Survival and Exopolysaccharide Production of Lactic Acid

Survival and Exopolysaccharide Production of Lactic Acid Bacteria

Grown on Grape Pomace _______________________________________

A Thesis

presented to

the Faculty of the Graduate School

at the University of Missouri

_______________________________________________________

In Partial Fulfillment

of the Requirements for the Degree

Master of Science

_____________________________________________________

by

WARREN E. AULD

Dr. Azlin Mustapha, Thesis Supervisor

DECEMBER 2010

© Copyright by Warren E. Auld 2010

All Rights Reserved

The undersigned, appointed by the dean of the Graduate School, have

examined the thesis entitled

SURVIVAL AND EXOPOLYSACCHARIDE PRODUCTION

OF LACTIC ACID BACTERIA GROWN ON GRAPE POMACE

presented by Warren E. Auld,

a candidate for the degree of master of science,

and hereby certify that, in their opinion, it is worthy of acceptance.

Dr. Azlin Mustapha, Food Science

Dr. Ingolf U. Grün, Food Science

Dr. Mengshi Lin, Food Science

Dr. Mark Ellersieck, Statistics

To Becky

What a long strange trip it's been. The Grateful Dead

ii

ACKNOWLEDGEMENTS

I would like to thank Les Bourgeois Vineyards and their head

winemaker Cory Bomgaars for supplying the grape pomace used in

these experiments.

I would also like to thank Richard Oberto and Harold Huff who

patiently explained the care and feeding of a freeze drier.

My thanks to the Robert T. Marshall and Marion Fields Fellowships

for their generous assistance in funding my studies.

Many thanks to my lab mates for all your support and letting me

think out loud.

And to my commitee members, especially Dr. Mustapha, thank

you for putting up with me all these years.

iii

Table of Contents

ACKNOWLEDGEMENTS .................................................................. ii

LIST OF ILLUSTRATIONS .............................................................. ix

LIST OF TABLES .......................................................................... xi

Chapter

1. INTRODUCTION .................................................................... 1

2. LITERATURE REVIEW ............................................................. 4

2.1 Fiber .............................................................................. 4

2.1.1 Early definitions of fiber ........................................... 4

2.1.2 Current definitions of fiber ........................................ 6

2.1.3 Properties of fiber in a food system ............................ 9

2.1.3.1 Solubility ..................................................... 9

2.1.3.2 Viscosity ................................................... 10

2.1.3.3 Fermentability............................................ 11

2.1.4 Health benefits of fiber ........................................... 12

2.1.5 Fiber sources ........................................................ 13

2.2 Lactic acid bacteria (LAB) ................................................ 16

2.2.1 Lactic acid bacteria defined ..................................... 16

2.2.2 Divisions within LAB ............................................... 17

2.2.3 Lactic acid bacteria in food ..................................... 20

iv

2.2.4 LAB nutritional requirements ................................... 21

2.2.5 Lactobacillus culture media. .................................... 23

2.2.5.1 Nutritional and growth requirements of

Lactobacillus ......................................................... 24

2.3 Grape pomace ............................................................... 25

2.4 Exopolysaccharides (EPS) ............................................... 27

2.5 Composition of EPS from LAB used in this study................. 28

2.6 Fiber isolation and extraction methods .............................. 28

2.7 Structural analysis of fiber .............................................. 35

2.8 Fourier transform infrared (FTIR) spectroscopy .................. 36

2.8.1 Use of IR spectroscopy in the food industry .............. 37

2.8.1.1 Lipids ........................................................ 37

2.8.1.2 Milk .......................................................... 39

2.8.1.3 Protein ...................................................... 40

2.8.1.4 Carbohydrates ........................................... 40

2.8.2 IR spectroscopy and microbiology ............................ 42

2.8.2.1 Identification of bacterial species .................. 42

2.8.2.2 FTIR and EPS ............................................. 45

2.9 Objectives ..................................................................... 47

3. MATERIALS AND METHODS .................................................. 48

3.1 Bacterial strains ............................................................. 48

v

3.2 Grape pomace ............................................................... 48

3.3 pH measurement ........................................................... 49

3.4 Preliminary studies ......................................................... 49

3.4.1 Growth on pomace (broth) ..................................... 49

3.4.2 Growth on pomace (agar) ....................................... 50

3.4.3 Further growth and preliminary survival studies ........ 51

3.4.3.1 Growth of lactic acid bacteria in pomace

solution without pH adjustment ............................... 51


solution with pH adjustment ................................... 52

3.5 Primary studies .............................................................. 53

3.5.1 pH Adjustment ...................................................... 53

3.5.2 Survival study ....................................................... 54

3.5.3 Exopolysaccharide (EPS) production study ................ 55

3.5.3.1 Sample preparation .................................... 55

3.5.3.2 Soluble fiber extraction ............................... 55

3.5.3.3 Freeze-drying ............................................ 56

3.5.4 Fourier transform infrared spectroscopy ................... 56

3.5.5 Sample preparation ............................................... 57

3.5.6 Sample collection .................................................. 57

3.5.7 Growth and EPS production in tomato juice .............. 59

vi

3.6 Statistical and spectral analyses ...................................... 60

4. RESULTS ............................................................................ 61

4.1 Preliminary Studies ........................................................ 61








4.2 Primary Studies ............................................................. 69

4.2.1 pH adjustment ...................................................... 69

4.2.2 Survival Study ...................................................... 69




4.2.5 Growth and EPS production in tomato juice .............. 82

5. DISCUSSION ...................................................................... 84

5.1 Preliminary studies ......................................................... 84



vii






5.2 Primary studies .............................................................. 88

5.2.1 Survival study ....................................................... 88




5.3 Growth and EPS production in tomato juice ....................... 98

6. CONCLUSION ................................................................... 100

APPENDIX

A. Survivor regression lines .................................................... 102

B. FTIR spectra ..................................................................... 113

C. Programs ......................................................................... 116

C.1 survival-slopes-aov.r .................................................... 116

C.2 extract.r ..................................................................... 117

C.3 peak-compare.r ........................................................... 119

REFERENCES ........................................................................... 124

viii

ix

LIST OF ILLUSTRATIONS

Figure Page

3-1. AOAC method 991.43 as modified for this study.. ................... 58

3-2. Samples on a slide. ............................................................. 59

4-1. Growth of LAB on pomace (9-day trial) .................................. 64

4-2. Growth of LAB on pomace (83-day trial) ................................. 65

4-3. Growth of LAB on pH adjusted Pomace ................................... 67

4-4. Adjusted pH resulting from addition of 4M NaOH and subsequent autoclaving ............................................................... 68

4-5. Growth of L. paracasei on pH adjusted grape pomace .............. 68

4-6. Adjusted pH resulting from addition of 4 M NaOH to second

batch of pomace and subsequent autoclaving ................................ 69

4-7. Average survival curve (adjusted pH) .................................... 74

4-8. pH (pH adusted).................................................................. 75

4-9. Average survival curves (unadjusted pH) ............................... 76

4-10. pH (unadjusted) ................................................................ 77

4-11. Average slope of survivor count regression lines .................... 78

x

4-12. Average slope of pH regression lines .................................... 79

4-13. Average material recovered after extraction and freeze-drying 82

A-1. Survival and pH curves for L. paracasei ................................ 102

A-2. Survival and pH curves for L. acidophilus LA-2 ...................... 103

A-3. Survival and pH curves for L. fermentum ............................. 104

A-4. Survival and pH curves for Kefir Coccus AM-W1 .................... 105

A-5. Survival and pH curves for L. acidophilus ADH ...................... 106

A-6. Survival and pH curves for L. bulgaricus Lb-12 ...................... 107

A-7. Survival and pH curves for L. rhamnosus AM-L3 .................... 108

A-8. Survival and pH curves for L. paracasei AM-L4 ...................... 109

A-9. Survival and pH curves for L. paracasei AM-L6 ...................... 110

B-1. FTIR spectra of pH-adjusted extract samples ........................ 113

B-2. FTIR spectra of unadjusted extract samples .......................... 114

B-3. FTIR spectra of samples grown in tomato juice. .................... 115

xi

LIST OF TABLES

Table Page

2-1. Gums and Thickeners of microbial origin ................................ 14

2-2. Families and genera of LAB (adapted from Jay 2005). .............. 18

2-3. LAB used in AOAC vitamin bioassays. ..................................... 23

2-4. Nutritional requirements of species used in this study .............. 25

2-5. AOAC Methods for Determination of Dietary Fiber .................... 34

3-1. Bacterial strains used. .......................................................... 49

3-2. Pomace Agar ...................................................................... 51

4-1. Growth of lactic acid bacteria on pomace (agar) ...................... 62

4-2. Growth ranges (CFU/mL) of LAB in pomace solution for 83 d. ... 66

4-3. Extraction results after growth in diluted tomato juice. ............. 83

5-1. Peak Assignments ............................................................... 96

A-1. Tukey's HSD for the slopes of the survivor regression lines. .... 111

A-2. Tukey's HSD for pH-adjusted and pH-unadjusted subsets of the slopes of the survivor regression lines. .................................. 112

1

CHAPTER 1

INTRODUCTION

The food industry is continuously seeking new ingredients for use

as thickeners, emulsifiers, or dietary fiber. Historically, these

ingredients have largely come from plants. However, microbial

production of such ingredients have been explored. For example,

xanthan gum, which was approved by the Food and Drug

Administration (FDA) in 1969 for use as a thickening agent

(Sutherland 2004) is produced by growing the bacterium,

Xanthamonas campestris, on sucrose or glucose (Sutherland 2001).

Other thickeners produced by microorganisms include gellan and

curdlan which are also approved for use in food (Sutherland 2001).

Dietary fiber is a functional ingredient that has received much

attention recently. In addition to its properties within food, a diet high

in fiber is believed to provide a number of health benefits. These

include improved laxation, and possible reduction of serum cholesterol.

Fermentable fibers can supply nutrients for friendly (probiotic) bacteria

that reside in the colon. The Food and Nutrition Board of Institute of

2

Medicine (FNB/IOM) has recommended a daily intake of 14 g/1000

kcal of fiber based on epidemiological and clinical data suggesting that

adequate fiber may reduce the risk of coronary heart disease (Otten

and others 2006). Cardiovascular disease is the cause of 1.4 million

deaths and 865,000 heart attacks annually in the United States

(American Heart Association 2005).

The concept of converting food industry waste by-products into

something valuable is not new. In many cases, the waste is simply

milled into a powder for direct addition to food (Laufenberg and others

2003, Larrauri 1999). This approach has been considered for using

grape pomace as a food additive (Valiente and others 1995; Martin-

Carron and others 1997; Goni and others 2005). Sutherland (1996)

notes that many bacteria which produce EPS, including X. campestris,

do not have strict dietary requirements and are capable of secreting a

number of enzymes which can hydrolyze molecules such as cellulose,

pectin, amylose and protein. Whey from cheese production, at one

time considered an inconvenient waste product, is now exploited as a

source of whey protein, lactose, and as feedstock for biogas and

single-cell protein fermentations, as reviewed by Marwaha and

Kennedy (1988) and Gonzalez Siso (1996). More recently, work has

3

been done to utilize whey in alcohol production (Athanasiadis and

others 2002; Kargi and Ozmihci 2006).

There are two reasons for believing that microbial conversion is a

superior method for recycling such waste products back into the food

supply. First, the product is an odorless, tasteless, light colored

powder which does not impart the flavor of the substrate to whatever

it is added to. In addition, this product is consistent. Our input

substrates are agricultural products which vary from season to season

and field to field. Second, the enzymes which catalyze the catabolic

reactions which assemble EPS produce very specific stereochemistry

which can be difficult to replicate. There is evidence which suggests

that the quantity and composition of the product depends not only on

bacterial species (De Vuyst and Degeest 1999, Laws and others 2001,

Sutherland 2001) but the substrate as well (Mozzi and others 2001).

4

CHAPTER 2

LITERATURE REVIEW

2.1 Fiber

2.1.1 Early definitions of fiber

One of the earliest mentions of fiber as a necessary dietary

component is found in an account by Dr. William Beaumont (1834), a

US Army surgeon. The book recounts Dr. Beaumont's treatment of his

patient, Alexis St. Martin, who suffered a wound which, after healing,

left a permanent opening leading into his stomach. Regarding fiber,

Dr. Beaumont says,

The quality of nutriment is a matter of considerable importance in dietetic regulations. Bulk is, perhaps, nearly

as necessary to the articles of diet as the nutrient principle. They should be so managed that one should be

in proportion to the other. Too highly nutritive diet is probably as fatal to the prolongation of life and health, as

that which contains an insufficient quantity of nutriment. (Beaumont 1834)

What Beaumont is describing would later become known as the

dietary fiber hypothesis. More modern versions of the hypothesis

5

state that there is a link between a diet with adequate fiber intake and

a low incidence of diabetes, heart disease, some forms of cancer and

chronic bowel disorder (Dreher 2001).

Beaumont's idea had little impact on the scientific community for

some time. For instance, Atwater, working in the latter half of the 19th

century, included fiber, ash and moisture content in his food analyses

but discounted them as acaloric and therefore unimportant (Carpenter

2003). A century after Beaumont's book was published, Hipsley

(1953) suggested a link between dietary fiber and the pregnancy

diseases eclampsia and toxemia. In this paper, Hipsley coined the

term "dietary fiber" and defined it to mean lignin, cellulose and

hemicellulose, what we would now term insoluble dietary fiber.

The next major step was taken by Trowell (1972) who defined

dietary fiber as "the skeletal remains of plant cells that are resistant to

hydrolysis by the enzymes of man." At the time, Trowell believed that

dietary fiber components were only located in plant cell walls (Trowell

and Burkitt 1987). Trowell and others (1976) would later expand this

definition to include "plant polysaccharides and lignin which are

resistant to hydrolysis by the digestive enzymes of man".

6

2.1.2 Current definitions of fiber

At the time of this writing, there are numerous competing

definitions for dietary fiber. The Institute of Medicine/Food Nutrition

Board (IOM/FNB) included an extensive historical collection of

definitions of dietary fiber in their report on a proposed definition of

dietary fiber (IOM 2001). The variation reflects changes in our

understanding of dietary fiber over time and the particular interests

and weltanschauung (world-view) of the people writing the definitions

(Prosky 2001). Here we present a small sample of some of the

approaches taken.

The American Association of Cereal Chemists (AACC) has

adopted the following definition:

Dietary fiber is the edible parts of plants or analogous carbohydrates that are resistant to digestion and

absorption in the human small intestine with complete or partial fermentation in the large intestine. Dietary fiber

includes polysaccharides, oligosaccharides, lignin, and associated plant substances. Dietary fibers promote

beneficial physiological effects including laxation, and/or blood cholesterol attenuation, and/or blood glucose

attenuation. (AACC 2001)

The definitions given in the Dietary Reference Intakes published

by the Institute of Medicine/Food Nutrition Board in cooperation with

Health Canada are as follows:

7

Dietary fiber consists of nondigestible carbohydrates and

lignin that are intrinsic and intact in plants.

Functional fiber consists of isolated nondigestible carbohydrates that have beneficial physiological effects in

humans.

Total fiber is the sum of dietary fiber and functional fiber. (Otten and others 2006)

The AACC responded to this definition with an article titled "All

Dietary Fiber Is Fundamentally Functional" which included the

following summary:

The recent publication of definitions relating to Dietary

Fiber by the Institutes of Medicine are not supported by current science on the subject. These definitions will not

have the desired objective of increasing both consumer knowledge about and consumption of dietary fiber.

Instead, they will result in confusion regarding health messages about dietary fiber. AACC believes a science-

based definition and research directions are needed to

optimize use of resources to promote healthier life styles utilizing accurate and clear consumer information. (AACC

2003)

These definitions, while in many ways similar, illuminate major

differences between these groups originating from their underlying

organizational purposes. Both groups seek to advance their own

agenda through the adoption of their definition1.

1 No value judgement regarding either agenda is intended.

8

In the United States, the Food and Drug Administration pursuant

to the Nutrition Labeling and Education Act of 1990 (NLEA) has

given us a de facto definition of fiber:

The nondigestible dietary fiber will be determined by the

method "Total Dietary Fiber in Foods, Enzymatic

Gravimetric Method, First Action." [sic] in the Journal of the Association of Official Analytical Chemists (JAOAC),

68:399, 1985, as amended in JAOAC, 69:370, 1986. Both methods are incorporated by reference. (FDA 1987)

The method referred to in this definition is now called "AOAC

INTERNATIONAL Enzymatic-Gravimetric Method 985.29" (FDA 2007).

According to this definition, the insoluble carbohydrate fraction

combined with that portion of the soluble carbohydrate fraction which

can be precipitated with ethanol is considered dietary fiber (Horwitz

and Latimer 2007). The potential exists for large errors from this

method due to precipitation of non-SDF components as well as

incomplete precipitation of the SDF fraction (Mañas and Saura-Calixto

1993). Southgate (1969) highlighted the underlying problem by

saying "Procedures for the determination of available carbohydrates in

foods are concerned with the measurement of a relatively small

number of well-defined substances. This is not the case as far as the

unavailable carbohydrates are concerned; they form a group of

heterogeneous substances whose chemistry is, in many cases, poorly

9

understood." In addition, inulin and oligofructose which are often

classified as dietary fiber are not recovered by the AOAC method

(Coussement 1999) as well as resistant starch (Asp 1987). AOAC

Method 985.29 was designed as a compromise between speed of

analysis and accuracy of results (Prosky 2001).

2.1.3 Properties of fiber in a food system

Fiber has a number of interesting properties within a food

system. The most important of these are solubility, viscosity and

fermentability.

2.1.3.1 Solubility

Solubility in this context has the usual meaning in chemistry;

how much of one substance can be disolved in a given quantity of

another substance (Daintith 2000). Pectin, gums, and -glucans are

considered soluble fibers while cellulose, lignin, chitin and chitosan are

insoluble. Hemicellulosic fibers can fall into either category depending

on the material (Gropper and others 2009). Solubility plays a role

both in our definition (2.1.2) and analysis of fiber (2.6). From a

product development perspective, solubility can be an important

10

characteristic when selecting an additive. If the end product is

intended to be smooth and creamy or a liquid, insoluble fiber would

not be an appropriate choice. Bonnand-Ducasse and others (2010)

found that the addition of insoluble dietary fiber improved the water

holding capacity of dough and shorten the time for optimum dough

development while addition of soluble dietary fiber altered the

Newtonian viscosity and shear rate. Peressini and Sensidoni (2009),

working with inulin2, found changes in the viscoelastic and

breadmaking characteristics of dough depended on both the quantity

and type of additive used.

2.1.3.2 Viscosity

Viscosity is a measure of a material's resistance to shear stress

(Singh and Heldman 2009). Less formally, viscosity refers to the

thickness of a liquid; water, vegetable oil and honey have increasing

thickness and thus increasing viscosity. The food industry uses gums

and pectins to thicken and stabilize foods (Potter and Hotchkiss 1998).

In some instances viscous fiber serves to both improve the rheological

properties of the food product and fortify the food with additional fiber.

2 A soluble dietary fiber additive derived from chicory and Jerusalem

artichoke (Peressini and Sensidoni 2009).

