INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI

films the text directly from the original or copy submitted. Thus, some

thesis and dissertation copies are in typewriter face, while others may

be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of thecopy submitted. Broken or indistinct print, colored or poor qualityillustrations and photographs, print bleedthrough, substandard margins,

and improper alignment can adverselyaffect reproduction.

In the unlikely event that the author did not send UMI a complete

manuscript and there are missing pages, these will be noted. Also, ifunauthorized copyright material had to be removed, a note will indicate

the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand corner and

continuing from left to right in equal sectionswith small overlaps. Each

original is also photographed in one exposure and is included in

reduced form at the back of the book.

Photographs included in the original manuscript have been reproducedxerographically in this copy. Higher quality 6" x 9" black and white

photographic prints are available for any photographs or illustrations

appearing in this copy for an additional charge. Contact UMI directly

to order.

U·M·IUniversity Microfilms International

A Bell & Howell Information Company300 North Zeeb Road. Ann Arbor. M148106-1346 USA

313/761-4700 800/521-0600

Order Number 9312191

Fluorescent age pigment accumulation as a determinant ofchronological age in aquatic organisms

Hill, Kevin Thomas, Ph.D.

University of Hawaii, 1992

V·M·I300N. ZeebRd.AnnArbor,MI48106

_.,- _.._--_._-_.....----_ ..__ ._ ..

FLUORESCENT AGE PIGMENT ACCUMULATION AS A DETERMINANT

OF CHRONOLOGICAL AGE IN AQUATIC ORGANISMS

A DISSERTATION SUBMITIED TO THE GRADUATE DIVISION OF THEUNIVERSITY OF HAWArI IN PARTIAL FULFILU"ffiNT OF THE

REQUIREMENTS FOR THE DEGREE OF

DOcrOR OF PmLOSOPHY

IN

ZOOLOGY

DECEMBER 1992

BY

Kevin T. Hill

Dissertation Committee:

Christopher Z. Womersley, ChairpersonGregory Ahearn

Fred I. KamemotoJames D. Parrish

Christopher L. BrownJack R. Davidson

ACKNOWLEDGMENTS

I am grateful to my mentor, Dr. Christopher Womersley, for inviting me into his

laboratory, showing interest in this project, and for providing enlightenment, insight, and

support through the duration of this dissertation project. I would also like to thank my

other committee members-- Drs. Gregory Ahearn, Christopher Brown, Jack Davidson,

Fred Kamemoto, and James Parrish. Their ideas, perspectives, and critical reviews

contributed greatly to the quality of this project.

I thank those individuals who aided in the culture and maintenance of fish used in

this study: Paul Shiota and Torn Kazama of the National Marine Fisheries Service,

Honolulu Laboratory, maintained the fish used in Chapter 3 of this study; Miles

Yamamoto andAdrian Franke helped maintainthe experimentalaquarium system in

Edmondson Hall during my absence; and Alane Gresham rescued hundreds of rosy barb

fry during an island-wide power outage.

Special thanks go out to past and present members of Chris Womersley's research

group for enriching my daily life in the laboratory, in particular, Lynne Riga who shared

her friendship and philosophies and Alane Gresham for sharing insights on the chemistry

of aging.

Marty Martinelliand Dr. Andrew Taylor provided much needed statisticaladvice

and computer assistance with SAS. I am also indebted to the support staff of the

Department of Zoology--Lori Yamamura, Sally Oshiro, and Audrey Shintani-- for their

patience and expertise in dealing with matterspertaining to red tape. illustrative skill of

Sue Monden was much appreciated for figures in Chapter 2.

The research was made possible through generous support provided by: the

University of Hawaii Sea Grant College Program; the Research Corporation of the

University of Hawaii; the Universityof Hawaii Center on Aging; Sigma Xi - The

Scientific Research Society; the ARCS Foundation of Hawaii; the University of Hawaii

iii

Office of Research Administration; and the Department of Zoology, University of

Hawaii. I am very grateful for all financial contributions that enabled me to complete this

research.

All of this work was inspired through the tolerance and affection of my wife,

Astrid Ronke, who instilled in me the persistence needed to complete this work. I also

thank my parents, Randall and Geraldine Hill, for their support and patience through my

years in graduate school.

iv

ABSTRACT

Age pigments have been intensivelystudied by gerontologists hoping to define the

biochemical processes involvedin aging. More recently,age pigment accumulation has

been studied by field biologistshoping to estimatechronologicalage of animals based on

quantitiesof pigments measured in postmitotic tissues. The present research was

conducted to evaluate methods used for extractionand measurementof fluorescent age

pigments (FAP), describe patterns of FAP accumulation in two teleost fish speciesand

develop multivariateage predictionmodelsbased on FAP content and other somatic

variables, and determine the effects of some environmentaland physiologicalvariableson

FAP variability. Evaluation of procedures for handling specimensand extracts revealed

increases in FAP-like fluorophores in vitro in Oreochrornis mossambicus brain, heart,

and muscle tissues and theirextracts with increasedstoragetemperature (-20°C and

above) and time, particularlyin the chloroform/methonal solvent system. Ultrasonication

of tissue homogenatesgreatlyenhanced this effect and generated other fluorophores in

solution. Fluorescence assay temperaturealso affectedexpression ofFAP. Some or all

of these effects of handlingproceduresmay have produced variabilityof results reported

from previous FAP studies.

NonpolarFAP extractedfrom o. mossambicus brain accumulated in an

increasingly rapid manner with age. Nonpolarand polar FAP from Puntius conchonius

brain leveled off with increasingage. FAP in £. conchonius heart increased linearly with

age. Multiple regressionanalysesdemonstratedthe age-predictive value of FAP data for

both species under controlledrearing conditions. FAP content and otolith dimension data

consistentlyprovided slightlystrongerpredictive models than somaticdata alone.

Temperatureand body weight affectedFAP content in £. conchonius brain and heart

tissues. Brain FAP was inversely related to body weight, especially at the lower

temperatures. Heart FAP was inverselyrelated to body weight at 19°C, but positively

v

--_._.._-- .-

related at higher temperatures. Increasing temperature decreased brain FAP, but

increased heart FAP. Different levels of saturation of cellular lipid constituents at

different temperatures may affect the potential for FAP formation via lipid peroxidation.

FAP differences detected between non-sibling fish indicated genetic variation in the ability

to control FAP genesis.

VI

TABLE OF CONTENTS

ACKNOWI...EDGMENTS , , , .. iii

ABSTRACT v

LIST OF TABLES ix

LIST OF FIGURES x

CHAPTER 1. INTRODUCTIONOverview 1

Cellular genesis ofFAP 2Quantification ofFAP 4Age pigments in aquatic animals , " 5

Research problem 8Study objectives 10Literature cited 11

CHAPTER 2. CRITICAL ASPECfS OF FLUORESCENT AGE PIG?vIENT?vIETHODOLOGIES: MODIFICATION FOR ACCURATE ANALYSIS AND

AGE ASSESS?vIENTS IN AQUATIC ANIMALSIntroduction " 29Materials and methods '" , . 32

Model animal/tissues 32Tissue storage temperature " . 32Unmodified extraction procedures 33Incubation temperature of sample extracts 33Effects of ultrasonication on fluorescence yield 34Confirmation and removal of interfering compounds 34Assay temperature 34Modified extraction and assay procedures 35

Results 36Tissue storage temperature 36Incubation temperature of sample extracts 37Effects of ultrasonication , , 38Fluorescent contaminants 38Assay temperature 39FAP accumulation with age 39

Discussion 40Literature cited 46

CHAPTER 3. FLUORESCENT AGE PIGMENT ACCUMULATION ANDTHE USE OF MULTIPLE REGRESSION MODELS FOR AGE ESTIMATION

IN TWO FISH SPECIESIntroduction " 70Materials and methods , , 73

Specimen culture and preparation 73FAP extraction and spectrofluorometric assay 74Multiple regression analysis 76

vii

T ABLE OF CONTENTS (continued)

Results 76Somatic and otolith growth 76Organ growth and FAP accumulation with age 77Multiple regression analyses 79

Oreochromis mossambicus 79Puntius conchonius 80

Discussion 82References 88

CHAPTER 4. INTERACTIVE EFFEcrS OF SOME ENVIRONMENTAL ANDPHYSIOLOGICAL VARIABLES ON FLUORESCENT AGE PIGMENT

ACCUMULATION IN BRAIN AND HEART TISSUES OF ANAQUATIC POIKILOTIffiRM

Introduction 112Materials and methods , 115

Model animal 115Aquarium system 115Experimental design 116

Temperature/ration experiment 116Temperature/photoperiod experiment 117

Oxygen consumption , " 117Specimen/tissue preparation 118FAP extraction and assay 118Statistical analysis 120

Results , 120Temperature/ration experiment 120Temperature/photoperiod experiment ;; 123

Discussion 124Literature cited 131

CHAPTER 5. SUMMARY AND CONCLUSIONS 158

viii

~ .. __ .._--_.._--.... ""--

LIST OF TABLES

Table Page

2-1 Previous studies of fluorescent age pigment accumulationin aquatic animals " , ,.. 54

3-1 Regression coefficients and associated statistics from multipleregression models of age for Oreochromis mossambicus 96

3-2 Regression coefficients and associated statistics from multipleregression models of age for Puntius conchonius 98

4-1 Results of ANCOVA model for testing effects of fish body weight,temperature, photoperiod, and clutch on polar FAP frombrain tissues of Puntius conchonius 140

4-2 Results of ANCOVA model for testing effects of fish body weight,temperature, photoperiod, and clutch on polar FAP fromhean tissues of Puntius conchonius 141

ix

Figure

1-1

1-2

1-3

1-4

2-1

2-2

2-3

2-4

2-5

2-6

2-7

3-1

3-2

3-3

LIST OF FIGURES

Page

Modulation of superoxide radicals produced through normalcellular metabolism 22

Free radical-mediated lipid peroxidation and malondialdehydeproduction " 24

Fluorescent crosslinkage reactions of malondialdehyde 26

Summary schematic ofprocesses leading to fluorescentage pigment formation , 28

Effect of storage temperature on fluorescent products extractedfrom skeletal muscle of Oreochromis mossambicus 57

Effects of incubation temperature on fluorescence ofunfractionated chloroform/methanol extracts of brain, heart, andmuscle tissues of Oreochromis mossambicus 59

Effects of incubation temperature on fluorescence of finalchloroform extracts of brain, heart and muscle tissues ofOreochromis mossambicus '" 61

Effects of ultrasonication on fluorescence of homogenates ofbrain, heart, and muscle tissues of Oreochromis mossambicus 63

Fluorescence spectra of fluorescent age pigments and flavincontaminants extracted from brain tissues ofOreochromismossambicus , '" " 65

Effects of assay temperature on expression of fluorescenceof quinine sulfate standard and fluorescent age pigments extractedfrom brain tissues of Oreochromis mossambicus 67

Fluorescent age pigment content and brain weight as functions ofchronological age in Oreochromis mossambicus 69

Body size and otolith weight as functions of chronological agein Oreochromis mossambicus , 101

Body size and otolith weight as functions of chronological agein Puntius conchonius l03

Brain weight and fluorescent age pigment level as functions ofchronological age in Oreochromis mossambicus , 105

x

Figure

3-4

3-5

3-6

4-1

4-2

4-3

4-4

4-5

4-6

4-7

4-8

-_.__..~------ ....._-

LIST OF FIGURES (continued)

Page

Emission spectra of fluorescent age pigments extracted frombrain and heart tissues of Puntius conchonius , 107

Brain weight and fluorescent age pigment level as functions ofchronologicalage in Puntius conchonius 109

Heart weight and fluorescentage pigment level as functions ofchronologicalage in Puntius conchonius.. , 111

Mean fish weights of Puntius conchonius reared in thetemperature/ration and temperature/photoperiod experiments 143

Effects ofdifferent temperatureand ration levels on FAP in braintissues of Puntius conchonius 145

Brain polar FAP as a function of fish weight for Puntiusconchoniusreared in the temperature/ration experiment. 147

Effects of different temperatureand ration levels on FAP in hearttissues of Puntius conchonius 149

Heart FAP as a function of fish weight for Puntiusconchoniusreared in the temperature/ration experiment 151

Weight-specificrate of oxygen consumption as a function of fishweight and temperature in Puntius conchonius 153

Effects of different temperature and photoperiodlevels on FAPin brain tissues of Puntius conchonius 155

Effects of different temperature and photoperiodlevels on FAPin heart tissues of Funtius conchonius , " .157

xi

CHAPTER 1

INTRODUCTION

Overview

Managers of living aquatic resources are becoming increasinglyconcernedabout

the welfare of fishery stocks under continuedexploitation. An understandingof the

sustainable yieldof a stockis essential for draftingand implementing effectivefisheries

management policies. Determination of age andgrowth rates is perhaps the most

fundamental componentof this type of assessment (Ricker 1979). Age data, in

conjunction with length and weight measurements, can give informationon life history

parameters (natality, age at recruitmentand sexualmaturity, longevity, and mortality)

necessary to evaluatethe overall conditionof thestock.

Traditional methodsof age determination in fish have included such indirect

measuresas length-frequency analysisor the quantification of rhythmic events in hard

tissues (BagenalI974). However, results from manylength-frequencystudies have been

inconclusive because: 1) there is difficultyin obtaining adequate, representative numbers

of specimensin a full range of life stages;2) sampling introduces gear bias; 3) modes for

older fish are difficult to distinguish due to asymptotic growth and variability within

cohorts; and 4) cohorts are ill-definedin fish species withprotractedrecruitmentperiods

(Neal & Maris, 1985).

Use of rhythmic events registered in hard tissues (e.g., bands in scales, otoliths,

fin spines,vertebrae) has been the most productive technique for chronologicalage

estimation in manyfish species. These conventional techniques are often inadequate

when applied to tropical fishes (Brothers 1987). Even with the discovery of daily growth

increments in theotoliths of juvenile fishes (Pannella 1980), Prince & Lee (1991) have

shown that thecorrespondence betweendays andincrements can break down in older

1

individuals. These problems have led researchers to the realization that alternative,

innovative methods of age determination are desirable to obtain the information required

to manage fish populations.

The quantitative assessment of cellular metabolites accumulated during the process

of physiological aging has been suggested as an alternative method for accurate estimation

of chronological age. One class of metabolites, commonly termed 'fluorescent age

pigments' (pAP), occurs in all animals studied to date, and is considered by many to be

the most clearly discernible evidence of aging in the cell (Sohal & Wolfe, 1986). Not

degradable by normal enzymatic processes, FAP polymers accumulate in cells over time,

and their quantity may provide an index of the relative age of the cell, tissue, and

ultimately the whole animal. Thus far, the study ofFAP has centered mostly on

gerontological research, but interest is growing among field biologists in the possible

application of these methodologies to study rates of aging, estimate chronological age,

and predict longevity in wild stocks. Early attempts to determine the age of animals in the

field using this approach have met with limited success because of high variability in FAP

levels and inadequate sampling. A satisfactory relationship between known chronological

age and FAP content has yet to be validated.

Cellular genesis ofFAP

Events leading to the production ofFAP are generally well known. The term

FAP is used to describe a heterogeneous family of compounds (pre-ceroids, ceroids, and

lipofuscins) which accumulate in the cell as a result of both normal metabolic activity and

general molecular 'wear and tear'. First described by Hannover (1842), age pigments

can be observed as yellowish-brown, polymorphic granules in the cytoplasm. Chemical

analysis has revealed that they are highly polymerized collections of lipids, protein

fragments, and acid-hydrolysis resistant residues (Bjorkerud 1964). They are believed to

2

collect as the result of numerous intracellular digestions by lysosomes of peroxidized

lipids and damaged proteins and organelles (Tsuchida et al. 1987).

Free radical production during cellular metabolism is considered the starting point

for age pigment formation (see reviews by Hannan 1984, Porta 1991). Superoxide

radicals produced in the respiratory chain are converted to hydrogen peroxide in the

presence of the superoxide dismutase (Fig. 1-1). Under homeostatic conditions,

hydrogen peroxide is subsequently converted to oxygen and water by catalase or

glutathione peroxidase. It is inevitable that some hydrogen peroxide will, in the presence

of divalent metals, dissociate into a hydroxide ion and a hydroxyl radical (Fig. 1-1).

Hydroxyl radicals are highly reactive, and there are no known protective mechanisms

against their attack on biomolecules.

Hydroxyl radicals may damage many types of biomolecules, but the unsaturated

fatty acids and membrane phospholipids are probably the most important to consider in

terms of FAP formation. In the example illustrated in Fig. 1-2, attack on linolenic acid by

a hydroxyl radical results in the formation of a lipid radical, which is oxidized in

subsequent steps to form an endoperoxide. Endoperoxides dissociate into two smaller

lipid radicals and malondialdehyde, a powerful bifunctional crosslinking agent in the

cytoplasm. Free radical-mediated lipid peroxidation is further propagated by the two new

lipid radicals unless controlled by antioxidants such as alpha-tocopherol (vitamin E).

Malondialdehyde, a by-product of lipid peroxidation, can crosslink with the amine

groups of any two biological compounds to form autofluorescent condensation products

such as amino-imino propene Schiff base fluorophores (Tappel, 1975), 1,4­

Dihydropyridine-3,5-dicarbaldehydes (Kikugawa and Ido 1984), and

polymalondialdehyde (Gutteridge et al. 1977) (Fig. 1-3). Biomolecules crosslinked

through these reactions are inactivated, and the polymers are eventually engulfed by

autophagic lysosomes for intracellular digestion. Crosslinked regions are highly resistant

3

--_._ .._......

to degradation through normalenzymatic processes, and graduallyaccumulate in number

and size over time (Sohal and Wolfe 1986, Goebel 1987)as age pigment(pre-ceroid,

ceroid, and lipofuscin) granules (Fig. 1-4).

Manyfactors are thought to affect the rate ofFAP genesis. Theseinclude

metabolic rate (Sohal& Donato 1979), temperature (McArthur& Sohal 1982), degree of

fatty acid and membranephospholipid saturation (Sohalet al. 1984),production of

specificproteolytic enzymes (Katz& Shanker 1989), disease state (Hammer and Karakiri

1987,Koppang 1987,Hall et al. 1989), and in particular the efficiencyof antioxidant

defensemechanisms which intercept damaging freeradicals (Armstrong 1984,Brizzee et

al. 1984). Cell age may affect the abilityto preventfree radicaldamagethrough changes

in theexpression of antioxidant enzymes (Sohalet al. 1983); FAP may accumulate at a

faster rate in the tissues of senescent or diseasedanimals(Timiras 1974, Donato & Sohal

1978, Armstrong 1984, Nandy 1985,Porta 1987).

Quantification ofFAP

Agepigmentshave beenobserved at both the histological and biochemical levels.

Traditional histological methodshaveincluded directquantification of age pigment

granulesin stained tissues by lightmicroscopy (Strehler1964), epifluorescence

microscopy (Brizzee& Jirge 1981), and transmission electron microscopy (TEM)(Agius

& Agbede 1984). ProponentsofTEM methodshaveattemptedmorphometric

quantification throughdigitization of micrographs. Unfortunately, TEMmethods are time

consuming and expensive. More significantly, morphometric quantification is subject to

analytical biasesassociated with the selection of specific areas in the tissue and cell for

countinggranule numbersand for calculating percentvolumeof the celloccupied by the

granules.

An alternative method, introduced by Tappeland coworkers (Chio& Tappel

1969,Chio et al. 1969,Fletcher et al. 1973,TappeI1975), involves the measurement of

4

autofluorescent amino-imino propenes associated with age pigment polymers in whole

tissues. This method has gained widespread use in gerontological research on a wide

variety of animals, tissues, and model systems. It involves the homogenization,

extraction and fluorometric quantification of the soluble FAP present in the tissue. Early

studies used only the chloroform or lipid-soluble portion of these compounds, but it is

now known that some FAP's are soluble in more polar solvents as well (Tsuchida et al.

1987). Although the overall chemical composition of FAP products is diverse, the

presence of the autofluorescent crosslinks is universal. Thus, indirect measurement of

soluble FAP by fluorometry produces an index of the amount of all types of FAP present

in the tissue and provides a higher degree of overall replicability than light microscopic

and electron microscopic methods.

Eldred et al. (1982), Eldred (1987), and Porta (1987) have argued that Tappel's

fluorometric technique does not measure the fluorescence of whole lipofuscin granules,

but instead it measures direct products of lipid peroxidation not necessarily present in the

granules. This view has been based on comparison of the fluorescent properties of

solvent extracts with the properties of fluorescent granules visible by light microscopy.

The possibility exists that FAP measured fluorometrically represent more immediate

effects of lipid peroxidation associated with damaged membrane phopholipids (Tappe!

1975, Vladimirov 1986), damaged mitochondria (Dillard & TappeI1971, Tangeras

1983), or cross-linked proteins or oligonucleotides present in the cytosol, not yet

incorporated into large polymers in the lysosome (Kikugawa et al. 1989). Regardless of

the source of the measured fluorescence, Tappel's method has been successfully applied

to a large number of tissue systems over the past twenty years.

Age pigments in aquatic animals

Although age pigments have been described in a wide variety of animals, little

work has focused on aquatic animals such as fish and crustaceans. Some fishes are

5

potentially excellent experimental subjects for gerontological research (Comfort 1960,

1961, 1963; Woodhead 1979, 1980, 1985) because they have relatively short lifespans,

are easy to breed and maintain in captivity, and their metabolic rates can be manipulated.

Most researchers who have studyed FAP in fishes have used histological

procedures. Aboim (1946) reported age pigments in the interrenal cells of the

elasmobranch Raja clavata. Other studies have noted the presence of age pigment

granules in interrenal cells of Prionurus microlepidotus, Pterois lunulata, and Paralichthys

olivaceous (Oguri & Hibiya 1957), and more recently in the same tissues of Lophius

litulon and Triakis~ (Oguri 1986). Oguri (1986) commented on the abundant

occurrence of age pigment granules in fish specimens of large size and "advanced age".

Additional studies have focused on the accumulation of lipofuscin, haemosiderin, and

melanin pigments in macrophages, relating pigment content to fish health and age (Grove

1968; Agius 1980, 1981; Agius & Agbede 1984; Wolke et al. 1981; Brown & George

1985). Brown & George (1985) found that the number of macrophage age pigment

aggregates was more dependent on age than on the fish's health (as indicated by incidence

of parasitism and condition index). Aloj Totaro and coworkers have examined effects of

environmental pollutants on accumulation of lipofuscin granules in spinal ganglia of the

electric ray, Tomedo mannoratus, fmding higher levels in specimens from copper

polluted areas (Aloj Totaro et al. 1985; Aloj Totaro & Pisanti 1987). Unfortunately, no

consideration was given to the chronological age of their samples.

Fewer investigators have measured FAP in fish fluorometrically. Luo & Hultin

(1986) found that chloroform soluble FAP increased with age (estimated with unvalidated

scale annuli) in the skeletal muscle of winter flounder (pseudopleuronectes americanus).

Hammer & Madhusudhana-Rao (1987) found that presence or absence of dietary vitamin

E affected the final quantity of FAP in Oreochromis aurea brain, but had no effect on

levels in four other tissue types. Hammer & Karakiri (1988) measured FAP content in

6

field-captured Limanda limanda and found that individuals infected with lymphocystis

had significantly higher levels of the pigment Finally, Hammer (1988) tested effects of

temperature on FAP levels in larval Esox lucius, finding higher rates of accumulation in

larvae reared at suboptimal temperatures. Hill & Radtke (1988) measured FAP in

Dascyllus albisella brains, fmding a positive relationship between FAP content and

specimen size and age (estimated from unvalidated annuli in sagittae). Similar studies

have been conducted on dover sole and trout (Vemet et al, 1988) as well as larval

grunion, seabass, and halibut (Mullin & Brooks 1988) with variable results.

Ettershank (1983, 1984a, 1984b, 1985) was the first to apply FAP techniques to

crustaceans. He attempted to resolve age cohorts in Antarctic krill (Euphausia s\Werba),

which were previously thought to live three years based on length frequency

distributions. Ettershank found modal progressions of FAP content, assumed to be

annual, which led him to the conclusion that krill live up to six or seven years. Nicol

(1987) applied Ettershank's methods to field caught specimens of the euphausid,

Meganyctiphanes norvegica, and found a linear increase in FAP content with body

weight; however, fluorescence could not be compared to chronological age. More

significantly, Nicol pointed out a major flaw in Ettershank's use of formalin as a

specimen preservative by demonstrating enhanced fluorescence of samples after storage

in this powerful cross-linking agent. Subsequent investigators have committed similar

errors in the use of formalin to preserve spider crab larvae~ araneus; Hirche and

Anger 1987) and krill ill:. superba; Berman et al. 1989). Crossland et al. (1988) found an

increase in FAP in lobster (Panulirus cygnus) and scampi (Metanephrops andamanicus)

eye and tail muscle tissues with increasing size, however chronological age of their

specimens was unknown. Sheehy and Ettershank (1989) extracted FAP from whole

bodies ofDaphnia carinatl!, fmding an increase in absolute FAP content with increasing

age, but a decrease in the weight-specific content with increasing age. The decrease in

7

weight-specific FAP can be attributed to the dilutionof FAP by continuedtissuegrowth.

Most recently,Nicolet al. (1991)maintainedpopulations of E. superbain the laboratory

and measuredsignificant increases in absoluteand weight-specific FAP after one yearof

captivity.

PatternsofFAP accumulation have beenexaminedin only two molluscan species.

Nicol (1987)extractedFAP from mantle tissueof squid, nex illecebrosus, finding

increases in FAP contentwith increasingsize. Chronological age of his specimens was

not known. Clarkeet al. (1990)examinedFAP accumulation in field specimens of the

troehid gastropod Monodonta lineatawhose chronological age had beenestimatedfrom

shell annuli. Whole organFAP contentin brain and digestive gland increasedwith

estimated age, but in a variable manner. Weight-specific FAPvaluesdecreased with age,

a trend attributedto tissue growth.

Research problem

Little is yet understood about the relationship between FAP contentand

organismalage and growth in aquaticpoikilotherms. Even though the previously

mentioned studiesseempromising, results are preliminary at best. Each fluorescence

study has usedsomevariation of a technique originally developed. by Bligh & Dyer

(1959) for whole lipid extraction and later modified by Fletcheret al. (1973) for the

measurementofFAP. However,few of thesevariations have beenreplicable basedon

informationin the published reports of the studies. Many aspectsof the methodologies

have simplybeen assumed to workproperly, with no potentially detrimental effects to the

final expressionof fluorescence takeninto account. In addition, the manner in which

fluorescence results havebeenreported has beenobscure at best, with no standardization

or valid comparison possiblebetweenstudies.

8

As a result of the above limitations, it has yet to be demonstrated that FAP can be

used as an effective tool for predicting age of animals from the field. Only three

investigators (Vernet et al. 1988, Hammer 1988, and Mullin & Brooks 1988) have

examined FAP in fish of known age. Vernet et al. (1988) measured FAP content in three

year classes of laboratory reared Oncorhyncus mykiss (of mixed brood stock origin),

with highly variable results. Mullin & Brooks (1988) measured FAP in three larval fish

species (California grunion, Leuresthes~; white seabass, Atraetoscion nobilis; and

California halibut, Paralichthys californicus) up to 30 d in age, finding negative

correlations between age and FAP content. Hammer's (1988) use of methanol to

preserve larval pike specimens probably altered the final expression ofFAP fluorescence

in a fashion similar to Ettershank's krill samples (Nicol 1987). Even if a stable

relationship between age and FAP were verified, its usefulness for aging field specimens

would depend on a demonstration that information on FAP content can improve on the

accuracy of age estimates presently available for fish through other means.

Further - and thus far not considered in age estimations using FAP - a number of

environmental and physiological factors are likely to influence the rates ofFAP

accumulation in aquatic poikilotherms. Temperature, light, and diet are foremost among

the exogenous factors known to affect ontogenetic changes in growth (Kinne 1960;

Paloheimo & Dickie 1966; Brett 1979; Wurtsbaugh & Cech 1983; Cui & Wooton 1988).

Some of these factors have been implicated in determining the rate of FAP formation in

various poikilothermic (Sheldahl & Tappe11974; Ragland & Sohal 1975; Sohal & Donato

1979; Sohal et al. 1985) and homeothermic animals (porta 1987) and in liposomal model

systems (Petkau 1986). If these factors influence both growth and rates of FAP

accumulation in animals, they are likely to induce variability in this rate at a given age as

well as cwnulative amounts of FAP over longer periods of time. This is significant if

FAP quantities are to be used to predict the chronological ages of wild specimens.

9

Study objectives

In this dissertation, I study patterns of FAP accumulation and variation in a

freshwater <Puntiusconchonius Hamilton) and a euryhaline <Oreochromis rnossambicus

Peters) fish to evaluate the potential for accurate age estimation for management of

fisheries stocks. Specific objectives are: 1) to reassess aspects of FAP assay

methodologies and determine effects of these procedures on final FAP fluorescence

expression; 2) to establish the relationship between FAP content and known

chronological age in the selected species and to build age-predictive models using

multivariate analysis; and 3) to examine effects of specific factors (body size, ambient

temperature, food ration, and photoperiod) which may induce variability in the

relationship between age and FAP content.

Completing the first objective is critical to accomplishing the other two objectives.

