Insights Into the Evolution of Lungs and Swim Bladders Author(S): Christopher B

Home , Atlantic tarpon, Ginglymodi, Pond loach, Swim bladder

Division of Comparative Physiology and Biochemistry, Society for Integrative and Comparative Biology

The Origin and Evolution of the Surfactant System in Fish: Insights into the Evolution of Lungs and Swim Bladders Author(s): Christopher B. Daniels, Sandra Orgeig, Lucy C. Sullivan, Nicholas Ling, Michael B. Bennett, Samuel Schürch, Adalberto Luis Val, and Colin J. Brauner Source: Physiological and Biochemical Zoology, Vol. 77, No. 5 (September/October 2004), pp. 732-749 Published by: The University of Chicago Press. Sponsored by the Division of Comparative Physiology and Biochemistry, Society for Integrative and Comparative Biology Stable URL: http://www.jstor.org/stable/10.1086/422058 . Accessed: 08/11/2015 22:31

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The University of Chicago Press and Division of Comparative Physiology and Biochemistry, Society for Integrative and Comparative Biology are collaborating with JSTOR to digitize, preserve and extend access to Physiological and Biochemical Zoology.

http://www.jstor.org

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 732

The Origin and Evolution of the Surfactant System in Fish: Insights into the Evolution of Lungs and Swim Bladders*

Christopher B. Daniels1,† presence of a surfactant system and suggest that this system Sandra Orgeig1 may have originated in epithelial cells lining the pharynx. Here Lucy C. Sullivan1,‡ we present new data on the surfactant system in swim bladders Nicholas Ling2,§ of three teleost fish (the air-breathing pirarucu Arapaima gigas Michael B. Bennett3 and tarpon Megalops cyprinoides and the non-air-breathing Samuel Schu¨rch4 New Zealand snapper Pagrus auratus). We determined the pres- Adalberto Luis Val5 ence of surfactant using biochemical, biophysical, and mor- Colin J. Brauner6 phological analyses and determined homology using immu- 1School of Earth and Environmental Sciences, University of nohistochemical analysis of the surfactant proteins (SPs). We Adelaide, Adelaide, South Australia 5005, Australia; relate the presence and structure of the surfactant system to 2Department of Zoology, University of Auckland, Auckland, those previously described in the swim bladders of another New Zealand; 3School of Biomedical Sciences, Department of teleost, the goldfish, and those of the air-breathing organs of Anatomy and Developmental Biology, University of the other members of the Osteichthyes, the more primitive air- Queensland, St. Lucia, Queensland 4072, Australia; breathing Actinopterygii and the Sarcopterygii. Snapper and 4Department of Physiology and Biophysics, University of tarpon swim bladders are lined with squamous and cuboidal Calgary, Calgary, Alberta T2N 4N1, Canada; 5Instituto epithelial cells, respectively, containing membrane-bound la- Nacional de Pesquisas da Amazoˆnia (INPA), Manaus, mellar bodies. Phosphatidylcholine dominates the phospholipid Amazonas 69083, Brazil; 6Department of Zoology, University (PL) profile of lavage material from all fish analyzed to date. of British Columbia, 6270 University Boulevard, Vancouver, The presence of the characteristic surfactant lipids in pirarucu British Columbia V6T 1Z4, Canada and tarpon, lamellar bodies in tarpon and snapper, SP-B in tarpon and pirarucu lavage, and SPs (A, B, and D) in swim Accepted 11/19/03 bladder tissue of the tarpon provide strong evidence that the surfactant system of teleosts is homologous with that of other Online enhancement: color figure. fish and of tetrapods. This study is the first demonstration of the presence of SP-D in the air-breathing organs of nonmammalian species and SP-B in actinopterygian fishes. The extremely high cholesterol/disaturated PL and cholesterol/PL ra- ABSTRACT tios of surfactant extracted from tarpon and pirarucu bladders Several times throughout their radiation fish have evolved either and the poor surface activity of tarpon surfactant are charac- lungs or swim bladders as gas-holding structures. Lungs and teristics of the surfactant system in other fishes. Despite the swim bladders have different ontogenetic origins and can be paraphyletic phylogeny of the Osteichthyes, their surfactant is used either for buoyancy or as an accessory respiratory organ. uniform in composition and may represent the vertebrate Therefore, the presence of air-filled bladders or lungs in dif- protosurfactant. ferent groups of fishes is an example of convergent evolution. We propose that air breathing could not occur without the

* This article was presented at the symposium “How to Live Successfully on Land If One Is a Fish: The Functional Morphology and Physiology of the Introduction Vertebrate Invasion of the Land,” Sixth International Congress of Comparative Physiology and Biochemistry, Mount Buller, Victoria, Australia, 2003. Many species of fishes can breathe air. Fish utilize a number †Corresponding author; e-mail: [email protected]. of different structures for aerial gas exchange, including the ‡Present address: Department of Microbiology and Immunology, University of skin, gills, mouth and buccal cavity, intestine, and other spe- Melbourne, Victoria 3010, Australia. cifically evolved chambers. In particular, lungs appeared in- § Present address: Department of Biological Sciences, University of Waikato, dependently several times. Lung structure can vary from the Private Bag 3105, Hamilton, New Zealand. simple, transparent, baglike structures of rope fish and bichirs Physiological and Biochemical Zoology 77(5):732–749. 2004. ᭧ 2004 by The (Polypteriformes, Cladista) to more complex compartmental- University of Chicago. All rights reserved. 1522-2152/2004/7705-3061$15.00 ized structures in lungfish (Dipnoi). Lungs are likely to have