11

For example, Chen and others (2010) added soluble fiber derived from

okara as thickeners in dairy-based products. Viscous fiber can also be

used as a fat replacer. Inglett and others (2004) reported success

replacing coconut cream with Ricetrim in Thai foods. Morin and others

(2002) suggested the use of -glucan gum in the manufacture of

reduced-fat sausages. Nutritionally, viscosity is related to water-

binding and gel formation, both of which tend to slow gastric emptying

and digestion which increase the transit time through the digestive

system. At the same time, the increased viscosity also reduces

enzymatic access to nutrients in the lumen of the small intestine

leading to reduced absorption (Gropper and others 2009)3.

2.1.3.3 Fermentability

As was noted above, fiber is not digested by the enzymes of the

human small intestine. The large intestine, however, contains an

ecosystem made up of trillions of bacteria (Fraune and Bosch 2010)4

and some dietary fiber is fermentable by these bacteria (Dongowski

and others 2000; El Oufir and others 2000; Park and Floch 2007;

3 Insoluble fiber can also limit absorption by acting as a physical barrier (Gropper and others 2009). 4 This is roughly ten times the number of somatic cells in the human body (Fraune and Bosch 2010). As one commentator put it, "we're all just petri

dishes with shoes" (anyletter 2010).

12

Resta 2009; Sahakian and others 2010; Mirande and others 2010).

The products of this fermentation include lactate and short chain fatty

acids (SCFA). Both serve to acidify the intestinal lumen leading to a

reduction in secondary bile synthesis. SCFA increase sodium and

water absorption and encourage mucosal cell growth and possibly

reduce the risk of some cancers (Gropper and others 2009).

Fermentable fiber may also play a role in the control of obesity

(Keenan and others 2006; Venema 2010).

2.1.4 Health benefits of fiber

In addition to their structural and textural roles in food, fiber

may also confer health benefits on the consumer. Possible benefits

include improved laxation, reduced risk of obesity by promoting

satiety, normalization of blood lipids and reduced risk of coronary

heart disease (Otten and others 2006). Soluble fiber may also have a

prebiotic effect which promotes the the growth of probiotic bacteria

(reviewed by Slavin and others 2009).

13

2.1.5 Fiber sources

Most edible plant materials, fruits, nuts, vegetables, pulses and

grains, contain dietary fiber. This is true of both unrefined and refined

plant materials. However, the latter will generally have a lower fiber

content (Otten and others 2006). Processing operations such as

milling wheat into flour or juicing apples remove components such as

bran and apple pomace, thus lowering the fiber content of the final

product (Potter and Hotchkiss 1998). The material removed may be

added as an enrichment to the same or a different product; however,

the desired physiological effects will be reduced when compared to

those of the intact plant material (Gropper and others 2009).

As a rule, animal products are not sources of fiber. The

exceptions are chitin, made from the shells of marine crustaceans and

its deacetylated derivative chitosan (Aranaz and others 2009; Gropper

and others 2009)5. Japan and Korea permit the use of chitin and

chitosan as food additives (Korea 2004; JFCRF 2007; Aranaz and

others 2009). They are not permitted for use as food additives in the

United States, the EU or by the Codex Alimentarius (Aranaz and others

2009).

5 The first and second most abundant biopolymers found in nature are

cellulose and chitin (Aranaz and others 2009).

14

Many of the thickeners and gums used by the food industry are

also classified as dietary fiber. Among these are plant materials such

as pectins, galactomannans, plant exudates and products derived from

marine algae (Belitz and others 2004; Daniel and others 2007). Also

included in this group are a number of products derived from

microorganisms (Table 2-1).

Table 2-1 Gums and Thickeners of microbial origin

Product Source

Agrobacterium succinoglycan (e) Agrobacterium sp.

Bacillus natto gum (e) Bacillus natto6

Curdlan (a) Alcaligenes faecalis var. myxogenes

Gellan gum (b) Pseudomonas elodea

Levan (e) Bacillus subtilis

Macrophomopsis gum (e) Macrophomopsis sp.

Pullulan (d) Aureobasidium pullulans

Rhamsan gum (e) Alcaligenes sp.

Sclero gum (e) Sclerotium glucannicum

Welan gum (e) Alcaligenes sp.

Xanthan gum (c) Xanthamonas campestris

(a) FDA 2006, (b) FDA 2008a, (c) FDA 2008b, (d) Hayashibara 2002,

(e) JETRO 2009

6 Probably a synonym for Bacillus subtilis (Buchanan and Gibbons 1974).

15

Which microbial polysaccharides are permitted for use as food

additives depends on both where the food is made and where it is to

be sold. Xanthan gum, gellan gum, curdlan and pullulan have been

granted GRAS status in the United States (EAFUS 2010; Hayashibara

2002). These four additives are also listed as permissible in the Codex

Alimentarius in the General Standard for Food Additives (GFSA online

2010) and the Korea Food Additives Code (Korea 2004). Current

Canadian, Australian and New Zealand regulations only permit addition

of xanthan and gellan gums (Food Additive Dictionary 2006; Food

Standards Code 2010) while those in the European Union permit

xanthan and gellan gums and pullulan (European Parliament 2006).

In addition to the four microbial polysaccharides permitted in the

United States, Japanese regulations also permit the use of

agrobacterium succinoglycan, levan and bacillus natto,

macrophomopsis, rhamsan, sclero and welan gums (JFCRF 2007)7.

7 A USDA Foreign Agriculture Service report indicates that the Japanese

government is in the process of rescinding approval for levan and sclero gum for use as food additives. The basis for the change is the belief that these

additives are not in current use for foods sold in Japan (USDA 2009).

16

2.2 Lactic acid bacteria (LAB)

2.2.1 Lactic acid bacteria defined

The term lactic acid bacteria (LAB) refers to a group of Gram-

positive bacteria, all of which produce lactic acid as the primary end-

product of hexose fermentation (Jay and others 2005). Historically,

LAB were all classified within four genera: Lactobacillus, Leuconostoc,

Pediococcus and Lactococcus (Carr and others 2002). The use of 16S

rDNA to revise the phylogenetic tree (Woese and Fox 1977) has

increased the number from the original 4 to 13 genera (Jay and others

2005)8. These genera (Table 2-2) are all members of the order

Lactobacillales which along with the order Bacillaceae make up the

class Bacilli (De Vos and others 2009)9. The order includes six families.

In addition to the families shown in Table 2-2, Lactobacillales also

contains the family Aerococcaceae which does not include LAB. The

species Lactococcus, Leuconostoc, Streptococcus and Lactobacillus are

used in the dairy industry (Walstra and others 2006). Lactobacillus,

8 Aarnikunnas (2006), citing Axelsson, gives the number of genera as 20. 9 The class name, Bacilli, is derived from the Latin word baculum meaning a

rod or stick (Morris 1969) and refers to the appearance of many of these organisms when viewed through a microscope. Yet a number of the genera

included in this class are coccus, a name derived from the Greek word for berry (Morris 1969) and again referring to the appearance of the organism under the microscope. Originally the class Bacilli was expected to contain

only rod-shaped species. Later changes based on biochemical tests and 16S rDNA rather than morphology have made our system of nomenclature quite

perplexing at times.

17

Leuconostoc and Pediococcus cerevisae are used in the production of

fermented meat products (Pearson and Gillett 1999).

2.2.2 Divisions within LAB

LAB are divided into two groups, as suggested by Orla-Jensen,

based on the nature of their metabolic products when fermenting

glucose. The terms homo- and heterofermentive were later suggested

by Kluyver and Donker (Nelson and Werkman 1935).

Homofermentative LAB produce a single metabolite, lactic acid using

the Embden-Meyerhof (glycolysis) pathway to produce pyruvate (De

Vos and others 2009). Pyruvate is then reduced to lactate by the

oxidation of NADH to NAD+; the latter is then recycled for use in

further glycolysis (Nelson and Cox 2005). For each mole of glucose

consumed, two moles of lactate are formed.

Heterofermentative LAB ferment glucose using the 6-

phosphogluconate/phosphoketolase (6-PG/PK) pathway (Zalán and

others 2010)10. Here, glucose is decarboxylated to ribulose, yielding

one molecule of CO2. Ribulose is then cleaved by phosphoketolase

10 This pathway is also refered to as the Phosphogluconate pathway which is another name for the Pentose Phosphate Pathway. Both share a series of

enzymes which oxidize glucose-6-phosphate to 6-phosogluconate which is then decarboxylated to form ribulose-5-phosphate. In the Pentose

Phosphate Pathway, ribulose-5-phosphate goes through a series of catabolic

18

Table 2-2 Families and genera of LAB (adapted from Jay 2005).

Family Genus

Carnobactriaceae Carnobacterium

Lactosphaera11

Enterococcaceae Enterococcus

Tetragenococcus

Vagococcus

Lactobacillaceae Lactobacillus

Paralactobacillus

Pediococcus

Leuconostocaceae Leuconostoc

Oenococcus

Weissella

Streptococcaceae Streptococcus

Lactococcus

into a glyceraldhyde molecule (3 carbons) which can enter glycolysis

and ultimately become lactate and acetate (2 carbons) which can be

used to reoxidize NADH to NAD+ by converting the acetate to ethanol

reactions which yield sugars ranging in length from 3 to 7 carbons while in

the 6-PG/PK pathway, the 5 carbon molecule is split into a 3-carbon and a 2-carbon molecule. 11 Janssen and others (1995) proposed the transfer of Ruminococcus

pasteurii into a new genus, Lactosphaera. Subsequent work by Liu and others (2002) has led to the transfer of the sole member of this new genus,

Lactosphaera pasteurii, to the genus Trichococcus.

19

(Aarnikunnas 2006). Thus, for each mole of glucose consumed, 1

mole of CO2, 1 mole of lactate and 1 mole of ethanol are formed.

In addition to differing metabolites, these pathways yield

differing amounts of energy. Homofermentative LAB route both 3-

carbon halves of the original glucose through the Embden-Meyerhof

pathway and as a result, extract twice as much energy (Jay and others

2005; De Vos and others 2009).

LAB may be further divided into obligate homofermentative,

facultative heterofermentative and obligate heterofermentative

species. Obligate homofermentative species ferment hexoses to

lactate but cannot ferment pentoses. Facultative heterofermentative

species also ferment hexoses to lactate and they can also be induced

to ferment pentoses. The products of their pentose fermentation do

not include CO2 as there is no decarboxylation from hexose to pentose

required. Obligate heterofermentative species always ferment hexoses

into equimolar quantities of lactate, acetate (or its byproducts) and

CO2 (Jay and others 2005; Corsetti and Settanni 2006; De Vos and

others 2009).

20

2.2.3 Lactic acid bacteria in food

Preserving food from periods of plenty for use in periods of

scarcity has long been crucial in human survival. Among the various

preservation strategies adopted, fermentation is one of the oldest

(Steinkraus 1996a; Prajapati and Nair 2008). Even today,

fermentation by LAB is an important step in production and

preservation of many foods from around the world (reviewed in

Steinkraus 1996b). Many of these fermentations also serve as a

means of vitamin enrichment (Steinkraus 1998)12.

Numerous dairy products are made using LAB which contribute

flavor, help preserve the product, thicken and, in the case of cheese

manufacture, assist in curd formation (Oberman and Libudzisz 1998;

Stanley 1998; Walstra and others 2006; Bertolami and Farnworth

2008; Farnworth and Mainville 2008; Van de Water and Naiyanetr

2008; Miyazuki and Matsuzaki 2008; Vinderola and others 2008; Heller

and others 2008). Vegetables may also be fermented with LAB to

produce products such as sauerkraut and kimchee. Again the primary

goal of this fermentation is to preserve and add flavor (Harris 1998;

Surh and others 2008; Molin 2008; Holzapfel and others 2008; Hamdi

12 Many food fermentations involve mixed cultures or sequential fermentations making difficult to know without further study, which organism

is producing the vitamin.

21

2008; Hsieh and others 2008). LAB are used in the production of

fermented sausages to lower the pH of the product for better

preservation and their contribution of flavor compounds (Lücke 1998;

Pearson and Gillett 1999; Hammes and others 2008). LAB are used in

the production of grain products. In the United States, the most

familiar of these products is sourdough bread whose flavor and

keeping qualities are directly attributable to LAB (Hammes and Gänzle

1998; Vogel and others 1994; Corsetti and others 1998; Corsetti and

Settanni 2006; Kam and others 2007). LAB also play important roles

in the production of cocoa, coffee and tea (Fowler and others 1998),

the production of congeners for whiskey (Varnam and Sutherland

1994) and the malolactic fermentation of wine (Fleet 1998; Uthurry

and others 2006; Bloem and others 2007; García-Ruiz and others

2008).

2.2.4 LAB nutritional requirements

As discussed above, LAB can ferment hexose and

heterofermentative species can also ferment pentose. Some LAB can

make use of organic acids including citric, tartaric, succinic and malic

acids to produce energy for the cell (De Vos and others 2009). The

22

process by LAB of converting malic acid to lactate, referred to as

malolactic fermentation, is an important step in winemaking (Fleet

1998). The acids are usually degraded through oxaloacetic acid to

pyruvate. Lactobacillus fermentum lacks the malolactic enzyme (mle)

and instead converts malate to fumerate to succinate following a

reversal of part of the tri-citric acid cycle (De Vos and others 2009).

In addition to these sources of energy and carbon, LAB require other

growth factors. These are fastidious bacteria which have a long

history of being difficult to grow on well defined, simple media (Kulp

1927). Additional growth factors include amino acids, purine and

pyrimidine base and B vitamins (Jay and others 2005). Researchers

used this requirement to identify and characterize B vitamins such as

folate (Mitchell and others 1941)13 which was at one time refered to as

Lactobacillus casei factor (Johnson 1946). LAB continue to be used in

standard AOAC bioassay methods as shown in Table 2-3 (Horwitz and

Latimer 2007).

13 "Four tons of spinach have been extracted and carried through the first

stages of concentration." Mitchell and others 1941.

23

Table 2-3 LAB used in AOAC vitamin bioassays.

Species Analyte

Lactobacillus leichmannii (ATCC 7830) Cobalamin (B-12)

Streptococcus faecalis (ATCC 8043) Folic Acid

Lactobacillus plantarum (ATCC 8014) Niacin and Pantothenic Acid

Lactobacillus casei subsp. rhamnosus

(ATCC 7469)

Riboflavin (B-2)

2.2.5 Lactobacillus culture media.

A number different culture media have been tried for use with

LAB (reviewed in Schillinger and Holzapfel 2003). Growth is generally

poor on those media composed of only simple carbohydrates and

protein digests (Kulp 1927). It was found that supplementing agar

medium containing peptone with tomato juice produced satisfactory

growth due to an "accessory substance" in the tomato juice (Kulp

1927; Kulp and White 1932). Tomato juice agar using a slightly

revised formula14, along with tomato juice broth are continue to be

used for the maintenance and growth of Lactobacillus. The latter is

made with yeast extract rather than peptonized proteins and includes

a number of inorganic salts (Difco 1998). AOAC bioassays specify the

use of "Lactobacilli broth/agar AOAC" which are further refinements of

tomato juice broth and agar (Difco 1998; Horwitz and Latimer 2007).

14 Peptonized milk is added in addition to peptone.

24

De Man and others (1960) proposed a media which while still not

completely defined, more closely approximates the minimum

nutritional requirements of Lactobacillus. These media are today

refered to as MRS broth and agar (Difco 1998).

2.2.5.1 Nutritional and growth requirements of Lactobacillus

Of the nine species used in this study, eight belong to the genus

Lactobacillus representing five species15. Lactobacilli are generally

mesophillic and aerotolerant. They are also aciduric and some species

are capable of growing at a pH as low as 3.6. Most members of this

genus require pantothenic and nicotinic acid. Heterofermentative

species require thiamine. The specific nutritional requirements of the

species used here (which may vary by strain) as outlined in Bergey's

Manual of Systematic Bacteriology (De Vos and others 2009) are

shown in Table 2-4.

15 The ninth, a wild isolate from keffir, is of unknown genus.

25

2.3 Grape pomace

Pomace is the material leftover from juice or oil extraction

(Daintith 2000)16. In the manufacture of wine, grape pomace

primarily includes the skins and seeds which are separated from the

Table 2-4 Nutritional requirements of species used in this study

Fermentation Species Requirements

Homofermentative L. bulgaricus Cobalamin

Folic Acid Pantothenic Acid

Thymidine

L. acidophilus Folic Acid

Niacin

Pantothenic Acid Riboflavin

Facultative Heterofermentative

L. paracasei Folic Acid Niacin

Pantothenic Acid Riboflavin

L. rhamnosus Folic Acid

Niacin Pantothenic Acid

Riboflavin

Obligate

Heterofermentative

L. fermentum Niacin

Pantothenic Acid Thiamine

juice either before or after primary fermenation as is the case for white

and red wines respectively (Belitz 2004). Sufficient sugars remain in

the pomace for a second fermentation which can then be distilled

16 Pomace is also the name of a grade of olive oil produced by solvent

extraction.

26

producing grappa or marc17 (McGee 2004). Another biproduct of

pomace is grape seed oil (Pickett 1940) which contains a number of

antioxidants (Wie and others 2009; Xu and others 2010). Grape

pomace extract also contains antioxidants (Wang, Tong and others

2010) which may be of use in the treatment or prevention of diabetes

(Hogan and others 2010) and some cancers (Torres and others 2002;

González-Paramás and others 2004; Cai and others 2010). The use of

grape pomace as a dietary fiber additive has also been proposed

(Valiente and others 1995; Martin-Carron and others 1997; González-

Centeno and others 2010).

The composition of grape pomace varies depending on the

variety of grape used. Bravo and Saura-Calixto (1998), reported the

dry weight of red Cencibel grape skins contains an average of 54.16%

dietary fiber, 26.86% condensed tannins, 14.39% protein, 6.87% fat,

9.19% ash, 2.80% soluble sugars and 3.76% soluble polyphenols18.

For white Airén grape skins, the same group reported 59.04% dietry

fiber, 20.47% condensed tannins, 11.60% protein, 7.78% fat, 6.89%

ash, 2.71% soluble sugars and 4.48% soluble polyphenols. Airén

grape seeds contained 56.17% dietary fiber, 15.96% condensed

17 Grape marc is also a synonym for grape pomace (Cortés and others 2010). 18 Some proteins and condensed tannins are also included in the total dietary

fiber percentage.

27

tannins, 12.23% protein, 12.41% fat, 5.70% soluble sugars and

5.22% soluble polyphenols as percentages of dry weight19. Goni and

others (2005) reported the insoluble fraction of Airén grape skins to

contains 77.2% dietary fiber, 13.9% protein and 3.10% extractable

polyphenols. Seeds were found to contain 78.9% dietary fiber, 13.6%

protein and 3.61% extractable polyphenols (dry weight basis).

2.4 Exopolysaccharides (EPS)

Exopolysaccharides are polysaccharides produced by bacteria.