In Chapter 2, various aspects of sample handling and extraction procedures are

investigated for their effects on final expression of FAP fluorescence in tissue samples.

This work includes evaluation of the effects of sample storage, extraction, and extract

storage procedures for a variety of post-mitotic tissues of the selected fish species. Once

procedures for extraction and measurement of the compounds have been tested and

defined, the methods are applied to tissues of the study species to determine the basic

relationship between chronological age and FAP content (Chapters 2 and 3). After the

patterns ofFAP accumulation have been established, I incorporate fish size, otolith size,

organ size, and FAP data into various multiple linear regression models (Boehlert 1985)

to examine the utility ofFAP data for chronological age prediction (Chapter 3). In

Chapter 4, experiments are conducted to determine the degree to which FAP content may

vary in P. conchonius tissues under different environmental and physiological conditions.

10

Literature cited

Aboim, A. N. 1946. L'Organe interrenal des cyclostomes et des poissons. Portugal.

Acta. Biol, Ser. A. 1: 353-383.

Agius, C. 1980. Phylogenetic development of melano-macrophage centres in fish. J.

Zool. 191: 11-31.

Agius, C. 1981. Preliminary studies on the ontogeny of the melano-macrophages of the

teleost hemopoietic tissues and age-related changes. Dev. Compo Anat. 5: 597­

606.

Agius, C., and S. A. Agbede. 1984. An electron microscopical study on the genesis of

lipofuscin, melanin and haemosiderin in the haemopoietic tissues of fish. 1. Fish

BioI. 24: 471-488.

Aloj Totaro, E., F. A. Pisanti, and P. Glees. 1985. The role of copper level in the

formation of neuronal lipofuscin in the spinal ganglia of Tomedo m. Mar.

Environ. Res. 15: 153-163.

Aloj Totaro, E., and F. A. Pisanti. 1987. Lipofuscinogenesis and environmental stress:

study in Tomedo m. neurons. pp. 133-146 In: Aloj Totaro, E. and FA. Pisanti.

Advances in Age Pigments Research. Advances in the Biosciences. Pergamon

Press, N.Y.

Armstrong, D. 1984. Free radical involvement in the formation of lipopigments. pp. 129­

141 In: Armstrong, D., R.S. Sohal, R.G. Cutler and T.F. Slater (Eds.) Free

Radicals in Molecular Biology, Aging, and Disease. Aging. Raven Press, N.Y.

Bagenal, T. B.(ed.) 1974. The ageing of fish. Unwin Brothers Ltd., England, 234p.

Berman, M. S., A. L. McVey, and G. Ettershank. 1989. Age determination of Antarctic

krill using fluorescence and image analysis of size. Polar Biol. 9: 267-271.

Bjorkerud, S. (1964). Isolated lipofuscin granules - survey of a new field. Adv.

Gerontol. Res. 1: 257-288

11

----_ .._-_ ..

Bligh, E. G., and W. 1. Dyer. 1959. A rapid method of total lipid extraction and

purification. Can. J. Biochem. and Physiol, 37: 911-917.

Boehlert, G. W. 1985. Using objective criteria and multiple regression models for age

determination in fishes. Fish. Bull., U.S. 83:103-117.

Brett, J. R. 1979. Environmental factors and growth. pp. 599-675 In: Hoar, W.S., DJ.

Randall and J.R. Brett, (Eds.) Fish Physiology. Academic Press, Inc.

Brizzee, K. R., D. E. Eddy, D. Harman, and 1. M. Ordy. 1984. Free radical theory of

aging: effect of dietary lipids on lipofuscin accumulation in the hippocampus of

rats. Age 7: 9-15.

Brizzee, K. R., and S. K. Jirge. 1981. Fluorescent microscope techniques for the

visualization and histological quantification of autofluorescent lipofuscin bodies in

brain tissues. Johnson, J. E. (Ed.) Current Trends in Morphological Techniques.

CRC Press.

Brothers, E. B. 1987. Methodological approaches to the examination of otoliths in aging

studies. pp.319-330 In: Summerfelt, R. C. and G. E. Hall (Eds.) The Age and

Growth of Fish. The Iowa State University Press, Ames, Iowa.

Brown, C. L., and C. 1. George. 1985. Age-dependent accumulation of macrophage

aggregates in the yellow perch, Perca flavescens (Mitchill). J. Fish Diseases 8:

135-138.

Chio, K. S., and A. L. Tappel, 1969. Synthesis and characterization of the fluorescent

products derived from malondialdehyde and amino acids. Biochem 8: 2821­

2827.

Chio, K. S., U. Reiss, B. Fletcher, and A. L. Tappel, 1969. Peroxidation of subcellular

organelles: formation of lipofuscin-like fluorescent pigments. Science 166: 1535­

1536.

12

Comfort, A. 1960. The effect of age on growth-resumption in fish CLebistes) checked by

food restriction. GerontoI. 4: 177-186.

Comfort, A. 1961. The longevity and mortality of a fish CLebistes reticulatus Peters) in

captivity. Geronto1.5: 209-222.

Comfort, A. 1963. Effect of delayed and resumed growth on the longevity of a fish

CLebistes reticulatus, Peters) in captivity. Gerontol. 8: 150-155.

Clarke, A., M. A. Kendall, and D. J. Gore. 1990. The accumulation of fluorescent age

pigments in the trochid gastropod Monodonta lineata. J. Exp. Mar. BioI. Ecol.

144: 185-204.

Crossland, C. J., G. Denby, B. F. Phillips, and R. Brown. 1987. The use of fluorescent

pigments (lipofuscin) for ageing western rock lobster (Panulirus cygnus) and

scampi CMetanephrops andamanicus): a preliminary assessment CSIRO Mar.

Lab. Rep. No. 195: 1-21.

Cui, Y., and R. J. Wootton. 1988. Bioenergetics of growth of a cyprinid, Phoxinus

phoxinus (L.): the effect of ration and temperature on growth rate and efficiency.

J. Fish BioI. 33: 763-774.

Dillard, C. J., and A. L. Tappel. 1971. Fluorescent products of lipid peroxidation of

mitochondria and microsomes. Lipids 6: 715-721.

Donato, H., and R. S. Sohal. 1978. Age-related changes in lipofuscin-associated

substances in the adult male housefly, Musca domestic;!. Exp. GerontoI. 13: 171­

179.

Eldred, G. E. 1987. Questioning the nature of the fluorophores in age pigments. pp.23­

36 In:. Aloj Totaro, E. and F. A. Pisanti (Eds.) Advances in Age Pigments

Research. Advances in the Biosciences. Pergamon Press, N.Y.

Eldred, G. E., G. V. Miller, W. S. Stark, and L. Feeney-Burns. 1982. Lipofuscin:

resolution of discrepant fluorescence data. Science 216: 757-759.

13

Ettershank, G. 1983. Age structure and cyclical annual size change in the Antarctic krill,

Euphausia superba. Polar BioI. 2: 189-193.

Ettershank, G. 1984a. A new approach to the assessment of longevity in the Antarctic

krill Euphausia superba. J. Crust. BioI. 4: 295-305.

Ettershank, G. 1984b. Methodology for age determination of Antarctic krill using the age

pigment lipofuscin. BIOMASS Handb. No. 26: 1-14.

Ettershank, G. 1985. Population age structure in males and juveniles of the Antarctic

krill, Euphausia superba Dana. Polar BioI. 4: 199-201.

Fletcher, B. L., C. J. Dillard, and A. L. TappeI. 1973. Measurement of fluorescent lipid

peroxidation products in biological systems and tissues. Anal. Chern. 52: 1-9.

Goebel, H. H. 1987. Ultrastructural features oflipopigments in neuronal ceroid

lipofuscinoses, especially in biopsies. pp. 249-275 In: Aloj Totaro, E., P. Glees,

and F. A Pisanti (Eels.) Advances in Age Pigments Research. Advances in the

Biosciences. Pergamon Press, N.Y.

Grove, J. H. 1968. Hemosiderin in bluegill spleens. Trans. Am. Fish. Soc. 97: 48-50.

Gutteridge, J. M. C. 1987. Oxygen radicals, transition metals and aging. pp. 1-22. In:

Aloj Totaro, E., P. Glees, and F. A Pisanti (Eds.) Advances in Age Pigments

Research. Advances in the Biosciences. Pergamon Press, N.Y.

Hall, N. A, B. D. Lake, D. N. Palmer, R. D. Jolly, and A D. Patrick. 1989.

Glycoconjugates in storage cytosomes from ceroid-lipofuscinosis (Batten's

disease) and in lipofuscin from old-age brain. pp. 225-242 In: Porta, E.A (ed),

Lipofuscin and Ceroid Pigments. Plenum Press, New York.

Hammer, C. 1988. Accumulation of fluorescent age pigments in embryos and larvae of

pike, Esox lucius (Esocidae, Teleostei) at different incubation temperatures.

Environ. BioI. Fish. 22:91-99.

14

Hammer, C. and M. Karakiri. 1988. Disease dependent accumulation of fluorescent age

pigments in the common dab <Limanda limanda, Teleostei). pp. 445-446 In: 2s.­

Nagy, I. (00.) Lipofuscin - 1987: State of the Aart, Akademiai Kiad6, Budapest

and Elsevier Science Publishers, Amsterdam.

Hammer, C., and G. Madhusudhana-Rao. 1987. Individual and feed dependent

accumulation of fluorescent age pigments in five organs of tilapia (Oreochromis

~. Zeitschr. Angew. Z001. 74: 275-292.

Hannover, A. 1842. Mikroskopische Undersolgeser afNervensystemet. K. Dan.

Vidensk. Selsk. Nat. Math. Afb. 10: 1-112.

Harman, D. 1984. Free radicals and the origination, evolution, and present status of the

free radical theory of aging. pp. 1-12 in Armstrong, D., R. S. Sohal, R. G.

Cutler, and T. F. Slater (OOs.) Free Radicals in Molecular Biology, Aging, and

Disease. Raven Press, N.Y.

Hill, K. T., and R. L. Radtke. 1988. Gerontological studies of the damselfish, Dascyllus

albisella. Bull. Mar. Sci. 42: 424-434.

Hirche, H.-J., and K. Anger. 1987. The accumulation of age pigments during larval

development of the spider crab, Hyas araneus (Decapoda, Majidae). Compo

Biochem. Physio1. 88B: 777-782.

Katz, M. L., and M. J. Shanker. 1989. Development of lipofuscin-like fluorescence in

the retinal pigment epithelium in response to protease inhibitor treatment. Mech.

Age. Dev. 49: 23-40.

Kikugawa, K., Ido, Y. 1984. Studies on peroxidized lipids. V. Formation and

characterization of 1,4-dihydropyridine-3,5-dicarbaldehydes as model of

fluorescent components in lipofuscin. Lipids 19: 600-608

15

Kikugawa, K., T. Kato, and A. Iwata. 1989. A tetrameric dialdehyde formed in the

reaction of butyraldehyde and bebenzylamine: a possible intermediary component

for protein cross-linking induced by lipid oxidation. Lipids 24: 962-969.

Kinne, O. 1960. Growth, food intake, and food conversion in a euryplastic fish exposed

to different temperatures and salinities. Physiol. Zool. 33: 288-317.

Koppang, N. 1987. Hereditary ceroid lipofuscinogenesis in English setter dogs. A model

for human neuronallipofuscinogenesis and aging. pp. 185-195 In: Aloj Totaro,

E., P. Glees, and F. A. Pisanti (Eds.) Advances in Age Pigments Research.

Advances in the Biosciences. Pergamon Press, N.Y.

Luo, S.-W., and H. O. Hultin. 1986. Effect of age of winter flounder on some properties

of the sarcoplasmic reticulum. Mech. Age. Dev. 35: 275-289.

McArthur, M. C., and R. S. Sohal. 1982. Relationship between metabolic rate, aging,

lipid peroxidation, and fluorescent age pigment in milkweed bug, Oncopeltus

fasciatus (Hemiptera). J. Gerontol. 37: 268-274.

Mullin, M. M., and E. R. Brooks. 1988. Extractable lipofuscin in larval marine fish.

Fish. Bull., U.S. 86: 407-415.

Nandy, K. 1985. Lipofuscin as a marker of impaired homeostasis in aging organisms.

pp. 139-148 In: Davis, B.B. and W.G. Wood (Eds.) Homeostatic Function and

Aging. Raven Press, N.Y.

Neal, R. A. and R. C. Maris. 1985. Fisheries biology of shrimplike animals. pp.2-11O

In: Provezano, A.J., Jr. (Ed.), The Biology of Crustacea, Vol. 10, Economic

Aspects: Fisheries and Culture. Acad. Press, New York.

Nicol, S. 1987. Some limitations on the use of the lipofuscin ageing technique. Mar.

BioI. 93: 609-614.

16

Nicol, S., M. Stolp, and G. W. Hosie. 1991. Accumulation of fluorescent age pigments

in a laboratory population of Antarctic krill Euphausia superba Dana. J. Exp. Mar.

BioI. Ecol. 146: 153-161.

Oguri, M. 1986. On the yellowish brown pigment granules in the interrenal cells of

anglerfish and leopard shark. Bull. Jap. Soc. Sci. Fish. 52: 801-804.

Oguri, M., and T. Hibiya. 1957. Studies of the adrenal glands ofteleosts - II. On the

adrenal tissues of 15 species of fishes. Bull. Jap. Soc. Sci. Fish. 23: 144-149.

Paloheimo, J. E., and L. M. Dickie. 1966. Food and growth of fishes. II. Effects of

food and temperature on the relation between metabolism and body weight. J.

Fish. Res. Bd. Can. 23: 869-908.

Pannella, G. 1980. Growth patterns in fish sagittae. pp. 519-560 In: Rhoads, D.C. and

R.A. Lutz (Eels.). Skeletal Growth of Aquatic Organisms: Biological Records of

Environmental Change. Plenum Press, New York. 588p.

Petkau, A. 1986. Protection and repair of irradiated membranes. pp.481-508 In:

Johnson, J.E., R. Walford, D. Harman, and 1. Miguel (Eds.) Free Radicals,

Aging, and Degenerative Processes. Modern Aging Research. Alan R. Liss, Inc.,

New York.

Porta, E. A. 1987. Tissue lipoperoxidation and lipofuscin accumulation as influenced by

age, type of dietary fat and levels of vitamin E in rats. pp. 37-74 In: Aloj Totaro,

E. and F. A. Pisanti (Eds.) Advances in Age Pigments Research. Advances in

the Biosciences. Pergamon Press, N.Y.

Porta, E. A. 1991. Advances in age pigment research. Arch. Gerontol. Geriatr. 12: 303­

320.

Prince, E. D., D. W. Lee, J. R. Zweifel, and E. B. Brothers. 1991. Estimating age and

growth of young Atlantic blue marlin Makaira nigricans from otolith

microstructure. Fish. Bull., U.S. 89: 441-459.

17

Ragland, S. S., and R. S. Sohal. 1975. Ambient temperature, physical activity and aging

in the housefly, Musca domestica. Exp. Gerontol. 10: 279-289.

Ricker, W. E. 1979. Growth rates and models. pp. 677-743 In: Hoar, W.S., D.J.

Randall and J.R. Brett (Eds.) Fish Physiology. Academic Press, Inc., New

York.

Sheehy, M. R. J., and G. Ettershank. 1989. Extractable age pigment-like

autofluorescence and its relationship to growth and age in the water-flea Daphnia

carlnata King (Crustacea: Cladocera). Aust. 1. Zool. 36: 611-625.

Sheldahl, J. A, and A. L. Tappe1. 1974. Fluorescent products from aging Drosophila

melanogaster: An indicator of free radical lipid peroxidation damage. Exp.

Gerontol. 9: 33-41.

Sohal, R. S., and H. Donato Jr. 1979. Effect of experimental prolongation of lifespan on

lipofuscin content and lysosomal enzyme activity in the brain of the housefly,

Musca domestica. 1. Gerontol. 34: 489-496.

Sohal, R. S., K. J. Farmer, R. G. Allen, and N. R. Cohen. 1983. Effect of age on

oxygen consumption, superoxide dismutase, catalase, glutathione, inorganic

peroxides and chloroform-soluble antioxidants in the adult male housefly, Musca

domestica. Mech. Age. Dev. 24: 185-195.

Sohal, R. S., R. G. Bridges, and E. A Howes. 1984. Relationship between lipofuscin

granules and polyunsaturated fatty acids in the housefly, Musca domestica. Mech.

Age. Dev. 25: 355-363.

Sohal, R. S., A MUller, B. Koletzko, and H. Sies. 1985. Effect of age and ambient

temperature on n-pentane production in the adult housefly, Musca domestica.

Mech. Age. Dev. 29: 317-326.

18

Sohal, R. S., and L. S. Wolfe. 1986. Lipofuscin: characteristics and significance. pp.

171-183 In: Swaab, D.F., E. Fliers, M. Mirmiran, W.A. Van Gool, and F. Van

Haaren (Eds.) Prog. Brain Res. Elsevier Science Publishers, Amsterdam.

Strehler, B. L. 1964. On the histochemistry and ultrastructure of age pigment. pp. 343­

384 In: Strehler, B.L. (Ed.) Adv. Gerontol. Res .. Academic Press, N.Y.

Tangeras, A. 1983. Iron content and degree of lipid peroxidation in liver mitochondria

isolated from iron-loaded rats. Biochim. Biophys. Acta 757: 59-68.

Tappel, A. L. 1975. Lipid peroxidation and fluorescent molecular damage to membranes.

pp. 145-170 In: Trump, B. F. and A. V. Arstila (Eds.) Pathobiology of Cell

Membranes.

Timiras, P. S. 1974. Degenerative changes in cells and cell death. pp. 429-449 In:

Timiras, P.S. (Ed.) Developmental Physiology and Aging. MacMillan, N.Y.

Tsuchida, M., T. Miura, and K. Aibara. 1987. Lipofuscin and lipofuscin-like

substances. Chern. Phys. Lip. 44: 297-325.

Vernet, M., J. R. Hunter, and R. D. Vetter. 1988. Accumulation of age pigments

(lipofuscin) in two cold-water fishes. Fish. Bull.,U.S. 86: 401-407.

Vladimirov, Y. A. 1986. Free radical lipid peroxidation in biomembranes: mechanism,

regulation, and biological consequences. pp. 141-195 In: Johnson, T.E., R.

Walford, D. Harman and J. Miquel (eds.) Modern Aging Research: Free

Radicals, Aging, and Degenerative Processes. Alan R. Liss, Inc., New York.

Wolke, R. E., C. J. George, and V. S. Blazer. 1981. Pigmented macrophage

accumulations (MMC, PMB): possible monitors of fish health. Proceedings of

the US-USSR Symposium on Parasites and Pathogens of the World Oceans,

Leningrad

Woodhead, A. D. 1979. Senescence in fishes. Symp. Zool. Soc. Lond. 44: 179-205.

Woodhead, A. D. 1980. Fish in studies of aging. Exp. Gerontol. 13: 124-140.

19

Woodhead, A. D. 1985. Feral Fishes. In: Lints, F. A. (ed.). Non-mammalian models for

research on aging. Interdiscipl. Topics Geront. 21: 22-50. Karger, Basel.

Wurtsbaugh, W. A., and J. 1. Cech. 1983. Growth and activity of juvenile

mosquitofish: temperature and ration effects. Trans. Am. Fish. Soc. 112:

653-660.

20

Fig. 1-1. Modulation of superoxide radicals produced through normal cellular

metabolism. Adapted from Zs.-Nagy (1987).

21

_..__.._---------_._----_.- -.

Catalase

Glutathioneperoxidase

Superoxidedismutase

SuperoxideRadical

120~ I2H+

°2IH2021----->~H20 +~ O2

Fe (II)

Fe(III)

IOH"1Hydroxyl Radical

22

Fig. 1-2. Free radical-mediated lipid peroxidation and malondialdehyde production.

Adapted from Petkau (1986).

23

_•.__.._-_._-_ _-

-

Hydroxyl Radical Attack

OOH Linolenic Acid

OOH Lipid Radical

OOH Diene Conjugate

OOH Peroxy Radical

OOH Endoperoxide

OOH

0-0-

~ 0-0

Fe complexes ~

~ 0- 0-

~__~OOH

~rn Malondialdehydeo 0

24

Fig. 1-3. Fluorescent crosslinkage reactions of malondialdehyde. Adapted in part from

Gutteridge (1986).

25

MalondialdehydeOHC-CH 2 -CHO

OHi1;H0

NI

R

1,4-Dihydropyrldine­3,5-dicarbaldehydes

(390/450nm)

II+[OHC-C-CH=ln

Polymalondialdehyde(370/480nm)

~N=CH-CH=CHOHr+ R,NH2

"N=CH-CH=CH-NHR 2

Amino-imino propeneConjugated Schiff Base Fluorophores

(EX 340·360nm. EM 410·470nm)

26

Fig. 1-4. Summary schematic of processes leading to fluorescent age pigment

formation.

27

Unsaturated Fatty Acids & Phospholipids

Free Radical Attack

Peroxidation

Malondialdehyde & OtherCarbonyl Compounds

Crosslinkages

Primary Amine Groups- phospholipids- proteins- free aminoacids- nucleicacids

Cross-Linked Macromolecules ContainingConjugated Schiff Base Fluorophores

-(N-C=C-C=N)-

fusion withIysosomes

Fluorescent Age Pigments(Pre-ceroid, Ceroid, Lipofuscin)

28

CHAPTER 2

CRITICAL ASPECTS OF FLUORESCENT AGE PIGMENTMETHODOLOGIES: MODIFICATION FOR ACCURATE ANALYSIS AND

AGE ASSESSMENTS IN AQUATIC ORGANISMS

Introduction

Age pigments, first described by Hannover (1842), are now recognized as being

one of the most clearly discernible markers of aging in cells (see review by Hammer and

Braum 1988). Age pigments form under normal and pathological conditions in cells as

the result of polymerization reactions between oxidized lipids and proteins (Bjorkerud

1964, Davies 1988, Katz 1990). The residues from these reactions have been variously

classified by their chemical and physical properties, solvent solubilities, fluorescent

properties, and stage of formation, being assigned names such as "lipofuscin" (Hueck

1912), "wear and tear pigment" (HamperlI934), and "ceroid" (Lillie et al. 1941).

Derived from a diverse array of sources, age pigments can now be characterized as

containing numerous autofluorescent components, ranging in emission color from blue

(Hyden and Lindstrom 1950, Barden 1980, Porta et al. 1988) to yellow (Eldred et al.

1982, Katz et al. 1984, Eldred and Katz 1988), many of which have been found to

accumulate with age (Shimasaki et al. 1980, Eldred et al. 1982, Sohal et al. 1983,

Csallany et al. 1984). Many attempts have been made to identify the fluorescent

components of age pigments, however, due to their complex chemical composition, this

topic is likely to remain controversial for some time. Possible sources of fluorescence

include conjugated Schiff bases (N,N' disubstituted-l-amino-3-iminopropenes; Chio and

Tappe11969, TappeI1975), polymalondialdehyde (Gutteridge et al. 1977), 1,4­

dihydropyridine-3,5-dicarbaldehydes (Kikugawa and Ido 1984), and membrane bound

flavins (Kitani et al. 1988, Nokubo et al. 1988, 1989).

29

In 1973,Fletcher et aI. introduceda relatively simple spectrofluorometric assay

for the measurementof blue-emitting fluorophores, assumed to be related to age

pigments, from whole tissue samples. Numerous investigators have since adopted

Fletcher et al. 's method, demonstrating a relationship betweenorganismal age and

fluorescence of tissue extracts (Tappelet al, 1973,Miquel et aI. 1978, Bridges and Sohal

1980, McArthur and Sohal 1982, Ettershank et aI. 1983). The use of the terms

"lipofuscin" or "lipofuscin-like" to describesolvent-extractedfluorophores is currently

disputed (Eldred 1987,Eldred and Katz 1989)due to the differentialpropertiesof

fluorescence in tissue extracts comparedto histologically observed preparations. It

should at this point be recognized that Fletcher et al.'s method is probablynot specific to

a single type or class of age pigment; rather it is a more general assay for the blue-emitting

components of age-related fluorophores generated by lipid peroxidationprocesses in the

cell. Thus, for the purposes of this study, we will adopt the term 'fluorescent age

pigment', hereinafter referred to as 'FAP' (Sohal and Buchan 1981,Hammer and Braum

1988), to describe age-relatedfluorophores which have been measured

spectrofluorometrically. Use of more specific terms such as 'lipofuscin' or 'ceroid'

should be reserved for describing age pigments observed histologically.

Spectrofluorometric measurement of metabolically accumulatedFAP compounds

has recently been proposed as a means for estimatingage in wild specimens of aquatic

species. Ettershank (1983, 1984 a.b, 1985)first applied Fletcher et aI. 's (1973) FAP

assay to delineate age cohorts in Antarctickrill <Euphausia superba). Modifications of

Fletcher et al.'s (1973) and Ettershank's (1983, 1984 a.b) protocols have since been

applied to a number of teleost (Luo and Hultin 1986,Hammer and Madhusudhana-Rao

1987,Hammer 1988, Hammer and Karakiri 1988, Hill and Radtke 1988, Mullin and

Brooks 1988, Vernet et aI. 1988) and crustacean (Hirche and Anger 1987, Nicol 1987,

Berman et aI. 1989, Sheehy and Ettershank 1989) species.

30

.__.-----....._--

Results from these preliminarystudieshave been cursory and inconclusive with

regard to the applicationof this methodtowardage estimationfor population analyses.

For example, relativelyfew investigators have tested the method on animals of known

chronological age. Conversely, those studies using sample tissue of known age have

used samplingperiods insufficientto determine long-termontogeneticchanges in FAP

content. In addition, a survey of the methods thus far employed reveals a number of

unsubstantiated modifications to the extraction procedureand a lack of standardization for

reportingFAP data, thus negating the directcomparabilityof results between studies.

Using three marine invertebrate species, Nicol (1987) was the first to test specific

aspects (i.e., sonication and use of tissue preservatives) of Ettershank's (1983, 1984 a.b)

protocol. Of greatest significancewas his finding that tissue preservatives (formalin and

ethanol) caused in vitro elevationof samplefluorescence. These results were

corroborated by Mullin and Brooks (1988), who induced similareffects in formalin­

preserved fish larvae. Both of these fmdings raised questions not only as to the source of

fluorescence (native FAP vs formalin induced)measured in Ettershank's (1983, 1984

a.b, 1985) and Berman et al.' s (1989) krill samples, but also in Hirche and Anger's

(1987) spider crab larvae, and Hammer's (1988) pike larvae. Further, additional

questions with regard to the stabilityof extractableFAP compounds under different

storage and handling conditions remainedunanswered, and in essence have largely been

ignored.

The purpose of the present study was two-fold. The first objective was to

evaluate various aspects (i.e., tissue storageand extract incubation temperature and time,

sonicationof tissue homogenates,assay temperature and the use of aqueous-methanol

extracts)of the methodologies currentlyadoptedfor the spectrofluorometric quantification

ofFAP and to determine their effects on the expression of and variation in sample

fluorescence. For this purpose we chose Oreochromis mossambicusas the model

31

animal, concentrating on postmitotic tissues only. The second objective involved the

development of modified handling, extraction, and assay procedures to negate in-sample

fluorescence variation and the application of these procedures to brain tissues of O.

mossambicus to determine whether a distinct relationship between FAP content and

chronological age exists.

Materials and methods

Model animal/tissues

Known-age specimens of Oreochromis mossambicus, stocked as fry, were reared

for up to 886 d at the Kewalo Basin Research Facility of the National Marine Fisheries

Service, Honolulu Laboratory. Fish were held in circular fiberglass holding tanks at

ambient temperature with daily water changes, and fed ad libitum rations of trout chow

twice daily. Subsampled specimens were transported live to the laboratory, sacrificed by

ice bath immersion, weighed to the nearest O.OIg, and standard and fork lengths

measured to the nearest 0.05 mm. Fish were bled from the caudal region before

removing tissues. A preliminary survey of numerous postmitotic tissues indicated that

brain, ventricle, and skeletal muscle extracts contained detectable levels ofFAP

fluorescence with the least amount of interfering compounds; therefore, these were

chosen as model tissues for the remainder of the study. Dissected tissues were placed in

Eppendorf vials and immediately frozen in liquid nitrogen until analysis.

Tissue storage temperature

Skeletal muscle tissues were placed at various short- and long-term storage

temperatures to detect changes in sample fluorescence over time. White skeletal muscle

was dissected from the left anterior epaxial region of three sibling fish, briefly minced

with a Virtis motorized blade, and stored in Eppendorf vials either at room (23°C) or

refrigerated (3°C) temperatures for 0.5, 1,2,4,8, and 12 h, or frozen at -20 or -80°C for

32

5, 15, 30, 60, 90, and 120 d. Sample vials were frozen in liquid nitrogen at the end of

each time point. Sample fluorescence was compared to that of absolutely fresh tissue.

Tissues were analyzed using the modified extraction procedure, and assayed as described

in later sections.

Unmodified extraction procedures

Extraction and assay procedures adopted for the first part of this study were based

on protocols outlined by Ettershank (1983, 1984 a.b). These included: maceration of

tissues in spectroscopic grade chloroform/methanol (CHCI3!MeOH, 2:1 v/v), sonication,

mixing with a magnesium chloride solution [100 roM in deionized glass distilled water

(GDW)], and centrifugation for 20 min at RCF =3000 x g. Fluorescence of the lower

CHCl3 layer was compared to a standard solution of quinine sulfate (0.1 mg/L IN

sulfuric acid). Further details on extraction procedures are provided, when appropriate,

in the following subsections.

