Tissue culture and electrophoretic studies of pineapple (Ananas comosus) and related species


Material Information

Tissue culture and electrophoretic studies of pineapple (Ananas comosus) and related species
Ananas comosus
Physical Description:
xiv, 141 leaves : ill. ; 28 cm.
DeWald, Maria Grazia, 1948-
Publication Date:


Subjects / Keywords:
Pineapple -- Propagation -- In vitro   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1987.
Includes bibliographical references (leaves 130-140).
Statement of Responsibility:
by Maria Grazia DeWald.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000950585
notis - AER2773
oclc - 16928890
System ID:

Full Text







This dissertation is dedicated to my husband, Steve,

for his constant support, understanding, love, and help

through these last years while we were students and parents.


I express gratitude to Dr. Wayne Sherman for his

tireless encouragement, thoughtful guidance, and friendship

during all these years. Special thanks are extended to Dr.

Gloria Moore for allowing the use of laboratory, continuous

support, and suggestions for this dissertation. Appreciation

is extended to Dr. Richard Litz for allowing use of

laboratory and equipment and advice during the tissue

culture studies. Sincere appreciation is given to Dr. Paul

Lyrene and Dr. Walter Judd for their help in reviewing this

dissertation and for the knowledge I gained when attending

their courses. The help of Anne Harper with laboratory work

is appreciated.

Financial support provided by CDCH and Universidad

Central de Venezuela is also acknowledged.

Many thanks go to my parents, Francesco and Antonina

Antoni, for their love and support over the years.

Finally, I extend my appreciation to the graduate

students, the service staff, and the faculty of the

University of Florida Fruit Crops Department, Gainesville,

for the many ways in which they have helped.



ACKNOWLEDGEMENTS ........................................iii

LIST OF TABLES ................. ................................vi

LIST OF FIGURES... ............................ .................viii

KEY TO ABBREVIATIONS.................................. .... x

ABSTRACT.................... .......... ............. ..... xii


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

LITERATURE REVIEW............................. 3

Ananas Taxonomy.................................3
Pineapple Propagation......................... 4
Pineapple Tissue Culture...................... 6
Somaclonal Variation............................8
Application of Isozyme Analysis in
Horticultural Plant Systematics.............12


Materials and Methods.........................18
Genotypes Cultured ............................18
Sterilization and Initiation of Cultures....18
Culture Multiplication......................19
Establishment in Soil........................20
Results and Discussion.........................21


Introduction........................................ 35
Materials and Methods.........................36
Genotypes Used for Electrophoretic Studies..36
Starch Gel Preparation......................38






Protein Extraction and Gel Loading..........40
Electrophoretic Buffers and Staining
Enzyme Activity Staining.....................45
Analysis of Banding Patterns................45
Results and Discussion........................46
Selection of Electrophoretic Buffers
and Enzyme Systems Useful for Studying
Isozymes in the Genus Ananas...............46
Monomorphic Enzymes........................ 48
Enzyme Systems with Variable Banding
Patterns................................... 49
Linkage Analysis for Enzyme Loci............55

BY ISOZYME PHENOTYPES.........................77

Introduction...................... ...........77
Materials and Methods........................78
Results and Discussion.......................78
Cultivar Identification..................... 79
Species Identification...................... 83
Feral Type Identification....................86
Conclusions ..................... ....... ..... 88


Introduction................................ 101
Material and Methods........................103
Results and Discussion.......................104

PINEAPPLE PLANTS ........................... 116

Materials and Methods.......................117
Results and Discussion..................... 118
Electrophoretic Characterization..........118
Morphological Characterization.............119

VIII SUMMARY AND CONCLUSIONS ...................... 124

LITERATURE CITED.........................................130

BIOGRAPHICAL SKETCH............................ ..........141


Table Page

3.1 Murashige and Skoog salts formulation..............23

3.2 In vitro pineapple plantlet production for
each initiated bud................................24

3.3 In vitro pineapple plantlet production for
each initiated bud..................................25

3.4 In vitro pineapple plantlet production in
response to the presence or absence of plant
growth regulators in the culture medium.............26

4.1 Ananas genotypes surveyed for isozyme banding
patterns ............... .......... ... .. ........ 58

4.2 Electrophoretic buffer systems used for isozyme
detection in Ananas...................... ...... 59

4.3 Enzyme activity stains and buffer systems tested
for isozyme detection in Ananas....................60

4.4 Well resolved enzyme activity staining systems
and electroporetic parameters used in Ananas.......61

4.5 Peroxidase and phosphoglucomutase banding
patterns observed for seedlings derived from
an open-pollinated 'Smooth Cayenne' fruit..........62

4.6 Peroxidase and phosphoglucomutase banding
patterns observed for seedlings derived from
an open-pollinated 'Cambray' fruit.................63

4.7 Genotypic ratios and chi-square goodness of
fit values for 6 isozyme loci in pineapple..........64

4.8 Segregation ratios and chi-square goodness of
fit values for independent inheritance of
peroxidase and phosphoglucomutase loci in
'Smooth Cayenne'-derived seedlings..................65

Table Page

4.9 Segregating ratios and chi-square goodness
of fit values for independent inheritance
of peroxidase and phosphoglucomutase loci
in 'Cambray'-derived seedlings.....................66

5.1 Peroxidase and phosphoglucomutase isozyme
banding patterns observed for pineapple
cultivars .......................................... 91

5.2 Pineapple cultivar classification based on
peroxidase banding patterns........................92

5.3 Pineapple cultivar classification based on
phosphoglucomutase banding patterns.................93

5.4 Pineapple cultivar identification based on
peroxidase (Per) and phosphoglucomutase (Pgm)
isozyme banding patterns ..........................94

5.5 Peroxidase and phosphoglucomutase isozyme
banding patterns for Ananas species.................95

5.6 Phosphoglucomutase isozyme banding patterns
observed for various Ananas genotypes...............96

6.1 Designation of genotypes analyzed
electrophoretically... .......................... 109

6.2 Presence-absence data matrix for electrophoretic
analysis of Ananas genotypes......................111

7.1 Regenerate pineapple plants evaluated for
electrophoretic and morphological variation.......122


Figure Page

3.1 Initiation of axillary bud culture..................27

3.2 Proliferation of culture 4 months after

3.3 Actively dividing culture ready for subculture.....29

3.4 Regenerated plantlets established in individual
pots or in flats.............. ....................30

3.5 Comparison of multiplication cultures with
and without plant growth regulators................ 31

3.6 Ananas bracteatus var. tricolor.....................33

4.1 Non-variable isozyme banding patterns observed
for Ananas genotypes..............................67

4.2 Schematic representation of peroxidase isozymes
in 2 pineapple segregating populations.............70

4.3 Peroxidase isozyme patterns observed in pineapple..71

4.4 Schematic representation of phosphoglucomutase
isozyme in Ananas.................................. 72

4.5 Schematic representation of phosphoglucomutase
isozyme in 2 pineapple segregating populations.....73

4.6 Phosphoglucomutase banding pattern in pineapple....74

4.7 Malate dehydrogenase patterns observed in

5.1 Schematic representation of peroxidase (Per)
and phosphoglucomutase (Pgm) isozymes in
pineapple cultivars................................97


Figure Page

5.2 Schematic representation of peroxidase (Per)
and phosphoglucomutase (Pgm) isozymes in
Ananas species......................................98

5.3 Variation in phosphoglucomutase isozyme in
Ananas genotypes leaf extracts.....................100

6.1 Cluster analysis dendogram of 27 pineapple
cultivar isozyme phenotypes........................113

6.2 Cluster analysis dendogram of isozyme phenotypes
of Ananas species and feral types.................114

6.3 Cluster analysis dendogram of the isozyme
phenotypes of 62 Ananas genotypes................115

7.1 Abnormal pineapple regenerated plants............123




















adenosine 5' triphosphate disodium salt

6-benzylaminopurine (N6 benzyladenine)

coconut water

histidine buffer

indole-3-acetic acid

indole-3-butyric acid

K buffer


lithium borate/tris citrate buffer

Murashige and Skoog basal medium formulation

3-(4,5-dimethylthiazol-2-yl)-2,5 diphenyl-

tetrazolium bromide

2-naphthaleneacetic acid

beta-nicotinamide adenine dinucleotide

beta-nicotinamide adenine dinucleotide phosphate

disodium ethylenediaminetetracetic acid

nitro blue tetrazolium

phenazine methosulfate

band migration relative to band sample

band migration relative to borate front

TB: tris borate buffer

TC: tris citrate buffer

Tris: tris-(hydroxymethyl) amino methane

2,4-D: (2,4-dichlorophenoxy)acetic acid

Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




May, 1987

Chairman: Wayne B. Sherman
Cochairman: Gloria A. Moore
Major Department: Horticultural Science (Fruit Crops)

Tissue culture studies were conducted with pineapple

(Ananas comosus) and A. bracteatus var. tricolor in order to

evaluate the feasibility of in vitro techniques as a method

for the rapid multiplication of desirable genotypes. A

system utilizing shake cultures derived from axillary buds

of crowns, slips, and stems was devised. This system

resulted in the efficient production of large numbers of

pineapple plantlets that were easily established in soil

under greenhouse conditions. The regenerated plants were

characterized morphologically and on the basis of isozyme

banding patterns. High levels of phenotypic variability,

especially for leaf morphology, were observed shortly after

the small plants were transferred to the soil. After one

year only 4 variant phenotypes persisted in the 2000

regenerated plants. No isozyme variants were identified.

Electrophoretic characterization of Ananas with

emphasis on economically important pineapple cultivars was

also conducted to help establish phenetic relationships. A

method was developed whereby leaf sap could be quickly

extracted and analyzed for isozyme polymorphisms using

starch gel electrophoresis. One hundred and ninety

combinations consisting of 38 enzyme activity stains and 5

electrophoretic buffer systems were evaluated for isozyme

banding resolution. From this survey, 8 enzyme staining

systems giving good banding resolution were identified.

Three of these, peroxidase, phosphoglucomutase, and malate

dehydrogenase, were polymorphic. Two segregating seedling

populations were utilized to establish a genetic model for

the 3 polymorphic enzymes. The genetic analysis revealed

that both peroxidase and phosphoglucomutase were controlled

.by 3 independent, monomeric loci.

A taxonomic survey of 67 Ananas genotypes was conducted

utilizing the peroxidase and phosphoglucomutase staining

systems. Fifteen of the 27 pineapple cultivars and 5 of the

7 species analyzed could be uniquely distinguished using the

2 enzyme systems. Ananas comosus, one of the species that

could not be unambiguously distinguished, was highly

polymorphic due to the large number of cultivars included in

the species. Ananas monstrosus, the other species without a

unique isozyme phenotype, was confirmed to be a simple crown


mutation of A. comosus that occurs in many cultivars and

therefore is not considered a valid species. Most of the 33

feral types analyzed could be included under species with

similar banding patterns. Phenetic relationships of the

genotypes based on cluster analysis dendograms were



Pineapple is one of the most important fruit crops in

the world. Nearly 12 million metric tons are produced

annually in the tropical and subtropical regions of the

world (FAO Production Yearbook, 1984). It is particularly

well suited to arid and semiarid tropical locations where

few other crops do well.

One reason for the popularity of pineapple is that it

has proved very amenable to large-scale commercial

production and has good canning qualities. Most of the world

production can be attributed to 'Smooth Cayenne' and its

sports. 'Smooth Cayenne' is characterized by its relatively

smooth leaves (spiny tip) and large cylindrical fruit with

good canning qualities. However, it is susceptible to

several economically important diseases and does not produce

slips, which are needed for propagation. Numerous high

quality spiny cultivars also exist that are more adapted to

tropical regions, where 'Smooth Cayenne' does not develop

good fruit qualities, or are better suited for the fresh

market. The presence of spines and the scarcity of planting

material on these cultivars severely restricts their

suitability to large scale planting. Pineapple breeding

efforts have been directed mainly at the improvement of

'Smooth Cayenne', because most of the breeding programs have

been conducted by large commercial companies whose

production is dependent on the specific canning

characteristics of this cultivar.

