Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

Contents lists available at ScienceDirect

Comparative Biochemistry and Physiology, Part A

j ourna l homepage: www.e lsev ie r .com/ locate /cbpa

High blood oxygen affinity in the air-breathing swamp eelMonopterus albus

Christian Damsgaard a,⁎, Inge Findorf a, Signe Helbo a, Yigit Kocagoz a, Rasmus Buchanan a,Do Thi Thanh Huong b, Roy E. Weber a, Angela Fago a, Mark Bayley a, Tobias Wang a

a Zoophysiology, Department of Bioscience, Aarhus University, 8000 Aarhus C, Denmarkb College of Aquaculture and Fisheries, Can Tho University, Can Tho City, Vietnam

Abbreviations: [CO2]pl, Total concentration of carbonHemoglobin; [Hb-O2], Concentration of oxygen bound toMb, Myoglobin; n50, Hill’s cooperativity coefficient at halcentration of oxygen in blood; P50, Partial pressure of oPartial pressure of carbon dioxide; PO2, Partial pressureY, Fractional saturation; αCO2, Solubility coefficient of cblood, Solubility coefficient of oxygen in blood; φ, Bohr fac⁎ Corresponding author at: Zoophysiology, Departmen

3, Aarhus University, DK-8000 Aarhus C, Denmark. Tel.: +E-mail address: cdamsg@gmail.com (C. Damsgaard).

http://dx.doi.org/10.1016/j.cbpa.2014.08.0011095-6433/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 25 May 2014Received in revised form 4 August 2014Accepted 12 August 2014Available online 17 August 2014

Keywords:Air-breathingBloodHemoglobinMyoglobinRespirationSwamp eel

The Asian swamp eel (Monopterus albus, Zuiew 1793) is a facultative air-breathing fish with reduced gills.Previous studies have shown that gas exchange seems to occur across the epithelium of the buccopharyngealcavity, the esophagus and the integument, resulting in substantial diffusion limitations that must be compensat-ed by adaptations in others steps of the O2 transport system to secure adequate O2 delivery to the respiringtissues. We therefore investigated O2 binding properties of whole blood, stripped hemoglobin (Hb), two majorisoHb components and themyoglobin (Mb) fromM. albus. Whole bloodwas sampled using indwelling cathetersfor blood gas analysis and determination of O2 equilibrium curves. Hbwas purified to assess the effects of endog-enous allosteric effectors, and Mb was isolated from heart and skeletal muscle to determine its O2 binding prop-erties. The blood ofM. albus has a high O2 carrying capacity [hematocrit (Hct) of 42.4 ± 4.5%] and binds O2 withan unusually high affinity (P50= 2.8± 0.4mmHg at 27 °C and pH 7.7), correlatingwith insensitivity of the Hb tothe anionic allosteric effectors that normally decrease Hb-O2 affinity. In addition, Mb is present at high concen-trations in both heart and muscle (5.16 ± 0.99 and 1.08 ± 0.19 mg ∙ g wet tissue-1, respectively). We suggestthat the high Hct and high blood O2 affinity serve to overcome the low diffusion capacity in the relatively ineffi-cient respiratory surfaces, while highHct andMb concentration aid in increasing theO2flux from the blood to themuscles.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

The Asian swampeel (Monopterus albus) is an air-breathingmemberof the Synbranchidae that is widely distributed across South-east Asia(Rosen and Greenwood, 1976), where it inhabits slow flowing andoften hypoxic waters. Air-breathing is believed to have evolved as anadaptation to aquatic hypoxia and/or seasonal water level fluctuations,and the vast majority of extant air-breathing fish are found in tropicalhypoxic waters (Graham, 1997). Unlike most air-breathing fish,M. albus lacks a distinct air-breathing organ (ABO) and relies on extra-branchial gas exchange using a highly vascularised epithelium in thebuccopharyngeal cavity as well as a vascularized esophagus and

dioxide in blood plasma; Hb,hemoglobin; Hct, Hematocrit;

f saturation; [O2]total, Total con-xygen at half saturation; PCO2,of oxygen; RBC, Red blood cell;arbon dioxide in plasma; αO2,

tor.t of Bioscience, C.F. Møllers Alle45 87 15 43 27.

integument (Taylor, 1831; Liem, 1967; Lefevre et al., 2014). The buccalcavity of M. albus expands during air-breathing remaining initiallyinflated during submergence and exhalation occurs both under waterand at the surface prior to the next inhalation (Wu and Kung, 1940).Given the reduced gills,M. albus was originally classified as an obligateair-breather, like its close relative M. cuchia (Carter, 1931; Wu and Lui,1943; Lomholt and Johansen, 1976), but because M. albus maintainsblood O2 concentrations during forced submersion in normoxic water,it was argued to be a facultative rather than obligate air-breather(Iversen et al., 2013). The pharyngeal air-breathing structures arecharacterised by respiratory islets divided by non-respiratory section(Iversen et al., 2013), where only the former is perfused by intra-epithelial capillaries (Liem, 1967). This, in combination with a smallerrespiratory surface area and presumably con-current gas exchangerepresent a considerably lower diffusion capacity and lower gasexchange efficiency than normal piscine gills (Hughes, 1972). Thus, tosustain O2 uptakewith a low diffusion capacity it would seem beneficialto maintain a large PO2 (partial pressure of oxygen) gradient across thisepithelium, which can be achieved by high O2 affinity and O2 carryingcapacity of the blood (Hlastala and Berger, 2001).