Incubation temperature of sample extracts

The stability ofFAP products in solvent was examined in intermediate-stage

CHCI3!MeOH (2:1 v/v) and final CHCl3 fractions of tissue extracts at room temperature

(23°C) and -20°C for up to 8 h. Lyophilized brain, heart, and muscle tissues from

siblings were pooled by pulverizing and mixing dried tissues in 15 ml glass centrifuge

tubes with a teflon pestle. Stability of intermediate stage extracts was tested by extracting

triplicate aliquots of 20 mg of dry tissue in 2 ml CHCI3/MeOH, centrifuging the

homogenate for 20 min at a relative centrifugal force (RCF) of 3000 x g and 0 to 5°C, and

collecting and storing the CHCI3/MeOH supernatant in amber glass vials at the defined

incubation temperatures. Final stage CHC13 extracts were prepared by adding an equal

volume of 100 roM magnesium chloride solution in GDW to the CHC13!MeOH

homogenates, centrifuging as defined above, and transferring the CHCl3 hypophase to

amber glass vials for incubation.

33

----------- -

Effects of ultrasonication on fluorescence yield

Ultrasonication of tissue homogenates (widely adopted for the disruption of

biological membranes and thus increased exttaetion efficiency) was tested to determine

the overall effects on fluorescence yield ofCHCl3-extractable FAP compounds. Eighty

mg aliquots of lyophilized and pooled brain, heart, or muscle tissues were glass

homogenized in 8 ml CHCI3!MeOH (2:1), transferred to a 15 ml glass centrifuge tube,

and sonicated with a Sonics & Materials VibraCeli model VC300 microtip probe sonicator

for up to 35 min. Sonication output was 25 W at a 50% duty cycle with sample tubes

held in ice and water to absorb heat. 1 ml aliquots of CHCI3!MeOH homogenate were

removed at 5 min intervals, vortex mixed with an equal volume of the 100 mM MgCI

solution, centrifuged, and the CHCl3 layer assayed for FAP content.

Confirmation and removal of interfering compounds

Fletcher et al. (1973) mentioned the importance of removing interfering

fluorescent compounds from CHCI3!MeOH tissue extracts. Retinol fluorescence

(excitation 325 to 340 nm, emission 475 nm), present in the final CHCl3 extract, could

be removed by a brief (30 s) exposure of the solvent to high intensity ultraviolet light.

Flavin compounds fluoresce at 510 to 520 nm with excitation maxima at 280,350, and

450 nm and are present in the CHCI3/MeOH homogenate, removable by a water wash

step. In the present study, complete excitation and emission fluorescence spectra were

obtained before all sample assays to detect the presence of fluorescent contaminants. In

addition, the fluorescent properties of all aqueous/methanol phase extracts were checked

for the presence of these contaminants.

Assay temperature

Temperature dependence of sample and standard fluorescence has been

established (Guilbault 1973), however, most researchers have conducted assays at

ambient temperature. The importance of assay temperature control was tested on

34

Oreochromis mossambicus brain, heart, and muscle extracts by measuring sample

luminescence at temperatures ranging from 0 to 42°C at 3°C intervals. Temperature of the

sample solvent was maintained using a thermostatically controlled turret microcell holder

connected to a Brinkmann Lauda Model RM6 refrigerated circulating water bath.

Resulting changes in the calculated relative fluorescence intensity values were noted.

Modified extraction and assay procedures

Basefl on the results from the above experimental analyses a modified procedure

for the extraction and assay of FAP was implemented. Tissues were prepared for

extraction by immersion in liquid nitrogen and lyophilization to obtain the most accurate

sample weight. Dry tissues were weighed on a Sartorius Model 4503 MP6 electronic

microbalance to the nearest microgram, placed in glass tubes, and returned to the

lyophilizer until extraction. Tissues were macerated at room temperature for 4 min in a

Potter-Elvehjem all-glass tissue homogenizer using at least 1 ml of CHCl3 per 10 mg dry

tissue. Approximately one-third of the solvent typically evaporated during

homogenization. Because of this the initial meniscus was marked and the homogenate

adjusted back to its original volume after treatment. Homogenates were then transferred

to a glass centrifuge tube and vortex mixed for 2 min with an equal volume of a solution

containing 100 roM MgCI in GDW and methanol at a ratio of 3:1. Samples were

centrifuged at D-5°C for 20 min at RCF =3000x g. The lower CHCl3 layer was

immediately removed and assayed for FAP content.

Chloroform extracts were placed in matched quartz silicate microcuvettes with a 4

mm pathlength, sealed with a teflon stopper, and assayed at 18°C in a Perkin Elmer model

LS-5 fluorescence spectrophotometer equipped with a Xenon light source and standard S­

5 photomultiplier tube. Excitation and emission slits were both set at 10 nm for standard

assays, and luminescence readings were at a fixed scale setting of 1.000. Excitation and

emission maxima were determined using PreScan mode, and the maximum luminescence

35

at these settings was recorded. Fluorescence spectra reported herein were not corrected

for the spectral emission characteristics of the light source, nor the spectral efficiencies of

the monochromators, optics, and the photomultiplier tube. Quinine sulfate at a

concentration of 0.1 mg/11H sulfuric acid (H2S04) was used as a fluorescence standard,

and the sample readings were expressed as a percentage of the standard luminescence at

an excitation setting of 345 om and emission setting of 450 om. Sample FAP

concentrations are expressed either as relative fluorescence intensity (RFI), where:

sample luminescence solvent volume (ml)RFI = X X 100

standard luminescence sample dry weight (mg)

or as whole organ %fluorescence (%FL), where:

sample luminescenceWhole Organ %FL = d d I' X solvent volume (ml) X 100

stan ar ummescence

Sample luminescence spectral scans were plotted for all samples with a Perkin

Elmer model R100A chart recorder. Spectral scans presented herein were produced by

recording luminescence at 1 nm intervals with the complementary excitation/emission

wavelengths at a fixed setting.

Results

Tissue storage temperature

Chloroform extracts from fresh (t=O) pooled Oreochromis mossambicus muscle

had an average RFI value of 0.4412 (±O.0275) at excitation maxima of 350 to 355 nm

and emission maxima of 425 to 430 nm. Muscle samples stored at elevated (>O°C)

temperatures exhibited linear increases in fluorescence after a matter of hours with no

apparent shifts in excitation/emission maxima Muscle samples held at room temperature

(23°C) had elevated levels of fluorescence after only I h. A mean increase (.1RFI) of

9.3% was evident after only 2 h of incubation, and this increased to as much as 87%

36

-----------

~RFI at the end of 12 h (Fig. 2-1 A) (%dRFI = -1.754 + 6.303 (Hours), r2 = 0.96).

Although refrigeration at 3°C slowed the rate of fluorophore accumulation, a mean

increase of 9.1% ~RFI was measured after 8 h incubation (%~RFI=0.093 + 0.722

(Hours), r2 =0.56). Tissues stored for longer periods at -20°C also exhibited linear

increases in fluorescence, with a mean ~RFI of 9.7% after 15 d (Fig. 2-1 B) (%dRFI =

-0.087 + 0.638 (Days), r2 =0.98). No significant changes in fluorescence were

observed in muscle samples stored at -80°C after 120 d (%8RFI =0.621 + 0.0029

(Days), r2 = 0.003).

Incubation temperature of sample extracts

Fluorescence of CHCI3!MeOH- and CHCl3-soluble extracts of Oreochrornis

mossambicus brain, heart, and muscle increased over time at both room (23°C) and

reduced (-20°C) temperatures (Figs. 2-2 and 2-3). Increases in CHCI3!MeOH extract

fluorescence were accompanied by shifts in the excitation/emission maxima from

approximately 355/430 to 380/455 nm, while fluorescence maxima of CHCl3 extracts

remained unchanged. The change in CHCI3!MeOH extract fluorescence wavelengths

occurred by 3 h at 23°C and by 4 h at -20°C. Chloroform/methanol extracts incubated at

23°C exhibited the most rapid rises in fluorescence, with heart extracts increasing by an

average of 147% ~RFI after 3 h and as much as 880% ~RFI at the end of 8 h (Fig. 2-2

A). Final water-washed CHCl3 extracts of all three tissues increased in fluorescence on

the average of 10 to 16% 8RFI after only 2 h incubation at 23°C with heart extracts again

increasing more than brain or skeletal muscle (Fig. 2-3 A). Incubation at lowered

temperature (-20°C) slowed in vitro changes in sample fluorescence. Even so, increases

were observed within 2 to 4 h of incubation (Figs. 2-2 Band 2-3 B). Brain CHCl3

extracts were the least stable at -20°C, with an average dRFI of7.2% after 4 h and up to

18.5% after 8 h.

37

Effects of ultrasonication

Sonication of CHCI3!MeOH homogenates at 25 W of output caused dramatic

increases in fluorescence for all three tissues, with mean LlRFI values of 200, 119, and

83% for brain, heart, and muscle, respectively, after only 5 min sonication (Fig. 2-4 A).

Continued sonication for up to 30 min caused further increases in fluorescence, after

which levels decreased or stabilized. Of the tissues studied, brain homogenates had the

greatest change with an average LlRFI of 902% after 30 min (Fig. 2-4 A). Following a

sonication period of 5 min, the uncorrected fluorescence spectra of all tissue homogenates

also revealed shifts in excitation/emission maxima similar to those described for

CHCI3!MeOH extracts in the previous subsection. Results are presented for brain tissue

only (Fig. 2-4 B).

Fluorescent contaminants

Uncorrected fluorescence spectra from some brain, heart, and skeletal muscle

extracts showed the presence of a second emission peak in the 510 to 520 nm range in the

CHCI3/MeOH of the intermediate extraction stage as well as in the aqueous phase final

wash. Excitation scans of these samples at an emission setting of 520 nm revealed

excitation maxima at 286,351, and 447 nm (Fig. 2-5 B), confirming the presence of

fluorescent flavin contaminants in the CHCI3!MeOH and aqueous phase solvents. The

second 510 to 520 nm emission peak was not obvious in all CHCI3!MeOH and aqueous

extracts (Fig. 2-5 A). However, excitation scans of all samples at emission =520 nm

confirmed the presence of the three excitation maxima (Fig. 2-5 B). Flavin contaminants

were not present at detectable levels in the final CHCl3 sample solvent, and a second

water wash of the CHCl3 layer contained no flavin or FAP fluorescence. Thus, a single

water wash was sufficient to remove all flavin contaminants. There was no evidence of

retinol contaminants in any of the three chosen tissues.

38

Assay temperature

Absolute luminescence values of both standards and samples decreased with

increasing assay temperature (Fig. 2-6 A). The relationship between assay temperature

(T in "C) and brain extract luminescence (BL, as raw luminescence values) was linear,

where:

BL =0.61109 - 0.01021 (T), r2=0.969

and the relationship between assay temperature (T) and quinine sulfate luminescence (QL)

was best described as:

QL =1.6532 - 0.00467 (T), r2=0.886

The effect of temperature on fluorescence data is better observed when RFI values are

considered (Fig. 2-6 B). A three-fold decrease in RFI occurred over the temperature

range of 0 to 42°C, and inflections were apparent in this relationship (Fig. 2-6 B) within

the working range of room temperatures (20 to 24°C).

FAP accumulation with age

The extraction and assay of whole brain FAP content (% FL) using our modified

procedures resulted in a positive correlation with chronological age (Fig. 2-7 A). This

increase was evident both in immature fish and in mature male and female specimens, and

for the portion of the lifespan examined, it was well described by second order

polynomial regressions (Fig. 2-7 A) (Immature + males: %FL =7.407 - 4.586*10-2

(Days) + 2.33*10-4 (Days)2, r2 =0.926, P< 0.0001, n =101; Immature + females: %FL

=6.347 - 1.41*1O-2(Days) + 1.074*10-4 (Days)2, r2 =0.936, P< 0.0001, n =99). FAP

content (% FL) accumulated gradually during the first 370 d, after which it increased

rapidly and resulted in a marked difference between males and females.

Relative fluorescence intensity values decreased rapidly between 32 and 92 d,

remained low until approximately 330 d, and gradually increased thereafter, also

differentiated by sex (Fig. 2-7 B). After about the first 80 d, relationships between age

39

and RFI were well described using second order polynomial equations (Fig. 2-7 B)

(Immature + males: RFI = -0.366 + 2.033*10-3 (Days) + 3.843*10-6 (Days)2, r2 =

0.921, P< 0.0001, n = 101; Immature + females: RFI = 0.866 - 2.678*10-3 (Days) +

5.872*10-6 (Days)2, r2 =0.935, P< 0.0001, n =99).

Brain growth with age was most rapid initially, slowing after 300 d, and mirrored

the relationship between age and RFI. Relationships between age and brain dry weight

(BW in mg) were well described by logarithmic equations (Fig. 2-7 C) (Immature +

males: BW =-36.488 + 23.333 * loglQ(Days), r2 =0.943, P< 0.0001, n = 123;

Immature + females: BW =-32.101 + 21.067 * 10glQ(Days), r2 =0.914, P< 0.0001, n =121).

Discussion

It is apparent from the results of this study that the temperatures at which tissues

and extracts are handled and the time frame involved markedly affect expression of FAP

fluorescence and can introduce considerable error unless controlled. Similarly,

application of ultrasonication to tissue homogenates drastically alters the properties and

levels of sample fluorescence. These observed changes in tissue and extract fluorescence

suggest the in vitro formation ofFAP-like substances caused by lipid peroxidation (Luo

and Hultin 1986), the production of cross-linking agents (e.g., malondialdehyde and

formaldehyde), and the subsequent genesis of fluorophores in the sample solvent.

The effects of storage, handling, and assay temperatures on the accurate analysis

ofFAP are not surprising. The chemical instability of fish tissues under different

freezing conditions (as related to food quality) has been well documented. Changes

occurring during freezing include "browning" reactions between free amino acids and

sugars (Jones 1959), denaturation of proteins (Sawant and Magar 1961), dissociation of

trimethylamine to dirnethylamine and formaldehyde (Amano and Yamada 1964,

40

Tokunaga 1964, Castell et al. 1971), loss of membrane phospholipid through

autoxidation (Hardy et al. 1979), and increases in muscle fluorescence (Davis 1982).

Davis (1982) used fluorometric methods to document effects of freezing (-5 to -30°C) on

muscle tissues. In a companion paper, increases were attributed to formaldehyde

production and browning reactions between fructose, ribose, and free amino groups

(Davis and Reece 1982). The results reported for skeletal muscle of Oreochromis

mossambicus in the present work indicate similar changes affecting levels of apparent

FAP content, and demonstrate the necessity for proper freezing temperatures from the

moment the animal is sacrificed (i.e., immediate placement and storage in liquid nitrogen,

or at least at temperatures of -80aC or lower).

Because temperature would seem to be a critical factor affecting the native FAP

content of tissues, the increases observed with time in this research suggest that they are

probably due to the in vitro formation of other chemical complexes. We suspect that

these complexes are similar to those described in studies seeking to determine the

chemical nature of FAP fluorophores (Chio and Tappel 1969). Their formation is

presumably initiated by lipid peroxidation, which has been well defmed for muscle

microsome and sarcoplasmic reticulum (SR) model systems of fish. Lipid peroxidation,

measured by malondialdehyde (MDA) production, has been demonstrated for muscle

microsomal fractions from red hake (McDonald et al. 1979) and herring (Slabyj and

Hultin 1982). MDA production was dependent upon the levels of peroxidizing cofactors

(NADH, ATP, and Fe3+) and incubation temperature, and occurred under all conditions

over short periods of time (0 to 60 min) in both studies (McDonald et al. 1979, Slabyj

and Hultin 1982). Related studies using SR model systems have found similar cofactor

effects (Luo and Hultin 1986) as well as inhibition of peroxidation by phospholipase A2

(Shewfelt and Hultin 1983). Luo and Hultin (1986) incubated SR from winter flounder,

documenting increases in both MDA and fluorescence after only 1 and 3 h, respectively.

41

Once formation of MDA has occurred, a cascade of chemical reactions has been proposed

that eventually results in the genesis of blue and blue-green emitting fluorophores.

Chio and Tappel (1969) first described the cross-linking reaction between MDA

and amino acids to form N,N'-disubstituted l-amino-3-iminopropenes which fluoresce at

wavelengths between 400and 550 nm. Similar cross-linking reactions have now been

established between MDA and the primary amine groups of phospholipids (Dillard and

TappeI1971, Bidlack and Tappe11973, Trombly and TappeI1975), proteins (Shimasaki

et al. 1982, Fukuzawa et al. 1985, Kikugawa and Beppu 1987), and nucleic acids (Nair

et al. 1986). Lipid-soluble extracts used for FAP studies represent complex mixtures of

phospholipids, fatty acids, non-polar amino acids, proteins, and many other biological

compounds which are highly susceptible to peroxidation and polymerization reactions.

Our results support the conclusion that temperature and time are important in the

accelerated formation of fluorescent complexes. To eliminate further synthesis of

fluorescent adducts, tissue extracts should be handled expediently within a critical

window of time (from beginning of extraction through fluorescence assay) and assayed at

controlled temperature.

The fluorescence maxima shifts observed in CHCI3/MeOH extracts (whether or

not ultrasonication was applied to the homogenate) suggest the formation of compounds

similar to 1,4-dihydropyridine-3,5-dicarbaldehydes first isolated by Kikugawa and Ido

(1984) through the reaction of primary amines with the hydrolysate of

tetramethoxypropane; however, this process has yet to be substantiated. Regardless of

the source of change, CHCI3/MeOH tissue homogenates are highly unstable under

normal extraction conditions, and ultrasonication of homogenates greatly enhanced the

rates of fluorophore production. This may be due either to generation of free radicals in

sample solution through cavitation or to the localized generation of heat within

42

homogenates. Our results indicate that ultrasonication should be avoided because of the

rapid alteration of sample fluorescence properties.

Overall, our data raise questions as to the nature of fluorescence described in

previous investigations on FAP in aquatic animals (Table 1). A survey of the methods

adopted in these studies emphasizes the fact that few used fresh tissues or stored tissues

at temperatures adequate for stabilization of cellular constituents, most did not mention the

time elapsed between freezing and analysis, and none took extract incubation time and

temperature into full consideration. Additionally, ultrasonication was applied to tissue

homogenates in at least half of these studies (Table 1), and whereas the majority

employed sonication for less than 5 min, most applied a power output four times greater

than that found to generate additional fluorophores under our experimental procedures.

Conversely, Nicol (1987) found no significant differences between sonified and non­

sonified Meganyctiphanes norvegica individuals. This was probably because the samples

were preserved in formalin and had already increased in fluorescence prior to extraction.

Moreover, several of the studies cited in Table 1 (i.e., Hammer and Madhusudhana-Rao

1987, Hirche and Anger 1987, Hammer 1988, Hammer and Karakiri 1988) used the

polar extract phase to quantify FAP but failed to verify the absence of flavin contaminants

- which we found to comprise as much as half of the total sample fluorescence. Sheehy

and Ettershank (1989) also reported considerable quantities of flavin-like fluorescence in

CHCI3/MeOH and polar phase extracts from whole Daphnia carlnata. Ettershank (1983,

1984 a.b, 1985) was the only investigator to control temperature during fluorometric

analysis; the others conducted assays at room temperature.

When all these critical aspects ofFAP analysis are considered, it is not surprising

that the data from previous investigations have been too inconsistent for direct

comparison. More importantly, the main goal of most previously published research has

been to establish a relationship between FAP content and chronological age; however,

43

becauseof the highdegreeof variability and thefact that the techniques wereapplied to

animalsof known age in only a few studies (Table1), the relationships reportedare open

to question. In all probability, theyreflect fluorescence modifiedby analytical artifacts

rather than native tissue fluorescence. Of those studieson fishes that have taken

chronological age into account, the most complete is probably that of Vernet et al. (1988),

who measuredFAP content in adultOncorhynchus mykiss of known age (3 mo to 3 yr).

However, they sampledonly 21 individuals derivedfrom different broodstockanddid

not take into accountextractincubation or assaytemperature and time.

In contrast to theseprevious investigations, our study not only took into account

sourcesof variabilityin most aspectsof the extraction and assay procedure,but also

provided a chronological range of samples (from33 to 886 d), with more stringent

samplingprocedures. The presentresearchis the first to describe a clear relationship

betweenblue-emitting FAP accumulation and chronological age in an aquaticorganism.

However,other sourcesof variability have been subsequently exposed that require

further research, in order that the technique can be validated for accurate assessment of

age. For example, initialdecreasesin the weight-specific FAP concentrations in juvenile

Oreochromis mossambicus appear to correspondto a period of rapid brain growth,

suggestingthat althoughFAP accumulates in this tissue, growth of the whole organ

exceeds the rate of accumulation at this developmental stage. Siakotoset al. (1977) found

similarpatternsof accumulation of humanbrain lipofuscin which wasextractedand

purified by differential centrifugation, and speculated that the largepool of lipofuscin

found in youngindividuals may be related to the lysosomesderivedfrom the remodeling

of cellular structures duringearlydevelopmental stages. The markeddifferences in FAP

accumulation between male and female Q. mossambicus have yet to be explained, but will

have direct bearingon the applicability ofFAP analysis for accurateage assessment of

wild stocks.

44

An additional problem must be overcome before spectrofluorometric assays for

FAP can be applied successfully to crustaceans. Many crustacean species contain

considerable quantities of carotenoid compounds such as astaxanthin and canthaxanthin

(Thommen and WackernageI1964, Herring 1968 a.b, Goodwin 1984), that have light

absorptive properties within the emission range ofFAP compounds (Zeller et al. 1959,

Cooper et al. 1975) and are highly soluble in nonpolar solvents such as chloroform.

Thus, these carotenoid compounds may cause internal quenching of FAP fluorescence.

Preliminary investigations of FAP quantification in Penaeus vannamei (Hill, Moss, and

Womersley unpublished data) and Heterocarpus laevigatus (Hill, Shiota, and Womersley

unpublished data) have revealed large quantities of orange colored pigments in nonpolar

extract fractions which have UV/visual absorption maxima similar to those reported for

astaxanthin and canthaxanthin, Increases in orange pigment concentrations were

accompanied by decreases in FAP concentrations. If the observations indicate quenching

phenomena, they may explain the negative correlations between age and weight-specific

FAP fluorescence described by Hirche and Anger (1987) for Hyas araneus and by

Sheehy and Ettershank (1989) for Daphnia Carinata.

Clearly, the fluorometric assay for FAP holds a great deal of promise. However,

further investigation is necessary to determine the source in variations in FAP content

between sexes and the effects of environmental factors (e.g., temperature, light, food

ration) on tissue fluorescence. These and other problems must be resolved before

measurement ofFAP can become a standard method for aging aquatic animals.

45

Literature cited

Amano, K, Yamada, K (1964). A biological formation offonnaldehyde in the muscle

tissue of gadoid fish. Bull. Jap. Soc. scient. Fish. 30: 430-435

Barden, H. (1980). Interference filter microfluorimetry of neuromelanin and lipofuscin in

human brain. J. Neuropathol. expo Neurol. 39: 598-605

Berman, M. S., McVey, A. L., Ettershank, G. (1989). Age determination of Antarctic

krill using fluorescence and image analysis of size. Polar BioI. 9: 267-271

Bidlack, W. R., Tappel, A. L. (1973). Peroxidation of subcellular organelles: formation

of lipofuscin-like pigments. Lipids 8: 203-207

Bjorkerud, S. (1964). Isolated lipofuscin granules - survey of a new field. Adv. geront.

Res. 1: 257-288

Bridges, R. G., Sohal, R. S. (1980). Relationship between age-associated fluorescence

and linoleic acid in the housefly Musca dornestica. Insect Biochem. 10: 557-562

Castell, C. H., Smith, B., Neal, W. (1971). Production of dimethylamine in muscle of

several species of gadoid fish during frozen storage, especially in relation to

presence of dark muscle. J. Fish. Res. Bd. Can. 28: 1-5

Chio, K S., Tappel, A. L. (1969). Synthesis and characterization of the fluorescent

products derived from malondialdehyde and amino acids. Biochemistry (Am.

chern. Soc.) Easton, Pa. 8: 2821-2827

Cooper, R. D. G., Davis, J. B., Leftwick, A. P., Price, C., Weedon, B. C. L. (1975).

Carotenoids and related compounds. Part XXXII. Synthesis of astaxanthin,

phoenicoxanthin, hydroxyechinenone, and the corresponding diosphenols. 1.

chern. Soc. Perkin Trans. (I: Org. bio-org. Chern.) 21: 2195-2204

Csallany, A. S., Ayaz, K L., Menken, B. Z. (1984). Effects of dietary vitamin E upon

fluorescent compounds of the rat uterus. Lipids 19: 911-915

46

Davies, K. J. A. (1988). Protein oxidation, protein cross-linking, and proteolysis in the

formation of lipofuscin. In: Zs.-Nagy, I. (ed.) Lipofuscin - 1987: state of the art.

Akademiai Kiad6, Budapest and Elsevier Science Publishers, Amsterdam, p. 24­

25

Davis, H. K. (1982). Fluorescence of fish muscle: description and measurement of

changes occurring during frozen storage. J. Sci. Fd Agric. 33: 1135-1142

Davis, H. K., Reece, P. (1982). Fluorescence of fish muscle: causes of change occurring

during frozen storage. 1. Sci. Fd Agric. 33: 1143-1151

Dillard, C. J., Tappel, A. L. (1971). Fluorescent products of lipid peroxidation of

mitochondria and microsomes. Lipids 6: 715-721

Eldred, G. E. (1987). Questioning the nature of fluorophores in age pigments. In: Aloj

Totaro, E., Glees, P., Pisanti, F. A. (eds.) Advances in age pigments research.

Pergamon Press, Oxford, p. 23-36

Eldred, G. E., Katz, M. L. (1988). Fluorophores of the human retinal pigment

epithelium: separation and spectral characterization. ExpI. Eye Res. 47: 71-86

Eldred, G. E., Katz, M. L. (1989). The autofluorescent products of lipid peroxidation

may not be lipofuscin-like. Free radical BioI. Med. 7: 157-163

Eldred, G. E., Miller, G. V., Stark, W. S., Feeney-Burns, L. (1982). Lipofuscin:

resolution of discrepant fluorescence data. Science, N.Y. 216: 757-759

Ettershank, G. (1983). Age structure and cyclical annual size change in the Antarctic krill,

Euphausia superba. Polar BioI. 2: 189-193

Ettershank, G. (1984a). A new approach to the assessment of longevity in the Antarctic

krill Euphausia superba. 1. Crustacean BioI. 4 (Spec. No.1): 295-305

Ettershank, G. (1984b). Methodology for age determination of Antarctic krill using the

age pigment lipofuscin. BIOMASS Handb. Ser. 26: 1-14

47

Ettershank, G. (1985). Population age structure in males and juveniles of the Antarctic

krill, Euphausia superba Dana. Polar BioI. 4: 199-201

Ettershank, G., Macdonnell, 1., Croft, R. (1983). The accumulation of age pigment by

the fleshfly Sarcophaga bullata Parker (Diptera: Sarcophagidae). Aust. J. Zool.

31: 131-138

Fletcher, B. L., Dillard, C. J., Tappel, A. L. (1973). Measurement of fluorescent lipid

peroxidation products in biological systems and tissues. Analyt. Biochem. 52: 1-9

Fukuzawa, K., Kishikawa, K., Tokumura, A., Tsukatani, H., Shibuya, M. (1985).

Fluorescent pigments by covalent binding of lipid peroxidation by-products to

protein and amino acids. Lipids 20: 854-861

Goodwin, T. W. (1984). The biochemistry of the carotenoids, Vol. 2, Animals.

Chapman and Hall, London

Guilbault, G. G. (1973). Practical fluorescence: theory, methods and techniques. Marcel

Dekker, New York

Gutteridge, J. M. C., Heys, A. D., Lunec, 1. (1977). Fluorescent malondialdehyde

polymers from hydrolysed 1,1,3,3-tetramethoxypropane. Analytica chim. Acta

94: 209-211

Hammer, C. (1988). Accumulation of fluorescent age pigments in embryos and larvae of

pike, Esox lucius (Esocidae, Teleostei) at different incubation temperatures.

Envir. BioI. Fish. 22: 91-99

Hammer, C., Braum, E. (1988). Quantification of age pigments (lipofuscin). Compo

Biochem. Physiol. 90B: 7-17

Hammer, C., Karakiri, M. (1988). Disease dependent accumulation of fluorescent age

pigments in the common dab CLimandalimanda, Teleostei). In: Zs.-Nagy, I. (ed.)

Lipofuscin - 1987: state of the art. Akademiai Kiad6, Budapest and Elsevier

Science Publishers, Amsterdam, p. 445-446.

48

_..__.._-_._-_ .. -------_._•..- -.