The genus Ananas, to which pineapple belongs, is poorly

characterized taxonomically. One explanation for this may be

that the center of origin and diversity of this genus is

located in the isolated regions of the upper Amazonian River

basin. This lack of characterization limits the ability of

plant breeders to utilize the variability present in both A.

comosus and related species.

The objectives of this research were 1) to develop an

efficient in vitro propagation system for pineapple; 2) to

morphologically and electrophoretically evaluate the genetic

fidelity of the regenerated plants; 3) to identify

techniques that are useful for visualizing isozymes in

Ananas; 4) to survey the available Ananas germplasm to

determine if variability in isozyme banding patterns exists;

5) to study the possible genetic basis for variability in

isozyme banding patterns; 6) to uniquely identify Ananas

species, including pineapple cultivars, and to categorize

feral types into species using isozyme polymorphism; and 7)

to establish phenetic relationships among Ananas genotypes

based on isozyme banding patterns.


Ananas Taxonomy

Pineapples [Ananas comosus (L.) Merr.] have been

studied and characterized by naturalists and botanists for

hundreds of years; however, there has been much variation in

their taxonomic treatment. Some botanists believe Ananas is

a monotypic genus, i.e., a single species, comprised of many

varieties. Others have considered the genus to include many

species, which are often separated on the basis of a few,

slight character variations (Collins, 1960).

The latest classification of the genus Ananas was by

Smith and Downs (1979). Smith was the botanist who also

classified most of the Bromeliaceae (the Ananas family). The

species considered valid in this latest treatment are A.

ananassoides, A. bracteatus, A. comosus, A. fritzmuelleri,

A. lucidus, A. nanus, and A. parguazensis. Classification is

based upon the presence or absence of a crown (coma

foliaceous), syncarp size, presence and direction of spines,

and floral bract characteristics. This is the most

comprehensive survey of the pineapple genus to date,

providing botanists for the first time with a taxonomic key.

A chronological review of pineapple taxonomy was later

made by Antoni (1983), indicating that the present

classification of the genus Ananas is highly confused and

questioning the validity of some species, e.g., A.

monstrosus. The same author stressed the importance of using

techniques such as tissue culture and electrophoretic

separation of isozymes to differentiate species.

Pineapple Propagation

Pineapple is a self-incompatible perennial crop that is

vegetatively propagated. The crop is usually cultivated in

high density fields. The spineless cultivar 'Smooth Cayenne'

is planted in Hawaii at 60,000 to 80,000 plants per ha. (Py

and Tisseau, 1965). In other regions where spiny cultivars

are used, only 15,000 to 35,000 plants are planted per ha.

(Py and Tisseau, 1965). At such high planting densities,

propagules are a major limiting factor in the establishment

of new fields. The problem can be even more extreme in the

case of newly developed cultivars. The number of propagules

per pineapple plant varies with the cultivar. However, a

typical plant produces between 6-10 propagules for crop

cycle (18-22 months). These propagules include the crown

(produced on the fruit apex), slips (produced on the fruit

peduncle), and suckers (produced on the plant stem). A

solitary sucker is usually left in place to produce the

subsequent ratoon fruit. Culturally, slips are the most

desired propagule. They are more favorable than crowns

because they are uniform, root faster, and can be planted

deeper, giving rise to a more uniform stock of plants.

The most economically important cultivar, 'Smooth

Cayenne', has low slip production. Slip production is

further reduced in this cultivar when it is planted at high

densities to produce a fruit size acceptable for efficient

processing (Ravcof and Yamane, 1970). Crowns are sold with

the fruit in fresh market situations, thereby further

reducing the amount of propagules.

Several methods have been developed over the past 40

years to increase the amount of propagules (Py, 1979). One

method, developed by Macluskie (1939), is to section the

mature stems. The individual pieces are placed in sand boxes

and incubated in a warm moist place. After several weeks the

axillary buds grow and form adventitious roots. This

technique was later tried by several researchers using

sections of crowns and slips (Tkatchenko, 1947; Collins,

1960; Py and Estanove, 1964). A second method consists of

propagating a single leaf with intact axillary bud dissected

from the crown or slip (Seow and Wee, 1970). This method is

laborious and produces a limited amount of plantlets per


The most recent advance in pineapple multiplication

employs plant morphactins (Sanford and Abdul, 1971). Plant

morphactins are a diverse group of synthetic chemicals which

dramatically alter the morphological development of the

plant. Morphactins in pineapple are used in combination with

flower-inducing ethylene compounds, e.g., ethrel. Mature

plants (12-16 month old) are first induced to flower. The

flowering process is then converted into a multiplicant

process by the application of a morphactin, e.g., Maintain

CF-125. Using this method, up to 12 "sliplets" per plant can

be produced (Ravcof and Yamane, 1970). This is an economical

and rapid (2 years) method that allows the production of

over half a million vigorous propagules similar to slips per

ha. Planting material for most large-scale commercial

plantings of 'Smooth Cayenne' types is now produced by this

method of multiplication (Krezdorn, personal communication,


All of these methods require large fields to establish

plants. Rapid and more efficient pineapple propagation

systems still limit the establishment of new commercial

plantings. Tissue culture appears to be a very attractive

alternative method of multiplication and has particular

utility in situations where clonal material is scarce.

Pineapple Tissue Culture

Pineapples have been regenerated in vitro from the

culture of shoot tips, axillary buds, and immature syncarps.

Direct plant production from excised axillary buds was first

reported by Aghion and Beauchesne (1960). A single plantlet

was obtained from each cultured axillary bud on solid, half

strength Knop's medium supplemented with 15% (v/v) coconut

water (CW). Sita et al-. (1974) also regenerated single

plants from individual slip shoot tips on a solid medium of

Knudson major salts and Nitsch minor salts supplemented with

5.4 uM NAA.

Mapes (1974) was the first to describe a method for

multiple shoot production directly from axillary buds

present in cultured shoot. The shoot tips were placed in

liquid Murashige and Skoog (MS) medium supplemented with

0.22 mM adenine to initiate proliferation. Subsequently,

they were transferred to solid MS medium supplemented with

0.15 mM adenine to allow shoot development. The

proliferating mass of shoots was termed a protocorm-like

structure. Similar results were obtained by Teo (1974) and

Pannetier and Lanaud (1976). Wakasa et al. (1978) reported

the initiation of shoots from protocorm-like calli derived

from immature syncarps, axillary buds of suckers and slips,

immature crowns, and immature slips cultured on solid MS

media containing 5.4-54.0 mM NAA and 4.4-44.0 uM BA. The

best results were obtained on medium supplemented with 10.8

uM NAA and 8.8 uM BA. Wakasa et al. (1978) described 2 types

of protocorm-like calli. Calli derived from axillary buds

were described as nodule-like in appearance, while the calli

derived from axillary buds were described as globule-like.

Mathews et al. (1976) demonstrated that liquid culture

medium appeared to be most suitable for shoot proliferation

from axillary bud cultures. They used an MS medium

containing 9.7 uM NAA, 9.8 uM IBA, and 9.2 uM KIN. Shoots

were initiated on solid medium supplemented with 4.4 uM BA

and 4.9 uM IBA (Mathews and Rangan, 1979). Later studies by

these authors indicated that callus could also be initiated

from axillary bud cultures on MS medium supplemented with

54.0 uM NAA, 15% (v/v) CW, and 0.04% (w/v) casein

hydrolysate. Shoot proliferation occurred on MS medium

supplemented with 9.7 uM NAA, 9.8 uM IBA, and 9.2 uM KIN.

Root formation occurred on White's medium containing 0.27 uM

NAA and 0.196 uM IBA. Zepeda and Sagawa (1981) reported

limited shoot proliferation from axillary buds cultured in

liquid MS medium supplemented with 25% (v/v) CW.

Shoot proliferation from hybrid embryo-derived callus

was reported by Srinivasa Rao et al. (1981) on MS medium

with 10.8 uM NAA, 9.8 uM IBA, and 11.0 uM BA. Immature leaf

explants have also been used for shoot proliferation on MS

solid medium containing 9.7 uM NAA, 9.8 uM IBA, and 48.4 uM

BA (Mathews and Rangan, 1979).

Drew (1980) published a report on pineapple tissue

culture theorizing the possibility of obtaining 100,000

plants from a single axillary bud in 12 months. There is no

mention in the literature in regard to the number of

regenerated plants established in soil, with the exception

of Wakasa et al. (1978) who reported growing 400 regenerated


Somaclonal Variation

The term somaclonal variation was coined by Larkin and

Scowcroft (1981) to describe the phenotypic variation

observed in plants regenerated from somatic tissue in vitro.

Somaclonal variation appears to be common to all in vitro

propagation systems. However, the quantity and quality of

the variation differs greatly with the genotype and

propagation protocol.

A number of reviews have been written on this subject

(Bayliss, 1980; Larkin and Scowcroft, 1981; Orton, 1983a;

1983b; 1984; Evans et al. 1984). Agromonically useful

somaclonal variation has been reported in many crop species,

e.g., alfalfa (Johnson et al., 1980), maize (Edallo et al.,

1981), oats (McCoy et al., 1982), potato (Shepard et al.,

1980), sugarcane (Heinz and Mee, 1969), tobacco (Chaleff and

Keil, 1981), tomato (Evans et al., 1984), and wheat (Larkin

et al., 1984).

The variation observed in tissue culture-produced

somaclones may be stable and heritable or it may be

transient, nonheritable, and epigenetic in nature. It is

particularly critical for the scientist to distinguish

between types of phenotypic variation. Epigenetic variation,

because it is nonheritable, has limited usefulness to

cultivar improvement.

A wide range of chromosomal alterations are possible in

tissue culture-regenerated plants. Single base pair

alterations or deletions (Edallo et al., 1981) represent one

end of the spectrum, while major karyotypic alterations,

including the doubling or deletion of complete sets of

chromosomes (Janick et al., 1977), represent the other


Stable, heritable variation implies genetic

variability. This genetic variability may be due to

pre-existing heterogeneity present in the cells of the

explant and/or de novo genetic alterations occurring during

the culturing process (Evans et al., 1984). Unfortunately,

scientists have not been able to determine the relative

amount of variation attributable to these 2 sources (Larkin

and Scowcroft, 1981).

The full range of genetic variability that can be

generated and expressed in tissue culture-regenerated plants

has not yet been established. It does appear, however, that

the full range of genetic variability present in cultured

cells is not present in the regenerated plants (Edallo et

al., 1981). The regeneration process may be serving as a

selective screen for those cells that are genetically

competent to form plants.

Perhaps the best documented source of variation is the

explant. Barbier and Dulieu (1980) observed higher levels of

somaclonal variation in tobacco plants that were regenerated

from protoplasts and newly initiated callus cultures in

comparison with plants regenerated from well established

callus cultures. Wakasa (1979) also reported significant

differences in the amount of somaclonal variation observed

for vegetative morphology in 'Smooth Cayenne' pineapple

plants regenerated from 3 explant sources. He found that

plants regenerated from syncarp (immature fruit) tissues

exhibited 100% variability, while plants derived from

axillary buds of crowns and slips were 34% variable. The

lowest levels of variation (7%) were observed in plants

regenerated from small crowns. Varibilities in leaf wax,

color, spines, and foliage density were observed in the

plants derived from the syncarp, while only spine

variability was observed in the plants from slips and

axillary buds. It should be noted that this study was done

on up to 2-year-old regenerated plants that were grown in

the greenhouse, and no long-term study on the stability of

this variation was conducted.