The correlation between blood O2 affinity across fish species andtheir natural environmental O2 availability was first noted by Krogh

103C. Damsgaard et al. / Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

and Leitch (1919). In addition, given the much greater O2 availability inair than in water, it has been postulated that the evolutionary transitionfrom water to air-breathing would be associated with a decrease inblood O2 affinity (McCutcheon and Hall, 1937; Johansen et al., 1978b),as documented in at least two pairs of closely related specieswith differ-ent breathing mode (Powers et al., 1979). Although there are largevariations in blood O2 binding affinities amongst air-breathing fish(Lenfant and Johansen, 1968; Lomholt and Johansen, 1976; Heisler,1982) an analysis of the O2 affinity of whole blood from 40 genera ofwater and air-breathers from the Amazon basin (Powers et al., 1979)revealed no evidence of any systematic difference associated to breath-ing mode, while noting a trend that O2 affinity in species inhabitingslow-flowing hypoxic water was higher compared to species living infast flowingwaters. While the evolution of blood with a high O2 affinitywould benefit branchial O2 uptake, the unloading of O2 in the tissuesbecomes more difficult (Brauner and Wang, 1997; Wang and Malte,2011). This can be countered in part by the Bohr effect, where protonbinding to Hb and stabilizes its low affinity T(ense)-state conformation,increases O2 unloading and causes blood PO2 to increase. Unloading ofO2 in the tissues can further be enhanced by increasing the O2 fluxfrom the blood to the tissues. Myoglobin (Mb) functions as an intracel-lular O2 carrier in the skeletal and heart muscles of most vertebrates(Wittenberg and Wittenberg, 2003) and when expressed at highconcentrations it would increase the flux of O2 from blood to themitochondria.

Teleost fishes display the most extensive heterogeneity in adultHb structure and function amongst vertebrates and can accordinglybe categorized (Weber, 2000) as class I species like plaice and carp(Weber and de Wilde, 1976) that have multiple (electrophoretically-)anodal Hbs with almost identical O2 affinities and Bohr effects, andClass II species (anquillid eels, salmonids and catfish) (Weber et al.,1976; Weber and Lykkeboe, 1978) that additionally expresselectrophoretically-cathodal isoHbs that have high isoelectric points(pI N 8.2), commonly exhibit high O2 affinities, and show reverse orno Bohr effects in the absence of anionic effectors. The latter isoHbshave variously been postulated to function as a blood O2 reservethat can be drawn upon during hypoxia or when blood pH decreases(e.g. due to increased physical activity in fast-flowing water) (Powers,1972; Weber, 1990). No information appears to be available on thefunctional consequences of Hb multiplicity in Monopterus albus.

We hypothesized that M. albus would exhibit high blood O2

carrying capacity, a high blood O2 affinity, and high Mb concentra-tions in the O2 consuming muscles. To examine these hypotheses,we measured hematocrit (Hct) and whole blood O2 equilibria attwo CO2 levels to determine blood O2 affinity and Bohr effect. Be-cause the O2 affinity of the Hb in blood depends on its intrinsicO2 affinity and its interaction with protons and red blood cell(RBC) anionic effectors (Weber and Fago, 2004), we also measuredO2 equilibrium curves of the stripped (cofactor-free) Hb and themajor isoHb components, variously in the absence and presenceof chloride, and of physiological levels of RBC co-factors (unstrippedHb solutions). Finally, we measured the concentration of Mb in theheart and skeletal muscle and determined O2 equilibrium curves ofpurified Mb.

2. Materials and Methods

2.1. Fish

Specimens of Monopterus albus (Zuiew, 1793) were obtainedfrom a local aquaculture facility in Can Tho City (Vietnam) andtransported to Aarhus University (Denmark). They were kept inlarge aquaria at 27 ± 0.5 °C, fed to satiation with mussels everythird day and were acclimated for three months prior to theexperiments.

2.2. Surgical procedure and experimental protocol

Seven eels with a mean body mass of 145 ± 25 g (mean ± s.e.m)were anesthetized with benzocaine (0.3 g L-1, dissolved in a smallvolume of acetone) and transferred to an operating table when reflexesto pinching subsided. The dorsal aorta was accessed ventrally through a5 cm incision anterior to the anus and cannulated with a PE50 cathetercontaining heparinized saline (50 IU mL-1). The fish were allowed torecover undisturbed for 24 h at 27 °C in normoxic water to allowblood gases to return to normal values before blood sampling. On thefollowing day, the arterial catheter was extended and after having leftthe fish undisturbed for another 2 h, a 0.4 mL blood was drawn forin vivo measures of concentration of oxygen bound to Hb ([Hb-O2]),Hct, pH and total concentration of carbon dioxide in plasma ([CO2]pl)as described below (Sections 2.5 and 2.6). Thereafter, an additional1.5mL bloodwas sampled for measurement of in vitro blood O2 equilib-rium curves.

2.3. Hemoglobin purification

RBC were separated from plasma by centrifugation, washed 5 timesin 0.9% NaCl and lysed in 4-fold volume of ice-cold distilled water.1mol L-1 Hepes (pH 7.4)was added to reach a final buffer concentrationof 10 mM and the mixture was left on ice for 15 min. The hemolysatewas centrifuged at 8,100 g for 10 min. The supernatant containingthe Hb was collected and Hb heterogeneity was evaluated onisoelectrofocusing (IEF) polyacrylamide gels (GE-Healthcare, pHgradient 3-10). To strip Hb from organic allosteric effectors, thehemolysate was mixed with AG 501-X8 mixed bed resin, centrifugedfor 10 min at 10,000 rpm and the supernatant dialyzed against10 mmol L-1 Hepes pH 7.4 for 24 h. IsoHb composition was moreoverinvestigated by preparative IEF on a 110 mL (Amersham Biosciences,type 8102) column as previously described (Larsen et al., 2003), usingAmersham ampholytes in pH ranges of 3-10.5, 5-8 and 6.7 – 7.7 (10,30 and 60%, respectively).

2.4. Myoglobin purification

Heart and muscle tissue were dissected out of euthanized fish,washed in ice-cold saline, snap-frozen in liquid nitrogen and stored at-80 °C until further use.

