CHARACTERIZATION OF CARBOMC ANHYDRASE AND ERYTHROCYTE

CHLORIDE/BICARBONATE EXCHANGE IN A P R M l l l W AR-BREATHING

FISH, BOWFIN (AMIA CALVA): IMPLICATIONS FOR CO2 TRANSPORT AND

EXCRETION.

Matthieu Roger Gervais

A thesis submitted to the Department of Biology

in confomity with the requirements for

the degree of Master of Science

Queen's University

Kingston, Ontario, Canada

July, L998

Copyright Q Matthieu Roger Gervais. July 1998

National Library l*l of Canada Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services services bibliographiques

395 Wellington Sueet 395. rue Wellington CXîawaON K 1 A W OttawaON K t A W canada canada

The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Lhrary of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distritbuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfichelfilm, de

reproduction sur papier ou sur format éiectronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.

ABSTRACT

This thesis examined the characteristics of carbonic anhydrase (CA) and the

erythrocyte CI'/HC03- exchanger (Band 3; AE 1) in a primitive air-breathing fish, bowfin

(Amia calva), in order to gain insights into the evolution of CA. CIiHCO,' exchange, and

CO2 transport and excretion in lower vertebrates. Subceiiular fmcuonation and inhibition

experiments reveaied thai, similar to most fish, the vast majority of CA from bowfin giiis

originated from the cytoplasm. In contrast, a significant percentage (27 8) of the total air

bladder CA was associated with the microsornal (membrane) fraction. This membrane-

bound CA in the bowfin air bladder was similar to the mammalian pulmonary CA IV

isozyme in that it had a low sensitivity to CA inhibitors and appeared to be linked to the

membrane via a GPI anchor. A significant amount of CA activity was mzasured in the

blood of bowfin and was restricted to the erythrocytes due to the presence of an

endogenous plasma CA inhibitor. The plasma CA inhibitor was much less effective.

however, against membrane-bound CA in the bowfm air bladder. The turnover number

(kt) of b o w h erythrocyte CA was intermediate between that of the slow type 1

erythrocyte CA isozyme in more primitive fish and that of the fast type II erythrocyte CA

isozyrne in more recent fish. The CA isozyme in bowfin erythrocytes resembled the fast

mammalian CA isozyme (CA II) in tems of its sensitivity to inhibitors. In addition to CA

activity, bowfin erythrocytes were also found to possess a high rate of DIDS-sensitive Cl-

/HC03- exchange. Analysis of bowfin erythrocyte membrane proteins by SDS-PAGE

revealed that bowfin Band 3 probably has a molecular weight ( 140 kDa) that is heavier than

that of the erythrocytes of teleosts and higher vertebrates (93- 10 1 kDa). Furthemore,

bowfin erythrocyte membrane proteins did not react with a polyclonal anti-trout AE 1

antibody. Carbonic anhydrase and the erythrocyte Cl-/HC03- exchanger in bowfin

therefore possess several unique chmcteristics that are likely a consequence of the unique

respiratory mode a d o r primitive phylogenetic position of these fish. Taken together,

however, these results also suggest that the stmtegy for CO2 transport and excretion in

bowfin is probably similar to that of most vertebrates.

ACKNOWLEDGMENTS

1 would f m t like to thank my supenisor and fishing buddy, Bruce Tufts. Much of what I've learned in the pst two years at Queen's 1 attribute to him. His excellent guidance and patience both with experiments and writing, as well as the occasional 'aftemoon fishng trips', have made my Master's experience an enjoyable one.

1 am grateful to p s t and present labrnates who made coming into to lab something 1 looked forward to. In the eady d a y s , the versatile and charrning Suzie Cunie introduced me to graduate life and initiated me into the department. In the summer of 1997, the summer students, Joel Weaver and Kim Schreibert, provided me my daily dose of laughs with the exchange of our witty rernarks. Mr. Fix-it, Bruce Carneron added colour to our lab with his spontaneous visits and stories about his old army days. Finally, 1 would like to thank my most recent labmates, Sue Lund, Matt Phillips, Paul Miller, and my 'understudy', Honour's student Mike Staebler for their genuine friendship. I'm really going to miss our weekly trips to the badminton court.

1 would like to thank my supervisory cornmittee members, Chris Moyes and Me1 Robertson, for their support. 1 also thank Keny Hill for agreeing to be my extemal examiner.

Special thanks goes out to my two families, the Gervais' (mom, Marc, Mélanie, Catherine, and France) and the Goffaux's (Guyleine, Alex, Phillipe, Crystel, and Daniel) for their support and the good times.

FinaiIy, 1 would Iike to thank rny long time friend and now wife, ma belle Véronique. 1 have wanted to acknowledge her contribution to my academic life for a long time. 1 am forever thankful for her unconditional love and encouragement, and for making rny life away from the lab an exceptionally happy one.

This thesis is dedicated to my father, the late Pierre Roger Gervais, who introduced me to the world of nature.

Thank you ail.

M.R.G.

TABLE OF CONTENTS

Page

Abstract

Acknowledgments

Table of Contents

List of Tables

List of Figures

List of Abbreviations

i

iii

iv

v i

v ii

ix

Chapter 1: General Introduction 1

Chapter 2: Subcellular distribution and isozyme characteristics 14

of carbonic anhydrase in the gills and air bladder of bowfin

(Amia calva).

Abstract 14

introduction 15

Materials and Methods 17

Results 20

Discussion 32

Chapter 3: Characterization of carbonic anhydrase and CÏ/HC03- 3 7

exchange in the erythrocytes of bowfin (Arnia calva).

Abstract 37

Introduction 38

Materials and Methods

Results

Discussion

Chapter 4: General Discussion

Literature Cited

Vita

LIST OF TABLES

Page

Table 2.1. Carbonic anhydrase activity of the gills, erythrocytes, and air bladder 2 1 of bowfin.

Table 2.2. Suifanilamide. blood plasma, and acetazolamide inhibition of 28 carbonic anhydrase from the gill and air bladder of bowfin.

Table 3.1. Kinetic properties of mammalian CA 1, CA II, and bowfin erythrocyte 48 CA.

Table 3.2. Acetazolamide (az), CU*, and r inhibition constants of erythrocyte 50 CA from mainmals (CA 1 and CA II) and bowfïn.

LIST OF FIGURES

Figure 1.1. The basic strategy for carbon dioxide m s p o n and excretion in marnrnals.

Figure 2.1. Cellular fractionation of bowfin gill homogenate via differential centrifugation.

Figure 2.2. Cellular fractionation of the bowfin air bladder homogenate via differentid centrifugation.

Figure 2.3. Effect of 0.005 % SDS on CA activity from the cytoplasmic and rnicrosomal fiactions of the gills and air bladder of bowfin.

Figure 2.4. Sulfanilamide inhibition of CA activity from the cytoplasmic and microsomal fractions of the gills and air bladder of bowfin.

Figure 2.5. Effect of blood plasma on CA activity from the cytoplasmic and microsomal fractions of the gills and air bladder of bowfin.

Figure 2.6. Effect of washing by mild sonication on CA activity of the bowfin gill and air bladder microsomal pellets.

Figure 2.7. Effect of phosphatidylinositol-specific phospholipase C (PI-PLC) on the CA activity of the microsomal pellet of the bowfin air bladder.

Figure 3.1. Activity of C.4 in the erythrocytes of bowfin and its inhibition by acetazolamide.

Figure 3.2. Representative trace of mamrnalian CA 1. CA II. and bowfin eiythrocyte CA inhibition by copper and iodide.

Figure 3.3. Effect of saline or bowfin plasma on the activity of CA from bol

erythrocytes.

Figure 3.4. Unidirectional tnnsfer of HC03- across the erythrocyte plasma membranes of bowfin and its inhibiton by DIDS.

Pagz

2

Figure 35. Representative Eassson-Stedman plot of the fractional inhibition 53 of C1-/HC03- exchange in bowfin erythrocytes by DIDS.

Figure 3.6. Visualization of rainbow trout and bowfin erythrocyte membrane 55 polypeptides on a SDS-polyacrylamide gel and Western blot analysis of rainbow trout and bowfin erythrocyte membrane polypeptides.

Figure 4.1. The proposed strategy for CO2 transport and excretion in bowfin. 65

LIST OF ABBREVIATIONS

ABO: air breathing organ Band 3 or AEI: CÏ/HC03- exchanger (anion exchanger) CA: carbonic anhydrase DIDS: 4.4'diisothiocyanos tilbene 2,2'-disulphonic acid &: total concentration of free enzyme GPI: glycosylphosphatidylinositol anchor Hb: hernoglobin i: fractional inhibition of enzyme activity at a given inhibitor concentration 1,: concentration of inhibitor that reduces activity by 50 %

1,: concentration of inhibitor : turnover number constant D a : kilo Daltons K.,: inhibition constant &: Michaelis constant P3: microsornai fraction Pco*: partial pressure of carbon dioxide PI-PLC: phosphatidylinositol-specific phospholipase C @CA: porcine plasma CA inhibitor 0 ~ + : rate of Cl'/HC03- exchange Sj: cytoplasmic fraction SDS-PAGE: sodium dodecyl sulfate - polyacrylamide gel electrophoresis SDS: sodium dodecyl sulfate V : maximal rate of enzyme activity

Chapter 1 : GENERAL INTRODUCTION

1.1 Carbon dioxide transport and excretion in mnmmals

Aerobic metabolism involves both the consurnption of oxygen and production of

carbon dioxide. CO2 is a weak acid, and animals must therefore match its excretion into the

environment with its production at the tissues in order to avoid an acid-base disturbance.

Since much of Our knowledge about how vertebrates transport and excrete CO2 is largely

denved from studies on maznmals. the mammalian strategy for blood CO2 transport and

excretion will be described prior to that of other vertebrates.

In mamrnals, the basic strategy for blood CO2 transport and excretion requires the

cooperation of three erythrocyte proteins: hemoglobin (Xb), carbonic anhydrase (CA), and

a Cl-MC03- exchanger (AH; Band 3). CO2 from the tissues diffuses down its

concentration gradient into the blood plasma and the erythrocytes where it is then

hydrolyzed by the enzyme CA to form HC03- and H+ (Figure 1.1 ). In order to continue the

hydrolysis of CO2 within the erythrocytes and provide a CO2 gradient that favors the

removal of CO2 from the tissues. the products of the reaction (i.e. HC03' and FIf) must be

removed from the cytoplasm of the erythrocytes. This is accomplished mainly by

hemoglobin and the c1-MCO3- exchanger. The majority of HC03- is removed from the ce11

cytoplasm in exchange for plasma Cl- by the passive CÏ/HCO3- exchanger. Most of the

total CO2 traveling from the tissues to the gas exchange surface is thus camied as plasma

HC0,-. The protons formed by the hydration of CO2 are removed from the erythrocyte

cytoplasm through buffenng by Hb. The reaction between these protons and Hb

contributes to the release of O2 by Hb while the blood is passing through the tissues. At the

gas exchmge surface, this cycle is basically reversed in that Hb and the CÏMC03-

exchanger supply the H+ and HCO~- required for the formation of CO2 which diffuses into

the external environment (Figure 1.1).

Figure 1.1. The basic strategy for carbon dioxide transport and excretion in mammais. CO, fmm the metabolizing tissues, which difises down its concentration gradient and into the erythrocytes, is hydrolyzed by carbonic anhydrase (CA) to form HC03- and H+. The protons generated are largely buffered by hemoglobin (Hb) which releases O2 to the tissues. Most of the bicarbonate anions are passively transporteci out of the cell in exchange for plasma chloride by the C1-/HC03- exchanger. Most of the total CO2 transported from the tissues to the lungs is thus carried as plasma HC03-. At the lung the cycle is basically reversed in that Hb and the cI*/HCO3' exchanger supply the H+ and HC03- required for the formation of CO2 which diffuses into the air.

Metabolizing tissues + Lung

Metabolizing tissues + Air

3

1.2 Carbonic anhydrase isozymes in rnammals

Carbonic anhydme is a zinc-containing enzyme (EC 4.2.1.1) which catalyses the

reversible hydratioddehydration of carbon dioxide/bicarbonate in the foilowing reaction:

CO2 + H 2 0 c-----> H+ + HC03' (1)

CA was first discovered in mammalian erythrocytes and up until the early 1960's was

thought to exist as only one isoform (Meldmm and Roughton 1932). It is now known that

CA exists in at least nine different isoforms (isozymes CA 1 to CA IX) and is found in

virtuaiiy aii organisrns, including bacteria, plants, invertebrates and vertebrates (Dodgson et

ai. 1991; Hewett-Emmett and Tashian 1996). It has also been shown that CA'S function in

mammals is not restricted to CO2 transport as was originally thought. Ln marnmals, CA is

present in severai different organ systerns and is involved in a wide variety of physiological

processes including acid-base regulation, ion transport, bone calcification, and muscle

contraction (for reviews see Maren 1967; Sly and Hu 1995; Henry 1996). The different

mammalian CA isozy mes can be identified based on their subceiiular distribution,

susceptibility to certain inhibitors, kinetic activity, and molecular characteristics (Maren and

Sanyai 1983; Sanyal 1984; Henry et al. 1986, 1993, 1997; Hewett-Ernmett and Tashian

1991).

Due to its abundance, CA in the erythrocytes of marnmals is relatively easy to

purify and has consequently k e n the most studied. In most marnmals, erythrocyte CA

exists as two different isozymes, CA 1 and CA II. Some mamrnals including cattle, s h e e ~ ,

dolphins, and cats appear to have only CA II (reviewed by Carter 1972). The main

difference between the two isozymes is that CA 1 is a stow turnover enzyme whereas CA II

is a fast turnover isozyme. CA 1 and CA II have turnover numbers of 30 000 and 250 000

molecules of CO, hydnted sec-' (at temperatures near O°C), respectively (Maren et al.

1980). CA 1, however, has a CO2 substnte affinity about twice that of CA II (Le. lower

4

K,,, than CA II) and is about five times more abundant in mammalian erythrocytes than CA

II (Maren 1967) making it the second most abundant protein in erythrocytes.

In addition to their kinetic activities, CA 1 and II can be differentiated based on their

sensitivities to certain inhibitors. The most cornmon pharmacological inhibitors used to

study CA are a group of unsubstituted aromatic compounds referred to as the sulfonamides

(Maren et al. 196û). One of the most cornrnon sulfonarnide inhibitors used in both

physiologicai and kinetic studies is acetazolamide. Acetazolamide is an effective inhibitor of

CA within red ceIis since it is highly permeable to ceIi membranes (Maren 1967). In

addition, since acetazolamide is about 20 times more potent against CA II than CA 1 (i.e. Ki

for CA 11 is lower), it has proven to be very useful in differentiating between the

erythrocyte isozymes (Maren et al. 1980; Sanyal 1984). Other inhibitors that are more

selective against one of the two erythrocyte isozymes inciude copper (Magid 1967). and

anions such as iodide, cyanide, and chloride which are al1 more effective against CA 1 than

CA II (Sanyd 1984; Maren et al. 1993).

in addition to CA 1 and CA II. a third CA isozyme which has been linked to the

respiratory system is the membrane-bound isozyme located on the mammalian lung, CA

N. It was fmt recognized as a unique isozyme when it was discovered that it was resistant

to sodium dodecyl sulfate (SDS), at a concentration (5 46) that inactivates CA II (Whitney

and Briggle 1982). It is now known that this resistance to SDS is due to the presence of

two disulfide bonds that stabilize the enzyme by linking two sets of cysteine residues

(Waheed et al. 1996). CA IV is anchored to the endothelial cells of the marnmalian lung via

a phosphatidylinositol-glycan linkage (Zhu and Sly 1990; Heming et al. 1993). Since its

active site faces into the blood (Heming et al. 1993). it is able to catalyze CO? reactions in

the plasma. For this reason, it was first thought that CA IV had a major role in CO,

excretion by catalyzing the dehydration of plasma HC03' (Whitney and Briggle 1982).

