery, sweet Floral, green, citrus Neryl acetate 1728 174 6 1742 Woody, rosy Fruity, floral *Database LRIs and aroma descriptors (Rouseff 2006) Table 72. Differentiating volati les which impact aroma with percent difference Volatile GCO LRI Aroma difference from control Day one percent difference from control (% standard error) Day four percent difference from control (% standard error) Hexanal 1074 Only i n control ( ) 100.0 0.0 ( ) 100.0 0.0 Myrcene 1151 Decrease 32.3 22.6 ( ) 100.0 0.0 Unidentified 1191 Only in control ( ) 100.0 0.0 ( ) 100.0 0.0 Ethyl hexanoate 1246 Decrease 12.3 5.9 0.0 0.0 terpinolene 1296 Only in inoculated 0.0 0.0 100.0 0.0 Octen 3 one, 1 1310 Only in control ( ) 100.0 0.0 ( ) 100.0 0.0 Octyl acetate 1497 Decrease 8.3 15.2 ( ) 100.0 0.0 Linalool 1545 Decrease ( ) 17.5 8.8 ( ) 68.4 23. 6 Neryl acetate 1728 Only in control ( ) 100.0 0.0 ( ) 100.0 0.0

PAGE 87

87 Figure 71. Trend in aroma impact compounds perceived intensity

PAGE 88

88 Figure 72. Control and inoculated aromagram at incubation day one

PAGE 89

89 Figure 73. Control and inoculated aromagram at incubation day four

PAGE 90

90 Figure 74. Difference in inoculated sample aroma intensity from control at incubation day one

PAGE 91

91 Figure 75. Difference in inoculated sample aroma intensity from control at incubation day two

PAGE 92

92 Figure 76. Differen ce in inoculated sample aroma intensity from control at incubation day three

PAGE 93

93 Figure 77. Difference in inoculated sample aroma intensity from control at incubation day four

PAGE 94

94 CHAPTER 8 CONCLUSIONS AND FURT HER WORK Overall, there was a decrease in volat ile quantity and aroma intensity in the inoculated sample when compared to the control. In addition, v arious differentiating volatiles were observed. Of most importance are the qualitative differences including methanol, ethyl propanoate, ethyl 2methylbutanoate, 3methylbutanol, (Z) 3 hexenol, and 1octen3 ol, which are only identified in the inoculated sample. Neryl acetate, another qualitative difference is also of importance, as this was the only volatile identified in the control, but not the inoc ulated. These volatile s, along with others, make a series of 44 differentiating volatiles that could be used to identify P digitatum contamination in single strength OJ under carefully control conditions The differences betw een control and inoculated p ercent difference for some volatiles, such as ethyl butanoate, demonstrate the ability to determine length of contamination. The day of maximum percent difference depends on specific storage conditions, such as temperature, and would fluctuate with alteration in conditions. Further study and validation is required to apply this concept. Sulfur compounds observed from the GC S analysis could also be classified as qualitative or quantitative differentiating volatiles Their application in determining the presence of contamination is the same as the differentiating volatiles from the GC MS analysis. The formation and breakdown of volatiles results from the fungal metabolic pathway and enzymatically driven reactions, including bioconversions. These reactions help explain relationships between volatile s, such as the increase in ethanol and the 11% reduction in Brix. The increase or decrease in a certain volatile is usually the

PAGE 95

95 result of multiple reactions. Enzymes also aid in the bioconversion of vol atile s, such as terpineol. Some of the volatiles that change are aroma impact compounds These volatiles, including neryl acetate and octyl acetate, decreased overall resulting in a decrease in perceived aroma intensity in the inoculated sample. A terpinolene was the only aroma impact compound found to increase over the incubation period. The qualitative and quantitative differences in volatiles, including sulfur volatiles and aroma impact compounds between control and inoculated samples demonstrate that contamination from P digitatum results in unique reactions These reactions have profound effects on the volatiles present, causing the development of differentiating volatiles. The differentiating volatiles consist of both qualitat ive and quantitative differences that are distinctive and can be used as indicators of P digitatum growth. Studies of this nature, however, are onl y examinations of potential (Hutchinson 1973) The results here do show some potential for industry application as a means to identify P digitatum in OJ This study, however, was performed under controlled laboratory conditions and validation under commercial conditions would be needed prior to application of this methodology for quality control. Further studies examining more organisms would be required to determine volatiles specific only to P digitatum In addition, running more replicates and exploring the results in different OJ would allow for greater determination of the specificity of the volatiles found here to P digitatum and to P digitatum grow n in Simply Orange OJ.