Mb concentrations were measured using a modified version of themethod developed by Reynafarje (1963). Heart ventricle and muscletissues were homogenized for 1 min in 40 mmol L-1 phosphate,pH 6.6 buffer (19.25 mL buffer/g wet tissue) on ice using anUltra-Turrax T25 homogenizer (IKA, Staufen, Germany). Sampleswere centrifuged (50 min, 15,000 g) and the supernatant collectedand equilibrated with CO gas for 2 min. A pinch of sodiumdithionite was added and the sample was equilibrated with COfor 2 min. Finally, absorbance at 538 and 568 nm was measuredin quartz cuvettes using a HP 8453 UV-visible spectrophotometerand theMb concentrations (mg protein/g wet tissue) were calculated as:

C Mbð Þ ¼ A538nm−A568nmð Þ � 117:3mg g wet tissue−1 ð1Þ

To purify Mb heart ventricles and muscle tissue were homogenizedon ice in buffer (~5 mL/g tissue, 50 mmol L-1 Tris, 0.5 mmol L-1 EDTA,0.5 mg mL-1 DTT, pH 8.3) and centrifuged (Helbo and Fago, 2011). Thesupernatant was collected and submitted to two rounds of ammoniumsulphate precipitation (40 and 80%) followed by desalting through aPD-10 column equilibrated with 50 mmol L-1 Tris, 0.5 mmol L-1 EDTA,0.5 mg mL-1 DTT, pH 8.3. Finally the sample was passed through aTricorn Superdex 75 10/300 GL fast protein liquid chromatography(FPLC) gel filtration column (Amersham Biosciences) equilibratedwith 50 mmol L-1 Tris, 0.5 mmol L-1 EDTA, 0.5 mg mL-1 DTT,150 mmol L-1 NaCl, pH 8.3 at a flow rate of 0.7 mL min-1, to separate

Table 1Arterial blood parameters inM. albus.

Hct 42.4 ± 4.5pH⁎ 7.5 ± 0.03[Hb-O2] (mM) 7.5 ± 1.2O2 saturation (%) 69.3 ± 5.4[CO2] (mM)⁎ 18.8 ± 1.6[HCO3

- ] (mM)⁎ 18.2 ± 1.6PCO2 (mmHg) 24.1

Data are mean ± s.e.m (n = 6).⁎ Indicate plasma values.

104 C. Damsgaard et al. / Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

Mb from contaminants. Protein purity was assessed by SDS-PAGE(GE-Healthcare, Phast gel 20%) and IEF (GE-Healthcare, pH gradient3-9). Prior to O2 equilibrium measurements the purified Mb wasreduced to the ferrous form by addition of sodium dithionite crystalsfollowed immediately by buffer exchange using Amicon Ultra centrifu-gal devices (3× 25min, 3000 g) in 50mmol L-1 Tris, 0.5mmol L-1 EDTA,0.5 mg mL-1 DTT, pH 8.3.

2.5. O2 equilibrium curves

Blood O2 equilibrium curves were determined using an Eschweilertonometer at 27 °C coupled to two serial-linked Wösthoff gas pumps(Bochum, Germany) for mixing pure N2 and O2 to generate water-saturated gas mixtures with known PO2 values. Initially, blood wasequilibrated with 30% O2 (PO2 ~ 217 mmHg) for 40 min to determinethe its O2 carrying capacity ([O2]total), while PCO2 (partial pressureof carbon dioxide) was kept constant at 1% (7.3 mmHg) or 3%(23.0 mmHg). Subsequently, [Hb-O2] was measured in the 20-80%saturation range by lowering the O2 fraction in the equilibration gasmixtures to 1.68%, 1.05%, 0.42% and 0.21% (PO2 tensions of 1.5, 3.1, 7.1and 12.3 mmHg). Blood was equilibrated to each O2 level for a mini-mum of 20 min. Blood O2 concentration was measured as describedby Tucker (1967) and blood O2 saturation (Y) was calculated as thefractional [Hb-O2] relative to the [Hb-O2] value at 30% O2, which wascalculated as:

Hb−O2½ � ¼ O2½ �total−αO2 ; blood � PO2 ð2Þ

where αO2,blood is the blood O2 solubility (Christoforides and Hedley-Whyte, 1969).

O2 equilibrium curves of Mb and purified Hbwere determined usinga modified diffusion chamber (Weber, 1981, 1992), where a thin smear(4 μl) of Mb/Hb solution was placed in a diffusion chamber equippedwith a photomultiplier (model RCA 931-A) and an Eppendorf model1100 M photometer coupled to a potentiometric linear recorder.Water-saturated mixtures of O2 or air and ultrapure (N99.998%)nitrogen gas (N2) prepared by two serial-coupled Wösthoff (Bochum,Germany) gas mixing pumps were allowed to equilibrate the sample.PO2 was increased in steps while light absorption was recorded at436 nm. Zero and 100%O2 saturation levelswere obtained by equilibrat-ing with pure N2 and O2, respectively. Mb O2 equilibrium curves weremeasured in 50 mmol L-1 Tris buffer, 0.5 mmol L-1 EDTA, 0.5 mg mL-1

DTT, pH 8.3 at 25 °C. Hb O2 equilibrium curves were determined atpH values in the range 6.9-7.8 in 100 mmol L-1 Hepes buffer at 25 °C.