Subsequent studies measuring CO2 excretion rates while inhibiting CA IV with a

5

membrane impermeable inhibitor (Bidani 199 1 ; S wenson et al. 1993) have revealed,

however, that CA IV's role in CO2 excretion in mammals is only minor (5-8 % of the total

CO2 excreted). This has k e n largely attributed to the fact that there is a lirnited amount of

H+ in blood plasma available to combine with plasma HC03- (Bidani and Heming 199 1;

Heming and Bidani 1992). It is now thought that the main role of CA IV is to maintain an

equilibrium between plasma CO2 and HC03- during transit of the blood through the lung

(Henry et al. 1986). This continuai maintenance of Pco2 and pH at equilibrium values may

be important to provide accurate information to peripheral chemorecepton that influence

ventilation rate (Heming et al. 1993). CA 1 and CA II have also been identified in the

alveolar ceils of the mamrnaiian lung (Ryan et al. 1982; Henry et ai. 1986) and are thought

to play a role in the regulation of pulrnonary surfactant and fluid secretion rather than in

CO, excretion (Fleming et al. 1994).

1.3 The rnammalian plasma carbonic anhydrase inhibitor

The blood plasma of rnany mammals has been shown to inhibit CA released frorn

the erythrocytes (Booth 1938). The presumed reason for this inhibition is to restrict the

catalysis of CO2 reactions to the erytiirocytes, dthough it is not known for certain why this

rnay be necessary for blood CO2 transport. Recently, the CA inhibitor in porcine plasma

(pICA) has been identified as a monomenc transferrin-like protein (Roush and Fierke 1992;

Wuebbens et al. 1997). The concentration of plCA in porcine plasma (= 1 pM) is at least

three ordea of magnitude greater than that required to inhibit half of the activity of CA II

(Ki = 0.2 nM). Any CA II that Ieakes out of erythrocytes is therefore cornplexed with pICA

and inactivated (Roush and Fierke 1992). Interestingly, the pICA K., for porcine CA IV

( 150 nM) is only marginally smaller than the concentration of pICA in porcine plasma

(Roush and Fierke 1992). The porcine plasma CA inhibitor therefore appears to be

effective in abolishing CA II activity arising from lysed erythrocytes while minimaily

affecting vascular CA N in the lung (Heming et al. 1993).

1.4 The erylhrocyte CC/HC03- exchanger in rnammals

Discovered in the mid 19703, the CI-/HC03- exchanger (Band 3) is the most

abundant integral membrane protein in erythrocytes (Steck 1974). The predominant form of

the exchanger in hurnan erythrocytes is a protein dirner (93 kDa monorners) which f o m a

charnel that exchanges chloride for bicarbonate with a stoichiometry of 1 : 1 (Lambert and

Lowe 1978; Knauf 1979; Casey and Reithmeier 1991). Although erythrocyte CA has ken

shown to be an important protein for blood CO2 transport (Swenson and Maren 1978;

Swenson et al. 1993), the rate-limiting step in the excretion of CO2 is thought to be

erythrocyte anion exchange (Perry 1986; Stabenau et al. 1991; Perry and Gilrnour 1993).

Because of the iimited capacity of plasma to cany dissolved CO2, the ability to vansport

most CO2 as plasma HC03-, due to the presence of the CISHC03- exchanger, results in a

five fold increase in the total CO2 carrying capacity of the b l d (Weith et al. 1 982). At

physiological temperatures (38 OC), the maximum Cl-/HC03- exchange activity of human

erythrocytes is 40-50 nmol sec*' cm-2 (Weith et ai. 1982). Although the presence of

erythrocyte c~-/Hco~' exchange has been well documented in mammals, only a lirnited

amount of information is available on the number of Band 3 copies in the erythrocytes of

different rnammalian species. Quantitative analyses of Band 3 copies in marnrnalian

erythrocytes have found that human and llama erythrocytes contain 7000 and 23,000 copies

pd, respectively (Knauf 1979; Khodadad and Weistein 1983). Due to the difference in

the size of llama (43 and human (142 pn2) erythrocytes. however. both have the

same nurnber of Band 3 copies per ce11 ( 1 x 1 0 ~ ) .

1.5 Carbon diuxùie transport and excretion in lower vertebrates

As previously explained, our knowledge of the mechanisms involved in CO,

transport and excretion in vertebrates is based largely on mammalian studies. In contrast,

relatively few studies on lower vertebrates have been carried out. These studies have

revealed, however, that some lower vertebrates do not adhere to the mammalian strategy of

COz transport and excretion. For exarnple, the agnathans, which are among the most

primitive vertebrates, are unique in that their erythrocytes lack a functional anion exchanger

(Nikinmaa and Railo 1987; Ellory et al. 1987; Brill et al. 1992). As a consequence,

agnathans such as lampreys do not carry the majority of HC03- in the blood plasma as do

other vertebrates, but carry it within their erythrocytes (Tu fts and Bou tlier 1 989). Another

interesthg exception is the group of Antarctic icefish which entirely lack erythrocytes and

hemoglobin in their circulation (Ruud 1965). Although circulating CA also appears to be

absent (Feller et al. 198 L), there is some evidence that membrane-bound CA may be

available to catalyze plasma CO2 reactions at the gus of icefish (Tufts, Staebler, and

Gervais, unpublished). Further study is rrquired, however, to determine whether this pool

of CA makes a significant contribution towards CO2 excretion in icefish. A final example is

the mudpuppy, a neotonic salamander in which very low levels of CA are present in both

erythrocytes and plasma (Tufts et al. 1998). Since mudpuppy erythrocytes possess high

levels of Cl-/HCO,- exchange (Tufts et al. 1998). the CO2 transport strategy in these

anirnals may represent another unique situation in which erythrocyte CA activity rather than

Cl-/HC03- exchange is the rate limiting step in blood CO2 transport.

1.6 Erythrocyte c ~ - / H C O ~ ' exchange in lower vertebrates

The erythrocytes of most lower vertebntes c m also carry out C1*/HCO3- exchange

(Obaid et al. 1979; Romano and Passow 1984; Stabenau et al. 199 1 ; Jensen and Brahm

1995). although some differences in the exchanger exist when cornpared to the rnammalian

erythrocyte Cl-/HC03- exchanger. For instance, the rate of Cl-/HC03* exchange in four

teleost fish species was found to be higher than in human erythrocytes at a similar

temperature (Jensen and Brahrn 1995). Similarly, both turtie and trout erythrocytes have

four and eight fold higher numben of Band 3 copies than human and l l m a erythrocytes,

respectively (Romano and Passow 1984; Stabenau et al. t 99 1). Finally, although the

presence of disulfonîc stilbene denvatives such as SITS or DIDS have been shown to

inhibit CI-/'CO,* exchange activity in mammalian erythrocytes, this may not be a cornmon

feaaire arnong lower vertebrates since anion exchange in carp erythrocytes was found to be

unaffecteci by high concentrations of DIDS (Jensen and Brahm 1995).

1.7 The CO-evolution of carbonic anhydrase and C r / E C 0 3 - exchange in f ï h

erythrocytes

Similar to mammals, CA activity occurs in the erythrocytes of al1 fish examined to

date. in contrat to mammais, however, erythrocyte CA in most fish occurs as only one

cytoplasrnic isozyrne rather than two (Carlsson et al. 1980; Maren et al. 1980; Sanyal et al.

1982; Hall and Schraer 1983; Kim et al. 1983). Due to the paucity of information on the

amino acid or nucleotide sequences of erythrocyte CA in fish, fish CA's have not yet been

officiaüy identified as a particular isozyme as have CA 1 to CA M in mammals. Erythrocyte

CA's in a few fish have k e n characterized, however, based on their kinetics and

sensitivities to certain inhibitors. To date, characterization of fish erythrocyte CA has

yielded some interesting patterns which may provide insights into the evolution of CO,

transport in early vertebrates.

In agnathans, erythrocyte CA is a slow isozyme with a turnover number (k,, =

15 000 molecules of C O hydnted sec") that is sirnilar ta that of rnamrnalian CA 1 (Maren

et al. 1980; Henry et al. 1993). A sirnilar low activity CA 1-like isozyme has also k e n

identified in the erythrocytes of elasmobranchs. In addition, elasmobranch erythrocyte CA

has inhibition properties resernbling those of mammaiian CA 1 (Maren et al. 1980). The

erythrocytes of these two primitive groups of fish differ, however, in that functiond Cl-

/HC03- exchange is absent in the agnathans (Nikinmaa and Railo 1987; Ellory et al. 1987;

Brill et ai. 1 992), but present in elasrnobranchs (Obaid et ai. 1979). In contrast to these

primitive fish, the erythrocytes of the more evolved teleosts al1 seem to possess rapid CI'

RICO3' exchange (Romano and Passow 1984; Jensen and Brahm 1995) and CA that

resembles the fast turnover marnrnalian CA II in terms of its kuietic activity and sensitivity

to inhibitors (Maren et al. 1980; Hail and Schraer 1983; Kim et al. 1983; Henry et al.

1993). Based on these observations, a theory regarding the CO-evolution of erythrocyte CA

and the Cl'/HCO,- exchanger in lower vertebrates has k e n proposed (Henry et al. 1993).

This theory suggests that the fmt vertebrate erythrocytes probably contained CA as the

slow isozyme but not rapid C1*/HC03' exchange. This strategy permitted animals such as

agnatlians to increase the total amount of CO2 carried by the blood by rapidly converting it

to HC03- (Henry et al. 1993). The elasmobranchs may represent another stage where the

slow type 1 CA isozyme persists but CI-MCO~' exchange has aiso k e n incorporated into

the erythrocyte membrane thus allowing HC03' to be stored in the plasma Finaily, in the

erythrocytes of the more recently evolved te1eost fish, both a fast type II CA isozyrne and

rapid CI*/HC03- exchange are present (Maren et al. 1980; Henry et al. 1993). Further snidy

is required, however, to explain the possible selection pressures that could have formed

this general evolutionary pattern. In addition, this theory is based on results from a very

lirnited number of species and there is a paucity of information about the charactenstics of

erythrocyte cÏ/HCO~- exchange and CA in many fish groups that are intermediate between

agnathans, elasmobranchs, and teleosts. Exarnples of these fish taxa where the

characteristics of erythrocyte CA and anion exchange have been undescribed include the

holocephali, chondrostei, and holosteans.

10

1.8 Carbonic unhydmse in the gus exchange organs of lower vertebrates

CA activity has been measured in the gasexchange organs of several lower

vertebrates, however, little is known about its subcellular distribution or isozyme type. The

gilis of both fish (Maren 1967; Conley and Mallet 1988; Henry et al. 1988. 1993) and

neotonic amphibians (Toews et al. 1978; Tufts, Gervais, Moss, Henry, unpublished) have

been found to contain high levels of CA activity that is located within the cytoplasm of gill

cells. This large pool of cytopiasmic CA was initiaily thought to be responsible for

catalyzing the conversion of blood plasma HC03- to CO2 which would then diffuse out of

the blood and into the water (Hasweil and Randall 1978). It has more recently been

demonstrated, however, that branchial cytoplasmic CA has no role in cataiyzing plasma

HC03' for CO2 excretion because the gill endothelial membrane is essentially impermeable

to HC03- (Perry et ai. 1982). Instead, the primary role of branchial CA in fish seems to be

in acid-base and ion regulation since it catalyzes a small fraction of the CO, difising from

the blood to HC03- and H+ within the gill ceils and thereby provides counterions for

transporters present on the gill epithelium (Perry and Laurent 1990).

It has k e n recently determined that, in addition to cytoplasmic CA, the gills of

elasmobranchs contain an endothelial membrane-bound CA which faces into the blood

(Swenson et al. 1995; Gilmour et al. 1997; Henry et al. 1997). The role of this membrane-

bound CA is at least in part to maintain CO2 and H+ at equilibrium since its inhibition

results in a disequilibrium of the CO2 reactions in the blood (Gilmour et al. 1997). Since

elasmobranchs are among the only group of fish that have a pronounced p WPco,

ventilatory drive (Graham et al. 1990; Wood et al. 1990). equilibration of CO2 and H+ is

thought to be important for the penpherai chemoreceptors that influence ventilation (Henry

et al. 1997). In addition to the activity onginating from gill membrane-bound CA. plasma

CA activity has also been documented in the b l d of elasmobranchs (Wood et al. 1994;

Henry et al. 1997). The source of this CA. which only represents 0.02 % of the total CA

I l

activity of the blood, is likely from the endogenous lysis of senescent erythrocytes (Henry

et al. 1997). Aithough erythrocyte turnover is a normal occurrence. CA activity in the

plasma of most other fish is absent because of the presence of a plasma inhibitor sirnilar to

that in mammals (Haswell et al. 1983; Dimberg 1994; Henry et al. 1997).

CA activity has also been measured in the primitive lungs or air-breathing organs

(ABO) of several air breathing fish species (Burggren and Haswell 1979; Heming and

Watson 1986). At present, however. no information is available regarding the subcellular

distribution or isozyme srpe of A B 0 CA. Furthemore, the functional significance of CA in

the primitive vertebrate lungs is unknown. Further study on the aspects of CA in the A B 0

of fish is therefore required and could provide important insights into the evolution of the

CA IV isozyrne of the mammalian h g . These studies may also make a vaiuable

contribution toward our understanding of the evolution of air-breathing in vertebrates.

1.9 BowJn: primitive air-breathing fish

The bowfin (Amia cafva), is one of ody two extant holostean fishes that, in

addition to gilis, use a modified swimbladder for facultative aenal respiration. Holosteans

are a group of mostly extinct fish that occupy a phylogenetic position intermediate between

that of elasmobranchs and teleosts. Bowfin are thus arnong the most primitive air-breathing

vertebrates. The primitive phylogenetic position and unique respiratory mode of bowfin

have made it an interesting study subject for both our understanding of the evolution of

aenal respiration and comparative physiology.

Previous studies on the bowfin have already revealed some relatively unique

aspects of their respiration physiology. Herning and Watson (1986) found that the AB0

(air bladder) of bowfin contained about five times the arnount of CA activity that is present

in the swimbladder of a unimodally water-breathing teleost, the trout. On a per gram

protein basis. the activity of the air bladder CA in bowfin is about 20 % of that found in the

erythrocytes (Heming and Watson 1986). This is relatively high in cornparison to

mamrnaiian lung CA activity which is only about 1 % of that in mamrnalian erythrocytes

(Maren 1967). Although relatively high levels of CA activity have k e n reported for the air

bladder of bowfin, nothing is known about its subcellular distribution. The subcellular

distribution of CA is critical for its function since it determines which pool of HC0,- the air

biadder CA will have access to. Furthemore, knowledge about the isozyme characteristics

of air bladder CA in bowfin could provide insight into the evolution of CA in the lungs of

vertebrates, but there is currently no information on the charactenstics of CA isozymes in

early vertebrate ABO's.

Although evidence for the presence of both anion exchange and CA activity in

bowfin erythrocytes has been previously reported (Heming and Watson 1986; Tufts et al.

1994). no information regarding the quantity or characteristics of either is avaiiable. Since

holosteaiis have a phylogeny intermediate between that of elasmobranchs and teleosts,

snidies on the bowfm erythrocyte CA isozyme may provide insight into the transition from

the slow type 1 CA isozyme of agnathans and elasmobranchs to the fast type II CA isozyme

in teleost etythrocytes. In addition, a more detailed examination of the erythrocyte Cl-

/HC03- exchanger in bowfin would also further Our knowledge of the CO-evolution of

erythrocyte CA and the erythrocyte CliHC03- exchanger.