Hammer, C., Madhusudhana-Rao, G. (1987). Individual and feed dependent

accumulation of fluorescent age pigments in five organs of tilapia (Oreochromis

aurea). Zeitschr. angew. Zool, 74: 275-292

Hamperl, H. (1934). Die Fluoreszenzmikroskopie menschlicher Gewebe. Virchows

Arch. path. Anat. Physiol, 292: 1-51

Hannover, A. (1842). Mikroskopische Undersolgeser af Nervensystemet. K. danske

Vidensk. Selsk. Skr. (Nat. Math. Afh.) 10: 1-112

Hardy, R., McGill, A. S., Gunstone, F. D. (1979). Lipid and autoxidative changes in

cold stored cod (Gadus morhua). 1. Sci. Fd Agric. 30: 999-1006

Herring, P. J. (1968a). The carotenoid pigments of Daphnia magma Strauss - I. The

pigments of animals fed Chiorella pyrenoidosa and pure carotenoids. Compo

Biochem. Physiol. 24: 187-203

Herring, P. J. (1968b). The carotenoid pigments of Daphnia magna Strauss - II. Aspects

of pigmentary metabolism. Compo Biochem. Physiol. 24: 205-221

Hill, K. T., Radtke, R. L. (1988). Gerontological studies of the damselfish, Dascyllus

albisella. Bull. mar. Sci. 42: 424-434

Hirche, H.-J., Anger, K. (1987). The accumulation of age pigments during larval

development of the spider crab, Hyas araneus (Decapoda, Majidae). Compo

Biochem. Physiol, 88B: 777-782

Hueck, W. (1912). Pigmentstudien. Beitr. path. Anat. 54: 68-232

Hyden, H., Lindstrom, B. (1950). Microspectrographic studies on the yellow pigment in

nerve cells. Discuss. Faraday Soc. 9: 436-441

Jones, N. R. (1959). "Browning" reactions and the loss of free amino acid and sugar

from lyophilized muscle extractives of fresh and chill-stored codling (Gadus

callarias). Fd Res. 24: 704-710

49

Katz, M. L. (1990). Incomplete proteolysis may contribute to lipofuscin accumulation in

the retinal pigment epithelium. In: Porta, E. A. (ed.) Lipofuscin and ceroid

pigments. Plenum Press, New York, p. 109-118

Katz, M. L., Robison, W. G., Herrmann, R. K., Groome, A. B., Bieri, J. G. (1984).

Lipofuscin accumulation resulting from senescence and vitamin E deficiency:

spectral properties and tissue distribution. Mechanisms Ageing Dev. 25: 149-159

Kikugawa, K., Beppu, M. (1987). Involvement of lipid oxidation products in the

formation of fluorescent and cross linked proteins. Cherny Phys. Lipids 44: 277­

296

Kikugawa, K., Ido, Y. (1984). Studies on peroxidized lipids. V. Formation and

characterization of 1,4-dihydropyridine-3,5-dicarbaldehydes as model of

fluorescent components in lipofuscin. Lipids 19: 600-608

Kitani, K., Nokubo, M., Ohta, M., Zs.-Nagy, I. (1988). Characterization of peroxide­

induced autofluorescence of hepatocyte plasma membrane proteins in relation to

age pigments in rats. In: Zs.-Nagy, I. (ed.) Lipofuscin - 1987: state of the art.

Akademiai Kiad6, Budapest and Elsevier Science Publishers, Amsterdam, p.

159-182

Lillie, R. D., Daft, F. S., Sebrell, W. H. (1941). Cirrhosis of the liver in rats on a

deficient diet and the effect of alcohol. Pub. Hlth Rep., Wash. 56: 1255-1258

Luo, S.-W., Hultin, H. O. (1986). Effect of age of winter flounder on some properties

of the sarcoplasmic reticulum. Mechanisms Ageing Dev. 35: 275-289

McArthur, M. C., Sohal, R. S. (1982). Relationship between metabolic rate, aging, lipid

peroxidation, and fluorescent age pigment in milkweed bug, Oncopeltus fasciatus

(Hemiptera). J. Geront. 37: 268-274

McDonald, R. E., Kelleher, S. D., Hultin, H. O. (1979). Membrane lipid oxidation in a

mucrosomal (sic) fraction of red hake muscle. J. Fd Biochem. 3: 125-134

50

Miquel, J., Lundgren, P. R, Johnson, J. E. (1978). Spectrophotofluorometric and

electron microscopic study of lipofuscin accumulation in the testis of aging mice.

J. Geront. 33: 5-19

Mullin, M. M., Brooks, E. R. (1988). Extractable lipofuscin in larval marine fish. Fish.

Bull. U.S. 86: 407-415

Nair, V., Cooper, C. S., Vietti, D. E., Turner, G. A. (1986). The chemistry of lipid

peroxidation metabolites: cross-linking reactions of malondialdehyde. Lipids 21:

6-10

Nicol, S. (1987). Some limitations on the use of the lipofuscin ageing technique. Mar.

BioI. 93: 609-614

Nokubo, M., Ohta, M., Kitani, K., Zs.-Nagy, I. (1989). Identification of protein-bound

riboflavin in rat hepatocyte plasma membrane as a source of autofluorescence.

Biochim. biophys. Acta 981: 303-308

Nokubo, M., Zs.-Nagy, I., Kitani, K., Ohta, M. (1988). Characterization of the

autofluorescence of rat liver plasma membranes. Biochim. biophys. Acta 939:

441-448

Porta, E. A., Mower, H. F., Moroye, M., Lee, Ch., Palumbo, N. E. (1988).

Differential features between lipofuscin (age pigment) and various experimentally

produced "ceroid pigments". In: Zs.-Nagy, I. (ed.) Lipofuscin - 1987: state of the

art. Akademiai Kiado, Budapest and Elsevier Science Publishers, Amsterdam, p.

341-374

Sawant, P. L., Magar, N. G. (1961). Studies on frozen fish. II. Some chemical changes

occurring during frozen storage. Fd Technol., Lond. 15: 347-350

Sheehy, M. R. J., Ettershank, G. (1989). Solvent extractable age pigment-like

autofluorescence and its relationship to growth and age in the water flea, Daphnia

carlnata. Aust. J. Zoo!. 36: 611-625

51

Shewfelt, R. L., Hultin, H. O. (1983). Inhibition of enzymic and non-enzymic lipid

peroxidation of flounder muscle sarcoplasmic reticulum by pretreatment with

phospholipase A2. Biochim. biophys. Acta 751: 432-438

Shimasaki, H., Veta, N., Privett, O. S. (1980). Isolation and analysis of age-related

fluorescent substances in rat testes. Lipids 15: 236-241

Shimasaki, H., Veta, N., Privett, O. S. (1982). Covalent binding of peroxidized linoleic

acid to protein and amino acids as models for lipofuscin formation. Lipids 17:

878-883

Siakotos, A N., Armstrong, D., Koppang, N., Muller, J. (1977). Biochemical

significance of age pigment in neurones. In: Nandy, K., Sherwin, I. (eds.) The

aging brain and senile dementia. Plenum Press, p. 99-118

Slabyj, B. M., Hultin, H. O. (1982). Lipid peroxidation by microsomal fractions isolated

from light and dark muscles of herring (Clupea harengus). J. Fd Sci. 47: 1395­

1398

Sohal, R. S., Buchan, P. B. (1981). Relationship between fluorescent age pigment,

physiological age and physical activity in the housefly, Musca domestica.

Mechanisms Ageing Dev. 15: 243-249

Sohal, R. S., Ragland, S. S., Allen, R. G., Farmer, K. (1983). Correlation between

oxygen consumption, lipofuscin, fluorescent age pigment, lipid peroxidation

potential and life span in insects. In: Greenwald, R. A, Cohen, G. (eds.) Oxy

radicals and their scavenger systems. Vol. II: Cellular and medical aspects.

Elsevier Science Publishing Co., Inc., New York, p. 361-364

Tappel, A L. (1975). Lipid peroxidation and fluorescent molecular damage to

membranes. In: Trump, B. F., Arstila, A V. (eds.) Pathobiology of cell

membranes. Vol. I. Academic Press, New York, p. 145-170

52

Tappel, A. L., Fletcher, B., Deamer, D. (1973). Effect of antioxidants and nutrients on

lipid peroxidation fluorescent products and aging parameters in the mouse. J.

Geront. 28: 415-424

Thommen, H., Wackernagel, H. (1964). Zum Vorkommen von Keto-Carotinoiden in

Crustaceen. Naturwissenschaften 51: 87-88

Tokunaga, T. (1964). Studies on the development of dimethylamine and formaldehyde in

Alaska pollack muscle during frozen storage. Bull. Hokkaido reg. Fish. Res.

Lab. 29: 90-97

Trombly, R., Tappel, A. (1975). Fractionation and analysis of fluorescent products of

lipid peroxidation. Lipids 10: 441-447

Vernet, M., Hunter, 1. R., Vetter, R. D. (1988). Accumulation of age pigments

(lipofuscin) in two cold-water fishes. Fish. Bull. U.S. 86: 401-407

Zeller, P., Bader, E, Lindlar, H., Montavon, M., MUller, P., RUegg, R., Ryser, o.,

Sancy, a., Schaeren, S. F., Schwieter, U., Stricker, K., Tamm, R., ZUrcher,

P., Isler, O. (1959). Synthesen in der Carotinoid-Reihe. 13. Mitteilung. Synthese

von Canthaxanthin. Helv. chim. Acta 42: 841-847

53

-_.-.._-_ ..

Table2-1.Previous studiesof spectrofluorometrically measured fluorescent agepigmentaccumulation in aquatic animals. Tissues:b =brain;ceph.=cephalothorax; h =heart; 1=liver; m =muscle; s =spleen. Storage: FA, preservedin formalin; FNS, preservedin formalin then transferred to storagesolution; refrig.,refrigerated. Solvent: C, chloroform; M, methanol. Sonic.: ultrasonication.rt: assayedat room temperature. -: no information available basedon the described methods. Abbreviations for speciesnamesare:E.. superba =Euphausia £,., fa. americanus =Pseudopleuronectes a., Q.. aureus =Oreochromis .lL., H.. araneus =.lhas..iL" .c...hyperbQreus =Calanus. h., M.. norvegica =Meganyctiphanes n...., 1.illecebrosus =m I, E...~ =fu.Qx 1.., L...limanda =Limanda1, .Q.. albisella =Dascyllus ,ih, L.. Imlli£ =Leuresthes 1.u &. nobilis =Atractoscion n, fa. californicus =Paralichthys~,Q..mykiss=Oncorhynchus IIh, M.. pacificus=Microstomus p,.., .Q.. carinate=Daphnia £...

Species Tissue Age Tissue storage Extractincub. Assay Source

known method time temp. time Sonic. Solv. temp.

E. superba ceph. no 5%FNS 4wk/- WOC - yes C WOC Ettershank (1983)VI~

E. superba ceph. no 5%FA/S 4wk/- 2QoC - yes C WOC Ettershank (1984a)

E. superba ceph. no 5% FAIS 4wk/- WOC - yes C WOC Ettershank (l984b)

E. superba ceph. no 5%FA/S 4wk/- 2Q°C - yes C WOC Ettershank (1985)

P. americanus m no fresh - - - no C - Luo and Hultin(1986)

O. aureus b.h.l.s.m yes fresh - - - yes C+M rt Hammer& M-Rao (1987)

H. araneus larvae yes 4% FA - - - no C+M rt Hirche& Anger (1987)

C. hyperboreus body no frozen(?°C) - - - no C rt Nicol (1987)

M. norvesica body no frozen(?°C) - - - yes C rt Nicol (1987)

1 illecebrosus mantle no frozen(?°C) - - - no C rt Nicol (1987)

Table 2-1. (Continued) Previous studies of spectrofluorometrically measured fluorescent age pigment accumulation in aquaticanimals. Tissues: b =brain; ceph. = cephalothorax; h = heart; I =liver; m =muscle; s =spleen. Storage: FA, preserved informalin; FA/S, preserved in formalin then transferred to storage solution; refrig., refrigerated. Solvent: C, chloroform; M,methanol. Sonic.: ultrasonication. rt: assayed at room temperature. -: no information available based on the described methods.Abbreviations for species names are: E... superba =Euphausia s.., ~ americanus =Pseudopleuronectes iL., Q.. aureus =Oreochromis a, lL. araneus =~ a, !: hyperboreus =Calanus h.., M.. norve(:ica =Me(:anyctiphanes 11&, 1.illecebrosus =I!gi.., Ii~=~1,:b limanda =Limanda 1,Ik albisella =Dascyllus a.., L...~=Leuresthes L,~ nobilis =Atractoscionn..., .e californicus =Paralichthys £..,!1 mykiss =Oncorhynchus m..,~ pacificus =Microstomus lk Ik carl nata =Daphnia £...

Species Tissue Age Tissue storage Extract incub. Assay Source

known method time temp. time Sonic. Solv. temp.

klimanda b no - - - - yes C+M rt Hammer & Karakiri (1988)VIVI

D. albisella b no -WOC lOmo WOC 24h yes C rt Hill & Radtke (1988)

L. tenuis larvae yes -15 & -70°C - 4°C 3-4 h no C rt Mullin & Brooks (1988)

A. nobilis larvae yes -15 & -70°C - 4°C 3-4 h no C rt Mullin & Brooks (1988)

P. califomicus larvae yes -15 & -70°C - 4°C 3-4 h no C rt Mullin & Brooks (1988)

O. mykiss b.h.l yes -800C - refrig. - no C rt Vernet et al. (1988)

M. pacificus b no -800C - refrig. - no C rt Vernet et al, (1988)

E. superba ceph. no 5%FNS 4wk/- - - yes C rt Berman et al. (1989)

D. carinata body yes fresh - - - no C+M rt Sheehy & Ettershank (1989)

Fig. 2-1. Oreochromis mossambicus: Effect of storage temperature on relative

fluorescence intensity of skeletal muscle. (A) short-term storage of tissues at 23 and 3°e;

(B) long-term storage of tissues at -20 and -80°C. Uncorrected excitation and emission

maxima were 350 to 355 nm and 425 to 430 nm, respectively. Maxima did not shift as

intensity increased. Data presented are means ± standard deviations for triplicate

analyses.

56

-'- ......- ..._--~---_ ..- -....~----_...-., ...~.....

90A

80

70

60

50

40>.

30 ..-enc:Q) 20- 3°Cc:Q) 10 i/! !0c:Q) 0 •.j:.-. •0enQ) 0 1 2 3 4 5 6 7 8 9 10 11 12...0:::::l 90 Time (hours)u, SQ) 80 -20°C>- 70 ~!CI:lQ)

a: 60 .>'<I~ 50

40

~!30

20,/!10 -80°C

~/! !0 .~.-t-I !

-100 10 20 30 40 50 60 70 80 90 100 110 120

Time (days)

57

Fig. 2-2. Oreochromis mossambicus: Effect of incubation temperature on relative

fluorescence intensities of unfractionated chloroform/methanol-soluble extracts of brain,

heart, and muscle tissues. Extracts incubated at 23°C (A) and -20°C (B). Initial

uncorrected excitation and emission maxima were 355 and 430 nm, respectively, shifting

to 380 and 455 nm after 2 and 3 h of incubation at either temperature. Data presented are

means ± standard deviations for triplicate analyses.

58

--_._-_..---------------

873 4 5

Incubation T' 6rrne (hours)

21

o Brain

• Heart

• Muscle

A

4

2

0r---~-O)~-=;--~-

6

400

200

~ 300'enc(1)-c

59

Fig. 2-3. Oreochromis mossarnbicus: Effect of incubation temperature on relative

fluorescence intensities of final chloroform-soluble extracts of brain, heart, and muscle

tissues. Extracts incubated at 23°C (A) and -20°C (B). Uncorrected excitation and

emission maxima were 350 to 355 nm and 425 to 430 nm, respectively. Maxima did not

shift as intensity increased. Data presented are means ± standard deviations for triplicate

analyses.

60

_.._-_..._------_.- •.._---_ .._-- _.

8

-Q-r

765432

o Brain

• Heart

• Muscle

A

o

/1/ I

/1 ,A~·2 _

o~--_.,----I----I----I

120

>- 80-C/lc:Q)-c: 40Q)

oc:Q)

0oC/lQ)

""a.2 Bu,Q)

>-ellQ)

a:<l<fi

10

8

6

Incubation Time (hours)

61

Fig. 2-4. - Oreochromis mossambicus. Effects of ultrasonication at 25 W of output on

the expression of sample fluorescence. (A) change in relative fluorescence intensities of

sonified homogenates of brain, heart, and muscle, (B) uncorrected fluorescence spectra

of brain extracts before sonication (excitation/emission maxima =355/440 nm,

respectively) and after 5 min sonication (excitation/emission maxima =383/453 nm,

respectively). Data in (A) are means ± standard deviations for triplicate analyses.

62

353010 15 20 25Sonication Time (minutes)

5

o Brain

• Heart

A Muscle

A1000

>.-C/J 900c:2 800c:

Q) 700o:ii 600o~ 500...o::::l 400u,

Q) 300>-; 200Q)

a: 100<J~ 011II:;;,,--""---....1.---........----'-----1..__--"__---1

o

12B

550500450400350300250OL--.-.-:=...!....__...I..-__...L-__.......__....J-__--L.__-J

200

10

2

Q) 8oc:Q)oC/J 6Q)c:E::::l 4

..J

Excitation/Emission Wavelength (nm)

63

Fig. 2-5. - Oreochromis mossambicus. Uncorrected fluorescence spectra of

aqueous/methanol phase extracts of brain tissue. (A) excitation spectrum at emission =

455 om and emission spectrum at excitation =338 om; (B) excitation spectrum of same

sample at emission =520 nm. Excitation peaks at 286, 351, and 447 om indicate

contaminating flavin compounds.

64

1312 A 338 455

1 110

9- 8en- 7c:::l 6>.... 5~...

4-.c... 3~- 2Q)

o 1c:Q) 0oen 351Q) 7c: BE 6::l

5..J

43210250 300 350 400 450 500 550

Excitation/Emission Wavelength (nrn)

65

Fig. 2-6. - Oreochromis mossarnbicus. (A) Relationship between assay temperature and

luminescence of quinine sulfate standard (0.1 mg/11N H2S04) and chloroform extracts

from brain, and (B) resulting relative fluorescence intensity values from same sample

data.

66

Quinine Sulfa te Standard

12 18 24 30

( Oe )Temperature

6

0-0_0_0 Brain Extract-0-0-0_ _

o 0-0-0-0_0_0-0-

o

~~-.-.-.-.-.-.-.-.-.-.-.-.-.-.-

4

2

~...L...-1....-.~.L.-....L..--.L....J-....L-...---'---0.0

5

2.0A-en-c: •::l

>. 1.5~

~~-.0~

~

1.0-Q)0c:Q)0en

0.5Q)c:

E::l..J

>.-"Ciic:Q)-c:

Q)oc:Q)oQ)en...o::l

IJ..Q)

>-~Q)

a:

67

-._-..~ .._,~--_.---~ ..

Fig. 2-7. - Qreochrornis mossambicus. Fluorescent age pigment (FAP) accumulation

and brain weight as functions of chronological age. (A) Total FAP content of an entire

brain; (B) weight-specific FAP (relative fluorescence intensity) in brain; (C) brain dry

weight.

68

....;,.; ....... .:...--._"""-_--.4...:....:...... _

A :::

'" B

/Co ~ !'

....•. ..........•. ..... ..'.....:::l:::o =.•

/,' .......•" ",,0 ::: I ... ·

't ~ - ~~""a :J.,••••

a~ ~·~..?-:~:~·i ••l~-~-~·~·='Ii ...

160

~ 140Q)(J

c 120Q)(Jenell 1000::Ju, 80c:ell

cD 60Q)

0 40.J:

~20

a14

.:. Immature

o Male

• Female

- Male 1" Immature

.......Female 1" Immature

:::

:J

:J

:J

>- 12'iiic:Q)

.= 10

8

6

•

c

Q)(Jeell(Jenell'-o::J

u:ell>(;;Q)

a:

0

40

35

OJ 30.5:c 25OJQ)

~ 20>-'-0 15eell

cD 10

5

OL....E!._~--''----'---'_--I._--I._--'-_.........- .......o 100 200 300 400 500 600 700 800 900

Age (days)

69

CHAPTER 3

FLUORESCENT AGE PIGMENT ACCUMULATION AND THE USE OFMULTIPLE REGRESSION MODELS FOR AGE ESTIMATION IN TWO

FISH SPECIES

Introduction

For the past decade, aquatic biologists have attempted to establish whether a

relationship exists between chronological age and the accumulationof soluble fluorescent

age pigments (pAP) extracted from postmitotic tissues, with the ultimate goal of applying

this method to age prediction in population studies. To date, results from FAP studies

have been published for crustacean (Ettershank, 1983, 1984a,b, 1985; Hirche & Anger,

1987; Nicol, 1987; Crossland et aI., 1988; Berman et aI., 1989; Sheehy & Ettershank,

1989; Nicol et al., 1991), mollusk (Nicol, 1987; Clarke et al., 1990), and fish (Luo &

Hultin, 1986; Hammer & Madhusudhana-Rao, 1987; Hammer, 1988; Hammer &

Karakiri, 1988; Hill & Radtke, 1988; Mullin & Brooks, 1988; Vemet et aI., 1988; Hill

&Womersley, 1991) species. Results from some of these preliminary studies have been

variable, leading to skepticism about the applicability ofFAP assays for age estimation;

however, variability of results from some ofthese early studies has been explained in part

by the application of methodological procedures leading to inaccuracies in the FAP assay.

Nicol's (1987) discovery that formalin fixation of samples artificially induced FAP-like

fluorescence in the tissue essentially invalidated the results of numerous studies

(Ettershank 1983, 1984a,b, 1985; Hirche & Anger, 1987; Hammer, 1988; Berman et aI.,

1989). Furthermore, Hill &Womersley (1991; see Chapter 2) demonstrated that other

factors related to sample storage, extraction techniqueand handling time may have

affected variability of the FAP assay in numerousother studies.

An additional problem that has led some investigators to question the FAP assay

has been the presumption that weight-specific concentrationsof FAP (commonly referred

70

to as "relative fluorescence intensity", or RFI) should show some positive correlation

with age. On the contrary, most studies have shown that RFI values are either negatively

correlated with age (Mullin & Brooks, 1988;Vernet et al., 1988; Crossland et al., 1988;

Sheehy & Ettershank, 1989; Clarke et al., 1990),or decrease initially, increasing later

with advancing age (Clarke et al., 1990; Hill & Womersley, 1991). Only one study

reports consistent age-related increases in weight-specificFAP concentrations in an

aquatic species (Luo & Hultin, 1986). Earlier reports describe such a trend in other

vertebrates (e.g. - Tappel et al., 1973; Shimasaki et aI., 1977; Miquel et al., 1978;

Shimasaki et al., 1980). As many investigators have pointed out, any relationship

between tissue FAP concentration (RFI) and age will be inherently confused as long as

the rate of FAP genesis differs from the rate of accumulation of tissue mass (Hammer&

Madhusudhana-Rao, 1987; Vemet et aI., 1988; Hill & Womersley, 1991; Sheehy &

Roberts, 1991). Thus, RFI values are only useful for comparing FAP from areas within

the same tissue, FAP in different tissues from the same organism, FAP between different

animals of identical age, or for examining trends in FAP accumulationin animals whose

somatic or organ growth has approached some asymptote.

Questions have also been raised as to the vaildity of the FAP technique originally

developed by Fletcher et al. (1973). This controversy has centered mainly on

discrepancies in the autofluorescenceproperties of histologicallyobserved age pigment

granules (lipofuscins and ceroids) having emission maxima in the 500 to 630 nm

wavelength range, and soluble age-relatedfluorophores (FAP) which emit light in the

blue (400-490 nm) range. Eldred et al. (1982) suggested that this problem was related to

differential sensitivityof fluorescence instrumentation which biased fluorescencereadings

in the blue-green range unless corrected; however, Eldred & Katz (1989) later

demonstrated that this difference remained even after instrumentalbiases were corrected.

Numerous attempts have been made to characterize age pigment fluorescenceby

71

reproducing crosslinking reactions involved in the pathology of aging, and all have

resulted in modelfluorophores withautofluorescence in the 400-490 om range (Chio&

Tappel, 1969; Malshetet al., 1974; Gutteridge et al., 1977; Kikugawa & Ido, 1984;

Fukuzawa et al., 1985; Kikugawaet al., 1989; Yin & Brunk, 1991b). Despite this

controversy, a numberof studies haveclearlydemonstrated that solubleFAP is presentin

the tissuesof aquatic animals and that it accumulates in a mannerthat correlates with

chronological age (Luo & Hultin, 1986; Mullin& Brooks, 1988; Vemet et al., 1988; Hill

& Womersley, 1991; Nicolet al., 1991). Subsequently, Yin & Brunk (1991a) provided

data to support all FAP studies,and thus the FAP assay, by demonstrating that model

FAP fluorophores undergo a metachromatic shiftin fluorescence emissionfrom the

orangeand yellow rangeat highconcentration to the blue range at dilutedconcentrations.

This 'inner-filter' effect goes a long way towardexplaining the differencesbetween

fluorescence spectral properties of intactage pigmentgranules that are dense, light­

absorbing bodies (Katz et al., 1984; Docchioet al., 1991)and those that have been

isolated and solvent-extracted (Hendleyet al., 1963; Siakotos& Koppang, 1973).

The question of primary importance to mostfield biologists interested in the FAP

methodis whetherFAP content,which is a function of physiological agingprocesses,

can be used to predict chronological age. The presentstudy was designed to address this

issueby examining age-related patterns of FAP accumulation in laboratory-reared

populationsof a freshwater (Puntiusconchonius Hamilton) and a euryhaline

(Qreochromis mossambicus Peters) teleost The problemof age prediction was

approached by developing multiple linearregression models which include information

availablefrom somaticgrowth, otolithmorphometries, and FAP content, to establish

whetheradditional information available from theFAP methodcan improve the ability to

predictchronological age.

72

------- . "'--- ---- ..

Materials and methods

Specimenculture and preparation

Known-age specimens of O. mossarnbicus were reared for up to 886 d and sampled

under conditions described previously (Hill & Womersley, 1991). P. conchonius fry

were hatched at the Kewalo Basin ResearchFacilityof the NationalMarine Fisheries

Service, Honolulu, and reared for up to 538 d in cylindrical fiberglass tanks. Fish were

reared at room temperature (22 to 24°C), with an approximate 12 h photoperiod, bidaily

water changes, and fed ground trout chow twice daily. Specimens remaining after 538 d

were subsequently transferred to wet room facilities at the Department of Zoology,

Universityof Hawaii, and held under similar conditions until sacrifice at 1517 d.

Sampled P. conchonius individuals were transported live to the laboratory,

anesthetized by ice-bath immersion, weighed to the nearest 0.01 g, and standard length

measured to the nearest 0.05 mm using calipers. Specimens were labeled, wrapped in

aluminumfoil, frozen in liquid nitrogen, and either stored in liquid nitrogenor at -80°C

until dissection. Specimens were dissected while still in semi-frozen condition. Skull

tops were removed and the region surrounding the brain thoroughly rinsed with Ringer's

saline solution to remove blood and lipid deposits. Brain tissues removed included the

cerebral, inferior, optic, and cerebellar lobes with the medulla being severed posterior to

the tenth cranial nerve. Cranial nerves were severedat their point of attachment to the

brain. Hearts were excised and ventricle myocardial tissues separated and saved. Brain

and myocardial tissues were briefly rinsed in Ringer's solution, placed in Eppendorfvials

and immediately frozen in liquid nitrogenuntil analysis.

Sagittal otoliths from both species were removedfrom the otic cavity, cleaned of

extraneous tissue, rinsed in distilled water, soaked in 95% ethanol, and stored dry. O.

mossarnbicus otoliths were weighed to the nearest 0.001 mg using a Sartorius Model

4503 MP6 electronic microbalance, and otoliths from P. conchonius were weighed to the

73

nearest 0.0001 mg using a Cahn Model C-32 electronic microbalance. Otolith lengths

and widths for both species were measured to the nearest 0.01 mm using a stereo

dissecting microscope with an ocular micrometer.

FAP extraction and spectrofluorometric assay

Soluble FAP materials from brain and myocardial tissues were quantified following

the modified extraction and assay protocols outlined by Hill & Womersley (1991).

Tissues were lyophilized and their dry weight measured to the nearest 0.1 ug on a Cahn

Model C-32 electronic microbalance. Tissues were macerated at room temperature for 4

min in a Potter-Elvehjem glass tissue homogenizer using at least 1 ml of spectroscopic

grade chloroform (CHCI3) per 10 mg dry tissue. The initial meniscus was marked and

the volume reconstituted after homogenization. Homogenates were transferred to glass

centrifuge tubes and vortex-mixed for 2 min with an equal volume of a solution

containing 100 mM magnesium chloride in glass distilled water and spectroscopic grade

methanol at a ratio of 3:1. Samples were centrifuged at 0-5°C for 10 min at RCF =3000

x g. Nonpolar and polar fractions were immediately removed and assayed for FAP

content Sample extractions were performed singly and rapidly to prevent in vitro

production ofFAP-like fluorophores in the sample solvent (Hill & Womersley, 1991).