For O2 equilibrium curves P50 and n50 (partial pressure of oxygen athalf saturation and cooperativity coefficient, respectively) were calcu-lated from the zero-intercept and slope, respectively, of Hill plots(log(Y)/(1-Y) vs. logPO2, where Y is the fractional saturation, based on≥4 saturation steps. O2 equilibrium curves were derived from the Hillequation using the estimated P50 and n50 values:

Y ¼ POn502

POn502 þ Pn50

50

ð3Þ

The pH-sensitivity of O2 affinity was quantified as the Bohr factor(φ):

ϕ ¼ ΔlogP50

ΔpHð4Þ

2.6. Measurements and calculations of blood parameters

Arterial O2 saturation was calculated as the arterial [Hb-O2] relativeto the O2 carrying capacity for the individual fish. Plasma pH and [CO2]plas well as Hct were measured at 100% and 20-30% O2 saturation during

equilibration with 1% and 3% CO2. pH was measured on whole bloodusing a Radiometer BMS MK2 system with a microcapillary electrode(G299A) (Radiometer, Copenhagen, Denmark). [CO2]pl was determinedas described by Cameron (1971). Hct wasmeasured as volume percent-age of RBC after centrifugation of blood (12,000 rpm for 3 min.).All parameters were measured in duplicate. Plasma bicarbonate con-centrations [HCO3

-] were calculated as:

HCO−3½ � ¼ CO2½ �pl−αCO2

� PCO2 ð5Þ

where αCO2 is the temperature-compensated solubility of CO2 in troutplasma (Boutilier et al., 1985) and PCO2 is the partial pressure of CO2

in the gas mixture prepared by the Wösthoff pump. pK´ values werecalculated from the Henderson-Hasselbach equation as:

pK 0 ¼ pH− logHCO−

3½ �αCO2

� PCO2

!ð6Þ

In vivo PCO2 was calculated using the pH-compensated pK´ values:

PCO2 ¼ CO2½ �plαCO2

� 1þ 10pH−pK 0� � ð7Þ

3. Results

3.1. Blood oxygen binding

The in vivo arterial blood parameters for M. albus listed in Table 1reveal a high arterial PCO2 and low arterial Hb-O2 saturation. Bloodhas a high Hct (42.4 ± 4.5%) and binds O2 with low cooperativity(n50 = 1.3 ± 0.05) and high affinity (P50 = 2.8 ± 0.4 mmHg and4.1 ± 0.5 mmHg at pH 7.7 ± 0.1 (1% CO2) and 7.5 ± 0.1 (3% CO2),respectively) (1 mmHg= 133 Pa) (Fig. 1). The blood Bohr factor calcu-lated from the O2 equilibrium curves measured at 3% CO2 and 1% CO2

equals -0.79 (Fig. 1) compared to the Bohr factors of -0.27, -0.20 and-0.35 observed in the unstripped Hb and the stripped Hb in the absenceand presence of 100 mmol L-1 KCl, respectively (Figs. 2, 3).

An arterial O2 equilibrium curve derived from Eq. (3) and P50 andn50-values at the measured arterial pH (Table 1, Fig. 1) reveals a meanarterial bloodO2 saturation of 69.3± 5.4% corresponding to a calculatedmean arterial PO2 of 6.9 mmHg (Fig. 1). The calculated in vivo PCO2 was24.1 mmHg.

3.2. Hemoglobin heterogeneity and oxygen binding

Remarkably the measurements revealed the same O2 affinity in thestripped (organic effector-free) Hb in the absence and in the presenceof 100 mmol L-1 KCl, and in the unstripped Hb (hemolysate that doescontain organic phosphates) at high pH, but a higher O2 affinity in thestripped Hb in the absence of Cl- at low pH (Fig. 2). Cooperativity ofO2 binding is high in stripped Hb (n50 = 2.5-3.0) compared to wholeblood (n50 = 1.0-1.5).

Fig. 1. O2 equilibrium curves for M. albus blood equilibrated with 1% CO2 (black circles)and 3% CO2 (white circles) at 27 °C showing a high blood O2 affinity (P50 = 2.75 ± 0.4and 4.1 ± 0.5 mmHg for 1% CO2 and 3% CO2, respectively). The curves were generatedby applying average P50 and n50 values (n = 5) to Eq. (3). The red curve represents thearterial O2 equilibrium calculated using P50 and n50-values at arterial blood pH (7.5;Table 1, Eq. (3)). Horizontal and vertical red dashed drop lines indicate arterial O2 satura-tion and PO2, respectively. Horizontal black dotted line indicates half saturation. Datapoints and error bars represent mean ± s.e.m (n = 5).

105C. Damsgaard et al. / Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

Native IEF gels (not shown) indicated the presence of 5 distinctHb isoforms in M. albus. The marked heterogeneity is confirmed bypreparative IEF that revealed 6 isoHb components (Hbs I – VI inFig. 3A) with isoelectric points (pI) between 7.15 and 7.36 (Fig. 3A),indicating that M. albus lacks cathodic components. In accordancewith this heterogeneity pattern of multiple anodic components, thetwo major isoHbs, Hb I and Hb V, show virtually identical O2 affinities(P50 = 4.8 – 5.2 mmHg at 25 °C and pH 7.7) and similar, normal Bohreffects (φ = −0.24 and −0.21, respectively). The isolated Hbsshow the same cooperativity (n50 ~ 2.7) as in the composite Hb(stripped hemolysate). Strikingly, increasing CO2 tensions to 14.7 and44.1 mmHg (2% and 6% CO2, respectively) disclosed a distinct specific,pH-independent CO2 effect (Fig. 3B). The effect was greater at pH 7.7

Fig. 2. pH dependence of cooperativity coefficient (n50) (upper panel) and O2 affinity(logP50) (lower panel) for M. albus unstripped hemolysate (closed circles) and strippedHb measured at 25 °C, and in the presence of 100 mmol L-1 HEPES buffer and of100 mmol L-1 KCl (open triangles).