Interestingly, Heming and Watson (1986) found that, in contrast to teleosts, bowfin

lack a plasma CA inhibitor. This result suggests that, similar to elasmobranchs (Henry et

al. 1997). CA activity in bowfin blood rnay not be restricted to the erythrocytes. The

absence of a plasma C A inhibitor in bowfin requires further investigation, however, since

Heming and Watson (1986) suggested that their results may reflect a temporal rather than a

permanent condition.

The purpose of this thesis was. therefore, to study the components of CO2

transport and excretion in a primitive air-breathing fish, the bowfin. More specifically,

13

there were two major objectives that this thesis set out to achieve. The fmt objective was to

determine the subcellular distribution and isozyme characteristics of CA in the air bladder

and gills of bowfin (Chapter 2). It is hypothesized that this fint senes of experiments will

reveai a membrane-bound isozyme in the bowfin ABO. which could k an early

evolutionary fom of the marnmalian CA IV isozyme. The second major objective was to

determine the characteristics of CA and the CI-/HC03- exchanger in bowfin erythrocytes

(Chapter 3). For this second series of experiments. it is hypothesized that the CA and anion

exchanger in bowfin erythrocytes will exhibit unique properties, due to the intemediate

phylogenetic position of these fish, which may fil1 an important gap in our knowledge of

the co-evolution of the important etythrocyte proteins.

14

Chapter 2: Subcellular distribution and isozyme characteristics of carbonic

anhydrase in the gills and air bladder of bowfin (Amia calva).

ABSTRACT

The purpose of this study was to examine the subcellular distribution and isozyrne

characteristics of carbonic anhydrase (CA) from the gills and respiratory air bladder of

bowfin ( h i a calva), a primitive air breathing fish. Separation of subcellular fractions by

differential centrifugation revealed that the vast majority of CA from the gills of bowfh

originated from the cytoplasmic fraction. Washing of the giii microsornal pellet aiso

indicated that the CA onginaiiy associated with this pellet was largely due to contamination

from the cytoplasmic fraction. Experiments with the CA inhibitor, sulfanilamide, and the

plasma CA inhibitor from this species confmed that the bo6n gill probably contains o d y

one CA isozyme which had propertïes resembling those of CA II. In contrast to the

situation in the gus, a relatively large percentap (27 %) of the total air bladder CA was

associated with the microsomal fraction. Washing of the air bladder rnicrosomal pellet

removed little of the CA activity indicating that most of the CA in the rnicrosomal fraction

was associated with the membranes. Like the mammalian pulmonary CA IV isozyme,

rnicrosomal CA from the bowfin air bladder was less sensitive to the bowfin plasma CA

inhibitor, sodium dodecyl sulfate (SDS), and sulfanilamide than was cytoplasmic CA from

the air bladder. Microsomal CA from the bowfin air bladder aiso resembled CA IV in that it

appears to be anchored to the membrane via a phosphatidylinositol-glycan linkage which

could be cleaved by phosphatidylinositol-specific phosphoiipase C. Taken together, these

results suggest that a membrane-bound CA isozyme resembling marnmalian CA IV in terms

of inhibition characteristics and membrane attachment is present in the air breathing organ

of one of the most primitive air breathing vertebrates.

INTRODUCTION

Carbonic anhydrase (CA) is an enzyme which catalyses the revenible

hydrationldehydration of C02/HC03-. In addition to having an important function in the

transport of CO2 within the blood of vertebrates, CA has k e n found to play a significant

role in the homeostatic processes of a wide variety of tissues (Maren 1967; Sly and Hu

1995; Henry 1996). In mammals, nine CA isozymes have been identified in different

tissues and subceiiular fractions (Dodgson 1991, Hewett-Emmett and Tashian 1996).

Among these are cytoplasmic (CA 1, II, III), membrane-bound (CA IV), and mitochondnal

(CA V) isozymes (Dodgson 199 1). Identification of CA isozymes has been accomplished

using a number of experimental approac hes including molecular and kinetic

characterization, subcellular fractionation, and the use of specific inhibitors such as

sulfonarnides (Maren and Sanyal 1983; Sanyd 1984; Henry et al. 1986, 1993, 1997;

Hewett-Emmett and Tashian 199 1). In cornparison to mamrnals, much less is known about

the diversity of CA isozymes and their functions in lower vertebrates.

Both cytoplasmic (CA 1, II) and membrane-bound (CA N) CA have k e n found in

the lungs of mammals (Zhu and Sly 1990; Nioka and Forster 199 1). Pulrnonary CA IV is

located on the extracellular luminal surface of capillary endothelhl cells and is anchored

through a phosphatidylinosito1-glycan linkage (Zhu and Sly 1990; Heming et al. 1993).

Intravascular CA IV functions in establishing an equilibrium between CO2 and plasma

HC03* to maintain postcapillary blood pH (Crandall and O'Brasky 1978; Klocke 1980;

Bidani et al. 1983; Heming et al. 1993) and. to a lirnited extent, in dehydnting plasma

HC03' to CO2 during pulmonary gas exchange (Bidani 1991).

CA is also present in the gas exchange organs of lower vertebrates. High levels of

CA activity occur in the gills of water breathing vertebrates (Maren 1967; Girard and Istin

1975; Toews et al. 1978; Conley and Mallatt 1988; Henry et al. 1988, 1993). In the gills of

16

lampreys and teleost fish, CA appears to be restricted to the cytoplasm where it participates

in acid-base and ionic regulation rather than in C a excretion (Conley and Mallatt 1988;

Henry et al. 1988, 1993; Perry and Laurent 1990). More recently, it was found that the

gilis of the dogfish shark contain both cytoplasmic CA and a membrane-bound CA which

faces into the intravascular lumen (Swenson et al. 1995: Gilmour et al. 1997; Henry et al.

1997). It has b e n proposed that the primary function of membrane-bound CA in dogfish

gilis is probably to equilibrate post branchial CO2 and H? for chemorecepton that influence

ventilation rate (Gilmour et al. 1997; Henry et al. 1997), a role similar to that of CA N in

marnmalian lungs. Significant CA activity has also k e n measured in the air breathing organ

(ABO) of several air breathing fish species (Burggren and Haswell 1979; Heming and

Watson 1986). At present, however, v h a l l y nothing is known about the subceliular

distribution, isozyme characteristics, or function of CA in the ABO's of fish.

Heming and Watson (1986) suggested that the relatively high CA activity in the

A B 0 of bowfh, Amia caiva, may include a pool of intravascular membrane-bound CA, as

in the lungs of higher vertebrates. The bowfin is one of only two extant holostean fishes,

both of which use a modifed swirnbladder for facultative aerial respiration (Johansen et al.

1970; Rahn et ai. 197 1; Randall et al. 198 1). In the swimbladder of a more recent teleost,

the eel, a membrane-bound exiraceiiular CA has recently been found which it is thought to

be involved in buoyancy regulation (Pelster 1995). Several indirect lines of evidence

therefore suggest that the rnodified swimbladder (ABO) of early air breathing species such

as the bowfin may contain an intravascular membrane-bound CA isozyme, which could be

similar to the CA IV isozyme found in rnammaiian lungs. To date, however, this intriguing

possibility has not been further investigated. The main purpose of the present study was

therefore to determine the subcellular distribution and isozyme charactenstics of CA in the

bowfin ABO. To provide a basis for cornparison with the results from the ABO. and also

to provide further information regarding CA in lower vertebrate gills, these characteristics

17

were also exarnined for CA from the gills of these animais. Characterization of CA from

bowfin was accomplished by differential centrifugation to determine subcellular distribution

and by using pharmacological inhibitors as well as a recently discovered plasma inhibitor of

CA from this species (Gervais and Tufts, unpub.; see Chapter 3).

MATERIALS AND METHODS

Animal preparation and tissue collection

Bowfin, Arnia calva, (1 53.0 kg) were collected in the fa11 and spring from the Bay

of Quinte in southeastem Ontario. Fish were held in aerated dechlorinateci freshwater (8-

15°C) and were fed a diet of crayfïsh and dead minnows. There were no visual signs of

stress and no mortalities among the bowfin used in these experiments. Fish were not fed

during the two week period pnor to experimental use.

Individual fish were anesthetized in aerated water containing 250 rngl-' tricaine

methane sulphonate (MS-222; Sigma) buffered with 500 mg? NaHC03. Blood was

collected by blind caudal puncture into a heparinized (40 IU-ml-') synnge and then

transferred to a flask containing heparinized (40 nl-ml-') saline (in mmol-le': 124 NaCl, 10

NaHC03, 5.5 glucose, 5 KCl. 1.1 CaCL, 0.5 MgCl?). Erythrocytes were washed three

times, lysed in 200 volumes of distiIIed water, and then frozen for Iater measurement of CA

activity.

Pnor to harvest. the gills and air bladder of bowfin were perfused with saline to

remove erythrocyte CA. Access to the air bladder was made by a 20 cm mid-ventral

incision. Following cannulation of the nght air bladder vesse1 (from aortic arch VI) with PE

50 tubing. saline (described above) containing 7 rnniol-1-' EDTA at pH 7.8, was perfused

through the vessels of the air bladder with a 10 ml syringe. The left ductus Cuvieri (venous

circulation) was CU< to allow the perfusate (approx. 200 mi in total) to drain out of the air

bladder vessels. The muscle lining the air bladder and the areas that were not perfused

sufficientiy were carefully dissecteci away. The bulbus artenosus was cannulated with PE

200 tubing to perfuse the gills with approximately 200 ml of saline containing 7 mrnol-1-'

EDTA. Pink gill tissue was carefully dissected away to remove any contamination from

erythrocyte CA-

G a and air bln/l/lprfiactiomtion

Approximately 2.5 g of giII and air bladder tissue were added to 8 volumes of cold

bu ffer (described belo w ) and homogenized using a mo tor-drï ven Te flon-glass

homogenizer. The resulting cmde homogenates were analyzed for protein and CA activity

and then subjected to differential centrifugation (Henry et al. 1988; 1993). The fractionation

scherne was as foIlows: 1) low speed (275 g for 20 min, Beckman J2-2 1M centrifuge) to

produce a pellet containing intact ceIls. large ceil fragments, and nuclei; 2) superspeed

( 7 5 0 g for 20 min. Beckrnan J2-2 1M centrifuge) to produce a pellet containing

mitochondria; and 3) ultracentrifugation ( 100 000 g for 90 min, Beckman L8-55M

ultracentrifuge) to produce a microsornal pellet and a cytoplasrnic supernatant.

Centrifugation was always perfonned at 4OC. Pellets were then resuspended in 1- 10 mi of

cold buffer prior to analysis and each fraction was assayed for CA activity. A cytochrome

P450 assay was used to ensure that the 100 000 g pellet was the fraction that contained the

majority of the microsomes.

Memurernent of CA ac t ive and protein concentration

CA activity was measured by the electrometric ApH method (Henry 1991; Henry et

al. 1993). The reaction medium consisted of 10 ml of buffer (in mmol.l*': 225 mannitol, 75

sucrose, 10 TRIS base. adjusted to pH 7.4 with 10 % phosphoric acid) held at 4OC. After

addition of the enzyme source, the reaction was started with the addition of 400 pl of CO2

saturated distilled water (- 1 OC), delivered from a 1OOO pl gas tight Hamilton syringe. The

velocity of the reaction was rnerisured over a change of O. 15 pH units. To obtain the tme

cataiyzed rate in the charnber, the uncatalyzed rate was subtracted from the observeci rate

and the buffer capacity was taken into account to convert from pH units-time*' to mol

H'-tirne*'. pH was rneasured with a Radiometer GK240 1 C combined electrode connected

to a Radiometer PHM64 research pH meter. Protein was measured by the bicinchoninic

acid (BCA) method (Sigma) with bovine semm albumin (Sigma) as a standard.

Inhibition of CA

in order to determine the isozyme charactenstics of microsornai and cytoplasmic

CA, CA activity was measured in the presence of various inhibitors. Sensitivity of CA to

sodium dodecyl sulfate (SDS) was examined by measuring CA activity in the presence of

0.005 % SDS in the reaction chamber. Titrations of CA were also carried out with

increasing concentrations of sulfanilamide and the more potent inhibitor acetazolamide. The

inhibition constant (Ki) for sulfanilamide was calculated according to the method of Dixon

(1953). The inhibition constant for acetazolamide was calculated as the slope of the line

with the following equation:

where Eo is the total concentration of free enzyme in the reaction charnber, fo is the

concentration of inhibitor, and i is the fractional inhibition of enzyme activity at a given

inhibitor concentriütion (Easson and Stedman 1937). Sensitivity of gill and air bladder CA

to the bowfin plasma inhibitor of CA (Gervais and Tufts. unpub.; see Chapter 3) was also

de termined.

20

Washing and clenving

TO determine whether the CA activity observed in the rnicrosornai fractions was due

to membrane-bound CA, rnicrosomal pellets were washed three times to remove any

remaining cytoplasrnic CA. Pellets were resuspended in the reac tion bu ffer, agitated by

mild sonication (5 watts for 3 sec.), and then repelleted by ultracentrifugation as descnbed

above. CA activity of the pellet and supematant was determined after each wash.

After the finai wash, air bladder pellets were resuspended in the reaction buffer

(described above) and incubated (37°C for 90 min) with or without one unit of

phosphatidyhositol-specific phospholipase C (PI-PLC; Sigma) to see whether CA was

anchored to the membrane via a phosphatidyhositol-glycan anchor (Cross 1987; Low et

al. 1988). PI-PLC will cleave phosphatidylinositol-glycan anchored membrane-bound CA

and release it to the supernatant (Zhu and Sly 1990; Bottcher et ai. 1994). One unit of

Phosphatidylinositol-specific ph~spholipase C (PI-PLC) is defined as the amount that will

liberate one unit of acetylchoiinesterase per minute from a membrane-bound crude

preparation at pH 7.4 at 30°C. Following the 90 min incubation penod, the air bladder

sarnples were ultracentrifuged, and the resulting pellet and supematant CA activity were

determined.

RESULTS

Tissue CA activity and distribution

Signifiant levels of CA activity were found in al1 three tissues examined (Table

2.1). Measured per mg of protein, CA activity of the gills was approxirnately 2.5 times

higher than that of the erythrocytes and 14 times higher than that of the air bladder.

Subcellular fractionation of the gilis and air bladder produced different results. The

majority (7 1 %) of gill CA activity was found in the cytoplasmic fraction (S3; Figure 2.1).

Table 2.1. Carbonic anhydrase activity of the gills, erythrocytes, and air bladder of bowfin.

Tissue CA activity pal C a min-lmg protein- 1

Gill

Erythrocytes

Air bladder

Values are means I SE (N=6).

Figure 2.1. Cellular fractionation of bowfm gill hornogenate via differential centrifugation. Total CA activity for each fraction is reported as pmol CO2 min". Values are rnean t SE (N = 6). Percent recovery for each step is indicated in brackets.

Cnide Gill Homogenate

1 1,208 + 2,606

275 g, 20 min

Pl (ce11 debris) A S 1 (cytoplasm, mitochondria, microsomes)

2,005 k 892 7,905 f 1,397

7,500 g, 20 min

P2 (mitochondria) 1 S2 (cytoplasm, microsomes)

160f 51 6,566 + 1 ,O0 1

100,000 g, 90 min

P3 (microsornes) S3 (cytoplasm)

243 f 48 5,9 18 + 967

23

The microsornai fiaction (P3) of the gill contained only 3 % of the total gill activity. In

contrast, CA activity of the air bladder P3 fraction contained 27 %. and the cytoplasmic

fraction (S,) accounted for only 39 96, of the total activity (Figure 2.2). Measured per mg

of protein, the gill S3 and P3 f a t i ons had an activity of 252 and 54 pmol C O - min-' - mg" protein, respectively, whereas air bladder S3 and P3 fractions had an activity of 20

and 92 prnol CO2 - min-' mg-' protein, respectively.