EPS are found outside the cell wall and may remain close to the cell

wall forming a capsule or detach from the cell to form an extracellular

slime (Sutherland 1977). EPS are composed of repeated polymer

subuits assembled within the cell. A membrane-bound lipid carrier is

used as an assembly base and serves to tether the subunit to the

inside of the cell wall until it is completed and transported to the

exterior of the cell. Once outside, the EPS subunit is attached to

another subunit forming an extended polymer. Laws and others

(2001) provide a more detailed description of this process focusing

specifically on LAB. See Guo and others 2008 review the details of the

three primary pathways used for biosynthesis of most microbial

19 Pomace material used in this and the following study were not fermented.

28

polysaccharides. There are a number of functions attributed to EPS.

It can moderate the effects of of dessicating conditions and protect the

cell from toxins and attack. It also provides the mechanism whereby a

cell can attach to surfaces (De Vuyst and Degeest 1999; Wingender

and others 1999).

2.5 Composition of EPS from LAB used in this study

LAB produces one of two types of EPS: homopolysaccharides and

heteropolysaccharides (De Vuyst and Degeest 1999). The former

group includes glucans and fructans which are each composed of a

single type of monosaccharide. The literature includes descriptions of

strains of four of the species used in this study.

2.6 Fiber isolation and extraction methods

AOAC has approved six methods related to determination of

dietary fiber content (Horwitz and Latimer 2007). These methods are

listed in Table 2-5. All of those listed except 993.2120 are enzymatic-

gravimetric. One of the methods is designed to determine the amount

20 This nonenzymatic method is intended for foods containing more than 10% fiber and less than 2% starch (dry basis) such as fruits, vegetables and fiber

isolates (Horwitz and Latimer 2007).

29

of insoluble dietary fiber (IDF) in a sample. Another is designed to

find the soluble dietary fiber (SDF). The remaining four methods are

intended to determine the total dietary fiber (TDF) in a sample. While

the details vary, all of these methods share a common logic.

To begin with, the sample is dried, ground and weighed. Fat

should be removed if it constitutes a significant fraction (>10%

usually) of the dry sample weight. Enzymatic methods digest the

sample with alpha-amylase, amyloglucosidase and protease to

solublize starch and proteins. Filtration of the sample at this point will

capture the IDF in the retentate which can be washed, dried and

weighed. SDF in the filtrate can be recovered by ethanol precipitation,

filtered, dried and weighed. If proximate analysis is required, the

extracts may be ashed and reweighed.

In addition to the AOAC methods, other methods have been

investigated for soluble fiber extraction and analysis. Theander and

Westerlund (1986) proposed a method in which the sample is first

dried and ground, then sequentially sonicated with ethanol and

petroleum ether to remove simple sugars and lipids. The sample is

rehydrated and treated with amylase and amyloglucosidase to

hydrolyze any starches present and then centrifuged. The solids

30

L. paracasei (cited in DeVuyst and Degeest 1999)

3)--D-GalpNAc-(14)--D-Galp-(16)--D-Galp-(16)--D-Galp-(1 3

sn-glycerol-3-phosphate

L. paracasei (cited in Laws and others 2001)

-D-Glcp 1

6

6)--D-Galp-(13)--L-Rhap-(14)--D-Glcp-(14)--D-GlcpNAc-(1 3

1

-L-Rhap

Glc glucose GlcNAc N-acetyl glucosamine Gal galactose GalNAc N-acetyl galactosamine

Rha rhamnose f furanose

p pyranose

31

L. acidophilus LMG9433 (cited in DeVuyst and Degeest 1999)

-D-GlcpNAc 1

3

4)--D-GlcpA-(16)--D-Glcp-(14)--D-Galp-(14)--D-Glcp-(1

L. acidopilus 5e2 (Laws and others 2008)

-D-Galp 1

6

6)--D-Glcp-(13)--D-Glcp-(13)--D-GlcpNAc-(13)--D-Galp-(1 6

1

-D-Galp-(14)--D-Glcp

32

L. delbrueckii subsp. bulgaricus 291 (cited in Laws and others 2001)

-D-Galp-(14)--D-Glcp 1

6

4)- -D-Galp-(14)--D-Glcp-(14)--D-Glcp-(1

L. delbrueckii subsp. bulgaricus rr (cited in DeVuyst and Degeest 1999)

-D-Galp -D-Galp -L-Rhap 1 1 1

3 4 3

2)--D-Galp-(13)--D-Glcp-(13)--D-Galp-(14)--D-Galp-(1

L. delbruekii ssp bulgaricus LBB.B26 (Sánchez-Medina and others 2007) 1

6

3)--D-Galp-(13)--D-Galp-(14)--D-Glcp-(13)--D-Galf-(1

33

L. rhamnosus RW 9595 M and R (cited in Laws and others 2001)

HOOC 6

R -D-Galp

H3C 4 1

2

3)--L-Rhap-(13)--L-Rhap-(12)--D-Glcp-(13)--L-Rhap-(13)--D-Glcp-(13)--L-Rhap-(1

L. rhamnosus C83 (cited in Laws and others 2001)

6)--D-Galp-(16)--D-Glcp-(13)--D-Galf-(13)--D-Glcp-(12)--D-Galf-(1

L. rhamnosus GG (cited in Laws and others 2001)

-D-Galf 1

6

3)--L-Rhap-(13)--D-Galp-(13)--D-Galf-(13)- -D-Galp(14)--D-GlcfNAc-(1

34

Table 2-5 AOAC Methods for Determination of Dietary Fiber

AOAC 991.42 Insoluble Dietary Fiber in Food and Food Products

AOAC 993.19 Soluble Dietary Fiber in Food and Food Products

AOAC 991.43 Total Soluble and Insoluble Dietary Fiber in Food

AOAC 992.16 Total Dietary Fiber

AOAC 985.29 Total Dietary Fiber in Foods

AOAC 993.21 Total Dietary Fiber in Food and Food Products with

<2% starch

consist of the IDF fraction of the sample. To recover the SDF, the

supernatant is dialysed using filters with a molecular cut-off of 12000-

14000 Daltons. This addition of a protein digestion step to the method

was shown to increase the quantity of SDF recovered (Marlett and

others 1989). The Southgate method provides a more detailed

analysis of a sample with results including free sugars, the hexose,

pentose, uronic acid and starch composition of the soluble fiber

fraction and a similar compositional analysis of the insoluble fiber

fraction including also lignin, cellulose and ash (Southgate 1969). As

noted earlier, these more complete analytical methods were passed

over in favor of a more expeditious method (Prosky 2001).

35

2.7 Structural analysis of fiber

Cellulose, starch and EPS are all polysaccharides. What

differentiates them is the proportion of monosaccharides in their

composition and how those molecules are connected (Nelson and Cox

2005). Hydroxyl groups on the monosaccharide units may be

substituted with structures including acetyl, phosphate, sulphate or

amino groups (Cui 2005; Nelson and Cox 2005). To determine the

monosaccharide composition of a polysaccharide, an inorganic acid is

used to hydrolyze the linkages betweeen monosaccharide units (Pazur

1986; Brummer and Cui 2005). The composition may then be

analyzed using paper or thin-layer chromatography, ion-exchange

(Pazur 1986), GLC or HPLC (Pazur 1986; Brummer and Cui 2005).

The composition and sequence of short repeating polysaccharides

(such as those from bacterial) may also be determined using NMR

(Sletmoen and others 2003).

Once the constituents are established, the next tasks are to

determine the sequence of the components, their linkages and

anomeric configurations and the locations of substitutions (Cui 2005).

Unlike properly formed proteins which have fixed regular structures,

those of polysaccharides can vary. Glycogen includes branches which

occur every 8 to 12 units (Nelson and Cox 2005). Branches in

36

amylopectin occur in clusters rather than at regular intervals (Jane

2007). Linkage patterns are typically determined by methylation of

the free hydroxyl groups in the intact polysaccharide which is then

hydrolyzed and the residues analyzed using GC-MS (Pazur 1986;

Sletmoen and others 2003; Cui 2005). Periodate oxidation in

conjuction with TLC or GC (Pazur 1986) or FAB-MS, MALDI-MS or NMR

(Cui 2005) may be used for sequence determination and to gain

further structural and linkage information. Anomeric configuration

may be determined by enzymatic hydrolysis, chromium trioxide

oxidation (Pazur 1986; Cui 2005) or Smith degradation21 (Cui 2005).

Anomeric configuration may also be determined by NMR analysis of

oligosaccharide fragments of the original polysaccharide (Sletmoen

and others 2003).

2.8 Fourier transform infrared (FTIR) spectroscopy

Spectroscopy is the "study of interaction of electromagnetic

waves and matter" (Banwell 1972). Infrared (IR) spectroscopy is

concerned with the study of electromagnetic waves in the infrared

portion of the spectrum interact with matter. The infrared portion of

the spectrum extends from the near-infrared (NIR) which abuts the

21 Periodation, reduction and mild acid hydrolysis (Cui 2005).

37

visible portion of the spectrum, to the far-infrared (FIR) which is just

below the microwave region of the spectrum (Günzler and Gremlich

2002). FTIR is distinguished from other forms of IR spectroscopy by

its instrumentation. FTIR spectrophotometers use a broadband IR

source, modulated by an interferometer, to collect data in the time

domain (mirror displacement vs intensity). A Fourier transform

algorithm is then used to convert the data to a spectrum in the

frequency domain (wavenumber vs absorbance) (Smith 1996).

Multiple spectra are then averaged to improve the signal to noise

ratio22.

2.8.1 Use of IR spectroscopy in the food industry

IR spectroscopy can be used for both quantitative and qualitative

analysis of food (Stuart 2004).

2.8.1.1 Lipids

The earliest application of IR spectroscopy to food appears to be

Gibson (1920) who reported on a number of invariant peaks observed

22 Earlier types of IR spectrophotometers used optics to select a single monochromatic frequency for sampling. Collection of a complete spectrum

required sampling at each frequency (Smith 1996).

38

in the IR spectra of vegetable oils despite their obvious differences in

oils in visible light. IR spectroscopy can be used to monitor lipid

oxidation and hydrogenation and measure the amounts of cis and

trans fatty acids in a food products (Eskamani 1977). Cis double

bonds produce characteristic peaks at 840 and 770 cm-1 while trans

double bonds show a characteristic peak at 996 cm-1. A further peak

at 1163 cm-1, attributable to the ester bonds between the fat acids and

the glyerol backbone, is used as an internal reference for quantitative

analysis (Stuart 2004). The current method of total trans fat

determination is AOCS Cd 14e-09. This ATR-FTIR23 method measures

the characteristic peak at 966 cm-1 to determine the trans fat content

of a food sample. The method is sufficiently sensitive (>1.8% trans

fat as a percentage of total fat24) to meet current regulatory

requirements (Mossoba and others In Press).

23 Attenuated total reflectance. This technique makes use of the property of

an evanescent wave passing through a wave guide. If a sample is pressed against the outer reflecting surface of the wave guide, the sample will alter the evanescent wave providing information about the sample (Smith 1996). 24 Mossoba and others (In Press) arrived at this figure by dividing the statutory lower reporting limit for trans fats by the fat grams in a serving of

high fat food (ranch dressing).

39

2.8.1.2 Milk

The initial prototypes for an infrared milk analyzer (IRMA) were

introduced in 1964. These machines measured absorptions attributable

to ester bonds, peptide bonds and hydroxyl groups allowing a

determination of fat, protein and lactose contents of liquid milk (Biggs

1972). Barbano and Dellavalle (1987) suggested that milk casein

could also be determined using an IRMA by splitting the sample. One

portion of the sample is tested normally while the second is acidified to

precipitate the casein and the filtrate resampled for protein. The

difference between these two measurements represents the casein

fraction. IRMAs continue in use today, governed by AOAC Method

972.16 (Min and Boff 2003). There has been discussion of a need to

improve the standardization of calibration sample sets (Kaylegian and

others 2006a,b). Questions have also been raised about the

susceptibility of this method to allowing adulteration in particular with

melamine (Moore and others 2010). Monitoring moisture, protein and

fat using IR spectroscopy for on-line quality control in feta cheese

production was demonstrated by Adamopoulos and others (2001).

40

2.8.1.3 Protein

In addition to estimating the protein content milk, the protein in

wheat can be determined based on the intensities of peaks in the

amide I to IV bands. Accuracy is sufficient to distingish between

different type of flour (Eskamani 1977). IR spectroscopy can also

provide information about the secondary structures of proteins.

Carbonaro and Nucara (2010) provide an extensive review.

2.8.1.4 Carbohydrates

IR spectroscopy has been well received by the carbohydrate

chemists. The use of IR for carbohydrate analysis began in 1950

(Eskamani 1977). Kuhn (1950) presented the spectra of cellulose and

a group of sugars along with their derivatives25 as an aid in

identification. Differences in the anomeric forms were also noted.

Whistler and House (1953) determined that while the peaks associated

with the anomeric forms generally fell into bands, the positions and

amplitudes were insufficiently to reliably determine from without a

reference. White and others (1958) presented spectra for

disaccharides and their octaacetates. In 1961, Goulden proposed

25 The paper includes spectra for seventy nine substances.

41

using IR to measure the lactose content of milk which led to the

development of IRMA (Biggs 1972; Eskamani 1977).

IR spectroscopy can be used to monitor gel formation and to

measure the methyoxl content of pectins (Eskamani 1977). Monsoor

and others (2001) used DRIFTS to measure the polygalacturonic acid

content of pectin. Results were comparable to those from colorometric

and HPLC methods. Advantages of this method are minimal sample

preparation. Assifaoui and others (2010) compared the FTIR spectra

of calcium and zinc pectinate films. Ca2+ was found to have greater

interaction than Zn2+ with the functional groups of the pectin leading

to differences in gel formation.

Sucrose production from cane juice can be monitored with IR

spectroscopy by following the extinction of reducing sugars (Eskamani

1977). Cadet and Offmann (1997) proposed us the use of FTIR-ATR to

determine the raw sugar content of cane juice. The spectrum of cane

juice samples were were compared to the spectra of reference samples

using principle component analysis and principle componenet

regression. The method was found to be more accurate than

polarimetric techniques and avoids the need for purification which is

required before polarimetric and HPLC methods. Tewari and Irudayaraj

(2003) proposed a partial least squares predictive model. Input for

42

the model consisted of IR spectra from samples of known sugar

content. Results were again more accurate than polorimetric methods

and obviated the need for purification.

The carbohydrate fractions of soy beans and wheat can be

measured using ATR-FTIR (Eskamani 1977). Goodfellow and Wilson

(1990) used FTIR to track the gelation and retrogradation of starch as

the molecules uncoil and recoil. Amylose and amylopectin were both

observed to go through a brief retrogelation period of less than an

hour. Amylopectin continued through a further process which lasted

up to two weeks.

2.8.2 IR spectroscopy and microbiology

2.8.2.1 Identification of bacterial species

Efforts to differentiate bacteria using infrared spectroscopy have

been ongoing since at least the 1950's (reviewed by Arai and others

1963). The majority of their paper is concerned with developing a IR

spectrographic method for differentiating Nocardia, Mycobacterium and

Streptomyces using lyophilized cells mixed with KBr powder.

Combinations of absorption patterns in four regions of the specta were

then used to differentiate of the test strains to the species level.

43

Amiel and others (2000) proposed using FTIR for the

identification of LAB used in the dairy industry. The intial library was

constructed for use with a two stage algorithm. The first stage, using

a spectral pattern match, identifies the genus of the unknown. The

second stage determines the Mahalanobis distance to determine the

species. The method accurately identified the genus of wild-type

bacterial but was only 69% successful at identifying the species. The

same author also proposed the use of FTIR in addition to genomics for

bacterial classification (Amiel and others 2001).

Whittaker and others (2003) proposed a method of

differentiating bacteria based on the lipid content of their cell walls.

Fatty acids were extracted from the bacterial cell walls and converted

to fatty acid methyl esters (FAME). Spectra were collected and

analyzed using principle component analysis. Both Gram positive and

negative genera and species were correctly differentiated.

Sahu and others (2006) demonstarted the use of FTIR to

distinguish Streptococcus pneumoniae serotypes based on the

bacterias' absorbance spectra between 900 and 1185 cm-1. Cluster

analysis was used to automate the identification.

Dziuba and others (2006) developed a spectral library of LAB.

Pearson's correlation coefficients were calculated for each spectrum

44

which in turn were used to calculate a differentiation index which in

turn was used to distinguish between species. The authors found

incubation time and temperature affect the resuts and recommend

standardized culture conditions. In a subsequent paper (Dziuba and

others 2007a), the authors applied their methodology to the

identification of LAB. By selection of suitable regions, they were in

many cases able to correctly differentiate bacterial strains. They

conclude that a more sophisticated method is required which they

provide in their next paper (Dziuba and others 2007b). Artificial

neural networks were constructed using previously collected spectra26.

It was found that a network trained using complete spectra or trained

using certain combinations of specfic regions could correctly identify

unknowns. Savić and others (2008) describe the use of hierarchical

cluster analysis (HCA), principle component analysis (PCA) and multi-

dimensional scaling (MDS) for the differentiation of Lactobacillus

strains. All three could differentiate the strains used however HCA

would be unsuited to a significantly larger sample set. Based on their

results, the authors conclude that FTIR not a reliable tool for

taxonomic studys.

26 The spectra were segragated into nonoverlapping training, validating and

testing sets.

45

Increasing awareness and mounting losses both human and

economic, have made rapid detection of food-borne pathogens an

important on-going research topic. The use of FT-NIR for detection and

identification of bacteria in liquids was evaluated by Rodriguez-Saona

and others (2001). Principle component analysis permited the species

identification in pure cultures. High cell concentrations of 108 or 109

were required to sufficently cover the filter for detection.

Davis and others (2010) proposed two rapid methods for

detecting E. coli O157:H7 in ground beef: a filter method and an

immunomagnetic separation method. Both methods use a partial least

squares regression model to determine the the pathogen load of a

sample. The detection limit for both methods is 105 CFU/g of sample.

Following a 6 h incubation, the detection was reduced to 10 CFU/g.

2.8.2.2 FTIR and EPS

Verhoef and others (2005) tested the use of FTIR spectroscopy

for EPS sugar compositional analysis. Spectra from EPS isolates were

reduced using partial least squares followed by linear discriminant

analysis. The authors concluded this method could simplify EPS

compositional analysis given a sufficient sample set.

46

Marcotte and others (2007) proposed an FTIR method to

quantify purified EPS samples. Their method is an alternative to

existing colormetric methods which are sensitive to the sugar

composition of the sample Total sugar content is estimated using the

area under the region associated with sugars (970-1182 cm-1)

normalized using either the area under the C-H stretch region (an

estimate of total mass) or the amide II region (an estimate of total

protein). Results were comparable to those obtained from

colorometric assays.

Other constituents can interfere with FTIR quantification of EPS.

Michel and others (2009) reported that the presence of iron and

calcium sulphates lead to an overestimation of the EPS and that it is

difficult to compensate for their effects.

In EPS analysis, FTIR is used to identify functional groups that

make up the material. Wang Y and others (2010), working with EPS

isolated from kefir, noted the probable presence of hydroxyl, amide

and methyl groups along with absense of peaks indicative of glucuronic

acid or diacyl ester. Kodali and others (2009) also found evidence of

hydroxyl groups along with that for carbonyl groups. Information

about the linkage pattern was also derived. This information was

subsequently used to support further analysis by other methods.