Extract solvents were placed in quartz-silicate microcuvettes with a 4 mm pathlength

and sealed with a teflon stopper to prevent evaporation. Temperature of the sample

solvent was maintained at 18°C using a thermostatically controlled turret microcell holder

connected to a Brinkmann Lauda Model RM6 refrigerated circulating water bath. A

Perkin Elmer LS-5B luminescence spectrometer, Acer computer, and Perkin Elmer CFS

PC-control software (version 3.0) were used for the fluorometric assay. The LS-5B

luminescence spectrometer was equipped with a Xenon light source and a Hamamatsu

Type R928 red-sensitive photomultiplier tube to reduce possible bias in the 500-700 nm

region (Eldred et al., 1982). Excitation and emission slits were set at 5 and 20 nm,

74

------------ .._----- ----_ ... -.

respectively. Fluorescence spectra reported herein were corrected for the spectral

emission characteristics of the source lamp, the spectral efficiencies of the excitation and

emission monochromators, optics, and the photomultiplier tube using the method of

Rhys-Williams et al. (1983) and the emission correction routine in the CFS PC-control

software. All sample readings were expressed as a percentage of Quinine sulfate (0.1

mg/11N sulfuric acid) standard luminescence at an excitation setting of 345 nm and an

emission setting of 450 om. Sample FAP concentrations were expressed as whole-organ

%fluorescence (%FL), where:

sample luminescence%FL = x solvent volume (ml) x 100

standard luminescence

Only brain tissues were available for Q. mossarobicus, and only their chloroform­

soluble (nonpolar) FAP was quantified due to high levels of interfering flavin compounds

in the methanol/water (polar) fraction (Hill & Womersley, 1991).

Preliminary assays of P. conchonius brain and heart tissues revealed large quantities

ofFAP in the polar extract fraction. Excitation scans (250 to 500 om) at an emission

setting of 520 om (Fletcher et al., 1973) revealed low quantities of flavin compounds in

the polar fraction which, in most cases, contributed less than 5% of the total luminescence

at the FAP maxima settings. A simple algorithm was applied to polar fraction FAP data

to subtract the luminescence contributed by flavin compounds. Flavin luminescence

contribution at the FAP emission maxima was approximated and corrected by; 1)

assuming that the flavin luminescence spectrum is symmetrical in form, 2) measuring

luminescence at a point equidistant from the wavelength difference between the FAP

emission maximum and the flavin emission maximum (520 nm), and 3) subtraction of

this luminescence from the FAP raw luminescence value. For example, if the FAP

emission maximum was at 425 nm, then luminescence was also measured at 615 nm (=

520-425+520 nm) and subtracted from the FAP raw luminescence at 425 nm.

75

-'_ ......._.-----~---_ ...._---_ •.._...-... ... -"

Multiple regression analysis

Multiple regression models for fish, otolith, and FAP measurements were fitted to

the following form:

loglO (Age) =a + bj X; + b2X2 + b3X3 +...+ bnXn

where age (days) is known, a is the intercept, bi =regression coefficients, Xi =loglO

independent variables. Regressions were fitted in a stepwise manner with the inclusion

level for all variables set at £ =0.05. Independent variables included fish size [fish

weight (W) and standard length (SL)], otolith size [otolith weight (OWt), otolith length

(OL), and otolith width (OW)], brain and heart dry weight (BW and HW) and FAP

content (%FL). All variables were log transformed to meet the asswnptions of normality

and homogeneous variances. Four model types were developed in order to compare the

usefulness of information available from combinations of fish size, otolith size, and FAP

data. These models were developed for males and females separately, with data from

immature specimens included in both.

Results

Somatic and otolith growth

Somatic growth of O. mossambicus differed by sex, with males reaching a mean

standard length (SL) of 246.3 ± 15.3 mm and a mean weight (W) of 462.6 ± 101.9 g at

an age of 886 d (Fig. 3-1a,b). The largest male weighed 576.97 g with a SL of 264 mm.

Females reached a mean 185 ± 10.8 mm SL and 212.1 ± 44.3 g W by 886 d, with the

largest specimen being 253.85 g at a SL of 194 mm. Quadratic growth models

determined by regression and some associated statistics areprovided in Fig. 3-1. Growth

in SL was asymptotic in nature up to 886 d (Fig. 3-1a) for both males and females,

whereas the growth rate ofW continued to increase with age (Fig. 3-1b).

76

O. mossambicus males had larger otoliths than females (Fig. 3-lc). The best

quadratic models describing the relationship of otolith weight (OWt) to age for both males

and females are shown in Fig. 3-1c. The rate of growth in otolith weight did not decrease

significantly during this experiment. At 886 d, males had a mean OWt of 49.149 ± 1.299

mg and females had a mean OWt of 33.955 ± 3.661 mg.

Somatic growth in P. conchonius differed little between sexes for the first 538 d of

age (Fig. 3-2a,b). Logarithmic models described the relationship between SL and age

reasonably well up to 538 d. Growth was asymptotic in nature for both males and

females (Fig. 3-2a). Growth in W, also modeled using logarithmic equations (Fig. 3­

2b), was most rapid between 60 and 300 d, leveling off between 300 and 538 d.

Individuals held until 1517 d continued to grow, with females attaining a mean W of

16.59 ± 3.56 g (max. =18.86 g) and males having mean W of 7.43 ± 1.93 g (max. =9.05 g). Otolith growth (OWt) in P. conchonius was well described by logarithmic

regressions for both males and females and did not vary greatly between the sexes (Fig.

3-2c).

Organ growth and FAP accumulation with age

Spectral characteristics of nonpolar FAP extracted from O. mossambicus brain

tissues have been previously reported (Hill & Womersley, 1991). Brain growth and

whole-brain FAP (%FL) has also been previously described (Hill & Womersley, 1991).

Regression models describing brain growth and nonpolar FAP accumulation (%FL) with

age are presented again for comparative purposes (Fig. 3-3). In brief, whole brain

nonpolar FAP (%FL) accumulated gradually during the first 370 d, after which it

accumulated more rapidly, with males (mean %FL =140.6 ±37.3) having almost twice

as much FAP as females (mean %FL of 74.3 ± 8.51) by 886 d (Fig. 3-3b).

Characteristics of corrected emission (EM) spectra of FAP extracted from P.

conchonius brain and heart tissues are presented in Fig. 3-4. Brain tissues had far greater

77

proportions of polar to nonpolar FAP than did heart tissues. Polar FAP extracted from

both tissue types had similar corrected emission maxima, which ranged from 415 to 420

nm when excited (EX) at 340 nm wavelength. There was no obvious evidence of

fluorescence from interfering compounds (e.g., retinols and flavins), however, excitation

scans of polar extracts at EM =520 revealed minor levels of flavin fluorescence, and

steps were taken to account for this contribution to the total fluorescence at the EX and

EM maxima of the FAP (see Methods section).

Brain growth and FAP accumulation with age in P. conchonius were expressed with

logarithmic models (Fig. 3-5). Brain growth was most rapid during the first year of life,

and was similar for males and females (Fig. 3-5a). At 1517 d, males had a mean brain

dry weight (BW) of 9.730 ± 1.259 mg, and females had a mean BW of 8.236 ± 2.050

mg.

Soluble brain FAP in P. conchonius was detected in both the nonpolar and polar

solvent fractions; levels were up to 100-fold greater in the polar fraction (Fig. 3-5b,c).

Nonpolar and polar FAP were detected in immature fish sampled at 60 d and gradually

increased in quantity with age in both sexes. The relationship of brain nonpolar FAP to

age, modeled up to 538 d by logarithmic regression, was variable for both males (r2 =

0.615, P < 0.0001) and females (r2 =0.682, P < 0.0001) at each age, and the rate of

accumulation declined with increasing age (Fig. 3-5b). The highest mean values of

nonpolar FAP were present in the oldest individuals sampled. Brain polar FAP was also

present in large quantity in brain tissues at 60 d, and they slowly increased with time.

Regression models were highly variable for both males (r2 =0.312, P < 0.0001) and

females (r2 =00405, P < 0.0(01).

Relationships of heart weight and FAP content to age in P. conchonius were

modeled using logarithmic regression (Fig. 3-6). Heart weight increased continuously

but became more variable with increasing age (Fig. 3-6a).

78

FAP extracted from the ventricle myocardium was also more soluble in polar than

nonpolar solvents (Fig. 3-6b,c), but the difference was not as great as was found for

brain tissues. Both nonpolar and polar FAP were detectable in heart tissues from the

youngest individuals (60d) which had an average ventricle dry weight of 0.147 mg.

Nonpolar heart FAP increased linearly with increasing age (Fig. 3-6b), but with

considerable variability in the data (male r2 =0.605, P < 0.0001; female r2 =0.517, P <

0.0001). Similar trends were apparent for polar heart FAP (Fig. 3-6c), which was

approximately four times as abundant as nonpolar FAP extracted from the same tissues.

Multiple regression analyses

Oreochromis mossambicus

Multiple linear regression models using various combinations of independent

variables from fish size, FAP, and otolith morphometric data provided strong predictions

of age for O. mossambicus (Table 3-1). Four model types were developed in order to

compare the relative usefulness of information available from fish size, FAP, and otolith

data. For the first type considered, data included fish size (SL and W) and otolith

morphometries; for the second type, fish size and organ weight and FAP data were

considered; for the third type, all variables except FAP were considered; for the fourth

type, all these independent variables were considered.

Stepwise regression analysis of fish size and otolith data (organ size and FAP data

excluded) demonstrated that fish weight and the three otolith measurements (OWt, OL,

and OW) were the most important variables in predicting age for both males and females

(Table 3-1). All of the correlations were highly significant, and the included variables

explained 97.99% of the variation in age for males and 97.20% for females, as measured

by the adjusted r2 values.

Stepwise regression analysis of fish size, organ size and FAP data (otolith data

excluded) demonstrated that fish weight, brain dry weight, and nonpolar FAP (%FL)

79

-------._---- --

were the best predictorsof age for bothsexes (Table3-1). Again,coefficients for each

variablewere highlysignificant, and coefficients of determination were only slightly

higher for both males (r2 =0.9820)andfemales (r2 =0.9830) than in the model that

excluded brain and FAP information. Thus, in the absenceof information on otolith size,

a model includingFAP contentprovided more information on age than did bodysize

alone.

When all independent variables exceptFAP wereconsidered, it was determined that

fish weight, brain weight, otolithweight and otolithlength were the best predictor

variables for males (r2 =0.9815), and that brain weight, otolith weightand otolith length

were best for females (r2 =0.9798). The fit of these models was slightly poorer than

those includingsomaticandFAP data,and slightlybetter than those including only

somaticand otolithdata.

When information from all independent variables was includedin the stepwise

multiple regression analysis, it was determined that brain weight, nonpolarFAP (%FL),

and otolith weightwere the best variables for predictingage of males,and that these same

variablesalong with otolith lengthwere bestfor ageprediction in females. Coefficients

of determination were marginally highest for these models, explaining 98.43%of the

variation in age for males and 98.60% for females.

Puntius conchonius

The four multipleregression model typesdevelopedfor P. conchonius were similar

to thosedescribed for O. mossambicus, with the additionof information on heart weight

as well as polar and nonpolarFAP (%FL) data from both brain and heart. As for O.

mossambicus, fish size, otolith morphometries, organ weight andFAP data all proved

useful to variousdegrees for the prediction of chronological age (Table3-2).

Through stepwiseregression analysis of fish size and otolithdata (minusorgan and

FAP data) it was determined that fish weight and otolithweightwere the most important

80

variables for age prediction in male P. conchonius, as were fish standard length, otolith

weight, and otolith width for females (Table 3-2). These variables explained 95.35% of

the variation of age for males, and 97.12% for females, and the coefficients of all the

variables were significant (Table 3-2).

Stepwise multiple regression of fish size, organ weight, and FAP data revealed that

brain weight and heart nonpolar and polar FAP (%FL) were the best predictors of age in

males (r2 =0.9406) and that fish weight, brain weight, and brain and heart nonpolar FAP

(%FL) were the best variables for females (r2 =0.9402) (Table 3-2). Coefficients of

determination were slightly less than for the fish size/otolith size model, however

information on FAP content strengthened the prediction of age over that of somatic

variables alone.

When all independent variables except FAP were considered, brain weight, heart

weight, and otolith weight were the best predictor variables for males (r2 =0.9604), and

fish standard length, otolith weight, and otolith width were best for females (r2 =

0.9712). These models provided equal or better age prediction than those including

somatic and otolith data or somatic and FAP data.

As was the case for O. mossambicus, the strongest models for age prediction in P.

conchonius were the ones including some information from each of the data types (Table

3-2). Stepwise regression analysis of all independent variables determined that brain

weight, heart nonpolar and polar FAP, and otolith weight were the most important

variables for age prediction in males (r2 =0.9633). For females, the best variables for

age prediction included fish standard length, heart polar FAP, otolith weight, and otolith

width (r2 =0.9725).

81

Discussion

Results from this study provideadditionalevidence that FAP accumulatesin brain

and heart tissues offish over time and, more importantly, that informationon soluble­

FAP content can, at the very least, strengthen the prediction of chronologicalage. This

result is particularlyrelevant to researchers wishing to apply an age pigment technique,

eitherhistologicalor biochemical, as an alternativeor supplement to standardmethodsof

agedetermination. Patterns ofFAP accumulationreported in this researchclearly show

that rates of FAP genesis can be both species- and tissue-specific. NonpolarFAP

fractions extracted from brain tissue of Q. mossambicusand P. conchoniusincreased

with age, but the rate of FAP genesisapparently began to slowfairly early in life in P.

conchonius. The rate of accumulation was increasinglyrapid to the maximum age

sampledin Q. mossambicus. FAP accumulation differed in patternbetween brain and

heart tissues of£.. conchonius,with the rate of accumulationof heart FAP remaining

more constant with age. Within a given tissue type, considerablesimilaritywas evident

in the pattern of accumulationof nonpolar and polar FAP in £. conchonius,despite vast

differencesin the relative solubilities. Resultsof this nature emphasize the importanceof

examiningmore than one type of tissue when initiating studieson FAP accumulationfor

age prediction.

The decline in rate of accumulation ofFAP observed in brainsofP. conchonius with

age appears inconsistentwith the assumption that age pigmentsaccumulate in a linear

fashion throughout an animals' lifespan (Strehler et al., 1959); however, the assumption

of linearage pigment accumulation may not be entirelyvalid ifontogenetic changes in

metabolicrate are involved. While informationon such changes is presentlyunavailable

for P. conchonius, studies on other fish species have demonstratedage-relateddeclines in

weight-specific oxygen consumption, a trendrelated to the decreasedratio of area to

volume at greater age (De Silva et al., 1986; Oikawa et al., 1991). In one of the few

82

studies on FAP accumulation in aquatic species, Vemet et al. (1988) also reported a

leveling-off in the apparent rate offormation ofFAP in brain tissue of Dover sole

(Microstomus pacificus). Miquel et al. (1978) reported a similar decline in FAP extracted

from mouse testes.

A surprising result of the present study was the extreme ratio (ca. 100: 1) of polar to

nonpolar FAP extracted from P. conchonius brain. This is much higher than the ratios

reported in brain tissues of either Oreochromis aureus (Hammer & Madhusudhana Rao,

1987) or Q.. mossambicus (Hill & Womersley, 1991). Polar FAP was also higher in P.

conchonius heart extracts, but only by a ratio of ca. 4:1. We were only able to provide

these data for ~ conchonius because of relatively low levels of flavins in tissue samples.

Such was not the case for Q.. mossambicus, where high levels of contaminating flavin

compounds are present in brain tissue (Hill & Womersley, 1991). Clearly, flavin

contamination prevents useful measurement of polar FAP in some cases, but where it is

not masked by such interference, the polar fraction should not be ignored.

The extraction of age-related FAP which is more polar in solubility is not new (Desai

et aI., 1975; Taubold et al., 1976; Klass 1977; Davis et al., 1982), but the subject

involves some controversy. For example, Sheehy & Roberts (1991), working

exclusively with fluorescent compounds extracted from insect tissues, asserted that most

of the nonpolar fluorescence measured in other studies is in fact due to polar pteridine

contamination and not to the presence ofFAP. This does not appear to be the case for P.

conchonius, however, because the ratios of polar:nonpolar fluorescence were so vastly

different between tissue types. Ifnonpolar fluorescence were merely an artifact of polar

contaminants, consistent ratios between fraction quantities would be expected. Sheehy &

Roberts (1991) also assumed that all lipofuscin compounds are lipid in nature and that

FAP should therefore be soluble only in nonpolar solvents; however, this is not the case.

Detailed studies on the biochemical composition of isolated lipofuscin and ceroid granules

83

have revealed that the relative protein and lipid composition of these residues can vary

greatly, with protein fragments comprising up to 70% of the total mass (Porta, 1991).

Furthermore, most investigators using malondialdehyde (a breakdown product of lipid

peroxidation) to synthesize crosslinked products with FAP-like fluorescence, have done

so in polar solvents using amino acids and proteins as substrates (Chio & Tappel, 1969;

Shimasaki et al., 1982; Kikugawa & Ido, 1984). Thus, it is not surprising that much of

the FAP extracted from P. conchonius brain and heart, and the tissues of other animals,

(excluding insects) is polar in solubility.

The application of multiple linear regression analysis for chronological age prediction

is a recent development in fisheries biology (Boehlert 1985). This statistical methodology

appears to be effective for examining the ultility of FAP for age assessment Previous

studies have included variables obtained from measurement of otoliths (Boehlert, 1985)

or a combination of somatic and otolith data (Radtke et al., 1989; Radtke 1990; Beckman

et aI., 1991). Our multiple regression analyses show improvement in the prediction of

chronological age in both O. mossambicus and P. conchonius when FAP data are added.

This is the first time that such a utility has been demonstrated. FAP variables entered all

fitted models in the analysis when included in the initial variable list, however the fits (as

measured by r2) were never more than slightly better than models which included only

size, otolith, and organ weight variables.

The strongest model for Q.. mossambicus males and females was that which

incorporated all variables, with brain dry weight, brain nonpolar FAP (%FL), and otolith

weight (plus otolith length for females) being most important When otolith information

was arbitrarily excluded from male and female models, brain dry weight and FAP (%FL)

remained in the model with the addition of fish weight, to produce an age prediction

model with only slightly smaller coefficients of determination. The model with the

smallest coefficients of determination was the one that included only somatic and otolith

84

--. __.. _-----_ ..._-_ .. _..__ .- ..

data. In the absence of FAP information, fish weight and otolith weight and dimensions

were the most important age prediction variables, with fish standard length consistently

rejected from the model. Thus, for O. mossambicus, it would appear that FAP

information combined with fish and brain weight can provide a more reliable prediction of

age than otolith and somatic data alone, but that brain FAP and otolith weight data are the

best overall predictor variables.

The best age prediction models developed for f.. conchonius also included FAP data.

The most important variables for age prediction in males were heart nonpolar and polar

FAP, brain weight, and otolith weight Fish standard length, heart polar FAP, and

otolith weight and width were best for predicting female age. However. unlike O.

mossambicus, the model including only somatic and FAP data did not provide as strong a

prediction of age as the model including only somatic and otolith data. Nevertheless, the

model excluding otolith variables presented strong evidence that FAP information can

provide a more reliable prediction of age than somatic data alone. This was certainly the

case for P. conchonius males in which only brain weight, and heart nonpolar and polar

FAP data were included in the best predictive model.

From the above, it is clear that whole-organ FAP data can increase the accuracy of

age prediction in multivariate analyses of carefully standardized specimens. The use of

two teleosts species, with calcified structures traditionally used for aging, to demonstrate

the importance of FAP was an important aspect of the experimental design. This allowed

us to assess by comparison the importance of FAP in relation to calcified structures.

Thus, it was of particular interest when a combination of otolith and FAP information

provided the best predictive model. This held true even when the patterns ofFAP

accumulation were highly variable, as was the case for P. conchonius brains and hearts.

The retention of variables such as fish weight, otolith weight, organ weight, and

some FAP variables in the multiple regression models was probably due to their more

85

consistent patterns of increase with age as compared to some other variables. For

example, fish standard length was consistently excluded from multiple regression models

for O. mossambicus, whereas fish weight and otolith weight were, in most cases,

included in the models. Weight increases with age were consistent and fairly rapid for O.

mossambicus males and females, with sizes-at-age comparable to those reported by

Hodgkiss & Man (1977) for this species. The continuous pattern of growth in somatic

and otolith weight should not be surprising. O. mossambicus in this study were sampled

through less than half of their lifespan (Bruton & Allanson, 1974), and animals had only

reached 20% of their potential maximum weight (2953 g; Jubb, 1967).

Many animals of interest do not contain calcified skeletal structures suitable for

aging. Thus, it is encouraging that results of our models that excluded otolith data

demonstrated that FAP data can at least supplement, if not replace, somatic data for

predicting age. This analytical tool is especially pertinent to the problem of aging

commercially important fisheries species (e.g.- hagfish, lobster, shrimp, and squid) for

which tagging and size-based methods are currently the only alternatives.

With regard to field populations, further questions obviously must be addressed

before FAP assays or histological lipofuscin methods (Sheehy 1990b) can be used as

tools for age estimation. Age-related pigments (pAP and lipofuscin) are thought to form

as products of physiological aging processes related to metabolic rate. The rates of these

processes can vary greatly in poikilothermic animals, and little is known of the possible

effects of temperature or other environmental variables on the aging process and thus on

the rate of FAP genesis. Hammer (1988) tested the effect oftemperature on FAP

accumulation in pike (Esox lucius) larvae, but these results have been questioned due to

the use of formalin for specimen preservation. Sheehy's (1990a) histological study of

age pigment accumulation in crayfish <!'Jlerax cuspidatus) brain demonstrated a direct

effect of temperature on lipofuscin volume fraction in the olfactory lobe region. Studies

86

of temperature effects on "FAP" accumulation have been conducted on insects (Ragland

& Sohal, 1975; Sohal et al., 1981; McArthur & Sohal, 1982), but the fluorescence

extracted from insects is now suspected to be pteridine in origin (Lehane & Mail, 1985;

Sheehy & Roberts, 1991).

Application of FAP content as an aging tool for field populations will ultimately be

contingent on obtaining baseline information on age, growth, and accumulation patterns,

together with concurrent data on environmental and physiological variables that affect the

rate of FAP accumulation. The animals used for the present study were reared under

strictly controlled conditions in the laboratory. The resulting data represent a near

optimum which is probably unattainable with wild specimens. Sources ofFAP

variability must be determined and taken into account before multiple regression models

can be used for age prediction in wild stocks. A combination of mark-recapture and

laboratory growth studies will probably be required over relevant portions of the animal's

lifespan.

87

--- ------ . -_._------_ .

Literature cited

Beckman, D. W., A L. Stanley, J. H. Render & C. A Wilson, 1991. Age and growth

rate estimation of sheepshead Archosargus probatocephalus in Louisiana waters

using otoliths. Fish. Bull., U.£., Vol. 89, pp. 1-8.

Berman, M. S., A L. McVey & G. Ettershank, 1989. Age determination of Antarctic

krill using fluorescence and image analysis of size. Polar BioI., Vol. 9, pp. 267­

271.

Boehlert, G. W., 1985. Using objective criteria and multiple regression models for age

determination in fishes. Fish. Bull., us, Vol. 83, pp. 103-117.

Bruton, M. N. & B. R. Allanson, 1974. The growth of Tila,pia mossambica Peters

(Pisces: Cichlidae) in Lake Sibaya, South Africa. I. Fish Biol., Vol. 6, pp. 701­

715.

Chio, K. S. & A L. Tappel, 1969. Synthesis and characterization of the fluorescent

products derived from malondialdehyde and amino acids. Biochem., Vol. 8, pp.

2821-2827.

Clarke, A, M. A Kendall & D. J. Gore, 1990. The accumulation of fluorescent age

pigments in the trochid gastropod Monodonta lineata. I. Exp. Mar. BioI. Ecol.,

Vol. 144, pp. 185-204.

Crossland, C. 1., G. Denby, B. F. Phillips & R. Brown, 1987. The use of fluorescent

pigments (lipofuscin) for ageing western rock lobster (Panulirus cygnus) and

scampi (Metanephrops andamanicus): a preliminary assessment. CSIRO Mar.

Lab. Rep., No. 195, pp. 1-21.

Davis, B. 0., G. L. Anderson & D. B. Dusenbery, 1982. Total luminescence

spectroscopy of fluorescence changes during aging in Caenorhabditis elegans.

Biochem., Vol. 21, pp. 4089-4095.

88

Desai, 1. D., B. L. Fletcher & A. L. Tappel, 1975. Fluorescent pigments from uterus of

vitamin E-deficient rats. Lipids, Vol. 10, pp. 307-309.

De Silva, C. D., S. Premawansa & C. N. Keembiyahetty, 1986. Oxygen consumption in

Oreochromis niloticus (L.) in relation to development, salinity, temperature and

time of day. I. Fish BioI., Vol. 29, pp. 267-277.

Docchio, E, M. Boulton, R. Cebeddu, R. Ramponi & P. D. Barker, 1991. Age-related

changes in the fluorescence of melanin and lipofuscin granules of the retinal

pigment epithelium: a time-resolved fluorescence microscopy study. Photochem.

Photobiol., Vol. 54, pp. 247-253.

Eldred, G. E. & M. L. Katz, 1989. The autofluorescent products of lipid peroxidation

may not be lipofuscin-like. Free Rad. BioI. Med., Vol. 7, pp. 157-163.

Eldred, G. E., G. V. Miller, W. S. Stark & L. Feeney-Bums, 1982. Lipofuscin:

resolution of discrepant fluorescence data. Science, Vol. 216, pp. 757-759.

Ettershank, G., 1983. Age structure and cyclical annual size change in the Antarctic krill,

Euphausia superba. Polar BioI., Vol. 2, pp. 189-193.

Ettershank, G., 1984a. A new approach to the assessment of longevity in the Antarctic

krill Euphausia superba. I. Crust. BioI., Vol. 4, pp. 295-305.

Ettershank, G., 1984b. Methodology for age determination of Antarctic krill using the

age pigment lipofuscin. BIOMASS Handb., No. 26, pp. 1-14.

Ettershank, G., 1985. Population age structure in males and juveniles of the Antarctic

krill, Euphausia superba Dana. Polar BioI., Vol. 4, pp. 199-201.

Fletcher, B. L., C. J. Dillard & A. L. Tappel, 1973. Measurement of fluorescent lipid

peroxidation products in biological systems and tissues. Anal. Chern., Vol. 52,

pp. 1-9.

89

Fukuzawa, K, K. Kishikawa, A. Tokumura, H. Tsukatani & M. Shibuya, 1985.

Fluorescent pigments by covalent binding of lipid peroxidation by-products to

protein and amino acids. Lipids, Vol. 20, pp. 854-861.

Gutteridge, J. M. c., A. D. Heys, & J. Lunec, 1977. Fluorescent malondialdehyde

polymers from hydrolyzed 1,1,3,3,-tetramethoxypropoane. Analytica Chim.

Acta, Vol. 94, pp. 209-211.

Hammer, C., 1988. Accumulation of fluorescent age pigments in embryos and larvae of

pike, Esox lucius (Esocidae, Teleostei) at different incubation temperatures.

Environ. BioI. Fish., Vol. 22, pp. 91-99.

Hammer, C. & M. Karakiri, 1988. Disease dependent accumulation of fluorescent age

pigments in the common dab CLimandalimanda, Teleostei). In, Lipofuscin ­

.l281:~ of~ art, edited by I. Zs.-Nagy, Akademiai Kiad6, Budapest and

Elsevier Science Publishers, Amsterdam, pp. 445-446.

Hammer, C. & G. Madhusudhana-Rao, 1987. Individual and feed dependent

accumulation of fluorescent of fluorescent age pigments in five organs of tilapia

(Qreochromis~). Zeitschr. Angew. ZQQ!., Vol. 74, pp. 275-292.

Hendley, D. D., A. S. Mildvan, M. C. Reporter & B. L. Strehler, 1963. The properties

of isolated human cardiac age pigment I. Preparation and physical properties. I.

Gerontol., Vol. 18, pp. 144-150.

Hill, K T. & R. L. Radtke, 1988. Gerontological studies of the damselfish, Dascyllus

albisella. Bull. Mar. Sci., Vol. 42, pp. 424-434.

Hill, K T. & C. Womersley, 1991. Critical aspects of fluorescent age pigment

methodologies: modification for accurate analysis and age assessments in aquatic

organisms. Mar. BioI., Vol. 109, pp. 1-11.

90

Hirche, H.-J. & K Anger, 1987. The accumulation of age pigments during larval

development of the spider crab, Hyas araneus (Decapoda, Majidae). Compo

Biochem. Physiol., Vol. 88B, pp. 777-782.

Hodgkiss, J. I. & H. S. H. Man. 1977. Age composition, growth and body condition of

the introduced Sarotherodon mossambicus (Cichlidae) in Plover Cove Reservoir,

Hong Kong. Environ. BioI. Fish., Vol. 2, pp. 35-44.

Ikeda, T. & P. Dixon, 1982. Body shrinkage: a possible over-wintering strategy of the

Antarctic krill (Euphausia superba Dana). J:. Exp. Mar. BioI. Ecol., Vol. 62, pp.

143-151.

Ikeda, T. & P. G. Thomas, 1987. Longevity of the Antarctic krill (Euphausia superba

Dana) based on a laboratory experiment.Proc, Natl.!n.s.t. Polar Res. ~.,

Polar Bio!., Vol. 1, pp. 56-62.