Inhibition of microsorna2 and cytoplasmic CA

The response to 0.005 % SDS was essentially the same for both the gills and air

bladder. Both the gill and air bladder cytoplasrnic CA was sensitive to 0.005 1 SDS, but

the microsoxnai CA activity from these tissues was not significantly inhibited by this

treatment (Figure 2.3). CA from both tissues was aiso sensitive to the inhibitor

sulfanilamide (Figure 2.4). CA from the S3 and P3 fractions of the gills had similar

inhibition constants (Ki), however, CA from the S3 fraction of the bladder was 2.5 times

more sensitive to sulfanilamide than CA from the P3 of the bIadder and both gill fractions

(Table 2.2). CA from the S3 fraction of the bladder was also more sensitive to

acetazolamide than that of the S3 fraction of the gills (Table 2.2). Sensitivity of C A to blood

plasma was similar for both gill fractions, but varied between the bladder P3 and S3

fractions (Figure 2.5). As indicated by their ISo values (Table 2.2)- CA from the bladder P3

fraction was four tirnes less sensitive to plasma than CA from the bladder S3 fraction.

Washing and treatment with PI-PLC

Washing of the rnicrosomal pellets from the gill and air bladder also produced

different results (Figure 2.6). The first wash transferred about hdf of the gill pellet activity

to the supernatant. In contrast, washing of the bladder microsomal pellet did not

Figure 2.2. Cellular fractionation of the bowfin air bladder homogenate via differentiai centrifugation. Total CA activity for each fraction is reportecf as m o l CO2 min-'. Values are mean f SE (N = 6). Percent recovery for each step is indicated in brackets.

Crude BIadder Homogenate

P 1 (ce11 debris) A

S 1

20 min

(cytoplasm, mitochondria, microsomes)

471 +64

7,500 g, 20 min

P3 (microsomes)

146-t- 18 210 + 38

P2 (mitochondria) A S2 (cytoplasm, rnicrosomes)

88 I 13 352 f 32 (93%)

100.000 g, 90 min

Figure 2.3. Effect of 0.005 % SDS on CA activity from the cytoplasmic (S3) and microsomal (P3) fractions of the gus (A) and air bladder (B) of bowfin (means + SE , N = 6). A significant difference from the respective control value is indicated by an asterisk (Paired T-test; P<O.OS).

control

SDS

T [7 control

Figure 2.4. SuIfanilamide inhibition of CA activity from the cytoplasmic (S3) and rnicrosomal (P3) fractions of the gills (A) and air bladder (B) of bowfin (rneans + SE , N = 4).

[Sulfanilamide] (p M)

[Sulfanilamide] (y M)

Figure 2.5. Effect of blood plasma on CA activity from the cytoplasrnic (53) and microsomal (P3) fractions of the gills (A) and air bladder (B) of bo*n (means f SE , N = 4)

O 200 400 600

Volume of plasma (pl)

O 200 400 600

Volume of plasma (pl)

Table 2.2. Sulfanilarnide, blood plasma, and acetazolamide inhibition of carbonic anhydrase from the giil and air bladder of bowfin.

Gill S3 0.41 f 0.03 150 + 32 1.46 + 0.30

Gill P3 0.54 + 0.2 1 246 + 48 -

B ladder S3 O. 18 +_ 0.06 225 +, 16 0.42 4 0.28

B ladder P 0.46 + 0.08 900 k 155 -

Vaiues are mean f SE (N=4). K, Sulf and K, Az are the suifadamide and acetazolamide inhibition constants, respectively. Plasma Im is the approximate volume of plasma required to cause 50% inhibition. S3 and P3 are the cytoplasmic and rnicrosornal fractions, respectively.

Figure 2.6. Effect of washing by mild sonication on CA activity of the bobow gill and air bladder rnicrosomal pellets (means I SE , N = 4). A significant difference from the initial value is indicated by an astensk (Scheffe F-test; P~0.05).

GU Pellet Bladder Pellet

Initial 1 2 3

Wash #

30

significantly reduce the activity of the pellet until the final wash. After the final wash. 79 %

of the initial air bladder CA activity remained in the pellet, whereas only 32 % remained in

the gill pellet At least 83 % of the activity removed from the pellets by the first two washes

was recovered in the supernatant for both the gill and bladder. After the third wash, this

recovery was reduced to 50 %, indicating that the washing procedure itself probably caused

some reduction in total CA activity. Following the washing, air bladder pellets were

incubated with one unit of phosphatidylinositol-specific phospholi pase C (PI-PLC).

Treatment with PI-PLC transferred 84 % of the pellet CA from the air bladder to the

supernatant (Figure 2.7). In control non-treated sarnples, only 13 % of the CA activity was

transferred to the supernatant.

Figure 2.7. Effect of phosphatidylinositol-specific phospholipase C (PI-PLC) on the CA activity of the microsomal pellet of the bowfm air bladder (means + SE , N = 4). A significant difference from the respective control value is indicated by an asterisk (Paired T- test; P<O.OS).

Pellet * O Supernatant

*

Control PI-PLC

DISCUSSION

Substantial carbonic anhydrase (CA) activity was measured in the gills,

erythrocytes, and air bladder of the bowfin. Measured per mg of protein, the CA activity

ratio of bowfin gill:erythrocyte:air bladder found in this study ( 14:6: I . Table 2.1) is within

the range of that reported by Heming and Watson (1986) for bowfin (9:3: 1). The ratio of

bowfin air bladdererythrocyte CA activity (16) is also similar to the CA activity ratio

(AB0:eryttirocytes) of air-breathing teleosts, blue gourami (1:4) and walking catfish (1:s;

Burggren and Haswell 1979). but is about 13 times greater than the rat 1ung:erythrocyte

ratio ( 1 180; Maren 1 967).

SubceUular fractionation of the bowfin air bladder reveaied that the rnicrosomd

fraction, which consists of both the extemai and intemal membranes of cells, contained 27

% of the total CA activity (Figure 2.2). This value is much higher than the percent of CA

activity associated with the microsomal fraction in the gas exchange organs of other

vertebrates. rat lung (7 %; Henry et al. 1986) and gills of dogfish shark (1 %; Henry et al.

1997). Nonetheless, the cytoplasmic fraction of the bowfin air bladder stiil accounted for

the largest portion of the total CA activity (39 %: Figure 2.2). This finding is in agreement

with the results of Henry et al. (1986) where cytoplasrnic CA was found to account for

most of the CA activity in the rat lung (67 %). In the present study. it is possible that some

of the CA in the cytoplasrnic fraction of the air bladder homogenate originated from rninor

erythrocyte contamination of the sampies. If this is indeed the case, it is possible that an

even greater percentage of the total air bladder CA is membrane-bound.

in contrast to the air bladder, the v a t majority of CA activity in the gills of bowfin

appears to originate from the cytosol. Microsomal CA accounted for only 3 B of the total

activity of the gill (Figure 2.1 ). Moreover, while the majority (79 %) of the air bladder

microsomal CA remained in the membrane fraction after washing, less than half (32 %) of

33

the gill microsornai CA remained after washing (Figure 2.6). This indicates that rnost of the

gill rnicrosomal CA was probably only loosely associated with the membranes, rather than

membrane-bound. Thus, based on the relative activi ties of the various fractions, it appears

that little, if any, membrane-bound CA exists in the gilis of bowfin. A similar conclusion

has been reached for the gills of lamprey. trout. and catfish (Henry et al. 1988, 1993.

1997; Rahirn et al. 1988).

in addition to subcellular localization, CA isozymes cm be differentiated on the

bais of their sensitivity to specific CA inhibitors. The inhibition properties of CA isozymes

from the marnmalian lung have been extensively studied (Zhu and Sly 1990; Herning et ai.

1993; Maren et al. 1993; Fleming et al. 1994), but little is known about the types of CA

isozymes that are present in the respiratory organs of lower vertebrates. Another objective

of this study was therefore to examine the CA isozyme characteristics in the giils and air

bladder of bowfin. This kinetic characterization was aiso useful to further evaluate whether

CA associated with the rnicrosomal tissue fraction indeed represented a different CA

isozyme than that present in the cytoplasmic fraction from either the giil or air bladder. On

the basis of past studies, there are several characteristics of mamrnalian pulmonary CA IV

that distinguish it from the cytosolic isozyme CA II. These include the fact that mammalian

CA IV is 1) resistant to denaturation by sodium dodecyl sulfate (SDS). 2) less susceptible

to sulfonamide inhibition. 3) less susceptible to the endogenous plasma CA inhibitor, and

4) partiaily anchored by a phosphatidylinositol glycan linkage (Whitney and Bnggle 1982;

Zhu and Sly 1990; Maren et al. 1993; Herning et al. 1993). Each of these four

characteristics that distinguish the membrane-bound CA N isozyme from mammalian lungs

were investigated for bowfin CA.

The response to 0.005 8 SDS was similar for both the bowfin gill and air bladder

subcellular fractions. Cytoplasrnic CA activity was significantly inhibited in both tissues

upon treatrnent with SDS (Figure 2.3). This result provided further confirmation that CA

34

from the supernatant fraction of both tissues was indeed cytoplasrnic. The fact that

microsornal CA from these tissues was not significantly inhibited by SDS could be viewed

as evidence that both the gill and air bladder microsornai CA, like CA N, are more

stabilized by intramolecular disulfide bonds than is the cytoplasmic CA (Whitney and

Bnggle 1982; Waheed et al. 1996). It is important to note, however, that sensitivity to SDS

is not always effective in differentiating membrane-bound from cytoplasmic CA in non-

mammalian animals (Bottcher et al. 1994).

CA from the microsomal fraction of the bowfin air bladder was three times Iess

sensitive to the sulfonamide inhibitor, sulfadamide, than was cytoplasmic CA (Figure

2.4B, Table 2.2). This result provides further evidence that CA from the microsomal

fraction of the air bladder is a distinct isozyme from cytoplasmic CA. Both mammalian CA

IV and membrane-bound CA of lower vertebrates have dso k e n shown to be relatively

less sensitive to sulfanilamide than cytoplasmic CA (Zhu and Sly 1990; Maren et al. 1993;

Mafia et ai. 1996). In terms of absolute values of Ki, however, the K, for the microsomal

CA of the bowf~n air bladder is considerably lower than that of rnammalian CA IV (Maren

et al. 1993). In contrast to the air bladder, the gill microsomal fraction from the badin had

a sulfanilamide Ki that was similar to that of the cytoplasmic CA (Table 2.2). This result is

therefore consistent with the fractionation experiment which indicated thar the v a t majority

of bowfin gill CA most likely originates from the cytoplasm.

To our knowledge, no previous studies have attempted to characterize the isozyme

properties of CA in the gills of fish. On the basis of the inhibition properties, bowfin giI1

CA most closely resembles the CA II isozyrne of marnmals (Sanyai 1984; Maren et al.

1993). CA from the bowfin gill cytoplasmic fraction was 2 to 3 times less sensitive to

sulfanilamide and acetazolarnide than cytoplasrnic CA frorn the air bladder (Table 2.2).

These results suggest that CA from the cytoplasmic fraction of the gills and air bladder may

aIso be different isozymes, but further investigation is required to confinn fhis.

35

The presence of an endogenous CA inhibitor has been found in the blood plasma of

mammals (Roush and Fierke 1992; Wuebbens et al. 1997) and fish (Haswell et al. 1983;

Dimberg 1994; Henry et al. 1997). Heming et al. ( 1993) noted that rnarnrnalian CA N is

about thirty times less sensitive to the porcine plasma inhibitor as compared to mammalian

cytopiasmic CA II. In contrast to Heming and Watson (1986). we have recentiy found that

bowfin also have a plasma inhibitor (Gervais and Tu&, unpublished data), but in much

lower concentration than that in most other fish examined (Haswell et al. 1983; Dimberg

1994; Henry et ai. 1997). Nonetheless, the sensitivity of bowfin CA to the plasma inhibitor

might therefore be used to differentiate between CA isozymes (Roush and Fierke 1992;

Heming et al. 1993). In bowfin, CA from the giu cytoplasrnic and microsomal fractions,

and the cytoplasrnic fraction of the air bladder, was 4-5 times more sensitive to plasma than

the air bladder rnicrosomal CA (Table 2.2). These results again provide strong evidence

that CA from the rnicrosomal fraction of the bodn air biadder is a different isozyme from

that in the cytoplasm of the air bladder. In contras& our cornbineci results suggest that the

gills appear to have oniy one CA isozyme.

Since the results from our initial experiments provided strong evidence that the

bowfin air bladder contains a membrane-bound CA, we fùrther investigated whether the

enzyme was anchored to the membrane by a glycosylphosphatidylinositol (GPI) anchor

(Zhu and Sly 1990). The results from these experiments indicated that the majority of the

bowfin air bladder rnicrosomal CA appears to be GPI-anchored since treatment with

phosphatidylinositol-specific phospholipase C (PI-PLC) released most of this membrane-

associated CA to the supernatant (Figure 2.7). A similar result was obtained for rnembrane-

bound CA IV from the human lung (Zhu and SIy 1990).

In conclusion. the results of this study indicate the presence of a unique membrane-

bound CA isozyme in the air bladder of bowfin which resembles mammalian CA IV in

tems of its inhibition characteristics and membrane attachent. In contrast, the gills of

36

bowfin appear to contain a single cytoplasmic isozyme which is similar to mammdian CA

II in terms of its inhibition characteristics. Since bowfin are members of the primitive fish

order Holostei and are there fore arnong the rnos t pnmi tive air-breathing vertebrates, these

findings suggest that a CA IV-like isozyme was probably present in some of the earliest

MO'S. The function of this membrane-bound CA isozyme in the bowfin air bladder may

be similar to that of the mammalian lung CA N or, as suggested by Heming and Watson

(1986), it may be important for CO2 excretion across the air bladder during air-breathing.

Conversely, since the bowfin AB0 is a modified swimbladder, it is also possible that the

primary function of this membrane-bound CA isozyme is simply buoyancy regulation, as

suggested for the membrane-bund CA in the swimbladder of a more recent water

breathing teleost, the eel (Pelster 1995). Further study will therefore be necessary in order

to determine the primary hinction and evolutionary significance of membrane-bound CA in

the bowfin ABO.

37

Chapter 3: Characterization of carbonic anhydrase and CÏ/HC03- exchange

in the erythrocytes of bowfin (Amia calva).

ABSTRACT

The purpose of this study was to characterize carbonic anhydrase (CA) and the CÏ/HCO~'

exchanger (Band 3; AEl) in the eiythrocytes of bowfin (Amia calva), a primitive air-

breathing fish, in order to further understand the strategies of blood CO2 transport in lower

vertebrates and gain insights into the evolution of the vertebrate erythrocyte proteins. CA

and Band 3. A ~ i ~ c a n t amount of CA activity was measured in the erythrocytes of

bowfin (70 mm01 CO2 - min*' . ml"), although it appeared to be lower than that in the

erythrocytes of teleost fish. The reason for the lower CA activity can be aîtributed to the

fact that the concentration of CA in bowfin erythrocytes was about half that in erythrocytes

of teleosts such as trout. Furthemore, the turnover number (kt) of bowfin erythrocyte

CA was intermediate between that of the slow type 1 CA isozyme in agnathans and

eIasmobranchs and the fast type II CA in the erythrocytes of the more recent teleost fishes.

The inhibition properties of bowfin erytbrocyte CA were similar to the fast rnammalian CA

isozyme. CA II. In contrast to previous findings. a plasma CA inhibitor was found to be

present in the b l d of bowfin. Bowfin erythrocytes were also found to possess a high rate - I of CI-/KCO~- exchange (6 nrnoi HC03- . sec that was sensitive to DIDS.

Finally, visualization of both trout and bowfin membrane proteins by SDS-PAGE reveaied

a major band at the 100 kDa range for the trout and at the 140 kDa range in the bowfin. In a

Western blot analysis, a polyclonal anti-trout AE 1 antibody reacted strongly with the major

band in the trout lane. but no reaction was observed with the bowfin polypeptides. Taken

together. these results suggest that the strategy for blood CO2 transport in bowfin is

38

probably sirnilar to that in most other vertebrates despite several unique charactenstics of

erythrocyte CA and Band 3 in these primitive fish.