47

2.9 Objectives

The objectives of this study were to determine the long term

survival and ability of certain lactic acid bacteria to grow and produce

EPS on grape pomace, and to analyze the EPS extract by Fourier

Transform Infrared (FTIR) spectroscopy.

48

CHAPTER 3

MATERIALS AND METHODS

3.1 Bacterial strains

Bacterial strains used in this study are shown in Table 3-1.

Unless otherwise noted, cultures were prepared by subculturing 0.5 ml

of culture into 10 mL of MRS broth (Difco Labs, St. Louis MO) and

incubating anaerobically at 37C for 48 h. Lactobacillus rhamnosus GG

was used only in the pilot studies.

3.2 Grape pomace

Chambourcin grape pomace was obtained from Les Bourgeois

Vineyards (Rocheport MO). Pomace was removed from the primary

fermenter, pressed dry and stored at -16C until use. The preliminary

studies for this work used pomace of an earlier vintage than that used

for the primary studies.

49

3.3 pH measurement

The pH of cultures was measured using a Corning Model 220 pH

meter (Corning, Corning NY).

Table 3-1 Bacterial strains used.

Strain Source

Lactobacillus paracasei 25598 ATCC

Lactobacillus rhamnosus GG ConAgra

Lactobacillus acidophilus LA-2 Chr. Hansen

Lactobacillus acidophilus ADH Chr. Hansen

Lactobacillus bulgaricus Lb-12 Chr. Hansen

Lactobacillus fermentum 14931 ATCC

Lactobacillus rhamnosus AM-L3 Wild type from poultry litter

Lactobacillus paracasei AM-L4 Wild type from poultry litter

Lactobacillus paracasei AM-L6 Wild type from poultry litter

Kefir Coccus AM-W1 Wild type from kefir

3.4 Preliminary studies

3.4.1 Growth on pomace (broth)

In order to determine if grape pomace is a viable substrate for

lactic acid bacteria, a preliminary growth study was first conducted

50

with broth cultures. Two 40 mL screw cap tubes were filled with 10 g

grape pomace and 20 mL of distilled water. These tubes were

autoclaved at 121C and 15 psig for 15 m and cooled to room

temperature. An additional 10 mL of sterile distilled water was added

to each tube to make up for what was absorbed. The tubes were

inoculated with 1 mL of a 48-h culture of Lactobacillus paracasei or L.

rhamnosus GG and incubated anaerobically at 37C for 48 h.

3.4.2 Growth on pomace (agar)

To confirm the findings from the broth culture test in the

previous section, four types of agar plates were streaked: MRS agar

(Difco Labs), MRS-pomace agar, pomace agar and pH-adjusted

pomace agar. MRS-pomace agar was made from individual

ingredients replacing the dextrose with an equal mass of grape

pomace and adding 0.04 g of bromcresol purple. The recipe for

pomace agar is shown in Table 3-2. The addition of NaOH was

intended to promote the development of a harder agar medium and

shift the pH of the medium above 6.8 into the purple range of the

indicator. Pomace was ground in a commercial-grade blender (Waring

Commercial, Torrington CT) and passed through a 1.4 mm screen

before being added to the agar. The agar plates were streaked with a

51

24-h culture of L. rhamnosus GG, L. bulgaricus LB12, L. paracasei, L.

acidophilus LA-2, L. fermentum, L. acidophilus ADH or Kefir Coccus

AM-W1 and incubated at 37C for 24 h.

Table 3-2 Pomace Agar

Ingredient Amount per L

Grape pomace 20 g

Agar 15 g

Bromcresol purple 0.04 g

Distilled water 1 L

0.1 M NaOH

(pH-adjusted)

48 mL

3.4.3 Further growth and preliminary survival studies

3.4.3.1 Growth of lactic acid bacteria in pomace solution

without pH adjustment

Having demonstrated qualitative growth, the next step was to

collect quantitative data. A set of 40 mL screw capped tubes were

filled with 5 g of grape pomace and 20 mL of distilled water,

autoclaved and cooled to room temperature. The average pH of this

solution was 3.5. Each group was then inoculated with 1 mL of a 24-h

52

culture of L. paracasei, L. fermentum or Kefir Coccus AM-W1. Two

uninoculated tubes were prepared in the same fashion as controls. All

of the tubes were aerobically incubated at 23C. Tubes were removed

on days 0, 3, 6 and 9 for plating and enumeration. The controls were

plated on days 0 and 9. After plating, the pH of each suspension was

measured. The experiment was repeated using seven species (L.

paracasei, L. fermentum, Kefir Coccus AM-W1, L. bulgaricus LB12, L.

acidophilus LA-2, L. acidophilus ADH or L. rhamnosus LGG) and

removing tubes for plating on days 0, 11, 33 and 83.

3.4.3.2 Growth of lactic acid bacteria in pomace solution with

pH adjustment

In order to determine if increasing the pH of the grape pomace

solution to near neutral would result in improved bacterial growth, the

experiment was repeated using seven strains. To determine the

quantity of base required, aliquots of 1 M NaOH were added to one of

the tubes until the pH exceeded neutral. This volume, 600 µL 1 M

NaOH, was then added to the remaining tubes before autoclaving.

Plating was done on days 0, 3 and 6. Autoclaving the suspension in a

pH of 4. Therefore, to determine how much base was required to

produce a near neutral pH after autoclaving, 40 mL tubes were filled

53

with 5 g of pomace, 30 mL of distilled water and aliquots of 4 M NaOH

ranging from 350 to 600 µL. The tubes were then autoclaved and

cooled. Using these results, the experiment was again repeated with a

single strain, L. paracasei. An aliquot of 575 µL of 4 M NaOH was

added to each tube before autoclaving. Plating was done on days 0, 3,

6 and 9.

3.5 Primary studies

The results from the preliminary studies demonstrated that

grape pomace can be used as substrate for the growth of lactic acid

bacteria (LAB). They also led us to conclude that the pH of the pomace

suspension has an important effect on the growth of LAB which was

further investigated in the primary study described below. The results

also led to the question of how long the LAB can survive in these

environments.

3.5.1 pH Adjustment

Studies were done both with and without pH adjustment. A later

vintage of grape pomace was used in the primary studies requiring a

recalibration of the quantity of base needed to adjust the pH to

approximately 6, the pH of MRS broth (Difco 1998). The quantity of

54

NaOH used was chosen by preparing six 40 mL tubes with 5 g of

pomace, 30 mL of distilled water and between 275 and 525 µL of 4 M

NaOH (Fisher, Fairlawn NJ). The tubes were autoclaved and their pH

measured after cooling to room temperature. Based on the results of

this test, each container in pH-adjusted replicates was treated with

350 µL (Survival Study) or 3.5 mL (Exopolysaccharide Production

Study) of 4 M NaOH.

3.5.2 Survival study

A group of ten 40 mL test tubes was prepared for each strain.

Each tube contained 30 mL deionized water and 5 g grape pomace. In

addition, pH-adjusted tests received 350 µL of 4 M NaOH. Tubes were

autoclaved, cooled overnight to room temperature, inoculated with

500 µL of a 48-h culture and incubated aerobically at 23C. Tubes

were removed on days 0, 1, 2 and 3 and periodically thereafter for

plating and enumeration. The pH of each tube was also measured

after plating. Samples were plated by serial dilution in 0.1% peptone

water (Difco Labs, Sparks MD) and pour-plated with MRS agar (Difco

Labs). Plates were anaerobically incubated at 37C for 48 h and

colonies enumerated.

55

3.5.3 Exopolysaccharide (EPS) production study

3.5.3.1 Sample preparation

A 375 mL jar containing 300 mL deionized water and 50 g grape

pomace was prepared for each strain. In addition, pH-adjusted tests

received 3.5 mL of 4 M NaOH. Tubes were autoclaved, cooled

overnight to room temperature, inoculated with 5 mL of a 48-h culture

and incubated aerobically at 23C for 4 d. Jars were them immediately

transferred to a 90C water bath and held there for 60 min. The jars

were stored at 4C prior to extraction.

3.5.3.2 Soluble fiber extraction

A modified version of AOAC Method 991.43 (Horwitz and Latimer

2007) was used to extract the soluble fiber fraction from the samples

(Figure 3-1). Solids were removed by filtering through a Büchner

funnel fitted with a 15 cm Whatman #3 filter (Whatman, Maidenstone

Kent UK) and 3 g of Celite 421 (Aldrich, Milwaukee WI) into a 1 L

vacuum flask (a). The retentate was rinsed twice with ~300 mL of

deionized water. The filtrate was then quantitatively transferred to a 1

L beaker and boiled down to approximately 100 mL (b). While still

hot, approximately 300 mL of 95% ethanol was added to the boiled

filtrate (c). This mixture was then filtered through a Büchner funnel

56

fitted with a 7 cm Whatman #2 filter and 1 g of Celite 421 (d). The

retentate was rinsed three times with 70% ethanol. The retentate and

filter aid were then redissolved/resuspended in approximately 250 mL

of deionized water and filtered through a Büchner funnel fitted with a 7

cm Whatman #2 filter (e, f). The retentate was rinsed three times

with deionized water and the filtrate was then quantitatively

transferred to a 500 mL beaker and boiled down to ~50mL (g). This

sample was then quantitatively transferred to a 100 mL beaker,

covered with foil and cooled to -18C.

3.5.3.3 Freeze-drying

Frozen samples from above (3.5.3.2) were packed in groups of

four into 2 L freeze-drying flasks (Labconco, Kansas City MO) and

dried for 5 d using a Labconco Lyophlock 12 Freeze Dry System.

Samples were then weighed and transferred to sterile 50 mL

centrifuge tubes and stored at room temperature.

3.5.4 Fourier transform infrared spectroscopy

Infrared spectra were collected using a Nicolete 380 FTIR

spectrophotometer (Thermo Scientific, Waltham MA) equipped with a

57

VeeMAX II spectral reflectance attachment (Pike Technologies,

Madison WI) fitted with a 5/8 inch mask plate. Following trial and

error testing, the angle of incidence was set to 50. Data collection

was done using OMNIC, a software package supplied by the

spectrometer manufacturer.

3.5.5 Sample preparation

Samples were prepared by making a 1% solution (w/v) of

extracted material in deionized water. Beads of liquid (200 µL each)

were then pipetted onto gold-coated slides (Biogold – Gold Coated

Microarray Substrates, Themo Scientific). Six beads were pipetted for

each sample with three beads per slide (Figure 3-2).

3.5.6 Sample collection

Following an initial warm-up period of at least 15 min, a

background spectrum was collected using a clean gold-coated slide

placed directly on the mask. Slides were fitted in a simple cardstock

frame to prevent the sample from contacting the mask. Spectrum

collection parameters were 64 scans and 2 cm-1.

58

Figure 3-1 AOAC method 991.43 as modified for this study. Details of

each step are provided in section 3.5.3.2.

Boil down to 100 mL

Precipitate with Ethanol

Redissolve/resuspend retentate

Filter

Filter

Filter

Boil down to 50 mL

(a)

(b)

(c)

(d)

(e)

(f)

(g)

59

3.5.7 Growth and EPS production in tomato juice

The extract from the samples prepared in section 3.5.3.1

contained components derived from both the bacteria and the

substrate. In order to minimize the latter, another set of samples was

prepared using tomato juice. Tomato juice was chosen for two

reasons: it has long been recognized as a suitable substrate for LAB

(Kulp 1927) and because it has a low fiber content27. A 20% solution

of tomato juice (v/v) was made by combining 200 mL of canned

tomato juice from concentrate (obtained locally) in a 1 L graduated

cylinder with sufficient distilled water to make up the volume. Glass

27 The "Nutrition Facts" panel on the can indicates that a 240 mL serving

contains 2 g of dietary fiber.

Figure 3-2 Samples on a slide.

60

jars (375 mL capacity) were filled with 30 mL aliquots of this solution

and then handled as described in the preceeding sections. There were

no pH-adjusted replicates using tomato juice.

3.6 Statistical and spectral analyses

Regression and Honest significant difference were calculated

using R (R Development Core Team 2008). Scripts for data

normalization, Savitzky-Golay smoothing and derivation (Savitzky and

Golay 1964, Steinier and others 1972) and other transformations were

written by the author in Perl (Wall and others 1996). Source code for

these scripts is provided in APPENDIX C. Plotting was done using

Microsoft Excel (Redmond WA) and R (R Development Core Team

2008).

61

CHAPTER 4

RESULTS

4.1 Preliminary Studies


The tubes for L. paracasei and L. rhamnosus GG both showed

positive signs of growth. White precipitation / flocculation was

observed in both tubes.


As expected, all species grew well on MRS agar (Table 4-1). All

species also grew well on MRS-Pomace (P) agar, but the indicator

remained purple (pH > 5.2), indicating insufficient fermentation of the

pomace. None of the species grew on Agar+P, and the low pH kept

the indicator in the yellow range and made the agar soft and difficult

to streak. All species grew on pH adjusted Agar+P, but the color

remained purple. Colonies for all species were small, round and

colorless.

62

Table 4-1 Growth of lactic acid bacteria on pomace (agar)

MRS MRS-P Agar+P

Unadjusted pH

Agar+P

Adjusted pH

L. rhamnosus GG + + - +

L. bulgaricus Lb-12 + + - +

L. paracasei + + - +

L. acidophilus LA-2 + + - +

L. fermentum + + - +

L. acidophilus ADH + + - +

Kefir Coccus AM-W1 + + - +


4.1.3.1 Growth of lactic acid bacteria in pomace solution without pH adjustment

The LAB were grown in pomace solution without adjusted pH and

the results of the 9 day trial are shown in Figure 4-1. Counts for L.

paracasei, L. fermentum and AM-W1 had an average range28 of 0.39

log(CFU/mL) (SD=0.15). The average pH range over this same period

was 0.25 (SD=0.06). Low and high pH values for L. paracasei were

3.6 and 3.78 while those for L. fermentum and AM-W1 were 3.4 and

3.69. The results of the 83-day trial are shown in Figure 4-2. The

28

Average Range

max speciesi min speciesi i1

n

n

63

average range of counts was 1.92 log (CFU/mL) (SD=1.05).

Individual growth ranges are shown in Figure 4-2. The largest change

was observed with L. bulgaricus LB-12 whose population declined by

nearly four log cycles. L. acidophilus, Kefir coccus AM-W1 and L.

rhamnosus GG each declined by approximately two log cycles. L.

fermentum and L. acidophilus ADH declined by 1.28 and 1.42 log

cycles, respectively. L. paracasei showed the least change at 0.58 log.

The average pH range for this same period was 0.42 (SD=0.08). All

species exhibited a pH increase over the course of the experiment.

Start and ending pH of the controls were 3.68 and 3.62, respectively.

No colonies were observed when either tube was plated.


pH adjustment

The results of our initial attempt to adjust the pH of the pomace

solution are shown in Figure 4-3. The pH and colony counts remained

essentially unchanged for the duration of the experiment. The

average colony count range was 0.27 CFU/mL (SD=0.20) and the

average pH range was 0.13 (SD=0.10). The reader will recall that the

pH of all the tubes was adjusted to near neutral before autoclaving,

64

yet the average pH during the course of the experiment was 3.98

(SD=0.11). To compensate for this drop in pH, increasing quantities

Figure 4-1 Growth of LAB on pomace (9-day trial) of NaOH were added, producing the result shown in Figure 4-4.

Addition of each aliquot of base resulted in an incremental increase in

pH except the last between 550 and 600 µL which resulted in a sharp

increase.

65

Figure 4-2 Growth of LAB on pomace (83-day trial)

66

Table 4-2 Growth ranges (CFU/mL) of LAB in pomace solution for 83 d.

Species Range (CFU/mL)

L. paracasei 0.58

L. bulgaricus LB-12 3.94

L. acidophilus LA-2 2.03

L. fermentum 1.28

AM-W1 2.04

L. acidophilus ADH 1.42

L. rhamnosus GG 2.15

The results of the final preliminary pH adjusted experiment are

shown in Figure 4-5. The colony counts had a range of 0.31 CFU/mL.

Counts for L. paracasei rose from 7.81 to 8.11 CFU/mL between days

0 and 6 before falling back down to 7.87 CFU/mL on day 9. No

colonies were observed on the control plates. The pH range was 0.96

beginning at 6.43 and ending at 5.53. The pH of the beginning and

ending control tubes was 6.46 and 7.03, respectively.

67

Figure 4-3 Growth of LAB on pH adjusted Pomace

68

Figure 4-4 Adjusted pH resulting from addition of 4M NaOH and

subsequent autoclaving

Figure 4-5 Growth of L. paracasei on pH adjusted grape pomace

69

4.2 Primary Studies

4.2.1 pH adjustment

Results of pH adjustment calibration for the later vintage of

grape pomace used in the primary studies are shown in Figure 4-6.

Figure 4-6 Adjusted pH resulting from addition of 4 M NaOH to second

batch of pomace and subsequent autoclaving

4.2.2 Survival Study

Average survival and pH curves for adjusted pH replicated are

shown in Figure 4-7 and Figure 4-8. Curves for unadjusted replicates

are shown in Figure 4-9 and Figure 4-10. Missing values were

replaced by linear interpolation between two known points29. At the

outset, the pH-adjusted replicates of L. acidophilus LA-2 began an

29

y b ya yc ya xc xb

xc xa

where a b c

70

immediate steady decline. Other organisms show a slight increase

before declining in numbers. Kefir Coccus AM-W1, L. bulgaricus Lb-12

and L. paracasei AM-L4 rose by approximately 0.25 log CFU before

declining in numbers while L. rhamnosus AM-L3 and L. paracasei AM-

L6 rose by approximately 0.15 log and L. fermentum by 0.1 log CFU.

In most of these cases, the peak of the rise occurred on day 2.

However, the peak for L. AM-L6 came on day 1 while Kefir Coccus AM-

W1 did not peak until day 3. The remaining pH-adjusted replicates

show a pattern of decline, rebound and decline. In the case of L.

paracasei 25598, counts dropped by approximately 0.25 log, regained

most of the loss (~0.2 log) before following the overall pattern of

decline. The counts for L. acidophilus ADH dropped ~0.6 log, held

steady for two days before rising ~0.2 log and then declining. The pH

of these replicates declined by an average of 0.8 over the course of

this experiment. L. bulgaricus Lb-12, Kefir Coccus AM-W1 and L.

rhamnosus AM-L3 had the largest pH declines of 1.2, 1.1 and 1.0

respectively. The pH of L. fermentum declined by only 0.3. The

remaining strains fell by 0.9 (L. acidophilus LA-2), 0.8 (L. paracasei

AM-L4) and 0.7 (L. paracasei 25598, La ADH and L. paracasei AM-L6).