Jubb, R. A., 1967. Freshwater Fishes QfSouthern Africa. Cape Town: A. A. Balkema,

Rotterdam.

Katz, M. L., W. G. Robison Jr., R. K Herrmann, A. B. Groome, & J. G. Bieri, 1984.

Lipofuscin accumulation resulting from senescence and vitamin E deficiency:

spectral properties and tissue distribution. Mech. Age. Dev., Vol. 25, pp. 149­

159.

Kikugawa, K. & Y. Ido, 1984. Studies on peroxidized lipids. V. Formation and

characterization of 1,4-dihydropyridine-3,5-dicarbaldehydes as model of

fluorescent components in lipofuscin. Lipids 19, pp. 600-608.

Kikugawa, K, T. Kato, M. Beppu & A. Hayasaka, 1989. Fluorescent and cross-linked

proteins formed by free radical and aldehyde species generated during lipid

oxidation. In, Lipofuscin m:ill~ pigments, edited by E. A. Porta, Plenum

Press, New York, pp. 345-357.

91

Klass, M. R., 1977. Aging in the nematode Caenorhabditis elegans: major biological and

elemental factors influencing lifespan. Mech.~. Dev., Vol. 6, pp. 413-429.

Lehane, M. 1. & T. S. Mail, 1985. Determining the age of adult male and female

Glossina morsitans morsitans using a new technique. Ecol. Entomol., Vol. 10,

pp. 219-224.

Luo, S.-W. & H. O. Hultin, 1986. Effect of age of winter flounder on some properties

of the sarcoplasmic reticulum. Mech. Age. Dev., Vol. 35, pp. 275-289.

Malshet, V. G., A. L. Tappel & V. M. Burns, 1974. Fluorescent products of lipid

peroxidation: II. Methods for analysis and characterization. Lipids 9, pp. 328­

332.

McArthur, M. C. & R. S. Sohal, 1982. Relationship between metabolic rate, aging, lipid

peroxidation, and fluorescent age pigment in milkweed bug, Oncopeltus fasciatus

(Hemiptera). 1.. Gerontol., Vol. 37, pp. 268-274.

Miquel, J., D. R. Lundgren & J. Johnson, 1978. Spectrophotofluorometric and electron

microscopic study of lipofuscin accumulation in the testis of aging mice. ,I.

Gerontol., Vol. 33, pp. 5-19.

Mullin, M. M. & E. R. Brooks, 1988. Extractable lipofuscin in larval marine fish. Fish.

Bull., us, Vol. 86, pp. 407-415.

Nicol, S., 1987. Some limitations on the use of the lipofuscin ageing technique. Mar.

BioI., Vol. 93, pp. 609-614.

Nicol, S., M. Stolp & G. W. Hosie, 1991. Accumulation of fluorescent age pigments in

a laboratory population of Antarctic krill Euphausia superba Dana. J:. Exp. Mar.

Biol. Ecol., Vol. 146, pp. 153-161.

Oikawa, S., Y. Itazawa & M. Gotoh, 1991. Ontogenetic change in the relationship

between metabolic rate and body weight in a sea bream~ major (Temminck

.& Schlegel). 1. Fish BioI., Vol. 38, pp. 483-496.

92

Porta, E.A., 1991. Advances in age pigment research. Arch. Gerontol. Geriatr., Vol. 12,

pp. 303-320.

Radtke, R. L., 1990. Age determination of the Antarctic fishes Champsocephalus gunnari

and Notothenia rossii marmorata from South Georgia. Polar BioI., Vol. 10, pp.

321-327.

Radtke, R. L., T. E. Targett, A. Kellermann, J. L. Bell & K. T. Hill, 1989. Antarctic

fish growth: profile of Trematomus newnesi. Mm:. Ecol. frQg. ~., Vol. 57, pp.

103-117.

Ragland, S. S. & R. S. Sohal, 1975. Ambient temperature, physical activity and aging in

the housefly, Musca domestica. Exp. Gerontol., Vol. 10, pp. 279-289.

Rhys-Williams, A. T., S. A. Winfield & 1. N. Miller, 1983. Relative fluorescence

quantum yields using a computer-controlled luminescence spectrometer. Analyst,

Vol. 108, pp. 1067-1071.

Sheehy, M. R. J., 1990a. Individual variation in, and the effect of rearing temperature

and body size on, the concentration of fluorescent morphological lipofuscin in the

brains of freshwater crayfish, Cherax cuspidatus (Crustacea: Parastacidae).

Qmm. Biochem. Physiol., Vol. 96A, pp. 281-286.

Sheehy, M. R. J., 1990b. The potential of morphological lipofuscin age pigment as an

index of crustacean age. Mar. BioI., Vol. 107, pp. 439-442.

Sheehy, M. R. J. & G. Ettershank, 1989. Extractable age pigment-like autofluorescence

and its relationship to growth and age in the water-flea Daphnia carinata King

(Crustacea: Cladocera). Aust. I. Zool., Vol. 36, pp. 611-625.

Sheehy, M. R. 1. & B. E. Roberts, 1991. An alternative explanation for anomalies in

"soluble lipofuscin" fluorescence data from insects, crustaceans, and other aquatic

species. Exp. Gerontol., Vol. 26, pp. 495-509.

93

Shimasaki, H., T. Nozawa, O. S. Privett & W. R. Anderson, 1977. Detection of age­

related fluorescent substances in rat tissues. Arch. Biochem. Biophys., Vol. 183,

pp. 443-451.

Shimasaki, H., N. Ueta & O. S. Privett, 1980. Isolation and analysis of age-related

fluorescent substances in rat testes. Lipids, Vol. 15, pp. 236-241.

Shimasaki, H., N. Ueta & O. S. Privett, 1982. Covalent binding of peroxidized linoleic

acid to protein and amino acids as models for lipofuscin formation. Lipids, Vol.

17, pp. 878-883.

Siakotos, AN. & N. Koppang, 1973. Procedures for the isolation of lipopigments from

brain, heart and liver, and their properties: A review. Mech. Age. Dev., Vol. 2,

pp. 177-200.

Sohal, R. S., H. Donato & E. R. Biehl, 1981. Effect of age and metabolic rate on lipid

peroxidation in the housefly, Musca domestica L. Mech. ~. Dev., Vol. 16, pp.

159-167.

Strehler, B. L., D. D. Mark, AS. Mildvan & M. V. Gee, 1959. Rate and magnitude of

age pigment accumulation in the human myocardium. I. Gerontol., Vol. 14, pp.

430-439.

Tappel, A, B. Fletcher, & D. Deamer, 1973. Effect of antioxidants and nutrients on lipid

peroxidation fluorescent products and aging parameters in the mouse. 1­

Gerontol., Vol. 28, pp. 415-424.

Taubold, R. D., A. N. Siakotos & E. G. Perkins, 1974. Studies on chemical nature of

lipofuscin (age pigment) isolated from normal human brain. Lipids, Vol. 10, pp.

383-390.

Vernet, M., J. R. Hunter & R. D. Vetter, 1988. Accumulation of age pigments

(lipofuscin) in two cold-water fishes. Fish. BulL, us, Vol. 86, pp. 401-407.

94

Yin, D. & U. T. Brunk, 1991a. Microfluorometric and fluorometric lipofuscin spectral

discrepancies: a concentration-dependent metachromatic effect? Mech. Age. Dev.,

Vol. 59, pp. 95-109.

Yin, D. & U. T. Brunk, 1991b. Oxidized ascorbic acid and reaction products between

ascorbic acid and amino acids might constitute part of age pigments. Mech. ~.

Dev., Vol. 61, pp. 99-112.

95

._-_.._-_ _.__._ _-_ .

Table 3-1. - Coefficients and associated statistics from multiple regressions ofmultivariate linear models of age (logro) for Oreochromis mossambicus. Models werefitted in a stepwise manner using combinations of the independent variables including fishsize and otolith data, fish size, organ size, FAP data, and all variables combined. Datafrom immature specimens were included in both male and female analyses. All variableswere loglO transformed for analyses. Intercepts, coefficients and their standard errors(SE) are for log transformed data.

Independent ModelVariable Coefficient SE P Adj. r2

Males (n =92)

(fish size model) 0.9740Intercept (a) =1.729Fish weight 0.413 0.007 <0.0001

(fish size and otolith model) 0.9799Intercept (a) =2.153Fish weight 0.391 0.084 <0.0001Otolith weight 0.911 0.194 <0.0001Otolith length -1.165 0.282 <0.0001Otolith width -1.179 0.381 0.0027

(fish size and FAP model) 0.9820Intercept (a) =1.696Fish weight 0.425 0.023 <0.0001Brain dry weight -0.104 0.031 0.0012Brain nonpolar FAP 0.111 0.028 <0.0001

(all variables w/o FAP) 0.9815Intercept (a) =2.062Fish weight 0.262 0.080 0.0014Brain dry weight -0.148 0.035 <0.0001Otolith weight 0.591 0.156 0.0003Otolith length -0.745 0.291 0.0122

(all variables model) 0.9843Intercept (a) =1.886Brain dry weight -0.153 0.031 <0.0001Brain nonpolar FAP 0.147 0.025 <0.0001Otolith weight 0.566 0.028 <0.0001

96

Table 3-1. (Continued) - Coefficients and associated statistics from multiple regressionsof multivariate linear models of age (loglO) for Oreochromis mossambicus. Models werefitted in a stepwise manner using combinations of the independent variables including fishsize and otolith data, fish size, organ size, FAP data, and all variables combined. Datafrom immature specimens were included in both male and female analyses. All variableswere loglO transformed for analyses. Intercepts, coefficients and their standard errors(SE) are for log transformed data.

Independent ModelVariable Coefficient SE P Adj. r2

Females (n =90)

(fish size model) 0.9640Intercept (a) =1.724Fish weight 0.441 0.009 <0.0001

(fish size and otolith model) 0.9720Intercept (a) =2.069Fish weight 0.392 0.121 0.0017Otolith weight 0.929 0.214 <0.0001Otolith length -0.455 0.148 0.0028Otolith width -1.937 0.512 0.0003

(fish size and FAP model) 0.9830Intercept (a) =1.680Fish weight 0.492 0.022 <0.0001Brain dry weight -0.188 0.030 <0.0001Brain nonpolar FAP 0.168 0.030 <0.0001

(all variables wloFAP) 0.9798Intercept (a) =2.158Brain dry weight -0.259 0.034 <0.0001Otolith weight 0.882 0.049 <0.0001Otolith length -0.374 0.125 0.0037

(all variables model) 0.9860Intercept (a) =1.985Brain dry weight -0.213 0.029 <0.0001Brain nonpolar FAP 0.172 0.026 <0.0001Otolith weight 0.759 0.045 <0.0001Otolith length -0.313 0.103 0.0032

97

Table 3-2. - Coefficients and associated statistics from multiple regressions ofmultivariate linear models of age (loglO) for Puntius conchonius. Models were fitted in astepwise manner using combinations of the independent variables including fish size andotolith data, fish size, organ size, FAP data, and all variables combined. Data fromimmature specimens were included in both male and female analyses. Animals sampledat 1517 d were excluded from the analysis. All variables were loglO transformed foranalyses. Intercepts, coefficients and their standard errors (SE) are for log transformeddata.

Independent ModelVariable Coefficient SE P Adj. r2

Males (n = 90)

(fish size model) 0.8980Intercept (a) =2.213Fish weight 0.774 0.028 <0.0001

(fish size and otolith model) 0.9535Intercept (a) =2.870Fish weight 0.143 0.064 0.0280Otolith weight 0.772 0.075 <0.0001

(fish size and FAP model) 0.9406Intercept (a) =1.776Brain dry weight 0.718 0.064 <0.0001Heart nonpolar FAP 0.201 0.060 0.0013Heart polar FAP 0.152 0.046 0.0014

(all variables w/o FAP) 0.9604Intercept (a) =2.672Brain dry weight 0.246 0.085 0.0047Heart weight 0.151 0.066 0.0251Otolith weight 0.580 0.079 <0.0001

(all variables model) 0.9633Intercept (a) =2.500Brain dry weight 0.246 0.082 0.0035Heart nonpolar FAP 0.136 0.048 0.0060Heart polar FAP 0.079 0.038 0.0381Otolith weight 0.554 0.075 <0.0001

98

--_._-- ...._-_ .._._- ....

Table 3-2. (Continued) - Coefficients and associated statistics from multiple regressionsof multivariate linear models of age (loglO) for Puntius conchonius. Models were fittedin a stepwise manner using combinations of the independent variables including fish sizeand otolith data, fish size, organ size, FAP data, and all variables combined. Data fromimmature specimens were included in both male and female analyses. Animals sampledat 1517 d were excluded from the analysis. All variables were loglO transformed foranalyses. Intercepts, coefficients and their standard errors (S£) are for log transformeddata.

Independent ModelVariable Coefficient SE P Adj. r2

Females (n =89)

(fish size model) 0.9040Intercept (a) =1.100Fish weight 0.034 0.001 <0.0001

(fish size and otolith model) 0.9712Intercept (a) =3.774Standard length -0.464 0.195 0.0193Otolith weight 0.748 0.094 <0.0001Otolith width 1.048 0.242 <0.0001

(fish size and FAP model) 0.9402Intercept (a) =1.925Fish weight 0.211 0.092 0.0242Brain dry weight 0.515 0.119 <0.0001Brain nonpolar FAP 0.144 0.066 0.0311Heart nonpolar FAP 0.137 0.057 0.0191

(all variables w/o FAP) 0.9712Intercept (a) =3.774Standard length -0.464 0.195 0.0193Otolith weight 0.748 0.094 <0.0001Otolith width 1.048 0.242 <0.0001

(all variables model) 0.9725Intercept (a) =3.730Standard length -0.493 0.190 0.0114Heart polar FAP 0.082 0.036 0.0252Otolith weight 0.722 0.092 <0.0001Otolith width 1.059 0.236 <0.0001

99

-------._..-- ...

Figure 3-1. - Oreochromis mossambicus. Body size and otolith weight as functions of

chronological age (Days). Quadratic growth regressions for immature + males (n =101)

and immature + females (n =99) all have P < 0.0001. (A) Standard length (SL) as a

function of age for immature + males: SL =3.3813 + 0.51878(Days) - 2.8061*10­

4(Days)2, r2 = 0.976; and for immature + females: SL = 8.0858 + 0.45049(Days)­

2.8413*10-4(Days)2, r2 =0.983. (B) Fish wet weight (W) as a function of age for

immature + males: W =-15.692 + 0.21892(Days) + 3.8805*10-4(Days)2, r2 = 0.929;

and for immature + females: W =-17.042 + 0.26895(Days) + 3.2033* 10-5(Days)2, r2 =

0.940. (C) Otolith weight (OWt) as a function of age for immature + males: OWt =-2.5115 + 4.997*1O-2(Days) + 9.6809*1O-6(Days)2, r2 =0.950; and for immature +

females: OWt =-2.1908 + 4.7062*1O-2(Days) - 5.3785*1O-6(Days)2, r2 =0.973.

100

300 A•- 250 .sE

E--.c 200

o ~-0>c: 150J!2"Eca 100"Cc )( Immatureca- 50 • MaleCJ)

o Female

0

600 B •500 d'

-0> 400---.c0>

300'CD;: ~.c 200eniI

100

0

55 C if50

- 450>E 40-- 35-..c:C) 30

'CD25;:

..c: 20

.'!:0 15-0 10

5

00 100 200 300 400 500 600 700 800 900

Age (days)

101

Figure 3-2. - Puntius conchonius. Body size and otolith weight as functions of

chronological age (Days). Logarithmic growth regressions for immature + males (n =

90) and immature + females (n =89) all have P < 0.0001. Animals sampled at 1517 d

were excluded from the regression analyses. (A) Standard length (SL) as a function of

age for immature + males: SL = -23.418 + 25.536 * loglO(Days), r2 = 0.905; and for

immature + females: SL =-26.240 + 26.997 * loglQ(Days), r2 =0.906. (B) Fish wet

weight (W) as a function of age for immature + males: W =-5.6956 + 3.3363 *loglQ(Days), r2 =0.785; and for immature + females: W = -7.0110 + 4.0120 *loglQ(Days), r2 =0.828. (C) Otolith weight (OWt) as a function of age for immature +

males: OWt =-0.70798 + 0.41255 * loglQ(Days), r2 =0.898; and for immature +

females: OWt =-0.75479 + 0.43729 >Ie loglO(Days), r2 =0.908.

102

70 A0- 60E 0

<i •E 0- 50..c •-C> ifc 40~

"EtiS 30

"'Cc:: x ImmaturetiS- 20 • Maleen

o Female

10 ~

20 B18 8

16- 14C>-- 120

s:C>

10'~ •

8 •..cen 6 9u: • •

4 a2 i 'if0 ~

1.2 C •0- 1.0

C>E- 0.8 •-..cC>

'0) 0.6 •~0

•..c 0.40 0

,~

'0 I •-0 0.2

0.0 ~0 100 200 300 400 500 1517

Age (days)

103

Figure 3-3. - Oreochromis mossambicus. Brain weight and fluorescent age pigment

(FAP) level as functions of chronological age (Days). Logarithmic and quadratic

regressions for immature + males (n = 101) and immature + females (n = 99) all have P <

0.0001. (A) Brain dry weight (BW) as a function of age for immature + males: BW =

-36.488 + 23.333 * logIQ(Days), r2 =0.943; and for immature + females: BW =-32.101

+ 21.067 * logIQ(Days), r2 =0.914. (B) Brain nonpolar FAP (%FL) as a function of

age for immature + males: %FL =7.4078 - 4.5872*1O-2(Days) + 2.3313*1O-4(Days)2,

r2 = 0.926; and for immature + females: %FL = 6.3477 - 1.4104*1O-2(Days) +

1.0736*1O-4(Days)2, r2 =0.936.

104

-_._----_._-----_._----_._-- .

40 A •

- 35 6'0> •E 30 ~-..... e.!: 25 0.0> 0

Q)203:

~ 15-cc:: 10·co x Immature~

CO 5 - Maleo Female

0

180~ B- -6'-J 160u..~ 1400-Q. 120 •«u.. 100 -~

(\180 ~0c.

c:: 600c:: 40c::

"(\1 20~

CO0

0 100 200 300 400 500 600 700 800 900Age (days)

105

-----------------------

Figure 3-4. - Puntius conchonius. Corrected emission spectra for soluble fluorescent age

pigments (FAP) extracted from brain and heart tissues. (A) Brain polar and nonpolar

FAP corrected emission maxima were 415 and 420 nm when excited at 340 and 345 nm,

respectively. (B) Heart polar and nonpolar FAP corrected emission maxima were 415

and 425 nm when excited at 340 and 345 nm, respectively.

106

5.5 A5.0

4.5

4.0

3.5 Brain FAP

3.0-en 2.5.1::C::J 2.0

e- 1.5~"- 1.0+-' nonpolar fraction:c

0.5 ~"-~- 0.0(1)CJC 0.35 8(1)CJen(1) 0.30C

E Heart FAP::J

0.25-.J

0.20

0.15

0.10

0.05

0.00350 400 450 500 550

Emission wavelength (nm)

107

Figure 3-5. - Puntius conchonius. Brain weight and fluorescent age pigment (pAP) level

as functions of chronological age (Days). Logarithmic regressions for immature + males

(n = 90) and immature + females (n = 89) all have P < 0.0001. Animals sampled at 1517

d were excluded from the regression analyses. (A) Brain dry weight (BW) as a function

of age for immature + males: BW = -10.337 + 6.2202 * loglO(Days), r2 = 0.919; and for

immature + females: BW = -9.3086 + 5.6154 * loglO(Days), r2 = 0.926. (B) Brain

nonpolar FAP (%FL) as a function of age for immature + males: %FL =-4.5132 +

3.8772 * loglO(Days), r2 =0.615; and for immature + females: %FL =-5.6471 +

4.3583 * logIQ(Days), r2 = 0.682. (C) Brain-polar FAP (%FL) as a function of age for

immature + males: %FL = -234.80 + 200.43 * loglO(Days), r2 =0.312; and for

immature + females: %FL =-348.42 + 258.05 * logIQ(Days), r2 =00405.

108

---_ .._---.

12 A •- 100

C)

E •-- •.... 8 0

.l::C)

"056

0

:=~"0 4c:.~

2x Immature

CD • Male

o Female

0 ~

14 B--I 0

U- 12~ 00--a.. 10<C 0 0

U- 8 I S?•

I •~ 0 88as0

0 •6a. I • d'c • I la0 4 aO

c:c

"as 2~

CO0 ~

1200 C •--I 1000U-~0-- 800a.. 0<C •U- 600 •~

0 0as • I00

S? a0 400 x I I " s· 0

a. • 8 Ic: 0

I a. d'"as 200 i e I ~ §~

g •CO a 8 • a •~

100 200 300 400 500 1517

Age (days)

109

Figure 3-6. - Puntius conchonius. Heart weight and fluorescent age pigment (pAP) level

as functions of chronological age (Days). Linear and logarithmic regressions for

immature + males (n = 90) and immature + females (n = 89) all have P < 0.0001.

Animals sampled at 1517 d were excluded from the regression analyses. (A) Heart dry

weight (HW) as a function of age for immature + males: HW =-1.6957 + 1.0180 *

loglO(Days), r2 =0.574; and for immature + females: HW =-1.2374 + 0.77254 *

loglO(Days), r2 =0.566. (B) Heart nonpolar FAP (%FL) as a function of age for

immature + males: %FL = -1.0720*10-2 + 9.2545*10-3 (Days), r2 = 0.605; and for

immature + females: %FL = 0.16355 + 7.8694*10-3 (Days), r2 = 0.517. (C) Heart­

polar FAP (%FL) as a function of age for immature + males: %FL =4.0984 +

1.9406*10-2 (Days), r2 =0.313; and for immature + females: %FL =4.1528 +

1.6479*10-2 (Days), r2 =0.368.

110

.....;.; .• l-._... _....:...~_. .•• ~. . •. _.-.....-....0.-" _ .•~ .. ".+.

5 A 0

- )( Immature0> 4E • Male- o Female •- •..c !0> 3

"Q53: 0

e- 2 •"'C I if1:: 0 IIca •Q)

I0

900

0 ~

14 B-....J 00

LL. 12~ I •0- 0

a.. 10« •LL. 8 0 0'-

ifca 0

"0 6 •C.c:0 4c:1::ca 2Q)

I0

60 C •-....J 50LL. 0~0- 40a..« •LL. •30 •'- • .~as

! 0

"0 20 • ifc. 0 0 Ii 80

• I 0t:: " 0 •as10 I e 9Q) I ••I

0 ~

0 100 200 300 400 500 1517Age (days)

111

-_.__.._---_._--_....._-_. _..__ .- ..

CHAPTER 4

INTERACTIVE EFFECTS OF SOME ENVIRONMENTAL ANDPHYSIOLOGICAL VARIABLES ON FLUORESCENT AGE PIGMENT

ACCUMULATION IN BRAIN AND HEART TISSUES OF AN AQUATICPOIKILOTHERM

Introduction

The accumulation of age pigment complexes related to physiological aging is

considered one of the most consistent events in all senescing, nonmitotic cells (see

reviews by Sohal and Wolfe 1986, Tsuchida et al. 1987, Hammer and Braum 1988,

Porta 1991). Age pigments have been variously classified as pre-ceroid, ceroid,

lipofuscin, and fluorescent age pigment, depending upon their physical and chemical

properties determined by either histologic or biochemical methods. The age-related

accumulation of these pigments has been of interest to gerontologists who hope to define

biochemical aspects of aging and ultimately slow or arrest this process. More recently,

age pigment studies have been initiated by field biologists wishing to develop predictive

models, based on pigment content, for the estimation of chronological age for population

studies (Ettershank 1983, 1984 a.b, 1985; Brown and George 1985, Hirche and Anger

1987, Hill and Radtke 1988, Hammer 1988, Mullin and Brooks 1988, Vernet et al.

1988, Berman et al. 1989, Clarke et aI. 1990, Sheehy 1990a, Hill and Womersley 1991,

Nicol et al. 1991, Hill and Womersley, in review).

The majority of these field studies have examined soluble fluorescent age

pigments (FAP) using the extraction methods and spectrofluorometric assay described by

Fletcher et al. (1973). This method of age pigment quantification has recently been

criticized because of discrepancies between fluorescent emission properties of solvent-

extracted pigments (blue emission at 400 to 490 nm; Siakotos and Koppang 1973, Tappel

1975) and those observed in situ using fluorescence microscopy (green to yellow

112

emission at 500 to 630 nm; Eldredet al. 1982,Eldred and Katz 1989). This issue has

been partly resolved by Yin and Brunk(1991),who provided strong evidence for

concentration-dependent metachromatic shifts in fluorophore emission from yellow at

high density to blue under dilutedconditions. In addition to this controversy, Nicol

(1987)and Hill and Womersley (1991) defined flaws in the sample handling procedures

that most likely affected variability of FAP results in many of the early field studies. In

spiteof theseproblems,FAP has been shownto accumulateover time in nonmitotic

tissues and organs of numerous aquatic poikilotherms(Luo and Hultin 1986,Mullinand

Brooks 1988, Vernet et al. 1988,Hill and Womersley 1991,Nicol et al. 1991,Hill and

Womersley, in review) as well as mammalian species traditionally studied in

gerontological research (Tappel et al. 1973,Shimasaki et al. 1977, Miquel et al. 1978,

Shimasaki et al. 1980).

It is of primary importance to bothgerontologistsand field fish biologists that the

factors that lead to age pigment genesis and variabilitybe defined. Biochemicalevents

involvedwith age pigment formation are consideredto be stochasticin nature,determined

in part by rates of cellular metabolism and the generationof superoxideand hydroxyl

radicals in organelles and the cytoplasm (Harman 1984). Free radicals, in the presenceof

metalsand divalentoxygen, can initiatea cascadeof lipid peroxidation events leading to

malondialdehyde and other carbonylcross-linkingreactants (Gutteridge 1987). Amine­

containingbiomolecules (amino andnucleic acids, proteins, phospholipids,etc.) that

have been cross-linkedbecomedeactivated and eventuallydigestedin the lysosomal

recycling system (Harman 1989). Cross-linked portions that are nondigestable remainin

the cytoplasm as age pigment residuesand have autofluorescentproperties. Thus, there

are numerous criticalfactors that maydetermine the rate of age pigment genesis at the

cellular level, including: 1) the rate of free radical productionthrough oxidative

metabolism (Fridovich 1976,Armstrong 1984,Freeman 1984),2) the efficiency of

113

antioxidant mechanisms responsible for free radical control (Sohal et al. 1990, Cutler

1991), 3) the susceptibility of cellular lipid components to free radical-mediated oxidation

(Packer et al. 1967, Tappe11975, Mead 1976),4) presence of catalytic factors such as

oxygen (Sohal and Brunk 1989, Sohal et al. 1989) and metals (Aloj Totaro et al. 1985,

Gutteridge 1985), and 5) the efficiency oflysosomal digestive enzymes (Davies 1988,

Marzabadi et al. 1991)

Relatively few studies have examined the effects of metabolism on age pigment

formation. Rearing DrosQphila melano~ster at different temperatures, Sheldahl and

Tappel (1974) first demonstrated a direct effect on the quantity of fluorescent products

extracted from the flies. Other studies on FAP in insects followed, testing the effects of

oxygen consumption ill. melanogaster, Miquel et al. 1976), physical activity level

(Musca domestica, Sohal and Donato 1979, Sohal and Buchan 1981), and ambient

temperature (Oncweltus fasciatus, McArthur and Sohal 1982). Questions have now been

raised as to whether the fluorescent products extracted in these studies represent FAP or

pteridine metabolites that are ubiquitous in insects (Ziegler and Harmsen 1969, Lehane et

al. 1986, Sheehy and Roberts 1991). Nonetheless, Sohal and Donato (1979) and Sohal

(1981) confirmed the positive effect of activity level on lipofuscin accumulation in M.

domestica using electron microscopy. Activity level has also been found to affect rates of

age pigment accumulation in rats (Basson et al, 1982) and hamsters (papafrangos and

Lyman 1982). Most recently, Sheehy (l990b) reared crayfish (Cherax cuspidatus) at

three temperatures, finding greater quantities of lipofuscin in the olfactory lobe cell mass

of animals reared at higher temperature.

The following investigation was conducted to address additional questions related

to metabolic and growth-related effects on FAP genesis in nonmitotic tissues, using a

vertebrate aquatic poikilotherm as the study organism. Our primary objectives were: 1) to

determine the degree to which FAP content may vary within and between sibling and

114

non-sibling fish of identical age, and 2) to determine whether body weight, rearing

temperature, ration level, photoperiod, or a combination of these factors, can influence

FAP genesis, ultimately affecting the ability to predict chronological age.

Materials and methods

Model animal

We chose the tropical freshwater fish Puntius conchonius (Cyprinidae), the rosy

barb, as our study organism. ~. conchonius is small, easy to breed, and tolerant to a

wide range of physiological temperatures (15 to 33° C), lending itself well to

physiological studies requiring large numbers of individuals in limited laboratory space.