The amount of time in culture can also influence the

variability of regenerated plants. Karyotypic changes have

been correlated with increasing time in culture (Bayliss,

1980). Increasing time in culture has been shown to increase

somaclonal variation in tobacco (Syono and Furuya, 1972),

Chrysanthemum (Bush et al., 1976; Sutter and Langhans,

1981), and oats (McCoy, 1980 cited by Reisch, 1983). In

contrast, long-term cultures of corn were reported to be

stable (McCoy, 1980 cited by Reisch, 1983).

The phytohormones used in the culture medium are

another possible factor affecting somaclonal variation. High

concentrations (18 uM) of the auxin 2,4-D were reported to

produce unique variation in regenerated barley plants that

was not observed in plants regenerated using lower

concentrations (Deambrogio and Dale, 1980). Similarly, a

medium supplemented with NAA (27 uM) and KIN (0.5 uM) was

shown to produce significantly more karyotypic alterations

in regenerated Haworthia plants than a medium supplemented

with 0.6 uM IAA (Ogihara, 1981).

The genotype has also been shown to influence

somaclonal variation. Liu and Chen (1976) evaluated 8

sugarcane cultivars and found that regenerated plants from

the cultivars ranged from 1.8% variable to 34% variable.

Skirvin and Janick (1976) found somaclonal variation

cultivar differences in Pelargonium sp. DeWald and Moore

(1987) indicated in their review that certain crops, e.g.,

polyploids and asexually propagated crops in which sexual

fertility is not important, appear to tolerate higher levels

of chromosomal variation produced via somaclonal variation

than other crops.

Applications of Isozyme Analysis in Horticultural Plant


Isozymes are multiple molecular forms of an enzyme

having the same substrate specificity (Markert and Moller,

1959). Hunter and Markert (1957), using the starch gel

electrophoresis technique developed by Smithies (1955),

demonstrated that isozymes can be visualized on starch gels

stained with specific histochemical stains. Bands appearing

on a starch gel correspond to different alleles and/or loci

coding for a specific enzyme.

Isozymes have been used extensively as genetic markers

in plants. They have a number of advantages over

conventional markers. Isozymes are generally not affected by

the morphological or physiological status of the plant. They

interact codominantly at a single locus, rarely exhibit

epistatic interactions, and can be detected in a variety of

tissues (Tanksley and Rick, 1980; Tanksley et al., 1982;

Tanksley, 1983).

The use of electrophoretic analysis to characterize

isozyme variation as an aid for plant systematics is now

widely applied and has been the subject of several reviews

(Gottlieb, 1977; Crawford, 1983). The electrophoretic

phenotype can be directly equatable with genotype, unlike

many morphological characters, which may demonstrate a great

deal of environmental or developmental plasticity (Crawford,

1983; Gottlieb, 1977).

In plant systematics, isozyme electrophoresis has been

successfully utilized for both cultivar and species

identification by providing a way to quantitatively assess

the genetic variation among conspecific populations and the

genetic divergence among congeneric species. Other important

applications include the resolution of polyploid taxa and

insight into their origins, the investigation of the origin

of cultivated crops, and the testing of broad systematic or

evolutionary hypotheses.

Reliable identification of cultivars using classical

methods based on morphological and physiological characters

has become increasily difficult, partially due to the large

number of lines or varieties now available. Isozymes as

bio-chemical markers for cultivar identification have proven

reliable, consistent, and essentially unaffected by

environmental conditions (Bailey, 1983). They can be used as

"fingerprints" to aid in the visual identification of

cultivars (Pierce and Brewbaker, 1973). The utility of this

technique has already been demonstrated in a number of

species, e.g., Anthurium (Kobayashi et al., 1987), alfalfa

(Quiros, 1980), apple (Menendez et al., 1986; Weeden and

Lamb, 1985), avocado (Torres et al., 1978a), barley

(Bassiri, 1976), bean (Bassiri and Rouhani, 1977; Weeden,

1984), Camellia (Wendel and Parks, 1983), Citrus (Torres et

al., 1978b), grape (Schwennesen et al., 1982), grasses

(Wilkinson and Beard, 1972), maize (Cardy and Kannenberg,

1982), Musa (Jarret and Litz, 1986b), peach (Carter and

Brock, 1980), pear (Santamour and Demuth, 1980), pecan

(Mielke and Wolfe, 1982), poinsettia (Werner and Sink,

1977), rice (Sarkar and Bose, 1984), strawberry (Bringhurst

et al., 1981), and wheat (Menke et al., 1973).

The use of isozymes in the investigation of plant

taxonomy, particularly species circumscription is relatively

recent. Zymogram techniques have been used most extensively

in Nicotiana. Characteristic enzyme patterns for most

Nicotiana species have been observed using proteins (Hart

and Bathia, 1967), peroxidase alone (Sheen, 1970), and in

combination with esterases (Smith et al., 1970). Species

identifications using isozymes have also been reported in

Datura (Conklin and Smith, 1971), Citrus (Button et al.,

1976), Musa (Jarret and Litz, 1986c), and Pistacia (Loukas

and Pontikis, 1979).

Specific isozymes or proteins have been used to

identify particular genomes in polyploids. In bananas,

Jarret and Litz (1986a) confirmed the bispecific origin of

the triploid 'Chato' from Musa acuminata and Musa balbisiana

using leaf isozyme polymorphisms. An allopolyploid origin

for the apple has been proposed based on inheritance data of

7 pollen enzymes (Chevreau et al., 1985). Isozyme analysis

of Gossypium hirsutum, a natural allopolyploid, has shown

that esterases exhibit greater intergenomic variability than

intraspecific or intravarietal variability (Cherry et al.,

1970). In a later study, Cherry et al. (1971) showed that

evolutionary changes had occurred in the species that were

reflected in isozyme differences.

One of the most common uses of isozymes has been to

assess the validity of proposed parentages of cultivars and

progeny. This has been done in apple (Chyi and Weeden, 1984;

Weeden and Lamb, 1985), avocado (Golching et al., 1985;

Torres and Bergh, 1980; Torres et al., 1978a), Citrus

(Torres et al., 1978b), and date palm (Torres and Tisserat,


Protein and isozyme analyses have proved particularly

helpful in deducing systematic relationships between groups

where morphological and cytological data have been


inconclusive. Electrophoretic studies have been used

successfully to establish phylogenetic relationships in

rice, wheat, barley, soybean, and cotton (Ladizinsky and

Hymowitz, 1979). More recently, similar studies have been

conducted with Amaranthus (Hauptli and Jain, 1984) and

Capsicum (Panda et al., 1986).



Pineapple is routinely propagated vegetatively by means

of lateral shoots, basal suckers, or crowns. Plant material

is often limited, especially for the 80,000 plants/ha needed

in the establishment of new plantations, and for the

propagation of improved cultivars or newly discovered


In vitro propagation of pineapple was first achieved in

1960 by Aghion and Beauchesne (1960), but it was not until

1974, with the simultaneous reports of Laksmi Sita et al.

(1974) in India, Teo (1974) in Malaya, and Mapes (1974) in

Hawaii that in vitro production of pineapple was considered

as a commercial alternative. Most recent reports (Pannetier

and Lanaud, 1976; Wakasa et al., 1978; Mathews and Rangan,

1979; Drew, 1980; Zepeda and Sagawa, 1981) deal only with

the factors involved in the establishment of axillary buds

in culture. They do not consider the number of plants

regenerated or the efficiency of the system for reduced

labor cost. Furthermore, high levels of variability were

observed in regenerated 'Smooth Cayenne' plants by Wakasa

(1979). Such variability would compromise the commercial use

of in vitro propagation of pineapple, since regenerated

plants could not be assured to be clonal in nature.

This study was conducted using several species and

cultivars to improve the technique of tissue culture of

axillary buds for rapid pineapple propagation and to

investigate the possible variation occurring in the

regenerated plants.

Materials and Methods

Genotypes Cultured

Two cultivars of the common pineapple Ananas comosus

(L.) Merr. were used: 'PR-1-67', a spiny commercial cultivar

grown for fresh fruit and canning in Puerto Rico and

'Perolera', a smooth-leaved cultivar grown in the Northen

Andes of South America and used for fresh fruit. The third

genotype was Ananas bracteatus var. tricolor Smith, a

small-fruited ornamental characterized by showy colored

bracts and tricolored leaves (Fig. 3.6A). Two plants of

PR-1-67 and 1 plant of the other 2 genotypes were used in

order to determine intracultivar and intercultivar

variability in response to tissue culture.

Sterilization and Initiation of Cultures

Crowns and stems were rinsed in water, defoliated and

surface-sterilized by agitating in a 20% sodium hypochlorite

solution with 2-3 drops of a surfactant (Tween 20 or Tween

80) for 20 min, followed by 3 rinses of 10 min each in

sterile deionized water. Terminal and axillary buds were

then excised aseptically and again surface-sterilized in a

2% sodium hypochlorite solution for 10 min followed by 3

more 10 min rinses in sterile deionized water (Fig. 3.1A).

Buds were inoculated onto a modified Murashige and Skoog

(1962) medium (Table 3.1), supplemented with 3% sucrose,

0.8% Difco Bacto-agar, 0.57 mM inositol, 1.2 uM thiamine

HC1, 10.8 uM NAA and 8.8 uM BA. Growth regulators were added

and the medium pH was adjusted to 5.7 prior to autoclaving

at 121 C, and 1.1 kg cm for 20 min. The medium was poured

into 60 X 15 mm sterile Petri dishes for initial bud

explants or into test tubes for subculture (Fig.3.1B).

Cultures were incubated at room temperature (24-27 OC)
-1 -2
with a 16 hr photoperiod of 76 umol s m provided by

cool-white fluorescent lamps. Explants were subcultured onto

fresh medium in test tubes at 6 week intervals until

adequate growth had occurred for transfer to liquid medium

(Fig. 3.2).

In the experiment described, 3 buds from each of the 2

plants of 'PR-1-67' and 1 plant of 'Perolera' and Ananas

bracteatus var. tricolor were established in this manner.

Culture Multiplication

Proliferating explants were multiplied in liquid

cultures consisting of 50 ml of the medium described above

in a 125 ml Erlenmeyer flask, maintained at 100 rpm on

gyrorotary shakers under 16 hr photoperiod (76 umol s- m 2)

provided by fluorescent lamps. Liquid shake cultures were

subcultured at approximately 4 week intervals, when flasks

had become tightly packed with plantlets (Fig. 3.3). In this

way, numerous flasks of callus and regenerating plants were

obtained from one original bud.

An experiment was conducted to determine the necessity

of transferring the callus and plantlets to a medium without

hormones to induce root formation and thus plant survival,

and to evaluate the influence of plant growth regulators on

the number and size of plants produced. Ten grams of an

actively dividing 'Perolera' culture were inoculated into

flasks containing 50 ml of basal medium with or without

hormones (10.8 uM NAA and 8.8 uM BA). The 2 treatments were

replicated 9 times.

Establishment in Soil

After each subculture plantlets 2.5 cm and larger were

harvested. They were transferred to individual pots or to

flats containing a commercial soil mixture and enclosed in

plastic bags, creating a greenhouse effect (Fig. 3.4). The

plants were incubated in a growth chamber at 28 C under

fluorescent lamps and gradually hardened off by removing the

plastic covers. Acclimated plants were transplanted to

larger pots and placed in a greenhouse under a black cloth

screen to prevent scorching. The screen was removed 3 weeks

later and surviving plants were counted. Fertilization was

initially done biweekly with an acid-forming plant

fertilizer and later on with a solution of 1/10 MS salts as

apparent nutrient deficiencies developed. Phenotypic

observations of the plants were made periodically.