INTRODUCTION

In most vertebrates. blood CO2 transport and excretion follows the same basic

strategy (Cameron 1979; Swenson 1990). CO, from the tissues diffuses down its

concentration gradient into the eryîhrocytes where it becomes catalyzed to HC0,- and H+

by carbonic anhydrase (CA). The protons generated are largely buffered by hernoglobin

whereas most of the bicarbonate anions are passively transported out of the celi in exchange

for plasma chioride. Most of the total CO2 ~ s p o r t e d from the tissues to the gas exchange

organ is thus camied as plasma HCO). At the capillaries of the gas exchange organ, the

cycle is essentially reversed. Two important components of this process are erythrocyte CA

and the erythrocyte Cl-/HC03- exchanger which is also known as Band 3 or AEI. The

cooperative functions of CA and the C1-/HC03- exchanger in blood CO2 transport and

excretion in most vertebrates has lead to the suggestion that their kinetic activities may have

coevolved (Henry et al. 1993). This coevolution hypothesis is derived from studies on

lower vertebrates where variations in the activities of CA and the Cl-/HC03- exchanger

seems to follow an evolutionacy pattern (Henry et al. 1993). At present, however, more

work is needed to confirm this hypothesis since it is based on relatively few species from

only three main fish taxa, the Agnatha, Elasmobranchii, and Teieostei.

In marnmals, at least nine isozymes of CA have k e n identified to date. These

isozymes can be differentiated based on specific activity, subcellular and tissue distribution.

and their sensitivities to certain inhibitors (Maren and Sanyal 1983: Sanyal 1984; Dodgson

1991). In most vertebrates. CA activity in the blood is restricted to the eiythrocytes. The

absence of CA activity in blood plasma is thought to be due to the presence of a

39

endogenous plasma CA inhibitor which may be present to confine any catalysis of HC03-

or CO2 to the erythrocytes (Booth 1938; Van Goor 1948; Roush and Fierke 1992; Henry et

ai. 1997). Bot . a low activity isozyme, CA 1, and a high activity isozyme, CA II, have

been identified in the cytosol of vertebrate eiythrocytes (Maren 1967). In addition to their

activities, CA 1 and II can be differentiated based on the? sensitivities to certain inhibitors

such as acetazolamide, iodide, and copper (Magid 1967; Maren et al. 1980; Sanyal 1984).

Most lower vertebrates examined possess oniy one cytoplasmic CA isozyme in their

erythrocytes (Carlsson et al. 1980; Maren et al. 1980; Sanyal et al. 1982; Hall and Schraer

1983; Kim et al. 1983). In agnathans and elasmobranchs, erythrocyte CA resembles the

slow type 1 isozyrne in both activity and sensitivity to inhibitors (Carlsson et al. 1980;

Maren et al. 1980; Henry et al. 1993). The more recently evolved teleosts, however,

possess a fast type II CA isozyme in their erythrocytes (Maren et al. 1980; Hall and Schraer

1983; Kim et al. 1983; Henry e: al. 1993). Although it is thought that the evolution from

the slow to the fast CA isozyme occurred during the transition from the elasmobranchs to

the teleosts (Maren et al. 1980; Henry et al. 1993), very littie is known about the

characteristics of CA in the erythrocytes of phylogeneticaliy intermediate fish.

Erythrocyte CI-HC03- exchange is generaily thought to be the rate-iimiting step for

blood CO, transport and excretion (Perry 1986; Perry and Gilmour 1993). It has been

shown, however, that a significant amount of variability exists in the characteristics of

anion exchangers arnong vertebrate erythrocytes. For instance, human and turtle 2

erythrocytes contain a sirnilar density of anion exchangers (7000-8000 copies/pm , Knauf

1979; Stabeneau et al. 199 l), whereas llama and trout erythrocytes contain considerably

more with 23,000 and 30,000 copies/pn2, respectively (Khodadad and Weinstein 1983;

Romano and Passow 1984). Among the fish examined, there aiso appears to be a

considerable degree of variability in the rates of anion exchange (Obaid et al. 1979, Jensen

and Brahm 1995; Peny et al. 1996). The situation is most extreme in the erythrocytes of

40

the primitive agnathans which possess very little or no CI-MC03- exchange activity

(Nikinmaa and Rai10 1987; Ellory et al. 1987; Brill et al. 1992). As a result, the blood CO2

transport properties in larnprey differ from those of most vertebrates (Tufts and Boutilier

1989). There is also some evidence that CI'/HC03- exchange in some lower vertebrate

erythrocytes rnay differ in their sensitivity to certain inhibitors. For example, although Cl-

/HC03' exchange in most vertebrate erythrocytes is extremely sensitive to DiDS, this does

not appear to be the case in carp erythrocytes (Jensen and Brahm 1995). Taken together,

many of these results suggest that there may also be significant differences in the molecular

characteristics of the anion exchanger in lower vertebrate erythrocytes, but to date, this

issue has not k e n thoroughly investigated.

The bowfin, Amia calva, is a primitive air-breathing fish which occupies a

phylogenetic position intemediate between that of elasmobranchs and teleosts. Previous

studies indicate that there may be some unique features of the CO2 transport system in these

animals. For example, it has been suggested that CA in the air-breathing organ (ABO) of

may have a role in CO2 excretion across the AB0 (Heming and Watson 1986).

Recent evidence indicating the presence of membrane-bound CA in this organ adds fuaher

support to this intriguing possibility (Gervais and Tufts 1998). In addition, bowh appear

to lack the blood plasma CA inhibitor, which is present in most other fishes examineci

(Haswell et al. 1983; Heming and Watson 1986; Dimberg 1994; Henry et al. 1997).

Evidence for the presence of both anion exchange and CA activity in bowfin erythrocytes

has been previously reported (Heming and Watson 1986; Tufts et al. 1994). At present,

however, virtually nothing is known about the characteristics of either of these erythrocyte

proteins in bowfin. The purpose of this study was therefore to determine the characteristics

of CA and Band 3 in the blwd of bowfin. We hypothesize that bowfin erythrocyte CA and

Band 3 may exhibit unique characteristics reflecting the phylogenetic position of this fish

which is intermediate between that of early vertebrates such as agnathans and

41

elasmobranchs and more recent teleost fish. In view of recent hypotheses regarding the CO-

evolution of CA and Band 3 in vertebrate erythrocytes (Henry at al. 1993). the results of

this study should fil1 an important gap in our understanding of the evolution of these

important erythrocyte proteins.

MATERIALS AND METHODS

Animal preparation and blood collection

Bowfm, Amin calva, (1.5-3.0 kg) were coliected in the fall from the Bay of Quinte

in southeastern Ontario. Fish were held in aerated dechlorinated freshwater (8- 15OC) and

were fed a diet of crayf5sh and dead minnows. There were no visual signs of stress and no

rnoaalities among the bowfm used in these experiments. Fish were not fed during the two

week period prior to experimental use.

Individual fish were anesthetized in aerated water containing 250 mgl-' tricaine

methane sulphonate (MS-222; Sigma) buffered with 500 mgT1 NaHC03. Blood was

collected by blind caudal puncture into a heparinized (40 I U ~ - ' ) syringe and then

transferred to a flask containing heparinized (40 TU-ml-') saline (in mmol-1-': 124 NaCl, 10

NaHC03, 5.5 glucose, 5 KCI, 1.1 CaCL, 0.5 MgC12). Erythrocytes collected for

measurement of CA activity were washed three tirnes in saline, lysed in 200 voIumes of

distilled water, and then frozen for later measurement of CA activity. Erythrocytes collected

for measurement of anion exchange were washed and equilibrated in a chlonde-rich

bicarbonate-free saline (in mmo14-': 124 NaCl, 5 glucose. 6 KCl. 1.5 CaCL, 1 HEPES. 5

EDTA, pH 7.8) prior to measurernent.

Memurement of erythrocyte CA activity

Erythrocyte CA activity was measured by the electrometric ApH method (Heruy

199 1 ; Henry et al. 1993, Gervais and Tufts, 1998). The reaction medium consisted of 10

ml of buffer (in rnmol-lm': 225 mannitol, 75 sucrose, 10 TRIS base, adjusted to pH 7.4

with 10 % phosphoric acid) held at 4OC. After addition of the enzyme source, the reaction

was started by the addition of 400 pl of CO2 saturated distilled water (- 1°C), delivered

frorn a 1 0 pl gas tight Hamilton syringe, for a final concentration of about 2.7 rnM CO2

inside the reaction chamber. The velocity of the reaction was measured over a change of

0.15 pH units. pH was measured with a Radiometer GK2401C combined electrode

connecteci to a Radiometer PHM64 research pH meter.

Inhibition and kinetic analysis of CA

In order to determine the isozyme characteristics of erythrocyte CA, CA activity

was measured in the presence of various inhibitors. Sensitivity of CA to acetazolamide,

copper, and iodide was exarnined by measuring CA activity in the presence of increasing

concentrations of these inhibitors. The inhibition constants (Ki) for copper and iodide were

calculated according to the method of Dixon (1953). Since acetazolamide is a

noncompetitive reversible inhibitor and potendy inhibits CA at nM concentrations, the

inhibition constant was determined by plotting the data on an Easson-Stedrnan plot (Maren

1967). The inhibition constant for acetazolamide was calculated as the dope of the line with

the following equation:

43

where Io is the concentration of inhibitor, i is the fractional inhibition of enzyme activity at a

given inhibitor concentration, and E, is the total concentration of free enzyme in the reaction

chamber (Easson and Stedman 1937). To further characterize erythrocyte CA, Km and

v~ were determineci by measuring the activity against increasing concentrations of CO,

and then plotting the recipmals to obtain a Lineweaver-Burk plot (Maren et al. 1980,

Henry et al. 1993). The active site turnover rate, Gt, was determined from the

relationship Vm/ E, (Maren et al. 1980). E, was also used to denve an estimate of the

concentration of CA in the erythmcytes of bowfin. These kinetic analyses were also

performed on mammalian CA 1 and II (Sigma) in order to provide a bais of cornparison

under identical experimental conditions.

Memurement of erythrocyte Ci/TiCO3- exchange

Erythrocyte Cl-/HC03- exchange was measured according to the method of

Stabenau et al. (199 1). This assay measures the time course of the change in extracellular

pH when Cl'-rich HC03'-free erythrocytes are placed into a ~CO~--r ich Cl--free medium.

The initial change in extracellular pH (dpHJdt) is then used to calculate the rate of HCO?*

uptake by the ceiis (see equation 2 below). The HC03'-rich Cl--free reaction medium

consisted of 10 mi of the test buffer (in rnmol-1-l: 320 sucrose, 2 NaHCO,, 2 HEPES,

0.005 bovine CA (380-520 Wilbur-Anderson units U/ml), pH 7.8) held at lO0C.

Following equilibration of bowfin erythrocytes with the chloride rich saline (see above).

100 pl of packed erythrocytes was introduced into the test buffer to start the reaction. The

initial transfer of acid-base equivalents across the erythrocyte surface, which represents the

rate of Cl-/HC03' exchange (OH+). was rhen calculated with the following equation:

44 -2 - I where 0 ~ + is the initial rate of CI-MCOS' exchange in nmol HC03' - cm - sec , Po is

the buffer capacity in -01 H+ pH unit-' - litef', Hct is the hematocrit of the reaction

chamber (Stabenau et al. 1991). Volume (V; 160.3 p3) and surface area (A; 257 of

the erythrocytes were detennined with a hemocytometer and light microscope, respectively

(Houston 1990, Stabenau et al. 199 1).

Inhibition and quant#ication of er ythrocyte Cl-MC03- exchange

Fractional inhibition of bowfin erythrocyte CI-/HC03- exchange was accomplished

by titrating with increasing concentrations of the potent inhibitor DIDS. To obtain an

estimate of the number of Cl-/HC03- exchangers, an Easson-Stedman plot of 1414) vs Ui

was plotted (Stabenau et al. 199 1) where I is the concentration of DIDS and i is the

fractional inhibition of BK+ calculated using the following two equations:

where ki and k, are the rate coefficients for C1-MCO3- exchange in the presence and

absence of DIDS, respectively, and fHC03-], is the initial concentration of HC03- in the

reaction medium. The i/i intercept of the Easson-Stedman plot, E,, was then used to

calculate the numkr of DIDS binding sites present in bowfin erythrocytes (Stabenau et al.

1991).

SDS-PAGE and Western Blot of erythrocyte membrane proteins

Erythrocyte membrane proteins were subjected to SDS-PAGE and Western blot

analysis in order to visually compare the Cl-MC03- exchanger in the bowfin with that in a

RESULTS

Erythrocyte CA activiiy and inhibition

Significant levels of CA activity were present in the lysate of bowfin erythrocytes

(Figure 3.1). Taking the lysate dilution into account. bowfin erydirocytes have a total CA

activity of about 70 m o l CO2 - min-' - ml erythrocyte<'. Based on an erythrocyte 3

volume of 160.3 prn . this corresponds to a CA activity of about 1.04 x IO-' ' mol CO2 - - 1 min erythrocyte". This apparent CA activity in the lysate was also inhibitable by the

specific CA inhibitor acetazolamide in the nM concentration range (Figure 3.1). Using E,

from the Easson-Stedman plot (see methods), the concentration of CA in the erythrocytes

of bowfin was estimated to be 0.66 f: O. 12 mM.

Figure 3.1. Activity of CA in the erythrocytes of bowfin and its inhibition by acetazolarnide. CA activity is in pmol CO2 . min-' ml lysate". Values are rnean I SE (N = 8).

47

Bowfin erythrocyte CA had a substrate affmity (K,,,) which was relatively close to

that of mammaiian CA II and was about four times higher than that of marnmalian CA 1

(Table 3.1). Erythrocyte CA from bowfin was also more similar to manundian CA II than

CA 1 in terms of its sensitivity to both copper and iodide (Figure 3.2 A & B). This is most

apparent when the copper and iodide inhibitor constant values (6) of bowfin CA are

compared with those for mammalian CA 1 and II (Table 3.2). The acetazolarnide inhibition

constant of bowfin CA was also very similar to that of marnmalian CA II (Table 3.2).

Erythrocyte CA from bowfin did, however, differ from both marnmalian CA 1 and II in

t e m of its specific turnover rate (K&J which was intexmediate between that of CA 1 and II

(Table 3.1 ).

Although CA activity was detected in bowfin erythrocytes, it does not appear to be

present in the blood plasma of this species (Figure 3.3). Bowfin erythrocyte CA was

inhibited by bowfin plasma whereas erythrocyte CA treated with saline was not

significantiy inhibited (Figure 3.3). The volume of plasma required to inhibit the bodn

erythrocyte CA activity in this expenment by 50 95 (ISo) was about 300 ~ L L

Quantification of er ythrocyte CZ'L~CO~' exchunge and Western blot analysis

Bowfh erythrocytes were found to possess a significant amount of C1-/HC03-

exchange (Figure 3.4). The uninhibited control value of C1-/HC03- exchange (OH+) was 6 - I

nmol HC03- sec . cm-2 membrane surface area. Based on an erythrocyte surface area 2 of 257 prn , this corresponds to an anion exchange activity of about 9.25 x 10-13 mol

- 1 HC03- - min - erythrocyte-'. Bowfin erythrocyte anion exchange was also found to be

sensitive to the specific CÏ/HCO3- exchanger inhibitor DIDS (Figure 3.4). Plotting of the

inhibition data on an Easson-Stedrnan plot (Figure 3.5). indicated that the average

concentration of DIDS binding sites (or CÏ/HC03' exchangers) present in the reaction

chamber (Eo) was found to be 6.4 p M (Figure 3.5). This yielded a density of CI-/HC03-

Table 3.1. Kinetic properties of mammalian CA 1, CA II, and bowfin erythrocyte CA.