All replicates had an initial pH drop of ~0.2 (L. paracasei AM-L6) to

~0.6 (L. acidophilus LA-2 and Kefir Coccus AM-W1) within this first few

71

days. The pH of some replicates (L. acidophilus LA-2, L. bulgaricus Lb-

12, L. paracasei AM-L4 and L. paracasei AM-L6) continued a more

gradual decline for the remainder of the experiment. Others, (L.

paracasei 25598, L. acidophilus ADH, Kefir Coccus AM-W1 and L.

rhamnosus AM-L3) rose slightly before beginning the final descent.

The remaining replicate, L. fermentum, rebounded after the initial drop

before settling in to a relatively constant pH of ~5 for the duration.

In replicates for which the initial pH was not adjusted, counts

again declined as time progressed. Counts for L. paracasei 25598, L.

acidophilus LA-2, L. acidophilus ADH and Kefir Coccus AM-W1 all

began an immediate decline. The counts for L. bulgaricus Lb-12, L.

rhamnosus AM-L3 and L. paracasei AM-L6 exhibited a slight pattern of

decline, rebound and decline with an amplitude of 0.2 log. L. paracasei

AM-L4 exhibited a more extreme oscillation, rising and falling by

roughly 0.5 log before settling into a stable fixed population. The

counts for L. fermentum follow an initial pattern of 0.5 log decline over

the first three days. Subsequent counts a month later were one log

cycle higher (0.5 log higher than the initial population). No

contamination was observed in the tubes nor was there anything

remarkable about the plated colonies. The average pH of all strains in

this group of replicates rose by 0.1 to 0.3 in the first two days followed

72

by a decline of similar magnitude. The average pH then remained

stable for the balance of the experiment. No unadjusted replicate

experiments were performed for Kefir Coccus AM-W1, L. acidophilus

Lb-12, or L. paracasei AM-L4. Data for the first replicate is presented

here but will not be included in further discussion.

The average slopes of the survivor and pH regression lines are

shown in Figure 4-11 and Figure 4-12 respectively30. In general,

colony counts declined over time with counts declining more slowly

when the pH was initially adjusted upward. The pH for these adjusted

experiments subsequently declined over time. In those experiments

where the pH was not adjusted, an increase in pH over time was

observed similar to that in the preliminary studies. Two-way ANOVA

was used to test for differences in the survivor and pH slopes. For the

population slopes we used the linear model

CFU ~ species adj species* adj31

where:

CFU = slope of the population regression line for a given

replicate

species = the strain used to inoculate the replicate

30 Graphs of the regression lines underlying this section along with the R

scripts used to analyse the data are presented in APPENDIX A. 31 The "*" denotes an interaction term between independent variables.

73

adj = was this replicate pH adjusted.

ANOVA indicated the presence of significant differences in both

independent variables (species: p=2.9e-5, adj: p=4.8e-6) and the

interaction term (p=7.00e-7). We then used Tukey's HSD to locate

these differences, the results of which are available in appendix

APPENDIX A. This same analysis was performed using the linear

model

pH ~ species adj species* adj

where:

pH = slope of the pH regression line for a given replicate

A significant difference was found between the levels of adj

(p=1.6e-6).

ANOVA was also applied to the pH-adjusted and unadjusted

subsets of the data. Significant differences were found between the

species in both subsets of the survior slopes. Tukey's HSD was used

to locate the differences. No differences were found within either

subset of the pH slope data.

74

Figure 4-7 Average survival curve (adjusted pH)

75

Figure 4-8 pH (pH adusted)

76

Figure 4-9 Average survival curves (unadjusted pH)

77

Figure 4-10 pH (unadjusted)

78

Figure 4-11 Average slope of survivor count regression lines

79

Figure 4-12 Average slope of pH regression lines

4.2.3 Exopolysaccharide (EPS) production study 4.2.3.1 Soluble fiber extraction

The quantities of material recovered after extraction and freeze-

drying are presented in Figure 4-13. The values shown are averages

of all pH adjusted or unadjusted replicates for a given species. Within

the unadjusted samples, the largest average was 1.74 g [SD=0.29,

n=2], recovered from Kefir Coccus AM-W1. The minimum recovered

80

was 1.21 g [SD=0.19, n=3] from L. paracasei. The remaining strains

fell between these values (L. acidopilus LA-2=1.50 g [SD=0.20, n=3],

L. fermentum=1.43 g [SD=0.62, n=3], L. acidopilus ADH=1.48 g

[SD=0.01, n=2], L. bulgaricus Lb12=1.34 g [0.63, n=2], L.

rhamnosus AM-L3=1.36 g [0.44, n=3], L. paracasei AM-L4=2.02 g32,

L. paracasei AM-L6=1.52 g [SD=0.56,n=3]). Samples containing only

water and pomace averaged 1.19 g [SD=0.31, n=3].

Taken by species, the average material recovered from adjusted

pH samples was less than that for unadjusted samples. The largest

quantity, 1.40 g [SD=0.54, n=2], was recovered from L. paracasei

AM-L4. The smallest, 0.88 g [SD=0.22, n=2], was recovered from L.

acidophilus ADH. The remaining strains again fall between these

values (L. paracasei=1.03 g [SD=0.38, n=3], L. acidopilus La2=1.16 g

[SD=0.30, n=3], L. fermentum=1.35 g [SD=0.45, n=2], Kefir Coccus

AM-W1=1.21 g [SD=0.51, n=2], L. bulgaricus Lb12=1.07 g [SD=0.12,

n=2], L. rhamnosus AM-L3=0.92 g [SD=0.45, n=3], L. paracasei AM-

L6=0.91 g [SD=0.21,n=3]). Samples containing only water, pomace

and NaOH averaged 1.00 g [SD=0.23, n=5].

32 The replicate was lost during extraction;the value given here is the result of extracting a single sample and is intended for reference only and will not

be used in further calculations.

81

ANOVA analysis of the extraction results using the linear model

extract ~ species adj species* adj

where:

extract = grams of extracted material

species = the strain used to inoculate the replicate

adj = was this replicate pH adjusted.

indicates a significant difference beween samples with and without

initial pH adjustment (p=1.8e-3) which we confirmed using Tukey's

HSD.


The spectra discussed in this paper are presented in APPENDIX

B. The spectra in Figure B-1 were collected from pH-adjusted extract

samples while those in Figure B-2 were collected from unadjust

samples. Potential of potential interest were identified by Peaks, an R

package which implements the algorithm described by Mariscotti

(1967). Peaks present in both the treatment and control samples

were assumed to arise from the substrate and were ignored. A

parameter, "Q", was used to specify the minimum required distance

between peaks before they would be considered different. In the

figures, Q=3 indicates a minimum peak spacing of 2.8929/cm.

82

Figure 4-13 Average material recovered after extraction and freeze-

drying

4.2.5 Growth and EPS production in tomato juice

Results for the extraction and freeze-drying of samples grown in

tomato juice are shown in Table 4-3. IR spectra for these samples

are presented in APPENDIX B.

83

Table 4-3 Extraction results after growth in diluted tomato juice.

Species grams

L. fermentum 0.21

L. acidophilus La-2 0.16

L. acidophilus ADH 0.20

L. rhamnosus AM-L3 0.11

L. paracasei AM-L6 0.16

control 0.22

84

CHAPTER 5

DISCUSSION

5.1 Preliminary studies


LAB are used to ferment a wide variety of plant materials for

foodstuffs (Steinkraus 1996b). They are also responsible for the

malolactic fermentation of wine (Fleet 1998) and they make important

contributions to the production of silage (Woolford and Pahlow 1998).

And silage can be made from grape pomace (Spanghero and others

2009). With this in mind, it is reasonable to predict that grape

pomace will make a suitable substrate for LAB.

The white precipitate observed in our initial experiment was

assumed to be clumps of bacterial cells settling out of suspension. This

behavior is consistent with our experience growing these species in a

broth culture.

85


To confirm our conclusion from the previous experiment, agar

plates were made up and streaked with bacteria. MRS agar was

developed specifically for the laboratory culture of LAB (De Man and

others 1960). Failure of the LAB to grow on this medium would have

been quite surprising. By substituting pomace for the dextrose in the

formulation, the presence of fermentable sugars available in grape

pomace could be determined. The pH of plates of both the MRS-

Pomace agar and the pH adjusted Agar+P remained above 5.2 after

the formation of colonies leading to the conclusion that the sugars

available in grape pomace for fermentation by LAB are limited and/or

not readily metabolized. In lieu of available sugars to ferment, the

bacteria have likely shifted to utilizing organic acids (Fleet 1998).

Grape pomace may provide other essential growth factors or these

may be supplied by autolyzed yeast cells left over from the primary

fermentation. The use of (pH adjusted) Agar+P was intended to

determine whether pomace alone could supply the required nutrients

for the culturing of LAB.

In the results section, we report the negative growth of LAB on

Agar+P which, given the preceding and following sections, may seem a

bit odd. It is quite possible that there really was no growth. It is

86

presumed, however, that the lack of observed colonies was due to the

difficulties of streaking onto very soft agar. The combination of low pH

and autoclaving probably hydrolyzed the media and prevented it from

setting properly (Khalid and others 1993). The loop dug into the agar

surface. If any colorless colonies such as those observed on the other

plates were present, they would have been hidden under the agar

surface. The formation of visible colonies indicates the likely absence

of growth inhibitors in the pomace.


5.1.3.1 Growth of lactic acid bacteria in pomace solution

without pH adjustment

The results from this section are substantially in line with those

of the primary studies, discussed below. At this point in the

discussion, it is sufficient to note that the bacteria in this pair of

experiments appear to be in the stationary (LP) or death phases of the

growth curve. There is no initial rise in number before the decline sets

in. These results also indicate an increase in pH over time which,

given the normal habits of LAB, is counterintuitive. Under normal

circumstances, LAB are acid producers who lower the pH of their

87

environment. Bacteria can and will attempt to adapt their

environment to best suit their needs (Jay and others 2005).

In previous experiments, cultures were incubated anaerobically

at 37C which is generally recognized as optimum conditions for LAB

growth (De Vos and others 2009). The optimum conditions for EPS

production are cooler De Vuyst and Degeest 1999; Laws and others

2001). The lower temperature and aerobic conditions were also

selected for practical reasons; incubators large enough to use the

findings of this work on a commercial scale would be quite expensive.


pH adjustment

Raising the pH of the pomace mixture to near neutral before

autoclaving produced an increase in pH of 0.5 after autoclaving. The

observed patterns of growth and pH were sufficiently similar to the

previous attempt that the experiment was abandoned after 6 d. We

theorized at the time that the pH change was due to the difference

between titratable and total acidity with the heat from autoclaving

forcing the reaction to proceed. However, based on work by Nef

88

(1914), the more likely explanation is the dissociation of glycosidic

bonds when heated in the presence of a strong base33.

The pH increase of 3.0 produced a remarkably different pattern

of growth. The bacterial population remained fairly constant but unlike

our previous experiments, the pH of the mixture fell with time. This

decrease is unlikely to be due to unreacted hydroxide ions as a similar

decline was not observed in the controls.

5.2 Primary studies

5.2.1 Survival study

During the first 51 days of the experiment, the average colony

counts for most strains declined. On average, pH adjusted replicates

declined by 0.8 log CFU/mL (SD=0.36) while unadjusted replicates

declined by 1.3 (SD=0.69). A 10% remaining viability after roughly 9

weeks compares favorably with previously published results. Hull and

others (1984) reported >15% viability of L. acidophilus in yogurt after

4 weeks. art ne -Villaluenga and others (2006) report one and two

log reductions of L. acidophilus La-5 after 21 days in fermented milk

stored at 4˚C, respectively with or without the addition of raffinose

family oligosaccharides. Buriti and others (2010) report 1 to 3 log

33 Our discovery of this paper is a classic example of the Gruen Effect

wherein an important reference is found after the experiment is completed.

89

reductions over the course of 28 days of La-5 in guava mousse stored

at 4˚C.

Also of interest are the early reactions of the various strains.

Some L. paracasei, L. acidophilus ADH and Kefir coccus AM-W1 in the

unadjusted replicates and L. acidophilus LA-2 in all replicates, simply

declined in population with no apparent growth. These combinations

of media and environment were apparently unsuitable for the growth

of these bacteria. The counts for other pH-adjusted replicates, Kefir

coccus AM-W1, L. bulgaricus Lb-12, L. fermentum, Lactobacillus

rhamnosus AM-L3, Lactobacillus paracasei AM-L4 and Lactobacillus

paracasei AM-L6, increased before declining indicating that they

followed some approximation of a typical growth curve. The data

collection interval is too coarse to determine whether there is an initial

lag phase. The remaining pH-adjusted replicates, Lactobacillus

paracasei and Lactobacillus acidophilus ADH, and a majority of

remaining unadjusted replicates, Lactobacillus bulgaricus Lb-12,

Lactobacillus rhamnosus AM-L3, Lactobacillus paracasei AM-L4 and

Lactobacillus paracasei AM-L6, show a pattern of initial decline,

rebound and then decline, all over a period of three days. We

speculate this pattern is a lag phase possibly prolonged by the low

growth temperature. The apparent absence of a linear growth region

90

between the log and decline phases is probably due to coarse spacing

of data points relative to the phenomenon. A different effect was

observed with L. fermentum. For the first three days, the average

counts plummeted by ~0.5 log CFU. Yet when the next points were

collected a month later, the counts had not only recovered but

exceeded the initial counts. In addition, the trajectory of the counts

suggests an extended linear growth region which rolled into the

decline phase between 50 and 80 days (data not shown). In short, L.

fermentum liked these conditions.

In the case of average pH results of pH-adjusted replicates, the

pH drop was likely due to the fermentation of sugars by the bacteria.

The comparatively steep initial drop followed by a more gradual tail is

probably a matter of numbers. The largest numbers of bacteria in

each tube were present at the outset of the experiment. As the total

number of organisms declines, their ability to affect large-scale

changes on their environment also declines. The subsequent pH rise

observed in many of the adjusted pH replicates and all of the

unadjusted replicates is possibly due to malolactic fermentation, a

process which converts L-malic acid to L-lactic acid thereby raising the

pH by 0.3 to 0.5 (Fleet 1998). Decarboxylation of malate yields

lactate and H+ which are removed from the cell by a malate/lactate

91

antiport. The charge gradient across the cell membrane is increased,

permitting the assembly ATP (Poolman and others 1991). The rate of

malate transfer is pH dependent: lower pH results in increased malate

transport (Olsen and others 1991). This is a two-edged sword

however. At low pH, acids lose their charge and, as neutral molecules,

can pass through the cell membrane (Guzzo and Desroche 2009).

Once in the cytosol, the acid dissociates resulting in the denaturation

of proteins and inhibition of enzymes (Jay and others 2005).

The notion that two very different growth regimes are at play

here is supported by our finding of a significant difference between the

slopes of pH adjusted replicates (negative slope) and unadjusted

replicates (positive slopes). The sign of the slopes is also consistent

with unadjusted replicates making use of acid fermentation. The

ANOVA results for the slopes of the survivor curves indicate two

significant differences in the data. There is a statistically significant

difference between the adjusted / unadjusted replicates of AML6 and

those for L. fermentum. For, AML6, the difference is one of degree -

the viable populations of the unadjusted replicates declined more

quickly than those for which the pH was initially adjusted. However, in

the case of L. fermentum a completely different scenario is observed.

The population trends of the adjusted/unadjusted replicates are in

92

opposition; the adjusted replicates are in decline while the unadjusted

replicates are growing.

The growth of L. fermentum in unadjusted pomace may be due

to a number of factors. Some strains of this species have very high

tolerance for acidic condition. Bao and others (2010) reported 35 of

90 strains of L. fermentum tested grew over a period of 72 h at pH

3.0. Pereira and Gibson (2002) reported two strains capable of

surviving at pH 2. L. fermentum F53 declined approximately 2.5 log

over the course of 2 h while L. fermentum KC5b declined by less than

0.2 log over the same time period. Another factor may be the ability

by L. fermentum to reduce fructose to mannitol (Buchanan and

Gibbons 1974). Heterofermentative species such as L. fermentum,

use this pathway to oxidize NADH. Fructose is used as the final

electron acceptor (Vrancken and others 2008). None of the other

heterofermentative species used in this study reduce fructose to

mannitol.

The ANOVA results for subsets of the slopes of the survivor

regression lines partition both subsets into three groups. In the pH-

adjusted subset, each group spans at least 6 of the 9 species leading

us to wonder if a third replicate might not erase the difference. It is

interesting to note that all three L. paracasei strains are clustered

93

together. In the unadjusted pH subset, Lactobacillus paracasei AM-L4

and Lactobacillus fermentum. The behavior of Lactobacillus

fermentum has been discussed above. And as there is no replicate

data for Lactobacillus paracasei AM-L4, the observed behavior may be

an anomaly.

5.2.2 Exopolysaccharide (EPS) production study

5.2.2.1 Soluble fiber extraction

Based on the ANOVA results, pH adjustment has a significant

influence on the quantity of material recovered from the samples. Less

material was recovered from those samples in which the initial pH was

adjusted. This is consistent with Nef's (1914) findings of the

dissociation of polysaccharides when heated in the presence of a

strong base.

We now face the question as to how much of the material

extracted is derived from pomace and how much is derived from

bacterial action. To our knowledge there are no reports in the

literature specific to chambourcin grapes. There are reports for other

varietals. The soluble fiber fraction of Airèn grape pomace, a white

varietal grown in Spain, is 9.53% of dry weight (Valiente and others

1995). And 14.5% of the dry weight of Manto Negro grape pomace, a

94

red Spanish varietal, is made up of soluble fiber (Llobera and Cañellas

2007). Bravo and Saura-Calixto (1998), report soluble fiber of 4.43%

for Airèn grape skins, 3.23% for Airèn grape seeds and 2.34% for red

Cencibel grape skins. Using the soluble fiber percentage of the Airèn

grape seeds for both red and white, the total soluble fiber for these

pomaces is 7.66% and 5.57%. Based on this admittedly small

sample, we estimate chambourcin should contain approximately 9.3%

soluble fiber or 4.7 g / jar34. Clearly our results fall short of this bar.

Autoclaving was ruled out as a cause; the quantity of material

recovered from unautoclaved material falls within the range of results

for autoclaved materials (data not shown). The discrepancy is likely

due to differences in sample preparation and purpose. The studies

cited were interested in the composition of grape pomace as an end

and ground their samples before analysis. Pomace samples used in

our study were mostly intact, and as such, a large fraction of the

soluble fiber in the pomace remained unextracted. If we ignore the

fraction of soluble fiber contributed by the seeds, and use only the

data from Bravo and Saura-Calixto (1998), we estimate a soluble fiber

content of 3.4% or 1.7 g / jar.

34 Each jar contains 50 g of grape pomace and 300 mL of water.

95

Published work places the quantity of EPS produced anywhere

between 25 and 9800 mg/L depending on strain, growth conditions

and extraction method (reviewed by Behare and others 2009)35. Using

the median value from this dataset, 540 mg/L, we expect to recover

162 mg / jar or roughly one tenth of the expected soluble fiber

recovered from the substrate. Clearly any EPS recovered will not be a

pure isolate.

Returning to the bar chart, we assume samples whose standard

error36 overlaps the standard error of the corresponding pomace-only

sample will only contain material derived from the grape pomace. This

assumption implies that none of the pH adjusted samples will likely

contain appreciable quantities of EPS. Of the unadjusted samples,

Lactobacillus acidophilus LA-2 and Kefir Coccus AM-W1 are the likely

prospects for finding EPS.