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Surfactant in Air-Breathing Organs of Fish 733

first appeared in the placoderms. Moreover, the possible pres- Sullivan et al. 1998; Prem et al. 2000; Bourbon and Chailley- ence of lungs in placoderm fossils indicates that these animals Heu 2001) and gas exchange (gar; Smits et al. 1994), and also must also have had a breathing mechanism and possibly a in lunged fishes (rope fish, bichirs, and the three species of respiratory oscillator for responding to changes in blood and sarcopterygian lungfishes; Smits et al. 1994; Orgeig and Daniels lung gases. Recent evidence suggests that such mechanisms 1995; Sullivan et al. 1998). must have been in place before the lungs themselves evolved In this article, we use previously published biochemical and (Perry et al. 2001). However, placoderm lungs are not ho- morphological analyses of the surfactant system of primitive mologous with the original Osteichthyian lungs, because pla- fish and add new information on this system in teleosts (two coderm lungs were most likely derived from the anterior phar- freshwater-inhabiting air breathers, the pirarucu A. gigas and ynx (Denison 1941; Perry et al. 2001). Moreover, placoderm the tarpon M. cyprinoides, and one marine fish, the snapper lungs may have served a different function. They may have Pagrus auratus). This review (with new data) demonstrates that been important in promoting neutral buoyancy for these the surfactant system of fishes is most likely homologous with heavily armored fish. With the extinction of the placoderms, that of the tetrapods, despite the different ontogenetic origins lungs disappeared in the stem group and are lacking in all and functional aspects of the air-breathing organs (Rubio et al. cartilaginous fishes. Chondrichthyians are less dense than pla- 1996; Sullivan et al. 1998; Prem et al. 2000). In addition, the coderms, because they lack armor plates, and they use squalene high-cholesterol, low–saturated phospholipid composition of in their liver for buoyancy (Hickman et al. 2001). However, it surfactant found in fishes and in the ancient sarcopterygian, is likely that lungs reappeared in the stem ancestor of the Os- the Australian lungfish (Neoceratodus forsteri), represents a very teichthyes. In both branches of the Osteichthyes (Sarcopterygii, primitive, poorly surface-active mixture, which we term a “pro- i.e., lungfish and the most primitive Actinopterygii, i.e., tosurfactant” (Daniels and Skinner 1994; Daniels et al. 1995a; Polypteriformes), the lungs appear as paired ventral structures Orgeig and Daniels 1995; Daniels and Orgeig 2001). A prim- derived from the posterior pharynx and posterior to the gills itive, if not the original, function of fish surfactant is likely to (reviewed in Perry et al. 2001). Primitive Devonian ostei- be acting as an antiadhesive (Daniels and Skinner 1994), but chthyian fish may have breathed air in response to low envi- surfactant may also prevent fluid from entering the bladder or ronmental O2 and used the neutral buoyancy that an air-filled lung, prevent oxidative damage to the epithelial lining, and act lung provides to rest at the surface (Dehadrai and Tripathi 1976; as an antiseptic/antibiotic (Daniels and Orgeig 2001). We also Fange 1983). outline the evolution of the surfactant system in the fish, with However, it is now accepted that while lungs occur in the particular emphasis on postulating a mechanism whereby a most primitive fishes and tetrapods, their ontogenetic origin is homologous system could appear in structures derived inde- different from that of swim bladders (Perry et al. 2001). The pendently several times. swim bladder is an unpaired air-holding structure arising dor- sally from the posterior pharynx (Perry et al. 2001). Swim Material and Methods bladders are usually regarded as the primary buoyancy aid and Fish also support hearing by amplifying sound in the Ostariophysi (Johansen 1968). They are, however, employed in breathing in The tarpon Megalops cyprinoides occurs in a wide range of gars (Ginglymodi) and bowfin (Halecomorphi), where they are marine and freshwater habitats, including East Africa, Southeast termed “pulmonoid” because of the arterial supply from the Asia, Japan, Tahiti, Australia’s tropical seas, and the freshwaters sixth branchial arch. In some teleosts, including a basal teleost, of far northern Australia. The fish can tolerate a wide range of the pirarucu (Arapaima gigas), and a derived teleost, the tarpon habitats, including landlocked lagoons and oxbow lakes or oxeye herring (Megalops cyprinoides), the swim bladder, (termed billabongs in Australia), with waters ranging from pH which is supplied by a swim bladder artery, is respiratory (Perry 5.2 to 9.1 and temperatures between 23Њ and 34ЊC (Merrick et al. 2001). and Schmida 1984). It has been reported that M. cyprinoides Pulmonary surfactant plays a crucial role in the physical will die in aquaria if prevented from reaching the surface (Mer- forces acting at the air-liquid interface in the lungs during the rick and Schmida 1984), and the closely related Atlantic tarpon dynamic changes in surface area and volume that occur during (Megalops atlanticus) will regularly ventilate the swim bladder inspiration and expiration. Surfactant consists of disaturated by a characteristic surface roll, even in normoxic waters (Gra- phospholipids (DSP), unsaturated phospholipids (USP), neu- ham 1976). Therefore, air breathing may be essential for this tral lipids (predominately cholesterol), and surfactant proteins species, particularly in water with low oxygen levels (Graham (SPs). Four SPs have been described in mammals: SP-A, SP- 1997). B, SP-C, and SP-D. Surfactants have also been located and The pirarucu Arapaima gigas is one of the largest freshwater described in fish swim bladders, including those that are used fishes in the world, reaching up to 4.5 m and 250 kg (Graham for buoyancy (e.g., in goldfish, eels, and carp; Daniels and 1997). It is an obligate air-breathing teleost from the Amazon Skinner 1994; Orgeig and Daniels 1995; Rubio et al. 1996; that will drown in 10–20 min without access to normoxic air.