Fig. 3. A. Preparative isoelectric focusing of M. albus Hb showing optical density (O.D.) at540 nm (circles) and pH at 15 °C (triangles) of the collected fractions, and (horizontalbars) the fractions of isoHbs (I to VI) that were pooled. B. pH dependence of P50 and n50values of Hb I (solid symbols) and Hb V (open symbols), measured at 25 °C in 0.1 MHEPES buffer, in the absence of CO2 (squares), and in the presence of 14.7 mm Hg CO2

(inverted triangles) and 44.1 mmHg (pyramidal triangles) (2 and 6% CO2, respectively).

than 7.0 and correspondingly decreased the Bohr effect (φ = −0.18and −0.16 in Hbs I and V, respectively, at PCO2 of 44.1 mmHg) – ascharacterizes carbamino formation (Dahms et al., 1972; Weber et al.,2013).

3.3. Myoglobin oxygen binding

The concentrations of Mb in the heart and skeletal muscle were5.16 ± 0.99 and 1.08 0.19 mg g wet tissue-1, respectively (n = 4).Mb was successfully purified as judged from SDS gels (not shown),with a molecular mass estimated to ~16 kDa. The P50 of M. albusMbwas 1.11± 0.08mmHg and n50 0.96 ± 0.01 consistent with lackingcooperativity of the monomeric Mb (Fig. 4).

3.4. Whole blood buffering

The calculatedpK’-values of CO2 hydration (CO2+H2O⇌H++HCO3-)

were 6.1-6.3. [CO2]pl and were not affected by blood deoxygenationwhen equilibrated with 1% CO2, but increased during deoxygenation

Fig. 4. Hyperbolic fit of O2 equilibrium curve measured at 25 °C of myoglobin fromM. albus that binds O2 non-cooperatively (n50 = 0.96 ± 0.01) and with a high affinity(P50 = 1.11 mmHg). Data points and error bars represent mean ± s.e.m (n = 2).

106 C. Damsgaard et al. / Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

when equilibrated with 3% CO2, indicating the presence of a Haldaneeffect (Table 2). Correspondingly, plasma [HCO3

-] increased 1.3 fold com-pared to the arterial value during 70-80% deoxygenation when equili-brated with 3% CO2, but only showed a moderate increase duringdeoxygenation when equilibrated with 1% CO2 (Table 2). Based onin vitro measurements of pH and calculations of [HCO3

- ], the non-bicarbonate buffering capacity was -22.3 slykes (Δ[HCO3

-] ΔpH-1).

4. Discussion

4.1. Whole blood oxygen affinity

Our study documents thatM. albus has a very high blood O2-affinity(P50 = 2.8 mmHg at pH 7.7 and 27 °C) compared to other species ofSynbranchidae (P50 = 7.0 mmHg at pH 7.8 and 30 °C in Synbranchusmarmoratus (Heisler, 1982), P50 = 7.9 mmHg at pH 7.6 and 30 °C inMonopterus cuchia (Lomholt and Johansen, 1976)), other air-breathingfish with more complex and conceivably more efficient ABO’s(P50 = 10 mmHg at PCO2 = 6 mmHg and 25 °C in Protopterusaethiopicus (Lenfant and Johansen, 1968)), and even compared tohypoxia-acclimated carp (Cyprinus carpio) (P50 = 3.0 mmHg atpH = 7.9 and 20 °C (Weber and Lykkeboe, 1978)). As blood passesthrough the thick respiratory epithelium in the buccopharyngealcavity of M. albus, it is likely that the low PO2 of the blood serves tomaintain a large trans-epithelial PO2-difference facilitating O2

uptake. Thus, the high blood O2 affinity will to some extent counteractthe low diffusion capacity for O2 air-breathing structures (Liem, 1967).Wepropose a relationship between the complexity of the extrabranchialgas-exchange surfaces and blood O2 affinity, where O2 uptake acrosssimple respiratory surface with high trans-epithelial resistance benefits

Table 2In vitro respiratory parameters in blood and plasma equilibrated to O2 tensions of 217 and1.52 mmHg and 1% and 3% CO2.

PO2 = 217 mmHg PO2 = 1.52 mmHg

1% CO2

(7.3 mmHg)3% CO2

(22.0 mmHg)1% CO2

(7.3 mmHg)3% CO2

(22.0 mmHg)

pH⁎ 7.72 ± 0.05 7.51 ± 0.04 7.74 ± 0.03 7.54 ± 0.05[Hb-O2] (mM) 11.2 ± 2.0 9.7 ± 1.5 3.9 ± 1.1 2.1 ± 0.3[CO2] (mM)⁎ 11.1 ± 1.5 18.9 ± 0.9 11.2 ± 0.8 24.7 ± 2.5[HCO3

- ] (mM)⁎ 8.5 ± 0.3 17.9 ± 1.3 9.6 ± 0.5 23.8 ± 3.6Hb-O2 Saturation (%) 100 100 31.5 ± 3.5 22.3 ± 4.9

Data are mean ± s.e.m.⁎ Indicate plasma values.

from high blood O2 affinities, whereas a lower blood O2 affinity is bene-ficial for species with awell-developed ABO, allowing for the full exploi-tation of the high and stable PO2 in air. Species differences in thediffusion resistances in the ABO may accordingly explain some of thevariation in the blood O2 affinity between air-breathing fish and thehigh blood O2 affinity in M. albus.