In some instances it is possible to show the presence of or

quantify a substance in an FTIR spectrum despite of background noise

(Cadet and Offmann 1997). Using the peak assignments given in

35 The average reported yield was 1206 mg/L (SD=2339). 36

SE

N

96

Wang, Li and others (2010), the tentative peak assignments shown in

Table 5-1 can be made.

Both the pH-adjusted and unadjusted spectra of Lactobacillus

paracasei and Lactobacillus paracasei AM-L6 have peaks near 1543

cm-1 which Wang, Li and others (2010) ascribe to N-H bending of

proteins. The pH-adjusted spectra of Lactobacillus acidophilus LA-2

and Lactobacillus rhamnosus AM-L3 and the unadjusted spectra of

Lactobacillus fermentum and Kefir Coccus AM-W1 also have peaks

near 1543 cm-1.

Table 5-1 Peak Assignments

Wavenumber

(cm-1)

Substrate Attributed to

Pomace Tomato

pH-adjusted no pH

adjustment

1543 L. paracasei L. paracasei Bending of N-H

in proteins L. paracasei

AM-L6

L. paracasei

AM-L6

L. acidophilus

La-2

L. fermentum

L. rhamnosus

AM-L3

AM-W1

1448 L. paracasei

AM-L6

L. paracasei

AM-L6

L. paracasei

AM-L6

Asymetric

bending of C-

H2 and C-H3

groups in

proteins

L. bulgaricus

LB-12

Kefir Coccus

AM-W1

L. rhamnosus

AM-L3

L. paracasei

AM-L4

L. acidophilus

ADH

L. fermentum

1404 L. fermentum L. fermentum C-O stretch in

carboxylic acids L. rhamnosus

AM-L3

97

Both the pH-adjusted and unadjusted spectra of Lactobacillus

paracasei AM-L6 have peaks near 1448 cm-1 as do the pH-adjusted

spectra of Lactobacillus bulgaricus Lb-12 and Lactobacillus paracasei

AM-L4 and the unadjusted spectrum of Kefir Coccus AM-W1. This

peak is explained by Wang, Li and others (2010) as the result of

asymetric bending of C-H2 and C-H3 groups in proteins. Neither of

these peaks is present in the spectra of pomance-only samples. Taken

together these peaks suggest the presence of either a protein or

another nitrogenous compound. EPS from strains of L. paracasei, L.

acidophilus are known contain N-acetyl glucosamine and N-acetyl

galactosamine (DeVuyst and Degeest 1999; Laws and others 2001).

The spectra for Lactobacillus fermentum for both pH-adjusted

and unadjusted samples have a peak near 1404 cm-1. This peak is

attributed to C-O stretching of carboxylic acids (Wang, Li and others

2010). This peak is also present in the unadjusted pH spectrum of

Lactobacillus rhamnosus AM-L3. Again, this peak is not present in the

spectra of pomace-only samples. The EPS from some strains of L.

rhamnosus include a carboxylic acid (Laws and others 2001). Whether

these peaks are in fact evidence of the presence of EPS in these

samples will require further investigation.

98

5.3 Growth and EPS production in tomato juice

The object of this portion of the study was to collect additional

spectra against a different and possibly less cluttered background.

The average quantity of material recovered, 0.18 g, is approximately

36% of what one would expect based on the nutrition facts panel on

the juice can37. The value given on the nutrition facts panel is for total

fiber while our values are for soluble fiber. This data strongly suggests

that the fraction of the recovered material attributable to bacterial

action is quite small.

Using the peaks assignments discussed above, we find peaks

near 1448 cm-1 in the spectra of L. rhamnosus AM-L3, L. paracasei

AM-L6, L. acidophilus ADH and L. fermentum. This peak was not

present in the control. Peaks were not found at 1543 cm-1 or 1404

cm-1. This pattern suggests the presence of methylene and methyl

groups but is not suggestive of anything peculiar to EPS structures.

There are a number of possible causes for the differences in

peak patterns between the two datasets. Grape pomace may induce

EPS production while tomato juice does not. It's also possible the

37

2 g fiber

240 mL juice

200 mL juice

1000 mL solution

300 mL solution

jar

0.5 g fiber / jar

99

peaks observed in the grape pomace samples are phantoms. More

data will be required to select between these alternatives.

100

CHAPTER 6

CONCLUSION

Based on our results, we conclude that the majority of the

species used in this study can survive for a period of weeks in a

mixture of grape pomace and water. Most do not thrive however. The

one exception was L. fermentum which appeared to thrive in the low

pH environment. There are a several further avenues of research to

develop grape pomace as a suitable substrate for LAB. How would the

results differ had we sterilize the pomace and hydroxide separately?

Would a starting pH nearer neutral result in better growth and

survival? And finally, can a strong base in the presence of heat be

used to completely hydrolyze grape pomace and would the resulting

mixture be a suitable substrate for LAB?

The results of our EPS study were less clear cut. There are hints

in the FTIR data that could indicate the presence of EPS. Further

study is necessary. Future work in this area should include use of

more prolific EPS-producing LAB strains coupled with a less cluttered

101

background signal. This would permit the identification of the

characteristic peaks for each EPS. This data could then be used to

verify the presence of EPS against a background from grape pomace.

With suitable calibration, this data could also be used to estimate

quantity of EPS produced. An alternate EPS extraction method such as

ultrafiltration should also be considered.

102

APPENDIX A

Survivor regression lines

Figure A-1 Survival and pH curves for L. paracasei. () = Unadjusted

pH, () = Adjusted pH. Solid and filled markers indicate first and

second replicates. Regression lines for unadjusted pH runs are dashed. Those for adjusted pH runs are dot-dot-dashed.

103

Figure A-2 Survival and pH curves for L. acidophilus LA-2. () =

Unadjusted pH, () = Adjusted pH. Solid and filled markers indicate first and second replicates. Regression lines for unadjusted pH runs

are dashed. Those for adjusted pH runs are dot-dot-dashed.

104

Figure A-3 Survival and pH curves for L. fermentum. () = Unadjusted

pH, () = Adjusted pH. Solid and filled markers indicate first and second replicates. Regression lines for unadjusted pH runs are dashed.

Those for adjusted pH runs are dot-dot-dashed.

105

Figure A-4 Survival and pH curves for Kefir Coccus AM-W1. () =



106

Figure A-5 Survival and pH curves for L. acidophilus ADH. () =



107

Figure A-6 Survival and pH curves for L. bulgaricus Lb-12. () =



108

Figure A-7 Survival and pH curves for L. rhamnosus AM-L3. () =



109

Figure A-8 Survival and pH curves for L. paracasei AM-L4. () =



110

Figure A-9 Survival and pH curves for L. paracasei AM-L6. () =



111

Table A-1 Tukey's HSD for the slopes of the survivor regression lines.

species adj species:adj

AML6 a Y a AML6:N a LaADH a N b LB12:N a b

La2 a b LaADH:N a b AML3 a b c La2:N a b

Lb12 a b c d AML3:N b c LP b c d e AMW1:N a b c d

AMW1 c d e LP:N b c d Lferm d e Lferm:Y b c d

AML4 e LaADH:Y b c d AML3:Y b c d e

La2:Y b c d e LB12:Y c d e

AML4:Y c d e f AML6:Y d e f

LP:Y d e f

AMW1:Y d e f AML4:N e f

Lferm:N f

112

Table A-2 Tukey's HSD for pH-adjusted and pH-unadjusted subsets of

the slopes of the survivor regression lines.

pH-adjusted not pH-adjusted

L ferm a AML6 a

LaADH a b Lb12 a

AML3 a b c LaADH a

La2 a b c La2 a

Lb12 a b c AML3 a

AML4 a b c AMW1 a b

AML6 b c LP a b

LP c AML4 b c

AMW1 c Lferm c

113

APPENDIX B

FTIR spectra

Figure B-1 FTIR spectra of pH-adjusted extract samples

114

Figure B-2 FTIR spectra of unadjusted extract samples

115

Figure B-3 FTIR spectra of samples grown in tomato juice.

116

APPENDIX C

Programs

C.1 survival-slopes-aov.r

ANOVA and HSD program used to create the data for Table A-1 and

Table A-2.

setwd("~/research/Thesis Data")

cfu = read.csv("survival-slopes-cfu.csv",header=TRUE)

pH = read.csv("survival-slopes-pH.csv",header=TRUE)

cfu

pH

cfuaov = aov(cfu ~ species+adj+species*adj,data=cfu)

summary(cfuaov)

TukeyHSD(cfuaov,which="species",ordered=TRUE)

TukeyHSD(cfuaov,which="adj",ordered=TRUE)

TukeyHSD(cfuaov,which="species:adj",ordered=TRUE)

phaov = aov(pH ~ species+adj+species*adj,data=pH)

summary(phaov)

#TukeyHSD(phaov,which="species",ordered=TRUE)

TukeyHSD(phaov,which="adj",ordered=TRUE)

#

# Now look for between species differences within

# pH adjusted and unadjusted subsets

#

cfu.adj = read.csv("survival-slopes-cfu-

adj.csv",header=TRUE)

cfu.noadj = read.csv("survival-slopes-cfu-

noadj.csv",header=TRUE)

pH.adj = read.csv("survival-slopes-pH-adj.csv",header=TRUE)

117

pH.noadj =

read.csv("survival-slopes-pH noadj.csv",header=TRUE)

cfuaov.adj = aov(cfu ~ species ,data=cfu.adj)

summary(cfuaov.adj)

TukeyHSD(cfuaov.adj, ordered=TRUE)

cfuaov.noadj = aov(cfu ~ species ,data=cfu.noadj)

summary(cfuaov.noadj)

TukeyHSD(cfuaov.noadj, ordered=TRUE)

summary( aov(pH ~ species ,data=pH.adj) )

summary( aov(pH ~ species ,data=pH.noadj) )

C.2 extract.r

ANOVA and HSD program used to analyze extract quantities.


extract = read.csv("extract.csv",header=TRUE)

extract

extract.aov = aov(extract ~ species + adj + species*adj,

data = extract)

summary(extract.aov)

#TukeyHSD(extract.aov,which="species",ordered=TRUE)

TukeyHSD(extract.aov,which="adj",ordered=TRUE)

#TukeyHSD(extract.aov,which="species:adj",ordered=TRUE)

118

119

C.3 peak-compare.r

Program to locate possible peaks in IR spectra. Locations of peaks in treatment samples are compared with those identified in the untreated

controls. Matches are assumed to come from the substrate and are rejected. This program also plots the spectra.

library(Peaks)

#####

# UniquePeaks(spectrum, backgnd, tolerance)

# Find peaks unique to spectrum, not found in background.

# Tolerence determines how apart peaks can be and still

# be considered the same.

UniquePeaks <- function(spectrum, backgnd, tolerance=1) {

spec_olen = length(spectrum) # remember original lengths

bg_olen = length(backgnd)

UNIQ = array()

spec = sort(spectrum) # order and use side effect(NA removal)

bg = sort(backgnd)

bgidx = 1

uniqidx = 0

for (specidx in 1:length(spec)) { # index for spectrum

idx = 1 # index for background

repeat {

if ( spec[specidx] >= (backgnd[idx] - tolerance)

& spec[specidx] <= (backgnd[idx] + tolerance) ) {

# peak matches - ignore

break

}

idx = idx + 1

if (idx > length(bg)) {

uniqidx = uniqidx + 1

120

UNIQ[uniqidx] = spec[specidx]

break

}

}

}

length(UNIQ) = spec_olen

return(UNIQ)

}

##########################################

myplot = function(X,Y,jar,species) {

plot(X, Y,

xlab = "Wavenumber (1/cm)",

ylab = "% Reflectance",

xlim = (c(2000,500)),

ylim = (c(0,100)),

xaxs = "i",

yaxs = "i",

type = "l"

)

mtext(paste(jar,"-",species,sep=" "))

}

##########################################

pad_length = 100

X = NULL

Y = NULL

POS = NULL

COLUMNS = NULL


pj_path = "ftir-cooked/"

jar_list = read.csv("jar-list",header=TRUE,colClasses="character")

121

for (file in jar_list$jar) {

count = 6

realN = 0

for ( N in 1:count) {

filename = paste(pj_path,file,"-",N,".txt",sep="")

if(file.exists(filename)) {

spectrum = read.csv(filename,header=FALSE)

realN = realN + 1

}

else {

next

}

if (realN == 1) {

AVG = spectrum

}

else {

AVG = AVG + spectrum

}

}

AVG = AVG / realN

X = data.frame(cbind(X, AVG$V1))

Y = data.frame(cbind(Y, AVG$V2))

z = SpectrumSearch(100-AVG$V2,sigma=2,threshold=10)

z$pos = sort(z$pos)

length(z$pos) = pad_length

POS = data.frame(cbind(POS, z$pos))

print(file)

print(z$pos)

if (is.null(COLUMNS)) {

COLUMNS = file

}

else {

COLUMNS = cbind(COLUMNS, file)

}

}

colnames(POS) = COLUMNS

colnames(X) = COLUMNS

122

colnames(Y) = COLUMNS

#

# At this point, we have x's, y's and peakindicies in X, Y and POS.

# Also we've got a list of jars, matching pomace and species in

jar_list

#

# 1st, get remove peaks in sample and pomace

#

Q = 3 # minimum # of points before peaks considered different.

# Q = 1, one point to either side, Q = 2, two points....

# For 3631 points in 500-4000, data points are 0.9643

cm^-1 apart

# use length(a$b) to get number of rows - using length(a) gives

number of columns

for (i in 1:length(jar_list$jar)) {

jar = jar_list$jar[i]

pomace = jar_list$pomace[i]

species = jar_list$species[i]

if(jar != pomace) {

print(paste("jar =",jar,"pomace =",pomace,"species

=",species,sep=" "))

zpos = UniquePeaks(POS[[jar]],POS[[pomace]],Q)

length(zpos) = pad_length

POS[[jar]] = zpos

}

else {

print(paste("Skipping jar",jar,sep=" "))

}

}

pdf("spectra.pdf",width=6,height=3,onefile=TRUE)

for (i in 1:length(jar_list$jar)) {

jar = jar_list$jar[i]

pomace = jar_list$pomace[i]

species = jar_list$species[i]

p1543 = jar_list$p1543[i]



123

################

region = c("CHO [stretch]",

"Amide II (N-H)",

"Amide I (C=O)",

"Ketones&COOR"

)

range_lo = c(950,1500,1630,1700)

range_hi = c(1200,1650,1700,1750)

################

# plain plot

#### myplot(X[[jar]], Y[[jar]], jar, species)

################

# plain plot with peaks

opar = par( mar=c(4,4,2,1) )

myplot(X[[jar]], Y[[jar]], jar,

paste(species,"(Q=",Q,")",sep=" "))

# yes, but it's harder to read without the extra assignments

pos = sort( POS[[jar]] )

x = X[[jar]][pos]

y = Y[[jar]][pos]

print("~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~~~~~~~~~~~~~")

print(cbind(jar," ",species))

print(x)