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 734 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

The degree of aerial dependence is related to size. After hatch- Snapper (Pagrus auratus). Snapper (bodymass p 500 g) were ing, A. gigas is a water breather but quickly becomes an air captured by line fishing in the Hauraki Gulf, New Zealand. breather by the time it is about 18 mm long, 8–9 d posthatch Fish were killed by stunning and pithing, and a midventral (Graham 1997). To determine whether there is an effect of size incision was made from the pelvic girdle to the vent to expose on surfactant composition of the air-breathing organ, we col- the gut cavity. Both the internal and external surfaces of the lected two size classes, 30–70 g and 550–1,000 g. Unfortunately, swim bladder were fixed before excision to optimize tissue fix- these were the only sizes that could be obtained. Both sizes are ation and maintain the in situ arrangement of the delicate capable of surviving anoxic water by securing all O2 uptake mucosal surface. We ligated the esophagus and removed all from the air, and both groups secure approximately 80% of abdominal organs, taking care not to damage the swim bladder. their metabolic O2 requirement from air in normoxic water (C. The gut cavity was then filled with fixative (2% paraformal- J. Brauner and A. L. Val, personal observations; Stevens and dehyde, 2% gluteraldehyde, 0.1 M sodium cacodylate, 375 Holeton 1978). Clearly, air breathing is essential for this species. mOsm, pH 7.8). In order to fix the inside of the swim bladder The snapper Pagrus auratus is a demersal marine species in without altering its volume, the swim bladder gases were grad- the family Sparidae and is found in New Zealand, Australia, ually replaced with fixative. A 20-mL syringe was half filled and Japan. Species of Sparid snappers comprise major com- with fixative and the syringe needle introduced to the swim mercial aquaculture and wild fisheries in many areas of the bladder lumen through the epaxial muscle. One milliliter of world. Sparids are Perciform fishes possessing a euphysoclistous fixative was introduced, followed by the withdrawal of the same swim bladder, which is solely a hydrostatic organ and does not volume of swim bladder gas. This reciprocal replacement was perform any respiratory function. The swim bladder develops repeated until the lumen was entirely filled. Only material for as an extrusion of the gut wall and is first filled by swallowing TEM was collected for this species. air via the pneumatic duct at around 5 d after hatching. By around 30 d after hatching, the pneumatic duct has closed and Electron Microscopy and Immunohistochemistry the swim bladder loses its connection to the gut. After in situ fixation of the snapper swim bladder (see above) for 1 h, the ventral swim bladder wall was carefully excised Lavage Procedure from the body and cut into small strips of approximately Tarpon (Megalops cyprinoides). Four fish (body mass range: 2 # 5 mm and continued to fix for 24 h in 2% paraformal- 236–540 g) were captured from the Mary River (Northern Ter- dehyde, 2% gluteraldehyde, 0.1 M sodium cacodylate, 375 ritory, Australia) or the Brisbane River (southern Queensland, mOsm, pH 7.8. In the case of the tarpon, sections of the bladder Australia). We killed the fish by bathing them in a solution of wall and associated respiratory tissue bands forming the ac- 1% MS-222 and removed the tissues overlying the swim blad- cessory breathing organ were cut into small pieces and fixed der. A tube 2 mm in diameter was passed into the swim bladder for TEM, as previously described (Sullivan et al. 2001). via the pharyngeal opening, and the air was then sucked out In the tarpon, sections for immunohistochemistry were pre- with a 20 mL syringe. The bladder was lavaged with 5–6 mL pared by freeze substitution and immunolabelling, as previously of ice-cold 0.15 M NaCl. The saline was withdrawn, and the described (Sullivan et al. 1998). The primary antibodies to hu- lavaging process was repeated twice. The lavage was centrifuged man SP-A and SP-D and bovine SP-B were purchased (Chem- at 150 g for 10 min to remove cellular debris and lyophilized. icon Australia, Boronia, Australia). Material for biochemical analysis, transmission electron microscopy (TEM), and immunohistochemistry was collected Surfactant Composition from this species. Methods for measuring total phospholipids, disaturated phos- Pirarucu (Arapaima gigas). Six small fish (33–66 g) and three pholipids, and phospholipid headgroups have been published larger fish (550, 655, and 1,006 g) were obtained from Boutique previously (Daniels and Skinner 1994; Orgeig and Daniels 1995; do Peixe Vivo and Amazonas Ecopeixe S.A. and held at the Daniels et al. 1999). We measured cholesterol in the neutral National Institute for Research of the Amazon on a natural lipid fraction (reconstituted with heptane) with a high-perfor- light cycle for 3 wk before sampling. During this time, fish were mance liquid chromatography (HPLC) system (LC1500 HPLC fed goldfish at a ration of 1% body weight per day. Fish were pump, LC1610 autosampler, DP800 data interface; GBC In- killed in a solution of 1% MS-222, a tube was inserted into struments, Melbourne, Australia). We injected 30 mL onto a the pharyngeal opening of the swim bladder, and lavage was silica column (Alltech4.6 # 250 mm with 5 mm spherical silica) obtained as described above for tarpon. The entire ventral lower equilibrated with a hexane : isopropanol (99 : 1 v/v) mobile half of the swim bladder was removed and weighed before and phase (HPLC grade; APS Chemicals, Greenfields, Australia). after lyophilization. Only lavage material was collected from Samples were eluted (1 mL/min) and measured at 206 nm this species. (Model 2151 LKB Bromma variable-wavelength UV detector).

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Surfactant in Air-Breathing Organs of Fish 735

Figure 1. Appearance of the respiratory portion of the swim bladder of the tarpon Megalops cyprinoides. The surface consists of a thin layer of capillaries (C) and an overlying layer of epithelial cells (EC).AS p airspace .N p nucleus . Scale bar p 10 mm.