PAGE 96

96 APPENDIX A VOLATILE TABLES Table A 1. All peaks detected using GC MS LRI Volatile LRI Volatile LRI Volatile 714 Acetaldehyde 1262 Unidentified 1591 Unidentified 825 Unidentified 1267 terpinene 1599 Unidentified 829 Unidentified 1270 Unidentified 1608 Unidentified 900 Ethyl acetate 1276 Unidentified 1626 elemene 909 Methanol 1286 Hexyl acetate 1631 Unidentified 947 Ethanol 1292 Unidentified 1636 Terpineol, 4 974 Ethyl propanoa te 1296 P cymene 1646 caryophyllene 991 Unidentified 1300 terpinolene 1657 Unidentified 997 Unidentified 1310 Octanal 1666 Unidentified 1002 Methyl butyrate 1321 Unidentified 1671 Unidentified 1022 Unidentified 1354 Unidentified 1680 Citronellyl ac etate 1037 pinene 1359 Hexanol 1690 selinene 1050 Ethyl butanoate 1365 Unidentified 1697 Unidentified 1065 Ethyl2 methylbutanoate 1378 Unidentified 1704 Unidentified 1077 Unidentified 1383 (E) 2 hexenol 1720 Unidentified 1082 Unidentified 1388 U nidentified 1730 terpineol 1090 Unidentified 1398 (Z) 3 hexenol 1233 Unidentified 1098 Hexanal 1417 Nonanal 1738 Unidentified 1105 Unidentified 1427 Unidentified 1746 Neryl acetate 1120 pinene 1442 Perillene 1750 Unidentified 1132 Sabinene 1451 Et hyl octanoate 1753 Unidentified 1161 3 carene 1465 Octen 3 ol, 1 1760 Unidentified 1171 Myrcene 1468 Unidentified 1767 Unidentified 1175 Unidentified 1471 p dimethylstyrene 1770 Valencene 1180 Unidentified 1476 Unidentified 1779 selinene 118 6 Unidentified 1488 Unidentified 1784 Unidentified 1194 terpinene 1492 Octyl acetate 1788 Carvone 1203 Unidentified 1500 Pentadecane 1798 cadinene 1212 Methylbutanol, 3 1526 Decanal 1812 Unidentified 1228 Limonene 1531 copaene 1819 panasinsen 1240 phellandrene 1558 Linalool 1830 Unidentified 1245 Ethyl hexanoate 1569 Octanol 1842 Perillaldehyde 1249 E 2 hexenal 1575 Benzaldehyde 1856 Unidentified 1256 Unidentified 1580 Unidentified 1864 Cis carveol

PAGE 97

97 Table A 2. All peaks detected using G C S LRI Sulfur Volatile 695 Unidentified 738 Dimethyl sulfide 942 Unidentified 1093 Dimethyl disulfide 1413 Unidentified 1484 Unidentified 1557 Unidentified 1563 Unidentified 1597 Dithiane, 1,4 1670 Unidentified 1743 Unidentified 1779 Phene thyl mercaptan 1789 Unidentified 2125 Unidentified