Results and Discussion

Production of plantlets from axillary bud cultures

initiated on solid medium and multiplied in liquid shaker

cultures started 9 months after explanting. The total number

of harvestable plantlets doubled with each monthly

subculture after the 11th month (Table 3.2). From 300 to 350

plantlets could be obtained by the 13th month from a single

bud culture (Table 3.3). Approximately 25 plantlets could be

harvested from each 125 ml flask with each additional

subculture. In all genotypes, plantlets larger than 3.0 cm

had a survival rate of nearly 100% when transferred to soil.

A 2-fold increase in the total number of harvestable

plantlets was obtained in 6 weeks when plant growth

regulators were removed from the multiplication medium

(Table 3.4; Fig. 3.5). However, after 2 subcultures (14

weeks), all plantlets had been harvested and no further

callus production or multiplication occurred. Further,

transferring to a medium without plant growth regulators to

induce root formation was not necessary for plantlet

survival. Plantlets without roots produced in multiplication

medium with plant growth regulators also had a high survival

rate and, while initially smaller than plantlets produced in

hormone-free medium, attained the same level of growth after

2 months in soil. The cultures periodically transferred into

multiplication medium with plant growth regulators appear to

be capable of indefinite multiplication. Phenotypic

variation using leaf morphology as a marker was assessed in

approximately 2000 regenerated plants and will be discussed

in detail in Chapter VII.

Tissue culture of the species A. bracteatus var.

tricolor produced regenerated plants that were either

tricolor, green or albino (Fig. 3.6). This indicates the

chimeral origin of this genotype. It is clear that these

variegated plants should not be given formal varietal rank.

Table 3.1. Murashige and Skoog salts formulation.

Salts mg 1-1

Major Salts
NH NO3 1650.00
KNO 1900.00
MgS 4.7H20 370.00
CaCl 440.00
KH2 P4 170.00

Minor Salts
Na EDTA 37.30
FeSO4.7H O 27.80
MnSO4.4H O 22.30
ZnSO4.7H O 8.60
H BO 6.20
K 10.83
NaMoO ..2H 0 0.25
Co0S4.6H20 0.025
CuSO4.5H20 0.025



~.O C'4

1 V



4 r



4 -4


S00 -4



1 CC

-4 Mn-4












z 0 c

oz -H -H
r-4 4-J
4-' Cu k
Z -rq 'o

U 0 -H 4J

a) .1 a)
U0 44 r
-W u
LU )H a) 4

a) 4J
c-. 0. (0 0

W 4-4 444

m rI 4 r4

4J a) ) a)

a)-4 4- )
~ uU

tIH i-I xI 3
ka) u* U
a) ~ ow

a) -4r

a a

-4 ~ ) a)

N >iX :

0) r- '
C"4 0 LO
co r- r-


(N Cr N-

CN M -1

(N 0n

M~ M


N r- r
a) '0 0 '
a4 4
o -

p-I p-


















0 .q

a )


> .-r

QC (






r x
-I- a)

.H- 4-H

c4) r


N4 -4
4-) 2o
.4c) CU
cU C

W~ U

9 2
0O (U
00 Q)

U a)

a) 0

r U

a) N
> a)
4 >


CU a)





U 0H



4-1 (
C-q 4c
C -H




0 0
u a)




r- m


U) -.,-

0* a 9

0 C4
O r4 C

CO 0--
E *

C 4-)

a O U UO

r-I, .i ( U-

z 4
0 o< 0

r-1 < -H 0

N >1 x :

moCDMnooao -q VN



Figure 3.1. Initiation of axillary bud culture. A. Buds
aseptically removed; B. Bud expansion 8 days
after inoculation.

Figure 3.2. Proliferating culture 4 months after
inoculation and ready to be transferred
to liquid multiplication medium.


Figure 3.3. Actively dividing culture ready for
subculture. Plantlets larger than
2.5 cm will be transferred to soil.




U) U)
"-H 4-4



(0 E



4) u




Figure 3.5. A. Comparison of multiplication cultures
with (right) and without (left) plant
growth regulators.


Figure 3.5.

continued. B. Plantlets produced in medium
with plant growth regulators; C. Plantlets
produced in medium without plant growth


Figure 3.6. A. Ananas bracteatus var. tricolor.



Figure 3.6. continued. B. In vitro multiplication showing
resolution of chimera; C. Regenerated plants.
Albino plants are incapable of survival in



Electrophoretic characterization of plants based on

differential patterns of protein migration is a powerful

technique to separate and classify genotypes. Isozyme

analysis has aided taxonomists, breeders, and physiologists

in understanding the underlying causes of phenotypic

variation. The plant breeder is particularly interested in

classification and characterization studies in order to

better utilize existing germplasm resources. The purpose of

this study was to determine what enzyme and buffer systems

work well in Ananas and which enzymes show polymorphism. The

analysis was conducted with taxonomy, breeding, and cultivar

improvement in mind.

No previous reports of isozyme studies have been

published for pineapple. For this reason, an initial study

was undertaken using a small number of Ananas genotypes to

determine which enzyme staining systems would give well

resolved banding patterns. Horizontal starch gel

electrophoresis was used for isozyme characterization

because it is non-toxic and yields multiple slices from a

single gel, thus allowing the rapid screening of a large

number of enzyme staining systems. The procedures followed

were similar to those currently being used with Citrus by

Moore (personal communication, 1985). The enzyme staining

systems that resulted in well resolved banding patterns were

used to survey a larger number of Ananas genotypes to

identify and characterize isozyme polymorphism. Genetic

studies were performed using seedling populations to

determine the basis of variability observed among genotypes.

Information from these studies was used to formulate a

system by which cultivars, feral types, and species could be

identified and distinguished, and their phenetic affinities

could be established. These systems will be described in

Chapters V and VI. Selected enzyme staining systems were

used to determine phenotypic stability of tissue

culture-regenerated plants and will be discussed in Chapter


Materials and Methods

Genotypes Used for Electrophoretic Studies

A list of the genotypes used in this study is given in

Table 4.1. The population included all of the species

presently considered valid in the genus Ananas (Smith and

Downs, 1979; Antoni, 1983) with the exception of A.

fritzmuelleri, which was not available in any of the major

world collections. The A. comosus cultivars surveyed

represent the most important cultivars in use world-wide.

The Ananas genotypes indicated by numbers are introductions

made by the author and originally established at the

Facultad de Agronomia, Universidad Central de Venezuela, or

material from the collection at the USDA Subtropical

Horticultural Research Unit, Miami, Florida. These feral

types were collected in the regions of southern Venezuela

and northern Brazil, an area that has been suggested as the

center of origin for the genus Ananas (Leal and Antoni,

1980). The suspected taxonomic classification of the feral

types follows the introduction number. The identification

was based on the botanical characters considered valid for

each Ananas species by Smith and Downs (1979). High levels

of phenotypic variation were present for several characters

among the collected types, including presence or absence of

spines, leaf color, and fruit shape, size, and color.

Genetic studies were based on isozyme analysis from

both the Ananas germplasm survey and 2 seedling populations.

The seedling populations were obtained from a commercial

planting 'of 'Smooth Cayenne' and 'Cambray' growing in

adjacent fields in northwestern Ecuador. Since pineapple has

only been reported as self-incompatible within a cultivar

and cross-compatible between cultivars (Brewbaker and

Gorrez, 1967), these seedling populations were assumed to be

2 segregating populations representing reciprocal crosses

between 'Smooth Cayenne' and 'Cambray'. The seeds were taken

from mature fruit, rinsed in water, and air dried. Seeds

were scarified by a treatment of 30-60 sec in concentrated

sulphuric acid followed by a rinse in sterile water and were

planted in covered petri dishes containing a sterilized soil

mixture moistened with sterile water. Periodic sprays with a

1% Captan solution were made to avoid fungal contamination.

Germinated seeds were transferred to pots containing a

commercial soil mix composed of sphagnum peat moss,

horticultural grade vermiculite and perlite, composted pine

bark, and washed granite sand. The pots were covered with

polyethylene bags to prevent dessication. Transplants were

initially held in a growth chamber at 27 oC with 16 hr

-1 -2
fluorescent light (76 umol s m ), and were taken to the

greenhouse after acclimation.

All plants of the collection were kept at Gainesville,

Florida, under greenhouse conditions during the cold months

(Nov.-Apr.) and outside the rest of the year. Plants were

kept in 25 cm pots with the same soil mixture as the

seedlings and were periodically fertilized with a standard

water-soluble 20-20-20 fertilizer.

Starch Gel Preparation

A 10% (w/v) starch gel consisting of a 2:1 (w/w) mixture

of Sigma and Connaught hydrolyzed potato starch was used

throughout this study. All the starch was from the same lot

number. Thirty grams of the starch mixture along with 300 ml

of a gel buffer solution (Table 4.2) were poured into a 1000

ml side arm flask. The mixture was swirled vigorously by

hand and placed on a magnetic stir pad at medium speed for

10 min to assure a well-mixed suspension and to avoid the

formation of lumps in the starch. The flask was then placed

on a magnetic hot plate at high heat and stirred at medium

speed. The solution turned translucent as it reached maximum

viscosity. At this point, the flask was again swirled by

hand and the cofactor NADP (10 mg dissolved in 1 ml water)

was added for those staining systems requiring it. The flask

was heated again until large air bubbles appeared on the

surface of the solution.

The solution was degassed immediately by vacuum before

it started to cool, to avoid air bubbles forming in the gel

that could interfere with the migration of proteins. The

vacuum source was provided by attaching an aspirator to a

sink faucet and to the flask side arm. The solution was

ready to be poured when no small bubbles were present.

The molds used had inside dimensions of 19 x 19 x 0.5 cm

and were made from plexigas sealed with methylene chloride.

Any bubbles that formed while pouring the starch solution

into the molds were removed with a spatula before the gel

started to harden. The gel was allowed to cool for

approximately 30 min at room temperature, and then further

cooled by refrigeration (4 C) for 1 hr prior to loading the

samples. Gels could be kept for 12-16 hr at room temperature

before being placed in the refrigerator and they were

covered with polyethylene film to prevent dessication.

Protein Extraction and Gel Loading

Leaf sap was used as the main protein source throughout

this study because it was easily prepared, readily

available, and gave good results. All leaf samples were

taken from the fourth leaf at the center of the plant,

counting from the first visible leaf, and were used

immediately or stored in plastic bags at 4 C for later use.

Samples were used within 7 days of collection, although no

decrease in enzyme activity could be detected for up to 3

weeks in storage. Leaf sap was expressed from the proximal

portion of the adaxial surface of the lamina, using a pestle

and a filter paper wick (Whatman No. 3, 4 x 8 mm). The white

waxy epidermis and underlying layer of fibers were scraped

from the leaf surface to aid in leaf sap absorption into the

wick. The saturated wick was then inserted into a slit 5 cm

from the top (cathodal side) of the chilled gel for anodal

migrating enzymes and at 8 cm from the top for cathodal

migrating enzymes, e.g. peroxidase. Once all samples had

been inserted, glass rods were placed at the top and bottom

of the gel to ensure good contact between wicks and gel

during electrophoresis.

Pollen was collected from flowering plants just prior

to anthesis, dried at room temperature for 24 hr, and frozen

at -4 C until used in electrophoretic assays. Proteins were

extracted from the pollen u-sing the method outlined by

Weeden and Gottlied (1980). A small amount of frozen pollen

was placed in 0.1 ml of extraction buffer for 4 hr. The

supernatant was then absorbed into the wick.

Electrophoretic Buffers and Staining Systems

Five electrophoretic buffer systems were tested:

histidine citrate (H), lithium borate/tris citrate (LBTC),

tris borate (TB), tris citrate (TC), and K buffer (K) [Table

4.2], as were 38 enzyme activity stains. Tested combinations

are listed in Table 4.3.

The formulations used for the enzyme staining solutions

were derived from several authors. Endopeptidase (E.C.

3.4.22.?) was according to Melville and Scandalios (1972).