A backpressure regulator (100 psi) was fitted to the detector (140 mM NaCl, 10 mM HEPES, 2.5 mM CaCl2,pH6.9)toa outlet to minimize the formation of bubbles in the flow cell. concentration of 20 mg/mL of phospholipid. Retention time for cholesterol was 10.2 min, and total run time We used a leakproof system, the captive bubble surfactometer per sample was 25 min. (CBS), which enables the determination of surface tension, area, SP-B in lavage samples from tarpon and pirarucu was quan- and volume of a bubble with a surfactant film at the air-water tified in a competitive ELISA based on a previously published interface (Schu¨rch et al. 1992). We used a previously published method (Lesur et al. 1993). The primary antibody, rabbit an- method for the preparation of small sample volumes (Codd et tibovine SP-B polyclonal antibody (Chemicon Australia), and al. 2002). We measured the rate of adsorption for the first 5 the secondary antibody, antirabbit IgG-HRP conjugate (Sigma, min and then performed quasi-static compressions (Codd et Sydney, Australia), were both used at a dilution of 1 : 1,000. al. 2002). From these we determined the minimum surface

The reaction was developed in Sigma Fast tablets (Sigma), and tension upon ﬁnal compression (STmin), the maximum surface the absorbance at 492 nm was determined with a Titertek MCC tension before compression (STmax), and the percent surface

Multiskan (Titertek, Huntsville, Ala.). The amount of SP-B in area of compression (%SAcomp), which is a measure of the extent the lavage samples was compared to the amount of total pro- of compression required to achieve STmin. Measurements were tein, which was determined by the method of Lowry et al. performed both at room temperature (22ЊC) and at 37ЊC. We (1951). recorded the bubble continuously and calculated bubble volume, area, and surface tension from digitized bubble images, using height and diameter (Schoel et al. 1994). We compared Surface Activity these results to values obtained for other ﬁsh species using this Surface activity for an aliquot of lyophilized lavage from one and other techniques. tarpon was determined after the lavage was reconstituted in water and centrifuged at 40,000 g (Beckman Optima TLX Ul- Results tracentrifuge; Beckman, Sydney, Australia) for 30 min to pellet Electron Microscopy and Immunohistochemistry large aggregate material. A phosphorus assay of the pelleted material and the supernatant revealed that 190% of the phos- Tarpon. The respiratory surface of the swim bladder accessory pholipids remained in the supernatant. The supernatant was breathing organ consisted of a thin layer of blood capillaries lyophilized again and reconstituted in a buffered salt solution and an overlying layer of epithelial cells (Fig. 1). There were a

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 736 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

Figure 2. Cuboidal epithelial cells in the lining of the swim bladder of the tarpon Megalops cyprinoides. The cells have a large nucleus and contain either numerous electron-lucent structures (A, arrows, scale bar p 2mm) or electron-dense structures (B, arrows, scale bar p 1 mm). number of cuboidal epithelial cells containing a dark nucleus as simian and common in human Type II alveolar cells, al- and numerous electron-lucent structures (Fig. 2A) that appear though their description of this structure as having concentric to represent one form of lamellar body (Table 1). Other lamellar lamellae may be misleading and appears to depend on the plane bodies appeared as granular, electron-dense structures (Figs. of sectioning and possibly the developmental stage of the cy- 2B,3A), and still others consisted of concentric rings of lamellae tosome (Table 1). For example, a lamellar body with lamellae in a membrane-bound vesicle (Fig. 3B). Different structures fused at the poles may appear concentric if sectioned equa- for the lamellar bodies are also reported for the snapper (see torially. As with the tarpon, several different morphs of cyto- below) and may represent a developmental sequence. At a some were recognized and may represent a developmental se- higher magnification, SP-A, SP-B, and SP-D were located quence based on the assumption that all eruptant cytosomes within the tarpon swim bladder. SP-A and SP-B were located appeared to be of the concentric lamellate form. The proposed both within the cells and in the airspaces. The location of SP- developmental sequence for these cytosomes is described in A appeared to be quite diffuse (Fig. 3A), whereas SP-B was Figure 4C. found to be associated with lamellar bodies (Fig. 3B). A small Cytosomes appear to form from membrane-bound, electron- amount of SP-D staining was found in the air spaces only (Fig. dense aggregations of an amorphous material. Brooks (1970) 3C). reports formation of epithelial cytosomes in trout from similar electron-dense bodies and states that this material appeared to Snapper. The lining of the swim bladder lumen was composed be composed of very tightly packed membranes. The lamellate of squamous epithelial cells. These cells were generally in the structure of these bodies in the snapper could only be distin- region of 200 nm thick, except for a central thickening due to guished close to the site at which lamellae were separating from the presence of the nucleus and adjacent mitochondria. Cells the electron-dense material (Fig. 4C). These osmiophilic la- were joined peripherally by tight junctions at the lumenal sur- mellae begin to form and separate centrally but remain attached face and formed extensive interdigitations with adjoining cells peripherally in an equatorial, or possibly polar, adhesion plaque that extended for 1–2 mm from the position of cellular abut- (Fig. 4D). Total separation of the individual lamellae to form ment. The apical surface of the epithelium was interrupted by a truly concentric structure seems to occur just before eruption occasional microvilli that appeared more common toward the from the apical surface of the cell. cell periphery. This cell type was characterized by the presence The gas gland epithelium contains multilamellar bodies, or of numerous lamellar bodies or cytosomes. Cytosomes ap- cytosomes, of the form previously described for the toadfish peared to be more common in the region of central thickening gas gland (Morris and Albright 1975, 1977) and classified as close to the nucleus (Fig. 4A,4B). The cytosomes of the lumenal subsimian by Creasey et al. (1974). This cytosome has lamellae epithelium were of the form described by Creasey et al. (1974) that traverse from one side of the structure to the other (Fig.