CA 1 1.8 + 0.1

CA II 5.9 + 0.1

Bowfin 8.4 + 1.5

Values are mean f SE (N = 8 for bowfin and N = 3 for CA I and CA II).

Figure 3.2. Representative trace of rnammalian CA 1, CA II, and bowfin erythrocyte CA inhibition by copper (A) and iodide (E3).

100 CA11 I A Bowfin

n

# A I w

A Y .M

0 A

I= > -4 u

2 a 40 I

6 U

P o I O

20 .A I A 0

0 CA11

A Bowfin

Table 3.2. Acetazolamide (a). Cu*. and ï inhibition constants of erythrocyte CA fiom mamrnals (CA 1 and CA II) and bowfin-

CA 1 30.1 I 11.3 42.5 f 13.9 0.3 4 O. 1

CA II 1.7 f 0.4 0-6 + 0.3 10.4 f: 1.5

Bowfin 1.6 ': 0.2 0.4 k 0.2 5.6 t 0.7

Values are mean f SE (N = 8 for bowfin and N = 3 for CA 1 and CA Il).

Figure 3.3. EKect of saline or bowfm plasma on the activity of CA from bowfin erythrocytes. CA activity is in pmol CO2 - min-' . ml lysate-'. Values are rnean t SE (N =

4)

O saline

plasma

volume added (pl)

Figure 3.4. Unidirectional transfer of HCO, across the erythrocyte plasma membranes of bowfin and its inhibiton by DIDS. Bicarbonate transfer is in nmol HC03- cm 2

- 1 erythrocyte membrane surface ara-' - sec . Values are rnean IT SE (N = 8).

Figure 3.5. Representative Easson-Stedman plot of the fractionai inhibition (i) of Cl- /HCO3- exchange in bowfin erythrocytes by DIDS.

54 7 2

exchangers of about 6.8 x 10 copies per ce11 or 265,000 copies per pn of erythrocyte

membrane surface.

Electrophoretic sepmtion and staining of trout erythrocyte membrane proteins

resulted in a broad diffuse band at about the 100 kDa range (Figure 3.6A). Western blot

andysis of the trout membrane proteins revealed a high degree of reactivity with the anti-

trout AE I antibody and the reaction was strongest for the 100 kDa band (Figure 3.6B).

Electrophoretic separation and staining of bowf5n erythrocyte membrane proteins also

y ielded a diffuse band, but at a higher molecular weight ( 140 kDa) than that in the trout

(Figure 3.6A). No significant reaction of the bowfin membrane proteins with the anti-trout

AE 1 antibody was apparent (Figure 3 -6B). The absence of a reaction between the anti-trout

AEl antibody and the bowfin membrane proteins was not likely due to differences in the

arnount of protein available between species since twice-as rnuch total protein (5 pg) was

loaded in the bowfin lane, as compared to the trout lane.

Figure 3.6. Visualization of rainbow trout (Tr) and bowfin (BQ erythrocyte membrane polypeptides on a 7.5% SDS-polyacrylamide gel stained with Coomasie blue (A). Western blot andysis (B) of rainbow trout (Tr) and bowfin (Bf) erythrocyte membrane polypeptides, using rabbit anti-trout AE 1 polyclonal antibodies (Cameron et al. 1 996). The sizes of the markers (lane 1) are given in kDa.

DISCUSSION

As reported by Heming and Watson (1986), the erythrocytes of bowfin were found

to contain a significant amount of CA activity that was sensitive to the CA inhibitor

acetazolamide. Among lower vertebrates, few numbers are available for cornparison of CA

kinetic values. Moreover, c o m p ~ s o n of erythrocyte CA activity rates between studies is

difficult since these numbers may be influenced by specific experimental conditions which

Vary between studies. Under similar assay conditions, the total CA activity of bowfin

erythrocytes (70 mm01 CO2 min-' - ml-') was about 30 1 of the CA activity in the

erythrocytes of the more recent teleost, the rainbow trout (Gervais and Tufis unpubiished).

In some non-mammalian vertebrates, the presence of a reducing agent is required to reduce

the formation of disulfide bonds and maintain the activity of CA (Henry et al. 1993, Kim et

ai. 1983). The activity of bowfin erythrocyte CA was not likely affected by the formation

of disulfide bonds, however, since assaying in the presence of the reducing agent (5 mM

D m had no effect on CA activity. The relatively low activity of CA in bowfin eryttirocytes

can be attribut4 in part to the fact that the bowfin CA turnover number was lower than tha.t

of teleost erythrocyte CA (Maren et al. 1980; Maren and Friedland 1978) and marnmalian

CA II (Table 3.1). Interestingly, the turnover number of bowfin erythrocyte CA is

intermediate between the high activity CA of teleosts (Maren and Friedland 1978; Maren et

al. 1980; Sanyal et al. 1982) and the low activity CA of elasmobranchs and agnathans

(Maren et al. 1980; Henry et a.. 1993)- In addition to a slow erythrocyte CA turnover

number, the concentration of CA in bowfin erythrocytes was about half that in more recent

teleosts such as trout (1.1 + 0.4 1 mM; Gervais and Tufts, unpublished). The concentration

of substnte required to reach half of V,, (Km) of bowfin erythrocyte CA was Found to be

higher than that for both marnrnalian CA 1 and II (Table 3.1 ), but was also within the range

reported for other fish (Maren et al. 1980; Sanyal et al. 1982; Henry et al. 1993).

57

Aithough bowfin erythrocyte CA differed from both mammalian CA I and LI in

ternis of substrate affinity and specific activity (Table 3.1). it had inhibition charactenstics

that were similar to mammalian CA II (Figure 3.2, Table 3.2) and rainbow trout CA

(Henry et al. 1993, Gervais and Tufts, unpubiished). Based on sensitivities to certain

inhibitors, bowfïn erythrocyte CA rnay therefore have a structure which is very similar in

some aspects to rnarnmalian CA II. Interestingly, however, bowfm CA lacks the fast

catalytic activity that is thought to have arisen with the evolution of the teleosts (Maren et al.

1980; Henry et al. 1993). Thus, other aspects of the structure of bowfin CA rnay be more

sirniiar to that of the slower erytfirocyte CA isozyme. Similarly. Singer and Ballantyne

(1990) found that metabolic enzyme activities of bowfin were dso lower than those of most

teleosts. Given the interesting kinetic properties of bowfin CA, which seem to be

intermediate between those of rnammalian CA 1 and II, molecular studies on the CA

enzyme from this species rnay provide valuable insight into the evolution of CA in

vertebrates.

Although a significant amount of CA activity was measured in the erythrocytes of

bowfin, no CA activity was rneasured in the plasma. Moreover, in contrat to previous

results of Heming and Watson (1986), bowfin plasma was actually found to contain a CA

inhibitor (Figure 3.3). A plasma CA iiïkLvitcr W ~ S k e n found in the blood of several

teleosts, but in considerably higher concentrations than that in the bowfin (Haswell et al.

1983; Dirnberg 1994; Henry et al. 1997). Heming and Watson (1986) suggested that the

absence of the plasma CA inhibitor observed in their study of the bowfm may be a temporal

rather than a permanent condition. In order to test this idea, we examined whether the

activity of this inhibitor in bowfin was sirnilar under two different acclimation conditions.

Since we found that plasma from bowfin acclimated at both 5°C and 15°C contained a CA

inhibitor with sirnilar inhibitory activity, we cm at least conclude that the CA inhibitor in

bowfin is not affected by acclimation temperature and therefore is probably not largely

58

affected by season. The plasma inhibitor probably functions to inhibit CA released from the

normal lysis of erythrocytes (Booth 1938; Van Goor 1948; Roush and Fierke 1992)-

Interestingly, Gervais and Tufts (1998) hund that membrane-bound CA in the bowfin air-

breathing organ (ABO) was about three times less sensitive to the plasma CA inhibitor than

erythrocyte CA. As proposed for the mammalian plasma CA inhibitor (Heming et al.

1993). the bowfin plasma inhibitor may therefore be effective in scavenging and inhibiting

cytoplasmic CA released from lysed erythrocytes while minimaliy affecting membrane-

bound CA present in the B O .

Aithough bowfh erythrocytes possess less CA activity than the more recent

teleosts, the rate Limitiag step in the uptake and removal of CO2 in the blood is thought to

be the rate of anion exchange (Perry 1986; Perry and Gilmour 1993). Given this and the

fact that rates of anion exchange are quite variable in lower vertebrates. we also attempted

to quanti@ the rate of anion exchange in bowfin erythrocytes using the approach of

Stabenau et al. (1991). As found by Tu& et al. (1992). anion exchange is functionally

present in bodn erythrocytes and is sensitive to the specific inhibitor DIDS (Figure 3.4).

In the present study, we further determined that bowfin erythrocytes had a rate of anion

exchange which was about 6 times higher than that of turtle and human erythrocytes

(Stabenau et al. 1991). An estimation of the number of exchangers present on bowfin

erythrocytes yielded a vdue of 265,000 copies per pin2, a value that is significantly greater

than that predicted for human and turtle erythrocytes (Knauf 1979; Stabenau et al. 199 1). It

is important to note that this calculation is dependent on the accuracy of the ce11 volume and

surface area measurernents and the assumptions that i) DIDS is binding in a one-to-one

fashion, ii) HC03' movement across the ce11 membrane is unidirectional. and iii) HC03-

rnovements occur only through the anion exchanger. Errors in any of these factors may

have therefore adversely affected Our estimation of the number of copies of Band 3 in

bowfin erythrocytes. Nonetheless, based on the bowfin erythrocyte anion exchange rate

59

and copy number, this analysis suggests that the turnover number of the bowfin anion

exchanger is several fold lower than that of the human or turtie exchanger. A similar

conclusion was reached for &out which have comparable exchange rates to humans but

have 4.3 times more copies of Band 3 (Romano and Passow 1984; Stabenau et al. 199 1).

Determination of the kinetic rates of both CA and Band 3 in this study provided an

oppoaunity to evaluate whether erythrocyte CÏ/HC03* exchange or CA activity should be

the rate limiting step in blood CO2 transport and excretion in bowfin. Under similar

substrate concentrations (COz and HC03'), bowfin erythrocytes were found to have a CA

activity (1.04 e-" mol CO2 - min-' . erythrocyte-l, 4°C) which was about 10 tirnes faster - 1 than the rate of anion exchange (9.25 e'I3 mol HC03- - min - erythrocyte-', 10°C). This

difference would only be further increased if both assays were run at the same temperature.

Results from this study therefore support the previous conclusions of Perry (1986) and

Peny and Giimour ( 1993) for other fish and indicate that the rate of erythrocyte anion

exchange is probably also the rate limiting step for blood CO2 transport and excretion in

bowfin.

Elecîrophoresis of trout and bowfin erythrocyte membrane polypeptides yielded

different separation patterns (Figure 3.6A). The broad diffuse band at approximately 100

kDa in the trout irnmunoreacted with the anti-trout AE 1 antibody (Figure 3.6B). This

agrees with previous results which determined that, based on the amino acid sequence,

trout Band 3 had a molecuIar mass of 102 kDa (Hubner et al. 1992). As suggested by

Cameron et al. ( 1996). reaction of heavier (400 m a ) trout polypeptides with the anti-bout

AE1 antibody may indicate that the trout AEI protein exists in monomenc, dirneric, and

also in higher order oligomeric foms. A diffuse band, similar to that in the trout, was also

observed in the bowfin, although it was found at a much larger molecular weight (140

kDa). Moreover, the diffuse bowfin band failed to react with the anti-trout AEl antibody

(Figure 3.68). Thus, while bowfin erythrocytes clearly possess the ability to carry out

60

anion exchange, the molecular characteristics of the exchanger may be quite different from

those of teleosts and higher vertebraies. Further molecular study of Band 3 rnay therefore

provide important insights into the evolution of anion exchange in vertebrates.

In conclusion, bowfin are similar to most vertebrates in that they possess a

considerable arnount of CA activity, as well as anion exchange, in their erythrocytes. In

addition, the blood of bowfin contains a plasma CA inhibitor which would likely restrict

catalysis of CO2 reactions to wiihin the erythrocytes. Thus, the basic strategy for blood

CO2 transport in bowfm is likely very similar to that in other vertebrates. Results from this

study also suggest, however. that both CA and Band 3 in bowfin may have unique

characteristics that represent an interesting intemediate stage in the phylogenetic evolution

of these important respiratory proteins. Further study of the kinetic and molecular

properties of CA and Band 3 in this primitive fish group are therefore warranted.

Chapter 4: GENERAL DISCUSSION

The goal of my thesis was to characterize two proteins in bowfin which play a

central role in the process of CO2 transport and excretion, narnely carbonic anhydrase (CA)

and the erythrocyte chloridehicarbonate exchanger (Band 3). There are two main rasons

why CA and the erythrocyte C1-MC03- exchanger in bowfin may have characteristics that

are unique in cornparison to other vertebrates. Firstiy, bowfin are arnong the relatively few

species of vertebrates that are bimodal breathers, using giUs for aquatic respiration and an

air bladder for aerial respiration, and secondly, bowfin occupy a primitive phylogenetic

position which is intemediate between that of elasmobranchs and teleosts. The results

obtained in this thesis have shown that severai characteristics of CA and the erythrocyte Cl-

/HC03- exchanger in bowfin are in fact probably unique arnong vertebrates. Taken

together, however, my results indicate that the basic strategy for CO-, transport in bowfin

blwd appears to be very sirnilar to that of most other vertebrates.

4.1 Unique characteristics of CA and the erythrocyte C[/HCO3- exchanger

in bowfin

Several unique characteristics of CA and the erythrocyte C1-/HC03- exchanger in

bowfin were revealed in this study. Although CA has been extensively studied in the lungs

of marnmals. vinually nothing is known about CA in the lungs of lower vertebrates. It is

for this reason that the most intriguing discovery in this thesis was the presence of

membrane-bound CA in the bowfin air bladder which had isozyrne characteristics sirnilar to

CA IV, the marnrnalian lung CA isozyme. This is somewhat surprising considering that

bowfin are among the most primitive venebrates whereas mammals are the most recent. It

is very possible that membrane-bound CA on the bowfin air bladder is the ancestral

isozyrne of CA IV. Further studies examining the characteristics of CA in the air-breathing

62

organs (ABO) of lungfish, which gave rise to the tetrapods, are now wamted, as are

molecular studies to determine the amino acid sirnilarity between bowfin A B 0 CA and

mammalian CA IV. A second interesting finding regarding CA in a gas exchange tissue. is

that CA in the gilis of bowfrn, like that of teleosts (Conley and Mallatt 1988; Henry et al.

1988; 1993)- is almost entirdy cytoplasrnic. Thus, unlike elasmobranchs (Swenson et al.

1995; Henry et al. 1997), a significant pool of membrane-bound CA does not appear to

have been incorporated in the gills of bowfin.

Earlier studies on CA in the erythrocytes of fish have suggested that there is a

pattern in the evolution of CA in lower vertebrates. The most primitive vertebrates,

elasmobranchs and agnathans, appear to have a slow type 1 CA isozyme in their

erythrocytes. whereas the more recent teleosts seem to have evolved a faster type II CA

isozyme in their erythrocytes (Maren et al. 1980; Henry et al. 1993). This study found that

CA in the erythrocytes of bowfin had a turnover number that was between that of

elasmobranchs and teleosts. Interestingly, this probably reflects the intermediate

phylogenetic position of bowfin. Thus, it is likely that bowfin erythrocyte CA represents an

intermediate step in the evolution from the slow CA isozyme to the fast CA isozyme in

10 wer vertebrates.