Clutches used for the following experiments were obtained by breeding adult pairs in 19 I

all-glass aquaria for 8 to 12 h at 25° C. Adults were removed immediately following the

spawning event Larvae hatched within 3 to 4 d after fertilization, and yolk sack

absorption was complete after 3 to 4 d, at which time the larvae were fed a diet of 50-150

um diameter Hatchfry Encapsulon (Argent Chemical Laboratories) for 4 d. Fry were

subsequently fed a combination of larval Anemia salina and crushed flake food until

transfer to experimental aquaria at approximately 45 d of age.

Aquarium system

The experimental system consisted of 18 118 I all-glass closed-system aquaria.

Water quality was maintained using air-driven undergravel filters with 25% of the water

exchanged biweekly. Dissolved oxygen and pH were monitored weekly and temperature

was monitored daily. Aquaria were cooled to 19° C using a thermostatically controlled

refrigerated water bath which circulated chilled water through coiled glass tubing in each

aquarium. All 25° and 31° C aquaria were held at temperature using 100 W thermostatic

heaters. In addition, all 25° C aquaria were equipped with glass coils cooled by tap water

(22 to 23°C) to assure proper temperature on days when room temperature exceeded 25°

115

C. Photoperiod was controlled by enclosingeach aquariumin heavy gauge black plastic

sheeting to excludeexternallight. All aquaria were coveredwith full hoods equipped

with 15 W fluorescentbulbs controlled by timer boxes.

Experimental design

Two independent experimentswere conducted to assess the interactive effectsof

temperature, ration level,photoperiod,clutch, and individualsizeon FAP accumulation

in brain and heart tissues of f. conchonius. In the first, temperature and ration level were

varied while photoperiod was held constant In the secondexperiment, temperature and

photoperiod were variedwhile ration level was held constant. Possible interactions

between photoperiod. andration level were not addressedin this study.

Temperature/ration experiment: This experimentwas designedto test for

interactiveeffects of temperature, ration level, clutch, and individual body size on

variabilityof FAP. At an age of 45 d, three uniparentalclutchesof fry (referred to as CI,

C2, and C3) were randomlydivided into experimentalaquariaat a stockingdensity of 24

fish per aquarium. Temperaturelevels were 19°C± 0.2°C, 25°C± 0.2°C,and 31°C±

D.2°C. Within each temperature, fish were fed commercialflake food (Wardley'sBasic

Flake, The Wardley Corporation) either ill! libitum twice daily (highration, or Hi), or ad

libitum once on alternating days (low ration, or 1..0). This three-temperature, two-ration

design was repeated for each of the three clutches,which were acquiredfrom different

parental sources. All aquariawere held at a constant 12 h light: 12 h dark photoperiod.

Individualswere sacrificed for FAP measurementafter beingheld underexperimental

conditions for 8 mo, at an age of 280 to 290 d. The effects of temperature and body size

on metabolicrate weremeasuredon individualsfrom each of the six treatments in this

experiment Methodsfor oxygen consumptionmeasurement are describedin a later

subsection.

116

Temperature/photoperiod experiment: Interactive effects of temperature,

photoperiod, clutch, and body size on FAP content were tested using a design similar to

the temperature/ration experiment. Individuals were divided into aquaria at 45 d with 24

fish per aquarium. Treatments included three temperature levels (19, 25, and 31°C) and

photoperiod was held at two levels within each temperature; 18 h light: 6 h dark (referred

to as l8L), or 6 h light: 18 h dark (referred to as 6L). All treatments were fed high ration

levels. Three uniparental clutches, referred to as C4, C5, and C6 to avoid confusion with

the previous experiment, were reared under each of the six treatment conditions.

Individuals were held under experimental conditions for 8 mo, and sacrificed at 280 to

290 dofage.

Oxygen consumption

Laboratory determinations of oxygen consumption were made on three randomly

selected f. conchonius from each of the 6 temperature/ration treatments (C3 only)

approximately one month prior to the end of the experiment. Respiration was determined

by measuring changes in oxygen tension in the water medium using a polarographic

electrode. The respiration chamber consisted of a one liter specimen flask connected in

series to another one liter flask containing oxygen and temperature electrodes and a

mixing bar driven by a magnetic stirrer. Oxygen tension was measured directly as mg 02

per liter using an Omega Model PHH-71 dissolved oxygen/temperature probe. Water

medium was recirculated through the two flasks in series using a motorized pump. The

two flasks were held in a water bath whose temperature was maintained using a

submerged glass coil connected to a thermostatically-controlled Lauda Model RM5 water

circulator. Temperature was held at either 19,25, or 31°C ± 0.2°C throughout each

observation.

Triplicate oxygen consumption measurements were made for each individual,

with measurements lasting between 20 and 40 min, depending upon the rate of

117

consumption. Oxygen measurements were only taken within the range of 80 to 100% of

air saturation (Belman and Childress, 1973), and the medium was resaturated between

observations using an air stone. New water medium, sterilized by UV irradiation, was

used for each specimen, and background oxygen depletion due to microbial respiration

was negligible. Observations were made between 0900 and 1500 hrs to account for daily

changes in metabolic rhythms. Specimens were starved for at least 16 h prior to the

observations to minimize specific dynamic action (Saunder, 1963). Fish wet weight and

chamber water volume were measured at the end of the three replicate observations.

Specimen/tissue preparation

Puntius conchonius individuals were anesthetized by ice-bath immersion, and

weighed to the nearest 0.01 g. Upon removal of cranial bones, the brain cavity was

thoroughly rinsed with Ringer's saline solution to remove blood and lipid deposits.

Brain tissues removed for FAP assay included the cerebral, inferior, optic, and cerebellar

lobes, with the medulla being severed posterior to the tenth cranial nerve. Cranial nerves

were severed at their point of attachment to the brain. Hearts were excised and ventricle

myocardial tissues separated and saved for FAP analysis. Brain and myocardial tissues

were briefly rinsed in Ringer's solution, placed in Eppendorf vials, and immediately

frozen in liquid nitrogen until analysis.

FAP extraction and assay

Soluble FAP materials in tissues were quantified following the modified protocols

outlined by Hill and Womersley (1991). Tissues were lyophilized and their dry weights

measured to the nearest 0.0001 mg on a Cahn Model C-32 electronic microbalance.

Tissues were macerated at room temperature for 4 min in a Potter-Elvehjem glass tissue

homogenizer using at least 1 m1of spectroscopic grade chloroform (CHCI3) per 10 mg

dry tissue. The initial meniscus was marked and the volume reconstituted after

homogenization. Homogenates were transferred to glass centrifuge tubes and vortex

118

mixed for 2 min with an equal volume of a solution containing 100 roM magnesium

chloride (MgCI) in glass distilled water and spectroscopic grade methanol at a ratio of 3: 1.

Samples were centrifuged at 0-5°C for 10 min at RCF =3000 x g. Nonpolar and polar

fractions were immediately removed and assayed for FAP content Sample extractions

were performed singly and rapidly to prevent in vitro production ofFAP-like

fluorophores in the sample solvent (Hill and Womersley, 1991).

Extract solvents were placed in quartz-silicate microcuvenes with a 4 mm pathlength

and sealed with a teflon stopper to prevent evaporation. Temperature of the sample

solvent was maintained at 18°C using a thermostatically controlled turret microcell holder

connected to a Brinkmann Lauda Model RM6 refrigerated circulating water bath. A

Perkin Elmer LS-5B luminescence spectrometer, Acer computer, and Perkin Elmer CFS

PC-control software (version 3.0) were used for the fluorometric assay. The LS-5B

luminescence spectrometer was equipped with a Xenon light source and a Hamamatsu

Type R928 red-sensitive photomultiplier tube to reduce possible bias in the 500-700 nm

region (Eldred et al., 1982). Excitation and emission slits were set at 5 and 20 nm,

respectively. Fluorescence spectra reported herein were corrected for the spectral

emission characteristics of the source lamp, the spectral efficiencies of the excitation and

emission monochromators, optics, and the photomultiplier tube using the method of

Rhys-Williarns et aI. (1983) and the emission correction routine in the CPS PC-control

software. All sample readings were expressed as a percentage of Quinine sulfate (0.1

mglllN sulfuric acid) standard luminescence at an excitation setting of 345 nm and an

emission setting of 450 om. Sample FAP concentrations were expressed as whole-organ

%fluorescence (%FL), where:

sample luminescence%FL = x solvent volume (ml) x 100

standard luminescence

119

Statistical analysis

Unpaired two-tailed t-tests were applied to data within treatments to test for

differences in FAP content between sexes. Preliminary examination of the data revealed a

body weight effect on FAP content, thus it was necessary to standardize this effect by

applying Analysis of Covariance (ANCOVA) to the data. ANCOVA was used, whenever

possible, to evaluate the effects and interactions of temperature, ration level or

photoperiod, and clutch on FAP extracted from brain and heart tissues, with body weight

regressed as a covariate. Temperature, ration, photoperiod, and clutch were treated as

class variables, with clutch being random. All FAP data were log-transformed to

homogenize the variances. Computations were made using GLM procedures of the SAS

statistical software (SAS, 1985).

Results

Temperature/ration experiment

Both temperature and ration level had considerable effects on body weight in this

experiment by 280 to 290 d of age (Fig. 4-1a). At low ration level, mean (± 95%

confidence limits) fish weight ranged from 0.79 ± 0.18 g (CI-31°-Lo) to 1.33 ± 0.15 g

(C3-19°-Lo). High ration fish ranged in weight from 1.72 ±0.19 g (CI-19°-Hi) to 3.97

± 0.73 g (C2-31-Hi). Within temperatures, weight was greater in fish fed high rations,

and this effect was accentuated with increasing temperature (Fig. 4-1a). Mean weight

ratios between high and low ration treatments were approximately 2:1 at 19°C, 3:1 at

25°C, and 4: 1 at 31°C. Thus, under low ration conditions, temperature had little or no

effect on final size, whereas temperature had a positive effect on weight at high ration

levels.

Polar and nonpolar FAP extracted from P. conchonius brain tissues had

excitation and emission characteristics similar to those previously reported for this species

120

(Hill and Womersley, 1992). There were no apparent treatment effects on excitation and

emission maxima or the overall shape of the spectra. FAP extracted from brain was much

more polar in solubility with ratios of polar:nonpolar FAP varying great!y (ca. 6:1 to

50:1) within and between clutches and treatments (Fig. 4-2).

Brain polar FAP content varied greatly between all experimental treatments with

mean %FL values ranging from 52.1 ± 12.5 (CI-31°-Hi) to 541.7 ± 98.1 (C2-19°-Lo)

(Fig. 4-2a). The most notable trend was the inverse relationship between rearing

temperature and polar FAP content This trend held for all three clutches and for both

ration levels. There was a tendency for low ration treatments to have higher polar FAP

values than high ration treatments at both 19°C and 31°C in all three clutches (Fig. 4-2a).

The above trends were not apparent for nonpolar FAP extracted from the same brain

tissues (Fig. 4-2b). Brain nonpolar FAP, present in much lower quantity than polar

FAP, varied between clutches both within and between treatments. Within clutches,

nonpolar FAP tended to be higher in both 25°C treatments.

An attempt was made to apply ANCOVA models to the above brain FAP data

The data failed to meet one of the most basic assumptions of ANCOVA, which is a

parallel relationship between the dependent variable (in this case, brain FAP) and the

covariate (fish weight) under the different treatment conditions. Scatter plots of brain

polar FAP as a function of fish weight revealed a negative relationship between the two

variables, and this was evident for most treatments (Fig. 4-3); however, the nature of this

response varied under the different treatment conditions. For example, body weight

appears to have a far greater effect on brain polar FAP at 19°C (Fig. 4-3a) than for

animals of similar size at 31°C (Fig. 4-3c). Because of this, it was not possible to apply

further statistical analyses to these data to determine possible interactive effects of

temperature, ration level, or clutch on brain FAP. While ANCOVA was not feasible, it is

121

clear that brain polar FAP values are inversely correlated with temperature in fish of

similar size.

FAP extracted from heart tissues also had excitation and emission properties

similar to those previously reported for this species (Hill and Womersley, 1992).

Rearing conditions in this experiment did not affect the shape of the luminescence spectra

of either soluble fraction. The solubility of heart FAP in polar and nonpolar solvent

fractions varied between clutches, but was approximately equal or slightly greater in the

polar fraction.

Heart polar FAP varied between treatments and clutches, with %FL values

ranging from 2.44 ± 0.34 (C2-19°-Lo) to 11.77 ± 2.08 (C2-31°-Hi) (Fig. 4-4a). Over all

temperatures, except for C1-19° and C3-19°, FAP values were higher in fish fed high

ration levels. Excluding C3-19°-Lo, temperature had no clear effect on heart polar FAP in

low ration treatments. Temperature did, however, have a positive effect on polar FAP in

high ration animals, especially between 25° and 31°C (Fig. 4-4a). Heart nonpolar FAP

were variable between clutches both within and between treatments with no obvious

consistent trends related to rearing temperature or ration level (Fig. 4-4b).

Attempts to apply ANCOVA to heart FAP data failed for reasons similar to those

stated above for brain data. Plots of heart total FAP as a function of fish weight (Fig. 4­

5) revealed that FAP was negatively correlated with fish weight at 19°C, and positively

correlated at 25° and 31°C.

Metabolic rate, measured as oxygen consumption, varied with both fish weight

and temperature. Within the three temperatures, the weight-specific rate of oxygen

consumption was inversely related to fish weight, presumably due to the greater ratio of

surface area to volume at smaller body sizes (Fig. 4-6). The effect of body weight on

oxygen consumption was stronger at 19°C and 25°C than at 31°C. Temperature showed

the expected positive effect on metabolic rate.

122

TemperaturelphotQperlod experiment

Fish weight was not greatly affected by temperature, photoperiod, or clutch in this

experiment (Fig. 4-1b). Mean (± 95% confidence limits) fish weights ranged from 2.57

±0.28 g (C6-19°-6L) to 4.27 ±0.70 g (C6-31°-18L). There were no apparent

differences in weight between photoperiods within temperatures, and size differences

between temperatures were not as obvious as those for high ration treatments in the

temperature/ration experiment

Brain polar FAP content was inversely related to rearing temperature for all three

clutches, with mean %FL values ranging from 275.1 ±84.6 (C4-19°-6L) to 50.5 ± 16.6

(C5-31°-6L) (Fig. 4-7a). The relation of polar FAP to temperature was similar to that

reported for high ration animals in the temperature/ration experiment (Fig. 4-2a).

Photoperiod did not appear to have a consistent influence on polar FAP content.

Nonpolar FAP, present in much lower quantities than the polar fraction, did not appear to

be affected by either rearing temperature or photoperiod in C5 or C6, but was lower at 19°

and 25°C in C4 (Fig. 4-7b).

Results of the ANCOVA model developed for brain polar FAP data are presented

in Table 4-1. The ability to meet ANCOVA criteria was probably due to the smaller

magnitude of size differences resulting from the experimental treatments. The ANCOVA

test revealed that both temperature (F 2,219 =96.54, P < 0.0001) and clutch (F 2,219 =

203.96, P < 0.00(1), as well as their interactive term (F 4,219 = 147.96, P < 0.0001),

were significant factors affecting polar FAP content in P. conchonius brain (Table 4-1).

Neither fish weight or photoperiod alone significantly affected polar FAP. The interactive

term of photoperiod x clutch was significant (F 2,219 = 7.92, P = 0.0005). Brain

nonpolar FAP data failed to meet ANCOVA assumptions, and a model was not applied.

Heart polar FAP appeared to bepositively correlated to rearing temperature, with

mean %FL values ranging from 7.80 ±1.99 (C6-19°-18L) to 12.33 ± 4.75 (C5-31°-18L)

123

-_.__ ._-_._-_._ ..._--_._ ..__ .- ...

(Fig. 4-8a). Application of an ANCOVA model to the data revealed that fish weight (F

1,219 = 30.60, P < 0.00(1), temperature (F 2,219 = 20.05, P < 0.0001), and clutch (F

2,219 =12.29, P < 0.00(1) all significantly affected polar FAP content, and that

photoperiod had no general effect (F 1,219 =0.01, P =0.9202) (Table 4-2). Significant

interaction terms included temperature x photoperiod, temperature x clutch, and

temperature x photoperiod x clutch, indicating a highly mixed response of heart polar

FAP to each these factors. Application of ANCOVA procedures to heart nonpolar FAP

was not possible, however the most generalized trend was for %FL values to be lowest at

25°C (both photoperiods and all clutches), slightly higher at 19°C, and highest at 31°C

(Fig. 4-8b).

Discussion

The results of the present study provide strong evidence for a relationship

between temperature, body weight, metabolic rate, and FAP genesis in ~. conchonius.

Further, the differences observed between clutches suggest that this is not merely a

physiological effect, but that genetic factors are integrally involved. This is the first study

to demonstrate a temperature effect on FAP accumulation in a vertebrate organism,

confirming the usefulness of fish and other higher order poikilotherms for aging research

(Comfort 1963, Woodhead 1980).

Both temperature and body size showed differential effects on FAP accumulation

in brain and heart tissues. The most striking pattern, repeated for all clutches in both

experiments, was the inverse relationship between ambient temperature and brain polar

FAP. This result is entirely contrary to patterns reported by Sheehy (1990b) for

lipofuscin detected in the olfactory lobe cell mass of crayfish brains. Moreover, our brain

polar FAP data are contradictory to the precept that metabolic rate, free radical production,

and age pigment genesis should be positively correlated with temperature (Sohal 1984).

124

WhileMcArthur and Sohal (1982) havecautioned againstassuming that oxygen

consumption is strictly temperature dependent, our own data clearlyshow thatE.

conchonius has higher ratesof oxygen consumption at higher temperatures. In

comparison, heartdata fromthe sameanimals demonstrates a significant positive

relationship of FAP to ambient temperature, a result that agrees with previously published

age pigmentresearch (Sheehy 1990b).

The fact that differentpatterns of FAP accumulation occurin E. conchoniusbrain

and heart tissueshave led us to suspect the involvement of other unknown

physiologicallbiochemical factors in the genesisofFAP. Thesedifferences are better

understood when the determinants of age pigmentgenesisare considered in terms of the

adaptive physiologyof the organism. It is well established that poikilothermic animals

acclimate to colder ambient temperatures by increasing thedegreeof unsaturation of the

free fattyacids and fatty acidmoieties of phospholipids to maintain membrane

permeability and fluidity (Hazeland Prosser 1974,Cossins et al. 1977, Cossins and

MacDonald 1989,Hazel andWilliams, 1990). Homeoviscous adaptation could

theoretically predispose cold-adapted tissues to be more susceptible to freeradical attack

and peroxidative breakdown, leading to the increasedproduction of cross-linking

reactants in the cytoplasm. In theevent that increased lipid radical generation is not

compensated for by increases in non-enzymatic antioxidants (e.g., beta carotene, alpha

tocopherol, ascorbic acid), higherrates of age pigment formation would result. Thus,

temperature may affectage pigmentgenesis in twodifferentmanners: through increased

metabolism and free radicalproduction at higher temperatures, and by increased lipid

peroxidation potentialat lower temperatures.

Thoughspeculative, theabove theory may partiallyexplain the differences in FAP

accumulation observed in E. conchonius brain and heart tissuesunder the various

experimental conditions. IfFAP genesis is indeeddependenton some balancebetween

125

metabolism and lipid peroxidative potential, then it would be expected that brain (a lipid

rich tissue) would be more susceptible to lipid peroxidation events than heart tissues at

lower temperature. Consequently, if the potential for lipid peroxidation is great enough,

it could mask any positive temperature effects related to metabolism. The converse may

hold true for myocardial tissues, which possess a lower lipid content and whose

metabolic activity is directly affected by temperature (Bowler and Tirri 1990).

An additional factor important to the above considerations is the dependence of

metabolic rate and FAP genesis on body size. Both brain and heart FAP contents varied

with fish weight in this study. The nature of this relationship varied between tissues as

well as between temperatures within tissue type. Respiration measurements also

demonstrated that fish of smaller body size had higher rates of oxygen consumption per

unit of body weight at any given temperature. Higher weight-specific metabolism in

smaller specimens has been widely described in animals including fish (Wieser 1984,

Schmidt-Nielson 1984, Taylor 1987). Using freshwater carp (Cyprinus carpio), Itazawa

and Oikawa (1986) demonstrated a decrease in weight-specific metabolic rate with

increasing size that was due to a relative increase in quantity of tissue of low metabolic

activity (bone and white muscle). Based on this widely accepted relationship between

size and metabolism, it could be predicted that smaller fish would accumulate higher

quantities of FAP than larger fish of identical age, particularly smaller fish living at lower

temperatures where the potential for lipid peroxidation is higher. This is indeed the case

for both brain and heart tissues of f. conchonius. While no previous study has

considered weight-specific metabolic rate as a factor affecting age pigment accumulation

within a given species, comparisons have been made between species. Dogs, which have

a fivefold greater weight-specific rate of oxygen consumption than humans, also

accumulate myocardial lipofuscin five times faster than humans (Munnel and Getty 1968,

Strehler et al. 1959).

126

Restriction of caloric intake has been shown to increase mean life span and

maximum lifespan potential in numerous invertebrate and vertebrate animals (see reviews

by Weindruch 1984, and Finch 1990). There are many potential mechanisms by which

caloric restriction might slow the aging process, yet the precise manner in which it occurs

is still unknown. The broadest interpretation of most caloric restriction studies is that

reduced caloric intake acts to decrease oxidative metabolism, reducing free radical

production and associated molecular damage (Weindruch 1984). In addition, diet

restriction induces synthesis of catalase (Koizumi et al. 1987, Semsei et al. 1989, Rao et

al. 1990), free glutathione (Lang et al, 1990), glutathione peroxidase (Rao et al. 1990)

and superoxide dismutase (Semsei et al. 1989, Rao et al. 1990) in livers of diet-restricted

rodents. Superoxide dismutase catalyzes conversion of the superoxide radical to

hydrogen peroxide, while catalase and glutathione peroxidase are responsible for

converting hydrogen peroxide into oxygen and water, preventing its dissociation into

hydroxide and the highly-damaging hydroxyl radical.

We found no evidence that diet restriction lowers oxygen consumption in E.

conchonius. On the contrary, fish fed low rations had an overall smaller body size and

higher weight-specific rates of oxygen consumption. While fish held on low ration

attained less weight than their high ration counterparts, they were by no means

undernourished or starved. All low ration fish had well developed gonads and abundant

visceral fat deposits by the end of the experiment Thus, even though ANCOVA analysis

of FAP data was not possible for the temperature/ration experiment, we feel that any FAP

differences observed between dietary treatments were due mainly to the differential effects

of body size and the lipid peroxidative potential of the respective tissues rather than

malnourishment stress.

Photoperiod - frequently a determinant of growth and activity level in fishes­

appears to have no significant effect on FAP accumulation in E. conchonius. However,

127

_. __..•_------.. --"'---

our experiments were conducted only through the first quarter of the potential lifespan of

P. conchonius; longer-term effects may occur. This is the first study to test for effects of

photoperiod on accumulation of age pigments, but it is not the first to question the role of

the level of physical activity. Physical activity level has been shown to directly affect age

pigment levels in flies (Sohal and Donato 1979, Sohal 1981) and rodents (Basson et al.

1982, Papafrangos and Lyman 1982). Papafrangos and Lyman (1982) induced

hibernation in Turkish hamsters, slowing the rate of lipofuscin accumulation in the heart

and brain and prolonging their lifespan.

The significant differences in FAP observed between £. conchonius non-siblings

of identical age suggest that genetic factors may be involved in their formation.

Differential expression of genes related to metabolism, production of antioxidant

enzymes, or lysosomal digestive enzymes could all potentially influence the rate of age

pigment production. This is highly speculative with regard to £. conchonius, but the

topic deserves further investigation.

With respect to aging research, the results of this study provide further support

for Strehler's (1959) suggestion that age pigments represent physiological events

associated with the aging process. We have provided evidence that FAP production is

dependent upon temperature, body size, and oxygen consumption, and that the manner in

which these factors influence FAP differs between tissue types. While the

interdependence of these relationships is not completely understood, our results provide

indirect evidence that homeothermic adaptation may affect the susceptibility of cellular

lipid components to peroxidative decomposition, a process which may itself lead to

increased FAP production. This possibility has not previously been considered in studies

on temperature and FAP accumulation in poikilotherms.

The results reported in the present work are also relevant to biologists hoping to

apply FAP methodologies to models for chronological age estimation in wild stocks of

128

-------- -------- ----- --

aquaticpoikilotherms. Hill and Womersley (inreview) examinedthe timecourseofFAP

accumulation in brain and heart tissuesof P. conchonius rangingfrom 60 to 538 d of age.

Brain and heartFAP accumulated with time, but brainFAP was highly variable at any

given age, and less useful than heart FAP for purposes of age prediction. Hill and

Womersley's findings were based on animals that had beenreared in the laboratory under

stable temperature and dietaryregimes. In the presentstudy, the degree of brain and heart

FAP variationmeasured between temperatures, body sizes, and clutches in P. conchonius

was inducedin approximately half the timeframe of Hill and Womersley's (in review)

study and over a relatively short portionof the animals' potentiallifespan (ca.4 yr).

Presumably, thesevariations would increasewith time,and this significantvariability

must be taken into accountin considering the useof any age-predictive model basedon

FAP content The largeeffectsof interactions betweenvariables reportedin the present

studyindicatethat it will be difficult to account for individual effectsof variables and

accurately calibrate age againstFAP contentin wildspecimens. Althoughmanymarine

animalslive underrelatively stable thermal conditions, manyother speciesexperience

ranges of temperature that, according to our results, wouldproducemajor variationsin

FAP content The lifetimehistoryof exposure to temperature or other variables that affect

FAP content is usually unknown for a wildspecimen at the time of aging. Even without

temperature variability, sizevariabilitywithinage classesmay pose a problemif the rate

of FAP accumulation is dependentupon the weight-specific rate of oxygen consumption.

Variability in FAP accumulation mayoccuras a directmetabolic resultof variability in

ration level; suchan effect could not be separated from the effectof growth in our study.

FAP variability may be accentuated in freshwater species in temperate regions, or in

nearshore or migratory marinespecies thatexperience widefluctuations in temperature,

salinity, food availability, pollution (AlojTotaroet al, 1985), or other environmental or

physiological variables. All these factorsmust be given high priority if FAP

129

-_.__.._------_ .. -._-_._ ..__ ...

methodologies are to be used for chronologically aging natural populations of aquatic

animals. Based on the amount of variability we have encountered, the number of

variables involved, and the complexity of the interactions, successful estimation of age of

wild specimens is likely to be very difficult.

130

Literature cited

Aloj Totaro, E., F. A. Pisanti, and P. Glees. 1985. The role of copper level in the

formation of neuronal lipofuscin in the spinal ganglia of Tor,pedo m. Mar.

Environ. Res. 15: 153-163.

Armstrong, D. 1984. Free radical involvement in the formation of lipopigments. pp. 129­

141 in Armstrong, D., R. S. Sohal, R. G. Cutler, and T. F. Slater (eds.) Free

Radicals in Molecular Biology, Aging, and Disease. Raven Press, N.Y.

Basson, A. B. K., S. E. Terblanche, and W. Oelofsen. 1982. A comparative study on

the effects of aging and training on the levels of lipofuscin in various tissues of

the rat. Compo Biochem. PhysioI. 71A: 369-374.

Belman, B. W., and J. J. Childress. 1973. Oxygen consumption of the larvae of the

lobster Panulirus interruptus (Randall) and the crab Cancer productus (Randall).

Compo Biochem. PhysioI. 44A: 821-828.

Berman, M. S., A. L. McVey, and G. Ettershank. 1989. Age determination of Antarctic

krill using fluorescence and image analysis of size. Polar BioI. 9: 267-271.

Bowler, K., and R. Tirri. 1990. Temperature dependence of the heart isolated from the

cold or warm acclimated perch~ fluviatilis). Compo Biochem. PhysioI. 96A:

177-180.

Brown, C. L., and C. 1. George. 1985. Age-dependent accumulation of macrophage

aggregates in the yellow perch, Perea flavescens (Mitchill). J. Fish Diseases 8:

135-138.

Clarke, A., M. A. Kendall, and D. J. Gore. 1990. The accumulation of fluorescent age

pigments in the trochid gastropod Monodonta lineata. J. Exp. Mar. BioI. EcoI.

144: 185-204.

Comfort, A. 1963. Effect of delayed and resumed growth on the longevity of a fish

<Lebistes reticulatus, Peters) in captivity. Gerontol. 8: lSD-ISS.

131

Cossins, A. R., M. J. Friedlander, and C. L. Prosser. 1977. Correlations between

behavioral temperature adaptations of goldfish and the viscosity and fatty acid

composition of their synaptic membranes. J. Compo Physiol, 120: 109-121.

Cossins, A. R., and A. G. Macdonald. 1989. The adaptation of biological membranes to

temperature and pressure: fish from the deep and cold. J. Bioenerg. Biomem. 21:

115-135.

Cutler, R. G. 1991. Antioxidants and aging. Am. J. Clin. Nutrit. 53: 373-378.

Davies, K. J. A. 1988. Protein oxidation, protein cross-linking, and proteolysis in the

formation of lipofuscin. pp. 24-25 in Zs.-Nagy, I. (ed.). Lipofuscin - 1987: State

of the Art. Elsevier Science Publishers, Amsterdam.

Eldred, G. E., and M. L. Katz. 1989. The autofluorescent products of lipid peroxidation

may not be lipofuscin-like. Free Rad. BioI. Med. 7: 157-163.