PAGE 98

98 APPENDIX B AROMA IMPACT COMPOUNDS Table B 1. All aroma impact compounds detected using GC O LRI Observed Aroma Descriptor Volatile 836 Sulfuric, sour, fruity Unidentified 1031 Fruity, apple, swe et, cotton candy Unidentified 1048 Fruity, citrus, bubble gum, pineapple Unidentified 1074 Green, grassy, fresh Hexanal 1151 Musty, moldy, metallic, alcoholic Myrcene 1191 Leafy, pungent, spicy, floral Unidentified 1209 Rubber, soil Unidentified 1230 Sweet, plants, floral, spicy Unidentified 1246 Fruity, fresh, green, grassy Ethyl hexanoate 1284 Minty, sweet, citrus Unidentified 1296 Green, candy, fruity terpinolene 1310 Mushroom Octen 3 one, 1 1370 Musty, moldy, earthy, green bean Unidentified 1391 Celery, herbal Unidentified 1445 Cooked potato Unidentified 1497 Minty, fruity, pine like, bamboo, leaf Octyl acetate 1530 Dusty, earthy, damp, herbal Unidentified 1545 Floral, flowery, sweet Linalool 1660 Medicinal Unidentified 1728 Woody, rosy Neryl acetate

PAGE 99

99 LIST OF REFERENCES Achilea O, Fuchs Y, Chalutz E & Rot I. 1985. The contribution of host and pathogen to ethylene biosynt hesis in Penicillium digitatum infected citrus fruit. Physiologial Plant Pathology 27(1):5563. Acree TE & Arn H. 2004. Flavornet and human odor space. Gas chromatography olfactometry (GCO) of natural products Geneva, NY. Adams A, Demyttenaere JCR & D e Kimpe N. 2003. Biotransformation of (R) (+) and (S) ( ) terpineol by Penicillium digitatum -investigation of the culture conditions. Food Chemistry 80(4):525534. Alaa A, Alkorta I, Domnguez JB, Llama MJ & Serra JL. 1990. Pectin lyase activity in a Penicillium italicum strain. Applied Environmental Microbiology 56(12):37553759. Ariza MR, Larsen TO, Petersen BO, Duus J & Barrero AF. 2002. Penicillium digitatum metabolites on synthetic media and citrus fruits. Journal of Agricultural and Food Chemistry 50(22):63616365. Barmore CR & Brown GE. 1979. Role of pectolytic enzymes and galacturonic acid in citrus fruit decay caused by Penicillium digitatum Phytopathology 69(9):675678. Bush DA & Codner RC. 1968. The nature of macerating fac tor of Penicillium digitatum saccardo. Phytochemistry 7(5):863869. Bush DA & Codner RC. 1970. Comparison of the properties of the pectin transeliminases of Penicillium digitatum and Penicillium italicum Phytochemistry 9(1):8797. Caccioni DRL, Guizzard i M, Biondi DM, Agatino R & Ruberto G. 1998. Relationship between volatile components of citrus fruit essential oils and antimicrobial action on Penicillium digitatum and Penicillium italicum International Journal of Food Microbiology 43(12):73 79. Cai J, Liu B & Su Q. 2001. Comparison of simultaneous distillation extraction and solid phase microextraction for the determination of volatile flavor components. Journal of Chromatography A 930(12):1 7. Cevallos Cevallos JM, Reyes De Corcuera JI, Etxeberria E, Danyluk MD & Rodrick GE. 2009. Metabolomic analysis in food science: A review. Trends in Food Science & Technology 20(1112):557566. Cole ALJ & Wood RKS. 1970. Pectic enzymes associated with Penicillium digitatum decay of citrus fruit. Annals of Botany 34(1):211216.