print("~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

~~~~~~~~~~~~~~~~~~")

#### Uncomment to turn on marking all peaks

# segments(x, y , x , y + 10)

# segments(x, y + 10, x - 50, y + 12)

if (p1543 == "Y") {

xa = 1543; xb = xa

ya = Y[[jar]][1083]; yb = ya + 40

if(yb > 80) { yb = ya - 30 }

segments(xa, ya, xb, yb)

text(xb, yb, "1543",adj=c(0,0.5))

}

if (p1448 == "Y") {

124

xa = 1448; xb = xa

ya = Y[[jar]][984]; yb = ya + 30

if(yb > 80) { yb = ya - 20 }


text(xb, yb, "1448",adj=c(0,0.5))

}

if (p1404 == "Y") {

xa = 1404; xb = xa

ya = Y[[jar]][939]; yb = ya + 20

if(yb > 80) { yb = ya - 10 }


text(xb, yb, "1404",adj=c(0,0.5))

}

box(which="figure",lwd=3)

par(opar)

REFERENCES

AACC Report. 2001. The definition of dietary fiber. Cereal Foods World.

46(3):112-26. Available from: http://www.aaccnet.org/DietaryFiber/pdfs/dietfiber.pdf

AACC Report 2003. All dietary fiber is fundamentally functional. Cereal Foods World. 48(3):128-32. Available from:

http://www.aaccnet.org/news/pdfs/DFreport.pdf

Aarnikunnas J. 2006. Metabolic engineering of lactic acid bacteria and characterization of novel enzymes for the production of industrially

important compounds [DPhil thesis]. Kuopio, Finland: University of Kuopio. 67p. Available from:

https://oa.doria.fi/bitstream/handle/10024/777/metaboli.pdf

Adamopoulos KG, Goula AM, Petropakis HJ. 2001. Quality control during processing of feta cheese - NIR application. J Food Composition

and Analysis. 14:431-40.

American Heart Association. 2005. Heart Disease and Stroke Statistics

- 2005 Update. American Heart Association: Dallas, Texas.

Amiel C, Mariey L, Curk-Daubié M-C, Pichon P, Travert J. 2000. Potentiality of Fourier transform infrared spectroscopy (FTIR) for

125

discrimination and identification of dairy lactic acid bacteria. Lait.

80:445-59.

Amiel C, Mariey L, Denis C, Pichon P, Travert J. 2001. FTIR spectroscopy and taxonomic purpose: contribution to the classification

of lactic acid bacteria. Lait. 81:249-55. anyletter [sic]. What is the craziest fact you know that most people

won't know [Internet]. 2010. [Accessed 2010 Oct 20]. Available from: http://www.reddit.com/r/AskReddit/comments/awi5k/what_is_the_cra

ziest_fact_you_know_that_most/c0jrzak

Arai T, Kuroda S, Koyama Y. 1963. Infrared Absorbtion spectra of whole cells on Nocardia and related organisms. J General Applied

Microbiology. 9(1):119-136. Aranaz I, Mengíbar M, Harris R, Paños I, Miralles B, Acosta N, Galed G,

Heras A. 2009. Functional characterization of chitan and chitosan. Cur

Chem Biol. 3:203-30.

Asp N-G. 1987. Dietary fibre - definition, chemistry and analytical determination. Molecular Aspects of Medicine. 9:17-29.

Assifaoui A, Loupiac C, Chambin O, Cayot P. 2010. Structure of

calcium and zinc pectinate films investigated by FTIR spectroscopy. Carbohydrate Res. 345:929-33.

Athanasiadis I, Boskou D, Kanellaki M, Kiosseoglou V, Koutina AA.

2002. Whey liquid waste of the dairy industry as raw material for potable alcohol production by kefir granules. J Agricultural and Food

Chemistry. 50:7231-4.

Banwell CN. 1972. Fundamentals of molecular spectroscopy, 2nd ed.

London: McGraw Hill. 348 p.

Bao Y, Zhang Y, Zhang Y, Liu Y, Wang S, Dong X, Wang Y, Zhang H. 2010. Screening of potential probiotic properties of Lactobacillus

fermentum isolated from traditional dairy products. Food Control. 21:695-701.

Barbano DM, Dellavalle ME. 1987. Rapid method for determination of

milk casein content by infrared analysis. Journal of Dairy Science. 70(8):1524-1528.

126

Beaumont W. 1834. Experiments and observations on the gastric juice, and the physiology of digestion. Boston: Lilly, Wait and

Company. 280 p.

Behare PV, Singh R, Kumar M, Prajapati JB, Singh RP. 2009. Exopolysaccharides of lactic acid bacteria: a review. J Food Sci and

Technol. 46(1):1-11.

Belitz H-D, Grosch W, Schieberle. 2004. Food chemistry: translation from the fifth German edition by M. M. Burghagen, 3rd revised ed.

Berlin: Springer-Verlag. 1070 p.

Bertolami MC, Farnworth ER. 2008. The properties of Enterococcus faecium and the fermented milk product -- Gaio. In: Farnworth ER.

Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC

Press. p 71-88.

Biggs DA. 1972. Infrared milk analyzer. J Dairy Sci. 55(5):650-1.

Bloem A, Bertrand A, Lonvaud-Funel A, de Revel G. 2007. Vanillin production from simple phenols by wine-associated lactic acid bacteria.

Letters in Applied Microbiology. 44:62-7.

Bonnand-Ducasse M, Della Valle G, Lefebvre J, Saulnier L. 2010. Effect of wheat dietary fibres on bread dough development and rheological

properties. J Cereal Sci. 52:200-6.

Bravo L, Saura-Calixto F. 1998. Characterization of dietary fiber and the In Vitro indigestible fraction of grape pomace. Am J Enol and Vit.

49(2):135-41.

Brummer Y, Cui SW. 2005. Understanding carbohydrate analysis. In:

Cui SW, editor. Food carbohydrates: chemistry, physical properties and applications. Boca Raton: Taylor & Francis. p 67-104.

Buchanan RE, Gibbons NE. 1974. Bergey's manual of determinative

bacteriology, 8th ed. Baltimore: Williams & Wilkins Company. 1246 p.

Buriti FCA, Castro IA, Saad SMI. 2010. Viability of Lactobacillus acidophilus in synbiotic guava mousses and its survival under in vitro

127

simulated gastrointestinal conditions. Int'l J Food Micro. 137(2-3):121-

9.

Cadet F, Offmann B. 1997. Direct Spectroscopic Sucrose Determination of Raw Sugar Cane Juices. J Agric Food Chem. 45:166-

71.

Cai H, Marczylo TH, Teller N, Brown K, Steward WP, Marko D, Gescher AJ. 2010. Anthocyanin-rich red grape extract impedes adenoma

development in the ApcMin mouse: Pharmacodynamic changes and anthocyanin levels in the murine biophase. Euro J Cancer. 46(4):811-

7.

Carbonaro M, Nucara A. 2010. Secondary structure of food proteins by Fourier trasform spectroscopy in mid-infrared region. Amino Acids.

38:679-90.

Carpenter KJ. 2003. A short history of nutritional science: part 2

(1885-1912). J Nutr. 113:975-84.

Carr FJ, Chill D, Maida N. 2002. The lactic acid bacteria: a literature survey. Crit Rev Micro. 28(4):281-370.

Chen W, Duizer L, Corredig M, Goff HD. 2010. Addition of soluble

soybean polysaccharides to dairy droducts as a source of dietary fiber. J Food Sci. 75(6):478-84.

Corsetti A, Gobbetti M, Balestrieri F, Paoletti F, Russi L, Rossi J. 1998.

Sourdough lactic acid bacteria effects on bread firmness and staling. J Food Sci. 63(2):347-51.

Corsetti A, Settanni L. 2006. Lactobacilli in sourdough fermentation. Food Res Int'l. 40(5):539-58.

Cortés S, Salgado JM, Rodríguez N, Domínguez JM. 2010. The storage

of grape marc: limiting factor in the quality of the distillate. Food Control. 21(11):1545-9.

Coussement P. 1999. Inulin and oligofructose as dietary fiber:

analytical, nutritional and legal aspects. In: Cho SS, Prosky L, Dreher

128

M, editors. Complex carbohydrates in foods. New York: Marcel Dekker.

p 203-12.

Cui SW. 2005. Structural analysis of polysaccharides. In: Cui SW, editor. Food carbohydrates: chemistry, physical properties and

applications. Boca Raton: Taylor & Francis. p 105-60.

Daintith J, ed. 2000. A dictionary of chemistry, 4th ed. Oxford: Oxford University Press. 586 p.

Daniel JR, Yao Y, Weaver C. 2007. Carbohydrate: functional

properties. In: Hui YH, editor. Food Chemistry: Principles and Applications, 2nd ed. West Sacramento:Science Technology System. p

1-27. Davis R, Irudayaraj J, Reuhs BL, Mauer LJ. 2010. Detection of E. coli

O157:H7 from ground beef using Fourier transform infrared (FT-IR)

spectroscopy and chemometrics. J Food Sci. 75(6):340-6.

De Man JC, Rogosa M, Sharpe ME. 1960. A medium for cultivation of Lactobacilli. J Applied Bacteriol. 23(1):130-5.

De Vos P, Garrity GM, Jones D, Krieg NR, Ludwig W, Rainey FA,

Schleifer K-H, Whitman WB. 2009. Bergey's manual of systematic bacteriology, 2nd ed, vol 3: the firmicutes. Dordrecht German:

Springer. 1450 p.

De Vuyst L, Degeest B. 1999. Heteropolysaccharides from lactic acid bacteria. FEMS Micro Revs. 23:153-77.

Difco. 1998. Difco manual, 11th ed. Sparks Maryland:Difco

Laboratories. 862 p.

Dongowski G, Lorenz A, Anger H. 2000. Degradation of pectins with

different degrees of esterification by Bacteroides thetaiotaomicron isolated from human gut flora. Appl Environ Micro. 66(4):1321-7.

Dreher ML. 2001. Dietary fiber overview. In: Cho SS, Dreher ML,

editors. Handbook of dietary fiber. New York: Marcel Dekker Inc. p 1-16.

129

Dziuba B, Babuchowski A, Niklewicz M, Brzozowski B. 2006. FTIR

spectral characteristics of lactic acid bacteria - a spectral library. Milchwissenschaft. 61(2):146-9.

D iuba B, Babuchowski A, Nalec D, Nikłęwic . 2007a. Identificatin

of lactic acid bacteria using FTIR spectroscopy and cluster analysis. Int'l Dairy J. 17:183-9.

Dziuba B, Babuchowski A, Dziuba M, Nalecz D. 2007b. Identification of

lactic acid bacteria using FTIR spectroscopy an artificial neural networks. Milchwissenschaft. 62(1):28-32.

EAFUS. 2010. Everything added to food in the United States

[Internet]. Washington DC: Food and Drug Administration. [Updated 7

Sep 2010: Accessed 13 Oct 2010]. Available from: (www.accessdata.fda.gov/scripts/fcn/fcnNavigation.cfm?rpt=eafusListi

ng&displayAll=true).

El Oufir L, Barry JL, Flourié B, Cherbut C, Cloarec D, Bornet F, Galmiche JP. 2000. Relationships between transit time in man and in

vitro fermentation of dietary fiber by fecal bacteria. Euro J of Clin Nutr. 54:603-9.

Eskamani A. 1977. Food Industry. In: Brame EG Jr, Grasselli JG,

editors . Infrared and raman spectoscopy, part B. New York: Marcel Dekker. p 629-34.

European Parliament and Council. 2006. Directive no 95/2/EC of 20

February 1996 on food additives other than colours and sweeteners,

revised 15 August 2006 [Internet]. Accessed 13 Oct 2010. Available from EUR-Lex (eur-

lex.europa.eu/LexUriServ/LexUriServ.do?uri=CONSLEG:1995L0002:20060815:EN:PDF).

Farnworth ER, Mainville I. 2008. Kefir -- a fermented milk product. In:

Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 89-127.

130

FDA (US Food and Drug Administration). 1987. Nutrition labeling of

food; calorie content.Federal Register. 52 FR 28690-28691.

FDA (US Food and Drug Administration). 2006. Curdlan. 21 CFR 172.809

FDA (US Food and Drug Administration). 2007. Food labeling: revision

of reference values and mandatory nutrients. Federal Register. 72 FR 62149-62174.

FDA (US Food and Drug Administration). 2008a. Gellan gum. 21 CFR

172.665

FDA (US Food and Drug Administration). 2008b. Xanthan gum. 21 CFR 172.695

Fleet GH. 1998 The microbiology of alcoholic beverages. In: Wood BJB,

editor. Microbiology of fermented foods, 2nd ed, vol 1. London: Blackie Academic & Professional. p 217-62.

Food Additive Dictionary. 2006. Food additives permitted for use in

Canada [Internet]. Ottawa, Canada: Health Canada. [Updated 11 Dec 2006: Accessed 13 Oct 2010] Available from: (www.hc-sc.gc.ca/fn-

an/securit/addit/diction/dict_food-alim_add-eng.php)

Food standards code. 2010. Food standards Australia New Zealand [Internet]. Canberra, Australia:FSANZ. [Updated 7 Oct 2010: Accessed

13 Oct 2010] Available from (www.foodstandards.gov.au/foodstandards/foodstandardscode/)

Fowler MS, Leheup P, Cordier J-L. 1998. Cocoa, coffee and tea. In:

Wood BJB, editor. Microbiology of fermented foods, 2nd ed, vol 1.

London: Blackie Academic & Professional. p 128-47.

Fraune S, Bosch TCG. 2010. Why bacteria matter in animal development and evolution. Bioassays. 32:571-80.

García-Ruiz A, Bartolomé B, Martínez-Rodríguez AJ, Pueyo E, Martín-

Álvarez PJ, Moreno-Arríbas MV. 2008. Potential of phenolic compounds for controlling lactic acid bacteria growth in wine. Food Control.

19:835-41.

131

GFSA Online Food Additive Index [Internet]. Rome, Italy: FAO/WHO.

[Updated 2010: Accessed 13 Oct 2010] Available from: http://www.codexalimentarius.net/gsfaonline/additives/index.html.

Gibson KS. 1920. The infra-red absorption spectra of vegetable oils.

The Chemist's Section of the Cotton Oil Press. 4(5):53.

Goni I, Martín N, Saura-Calixto F. 2005. In vitro digestibility and intestinal fermentation of grape seed and peel. Food Chem. 90:281-6.

González-Centeno MR, Rosselló C, Simal S, Garau MC, López F,

Femeniaa A. 2010. Physico-chemical properties of cell wall materials obtained from ten grape varieties and their byproducts: grape

pomaces and stems. LWT-Food Sci and Technol. 43:1580-6. González-Paramás AM, Esteban-Ruana S, Santos-Buelga C, de

Pascual-Teresa S, Rivas-Gonzalo JC. 2004. Flavanol content and

antioxidant activity in winery byproducts. J Agric Food Chem. 52:234-8.

Gonzalez Siso MI. 1996. The biotechnological utilization of cheese

whey: a review. Bioresource Technol. 57(1):1-11.

Goodfellow BJ, Wilson RH. 1990. A fourier transform IR study of the gelation of amylose and amylopectin. Biopolymers. 30:1183-9.

Gropper SS, Smith JL, Groff JL. 2009. Advanced nutrition and human

metabolism, 5th ed. Belmont CA: Wadsworth. 600 p.

Günzler H, Gremlich H-U. 2002. IR spectroscopy: an introduction, translated by Mary-Joan Blümich. Weinheim, Federal Republic of

Germany. 361 p.

Guo H, Yi W, Song JK, Wang PG. 2008. Current understanding on

biosynthesis of microbial polysaccharides. Current Topics in Medical Chemistry. 8:141-51.

Guzzo J, Desroche N. 2009. Physical and chemical stress factors in

lactic acid bacteria. In: König H, Unden Gottfried, Frölich J, editors. Biology of microorganisms on grapes, in must and in wine. Berlin:

Springer-Verlag. p 293-306.

132

Hamdi M. 2008. New trends of table olive processing for quality control

and function properties. In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 413-31.

Hammes WP, Gänzle MG. 1998. Sourdough breads and related

products. In: Wood BJB, editor. Microbiology of fermented foods, 2nd ed, vol 1. London: Blackie Academic & Professional. p 199-216.

Hammes WP, Haller D, Gänzle MG. 2008. Fermented meat. In:


Harris LJ. 1998. The microbiology of vegetable fermentations. In:

Wood BJB, editor. Microbiology of fermented foods, 2nd ed, vol 1. London: Blackie Academic & Professional. p 45-72.

Hayashibara International. 2002. Pullulan GRAS Notification [Internet]. Accessed: 14 Oct 2010. Available from FDA:

(www.accessdata.fda.gov/scripts/fcn/gras_notices/215492E.PDF)

Heller KJ, Bockelmann W, Schrezenmeir J, de Vrese M. 2008. Cheese and its potential as a probiotic food. In: Farnworth ER, editor.

Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 243-66.

Hipsley EH. 1953. Dietary "fibre" and pregnancy toxaemia. Brit Med J.

V2(4833):420-2.

Hogan S, Zhang L, Li J, Sun S, Canning C, Zhou K. 2010. Antioxidant rich grape pomace extract suppresses postprandial hyperglycemia in

diabetic mice by specifically inhibiting alpha-glucosidase. Nutr & Metab

[serial online]. 7(71):1-9. Available from: http://www.nutritionandmetabolism.com/content/7/1/71. Posted 27

Aug 2010.

Holzapfel W, Schillinger U, Buckenhüskes H. 2008. Sauerkraut. In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed.

Boca Raton: CRC Press. p 395-412.

133

Horwitz W, Latimer GW Jr, eds. 2007. Official methods of analysis of

AOAC International, 18th edition, 2005, current through revision 2, 2007. Gaithersburg MD: AOAC International. 1 v. (various pagings)

Hsieh Y-H P, Pao S, Li J. 2008. Traditional Chinese fermented foods.

In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 433-73.

Hull RR, Roberts AV, Mayes JJ. 1984. Survival of Lactobacillus

acidophilus in yoghurt. Australian J Dairy Technol. 39:164-6.

Inglett GE, Carriere CJ, Maneepun Saipin, Tungtrakul P. 2004. A soluble fibre gel produced from rice bran and barley flour as a fat

replacer in Asian foods. Int'l J Food Sci and Technol. 39:1-10. IOM (Institute of Medicine). 2001. Dietary reference intakes proposed

definition of dietary fiber: a report of the panel on the definition of

dietary fiber and the standing committee on the scientific evaluation of dietary reference intakes. Washington DC: Nat'l Acad Press. 64 p.

Jane J-L. 2007. Carbohydrates: basic concepts. In: Hui YH, editor,

Food Chemistry: Principles and Applications, 2nd ed. West Sacramento: Science Technology System. p 1-22.

Janssen PH, Evers S, Rainey FA, Weiss N, Ludwig W, Harfoot CG,

Schink B. 1995. Lactosphaera gen. nov. a new genus of lactic acid bacteria and transfer of Ruminococcus pasteurii Schink 1984 to

Lactosphaera pasteurii comb. nov. Int'l J Systematic Bacteriology. 45(3):565-71.

Jay JM, Loessner MJ, Golden DA. 2005. Modern food microbiology, 7th

edition. New York NY: Springer. 790p

JETRO (Japan External Trade Organization). 2009. Specifications and

standards for foods, food additives, etc. Under the food sanitation act (abstracts) 2008 [Accessed 21-Oct-2010] Available from:

http://www.jetro.go.jp/en/reports/regulations/pdf/foodext2008e_100929.pdf

JFCRF. 2007. List of Existing Food Additives [Internet]. The Japan

Food Chemical Research Foundation. [Updated 8 June 2010: Accessed

134

13 Oct 2010] Available from

http://www.ffcr.or.jp/zaidan/FFCRHOME.nsf/pages/list-exst.add

Johnson BC. 1946. The activity of "Lactobacillus casei factor," and "Vitamin Bc" for Streptococcus faecalis and Lactobacillus casei. J Biol

Chem. 163:255-9.

Kam PV, Bianchini A, Bullerman LB. 2007. Inhibition of mold growth by sourdough bread cultures. RURALS: Review of Undergraduate

Research in Agricultural and Life Sciences. 2(1). Article 5. Accessed: 18-Oct-2010. Available from:

http://digitalcommons.unl.edu/rurals/vol2/iss1/5

Kargi G, Ozmihci S. 2006. Utilization of cheese whey powder (CWP) for ethanol fermentations: effects of operating parameters. Enzyme and

Microbial Technology. 38(5):711-8.

Kaylegian KE, Houghton GE, Lynch JM, Fleming JR, Barbano DM.

2006a. Calibration of infrared milk analyzers: modified milk versus producer milk. J Dairy Sci. 89:2817-32.

Kaylegian KE, Lynch JM, Houghton GE, Fleming JR, Barbano DM.

2006b. Modified versus producer milk calibration: mid-infrared analyzer performace validation. J Dairy Sci. 89:2833-45.

Keenan MJ, Zhou J, McCutcheon KL, Raggio AM, Bateman HG, Todd E,

Jones CK, Tulley RT, Melton S, Martin RJ, Hegsted M. 2006. Effects of resistant starch, a non-digestible fermentable fiber, on reducing body

fat. Obesity. 14(9):1523-34.