Formate dehydrogenase (E.C. was from Wendel and

Parks (1982). Creatine kinase (E.C. and hexokinase

(E.C. were as described by Shaw and Prasad (1970).

Malate dehydrogenase (E.C. was from Cardy et al.

(1981). The following were according to Vallejos (1983):

adenylate kinase (E.C., alcohol dehydrogenase (E.C., aldolase (E.C., alpha-amylase (E.C., ascorbate oxidase (E.C., catalase (E.C., diaphorase (E.C., fumarase (E.C., galactose dehydrogenase (E.C., beta-D-

galactosidase (E.C., using alpha-napthyl-D-

galactopyranosidase as a substrate, glucose-6-phosphate

dehydrogenase (E.C., beta-D-glucosidase (E.C., glutathione reductase (E.C., isocitrate

dehydrogenase (E.C., laccase (E.C.,


lactate dehydrogenase (E.C., 6-phosphogluconate

aenyarogenase (E.C. .i.i.'i),


&.iiilbe (iC;., shikimate dehydrogenase (E.C., triose

phosphate isomerase (E.C., urease (E.C.,

and xanthine dehydrogenase (E.C. In addition to

these, the following staining systems were modified from

Vallejos (1983) as follows:

Acid phosphatase (E.C.
100 mg alpha-napthyl acid phosphatase, Na salt
50 mg Fast Black salt
100 mg 0.1 M sodium acetate, pH adjusted to 4.7
with glacial acetic acid



phosphatase (E.C.
mg beta-naphthylacid phosphatase, Na salt
mg Fast Blue RR salt
mg 0.1 M tris HC1, pH 8.5

mg Fast Blue RR salt
ml H 0
ml 0.2 M NaH2PO
ml 0.2 M Na2 HP
ml of the following solution: 1 g alpha-
napthyl acetate and 1 g beta-napthyl acetate
dissolved in acetone and brought to 100 ml
with H20

Glutamate oxaloacetate transaminase (Aspartate amino-
transferase, E.C.
500 mg L-aspartic acid
70 mg alpha-keto-glutaric acid
50 mg pyridoxal-5'-phosphate
200 mg Fast Blue BB salt
100 ml 0.1 M tris HC1, pH 8.0

Leucine aminopeptidase (E.C.
40 mg L-leucine-beta-napthylamide HC1
100 mg Black K salt
50 ml 0,2 M tris + 0.2 M maleic anhydride
20 ml 0.2 N NaOH
30 ml H20

Phosphoglucomutase (E.C.
500 mg glucose-1-phosphate, Na2 salt
20 mg NADP
30 mg MTT
4 mg PMS
100 mg MgC1
100 ml 0.1 M tris HC1, pH 8.0
40 units glucose-6-phosphate dehydrogenase

Phosphohexose isomerase (Glucophosphate isomerase,
100 mg fructose-6-phosphate, Na2 salt
10 mg NADP
20 mg MTT
4 mg PMS
100 mg MgC1
100 ml 0.1 M tris HC1, pH 8.0
10 units glucose-6-phosphate dehydrogenase

Malic enzyme was modified from Cardy et al. (1981);

peroxidase and superoxide dismutase were modified from Moore

(unpublished) and Moore and Collins (1983), as follows:

Malic enzyme (E.C.
30 mg NADP
30 mg NBT
4 mg PMS
100 mg MgCI
20 ml 0.1 A malic acid, pH adjusted to 7.0
with NaOH
1 ml 0.2 M NaH2PO4
48 ml 0.2 M Na2HPO4
30 ml H20

Peroxidase (E.C.
250 mg para-phenylenediamine
50 mg MnSO4
5 ml 1.0 M sodium acetate, pH adjusted to 4.7
with glacial acetic acid
30 ml 95% ethanol
65 ml H 0
0.5 ml 30% H 02 added after above solution is
poured onto gel

Superoxide dismutase (E.C.
2 mg Riboflavin
75 mg Na EDTA
10 mg NBE
100 ml 0.1 M tris HC1, pH 8.0


The gel with sample wicks in place was set on top of 2

plexiglass buffer tanks. The electrode consisted of a 19 cm

length of platinum wire. Five hundred milliliter of buffer

solution was added to each electrode buffer tank, covering

the electrode. The bridges between the electrode buffer and

the gel consisted of 4-ply chromatography paper (Whatman No.

3). A glass plate was placed on top of the gel to assure

good contact between the gel and electrode bridges.

Electrophoresis was carried out under constant voltage

in a refrigerator at 4 0C. Power settings and running time

varied with each buffer and enzyme staining system. Gels

were typically run for the initial survey of buffers at 100

to 180 volts and approximately 20 milliamps for 16 to 18 hr

or at 200 to 250 volts and 25 to 50 milliamps for 4 to 6 hr.

Protein migration was monitored using several techniques. In

the gels containing borate buffers, a light brown line was

visible that migrated anodally during electrophoresis. In

gels using K buffer, a sample wick saturated with

bromophenol blue was used to follow the progress of

electrophoresis. Protein migration was monitored in gels

using histidine buffers by noting a pink color in the anode

electrode buffer solution. Protein migration could also be

monitored by a decrease in current readings when using a

constant voltage source during electrophoresis.

Enzyme Activity Staining

Gels were sliced horizontally into 4 slices (2 mm

thick), the top slice was discarded. The gel slices were

placed in plastic boxes and refrigerated until staining

(usually less than 2 hr).

Staining solutions were prepared fresh, poured over the

gel slice and swirled lightly to eliminate air bubbles

trapped underneath. Gels were placed in the dark and

observed every 15 to 30 min to follow the progress of

staining. When optimum resolution was observed, the gels

were rinsed in deionized water and examined. Gels were fixed

for long-term storage using a (7:7:1 v/v/v) solution of

methanol, deionized water, and glacial acetic acid. The gels

were then wrapped in polyethylene film and refrigerated.

Analysis of Banding Patterns

The banding patterns (zymograms) were recorded by

measuring the migration rate of each band. An internal

standard was created in each gel system to calculate

relative migration rates. The borate front produced in the

TB and LBTC buffer systems and the bromophenol blue band in

the K buffer systems served as an internal standard for

these systems. Buffer systems that did not produce a front

or a bromophenol blue band (H and TC) were evaluated by

placing a sample with a unique banding pattern in the gel,

one band of which served as the internal standard. The

migration of the band relative to the front (Rf value) or to

the unique band (Rb value) was calculated by dividing the

band migration distance by the migration distance of the

internal standard.

A genetic mechanism to explain variability was

hypothesized when zymogram polymorphism was observed among

the genotypes. The 2 segregating populations available were

used to test the genetic mechanism by comparing the observed

ratios with the expected ratios using chi-square analysis.

The possibility of genetic linkage was investigated by

analyzing the segregation ratios at more than 1 locus.

Linkage between loci could be suspected if chi-square

analysis revealed a significant deviation from ratios

expected with independent segregation.

Results and Discussion

Selection of Electrophoretic Buffers and Enzyme Systems
Useful for Studying Isozymes in the Genus Ananas

The results of the survey testing 38 enzyme staining

systems with 5 buffers can be seen in Table 4.3.

Well-resolved banding patterns were obtained with 8 enzyme

activity stains: isocitrate dehydrogenase, malate dehydro-

genase, peroxidase, phosphoglucomutase, 6-phosphogluconate

dehydrogenase, phosphohexose isomerase, superoxide dismu-

tase, and triose phosphate isomerase. Table 4.4 lists these

enzyme staining systems along with important electrophoretic

parameters. These systems were selected for further study

and are detailed below.

Twelve additional enzyme systems showed limited

resolution or enzyme staining activity and were grouped into

3 categories: those with activity but no discrete banding,

those with poorly resolved banding and no apparent

variability, and those with poorly resolved banding but some

apparent variability. Enzyme staining systems giving no

discrete banding included beta-D-galactosidase, format

dehydrogenase, and urease. Systems giving poorly resolved

banding with similar activity among genotypes were esterase

and glutathione reductase. Staining systems giving poorly

resolved banding patterns with some variable activity among

genotypes included acid phosphatase, glucose-6-phosphate

dehydrogenase, glutamate oxaloacetate transaminase, leucine

aminopeptidase, malic enzyme, 6-phosphogluconate dehy-

drogenase, and shikimate dehydrogenase. Although 2

additional protein extraction buffers were tried with these

enzyme staining systems, no improved resolution was


No activity was detected for the following 18 enzymes:

adenylate kinase, alcohol dehydrogenase, aldolase, alkaline

phosphatase, alpha-amylase, ascorbic oxidase, catalase,

creatine kinase, diaphorase, endopeptidase, fumarase,

galactose dehydrogenase, beta-D-glucosidase, hexokinase,

laccase, lactate dehydrogenase, pyruvate kinase, and

xanthine dehydrogenase.

The 8 enzyme activity stains that gave well-resolved

banding patterns were used to survey the 62 genotypes listed

in Table 4.1. In 5 of the staining systems, monomorphic

banding patterns were obtained. The other 3 staining systems

gave polymorphic banding patterns.

Monomorphic Enzymes

These enzymes included isocitrate dehydrogenase,

phosphohexose isomerase, superoxide dismutase and triose

phosphate isomerase. Banding phenotypes observed for these

monomorphic systems are illustrated in Fig. 4.1. and are

described below:

Isocitrate dehydrogenase. Two anodally migrating bands

were observed with this stain, 1 at 2.0 and the other at 2.5

cm from the origin (Fig. 4.1A). The second band appeared

less intense and was not always detected. No variability was

observed among the genotypes studied.

6-Phosphogluconate dehydrogenase. The banding pattern

obtained is shown in Fig. 4.1B. Two anodally migrating

bands, 1 major and 1 minor, were observed for all genotypes

investigated. The major band exhibited a mobility of 2.3 cm

from the origin. The fainter minor band migrated 1.1 cm from

the origin. Fainter poorly resolved bands were also present.

Phosphohexose isomerase. The banding pattern for this

enzyme consisted of 3 bands (Fig. 4.1C). The first and

second bands stained more intensely and migrated at 2.3 and

2.6 cm, respectively, from the origin. The third band, which

migrated at 3.0 cm, usually stained less intensely and was

not always detected. Another unresolved zone of activity was


Superoxide dismutase. The banding pattern is

illustrated in Fig. 4.2D. A monomorphic pattern consisting

of 2 bands was observed for all the genotypes in this enzyme

staining system. The acromatic bands migrated 3.7 and 4.7 cm

from the origin.

Triose phosphate isomerase. Three well resolved bands

were observed for this enzyme with relative migrations of

2.0, 2.5, and 2.8 cm from the origin (Fig. 4.2E). No

variability was detected among the genotypes surveyed.

Enzyme Systems with Variable Banding Patterns

Variable banding patterns were observed for 3 enzyme

staining systems: malate dehydrogenase, peroxidase, and

phosphoglucomutase. Hypotheses to explain the observed

variability in these enzyme systems were based on a genetic

analysis of 2 segregating populations of plants from

putative reciprocal crosses between 'Smooth Cayenne' and

'Cambray'. Isozyme patterns for peroxidase and

phosphoglucomutase observed for other genotypes surveyed

will be discussed in detail in Chapter V.

Peroxidase. Three regions, 2 cathodal (Per-1 and Per-2)

and 1 anodal (Per-3), of well resolved variable banding

patterns were observed for all the genotypes tested, and are

illustrated in Figs. 4.2 and 4.3. All 3 variable regions

exhibited the same general banding patterns, consisting of

either a single fast-migrating band (FF), a single

slow-migrating band (SS), or a combination of the 2 (FS).