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Table 1: Morphological characteristics of surfactant from fishes (Osteichthyes) Type II Cells Lamellar Tubular Species (and Analogues)a Bodies Myelin Other Forms and Comments References Dipnoi: Protopterus annectens ? ϩ ?1 Protopterus sp. ϩϩ? Osmiophilic inclusion bodies in cells and 2 in alveolar space Lepidosiren paradoxa ϩ(*) ϩ ?(*)Only one cell type 3 L. paradoxa ϩϩϪExtracellular material also present 4 Neoceratodus forsteri ϩ(*) ϩ ?(*)Only one cell type 3 N. forsteri ϩϩ? Type II cells can be isolated; secrete LP 5 Teleostei: Salmo gairdneri ϩϩϪNonciliated (columnar or cuboidal?) cells 6 that secrete mucus-like material into lumen of swim bladder Anguilla vulgaris ? ϩ ?7 Lota lota ϩϩ? Very isolated LBs in cells of the gas 8 gland Acerina cernua ? ϩ ?8 Coregonus lavaretus ? ϩ ?9 Fundulus heteroclitus ? ϩ ?10 Gadus callarias ? ϩ ?10 Opsanus tau ? ϩ ?10 Amphipnous cuchia ϩ(*) ϪϪ(*)Cuboidal cells; air sac instead of lung 11, 12 Anabas tetudineus ϪϪϪLabyrinthine organ 11, 12 Channa punctatus ϩ(*) ϪϪ(*)Cuboidal cells; suprabranchial 11, 12 chamber Channa striatus ϩ(*) ϪϪ(*)Cuboidal cells; suprabranchial 11, 12 chamber Clarias batrachius ϩ(*) ϪϪ(*)Cuboidal cells with vesicles; air sac 11, 12 Heteropneustes fossilis ϩ(*) ϪϪ(*)Cuboidal cells; air sac 12, 13 Misgurnus fossilis ϩ(*) ϩϪ(*)Goblet epithelial cell of respiratory 14 intestine Pagrus auratus ϩϩϩDifferent forms of LBs; they possibly 15 represent a developmental sequence, i.e., only eruptant LBs have concentric lamellae Megalops cyprinoides ϩϩ? Different forms of LBs (electron-lucent, 15 electron-dense, or with concentric lamellae); SPs A, B, D present in Type II cells and airspace Halecomorphi: Amia calva ? ϩ ?3 Polypteriformes: Polypterus senegalensis ϪϪϪElectron-opaque granules in rare cells in 16 secretory crypts Polypterus ornatipinnis ϩϩϪ“Classical-appearing” Type II cells 13 Note. A plus sign indicates presence, a minus sign absence, and a question mark unresolved or not attempted. LB p lamellar body. References: (1) Klika and Janout 1967; (2) DeGroodt et al. 1960; (3) Hughes 1973; (4) Hughes and Weibel 1976; (5) Wood et al. 2000; (6) Brooks 1970; (7) Dorn 1961; (8) Jasinski and Kilarski 1964; (9) Fahlen 1967; (10) Copeland 1969; (11) Munshi 1976; (12) Hughes and Munshi 1973; (13) Marquet et al. 1974; (14) Jasinski 1973; (15) this study; (16) Klika and Lelek 1967. a Asterisks are defined in the first entry in the “Other Forms and Comments” column.

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 738 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

4E). Eruption of this form of cytosome through the apical cell surface was not observed. Simian or concentrically lamellate cytosomes of the form common to the general lumenal epithelium were also observed and seen to occur in eruption from the epithelium. It was notable that eruption of this form of cytosome could occur either from the apical surface onto the surface of the lumen or through the basal surface into underlying circulatory sinuses. Extracellular material was occasionally observed on the lumenal surface of the gas gland, trapped between apical microvilli (Fig. 4F). The transverse structure of this material (Fig. 4G) resembles the cytosomal tubular myelin of the vertebrate lung. As the crosshatched structure of tubular myelin is highly characteristic of surfactant, it is likely that the material in the snapper gas gland is surfactant.

Surfactant Composition Tarpon. The disaturated phospholipid (DSP) content of the lavage material was very low and was measurable in only one of the four tarpon ﬁsh samples (%DSP/PL p 4.04 ). The cholesterol (Chol) to DSP ratio (mg/mg) in this one sample was 5.65. The cholesterol to phospholipid (PL) ratio (mg/mg) in the -SE ,n p 4 ; Table 2). Thin ע mean ) 0.069 ע lavage was0.258 layer chromatography (TLC) revealed that phosphatidylcholine (PC) was the predominant PL in lavage, accounting for 71% of total PL. The only other signiﬁcant lipids were sphingomyelin (18.4%) and a combination of phosphatidylserine and phosphatidylinositol (7.7%). Lysophosphatidylcholine (1%), phos- phatidylglyerol (1%), and phosphatidylethanolamine (0.3%) were present only in trace amounts (Table 2). The ELISA demonstrated the presence of SP-B in lavage from two of the tarpon. There were 2 pg of SP-B per mg of total protein in one animal and 17.6 pg of SP-B per mg of total protein in the other.

Pirarucu. The %DSP/PL in the lavage of the small fish was SE ,n p 6 ). The Chol/PL and Chol/DSP ע mean )3.35 ע 26.85 ע mean ) 0.08 ע and 1.03 0.02 ע ratios (mg/mg) were0.266 SE,n p 6 ), respectively (Table 2). The Chol/PL ratio of the SE , n p ע mean ) 0.114 ע lavage from the large fish was0.34 3) and did not differ from that of the small fish. We did not measure DSP in the large pirarucu. SP-B was measurable by ELISA in the lavage of one of the large pirarucu. The lavage of this fish contained 44.6 pg SP-B per mg of total protein.

Surface Activity The adsorption of surfactant from the tarpon swim bladder was extremely slow. At 22ЊC the surfactant adsorbed slightly

Figure 3. Immunogold labeling of SP-A (A; scale bar p 0.2 mm), SP- B(B; scale bar p 0.5mm), and SP-D (C; scale bar p 0.5 mm) in the arrows), whereas SP-B appeared to be primarily associated with la- respiratory swim bladder of the tarpon Megalops cyprinoides. SP-A was mellar bodies (arrows). The presence of SP-D was restricted only to localized in the air spaces (ﬁlled arrows) and in the epithelial cells (open the airspaces (arrows).

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Figure 4. Multilamellar bodies (cytosomes) and tubular myelin in the snapper (Pagrus auratus) swim bladder. A, Cytosome erupting from lumenal epithelium (scale bar p 1mm). B, Semilamellate cytosome (scale bar p 1 mm). C, Progressive development of cytosomes from electron- dense amorphous granules to semilamellate form (scale bar p 1mm). D, Semilamellate cytosome, lamellae fused equatorially (scale bar p 2 mm). E, Subsimian cytosome (scalebar p 250 nm). F, Lumenal surface material, tubular myelin (scale bar p 250 nm). G, Transverse structure of the material in F (scalebar p 100 nm).