Based on ourmeasurements of arterial Hb-O2 saturation, we predictthat the very high blood O2 affinity results from the low arterial PO2 ofaround 7 mmHg in normoxic water (see the modelled arterial O2

equilibrium curve in Fig. 1). Thus, even with high blood O2 affinity,diffusion rates may be insufficient to attain full saturation of arterialblood due to the high epithelial resistance and the high Hct. However,blood in the dorsal aorta of M. albus is admixed with venous bloodand it is thus difficult to discern the extent to which the lack of fullarterial blood O2 saturation is due to diffusion limitations of the ABOor venous arterial blood mixing. A previous study on fish reportinghigh O2 carrying capacities in the air-breathing fish Synbranchus andPterygoplichthys (Johansen et al., 1978a) similarly suggests that thereduced saturation results from mixing of afferent blood from the ABOwith deoxygenated venous blood before it enters the circulation.Furthermore, it is suggested that this low arterial O2 saturation necessi-tates high perfusion requirements tomaintain an adequate O2 supply tothe tissues. Arterial blood that is not fully saturated allows O2 transportto occur on the steepest part of the O2 equilibrium curve. Combinedwith a high O2 carrying capacity, this results in a very high O2 capaci-tance of the blood (illustrated as the functional slope of the bloodO2 equilibrium curve) and thereby a proportionally lowered convectiverequirement of the heart.

This study documents higher Hct and [CO2]pl as well as lowerpH compared to a previous study on this species (Iversen et al., 2013).Both studies identify large inter-individual variation in Hct, whichwith the small sample size in this study could result in the variant Hctvalues. The Hct value reported in this study, however, agrees with thehigh Hct found in S. marmoratus and M. cuchia (Graham, 1997).pH was lower in this study compared to the value reported by Iversenet al. (2013), which partly explains the high [CO2]pl resulting fromnon-bicarbonate buffering.

4.2. Effect of allosteric effectors

This study documents comparable O2 affinities for stripped Hb andhemolysate in the physiological pH range, but shows a higher affinityof stripped Hb in the absence of KCl at low pH (Fig. 2). In fish RBC, allo-steric effectors such as ATP and GTP typically bind to Hb and stabilize itslow affinity tense-state conformation (Perutz and Brunori, 1982;Weberet al., 1987; Weber and Jensen, 1988), thus decreasing Hb-O2 affinity.Normally in vertebrates, the P50 for stripped Hb is therefore markedlylower than that in the hemolysate that contains red cell anionicco-factors. Thus, it seems that the insensitivity of M. albus Hb to theseanionic effectors explains the unusually high O2 affinity of the blood.The current data does not allow us to document if Hb is insensitivetowards ATP/GTP, as Cl- that competes for the same binding sitescould mask possible ATP/GTP effects (Damsgaard et al., 2013). TheO2 affinity of the stripped Hb and the isolated major isoHbs was lowerthan that for whole blood, even when taking into account thatintraerythrocytic pH is lower than plasma pH. While this finding is un-expected it is consistent with the lack of a significant effect of allostericeffectors that normally reduce affinity. It may moreover reflect the factthat the O2 affinities of whole blood and purified Hb were determinedby different methods and that the very high affinity renders bothmethods very sensitive to small experimental errors.

Values for cooperativity of O2 binding (n50) show consistent differ-ences between stripped Hb and whole blood. This is also observed forblood andHb in other fish,where n50 valueswere higher for hemolysatecompared to whole blood (Jensen, 1991; Mandic et al., 2009).

107C. Damsgaard et al. / Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

Blood of M. albus exhibits a normal Bohr effect (φ = −0.79) deter-mined from the right-shift of the O2 equilibrium curve (Fig. 1) with in-creasing PCO2. The Bohr effect is high compared to other air-breathingfish and lies in the range expected for water breathing fish (Shartauand Brauner, 2014). Given the very high O2 affinity, the Bohr effectmay play a role in unloading of oxygen in the tissues.

The marked specific (pH-independent) CO2 effect is noteworthy,given the fact that physiologically-significant carbamino formation hasnot been observed in water-breathing fish that have low blood CO2

tensions and the fact that the N-terminal residue of the α-chains(a major CO2 binding site in vertebrate Hbs) commonly is blocked byacetylation in teleost fish Hbs (Kleinschmidt and Sgouros, 1987;Weber, 2000). Supporting this inference, stripped carp Hb exhibits amodest specific CO2 effect, that is reduced by ATP, which competes forbinding at the N-terminal residue of the β-chains, and is obliteratedby GTP that binds more strongly at the same site (Weber andLykkeboe, 1978). Taken together these findings predict that the N-terminal residues of the α-chains of M. albus Hb are free, and that CO2

binding at the β-chain N-terminal is not materially inhibited byphosphate binding, which is consistent with the lacking ATPsensitivity ofM. albusHb. The data call for deeper analysis of themolec-ular mechanism of the CO2 effect in swamp eels, and its occurrence andbiological significance in air-breathing fish generally.

4.3. Myoglobin concentration and oxygen binding

To maintain a sufficient O2 flux in the face of a low off-loading PO2,M. albus should have a high Mb-O2 affinity and/or high Mb concentra-tion compared to other fish. The P50 of M. albus Mb lies within therange expected for vertebrate Mbs (Nichols and Weber, 1989; Helboet al., 2013). However, Mb is found in very high concentration in boththe heart and the skeletal muscle (5.2 ± 1.0 and 1.1 ± 0.2 mg g wettissue-1, respectively) compared to other, non-tuna fish species(Giovane et al., 1980). Mb facilitates intra-cellular O2 transportand a high Mb concentration will thus decrease the diffusion resis-tance between blood and mitochondria. This illustrates that all levelsin the O2 transport cascade respond to changes in blood O2 affinity. Asa high resistance step is introduced into the O2 transport system(the buccopharangeal epithelial O2 exchange), resistance-counteringadaptations must be expected, exemplified here by the possession ofhigh blood O2 affinity and Mb being present in high concentrations.Mb also serves as a cellular O2 store that, in combination with highblood Hct, may contribute to M. albus‘s significant anoxia toleranceof more than 30 min at 30 °C (Iversen et al., 2013).