PAGE 100

100 Dav BA, Kaplan HJ & Petrie JF. 1981. The isolation of Penicillium digitatum Sacc. strains tolerant to 2 AB, SOPP, TBZ and benomyl. Proceedings of the Florida State Horticultural Societ 93:344347. Demyttenaere JCR, Van Belleghem K & De Kimpe N. 2001. Biotransformation of ( R ) (+) and ( S ) ( ) limonene by fungi and the use of solid phase microextraction for screening. Phytochemistry 57(2):199208. Doyle MP & Beuchat LR. 2007. Food microbiology: Fundamentals and frontiers. 3rd ed. Washington, DC: American Society for Microbiology Press. p. 1038. Droby S, Eick A, Macarisin D, Cohen L, Rafael G, Stange R, McColum G, Dudai N, Nasser A, Wisniewski M & Shapira R. 2008. Role of citrus volatiles in host recognition, germination and growth of Pen icillium digitatum and Penicillium italicum Postharvest Biology and Technology 49(3):386396. Duetz WA, Bouwmeester H, van Beilen JB & Witholt B. 2003. Biotransformation of limonene by bacteria, fungi, yeasts, and plants. Applied Microbiology and Biotech nology 61(4):269277. Eckert JW & Ratnayake M. 1994. Role of volatile compounds from wounded oranges in induction of germination of Penicillium digitatum conidia. Phytopathology 84(7):746750. Faville LW & Hill EC. 1951. Incidence and significance of mic roorganisms in citrus juices. Food Technology 5(10):423425. Filtenborg O, Frisvad JC & Thrane U. 1996. Moulds in food spoilage. International Journal of Food Microbiology 33(1):85102. French RC. 1985. The bioregulatory action of flavor compounds on fungal spores and other propagules. Annual Review of Phytopathology 23(1):173199. Fries N. 1973. Effects of volatile organic compounds on the growth and development of fungi. Transactions of the British Mycological Society 60(1):121. Huis in't Veld JHJ. 1 996. Microbial and biochemical spoilage of foods: An overview. International Journal of Food Microbiology 33(1):118. Hutchinson SA. 1973. Biological activities of volatile fungal metabolites. Annual Review of Phytopathology 11(1):223246. Janssens L, De Pooter HL, Schamp NM & Vandamme EJ. 1992. Production of flavours by microorganisms. Process Biochemistry 27(4):195215.

PAGE 101

101 metabolites. Letters in Applied Microbiology 36(5):263267. Kavanagh JA & Wood RKS. 1971. Green mould of oranges caused by Penicillium digitatum Sacc.; effect of additives on spore germination and infection. Annals of Applied Biology 67(1):3544. Lacey J. 1989. Preand post harvest ecology of fungi causing spoilage of foods and other stored products. Journal of Applied Microbiology 67(s18):11s 25s. Larsen TO & Frisvad JC. 1994. A simple method for collection of volatile metabolites from fungi based on diffusive sampling from Petri dishes. Journal of Microbiological Methods 19(4):297305. Larsen TO & Frisvad JC. 1995a. Characterization of volatile metabolites from 47 Penicillium taxa. Mycological Research 99(10):11531166. Larsen TO & Frisvad JC. 1995b. Chemosystematics of Penicillium based on prof iles of volatile metabolites. Mycological Research 99(10):11671174. Larsen TO & Frisvad JC. 1995c. Comparison of different methods for collection of volatile chemical markers from fungi. Journal of Microbiological Methods 24(2):135144. Lui L, Vikram A, Hamzehzarghani H & Kushalappa A. 2005. Discrimination of three fungal diseases of potato tubers based on volatile metabolic profiles developed using GC/MS. Potato Research 48(1):8596. Moalemiyan M, Vikram A & Kushalappa AC. 2007. Detection and discrimin ation of two fungal diseases of mango (cv. Keitt) fruits based on volatile metabolite profiles using GC/MS. Postharvest Biology and Technology 45(1):117125. Moalemiyan M, Vikram A, Kushalappa AC & Yaylayan V. 2006. Volatile metabolite profiling to detect and discriminate stem end rot and anthracnose diseases of mango fruits. Plant Pathology 55(6):792802. Narciso JA & Parish ME. 1997. Endogenous mycoflora of gabletop carton paperboard used for packaging fruit juice. Journal of Food Science 62(6):12231239. Nilsson T, Larsen TO, Montanarella L & Madsen J. 1996. Application of headspace solid phase microextraction for the analysis of volatile metabolites emitted by Penicillium species. Journal of Microbiological Methods 25(3):245255. Parish ME. 1991. Microbiological concerns in citrus juice processing. Food Technology 45(4):128136.