739

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 4 ! DSP/PL (%) lysophosphatidylcholine p phosphatidyl-ethanolamine p LPC DSP (mg/gWL) Chol/DSP phosphatidylglycerol PE p Chol (mg/gWL) Chol/PL not p phosphatidylcholine PG p C); determined. Њ –25 Њ grams PC roomtemperature ND p Percent of PLs PC PG PE SM PI PS LPC/Unk p PL (mg/gWL) phosphatidylserine p phospholipid gWL C) disaturated RT A/B Њ T ( p p 3 2334 .084 23 23 59.6 .22 .5 .61 14.4 15.3 73.9 1.8 10.2 80.2 6.9 3.06 7.3 3.0 04 RT 5.6 9.3 RT 7.9 ND ND .022 .8 ND ND .4 71 ND .242 ND 1 .010 ND .028 .006 .3 .044 18.4 ND .052 7.7 3.66 .059 .126 ND 1 8.9 .169 .266 .222 27.8 ND 20.4 ND .258 1.03 ND 26.85 5.65 n 1–3 RT ND1–3 46.0 8.0 RT1–3 19.0 RT 11.01–3 ND 0 RT1–3 ND RT 0 70 ND 69 0 ND 0 93 0 7 39 0 25 5 6 ND 3 25–8 0 23 7 0 2 12 242–9 ND 2 .50 2 232–8 5 8 23 0 0 ND 76.4 8 .23 1.4 0 .097 4.6 8 ND ND 5.9 ND ND ND ND ND 7.9 ND ND ND ND ND ND ND ND ND ND 3.0 ND ND ND ND ND ND .091 ND ND ND .044 ND .021 .175 ND ND .182 ND ND .208 .069 ND .022 ND .008 1.30 2.05 1.72 14.1 9.54 11.6 d b All use subject to JSTOR Terms and Conditions e cholesterol DSP e p a phosphatidylinositol PS b b c a b p (posterior e ; ; PL. (22 b ; ; (a single value for PI and PS indicates that the bands could not be resolved); ; This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM a ambient PL p c unknown PL Chol A/B T p sphingomyelin PI swim bladder) 9 23 .14 75 1.2 4.7 15.9 5.1 0 .051 .328 .036 1.41 20.8 p Daniels and Skinner 1994. Phleger and Saunders 1978. Orgeig and Daniels 1995. This study. Smits et al. 1994. Hoplias malabaricus Neoceratodus forsteri Lepisosteus osseus Calamoichthys calabaricus L. paradoxa Hoplerythrinus unitaeniatus Erythrinus erythrinus Arapaima gigas A. gigas Polypterus senegalensis Lepidosiren paradoxa Protopterus annectens Megalops cyprinoides Carassius auratus Note.a b c or bodyd temperature;e ; of wet lung mass; ; ; ; Teleostei: Table 2: Surfactant composition in ﬁshes (Osteichthyes) Species Dipnoi: Ginglymodi: Polypteriformes: SM unkwn

740 Surfactant in Air-Breathing Organs of Fish 741

Surfactant Proteins

The presence of the surfactant lipids and proteins (SP-A, SP- B, and SP-D) provides strong evidence that the surfactant system of all fishes is homologous with that of the tetrapods. Furthermore, the surfactant system located in the lungs or swim bladders of fish (lungfish, bichirs, gar, carp, eels, goldfish, snapper, pirarucu, and tarpon) is likely to be homologous (Daniels and Skinner 1994; Rubio et al. 1996; Sullivan et al. 1998; Prem

Figure 5. Rate of adsorption of tarpon (Megalops cyprinoides) surfactant (10 mg/mL PL) to the air-liquid interface of a bubble in the captive bubble surfactometer. The surface tension at the end of adsorption is the equilibrium surface tension (STeq).

Њ faster than at 37 C (Fig. 5), and it reached a slightly lower STeq (22ЊC: 38 mN/m; 37ЊC: 41 mN/m). Upon the ﬁrst quasi-static compression cycle, tarpon surfactant demonstrated relatively Њ Њ high values for STmax (22 C: 41.0 mN/m; 37 C: 41.1 mN/m), Њ Њ STmin (22 C: 19.5 mN/m; 37 C: 21.7 mN/m), and %SAcomp (22ЊC: 75.3%; 37ЊC: 75.8%; Fig. 6). Surface tension–lowering properties, including STmin and %SAcomp, were marginally better at 22Њ than at 37ЊC (Fig. 6). Film stability was poor, as seen from the slight progressive increase in STmin following successive quasi-static cycles (Fig. 6; Table 3).

Discussion

Morphology Cells containing lamellar bodies and surfactant lipids and proteins are easily observable in the swim bladder of the tarpon and the snapper. It is also clear that the function of the gas- holding structure (respiration or buoyancy) is irrelevant to the requirement for a surfactant system in fish (Table 1). The surfactant-secreting cells in the tarpon and the snapper do not Figure 6. Quasi-static (QS) surface tension–area relations for tarpon differ structurally from those of other fishes and possess the (Megalops cyprinoides) surfactant on the captive bubble surfactometer same characteristics as the alveolar Type II epithelial cells found at (A)22ЊC and (B)37ЊC; the first, second, and fourth QS cycles (qs1, in tetrapod lungs (Daniels and Orgeig 2001). They demonstrate qs2, and qs4, respectively) are given for comparison. The lower limb microvilli on the apical surface, and the lamellar bodies are of each curve represents compression and the upper expansion. The lowest point at the end of compression represents ST . The change membrane-bound and of similar size and appearance to those min in surface area compression (%SAcomp) required to reach STmin is an observed in all other vertebrate groups, including fishes (Table indication of the surface tension–lowering quality of the sample. Sam- 1; Hughes 1970, 1973; Hughes and Weibel 1978; Prem et al. ple concentration was 10 mg/mL PL. A color version of this figure is 2000; Wood et al. 2000; Daniels and Orgeig 2001). available in the online edition of the journal.