Like most vertebrates, bowfin erythrocytes were found to possess a high rate of

anion exchange. At low temperatures (- IO0C), the rate of HC03- flux was similar to that of

teleosts and higher vertebrates (Weith et al. 1982; Romano and Passow 1984; Stabenau et

al. 199 1; Tufts et al. 1998). Like most other vertebrates. C1-/HC03' exchange in bowfin

erythrocytes was highly sensitive to DIDS. Results from SDS-PAGE and Western blot

andysis, however, revealed that bowfin Band 3 may have different molecuIar

characteristics, as compared to that of other fish such as teleosts (Hubner et al. 1992) and

the more distant mammais (Jennings 1989). Further molecular studies on the erythrocyte

63

Cl-/HC03- exchanger in kmwfh and other primitive fishes may therefore be warranted to

determine the evolutionary significance of these unique Band 3 characteristics.

Based on studies that rneasured rates of CO2 excretion in vitro and in vivo, it has

b e n proposed that Cl-/HC03* exchange is the lirniting step in the transport and excretion of

CO, in fish (Perry 1986; Perry and Gilrnour 1993). This proposal is based, however, on

only a few species of fish. In this study, the measurement of CA activity and Cl-/HCO,-

exchange under similar assay conditions provided an oppominity to further examine this

idea in another fish species, the bowfin. Results from this study were found to support the

hypothesis put forth by Perry (1986), since the rate of CA activity was at least ten fold

higher than the rate of anion exchange in bowfin erythrocytes. Bowfin are therefore similar

to teleosts in that erythrocyte C1-/HC03- exchange is likely the Limiting step in blood CO2

transport and excretion.

For at Ieast a decade now, it has been thought that bowfin are among the few

vertebrates that lack a plasma CA inhibitor (Heming and Watson 1986). It was therefore

surprising that a plasma CA inhibitor was found in this study. It is likely that the previous

snidy by Heming and Watson ( 1986) did not use high enough volumes of plasma to inhibit

erythrocyte CA. As in other species, the role of the plasma CA inhibitor in bowfin may be

to restrict catalysis of CO2 reactions to the erythrocytes. Altematively, it has k e n proposed

that the CA inhibitor in vertebrates may be present to form a complex with CA which c m

then be picked up by the liver for recycling (Wuebbens et al. 1997) and inhibition of CA

activity may simply be a secondary consequence of this process. Interestingly, membrane-

bound CA in the air bladder was found to be at least three fold less sensitive to the plasma

CA inhibitor than erythrocyte CA. It is therefore conceivable that catalysis of CO2 reactions

in the plasma may occur in vessels of the air bladder. Further in vivo or in situ studies are

required, however. to determine whether air bladder membrane-bound CA faces into the

intmvascular lumen.

64

4.2 Mode1 of CO2 transport and excretion in bowfin: possible rolers) of

membrane-bound CA in the air bladder

n i e two main proteins involved in the transport and excretion of CO2 in the blood

of vertebrates are CA and the erythrocyte CÏMC03- exchanger. The knowledge gained by

this thesis about the characteristics of CA and the erythrocyte Cl-MC03- exchanger in

bowfin provides an oppoctunity to propose a mode1 for the transport and excretion of CO2

in this primitive air-breathing vertebrate (Figure 4.1).

The basic strategy that bowfin use for transporthg and excreting the majority of

CO2 under normal conditions is probably quite similar to that of most other vertebrates.

Since CA is not present in the plasma, CO2 diffusing out of the erythrocytes wiU not be

rapidy hydrolyzed untii it reaches the elythrocytes. In the erythrocytes, cytoplasmic CA

will catalyze the hydrolysis of CO2 to fom HC03- and H+. Since hemoglobin in bowfm

erythrocytes is very similar to that of teleosts (Weber et al. 1976), the protons generated

would likely be buffered by hemoglobin. The presence of a high degree of C1-MC03-

exchange in b w f m erythrocytes suggests that most of the bicarbonate formed by the

hydration reaction would be rapidy shuttled out of the erythrocytes in exchange for plasma

chloride. Since it is unlikely that the gills of bowfin have membrane-bound CA that faces

into the blood, plasma bicarbonate would have to enter back into the erythrocytes via the

anion exchanger to be catalyzed back to CO2 which would diffuse into the water.

There are several possibilities for the potential role of membrane-bound CA in the

air bladder of bowfin. The first, which does not involve respiration. is in the regulation of

buoyancy. It has been proposed that membrane-bound CA in the swimbladder of eels

contributes to buoyancy regulation by providing both CO2 and H+ from the gas gland cells

for the Root effect (Pelster 1995). In this case, CA faces into the extraceliular fluid of the

swimbladder tissue rather than into the b l d (Pelster 1995). This role is probabIy unlikely

since bowfin, which live in shallow bays (Scott and Crossman 1973). probably do not

Figure 4.1. The proposed strategy for CO2 transport and excretion in bowfïn. CO2 from the metabolizing tissues diffuses into the erythrocytes and becomes hydrolyzed by carbonic anhydrase (CA) to fom H C 4 ' and H+. The protons generated are largely buffered by hemoglobin (Hb) which releases O2 to the tissues. Most of the bicarbonate anions are passively transported out of the cell in exchange for plasma chlonde by the Cl-/HCO3- exchanger. At the gills, Hb and the CÏ/HC03- exchanger supply the H+ and HC03' required for the formation of CO2 which diffuses into the water. Catalysis of plasma CO2 reactions may occur in the vessels of the air bladder because of the presence of membrane- bound CA. Due to a limitation of plasma protons, only a smaii portion of plasma HCO, would be hydrolyzed by air bladder CA to yield CO2 which f i s e s into the air. The main pathway for CO2 excretion across the air bladder would involve catalyses of plasma HCO-, , that entered into the erythrocyte via a CI-/HC03- exchanger, to yield CO2. In addition air bladder CA facing into the blood would maintain an equilibnum between CO2 and H+. Because the vessels of the air bladder are in parailel with those of the metabolizing tissues, the majority of the CO2 produced by the tissues would be excreted at the giils. The relative importance of air bladder CA for CO2 excretion may be increased. however, under conditions of aquatic hypercapnia (see text for details).

66

regulate their buoyancy very much (D. J. Randall, pers. comm.). Furthemore, bowfin

hemoglobin does not have a significant Root effect (Johansen et ai. 1970). If membrane-

bound CA in the bowfin air bladder faces into the blood. a second possible function could

be sirnilar to that of CA IV in the mamalian lung. This function would be to equiiibrate

plasma CO2 and H' in the circulation of the air bladder and provide accurate information

for pH and Pco2 sensitive peripherd chemoreceptors located downstream of the air

bladder. Based on the studies by johansen et al. (1970) and McKenzie et al. (199 1). it has

k e n suggested that bowfin may have more than one group of C02/pH sensitive

chemoreceptors that control ventilation (Smatresk 1994). Aithough it is possible that one of

these groups of chemoreceptors is located just downstrearn of the air bladder, further study

is needed to systematically identify COî or pH sensitive chemoreceptors in air-breathing

fish.

A third possible function of membrane-bound CA in the bowfin air bladder may be

for aeriai CO2 excretion across the air bladder (Heming and Watson 1986). Air bladder CA

facing into the blood would be available to catalyze plasma HC03- back to CO2 which

could diffuse into the air bladder for excretion. Plasma. however, is unlike the cytoplasm

of erythrocytes in that there is no large reserve of H+ buffered by hemoglobin. Under

normal circumstances. it is therefore unlikely that there would be a significant reserve of H+

for HC03- dehydration in bowfin. Thus. the vast majonty of CO2 that is excreted across

the air bladder probably originates from plasma HC03- which is catalyzed to COî by

erythrocyte CA. Moreover, since the air bladder vessels are in parallel with those of the

metabolizing tissues. most of the CO2 excreted would diffuse out across gills rather than

across the air bIadder (Randail et al. 198 1)-

The role of membrane-bound air btadder CA may, however. become much more

significant for CO2 excretion during conditions such as aquatic hypercapnia (elevated levels

of CO2) which are known to occur iîi the shallow warm bays that bowfin occupy (Ultsch

67

1996). In this situation. the gills would be eliminated as a site for CO2 excretion because of

a reverse gradient in Pcp?. As a result an acidosis would also be created in the blood

perfusing the air bladder. This acidosis could provide the protons needed for air bladder

CA to catalyze a large arnount of plasma HC03- to CO2 for aerial excretion. Although

plausible, this possible role of air bladder membrane-bound CA in bowfin needs further

s tudy .

In conclusion. several unique features of CA and the erythrocyte CI-/HC03-

exchanger in bowfin have been identified in this thesis. These interesting features are likely

a consequence of the unique phylogenetic position and respiratory system of this primitive

vertebrate. Despite the unique characteristics of these respiratory proteins, however. the

basic strategy for CO2 transport and excretion in bowfin is probably very sirnilar to that in

most vertebrates. Nonetheless, there are numerous avenues which warrant further study in

this area. For example, molecular characterization of CA and erythrocyte Band 3 in bowfïn

may yield important insights into the evolution of these respiratory proteins. In addition.

furîher study should also be conducted to determine whether there are circurnstances (e.g.

aquatic hypercapnia) where AB0 CA may have an important role in CO2 excretion in these

primitive fish.

68

LITERATURE CITED

Beam K. G., S. L. Alper, G. E. Palade, and P. Greengard. 1979. Hormonally regulated phosphoprotein of turkey erythrocytes. J. Cell. Biol. 83: 1 - 15.

Bidani, A. 199 1. Analysis of abnonnalities of capillary CO2 exchange in vivo. J. Appl. Physiol. 70: 1 686- I 699.

Bidani, A. and T. A. Heming. 199 1. Effects of perfusate buffer capacity on capiIIary CO2- HCO~--H+ reactions: theory. J. Appl. Physiol. 7 1 : 1460- 1468.

Bidani, A.. S. I. Mathew, and E. D. Crandall. 1983. Pulmonary vascular carbonic anhydrase activity . J. Appl. Physiol. 5575-83.

Booth, V. H. 1938. The carbonic anhydrase inhibitor in serum. J. Physiol. 9 1:474-489

Bottcher, K., A. Waheed, and W. S. Sly. 1994. Membrane-associated carbonic anhydrase from the crab g ik purification, characterization, and cornparison with mammalian CAS. Arch. Biachem. Biophys. 3 l2:429-435.

Brill, S. R., M. W. Musch, and Goldstein, L. 1992. Taurine efflux, , and erythrocyte volume regulation of the hagfish (Myxine glutinosu) and lamprey (Petromyzon marinus). J . f i p . Zool. 264: 19-25.

Burggren, W., and S. Haswell. 1979. Aerial CO2 excretion in the obligate air breathing fish Trichogaster trichopteru: a role for carbonic anhydrase. J. Exp. Biol. 82:2 15 -225.

Cameron, B. A., S. F. Perry, C. Wu, and B. L. Tufts. 1996. Bicarbonate permeability and immunological evidence for an anion exchanger-like protein in the red blood cells of the sea larnprey, Petromyzon marinus. J. Camp. Physiol. 166: 197-204.

Cameron, J. N. 1979. Excretion of CO2 in water-breathing animals - a shon review. Mar. biol. k t t . 1:3- 13.

Carlsson, U., B. Kjellstrorn, and B. Antonsson. 1980. Purification and properties of c yclostome carbonic anhydrase frorn erythrocytes of hagfish. Biochim. Biophys. Acta. 6 1 2: 160- 170.

69 Carter, M. J. 1972. Carbonic anhydrase: isoymes, properties, distribution, and functional

significance. Biul. Rev. 47:465-5 13.

Casey, J R. and R. A. F. Reithmeier. 1991. Analysis of the oligomeric state of Band 3, the anion transport protein of the human eryihrocyte membrane, by exclusion high performance liquid chromatography. J. Biol. Chem. 266: 15726- 15737.

Conley. D. M., and J. Mallatt 1988. Histochemical localization of Na%' ATPase and carbonic anhydrase activity in gills of 17 fish species. Can. J. Zool. 66:2398-2405.

Crandall, E. D., and J. E. O'Brasky. 1978. Direct evidence for participation of n t lung carbonic anhydrase in CO2 reactions. J. Clin. Invest. 62:6 18-622.

Cross, G. A. M. 1987. Eukaryotic protein modification and membrane attachment via phosphatidylinositol. Cell. 48: 179- 18 1.

Dimberg, K. 1994. The carbonic anhydrase inhibitor in trout plasma: purification and its effect on carbnic anhydrase activity and the Root effect. Fish Physiol. Biochem. 12:381-386.

Dixon, M. 1953. The determination of enzyme inhibition constants. Biochem. J. 55: 170 -171.

Dodgson, S. J., R. E. Tashian, G. Gros. and N. D. Carter. (eds.) 1991. The Carbonic Anhydrnses, Cellular physiology and molecular genetics. Plenum Press. New York. 379 pp.

Dodgson, S. J. 199 1. The Carbonic anhydrases: Overview of their importance in cellular physiology and in molecular genetics. In The carbonic anhydrases. (ed. S . J. Dodgson, R. E. Tashian, G. Gros, and N. D. Carter), pp. 3-14. Plenum Press. New York.

Easson, L. H. and Stedman E. 1937. The absolute activity of choline esterase. froc. R. Soc. Land. Ser, B. 1 2 1 : 142- 1 64.

Ellory, J. C., M. W. Wolowyk, and J. D. Young. 1987. Hagfish (Eptatrefus sfouti) erythrocytes show minimal chloride trasnport activity. J. Exp. Biol. 129:377-383.

Feller, G., A. Péqueux, and G. Hamoir. 198 1. La présence d'anydrase carbonique chez two poissons de l'archipel des Kerguelen, Channichthys rhinoceratus exempt d'hémoglobine et Notothenia mageilmica de formule sanguine normale. Comptes Rendus Séances Acad- Sc. 293:395-397.

Fleming, R.E., M.A. Moxley, A. Waheed, E. C. Crouch, W. S. Sly, and W. J. Longmore. 1994. Carbonic anhydrase II expression in rat type II pneurnatocytes. Am. J. Respir. Cell M d Biol. 10,499-505.

Gervais, M. R. and B. L. Tufts. 1998. Evidence for membrane-bound carbonic anhydrase in the air bladder of bowfin (Amia clava), a primitive air-breathing fish. J. fip. Biol. in press.

Gilmour, K. M., R, P. Henry, C. M. Wood, and S. F. Perry. 1997. Extracellular carbonic anhydrase and acid-base disequilibnum in the b l d of the dogfish Squalus acunthias. J- Exp. Bioi. 200: 173-183.

Girard, J. P., and M. Istin. 1975. Isoenzymes de l'anhydrase carbonique d'un poisson euryhalin: variations en relation avec l'osmoregulation. Biochim Biophys. Acta. 38 1 :22 1 - 232.

Graham, M. S., J. D. Turner, and C. M. Wood. 1990. Control of ventilation in the hypercapnic skate Raja ocella: 1. Blood and extracellular fluid. Respir. Physiol. 80:259-277.

Hall, G. E. and R. Schraer. 1983. Characterization of a high activity carbonic anhydrase isozyme purïfied from erythrocytes of Salmo gairdneri. Comp. Biochem. Physiol. 7SB:8 1-92.

Haswell, M. S. and D. J. Randdl. 1978. The pattern of carbon dioxide excretion in rainbow trout Salmo gairdnerî. J. Exp. Bioi. 72: 17-24.

Haswell, M. S., J. P. Raffin, and C. LeRay. 1983. An investigation of the carbonic anhydrase inhibitor in eel plasma. Cornp. Biochem. & Physiol. 74A: 175- 177.

Heming T. A. and A. Bidani. 1992. Influence of proton availability on intracapillary CO, HCO~--H+ reactions in isolated rat lungs. J. Appl. Physiol. 72:2 140-2 148.

Heming, T. A., and T. A. Watson. 1986. Activity and inhibition of carbonic anhydrase in Amia calva, a birnodal-breathing holostean fish. J. Fish Biol. 28:385-392.