The slow band for Per-i was at Rf -0.26, while the fast band

was at Rf -0.28. The slow band for Per-2 was at Rf -0.04 and

the fast band was at Rf -0.11. The slow band for the

anodally migrating Per-3 was at Rf 0.04 and the fast band

at Rf 0.07. Zones of activity with no defined bands were

observed between Rf 0.11 to 0.25 and 0.32 to 0.38.

The variability detected for peroxidase may be

explained by a model of a monomeric enzyme system with 2

alleles (F and S) at each of 3 loci (Per-1, Per-2, and

Per-3). The distal band (F) and proximal band (S) are

produced by different alleles. Heterozygous genotypes would

produce both a F and S band, while homozygous genotypes

would produce either an FF or SS band. Progeny produced from

the cross with 'Smooth Cayenne' (SS) as the seed parent gave

all nonparental isozyme phenotypes (FS) for Per-1, which

would result if parents homozygous (SS, FF) for different

alleles at a locus were crossed. 'Cambray', the putative

pollen parent of the seedlings, had a Per-i phenotype of FF.

Progeny showed a 1:1 ratio of FS and FF phenotypes for both

Per-2 and Per-3, which would be expected if parents of the

seedlings were homozygous (FF) and heterozygous (FS)

respectively for the alleles at the loci. This was the case

with 'Smooth Cayenne' (FF) and 'Cambray' (FS) [Tables 4.5

and 4.7; Figs. 4.2 and 4.3A].

Seedlings obtained from open-pollinated crosses with

'Cambray' as the seed parent and 'Smooth Cayenne' as the

putative pollen parent were expected to show a similar type

of segregation, since this would be the reciprocal cross of

the one described above and only the 2 cultivars were

growing in the region. Instead, progeny from the 'Cambray'

female parent exhibited only the 'Cambray' (FF) banding

pattern for Per-1, as would be expected from a cross between

2 plants that were homozygous at this locus (Table 4.6 and

Fig. 4.2). Chi-square values (Table 4.7) showed a very close

fit to a 1:2:1 (FF:FS:SS) ratio for both Per-2 and Per-3,

which indicates a cross between plants heterozygous at these

2 loci (Fig. 4.2). The banding patterns obtained from this

progeny suggested that 'Smooth Cayenne' did not participate

in the cross and that the seedlings are most likely the

result of selfing.

Phosphoglucomutase. Three anodal regions (Pgm-1, Pgm-2,

and Pgm-3) of well resolved variable banding patterns were

observed in all the genotypes tested and are illustrated in

Fig. 4.4. The banding patterns for the Pgm-1 region were

unique in comparison to Pgm-2 and Pgm-3, and consisted of 5

total bands in the wide population of genotypes surveyed; F,

S, M, L, or P were present in various combinations of 2 of

these bands. Both Pgm-2 and Pgm-3 exhibited the same

variable banding pattern consisting of either a single

fast-migrating band (FF), a single slow-migrating band (SS),

or combination of the 2 (FS). Values for Rb were calculated

using P.I. 115, a feral genotype exhibiting a distinctive

banding pattern (Figs. 4.4 and 4.6) which served as an

internal standard. The 5 bands for Pgm-1 were at Rb 0.43

(L), 0.63 (S), 0.73 (M), 0.83(F), and 1.0 (P). The slow band

(S) for Pgm-2 was at Rb 0.84 and the fast band (F) was at Rb

0.98. The slow band (S) for Pgm-3 was at Rb 1.00 and the

fast band (F) was at Rb 1.19.

The majority of plant species reported in the

literature have contained either 1 or 2 loci governing the

expression of phosphoglucomutase, with the exception of

Camellia japonica, in which a third locus was reported by

Wendel and Parks (1982). In the present study, it was found

that the banding pattern observed in Ananas also may be

explained by hypothesizing a monomeric enzyme with 3 loci

(Pgm-1, Pgm-2, and Pgm-3). Locus Pgm-1 would be composed of

5 alleles, while Pgm-2 and Pgm-3 would each have 2 allelic


The 2 seedling populations represent only 3 (F, M, and

S) of the 5 possible Pgm-1 allelic forms. Heterozygous

genotypes would show combinations of these 3 bands, i.e.,

FM, FS, or MS, while homozygous genotypes would produce

either a FF, MM or SS band. Likewise, according to the

model, the other 2 loci, Pgm-2 and Pgm-3, produce a distal

band (F) and proximal band (S) as the result of different

alleles. Heterozygous genotypes would produce both a F and S

band, while homozygous genotypes would produce either a FF

or SS band (Tables 4.5, 4.6, and 4.7).

The 'Smooth Cayenne'-derived population showed a 1:1

ratio of phenotypes for Pgm-1 (FS:MS), Pgm-2 (FF:FS) and

Pgm-3 (FS:SS), which would be expected if the parents were

heterozygous and homozygous for each locus, i.e., Pgm-1: FM

x SS, Pgm-2: FF x FS, and Pgm-3: SS x FS (Table 4.7).

'Smooth Cayenne' is heterozygous for Pgm-1 (FM) and

homozygous for Pgm-2 (FF) and Pgm-3 (SS). Conversely,

'Cambray' is homozygous for Pgm-1 (SS) and heterozygous for

Pgm-2 and Pgm-3 (FS). The segregation pattern for the 3 loci

indicate that the seedlings are the results of a cross

between 'Smooth Cayenne' x 'Cambray' (Figs. 4.5 and 4.6A).

The 'Cambray'-derived seedlings again exhibited only

the 'Cambray' banding pattern for Pgm-1, as would be

expected for progeny from parents homozygous at this locus

(Table 4.6). Segregation was observed for Pgm-2 and Pgm-3,

and chi-square values (Table 4.7) showed a very close fit to

a 1:2:1 (FF:FS:FS) ratio., indicating a cross between 2

heterozygous types (FS). As discussed above, 'Smooth

Cayenne' is heterozygous for Pgm-1 (FS) and homozygous for

Pgm-2 (FF) and Pgm-3 (SS), so that it could not have been

the pollen parent in this cross. Thus, the

phosphoglucomutase banding patterns obtained for this

population exhibited segregation ratios which corroborate

the hypothesis that the seedlings are indeed the result of

selfing and not outcrossing (Figs. 4.5 and 4.6B).

Self-incompatibility in pineapple is due to inhibition

of pollen-tube growth in the upper third of the style. It is

gametophytically controlled by a single S locus with

multiple alleles (Brewbaker and Gorrez, 1967). In general,

pineapples set no seed when self-pollinated, but can set

seed when cultivars are crossed (Collins, 1960). Collins

(1960) reports isolated self-fertile 'Smooth Cayenne'

plants. Seed production was variable and germinated seed

gave rise to seedy offspring. These plants exhibited very

low vigor and continued selfing was not possible.

Pareja (1968) hypothesized that the frequent occurrence

of seeds in 'Cambray' may be due to self-compatibility with

pollination being accomplished by mites. The isozyme banding

patterns displayed by the seedling population obtained from

'Cambray' comfirms the self-compatibility of this cultivar.

Malate dehydrogenase. One large anodal region

consisting of many closely adjacent bands was observed for

all genotypes tested and is illustrated in Fig. 4.5. Six of

the bands were well resolved and migrated to 3.5, 3.8, 4.1,

4.5, 5.0, and 5.5 cm from the origin. The 2 fastest

migrating bands, 5.0 and 5.5, were the only ones to show

clear segregation. Banding patterns for the 'Smooth

Cayenne'-derived seedlings exhibited segregation for these 2

bands (Fig. 4.7A) while no segregation (all seedlings had 2

bands) was observed for the 'Cambray'-derived seedlings

(Fig. 4.7C). When the power setting during electrophoresis

was changed from 9 to 16, more separation of the bands was

observed and segregation for other bands seemed to appear

(Fig. 4.7B).

Malate dehydrogenase has been reported as a dimeric

enzyme in numerous diverse plant species, e.g., avocado

(Torres and Bergh, 1980), celery (Orton, 1983c), citrus

(Torres et al., 1982), eucalyptus (Moran and Bell, 1983),

and maize (Goodman and Stuber, 1983). Pollen zymograms from

different cultivars, including 'Smooth Cayenne' and

'Cambray', were compared with their respective leaf

zymograms to determine if intra-allelic heterodimer

formation was present, as would be expected in a polymeric

enzyme. Two bands were absent in pollen zymograms (Fig.

4.7D), implicating intra-allelic heterodimer formation. The

presence of these interaction bands suggests that malate

dehydrogenase functions as a polymeric enzyme in Ananas.

However, more segregating populations must be analyzed

before the genetics of this enzyme system can be clarified.

Linkage Analysis for Enzyme Loci

Linkage analysis was carried out between the

peroxidase and the phosphoglucomutase loci. The results are

given in Tables 4.8 and 4.9. Chi-square analysis by

contingency tables showed that the observed ratios fit

ratios expected for independent segregation.


1. Eight enzyme staining systems of the 38 investigated

were found to give well-resolved banding patterns.

2. Well-resolved non-variable isozyme banding patterns

were observed for isocitrate dehydrogenase, 6-phospho-

gluconate dehydrogenase, phosphohexose isomerase,

superoxide dismutase, and triose phosphate isomerase.

3. Well-resolved variable isozyme banding patterns were

observed for malate dehydrogenase, peroxidase, and


4. A genetic model was hypothesized to account for the

variability observed for peroxidase and phosphoglu-

comutase staining systems in the Ananas germplasm

surveyed and in segregating seedling populations from

'Smooth Cayenne' and 'Cambray' pineapple cultivars. The

model hypothesizes that both peroxidase and phospho-

glucomutase function as monomeric enzymes coded for

by 3 loci each: Per-1, Per-2, Per-3, Pgm-1, Pgm-2,

and Pgm-3. Each locus contains 2 allelic forms, F

and S, with the exception of Pgm-1, which exhibits

5 allelic forms: F, S, M, L, and P.

5. Independent segregation was found between all of the

peroxidase and phosphoglucomutase loci.

6. The isozyme studies confirmed that the 2 seedling

populations were the products of crosses between

'Smooth Cayenne' x 'Cambray' and 'Cambray' x self.


7. The existence of the 'Cambray' x self population calls

into question belief that domestic pineapple cultivars

are obligated self-incompatible.

Table 4.1. Ananas genotypes surveyed for isozyme banding

A. ananassoides
A. bracteatus
A. comosus
A. lucidus
A. nanus
A. monstrosus
A. parguazensis

A. comosus cultivars
Black Jamaica
Cabeza de Mono
Cacho de Venado
Dupuis Smooth
Monte Oscuro
Pina de Brazil
Red Spanish
Smooth Cayenne
Valera Amarilla

Feral TypesZ
046 A. ananassoides
064 A. nanus
072-1 A. comosus
079 A. ananassoides
083 A. ananassoides
085 A. comosus
086 A. comosus
092-1 A. comosus
092-2 A. comosus
093 A. comosus
094 A. comosus
095 A. comosus
097 A. parguazensis
108 A. parguazensis
110 A. comosus
115 ?
116 ?
117 ?
126 A. comosus
188 ?
189 ?
192 A. comosus
193 ?
196 A. comosus
25291 A. comosus
487439 A. comosus
487440 A. comosus
487442 A. comosus
487443 A. comosus
487444 A. comosus

ZNumbers correspond to introductions followed by the
suspected taxonomical classification. "?" indicates that
the botanical characteristics of the plant do not fit any
of the presently described species.

Table 4.2. Electrophoretic buffer systems used for isozyme
detection in Ananas.