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 742 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

Table 3: Surface activity of surfactant from fishes (Osteichthyes) Min ST %SA Stability Bubble Surpellic Property/ Њ a Species and TA ( C) Device (mN/m) Comp Ratios Clicking Surface Activity Dipnoi: Protopterus annectens: RT Stableb ϩc Fully surpellicb Lepidosiren paradoxa: RT Calculated from bubble shapeb !.1 Stableb ϩd Fully surpellicb 37 WBd 22 24 WBe 22 70 Very lowe Neoceratodus forsteri: RT ϩf Partially surpellicb Teleostei: Carassius auratus: RT Calculated from bubble shapeb 20 Ϫc Partially surpellicb 22 PBSg 27.4 Very lowe Megalops cyprinoides: 22h CBS 19.5 75.3 Very low 37h 21.7 75.8 Very low Hoplias malobaricus: 37 WBd 22 70e Very lowe Hoplerythrinus unitaeniatus: 37 WBd 22 80e Very lowe Arapaima gigas: 37 WBd 22 75e Very lowe Halecomorphi: Amia calva: RT Calculated from bubble shapeb 4 Ϫc Partially surpellicb Ginglymodi: Lepisosteus osseus: RT Calculated from bubble shapeb 2–6 Ϫc Partially surpellicb Very lowe 70 1.0 ע WBe 17.0 24 p p p Note.TA ambient temperature; Min ST minimum surface tension; %SA Comp percent surface area compression required to achieve min ST; RT p roomtemperature (22Њ–25ЊC). WB p Wilhelmy balance; PBS p pulsating bubble surfactometer; CBS p captive bubble surfactometer. The number of measurements performed regards whole lung extracts, not lavage. “Surpellic” is a term coined by Pattle to describe the ability of lavage material to exist as stable bubbles and demonstrate bubble clicking. “Stability ratio” is a measure of a bubble’s stability as determined by the change in surface area over a standard length of time after stabilisation. “Bubble clicking” is a phenomenon involving rapid changes in bubble shape and surface tension immediately after formation in water. It indicates that the fluid involved is capable of attaining very low surface tension. a A plus sign indicates presence and a minus sign absence. b Pattle 1976. c Hughes 1967. d Phleger and Saunders 1978. e Daniels et al. 1998b. f Hughes 1973. g Daniels and Skinner 1994. h This study. et al. 2000). Given that the lung and swim bladder have separate localize with these structures (Rubio et al. 1995; Engle and ontogenetic origins and arise from different pharyngeal regions Alpers 2001). Furthermore, SP-A antibodies also cross-react (Perry et al. 2001), we can conclude that the surfactant system with material from both the intestine and swim bladder of the must predate the appearance of both lungs and swim bladders. carp (Rubio et al. 1996). We therefore postulate that the sur- Significant recent evidence has demonstrated surfactant- factant system originated in the epithelial cells lining the phar- containing lamellar bodies secreted from epithelial cells in the ynx. Here the SPs may have played an important role in the mammalian gut (Engle and Alpers 2001). SP-A and SP-D co- innate immune system of the organism.

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Surfactant in Air-Breathing Organs of Fish 743