4.4. Whole blood buffer effect

Air-breathing is typically associated with a rise in blood PCO2

because the ventilation requirements are reduced due to the high O2

capacitance in air compared to water (Dejours, 1975). This is illustratedby the high blood PCO2 of 24.1 mmHg in vivo, which is similar to that ofother air-breathing fish but higher than most water breathing fish(Shartau and Brauner, 2014). A high buffering capacity of the bloodstabilizes blood pH and the high non-bicarbonate buffering capacity of-22.3 slykes in M. albus is consistent with the high Hct (Brauner andBerenbrink, 2007).

5. Conclusions

The high blood and Hb-O2 affinities in M. albus appear to compen-sate for the low diffusion capacity of the extra-branchial gas exchangestructures, while the high Mb-concentration will serve to maintainsteep a PO2 gradient in the second diffusion step of the O2 transportsystem to secure O2 transport in the face of high ABO diffusion resis-tance. This study identifies novel respiratory adaptations associatedwith the evolution of air-breathing in fish and shows how adaptations

at different steps in the O2 transport system are integrated to cope withphysiological complications associated with primitive air-breathingstructures.

Acknowledgements

This study was supported by DANIDA and the Danish Council forIndependent Research, Natural Sciences (grant 10-084565 to AF).

References

Boutilier, R.G., Iwama, G.K., Heming, T.A., Randall, D.J., 1985. The apparent pK of carbonicacid in rainbow trout blood plasma between 5 and 15 C. Respir. Physiol. 61, 237–254.

Brauner, C.J., Berenbrink, M., 2007. Gas Transport and Exchange. In: McKenzie, David J.,A.P.F., C.J.B. (Eds.), Fish Physiology, Primitive Fishes. Academic Press, pp. 213–282.

Brauner, C.J., Wang, T., 1997. The optimal oxygen equilibrium curve: A comparisonbetween environmental hypoxia and anemia. Am. Zool. 37, 101–108.

Cameron, J.N., 1971. Rapid method for determination of total carbon dioxide in smallblood samples. J. Appl. Physiol. 31, 632–634.

Carter, G.S., 1931. Aquatic and Aerial Respiration in Animals. Biol. Rev. 6, 1–35.Christoforides, C., Hedley-Whyte, J., 1969. Effect of temperature and hemoglobin

concentration on solubility of O2 in blood. J. Appl. Physiol. 27, 592–596.Dahms, T., Horvath, S.M., Luzzana, M., Rossi-Bernardi, L., Roughton, F.J.W., Stella, G., 1972.

The regulation of oxygen affinity of human haemoglobin. J. Physiol. Lond. 223,29P–31P.

Damsgaard, C., Storz, J.F., Hoffmann, F.G., Fago, A., 2013. Hemoglobin isoform differentiationand allosteric regulation of oxygen binding in the turtle, Trachemys scripta. Am. J.Physiol. Regul. Integr. Comp. Physiol. 305, R961–R967.

Dejours, P., 1975. Principles of comparative respiratory physiology. North-Holland Pub. Co.Giovane, A., Greco, G., Maresca, A., Tota, B., 1980. Myoglobin in the heart ventricle of tuna

and other fishes. Experientia 36, 219–220.Graham, J.B., 1997. Air-breathing fishes: evolution, diversity, and adaptation. Academic

Press, San Diego.Heisler, N., 1982. Intracellular and extracellular acid-base regulation in the tropical

fresh-water teleost fish Synbranchus marmoratus in response to the transition fromwater breathing to air breathing. J. Exp. Biol. 99, 9–28.

Helbo, S., Fago, A., 2011. Allosteric modulation by S-nitrosation in the low-O2 affinitymyoglobin from rainbow trout. Am. J. Physiol. 300, R101–R108.

Helbo, S., Weber, R.E., Fago, A., 2013. Expression patterns and adaptive functional diversityof vertebrate myoglobins. Biochim. Biophys. Acta Protein Proteomics 1834,1832–1839.

Hlastala, M.P., Berger, A.J., 2001. Physiology of Respiration. Oxford University Press.Hughes, G.M., 1972. Morphometrics of fish gills. Respir. Physiol. 14, 1–25.Iversen, N.K., Lauridsen, H., Huong, D.T.T., Van Cong, N., Gesser, H., Buchanan, R., Bayley,M.,

Pedersen, M., Wang, T., 2013. Cardiovascular anatomy and cardiac function in theair-breathing swamp eel (Monopterus albus). Comp. Biochem. Physiol. 164A, 171–180.

Jensen, F.B., 1991. Multiple strategies in oxygen and carbon dioxide transport byhaemoglobin. Physiol. Strateg. Gas Exch. Metab. , pp. 55–78.

Johansen, K., Mangum, C.P., Lykkeboe, G., 1978a. Respiratory properties of the blood ofAmazon fishes. Can. J. Zool. 56, 898–906.

Johansen, K., Mangum, C.P., Weber, R.E., 1978b. Reduced blood O2 affinity associated withair breathing in osteoglossid fishes. Can. J. Zool. 56, 891–897.

Kleinschmidt, T., Sgouros, J.G., 1987. Hemoglobin sequences. Biol. Chem. Hoppe Seyler368, 579–615.

Krogh, A., Leitch, I., 1919. The respiratory function of the blood in fishes. J. Physiol. 52,288–300.

Larsen, C., Malte, H., Weber, R.E., 2003. ATP induced reverse temperature effect iniso-hemoglobins from the endothermic porbeagle shark (Lamna nasus). J. Biol.Chem. 278 (33), 30741–30747.

Lefevre, S., Findorf, I., Bayley, M., Huong, D.T.T., Wang, T., 2014. Acclimation to near-futuretemperatures improves temperature tolerance, but not cardiovascular and respiratoryfunctions of the air-breathingAsian swampeelMonopterus albus. J. Fish Biol. (in review).

Lenfant, C., Johansen, K., 1968. Respiration in the African lungfish Protopterus aethiopicus I.Respiratory properties of blood and normal patterns of breathing and gas exchange.J. Exp. Biol. 49, 437–452.