Buffer Electrode Gel
system buffery buffery

H 0.065 M histidine 0.009 M histidine
0.02 M citric acid 0.003 M citric acid
pH adjusted to 5.7 pH adjusted to 5.7
with citric acid with citric acid

K 0.18 M trizma base 0.045 M trizma base
0.1 M boric acid 0.025 M boric acid
0.004 M Na2EDTA 0.001 M Na2EDTA
pH 8.6 pH 8.6

LBTC 0.016 M LiOH 0.0016 M LiOH
0.192 M boric acid 0.019 M boric acid
pH 7.2 0.007 M citric acid
0.046 M trizma base
pH 7.7

TB 0.038 M trizma base 0.03 M boric acid
0.002 M citric acid pH 8.5
pH 8.6

TC 0.05 M trizma base 0.017 M trizma base
0.016 M citric acid 0.005 M citric acid
pH adjusted to 7.0 pH adjusted to 7.0
with citric acid with citric acid

ZH = Histidine citrate (Cardy et al., 1981); K = K buffer
(Loukas and Pontikis, 1979); LBTC = Lithium borate/Tris
citrate (modified Scandalios, 1969); TB = Tris borate;
TC = Tris citrate (Moore, unpublished).

Y10 mg NADP was added to buffer in cathodal tank and to gel
buffer when staining for enzyme system that required NADP
as a cofactor.


Table 4.3. Enzyme activity stains and buffer systems tested
for isozyme detection in Ananas.

Buffer system

Enzyme activity stain LBTC H TC TB K

Acid phosphatase
Adenylate kinase
Alcohol dehydrogenase
Alkaline phosphatase
Ascorbate oxidase
Creatine kinase
Formate dehydrogenase
Galactose dehydrogenase
Glucose-6-phosphate dehydrogenase
Glutamate oxalloacetate transaminase
Glutathione reductase
Isocitrate dehydrogenase
Lactate dehydrogenase
Leucine aminopeptidase
Malate dehydrogenase
Malic enzyme
6-phosphogluconate dehydrogenase
Phosphohexose isomerase
Piruvate kinase
Shikimate dehydrogenase
Superoxide dismutase
Superoxide dismutase + KCN
Triose phosphate isomerase
Xanthine dehydrogenase










- *

- = no activity; + = poor migration or resolution;

++ = fair migration or resolution; +++ = well resolved
bands; = assay not performed.

YModified to 0.065 M, pH 6.3.








0 )

0 C
C- o





S 0.



El Q.



C =


a) )
4-4 4-)
4-4 W





~ e~~ G~

n o0 0 n 0 0 0 0

n omN c4 N m 'o.
-4 -4 r-1 l

Ln 0 Ln In 0 In q In

o o o 0 0 0 0 0
0o o In o 0 VIn I
mm c m cN

>1 >1
a a tc; o


0 )


u c

a) 0
u e4

?0 a

>4J 0

S 4-4
3 44

0 41


4 4-

co (a
CE 4)

U) -H-

s-I r
40 ( 4
rCiC C0

E- 0

Table 4.5. Peroxidase and phosphoglucomutase banding
patterns observed for seedlings derived from
an open-pollinated 'Smooth Cayenne' fruit.

Isozyme genotypes

Seedlings Per-i Per-2 Per-3 Pgm-1 Pgm-2 Pgm-3

Table 4.6. Peroxidase and phosphoglucomutase banding
patterns observed for seedlings derived
from an open-pollinated 'Cambray' fruit.

Isozyme genotypes

Seedlings Per-1 Per-2 Per-3 Pgm-1 Pgm-2 Pgm-3



















H (U
^ -P


(N (n (n

0- CN (N
0 C4 C
*** V

I I l r rl r
I ** ** ** ** **
r] l rol



0o o o-I m co C0

r-Il r- rV-


0) 0) (1


eU) n W


r-it CP
I<& iI

o Ln
o r-
* I

0 0


I *.N C
* C(




I *..



1 0 i I -
- -l

0 LAn c

U) CnU) )

V --- V
U) r-I NM C

> 0) (U U)


I ow O

r-4 N en

On m t7
P4 P 0
0 tr CD











0 0


0 0

0 -H

u 0

- 41
U- (a
U u




-H +


O > LnooLnminMmmaLn
Qc) a4

0fa O.- w CN

U0 0 0
3r-I r-4 r- r-i r-4 r-4 r-4 r-l r-A -4

-I 4 r- r- r-4 -
S.. .. .. .. .. .. .. .. ..
H U ( r- r-I v.-4 .-4
44 W

4-1 tQ E-i
m4 4J

0 .
E 0 -- -oo o- -- o-

0 0 w 1,m r- m %
0 0-) rJW W C W W b W CQ M M
00 4OC2www

mC -O

4 -4
0- - -

o Iz I I I I I I IQI I
0 Q 0
I (0 44 ~ U 5C/ rZz 44 rjj r4 M W .4
OH r

u w r.
(0 -

(0 0 U)r rM4M m N
*H *04 ao wU) ) -H
-Ik CC UrJ Cf l
(0 44 -H (I

V 0- r. U)

S0 w0l0C0(7
C U(D ) HNCDI -O ,-0 v-4 V
*-r U) (0 -- )
-Pu N U) U)4 nMU)U (nCA)C CncU) W
(z '- P U ) Q ) I w x w m x ww 0

U) -H V -W

* w tp tp tr tU) 0 tr '0'00 tp0 V

0 4 W 4 W 4 W W E E E I
(0 0 U w ) ) )U U w U t'p p T
E-4 4- a4 P404 04 134 0i ai P4 (i a4~ A4a

0 0 L LA LA L 0 0
0) (N 0 (N N H
P > (0 .
S-H > I I I I
S0 0 0 o Ln
W a 4 Ln in in N CN
0 I
04- a)
(U > 0
0 ( o o o oo
c o N o > o
r > m m m in o

m -l
0 C

00 *X

(0 U i -4l r-l W( r- ri l 1 r r-1 i l -l 1-4
> o ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** **
> 0 .. .. .. .. .. .. .. .. 5* .. .. 0. *. Co *.
4-0 L .. ** ** ** ** ** ** ** ** ** ** ** ...
-HO) W) -1 .-i .- .-4l --4 r- .-4 .- I r-4- r-4 V-i -l 4 -
4-4O E-

41 4-

C0 3

U) 0 Un z-1ti) MCI) fciZ.04 LUn y) 6CN
rX4 rT4 rL4a d rZ.4 r 4 rL| r24 CL| PLa

a) 0 0 -- -.- 1- -

O0. 0 (N ( (O (N O
o0 0 U)( U(n U) E U) ( UI (no U)

04 rici) FZd j) IT4 U) rZ4 W EI U)
00 0
44 U) rNL mm r34 -I r m c
0 U) 441 LIC w C(n W W 4 I2 CrO2 W CJ
1 3 r4 ) WzUi4 UM1s) r4 M44 M(0
W EM I II I ) XC4 nCa II an Ia CI U tO

O 0
0 x

0 a) -H -1-
00 n C14 U O0 (nMX MA M W MM MM
(U 44 (U ra a)

0 0 4I I mI
-HC ) --- --- -- --- (1)

O)sIN ON(N (N mmO i
4J (U ) U N raM U M U) W fi) (n W m w aw wm 44 (n U) r2 V) U)M M)
(-I -J H I I I I 6 W U)

41 1 N) 1 )
(W r--lN n r-4N (e NM m
*HI I I II I I I I I u
4 4 W E E EE E EEE H
< A 04 04 04 04 04 a4 94 P4 Pc PL C
W C ( N C% (N ce m m t AI

SU4 4 W w W 4 4 E E 'I
(13 0 () Q) () (1) a) (U Q) (U) 0>0l t
E-4 P4 04 PL4 a4 a4 14 a4 A A4ia4 0 a 4



Figure 4.1. Non-variable isozyme banding patterns
observed for Ananas genotypes. A. Isocitrate
dehydrogenase; B. 6-Phosphogluconate
dehydrogenase. Sample origin is at the bottom
of each photograph and band migration is

Figure 4.1. continued. C. Phosphohexose isomerase;
D. Superoxide dismutase.

I ~~~HUIlr I I I

Imsp't*3g "~9ffl

Figure 4.1. continued. E. Triose phosphate isomerase.


m-mmmm -m mmm


--a --- -

Pe r-2
um- mmm- mmm -

m mmmm

m Per-i

1 2 3 4 5 6 7 8 9 10 11 12 13

Figure 4.2. Schematic representation of peroxidase isozymes
in 2 pineapple segregating populations.
1. 'Smooth Cayenne'; 2. 'Cambray'; 3-6. Hybrids
of 'Smooth Cayenne' x 'Cambray' cross;
7-13. Hybrids of 'Cambray' x self cross.






~ I

I'Y ?


I e


Figure 4.3. Peroxidase banding patterns observed in
pineapple. A. From right to left, 'Smooth
Cayenne' ,'Cambray', and progeny from the
putative cross 'Smooth Cayenne' x 'Cambray';
B. 'Smooth Cayenne', 'Cambray', and progeny
from the putative cross 'Cambray' x self.
Sample origin is at the middle of each
photograph and band migration is both
anodal and cathodal.



m -m



1 2 3 4 5 6 7 8 9

Figure 4.4. Schematic representation of phosphoglucomutase
isozyme in Ananas. 1. P.I. 115; 2. 'Maipure';
3. 'Smooth Cayenne'; 4. P.I. 072; 5. P.I. 085;
6. 'Black Jamaica'; 7. P.I. 095; 8. P.I. 27281;
9. 'Brecheche'. Black bands = Pgm-1; cross-
hatched bands = Pgm-2; dotted bands = Pgm-3.

~S~i~B ~ g~ ~3~ ~Sg~two

1.1 -

........ ........


Pg m-2

m m m

I m

m m


m M

1 2 3 4 5 6 7 8 9

Figure 4.5. Schematic representation of phosphoglucomutase
isozyme in 2 pineapple segregating
populations. 1. 'Smooth Cayenne';
2. 'Cambray'; 3-6. Hybrids of 'Smooth
Cayenne' x 'Cambray' cross; 7-9. Hybrids
of 'Cambray' x self cross.

Am= As= am ~~8W

M =- ill 8 i I ME=H~s

*4* *fb* | ,

Figure 4.6.

Phosphoglucomutase banding pattern in
pineapple. A. From right to left, P.I. 115,
'Smooth Cayenne', 'Cambray', and progeny
from the putative cross 'Smooth Cayenne' x
'Cambray'; B. 'Smooth Cayenne', 'Cambray',
and progeny from the putative cross
'Cambray' x self. Sample origin is at the
bottom of each photograph and band migration
is anodal.


Figure 4.7.


Malate dehydrogenase patterns observed in
pineapple. A. and B. Gels run at different
power settings (9 and 16), from right to left,
'Smooth Cayenne', 'Cambray', and seedlings
from the putative cross 'Smooth Cayenne'
x 'Cambray'. Sample origin is at the bottom
of each photograph and band migration is




Figure 4.7. continued. C. Progeny from the putative
cross 'Cambray' x self; D. Pollen and leaf,
from right to left, 'Smooth Cayenne',
'Brecheche', and 'Cambray'.



The utilization of isozyme techniques for cultivar

identification requires high levels of clearly recognizable

inter-cultivar isozyme polymorphism, preferably with

monomorphic cultivar phenotypes. The phenotypes should be

stable over a variety of environmental and developmental

conditions. The techniques used to characterize the

phenotypes should be simple and efficient.

Ananas comosus appears to be an ideal candidate for

electrophoretic characterization using isozyme polymor-

phisms. The self-incompatible and outcrossing nature of the

species would be expected to generate high levels of

inter-specific genetic variation, while intra-clonal

variability would be expected to be relatively low due to

the apomictic nature of pineapple.

The taxonomy and origin of the genus Ananas has been

the subject of much debate and speculation. Species are

generally distinguished by the presence or absence and

direction of leaf spines, petal morphology, and fruit size.

The purpose of this study was to electrophoretically

characterize the genus Ananas using isozyme polymorphisms

with emphasis on gaining information on the origin of the

commercially important pineapple cultivars of the world.