PAGE 102

102 Parish ME. 1997. Public health and nonpasteurized fruit juices. Critical Reviews in Microbiology 23(2):109119. Parish ME & Higgins DP. 1989. Yeasts and molds isolated fr om spoiling citrus products and by products. Journal of Food Protection 52(4):261263. Patrick R & Hill EC. 1959. Microorganisms in juice from freeze damaged oranges of various acidities and their resistance to heat. Applied Microbiology 7(4):193195. Pe lser PdT & Eckert JW. 1977. Constituents of orange juice that stimulate the germination of conidia of Penicillium digitatum Phytopathology 67(6):747757. Pesis E & Marinansky R. 1990. Volatile production induced by Penicillium digitatum in orange fruit and in culture. Journal of Phytopathology 128(4):306314. Raccach M & Mellatdoust M. 2007. The effect of temperature on microbial growth in orange juice. Journal of Food Processing and Preservation 31(2):129142. Reineccius G. 2006. Flavor chemistry and t echnology 2nd ed. Boca Raton, FL: CRC Press. Rouseff R. 2006. Citrus flavor and color. In: Smoot, J., editor). Flavor database: Retention index wax Lake Alfred, FL: University of Florida Citrus Research and Education Center. Schnrer J, Olsson J & Br jesson T. 1999. Fungal volatiles as indicators of food and feeds spoilage. Fungal Genetics and Biology 27(23):209217. Stange RR, Midland SL, Sims JJ & McCollum GT. 2002. Differential effects of citrus peel extracts on growth of Penicillium digitatum P italicum and P expansum Physiological and Molecular Plant Pathology 61(5):303311. terpineol by immobilized Penicillium digitatum Applied Microbiology and Biotechnology 49(1):96101. Tan Q & Day DF. 1998b. Organic co solvent effects on the bioconversion of (R) (+) limonene to (R) (+) terpineol. Process Biochemistry 33(7):755761. Tan Q, Day DF & Cadwallader KR. 1998. Bioconversion of (R) (+) limonene by P digitatum (NRRL 1202). Process Biochem istry 33(1):2937. Tournas VH, Heeres J & Burgess L. 2006. Moulds and yeasts in fruit salads and fruit juices. Food Microbiology 23(7):684688. Vikram A, Lui LH, Hossain A & Kushalappa AC. 2006. Metabolic fingerprinting to discriminate diseases of stored carrots. Annals of Applied Biology 148(1):1726.

PAGE 103

103 Vikram A, Prithiviraj B, Hamzehzarghani H & Kushalappa AC. 2004a. Volatile metabolite profiling to discriminate diseases of McIntosh apple inoculated with fungal pathogens. Journal of the Science of Food and Agriculture 84(11):13331340. Vikram A, Prithiviraj B & Kushalappa AC. 2004b. Use of volatile metabolite profiles to discriminate fungal disease of Cortland and Empire apples. Journal of Plant Pathology 86(3):215225. Wady L, Bunte A, Pehrson C & Lars son L. 2003. Use of gas chromatography mass spectrometry/solid phase microextraction for the identification of MVOCs from moldy building materials. Journal of Microbiological Methods 52(3):325332. Wyatt MK & Parish ME. 1995. Spore germination of citrus juice related fungi at low temperatures. Food Microbiology 12:237243. Wyatt MK, Parish ME, Widmer WW & Kimbrough J. 1995. Characterization of mould growth in orange juice. Food Microbiology 12:3473 55. Zhang Z & Pawliszyn J. 1993. Headspace solidphase microextraction. Analytical Chemistry 65(14):18431852.

PAGE 104

104 BIOGRAPHICAL SKETCH Gabriel Louis Shook graduated summa cum laude from the University of Florida with a Bachelor of Science in food science and human nutrition in May 2009. Throughout his undergraduate years, he has authored a food microbiology undergraduate thesis entitled Permeability of Staphylococcus aureus through Latex Finger Cots, been accepted into The National Society of Collegiate Scholars, Delta Epsilon Iota, and Golden Key International Honor Society, and was the recipient of the food science departments Outstanding Senior award. In order to further his knowledge in food chemistry, he chose to study flavor chemistry under Dr. R ussell Rouseff at the University of Florida and entered into the Master of Science program in the food science and human nutrition department. Gabriels project was a mixture of food microbiology and food chemistry. While working towards his m aster s degree, Gabriel interned with Kraft Foods in East Hanover, New Jersey Upon graduation, he returned to East Hanover, NJ, to work as a food product developer for Kraft Foods, Nabisco.