Liem, K.F., 1967. Functional Morphology of the Integumentary, Respiratory, and DigestiveSystems of the Synbranchoid Fish Monopterus albus. Copeia 1967, 375–388.

Lomholt, J.P., Johansen, K., 1976. Gas exchange in the amphibious fish, Amphipnous cuchia.J. Comp. Physiol. 107, 141–157.

Mandic, M., Todgham, A.E., Richards, J.G., 2009. Mechanisms and evolution of hypoxiatolerance in fish. Proc. R. Soc. B Biol. Sci. 276, 735–744.

McCutcheon, F.H., Hall, F.G., 1937. Hemoglobin in the Amphibia. J. Cell. Comp. Physiol. 9,191–197.

Nichols, J.W., Weber, L.J., 1989. Comparative oxygen affinity of fish and mammalianmyoglobins. J. Comp. Physiol. B 159, 205–209.

Perutz,M.F., Brunori, M., 1982. Stereochemistry of cooperative effects in fish and amphibianhaemoglobins. Nature 299, 421–426.

Powers, D.A., 1972. Hemoglobin adaptation for fast and slow water habitats in sympatriccatostomid fishes. Science 177, 360–362.

Powers, D.A., Fyhn, H.J., Fyhn, U.E., Martin, J.P., Garlick, R.L., Wood, S.C., 1979. A comparativestudy of the oxygen equilibria of blood from 40 genera of Amazonian fishes. Comp.Biochem. Physiol. 62A, 67–85.

108 C. Damsgaard et al. / Comparative Biochemistry and Physiology, Part A 178 (2014) 102–108

Reynafarje, B., 1963. Simplified method for the determination of myoglobin. J. Lab. Clin.Med. 61, 138.

Rosen, D.E., Greenwood, P.H., 1976. A fourth Neotropical species of synbranchid eel andthe phylogeny and systematics of synbranchiform fishes. Bull. Am. Mus. Nat. Hist.157, 1–70.

Shartau, R.B., Brauner, C.J., 2014. Acid–base and ion balance in fishes with bimodalrespiration. J. Fish Biol. 84, 682–704.

Taylor, J., 1831. On the Respiratory Organs and Air Bladder of Certain Fishes of the Ganges.William Blackwood.

Tucker, V.A., 1967.Method for oxygen content and dissociation curves onmicroliter bloodsamples. J. Appl. Physiol. 23, 410–414.

Wang, T., Malte, H., 2011. O2 Uptake and Transport: The Optimal P50. In: Farrell, Anthony P.(Ed.), Encyclopedia of Fish Physiology. Academic Press, San Diego, pp. 893–898.

Weber, R.E., 1981. Cationic control of O2 affinity in lugworm erythrocruorin. Nature 292,386–387.

Weber, R.E., 1990. Functional significance and structural basis of multiple hemoglobinswith special reference to ectothermic vertebrates. In: Truchot, J.-P., Lahlou, B. (Eds.),Animal Nutrition and Transport Processes. 2. Transport, Respiration and Excretion:Comparative and Environmental AspectsMol. Comp. Physiol.. vol. 6. S. Karger, Basel,pp. 58–75.

Weber, R.E., 1992. Use of ionic and zwitterionic (Tris/BisTris and HEPES) buffers in studieson hemoglobin function. J. Appl. Physiol. 72, 1611–1615.

Weber, R.E., 2000. Adaptations for oxygen transport: Lessons from fish hemoglobins. In:Di Prisco, G., Giardina, B., Weber, R.E. (Eds.), Hemoglobin Function in Vertebrates,Molecular Adaptation in Extreme and Temperate Environments. Springer-VerlagItalia, Milano, pp. 23–37.

Weber, R.E., de Wilde, J.A.M., 1976. Multiple haemoglobins in plaice and flounder andtheir functional properties. Comp. Biochem. Physiol. 54B, 433–437.

Weber, R.E., Fago, A., 2004. Functional adaptation and its molecular basis in vertebratehemoglobins, neuroglobins and cytoglobins. Respir. Physiol. Neurobiol. 144, 141–159.

Weber, R.E., Jensen, F.B., 1988. Functional adaptations in hemoglobins from ectothermicvertebrates. Annu. Rev. Physiol. 50, 161–179.

Weber, R.E., Lykkeboe, G., 1978. Respiratory adaptations in carp blood influences ofhypoxia, red cell organic phosphates, divalent cations and CO2 on hemoglobin-oxygenaffinity. J. Comp. Physiol. 128, 127–137.

Weber, R.E., Lykkeboe, G., Johansen, K., 1976. Physiological properties of eelhaemoglobin: Hypoxic acclimation, phosphate effects and multiplicity. J. Exp.Biol. 64, 75–88.

Weber, R.E., Jensen, F.B., Cox, R.P., 1987. Analysis of teleost hemoglobin by Adair andMonod-Wyman-Changeux models. Effects of nucleoside triphosphates and pH onoxygenation of tench hemoglobin. J. Comp. Physiol. 157B, 145–152.

Weber, R.E., Fago, A., Malte, H., Storz, J.F., Gorr, T.A., 2013. Lack of conventional oxygen-linked proton and anion binding sites does not impair allosteric regulation of oxygenbinding in dwarf caiman hemoglobin. Am. J. Physiol. Regul. Integr. Comp. Physiol. 305(3), R300–R312.

Wittenberg, J.B., Wittenberg, B.A., 2003. Myoglobin function reassessed. J. Exp. Biol. 206,2011–2020.

Wu, H.W., Kung, C.C., 1940. On the accessory respiratory organ of Monopterus. Sinensia11, 59–67.

Wu, H.W., Lui, C.K., 1943. The bucco-pharyngeal epithelium as the principal respiratoryorgan in Monopterus javanensis. Sinensia 221–239.