Materials and Methods

Starch gels were prepared and electrophoresis was

conducted as described in Chapter IV. Two buffer systems

were used for electrophoresis, a histidine system at pH 6.3

and a K buffer at pH 8.6 (Table 4.2). The histidine buffer

was used to assay for phosphoglucomutase activity and the K

buffer for peroxidase activity. At least 2 plants of each

genotype (Table 4.1) were examined electrophoretically for

peroxidase and phosphoglucomutase polymorphisms, except that

no peroxidase determination was made for the feral types.

Assays were repeated until consistent results were obtained.

The banding patterns were recorded using both schematic

drawings and photographs. Band migration distances were

based on Rf and Rb values as calculated in Chapter IV.

Results and Discussion

Considerable isozyme polymorphism was observed among

the genotypes examined for both peroxidase and

phosphoglucomutase. In both systems, bands were

characterized for each locus according to inheritance

studies discussed in Chapter IV. Peroxidase showed either a

single or a double banded pattern at 2 migrating regions,

corresponding to the Per-2 and Per-3 loci. No clear bands

were present for Per-1. Phosphoglucomutase exhibited the

most useful isozyme polymorphisms. Bands corresponding to

the Pgm-1 locus gave consistent and clear resolution but

overlapping bands corresponding to Pgm-2 and Pgm-3 loci

complicated the characterization of certain genotypes.

Cultivar Identification

Peroxidase. Limited isozyme polymorphism was detected

for peroxidase. A total of 7 different banding patterns was

observed and designated A1 through A7 (Table 5.1; Figs. 5.1

and 5.2). Patterns A3, A4, and A7 were unique to 'Negrita',

'Pina de Brazil', and 'Maipure' respectively. 'Panare' and

'Queen' shared pattern A6, while the other cultivars showed

patterns A1, A2 or A5 (Table 5.2).

Phosphoglucomutase. Eleven different phosphoglucomutase

banding patterns were identified and designated B1 through

B11 (Tables 5.1 and 5.2; Figs. 5.1 and 5.2). Unique banding

patterns were observed for 'Black Jamaica', 'Rondon',

'Valera Amarilla', 'Cacho de Venado', and 'Cabezona'. These

patterns were B3, B4, B7, B and B10 respectively. The rest

of the cultivars were distributed over the patterns B1, B2,

B5, B6, B9, and B11 (Table 5.3).

The polymorphic enzyme systems of peroxidase and

phosphoglucomutase were very useful for cultivar

classification. Using only the peroxidase or the

phosphoglucomutase isozyme system, 8 cultivars could be

distinguished by their unique banding patterns (Tables 5.2

and 5.3). However, using information combined from both the

peroxidase and phosphoglucomutase isozyme systems, 20 unique

classes of banding patterns could be distinguished. These

classes permitted the identification of 15 cultivars out of

the total 27 examined (Table 5.4).

The origin and exact relationship of many of the

pineapple cultivars studied remains unclear. A wide variety

of clones and cultivars exists because of the prevalence of

inter-clonal fertility and apomixis accompanied by

self-incompatibility. Certain pineapple cultivars and clones

exhibit high rates of mutation and many unique sports exist;

e.g., 'Smooth Cayenne', the most economically important

cultivar, is really a group of clones (Collins, 1960). These

factors considerably complicate characterization and

classification schemes.

Based on the peroxidase and phosphoglucomutase isozyme

banding patterns, it appears that there is a very close

relationship among many of the cultivars tested. 'PR-1-67',

a reputed hybrid widely grown in Puerto Rico, was obtained

from an open-pollinationed seed of 'Red Spanish' growing

adjacent to a field of 'Smooth Cayenne' (Ramirez et al.,

1970 and 1972). The parental relationship of these 2

cultivars to 'PR-1-67' is corroborated by their isozyme

banding patterns (Table 5.1). 'Esmeralda', a cultivar grown

in Mexico (Samuels, 1970), and 'Dupuis Smooth' are presumed

to be sports of 'Smooth Cayenne'. The 3 cultivars exhibited

the same banding pattern, indicating a common origin.

'Cumanesa' is very similar morphologically to 'Red Spanish'

and is believed to be a sport of the latter (Leal and

Antoni, 1980b). The identical banding pattern showed by the

2 cultivars suggests that 'Cumanesa' could indeed have

originated as a sport of 'Red Spanish'.

'Bumanguesa' and 'Perolera' are the 2 cultivars

commonly grown in the Andean regions of Venezuela and

Colombia. Both cultivars have similar morphological

characteristics of plant growth and differ only in mature

fruit color. 'Bumanguesa' produces a bright red fruit, while

'Perolera' produces a yellow fruit. Because of the

similarities of these 2 cultivars, Leal et al. (1979)

hypothesized that 'Bumanguesa' was a sport of 'Perolera'.

However, the peroxidase and phosphoglucomutase banding

pattern observed for these 2 cultivars do not indicate a

close relationship, unless the mutation involved the Per-2,

Per-3, and Pgm-3 loci (Table 5.1).

The cultivars 'Queen' and 'Panare', which represent the

"Queen" group of pineapples, exhibited identical isozyme

banding patterns for both peroxidase and phosphoglucomutase.

They also are very similar morphologically. This is

remarkable since cultivars belonging to this horticultural

group are grown commercially only in South Africa and

Australia (Samuels, 1970). 'Panare' was described by Leal et

al. (1979) from a clone being cultivated in southern

Venezuela (Bolivar State). Local people reported that the

original plants were "wild" types found growing nearby.

Since the area including southern Venezuela and northern

Brazil has been suggested as the center of origin for the

Ananas genus (Leal and Antoni, 1980a) it seems likely that

the "Queen" group probably was derived from 'Panare'-like

wild types. This finding confirms the presence of "Queen"

types growing in the Western Hemisphere.

There seems to be some correlation among the Pgm-1

alleles and the horticultural groups of pineapple (Table

5.1). All cultivars within the "Cayenne", "Queen", and

"Maipure" groups exhibit identical banding patterns, i.e.,

FM. Phylogenetic relationships between these groups will be

discussed in Chapter VI. Most of the cultivars in the

"Spanish" group had a common banding pattern for Pgm-1,

i.e., SS. However, the 4 cultivars in the "Abacaki" group,

showed different allelic combinations for Pgm-1, i.e., FM,


Pineapple cultivars are classified into horticultural

groups without regard to their origin or genetic

relationships, e.g., 'PR-1-67' is included in the "Spanish"

group even though it is a known hybrid (Samuels, 1970; Leal

and Soule, 1977). Similarly, other cultivars could be of

hybrid origin, which would explain the isozyme banding

variability observed between individuals within a group.

'Caicara' is from the same region as 'Panare' and

'Brecheche' and it may have been derived from these. 'Black

Jamaica' may be a hybrid between 'Smooth Cayenne' x 'Red

Spanish', and 'Valera Amarilla' may be a hybrid of

'Perolera' x 'Red Spanish', since the parents and hybrids

are all cultivated in the same regions. 'Cacho de Venado',

which has spiny leaves and conical-shape fruit, has been

included in the "Abacaki" group created by Py and Tisseau

(1965). Based on isozyme analysis of the Pgm-1 locus, it

appears to be more related to the "Spanish" group.

Furthermore, 'Cacho de Venado' is from the same region as

'Monte Oscuro' and the feral type 086 (Monagas State,

Venezuela). The banding pattern of these cultivars also

suggests a hybrid origin for 'Cacho de Venado'.

Species Identification

Peroxidase and phosphoglucomutase banding patterns for

7 Ananas species are shown in Table 5.5, Figs. 5.2 and 5.3.

Based on the banding patterns of peroxidase and

phosphoglucomutase, 5 of the 7 species can be distinguished.

The 2 exceptions are A. comosus, which includes all of the

commercial pineapple cultivars, and A. monstrosus. Ananas

comosus, as indicated in the previous section, exhibited

high levels of isozyme polymorphism, particularly for the

Pgm-1 locus. Ananas monstrosus also exhibited isozyme

polymorphism in the 2 individuals tested.

The amount of polymorphism observed within A. comosus

may be attributed to the present very broad circumscription

of this species. Smith (1961) separates A. comosus from the

other Ananas species by the length of the fruit, i.e.,

greater than 15 cm. Thus, in this classification system,

this species includes all the edible pineapples regardless

of other morphological differences such as leaf color,

presence or absence of spines, and fruit morphology. The

same flexibility is not allowed in the classification of the

other species and only individuals that exactly fit the

description, which includes leaf, flower, and fruit

morphology and growth habit, are included.

Ananas monstrosus was originally described by Carriere

(1870) and included in the later classification by Smith

(1961). The description was based on a single individual

plant found. It differs from A. comosus (pineapple) only in

that the fruit lacks a crown. The validity of this species

has been questioned by Antoni (1983), who reported finding a

mutation of the cultivar 'Negrita' that mimics the A.

montrosus characteristic, i.e., lack of crown. Based on

growth habit and leaf and fruit morphology, the 2 plants

representing A. monstrosus used in this study appear to be

crown mutations of 'Perola' and 'Red Spanish' respectively.

The peroxidase and phosphoglucomutase banding patterns

observed for these plants are also typical of 'Perola' and

'Red Spanish'. The taxon thus seems to be of polyphyletic

origin. The questionable validity of A. monstrosus as a

species is further corroborated by the fact that crown

mutations have commonly been seen in other pineapple

cultivars by the author.

The lack of intra-specific variability observed in the

isozyme banding patterns of A. ananassoides, A. bracteatus

var. tricolor, A. lucidus, A. nanus, and A. parguazensis may

have been due to the low sample number (2) used to

characterize each species. However, when the electrophoretic

data from all 5 of these species are considered together, it

is quite surprising that low inter-specific variability was

detected as well. It may be that this lack of variability

reflects the narrow circumscription of these species. Many

of the feral types used in this study closely resembled

individuals fitting the definition of species recognized by

Smith (1961), but they were not included in any of the

prsently recognized species because they differed in 1 or 2

characters from the species descriptions given by Smith and

Downs (1979). These genotypes may belong within the limits

of some of the presently delimited species, thus giving this

taxa greater intra-specific variability, as in the case of

the different clones of pineapple (A. comosus). Perhaps the

most valid explanation for the high degree of intra-specific

homology observed in all the 'Ananas' species, except for A.

comosus, is that they are self-pollinated. In contrast, A.

comosus is self-incompatible, with obligate out-crossing

favoring a high degree of isozyme and allelic heterogenity.

Feral Type Identification

The phosphoglucomutase banding patterns for the 29

feral types characterized electrophoretically are shown in

Table 5.6, Fig. 5.3. The full range of isozymic polymorphism

at each of the 3 phosphoglucomutase loci was represented in

this diverse collection of genotypes. The Pgm-1 locus was

particularly polymorphic, with all of the 5 alleles

represented, i.e., L, S, M, F, and P. Two of these alleles,

L and P, were unique to this collection and had not been

observed in any of the species or pineapple cultivars

previously characterized. Nine different banding patterns

(SL, ML, SS, MS, MM, FM, FS, FF, and PS) were observed at

the Pgm-1 locus. Five of these (SL, ML, MM, FF, and PS) were

new allelic combinations not previously observed.

The FF Pgm-1 genotype was observed in P.I. 095, typical

pineapple plants that produce large, conical 3 kg fruit. The

2 sampled plants of this introduction were collected from a

small plot of plants being cultivated by the Piaroas Indian

River community of El Gavilan. The region is an isolated

portion of dense tropical forest located in the north

central portion of the Amazonian Territory in southern

Venezuela. The cultivated plants were reportedly collected

from the wild by the local Indians.

The Pgm-1 genotype MM was observed in the 2 plants of

the introduction P.I. 487443. These plants were collected

from the tropical rain forest region of San Carlos de Rio

Negro. This region is located in the southernmost portion of