7 1 Heming, T. A., C. G. Vanoye, E. K. Stabenau, E. D. Roush, C. A. Fierke, and A.

Bidani. 1993. Inhibitor sensitivity of pulmonary vascular carbonic anhydrase. J. Appl. Physiol. 75: 1642- 1649.

Henry, R. P. 199 1. Techniques for rneasunng carbonic anhydrase activity in vitro. In The Carbonic Anhydrases (ed. S . J . Dodgson, R. E. Tashian, G. Gros, and N. D. Carter), pp. 1 1 9- 1 3 1. New York: Plenum Press.

Henry, R. P. 1996. Multiple roles of carbonic anhydrase in cellular transport and metabolism. Annu. Rev. Physiol. 58523-538-

Henry, R. P., B. L. Tufts, and R. G. Boutilier- 1993. The distribution of carbonic anhydrase type I and Ki isozymes in lamprey and trout: possible coevolution with erythrocyte chloride/bicarbonate exchange. J. Comp. Physiol. 163:380-388.

Henry, R. P., K. M. Gilmour, C. M. Wood, and S. F. Perry. 1997. Extracellular carbonic anhydrase activty and carbonic anhydrase inhibitors in the circulatory system of fish. Physiol. Zool.. 70:650-659.

Henry, R. P., N. J. Srnatresk, and J. N. Carneron. 1988. The distribution of branchial carbonic anhydrase and the effects of giil and erythrocyte carbonic anhydrase inhibition in the channel catfish Ic ta lum punctarus. J. Exp. Biol. lW2O 1-2 1 8.

Henry, R. P., S. J. Dodgson, R. E. Forster, and B. T. Storey. 1986. Rat lung carbonic anhydrase:activity, localization, and isozymes. J. Appl. Physiol. 60~638-645.

Hewett-Emmett, D. and R. E. Tashian. 199 1. Structure and evolutionary ongins of the carbonic anhydrase multigene family. In The carbonic anhydrases. (ed. S . I. Dodgson, R. E. Tashian, G. Gros, and N. D. Carter), pp. 15-32. Plenum Press, New York.

Hewett-Emrnett, D. and R. E. Tashian. 1996. Functional diversity, conservation, and convergence in the evolution of the a-, p-. and y-carbonic anhydrase gene families. Mol. Phyl. Evol. 550-77.

Hill, E. P. 1986. Inhibition of carbonic anhydrase by plasma of dogs and rabbits. Appl. Physiol. 60: 19 1 - 197.

Houston, A. R. 1990. Blood and circluIation. In Methods for Fish Biology. (ed. C . B . Schreck and P. B. Moyle), pp. 273-334. Bethesda: American Fisheries Society.

72 Hubner, S., F. Michel, V. Rudloff, and H. Appelhans. 1992. Amino acid sequence of

band-3 pmtein from rainbow trout erythrocytes derived from cDNA. Biochem J. 285: 17-23.

Jennings, M. L. 1989. Structure and function of the red blood ce11 anion transport protein. Annu. Rev. Biophys. Chem. 1 8:397-430.

Jensen. F. B. and I. Brahm. 1995. Kinetics of chioride transport across fish red blood ceii membranes. J. Exp. Biol. 198:2237-2244.

Johansen. K., D. Hanson, aiid C. Lefant. 1970. Respiration in a primitive air breather, Arnia calva. Respir. Physiol. 9: 162- 174.

Khodadad. J. K. and R. S. Weinstein. 1983. The -rich membrane of llarna erythrocytes: studies on ce11 shape and the organization of membrane proteins. J. Membr. Biol. 72:161- 171.

Kim, J-S.. C. V. Gay, and R. Schraer. 1983. Purification and properties of carbonic anhydrase from salrnon erythrocytes. Comp Biochem. Physiol. 76B:523-527.

Klocke, R. A. 1980. Equilibnum of CO2 reactions in the pulmonary capillary. J. Appl. Physioi. 48:972-976.

Knauf, P. A. 1979. Erythrocyte anion exchange and the Band 3 protein: transport kinetics and molecular structure. In Current Topics in Membranes and Transport. vol 12 (ed. F. Bronner and A. KleinzeUer), pp. 249-363. Academic, New York.

Laemmli, U. K. 1970. Cleavage of stnictural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685.

Low, M. G., J. Steinberg, G. L. Waneck, R. A. Flavell, and P. W. Kincade. 1988. Cell- specific heterogeneity in sensitivity of phosphaiidylinositol-anchored membrane antigens to release by phospholipase C. J. Immunol. Merhods 1 13: 10 1 - 1 1 1.

Maffia, M., F. Trischitta, M. G. Lionetto, C. Storelli, and T. Schettino. 1996, Bicarbonate absorption in the eel intestine: evidence for the presence of membrane-bound carbonic anhydrase on the bmsh border membranes of the enterocyte. J. Exp. Zool. 275:365-373.

Magid, E. 1967. The activity ofcarbonic anhydrase B and C from human erythrocytes and the inhibition of the enzymes by copper. Scand. J. Huemat. 4:257-270.

73 Maren, T. H. 1967. Carbonic anhydrase: Chemistry, physiology, and inhibition. Physiol.

Rev. 47595-78 1.

Maren, T. H. and B. R. Freidland. 1978. Further studies on the phylogeny of vertebrate carbonic anhydrase in red cells and secretory organs. Bull. Mt. Desen Isf. Biol. Lab. 18:79-82.

Maren, T. H. and G. Sanyal. 1983. The activity of suifonamides and anions against the carbonic anhydrases of animals, plants, and bacteria. Ann. Rev. Phannacol. Toxicol. 23, 439-459.

Maren, T. H., A. L. Parcell, and M. N. Malik. 1960. A kinetic analysis of carbonic anhydrase inhibition. J. Phamaco l. Exp. Ther. 1 3O:3 89-400.

Maren, T. H., B. R. Freidland, and R. S. Rittrnaster. 1980. Kinetic properties of primitive vertebrate carbonic anhydrase. Comp. Biochem Physiol. 67B:69-74.

Maren, T. H., G. C. Wynns, and P. J. Wistrand. 1993. Chernical propenies of carbonic anhydrase N, the membrane-bound enzyme. Mol. Pharmacoi. 4490 1-905.

McKenzie, D. J., S. Aota, and D. J. Randall. 199 1. Cardiovasculiir and ventilatory responses to blood pH, PCO? blood O2 content and catecholamines in an air- breathing fish, the bowfin &nia calva). Physiol. ZooII 64: 193-203.

Nikinmaa, M. and E. Railo. 1987. Anion movements across larnprey (LampetrafZuviutilis) red ce11 membrane. Biochim. Biophys. Acfa 899: 134- 136.

Nioka, S. and R. E. Forster. 1991. Lung carbonic anhydrase. In The carbonic anhydrases. (ed. S. J. Dodgson, R. E. Tashian, G. Gros, and N. D. Carter), pp. 333-340. Plenum Press, New York.

Obaid, A. L., A. M. Critz, and E. D. Crandall. 1979. Kinetics of chIoride/bicarbonate exchange in dogfish eryhtrocytes. Am. J. Physiol. 237:R 132-R 138.

Pelster, B. 1995. Mechanisms of acid release in isolated gas gland cells of the European eet Anguilla anguilla. Am. J. Physiol. 269: R793-R799.

Perry, S. F. 1986. Carbon dioxide excretion in fishes. Can. J. Zool. 64565-572-

Perry, S. F. and K. Gilmour. 1993. An evaiuation of factors Iimiting carbon dioxide excretion by trout red blood cells in vitro. J. Exp. Biol. 180:39-54.

Peny, S. F., and P. Laurent. 1990. The role of carbonic anhydrase in carbon dioxide excretion, acid-base balance, and ionic regulation in aquatic gill breathers. In Animal nutririon and trunsport processes. 2. Trcmport, respiration and excretion: Comparative and environmental aspects. Comp. Physiol. (ed. J-P. Tmchot and B. Lahlou, pp. 39-57. Basel, Karger.

Perry, S. F., C. M. Wood, P. J. Walsh, and S. Thomas. 1996. Fish red blood ceIl carbon dioxide excretion in vitro: a comparative study. Comp. Biochem. Physiol. 1 l3A: 121-130.

Perry, S. F., P. S. Davie, and C. Daxboeck. 1982. A cornparison of CO2 excretion in a spontaneously ventilating blood-perfused trout preparation and saline-perfused gill preparations: Contribution of the branchial epithelium and the red blood celi. J. Exp. BioL 10 1 :47-60.

Rahim, S. M., J. P. Delaunoy, and P. Laurent. 1988. Identification and irnmunocytochemical localization of two different carbonic anhydrase isoenzymes in teleostean fish erythrocytes and gill epithelia. Histochemistry 89:45 1-459.

Rahn, H., K. B. Rahn, B. J. Howell, G. Gans, and S. M. Tenney. 197 1. Air breathing of the garfish, Lepisosteus osseus. Respir. Physiol. 2:433-466.

Randall, D. J., J. N. Cameron, C. Daxboeck, and N. Smatresk. 198 1. Aspects of bimodal gas exchange in the bow fin, Amia calva L. (actinoptery gii: amiiformes). Respir. Physiol. 43 :339-348.

Romano, L. and H. Passow. 1984. Characterization of anion transport system in uout red blood cells. Am. J. Physiol. 246 (Ce11 Physiol. 15): C330-C338.

Roush, E. D. and C. A. Fierke. 1992. Purification and characterization of a carbonic anhydrase II inhibitor from porcine plasma. Biochem. 3 1 : 12536- 12542.

Ruud, J. T. 1965. The icefish. Scientrjk American. 213: 108-1 14.

Ryan, U. S., L. Whitney. and J. W. Ryan. 1982. Localization of carbonic anhydrase on pulmonary artery endothelial cells in culture. J. Appl. Physiol.: Respir. Environ. Exercise Physiol. 53 :9 1 4-9 1 9.

Sanyal, G. 1984. Comparative carbon dioxide hydration kinetics and inhibition of carbonic anhydrase isozymes in vertebrates. Ann. N. Y. Acad. Sci. 429: 165- 178.

75 Sanyal, G., N. 1. Pessah, E. R. Swenson, and T. H. Maren. 1982. The carbon dioxide

hydration activity of purified teleost red ce11 carbonic anhydrase. Inhibition by sulfonarnides and anions. Cornp. Biuchem Physiol. 73%:937-944.

Scott, W. B. and E. J. Crossman. 1973. Freshwater fishes of Canada. Bull. Fish. Res. Board Can. 184:lll-116,

Singer, T. D. and J. S. Ballantyne. 1990. Metabolic organization of a primitive fish. the bowfin (Amia culva). Can. J. Fish. Aquat. Sci 48:6 1 1 -6 1 8.

Sly, W. S. and P. Y. Hu. 1995. Human carbonic anhydrases and carbonic anhydrase deficiencies. Annu. Rev. Biochem. 64:375-40 1 .

Smatresk, N. 1. 1994. Respiratory control in the transition from water to air breathing in vertebrates. Amer. Zool. 34:264-279.

Stabenau E. K., C. G. Vanoye, and T. A. Herning. 1991. Characteristics of the anion transport system in sea turtle erythrocytes. Am J. Phyiol. 26 1 (Regulatory Integrative Comp. Physiol. 30): R 12 1 8- 1 225.

Steck, T. L. 1974. The organization of proteins in the hurnan red blwd cell membrane. J. Cell Biol. 62: 1-19.

Swenson, E. R. 1990. Kinetics of oxygen and carbon dioxide exchange. In Advances in Comparative and Environmental Physiology, vol. 6 (ed. R. G. Boutilier). pp. 163- 2 10. Berlin, Heidleberg: Springer-Verlag.

Swenson, E. R. and T. H. Maren. 1978. A quantitative analysis of CO, transport at rest and during maximal exercise. Respir. P hysioL 35: 129- 159.

Swenson, E. R., J. Gronlund, J. Ohlsson. and M. Hlastala. 1993. In vivo quantitation of carbonic anhydrase and Band 3 pmtein contribution to pulmonary gas exchange. J. Appl. Physiol, 74:838-848.

Swenson, E. R., L. Lippincott, and T. H. Maren. 1995. Effect of membrane-bound carbonic anhydrase inhibition on branchial bicarbonate excretion in the dogfish shark. Biol. Bull. 34:94-95.

Toews, D., R. G. Boutilier, L. Todd, and N. Fuller. 1978. Carbonic anhydrase in the amphibia. Comp. Biochem. Physiol. 59A:2 1 1-2 1 3.

76 Tufts, B. L. and R. G. Boutilier. 1989. The absence of rapid chlorïde/bicarbonate

exchange in lamprey erythrocytes: implications for CO2 transport and ion distributions between plasma and erythrocytes in the blood of Petromyzon marinm. J. Exp. B i d 144:565-576.

Tufts, B. L. and R. G. Boutlier. 1990. CO2 transport properties of the blood of a primitive vertebrate, Myxine glutinosa (L.). &p. Biol. 48:34 1-347.

Tufts, B. L., M. R. Gervais, A.G. Moss, and R. P. Henry. 1998. Carbonic anhydrase and red blood ce11 anion exchange in the neotonic aquatic salamander, Nectum macu1osu.s. Physio. Zool. In press.

Tufts, B. L., R.C. Drever, B. Bagatto, and B. A. Cameron. 1994. In vitro analysis of volume and pH regulation in the red blood ceils of a primitive air-breathing fish, the bowfin, Amia calva. Can J. 2001. 72:280-286.

Ultsch. G. R. 1996. Gas exchange, hypercarbia and acid-base balance, pdeoecology, and the evolutionary transition from water-breathing to air-breathing among vertebrates. Palaeo. 123: t -27.

Van Goor, H. 1948. Carbonic anhydrase - its properties, distribution, and significance for CO, transport. Enqmologia 13:73- 165.

Waheed, A., T. Okuyarna, T. Heyduk, and W. S. Sly. 1996. Cabonic anhydrase IV: purification of a secretory form of the recombinant hurnan enzyme and identification of the positions and importance of its disulfide bonds. Arch. Biochem. Biophys. 333:432-438.

Weber, R. E., B. Sullivan, J. Bonaventura, and C. Bonaventura. 1976. The hemoglobin system of the primitive fish, Amia calva: Isolation and functional characterization of the individual hemoglobin components. Biochim. Biophys. Acta. 434: 18-3 1 .

Weith, J. O., O. S. Anderson. J. Brahm, P. J. Bjerrum. and C. L. Borders. 1982. Chloride-bicarbonate exchange in red blood cells: physiology of transport and chernical modification of binding sites. Phil. Tram R. Soc. Ser. B. 299:383-399.

Whitney, P. L., and T.V. Bnggle. 1982. Membrane-associated carbonic anhydrase purified from bovine h g . J. Biol. Chem. 257: 12056- 12059.

Wood, C. M., I. D. Turner, R. S. Munger, and M. S. Graham. 1990. Control of ventilation in the hypercapnic skate Raja ocella: II. Cerebrospinal fluid and intracellular pH in the brain and other tissues. Respir. Physiol. 80:277-298.

Wuebbens, M. W., E. D. Roush, C. M. Decastro, and C. A. Fierke. 1997. Cloning, sequencing, and recombinant expression of the porcine inhibitor of carbonic anhydrase: A novel member of the transfemn famil y. Biochem. 36,4327-4336.

Zhu. X. L., and W. S. Sly. 1990. Carbonic anhydnse IV from human lung. J. Biol. Chem. 265:8795-880 1.

IMACjt LVALUATION TEST TARGET (QA-3)

APPLIEO 2 IWGE . lnc 1653 East Main Street - -. - Rochester. NY 14609 USA -- --= Phone: 71 W482-0300 -- -- - - Far: 71 W88-5989

O 1993. Applied Image. lm. AN Ri$'~b Fiesennid