Khalid AM, Bhatti TM, Shoukat M, Qadir G, Hayat MN. 1993. A simple

apparatus for measuring hardness of agar and Gelrite plates. World J Micro and Biotechnol. 9:128-30.

Kodali VP, Das S, Sen R. 2009. An exopolysaccharide from a probiotic:

biosynthesis dynamics, composition and emulsifying activity. Food Res Int'l. 42:695-9.

Korea Food Additives Code [Internet]. Seoul, Korea: Korea Food and

Drug Administration. [Updated 2004: Accessed 13 Oct 2010. Available from: http://fa.kfda.go.kr/foodadditivescode.html.

135

Kuhn LP. 1950. Infrared spectra of carbohydrates. Analytical Chemistry. 22(2):276-83.

Kulp WL. 1927. An agar medium for plating L. acidophilus and L.

bulgaricus. Science. 66(1717):512-3.

Kulp WL, White V. 1932. A modified medium for plating L. acidophilus. Science. 76(1957):17-8.

Larrauri JA. 1999. New approaches in the preparation of high dietary

fibre powders from fruit by-products. Food Sci and Technol. 10:3-8. Laufenberg G, Kunz B, Nystroem M. 2003. Transformatin of vegetable

waste into value added products: (A) the upgrading concept; (B) practical implementations. Bioresource Technol. 87:167-98.

Laws A, Gu Y, Marshall V. 2001. Biosynthesis, characterisation, and design of bacterial exopolysaccharides from lactic acid bacteria.

Biotechnology Advances. 19:597-625.

Laws AP, Chadha MJ, Chacon-Romero M, Marshall VM, Maqsood M. 2008. Determination of the structure and molecular weights of the

exopolysaccharide produced by Lactobacillus acidophilus 5e2 when grown on different carbon feeds. Carbohydrate Res. 343:301-7.

Liu J-R, Tanner RS, Schumann P, Weiss N, McKenzie CA, Janssen PH,

Seviour EM, Lawson PA, Allen TD, Seviour RJ. 2002. Emended description of the genus Trichococcus, description of Trichococcus

collinsii sp. nov., and reclassification of Lactosphaera pasteurii as Trichococcus pasteurii comb. nov. and of Ruminococcus palustris as

Trichococcus palustris comb. nov. in the low-G+C Gram-positive

bacteria. Int'l J Syst and Evolut Micro. 52:1113-26.

Llobera A, Cañellas J. 2007. Dietary fibre content and antioxidant activity of Manto Negro red grape (Vitis vinifera): pomace and stem.

Food Chem. 101:659-66.

Lücke F-K. 1998. Fermented sausages. In: Wood BJB, editor. Microbiology of fermented foods, 2nd ed, vol 2. London: Blackie

Academic & Professional. p 441-83.

136

Mañas E, Saura-Calixto F. 1993. Ethanolic precipitation: a source of

error in dietary fibre determination. Food Chem. 47:351-5.

Mariscotti MA. 1967. A method for automatic identification of peaks in the presence of background and its application to spectrum analysis.

Nuclear Instruments and Methods. 50:309-20.

Marcotte L, Kegelaer G, Sandt C, Barbeau J, Lafleur M. 2007. An alternative infrared spectroscopy assay for the quantification of

polysaccharides in bacterial samples. Analyt Biochem. 361:7-14. Marlett JA, Chesters JG, Longacre MJ, Bogdanske JJ. 1989. Recovery of

soluble dietary fiber is dependent on the method of analysis. J Clin Nutr. 50:479-85.

Martin-Carron N, Garcia-Alonso A, Goñi I, Saura-Calixto F. 1997.

Nutritional and physiological properties of grape pomace as a potential

food ingredient. Am J Enol and Vit. 48(3):328-32.

Martiníz-Villaluenga C, Frías J, Gómez R, Vidal-Valverde C. 2006. Influence of addition of raffinose family oligosaccharides on probiotic

survival in fermented milk during refrigerated storage. Int'l Dairy J. 16:768-74.

Marwaha SS, Kennedy JF. 1988. Whey-pollution problem and potential

utilization. Int'l J Food Sci and Technol. 23(4):323-36.

McGee H. 2004. On food and cooking: the science and lore of the kitchen. New York: Scribner. 884 p.

Michel C, Bény C, Delorme F, Poirier L, Spolaore P, Morin D, d'Hugues

P. 2009. New protocol for the rapid quantation of exopolysaccharides

in continuous culture systems of acidophilic bioleaching bacteria. Appl Micro and Biotechnol. 82:371-8.

Min DB, Boff JM. 2003. Crude fat analysis. In: Nielsen SS, editor. Food

analysis. New York: Springer. p 113-29.

Mirande C, Kadlecikova E, Matulova M, Capek P, Bernalier-Donadille A, Forano E, Béra-Maillet C. 2010. Dietary fibre degradation and

fermentation by two xylanolytic bacteria Bacteroides xylanisolvens XB1AT and Roseburia intestinalis XB6B4 from the human intestine. J

137

Appl Micro. 109:451-60.

Mitchell HK, Snell EE, Williams RJ. 1941. The concentration of "folic

acid". J Am Chem Soc. 63(8):2284.

Miyazaki K, Matsuzaki T. 2008. Health properties of milk fermented with Lactobacillus caseii Shirota (LcS). In: Farnworth ER, editor.

Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 165-208.

Molin G. 2008. Lactobacillus plantarum: the role in foods and in human health. In: Farnworth ER, editor. Handbook of fermented functional

foods, 2nd ed. Boca Raton: CRC Press. p 353-93.

Monsoor MA, Kalapathy U, Proctor A. 2001. Determination of polygalacturonic acid content in pectin extracts by diffuse reflectance

Fourier transform infrared spectroscopy. Food Chem. 74:233-8.

Moore JC, DeVries W, Lipp M, Griffiths JC, Abernethy DR. 2010. Total

protein methods and their potential utility to reduce the risk of food protein adulteration. Comprehensive Reviews in Food Science and

Food Safety. 9:330-57.

Morin LA, Temelli F, McMullen L. 2002. Physical and sensory characteristics of reduced-fat breakfast sausages formulated with

barley β-glucan. J Food Sci. 67(6):2391-6.

Morris William ed. 1969. The American heritage dictionary of the English language. New York: American Heritage Publishing Co. 1550 p.

Mossoba MM, Seiler A, Steinhart H, Kramer JKG, Rodrigues-Saona L,

Griffith AP, Pierceall R, van de Voort FR, Sedman J, Ismail AA, Barr D,

Da Costa Filho PA, Li H, Zhang Y, Liu X, Bradley M. Forthcoming. Regulatory infrared spectroscopic method for the rapid determination

of total isolated trans fat: A collaborative study. J Am Oil Chem Soc.

Mozzi F, Rollan G, Savoy De Giori G, Font de Valdez G. 2001. Effect of galactose and glucose on the exopolysaccharide production and the

activities of biosynthetic enzymes in Lactobacillus casei CRL 87. J Appl Micro. 91:160-7.

138

Nef JU. 1914. Dissoziationsvorgänge in der Zuckergruppe. Justus

Liebig's Annalen der Chemie. 403:204-83.

Nelson DL, Cox MM. 2005. Lehninger principles of biochemisty, 4th ed. New York: WH Freeman and Co. 1 v. (various pagings).

Nelson ME, Werkman CH. 1935. Dissimilation of glucose by

heterofermentative lactic acid bacteria. J Bacteriology. 30(6):547-57. Oberman H, Libudzisz H. 1998. Fermented milks. In: Wood BJB,

editor. Microbiology of fermented foods, 2nd ed, vol 1. London: Blackie Academic & Professional. p 308-50.

Olsen EB, Russell JB, Henick-Kling T. 1991. Electrogenic L-malate

transport by Lactobacillus plantarum: a basis for energy derivation from malolactic fermentation. J Bacteriology. 173(19):6199-206.

Otten JJ, Hellwig JP, Meyers LD, editors. 2006. Dietary reference intakes: the essential guide to nutrient requirements. Washington DC:

The National Academies Press. 543 p.

Park J, Floch MH. 2007. Prebiotics, probiotics, and dietary fiber in gastrointestinal disease. Gastroenterology Clinics of North America.

36:47-63.

Pazur JH. 1986. Neutral polysaccharides. In: Chaplin MF, Kennedy JF, editors. Carbohydrate analysis: a practical approach. Oxford, England:

IRL Press. p 55-96.

Pearson AM, Gillett TA. 1999 Processed Meats, 3rd ed. Gaithersburg MD: Aspen Publications. 448 p.

Pereira DIA, Gibson GR. 2002. Cholesterol assimilation by lactic acid bacteria and Bifobacteria isolated from the human gut. Applied and

Environmental Microbiology. 68(9):4689-93.

Peressini D, Sensidoni A. 2009. Effect of soluble dietary fibre addition on rheological and breadmaking properties of wheat doughs. J Cereal

Sci. 49(2):190-201.

Pickett TA. 1940. A Note on Muscadine Grape Seed Oil. Oil & Soap. 17(11):246.

http://www.scopus.com/source/sourceInfo.url?sourceId=18500162500&origin=resultslist

139

Poolman B, Molenaar D, Smid EJ, Ubbink T, Abee T, Renault PP, Konings WN. 1991. Malolactic fermentation: electrogenic malate

uptake and malate/lactate antiport generate metabolic energy. Journal of Bacteriology. 173(19):6030-7.

Potter NN, Hotchkiss JH. 1998. Food science, 5th ed. Gaithersburg MD:

Aspen Publishers, Inc. 608 p. Prajapati JB, Nair BM. 2008. The history of fermented foods. In:


Prosky L. 2001. What is dietary fiber? A new look at the definition. In:

McCleary BV, Prosky L, editors. Advanced dietary fibre technology. Oxford, England: Blackwell Science. p 63-76.

R Development Core Team. 2008. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna,

Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.

Resta SC. 2009. Effects of probitics and commensals on intestinal epithelial physiology: implications for nutrient handling. J Physiol.

587(17):4169-74.

Rodriguez-Saona LE, Khambaty FM, Fry FS, Calvey EM. 2001. Rapid detection and identification of bacterial strains By Fourier transform

near-infrared spectroscopy. J Agric Food Chem. 49:574-9.

Sahakian AB, Jee S-R, Pimentel M. 2010. Methane and the gastrointestinal tract. Digestive Diseases and Sciences. 55:2135-43.

Sahu RK, Mordechai S, Pesakhov S, Dagan R, Porat N. 2006. Use of FTIR spectroscopy to distinguish between capsular types and capsular

quantities in Streptococcus pneumoniae. Biopolymers. 83:434-42.

Sánchez-Medina I, Gerwig GJ, Urshev ZL, Kamerling JP. 2007. Structure of a neutral exopolysaccharide produced by Lactobacillus

delbrueckii ssp. bulgaricus LBB.B26. Carbohydrate Res. 342:2430-9.

http://www.r-project.org./

140

Savić D, Jaković N, Topisirović L. 2008. ultivariate statistical methods

for discrimination of lactobacilli based on their FTIR spectra. Dairy Sci and Technol. 88:273-90.

Savitzky A, Golay MJE. 1964. Smoothing and differentiation of data by

simplified least squares procedures. Anal Chem. 36(8):1627-39.

Schillinger U, Holzapfel WH. 2003. Culture media for lactic acid bacteria. In: Corry JEL Curtis GDW, Baird RM, editors. Handbook of

culture media for food microbiology. Burlington MA: Elsevier. p 127-140.

Singh RP, Heldman DR. 2009. Introduction to food engineering, 4th

ed. Burlington MA: Elsevier. 841 p.

Slavin JL, Savarino V, Paredes-Diaz A, Fotopoulos G. 2009. A review of

the role of soluble fiber in health with specific reference to wheat dextrin. J Int'l Med Res. 37:1-17.

Sletmoen M, Maurstad G, Sikorski P, Paulsen BS, Stokke BT. 2003.

Chracterisation of bacterial polysaccharides: steps towards single-molecular studies. Carbohydrate Res. 338:2459-75.

Smith BC. 1996. Fundamentals of Fourier transform infrared

spectroscopy. Boca Raton: CRC Press. 202 p.

Southgate DAT. 1969. Determination of carbohydrates in foods II. --- unavailable carbohydrates. J Sci of Food and Agric. 20:331-5.

Spanghero M, Salem AZM, Robinson PH. 2009. Chemical composition, including secondary metabolites, and rumen fermentability of seeds

and pulp of Californian (USA) and Italian grape pomaces. Ani Feed Sci and Technol. 152:243-55.

Stanley G. 1998. Cheeses. In: Wood BJB, editor. Microbiology of

fermented foods, 2nd ed, vol 1. London: Blackie Academic & Professional. p 263-307.

141

Steinier J, Termonia Y, Deltour J. 1972. Comments on smoothing and

differentiation of data by simplified least square procedure. Anal Chem. 44(11):1906-9.

Steinkraus KH. 1996a. Introduction to indigenous fermented foods. In:

Steinkraus KH, editor. Handbook of indigenous fermented foods, 2nd ed. New York: Marcel Dekker, Inc. p 1-5.

Steinkraus KH, editor. 1996b. Indigenous fermented foods involving an acid fermentation: preserving and enhancing organoleptic and

nutriitonal qualities of fresh foods. In: Handbook of indigenous fermented foods, 2nd ed. New York: Marcel Dekker, Inc. p 111-347.

Steinkraus KH. 1998. Bio-enrichment: production of vitamins in

fermented foods. In: Wood BJB, editor. Microbiology of fermented

foods, 2nd ed, vol 2. London: Blackie Academic & Professional. p 603-21.

Stuart BH. 2004. Infrared spectroscopy: fundamentals and

applications. Chichester, West Sussex England: John Wiley & Sons, Ltd. 224 p.

Surh J, Kim Y-K L, Kwon H. 2008. Korean fermented foods: kimchi and

doenjang. In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 333-51.

Sutherland IW. 1977 Bacterial exopolysaccharides - their nature and

production. In: Sutherland I, editor. Surface Carbohydrates of the Prokaryotic Cell. London: Academic Press. p 27-96.

Sutherland IW. 1996. Microbial biopolymers from agricultural products: production and potential. Int'l Biodeterioration and

Biodegradation. 38(3-4):249-61.

Sutherland IW. 2001. Microbial polysaccharides from Gram-negative bacteria. Int'l Dairy J. 11:663-74.

Sutherland IW. 2004. Microbial exopolysaccharides. In: Dumitriu S,

editor. Polysaccharides: structural diversity and functional versatility. Boca Raton: CRC Press. p 459-74.

142

Suzuki K, Asano S, Iijima K, Kitamoto K. 2008. Sake and beer spoilage lactic acid bacteria – A Review. J Institute of Brewing. 114(3):209-23.

Tanasupawat S, Visessanguan W. 2008. Thai fermented foods:

microorganisms and health benefits. In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC

Press. p 495-511.

Tewari J, Irudayaraj J. 2003. Rapid estimation of Pol content in sugarcane juice using FTIR-ATR spectroscopy. Sugar Technology.

5(3):143-8.

Theander O, Westerlund EA. 1986. Studies on dietary fiber. 3. improved procedures for analysis of dietary fiber. J Agric Food Chem.

34:330-6.

Torres JL, Varel B, Garcia MT, Carilla J, Matito C, Centelles JJ,

Cascante M, Sort X, Bobet R. 2002. Valorization of grape (Vitis vinifera) byproducts. Antioxidant and biological properties of

polyphenolic fractions differing in procyanidin composition and flavonol content. J Agric Food Chem. 50:7548-55.

Trowell H. 1972. Ischemic heart disease and dietary fiber. Am J Clin

Nutr. 25:926-32.

Trowell HC, Burkitt DP. 1987. The development of the concept of dietary fibre. Molecular Aspects of Medicine. 9:7-15.

Trowell H, Southgate DAT, Wolever TMS, Leeds AR, Gassull MA,

Jenkins DJA. 1976. Dietary fibre redefined. Lancet. 1:967.

USDA. 2009. Japan to ban the use of 125 food additives. GAIN Report

Number: JA9078. [Accessed 21 Oct 2010]. Available from: http://gain.fas.usda.gov/Recent%20GAIN%20Publications/Plan%20to

%20delete%20125%20existing%20food%20additives%20from%20current%20list_Tokyo_Japan_2009-12-28.pdf

Uthurry CA, Suárez Lepe JA, Lombardero J, García Del Hierro JR. 2006.

Ethyl carbamate production by selected yeasts and lactic acid bacteria in red wine. Food Chem. 94:262-70.

143

Valiente C, Arrigoni E, Esteban RM, Amado R. 1995. Grape pomace as a potential food fiber. J Food Sci. 60(4):818-20.

Van de Water J, Naiyanetr P. 2008. Yogurt and immunity: the health

benefits of fermented milk products that contain lactic acid bacteria. In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd

ed. Boca Raton: CRC Press. p 129-164. Varnam AH, Sutherland JP. 1994. Beverages: technology, chemistry

and microbiology. London: Chapman & Hill. 464 p.

Venema K. 2010. Role of gut microbiota in the control of energy and carbohydrate metabolism. Current Opinion in Clin Nutr and Metabolic

Care. 13:432-8.

Verhoef R, Schols HA, Blanco A, Siika-aho M, Rättö M, Buchert J,

Lenon G, Voragen AGJ. 2005. Sugar composition and FT-IR analysis of exopolysaccharides produced by microbial isolates from paper mill

slime deposits. Biotechnol and Bioeng. 91(1):91-105.

Vinderola G, Moreno de LeBlanc A, Perdigón G, Matar C. 2008. Biologically active peptides released in fermented milk: role and

functions. In: Farnworth ER, editor. Handbook of fermented functional foods, 2nd ed. Boca Raton: CRC Press. p 209-241.

Vogel RF, Böcker G, Stolz P, Ehrmann M, Fanta D, Ludwig W, Pot B,

Kersters K, Schleifer KH, Hammes WP. 1994. Identification of Lactobacilli from sourdough and description of Lactobacillus pontis sp.

nov. Int'l J Sys Bacteriol. 44(2):223-9.

Vrancken G, Rimaux T, De Vuyst L, Leroy F. 2008. Kinetic analysis of

growth and sugar consumption by Lactobacillus fermentum IMDO 130101 reveals adaptation to the acidic sourdough ecosystem. Int'l J

Food Micro. 128:58-66.

Wall L, Christiansen T, Schwartz R. 1996. Programming Perl, 2nd ed. Sebastopol, CA: O'Reilly Media. 645 p.

Walstra P, Wouters JTM, Geurts TJ. 2006. Dairy science and

technology, 2nd ed. Boca Raton: CRC Press. 782 p.

144

Wang X, Tong H, Chen F, Gangemi JD. 2010. Chemical

characterization and antioxidant evaluation of muscadine grape pomace extract. Food Chem. 123(4):1156-62.

Wang Y, Li C, Liu P, Ahmed Z, Xiao P, Bai X. 2010. Physical

characterization of exopolysaccharide produced by Lactobacillus plantarum KF5 isolated from Tibet Kefir. Carbohydrate Polymers.

82:895-903. Whistler RL, House LR. 1953. Infrared spectra of sugar anomers. Anal

Chem. 25(10):1463-5.

White JW Jr, Eddy CR, Petty J, Hoban N. 1958. Infrared identifiction of disaccharides. Analytical Chemistry. 30(4):506-510.

Whittaker P, Mossaba MM, Al-Khaldi S, Fry FS, Dunkel VC, Tall BD,

Yurawecz MP. 2003. Identification of foodborne bacteria by infrared spectroscopy using cellular fatty acid methyl esters. J Micro Methods.

55:709-16.

Wie M, Sung J, Choi Y, Kim Y, Jeong H-S, Lee J. 2009. Tocopherols and tocotrienols in grape seeds from 14 cultivars grown in Korea. Euro

J Lipid Sci and Technol. 111(12):1255-8.

Wingender J, Neu TR, Flemming H-C editors. 1999. What are bacterial extracellular polymeric substances. In: Microbial Extracellular

Polymeric Substances: characterization, structure and function. Berlin: Springer. p 1-19.

Woese CR, Fox GE. 1977. Phylogenetic structure of the prokaryotic

domain: The primary kingdoms. Proc Nat'l Acad of Sci USA.

74(11):5088-90.

Woolford MK, Pahlow G. 1998. The silage fermentation. In: Wood BJB, editor. Microbiology of fermented foods, 2nd ed, vol 1. London: Blackie

Academic & Professional. p 73-102.

Xu C, Zhang Y, Wang J, Lu J. 2010. Extraction, distribution and characterisation of phenolic compounds and oil in grapeseeds. Food

Chem. 122:688-94.

145

Zalán Z, Hudáček J, Štětina J, Chumchalová J, Halás A. 2010.

Production of organic acids by Lactobacillus strains in three different media. Eur Food Res Technol. 230:395-404.

Documents

Survival and Exopolysaccharide Production of Lactic Acid