Eldred, G. E., G. V. Miller, W. S. Stark, and L. Feeney-Burns. 1982. Lipofuscin:

resolution of discrepant fluorescence data. Science 216: 757-759.

Ettershank, G. 1983. Age structure and cyclical annual size change in the Antarctic krill,

Euphausia superba. Polar BioI. 2: 189-193.

Ettershank, G. 1984a Methodology for age determination of Antarctic krill using the age

pigment lipofuscin. BIOMASS Handb. No. 26. 14 pp.

Ettershank, G. 1984b. A new approach to the assessment of longevity in the Antarctic

krill Euphausia superba. J. Crust. BioI. 4: 295-305.

Ettershank, G. 1985. Population age structure in males and juveniles of the Antarctic

krill, Euphausill superba Dana. Polar BioI. 4: 199-201.

Finch, C. E. 1990. Longevity, senescence, and the genome. The University of Chicago

Press, Chicago. 922 pp.

Fletcher, B. L., C. J. Dillard, and A. L. Tappel. 1973. Measurement of fluorescent lipid

. peroxidation products in biological systems and tissues. Anal. Chern. 52: 1-9.

132

Freeman, B. A. 1984. Biological sites and mechanisms of free radical production. pp.

43-52 in Armstrong, D., R. S. Sohal, R. G. Cutler, and T. F. Slater (OOs.) Free

Radicals in Molecular Biology, Aging, and Disease. Raven Press, N.Y.

Fridovich.L 1976. Oxygen radicals, hydrogen peroxide, and oxygen toxicity. pp. 239­

277 in Pryor, W. A. (00.) Free Radicals in Biology. Academic Press, N.Y.

Gutteridge, J. M. C. 1985. Age pigments and free radicals: fluorescent lipid complexes

formed by iron- and copper-containing proteins. Biochim. Biophys. Acta 834:

144-148.

Gutteridge, J. M. C. 1987. Oxygen radicals, transition metals and aging. pp. 1-22 in Aloj

Totaro, E., P. Glees, and F. A. Pisanti (eds.) Advances in Age Pigments

Research. Advances in the Biosciences. Pergamon Press, N.Y.

Hammer, C. 1988. Accumulation of fluorescent age pigments in embryos and larvae of

pike, Esox lucius (Esocidae, Teleostei) at different incubation temperatures.

Environ. BioI. Fish. 22: 91-99.

Hammer, C., and E. Braum. 1988. Quantification of age pigments (lipofuscin): Mini­

Review. Compo Biochem. PhysioI. 90B: 7-17.

Harman, D. 1984. Free radicals and the origination, evolution, and present status of the

free radical theory of aging. pp. 1-12 in Armstrong, D., R. S. Sohal, R. G.

Cutler, and T. F. Slater (eds.) Free Radicals in Molecular Biology, Aging, and

Disease. Raven Press, N.Y.

Harman, D. 1989. Lipofuscin and ceroid formation: the cellular recycling system. pp. 3­

15 in Porta, E. A. (ed.). Lipofuscin and Ceroid Pigments. Plenum Press, N.Y.

Hazel, 1. R., and C. L. Prosser. 1974. Molecular mechanisms of temperature

compensation in poikilothenns. Physiol. Rev. 54: 620-677.

Hill, K. T., and R. L. Radtke. 1988. Gerontological studies of the damselfish, Dascyllus

albisella. Bull. Mar. Sci. 42: 424-434.

133

Hill, K. T., and C. Womersley. 1991. Critical aspects of fluorescent age-pigment

methodologies: modification for accurate analysis and age assessments in aquatic

organisms. Mar. BioI. 109: 1-11.

Hill, K. T., and C. Z. Womersley. Fluorescent age pigment accumulation and the use of

multiple regression models for age estimation in two fish species. J. Exp. Mar.

Biol, Ecol. (in review).

Hirche, H.-J., and K. Anger. 1987. The accumulation of age pigments during larval

development of the spider crab, Hyas araneus (Decapoda, Majidae). Compo

Biochem. Physiol. 88B: 777-782.

Itazawa, Y and S. Oikawa. 1986. A quantitative interpretation of the metabolism-size

relationship in animals. Experentia 42: 152-153.

Koizumi, A, R. Weindruch, and R. L. Walford. 1987. Influences of dietary restriction

and age on liver enzyme activities and lipid peroxidation in mice. 1.Nutrit. 117:

361-367.

Lang, C. A, T. S. Chen, and B. J. Mills. 1990. Glutathione status and longevity are

enhanced by dietary restriction. FASEB J. 4: abstract 601.

Lehane, M. J., J. Chadwick, M. A Howe, and T. S. Mail. 1986. Improvements in the

pteridine method for determining age in adult male and female Stomoxys

calcitrans (Diptera: Muscidae). 1. Econ. Entomol. 79: 1714-1719.

Luo, S.-W., and H. O. Hultin. 1986. Effect of age of winter flounder on some properties

of the sarcoplasmic reticulum. Mech. Age. Dev. 35: 275-289.

Marzabadi, M. R., R. S. Sohal, and U. T. Brunk. 1991. Mechanisms of

lipofuscinogenesis: effect of the inhibition of lysosomal proteases and lipases

under varying conditions of ambient oxygen in cultured rat neonatal myocardial

cells. APMIS 99: 416-426.

134

_.._- ._._-_._--_...--_ .. "_.'- - .

McArthur, M. C., and R. S. Sohal. 1982. Relationship between metabolic rate, aging,

lipid peroxidation, and fluorescent age pigment in milkweed bug, Oncopeltus

fasciatus (Hemiptera). J. Gerontol, 37: 268-274.

Mead, J. F. 1976. Free radical mechanisms of lipid damage and consequences for cellular

membranes. pp. 51-68 in Pryor, W.A. (ed.) Free Radicals in Biology. Academic

Press, N.Y.

Miquel, 1., R. R. Lundgren, K. G. Bensch, and H. H. Atlan. 1976. Effects of

temperature on life span, vitality and fine structure of Drosophila melanogaster.

Mech. Ageing Dev. 5: 347-370.

Miquel, J., D. R. Lundgren, and J. Johnson. 1978. Specrophotofluorometric and

electron microscopic study of lipofuscin accumulation in the testis of aging mice.

J. Gerontol, 33: 5-19.

Mullin, M. M., and E. R. Brooks. 1988. Extractable lipofuscin in larval marine fish.

Fish. BulL,U.S. 86: 407-415.

Munnel, J., and R. Getty. 1968. Rate of accumulation of cardiac lipofuscin in the aging

canine. J. Gerontol, 23: 154-158.

Nicol, S. 1987. Some limitations on the use of the lipofuscin ageing technique. Mar.

BioI. 93: 609-614.

Nicol, S., M. Stolp, and G. W. Hosie. 1991. Accumulation of fluorescent age pigments

in a laboratory population of Antarctic krill Euphausia superba Dana. 1.Exp. Mar.

BioI. £Col. 146: 153-161.

Packer, L., D. W. Deamer, and R. L. Heath. 1967. Regulation and deterioration of

structure in membranes. pp. 77-120 in Strehler, B. L. (ed.) Advances in

Gerontological Research. Academic Press, N.Y.

Papafrangos, E. D., and C. P. Lyman. 1982. Lipofuscin accumulation and hibernation in

Turkish hamster, Mesocricetus brandti. J. Gerontol, 37: 417-421.

135

Porta, E. A. 1991. Advances in age pigment research. Arch. Gerontol. Geriatr. 12: 303­

320.

Rao, G., E. Xia, and A. Richardson. 1990. Effect of dietary restriction and aging on the

expression of antioxidant enzymes. pp. 391-404 in Finch, C. E., and T. E.

Johnson (eds.). Molecular Biology of Aging. Alan R. Liss, Inc. N.Y.

Rhys Williams, A. T., S. A. Winfield, and J. N. Miller. 1983. Relative fluorescence

quantum yields using a computer-controlled luminescence spectrometer. Analyst

108: 1067-1071.

SAS (Statistical Analysis System). 1985. Version 6. A series of user's manuals,

statistical procedures and software for the personal computer. SAS Inst. Inc.,

Cary, N.C.

Saunder, R. L. 1963. Respiration of the Atlantic cod. J. Fish. Res. Bd. Can. 20: 373­

386.

Schmidt-Nielson, K. 1984. Scaling: why is animal size so important? Cambridge:

Cambridge University Press.

Semsei, I., G. Rao, and A. Richardson. 1989. Changes in the expression of superoxide

dismutase and catalase as a function of age and dietary restriction. Biochem.

Biophys. Res. Comm. 164: 620-625.

Sheehy, M. R. J. 1990a. The potential of morphological lipofuscin age pigment as an

index of crustacean age. Mar. BioI. 107: 439-442.

Sheehy, M. R. J. 1990b. Individual variation in, and the effect of rearing temperature and

body size on, the concentration of fluorescent morphological lipofuscin in the

brains of freshwater crayfish, Cherax cuspidatus (Crustacea: Parastacidae).

Compo Biochem. Physioi. 96A: 281-286.

136

._--_._----

Sheehy, M. R. J., and B. E. Roberts. 1991. An alternative explanation for anomalies in

"soluble lipofuscin" fluorescence data from insects, crustaceans, and other aquatic

species. Exp. Gerontol. 26: 495-509.

Sheldahl, J. A., and A. L. Tappel. 1974. Fluorescent products from aging Drosophila

melano~aster:An indicator of free radical lipid peroxidation damage. Exp.

Gerontol. 9: 33-41.

Shimasaki, R., T. Nozawa, O. S. Privett, and W. R. Anderson. 1977. Detection of age­

related fluorescent substances in rat tissues. Arch. Biochem. Biophys. 183: 443­

451.

Shimasaki, R., N. Ueta, and O. S. Privett. 1980. Isolation and analysis of age-related

fluorescent substances in rat testes. Lipids 15: 236-241.

Siakotos, A. N., and N. Koppang. 1973. Procedures for the isolation of lipopigments

from brain, heart and liver, and their properties: A review. Mech. Age. Dev. 2:

177-200.

Sohal, R. S. 1981. Relationship between metabolic rate, lipofuscin accumulation and

lysosomal enzyme activity during aging in the adult housefly, Musca domestica.

Exp. Gerontol. 16: 347-355.

Sohal, R. S. 1984. Metabolic rate, free radicals, and aging. pp. 119-127 in Armstrong,

D., R. S. Sohal, R. G. Cutler, and T. F. Slater (eds.) Free Radicals in Molecular

Biology, Aging, and Disease. Raven Press, N.Y.

Sohal, R. S., and U. T. Brunk. 1989. Lipofuscin as an indicator of oxidative stress and

aging. pp. 17-29 in Porta, E. A. (ed.) Lipofuscin and Ceroid Pigments. Advances

in Experimental Medicine and Biology. Plenum Press, N.Y.

Sohal, R. S., and P. B. Buchan. 1981. Relationship between fluorescent age pigment,

physiological age and physical activity in the housefly, Musca domestica Mech.

.Age. Dev. 15: 243-249.

137

_..__.._------_ .... _---_ ..__ ..

Sohal, R. S., and H. Donato Jr. 1979. Effect of experimental prolongation of life span

on lipofuscin content and lysosomal enzyme activity in the brain of the housefly,

Musca domestica. J. GerontoI. 34: 489-496.

Sohal, R. S., M. R. Marzabadi, D. Galarls, and U. T. Brunk. 1989. Effect of oxygen

concentration on lipofuscin accumulation in cultured rat heart myocytes - a novel

in vitro model of lipofuscinogenesis. Free Rad. BioI. Med. 6: 23-30.

Sohal, R. S., B. H. Sohal, and U. T. Brunk. 1990. Relationship between antioxidant

defenses and longevity in different mammalian species. Mech. Age. Dev. 53:

217-227.

Sohal, R. S., and L. S. Wolfe. 1986. Lipofuscin: characteristics and significance. pp.

171-183 in Swaab, D. F., E. Fliers, M. Mirmiran, W. A. Van GooI, and F. Van

Haaren (eds.) Prog. Brain Res.. Elsevier Science Publishers, Amsterdam.

Strehler, B. L., D. D. Mark, A. S. Mildvan, and M. V. Gee. 1959. Rate and magnitude

of age pigment accumulation in the human myocardium. J. GerontoI. 14: 430­

439.

Tappel, A. L., B. Fletcher, and D. Deamer. 1973. Effect of antioxidants and nutrients on

lipid peroxidation fluorescent products and aging parameters in the mouse. J.

Gerontol. 28: 415-424.

Tappel, A. L. 1975. Lipid peroxidation and fluorescent molecular damage to membranes.

pp. 145-170 in Trump, B. F., and A. V. Arstila (eds.) Pathobiology of Cell

Membranes. Vol. I. Academic Press, N.Y.

Taylor, C. R. 1987. Structural and functional limits to oxidative metabolism: Insights

from scaling. Ann. Rev. Physiol. 49: 135-146.

Tsuchida, M., T. Miura, and K. Aibara. 1987. Lipofuscin and lipofuscin-like

substances. Chem. Phys. Lip. 44: 297-325.

138

Vemet, M., J. R. Hunter, and R. D. Vetter. 1988. Accumulation of age pigments

(lipofuscin) in two cold-water fishes. Fish. Bull.,U.S. 86: 401-407.

Weinclruch, R. 1984. Dietary restriction and the aging process. pp. 181-202 in

Armstrong, D., R. S. Sohal, R. G. Cutler, and T. F. Slater (OOs.) Free Radicals

in Molecular Biology, Aging, and Disease. Raven Press, N.Y.

Wieser, W. 1984. A distinction must be made between the ontogeny and the phylogeny

of metabolism in order to understand the mass exponent of energy metabolism.

Respir. Physiol. 55: 1-9.

Woodhead, A. D. 1980. Fish in studies of aging. Exp. Gerontol. 13: 124-140.

Yin, D., and U. T. Brunk. 1991. Microfluorometric and fluorometric lipofuscin spectral

discrepancies: a concentration-dependent metachromatic effect? Mech. Age. Dev.

59: 95-109.

Ziegler, I., and R. Harmsen. 1969. The biology of pteridines in insects. pp. 139-203 in

Beament, J. W. L., J. E. Treherne, and V. B. Wigglesworth (eds.). Advances in

Insect Physiology. Academic Press, Inc., London.

139

--------------- --------_.-..__ .....

Table 4-1. Results of the ANCOVA model testing the effects of fish weight (ascovariate), temperature (Temp), photoperiod (Phot), and clutch (Clutch) on polar FAPextracted from Puntius conchonius brain tissues. FAP data were log-transformed tostabilize variances. Model r 2 (0.895) significant at P < 0.0001.

Dependentvariable Source df SS MS F p

Brain polar FAP Fish weight 1 0.0002 0.0002 0.04 0.8434

Temp 2 0.8879 0.4440 96.54 <0.0001

Phot 1 0.0017 0.0017 0.37 0.5444

Clutch 2 1.8758 0.9379 203.96 <0.0001

TempxPhot 2 0.0155 0.0078 1.69 0.1869

Temp x Clutch 4 2.7213 0.6803 147.96 <0.0001

Phot x Clutch 2 0.0728 0.0364 7.92 0.0005

Temp x Phot x Clutch 4 0.0171 0.0043 0.93 0.4881

Error 219 1.0071 0.0046

140

Table 4-2. Results of the ANCOVA model testing the effects of fish weight (ascovariate), temperature (Temp), photoperiod (Phot), and clutch (Clutch) on polar FAPextracted from Puntius conchonius hean tissues. FAP data were log-transformed tostabilize variances. Model r 2 (0.532) significant at P < 0.0001.

Dependentvariable Source df SS MS F P

Heart polar FAP Fish weight 1 0.3623 0.3623 30.60 <0.0001

Temp 2 0.4747 0.2373 20.05 <0.0001

Phot 1 0.0001 0.0001 0.01 0.9202

Clutch 2 0.2909 0.1455 12.29 <0.0001

TempxPhot 2 0.0844 0.0422 3.56 0.0300

Temp x Clutch 4 0.2870 0.0718 6.06 <0.0001

Phot x Clutch 2 0.0066 0.0033 0.28 0.7580

Temp x Phot x Clutch 4 0.1864 0.0466 3.94 0.0042

Error 219 2.5926 0.0118

141

--_.._---_ __ _ .

Fig. 4-1. - Puntius conchonius. Mean fish weight ±95% confidence limits at 280 to 290

d of age as affected by experimental treatments: (A) temperature/ration experiment with

low and high rations at each temperature (QC), and (B) temperature/photoperiod

experiment with 6 hand 18 h light (6L and 18L) per 24 h period at each temperature eC).

142

6L 18L31°

6L 18L25°

Treatment

• Clutch 1~ Clutch 2

II1iI Clutch 34

2

o

3

5 A

1

3

2

...-..C)

1'-"+-'..cC) 0-- Low High Low High Low HighQ)

~ 19o 25° 31°5 8 • Clutch 4..c

(/) ~ Clutch 5--LL 4 II Clutch 6

143

Fig. 4-2. - Puntius conchonius. Effects of temperature eC) and ration level (low and

high) on brain FAP (%FL) soluble in (A) polar and (B) nonpolar solvent fractions. Bars

represent means ± 95% confidence limits for Clutches 1, 2, and 3.

144

600 • Clutch 1 A-....J ~ Clutch 2LL 500~ II Clutch 30-o, 400«LL

300"-ro-0 200a.c

"(ij 100"-

CD0

Low High Low High Low High19' 25° 31°

- 80 B....JLL~0- 60o,<LL"-

40ro0a.c0 20ccOro"-co 0

Low High Low High Low High19' 25° 31°

Treatment

145

Fig. 4-3. - Puntius conchonius. Brain polar FAP (%FL) as a function of fish weight for

the temperature/ration experiment, Clutch 1 animals. Solid and open circles represent

fish fed low and high ration levels, respectively.

146

147

Fig. 4-4. - Pontius conchonius. Effects of temperature (0C) and ration level (low and

high) on heart FAP (%FL) soluble in (A) polar and (B) nonpolar solvent fractions. Bars

represent means ±95% confidence limits for Clutches 1, 2, and 3.

148

.-- ._-_ .....__ ..•.._.....

14 A---I 12u,"::R. 100-o,« 8LL~ 6CO-0c.. 41::CO

2Q)

:c0

Low High Low High Low High1~ 25° 31 0

22- B • Clutch 1--I 20u, FIa Clutch 2"::R. 180II Clutch 3- 16o,

« 14u,

12~

CO 10-0c.. 8c:

60c:

41::eu 2Q):c 0

Low High Low High Low High1~ 25° 31°

Treatment

149

Fig. 4-5. - Puntius conchonius. Heart total FAP (%FL) as a function of fish weight for

the temperature/ration experiment, Clutch 1 animals. Solid and open circles represent

fish fed low and high ration levels, respectively.

150

35A 19°C

30 • o High Ration

25 • Low Ration

20 • 0..,..15 ••• • 0 010 ...-. ~ 005 ••• 0 0000 0

CDO 0

0

- 35 8 25°C.....Ju, 0

~30 • 0

0 0 0- 25 0

a. 0 0

« • 0<020 • 0u, 0 0

ctS 15 .: 0cP 0

+-' • 0 00 10 •••• 0+-'

.1 G--.t::= 0

ctS 5 •Q)

I 0

35 C 31°C30 0

25 00

020 0 0

0 0

1500 o 0

• 0 o 0 0

• 010 • • °r) •5 • • •

00 2 3 4 5

Fish weight (9)

151

Fig. 4-6. - Puntius conchonius. Weight-specific rate of oxygen consumption as a

function of fish weight and ambient temperature. Data are means ± 95% confidence

intervals for triplicate determinations on individuals subsampled from Clutch 3,

temperature/ration experiment.

152

.....................:~.........•...... ..·..·..·..·..·..·..·..·!···I 31°I'·'.".{.••......·····

1'119°

3.5

-.c:~

3.0..c.2>Q)

3:.c: 2.5en'+=~0>

2.0E-c:0

:;::::::1.5a.

E::senc: 1.00oc:Q)0> 0.5>0-x0

4.03.51.5 2.0 2.5 3.0

Fish weight (g)1.0

0.0 -I-----..---.-------.-----r---....----.......,.....----.0.5

153

Fig. 4-7. - Puntius conchonius. Effects of temperature (0C) and photoperiod (6L and

18L) on brain FAP (%FL) soluble in (A) polar and (B) nonpolar solvent fractions. Bars

represent means ± 95% confidence limits for Clutches 4, 5, and 6.

154

-_._---------- -------

400 A• Clutch 4--J ~ Clutch 5LL~ 300 II Clutch 60-a..«LL 200~

ctS0a.

100c:"ffi~

CD0

6L 18L 6L 18L 6L 18L1S' 25° 31°

12 B--JLL 10~0-a.. 8<LL~ 6ctS0a. 4c:0c:c: 2

"as~

CD 06L 18L 6L 18L 6L 18L

1S' 25° 31°Treatment

155

Fig. 4-8. - Fundus conchonius. Effects of temperature (0C) and photoperiod (6L and

18L) on heart FAP (%FL) soluble in (A) polar and (B) nonpolar solvent fractions. Bars

represent means ± 95% confidence limits for Clutches 4, 5, and 6.

156

18 A • Clutch 4- 16....J~ Clutch 5LL.

~ 14 II Clutch 60-o, 12

~ 10~

8ctS0 6a.t 4ctSQ)

2J:0

6L 18L 6L 18L 6L 18L1~ 25° 31°

- B....J 6LL.~0 5-o,« 4LL~

ctS 3-0a.c: 20c:t 1ctSQ)

:::c 06L 18L 6L 18L 6L 18L

19" 25° 31°Treatment

157

-----------

CHAPTER 5

SUMMARY AND CONCLUSIONS

This dissertation has examinedthree important aspectsof FAP complexes formed

in postrnitotic tissuesof aquaticpoikilothenns. It providesnew insights relative to

applications in fisheries biology and biomedical gerontology. The primaryobjective of

thisresearch was to critically evaluate the potential for usingFAP methodologies for

accurate estimation of age in fish (andultimately in otheraquaticanimals suchas

mollusksand crustaceans). This goal was accomplished by 1) reassessing various

aspectsof methodologies used to quantifyFAP in animal tissues; 2) establishing the

natureof the relationship between measured levels of FAP and chronological age in two

fish species,and 3) examiningthe effects of several environmental and physiological

variables that can alter the rate ofFAP accumulation and inducevariability in the

relationship to chronological age.

In the first part of this research (Chapter2), it was found that the temperature and

time involvedin storage, incubationand analysis of tissuesand their solventextracts can

criticallyaffectFAP expression throughthe in vitroproduction of additionalFAP-like

fluorophores in tissuesof Oreochromismossambicus. Suchproduction was drastically

increasedif samplehomogenates were ultrasonicated. The finding that FAP-like

fluorophores can form in dead tissuesor samplesolvent is not surprisingin view of the

biochemical processesleading to fluorophore formation. The complex and unstable

biochemical composition of dead tissuesand theirextracts predisposes themto lipid

peroxidation, malondialdehyde production,and fluorophore formation as describedin

Chapters 1 and2. Ulttasonication, commonly applied to disintegratecells, generates free

radicals in solution and expedites these reactions. The fmdingsdescribed in Chapter2

can have profoundimplications in termsof the sources of fluorescencedescribed in all

158

previous FAP studies. In most FAP studieson aquatic animals (Table 2-1), some or all

of these critical sourcesof error have been ignored. This has probably led to artifactual

fluorescenceand high variability in the results of some of these early studies, as well as

erroneous conclusionsas to the usefulnessof the technique. My study did not attempt to

address such problems for the hundreds of pathologicaland gerontological studies

employing FAP methods over the past twenty years, but similaroversights have probably

occurred.

An additionalproblem, common to most of the preliminarystudies of FAP in

aquatic species, has been the use of animals whose chronologicalage was estimated, but

not validated by other methods. In some cases, age was not even estimated, and FAP

content was only compared with body size. Several studieshave attempted to examine

FAP accumulation in whole animalswithout regard for differential rates of accumulation

in different tissue types. The sampling frequency and time scale of many studies have

been inadequatefor accuratelydescribingontogeneticpatternsof FAP accumulationin the

study animals.

In the second phase of this dissertationresearch (Chapter3), the above problems

were addressedby examining patterns of FAP accumulation for extended periodsof time

in specimens of two fish species of known chronologicalage. Nonpolar FAP

accumulated in brain tissues of Oreochromismossambicusduring the first 2.5 years of

life, and it is assumed that accumulationwouldcontinue throughthe latter half of the

lifespan as well. FAP also accumulated in brain and heart tissues of Funtius conchonius

over the first 1.5 years, and the highest values in both tissues were found in isolated

measurementsof the oldest animals sampled (> 4 yearsold). Patterns of accumulation

were different in the two tissue types over the first 1.5 years; brain tissues had greater

and more variablequantitiesof FAP than heart tissues. Brain FAP tended to level off

with increasingage, whereas heart FAP increased in a more linear pattern. Reasons for

159

these differences are not clear; it is assumed that the rates of accumulation are affected by

ontogenetic changes in tissue metabolism as well as changes in the activity of the

antioxidant enzymes responsible for the control of free radicals. Differences were also

found in the relative solubilities of FAP in polar and nonpolar solvents within and

between tissue types, emphasizing the importance of examining both fractions whenever

possible.

This is the first study to incorporate FAP data into multiple regression models for

prediction of chronological age. Of the several types of multiple regression models

developed in Chapter 3, the best models for age prediction in both species were those that

included all types of data (body size, otolith size, organ weight, and FAP). However,

these best models provided only slightly stronger age prediction than similar models

constructed without FAP data. This result raises the question of whether obtaining the

additional information available from FAP measurement would be a cost effective

addition to gathering the fish size, otolith size, and organ weight data; however, a major

goal of FAP research is to use FAP to age animals lacking skeletal structures such as

otoliths. When otolith data were excluded from the multiple regression models in this

study, inclusion ofFAP still provided a better age-predictive model than somatic data

alone. This was especially evident for male P. conchonius, for which organ weight and

FAP data were the only significant variables remaining in the final fitted regression

model. The relationship between FAP content and chronological age was variable in this

study, but FAP variables consistently remained in the [mal multiple regression models.

The animals used for these multiple regression studies were reared under closely similar,

strictly controlled conditions in the laboratory, so the resulting data represent a near

optimum which is probably unattainable with field specimens.

The final goal of this dissertation was to determine the possible effects of some

environmental (temperature, photoperiod) and physiological (body weight, weight-

160

specific metabolic rate, ration level, genome) variables on FAP accumulation in brain and

heart tissues of P. conchonius (Chapter 4). These effects were tested in two experiments:

one tested for effects of temperature, ration level, and clutch; the other tested for effects

of temperature, photoperiod, and clutch. Temperature and body weight affected FAP

content in brain and heart in both experiments, but in a different manner in each organ.

Brain FAP was inversely related to body weight in both experiments, especially at the

lowest rearing temperature (19°C). Heart FAP was inversely related to body weight at

19°C, but positively related to weight at higher temperatures. Increasing temperature

decreased brain FAP in both experiments, but increased heart FAP content Ration level

appeared to affect FAP level by modulating body size, however statistical analyses and

separation of possible direct effects were not possible. Photoperiod had no significant

effect on FAP in either tissue. Significant differences in FAP were detected between

non-sibling fish, implicating genomic variation in the ability to control FAP genesis. The

mixed effects of body size, temperature, and weight-specific oxygen consumption for

each tissue indicated involvement of factors in addition to metabolic rate in the formation

of FAP products. Homeothermic adaptation via modification of cellular lipid constituents

may predispose lipid-rich brain tissues to lipid peroxidation at lower temperatures. In

contrast, heart tissues, which are lower in lipid content, are more susceptible to the

positive effects of temperature on oxygen consumption, and thus have higher FAP values

at higher temperatures. This is the first study to demonstrate these environmental and

physiological effects on FAP accumulation in a vertebrate poikilotherm.

Biologists hoping to apply FAP methodologies for age estimation offield

populations will be faced with a number of serious complications. The ability to apply

FAP as an aging tool will ultimately be contingent on obtaining experimental baseline

information on age, growth, and FAP accumulation patterns which can be related to

animals collected from the natural populations. Measuring FAP fluorescence in a tissue

161

-------_.-- -

is a relatively simple procedure to perform, however interpretation of the data for age

prediction is not nearly as straightforward a procedure as counting growth increments in

skeletal parts. Age pigment content of a tissue, whether measured as extractable FAP or

quantified using histological! morphometric methods, will always be dependent upon the

physiological and environmental life history of the animal. Physiological aging

processes are correlated only roughly with chronological age. Thus, investigators

wishing to apply FAP methods to natural populations will be required to conduct baseline

research defining the relationship between FAP accumulation and chronological age in

combined laboratory growth and mark-recapture studies, through relevant portions of the

animal's lifespan. Laboratory growth studies should take into account variation in FAP

induced by temperature, body size, nutrition and genetic factors, which were shown to be

important in this study. Other factors that may be important to consider (but not studied

here) include disease history and exposure to salinity fluctuations and radiation,

xenobiotics such as heavy metals, or other environmental extremes. Based on the amount

of variability and the complexity of the interactions encountered in this study, successful

estimation of age of wild specimens using FAP is likely to be difficult

162