The hydrophilic SPs, SP-A and SP-D, belong to the family factant film (Haagsman and Diemel 2001). It is likely that both of proteins known as collectins that are characterized by C- the biophysical and immune functions of the SPs are important terminal carbohydrate recognition domains. These proteins are in the lungs and swim bladders of fish. Moreover, the cross- thought to function as molecules of the innate immune system reactivity between the mammalian antibodies and the native, by recognizing a broad spectrum of pathogens, including vi- in situ fish proteins implies that the tertiary structures of all ruses, bacteria, and fungi, as well as allergens, such as pollen three proteins are highly conserved. This is remarkable, because grains and mite allergens (Haagsman and Diemel 2001). In in spite of major differences in the structure and function of addition to their presence in gut epithelium, SPs have also been gas-holding structures, both within a lineage and between dis- described in mucosal surfaces of other tissues, for example, the tantly related animals, the structure of these proteins and their middle ear duct, esophagus, stomach, and peritoneal cavity retention as part of the surfactant system have not been affected. (Van Rozendaal et al. 2001), as well as in the genitourinary We conclude that the proteins must be essential to both the tract (Bourbon and Chailley-Heu 2001). Hence, these proteins function of the surfactant system and the operation of lungs may perform a generalized function in mucosal immunity, as and swim bladders. The genes for these proteins, therefore, may these sites are directly exposed to the environment. Mucosal have extremely low mutation rates, or it may be that any mu- immunity involves both infectious defense and tolerance tation is extremely deleterious and is selected out. This evidence against environmental and dietary antigens (Haagsman and powerfully supports the hypothesis that a functional surfactant Diemel 2001). Furthermore, given that the epithelial lining of system is a crucial prerequisite for the evolution of aerial gas these structures arises from different regions of the pharynx, exchange using a lung or swim bladder. mucosal immunity may represent the ancient function of SPs in the pharynx. Surfactant Lipids However, it is not only the SPs that are expressed in gut mucosa. Surfactant-like lipid material enriched in phosphati- The predominant lipid in all vertebrates examined is phos- dylcholine and demonstrating surface tension–lowering capa- phatidylcholine (PC; Daniels and Orgeig 2001), and the disat- bilities has also been reported as lining the gut epithelium urated form of this lipid, dipalmitoylphosphatidylcholine (Engle and Alpers 2001). Although the function of these lipids (DPPC), is usually regarded as responsible for the reduction in is not clear, they are likely to have a role in protecting the surface tension (Veldhuizen et al. 1993). However, the relative underlying epithelial cells from both mechanical and chemical abundance of DSP, USP, and cholesterol varies greatly between injury and may be involved in the binding of molecules for different species (Daniels et al. 1995a). The surfactant of mam- biochemical processing (Engle and Alpers 2001). It is also pos- mals, birds, and reptiles is rich in DSP. However, the surfactant sible that the lipids may have a role in dealing with interfacial in the lungs of the air-breathing, ray-finned fishes (Polypteri- tension of liquids of different viscosities in the gastrointestinal formes and Ginglymodi) and the teleost swim bladders (Fig. tract and other mucosal surfaces (Sullivan et al. 1998). 7; Table 2) have a Chol/DSP ratio that is up to 15 times higher Hence, it is possible that the surfactant system could have than that of the tetrapods (Smits et al. 1994). Similarly, the migrated with the epithelial cells during the out-pouching most primitive of the lobe-finned fishes, the Australian lungfish events associated with the generation of either lungs or swim Neoceratodus forsteri has DSP/PL, Chol/DSP,and Chol/PL ratios bladders from different regions of the pharynx. To our knowl- of the same order as those of the ray-finned, air-breathing fishes edge, no analysis has been performed on the presence of SPs (Daniels et al. 1995a; Orgeig and Daniels 1995; Table 2). How- in other gas-exchange organs of fish, such as the gills, buccal ever, two derived species of lobe-finned fishes, the South Amer- cavity, or specialized respiratory regions of intestine. However, ican and African lungfishes Lepidosiren paradoxa and Protop- given that these structures also originate from the pharynx, it terus annectens, have a surfactant that is between two and three is likely that SPs are present in these structures. times richer in DSP than N. forsteri or the primitive, air- This study is the first demonstration of the presence of SP- breathing, ray-finned fishes (Polypteriformes and Ginglymodi; D in the lungs of nonmammalian species and the first to dem- Orgeig and Daniels 1995; Table 2). The derived lungfish also onstrate SP-B in actinopterygian fishes. SP-B has been previ- have Chol/PL ratios that are between one-third and one-fifth ously detected in chicken and salamander lungs (Zeng et al. those of N. forsteri, the Polypteriformes, and the Ginglymodi. 1998; Miller et al. 2001), and SP-A is well documented in fish Furthermore, on the basis of their Chol/PL and Chol/DSP ra- (Rubio et al. 1996; Sullivan et al. 1998; Prem et al. 2000). The tios, the surfactants of the teleost fishes (pirarucu, tarpon, and functions of the SPs in mammals are not yet fully understood. goldfish) group more closely with those of the primitive, SP-A and SP-D are thought to function predominantly in im- lunged, air-breathing, ray-finned fishes and N. forsteri. Collec- mune defense by promoting the uptake of numerous pathogens tively, these species and groups (N. forsteri, Teleosts, Polypter- by macrophages (Haagsman and Diemel 2001). SP-B, on the iformes, Ginglymodi) have Chol/PL ratios that are between other hand, is intricately involved with the surfactant PLs and three- and sevenfold greater and Chol/DSP ratios that are be- is important in regulating the biophysical activity of the sur- tween four- and 25-fold greater than those in the derived lung-

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 744 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

Figure 7. Schematic diagram of the evolutionary sequence of air-breathing organs among the ﬁshes. The ontogenetic origin (i.e., ventral or dorsal) and the evolution of lungs, swim bladders, and their blood supply and the loss of respiratory function are indicated in italics. Resp. p respiratory;fn p function . Figure modiﬁed from Perry et al. (2001).

fish (L. paradoxa and P. annectens; Table 2). Therefore, on a swim bladders, a highly fluid (i.e., high-Chol, high-USP, low- holistic scale, it appears that the surfactant lipid composition DSP) mixture may assist in the easy spreading of the surfactant of all the ray-finned and primitive lobe-finned fishes is relatively from (presumably isolated) clumps of secretory cells to cover similar, in that they have a high-cholesterol, low-PL, low-DSP the entire surface (Daniels et al. 1995b). Furthermore, such a mixture. On the other hand, the derived lobe-finned fishes (P. mixture is adequate to perform the antiadhesive function that annectens and L. paradoxa) have a surfactant composition much we have demonstrated for goldfish swim bladders, ray-finned, more similar to that of the tetrapods. However, upon closer air-breathing fish (Ginglymodi and Polypteriformes), and many examination it is apparent that among the ray-finned and prim- reptiles and amphibians (Daniels et al. 1995a). It is also possible itive lobe-finned fishes there are some subtle differences in that the unusual combination of lipids, in particular the large surfactant composition that may be related to the function of amount of cholesterol, in fish swim bladder surfactant may the gas-filled organ. For example, the goldfish (Table 2) is not serve as an antioxidant (Szebeni and Toth 1986), as the swim an air breather, and it has the highest Chol/PL ratio of all the bladders of most physoclist teleosts are exposed to hyperbaric fish. Hence, it appears that the less the air-breathing organ is oxygen (Pelster 2001). used for gas exchange, the more cholesterol its surfactant con- There are also differences in the relative abundance of the tains. This high-cholesterol, low-PL, low-DSP mixture may rep- PL head groups between the fishes and the tetrapods. In all resent the protosurfactant (Daniels et al. 1995a). cases, PC is the dominant head group, but there is significant The functions of this mixture will differ significantly from variation in the presence and abundance of the minor PLs. those of the surfactants of tetrapods, because it is poorly surface Surfactant isolated from mammals contains about 10% phos- active and highly fluid (see below). We have previously argued phatidylglycerol (PG) and only trace amounts of phosphati- that given the extremely smooth, nonseptated lungs of the po- dylethanolamine (PE) and sphingomyelin (Daniels and Orgeig lypterid fishes, the smooth swim bladder walls of the goldfish, 2001). PG is relatively uncommon in vertebrate surfactants and the comparatively large respiratory units of the gar, lung- (restricted only to mammals, a few amphibians, the lungfishes, fish, and other teleosts, a highly surface-active mixture may not and one of three snake species that have been examined; Daniels be required. It is possible that the larger size of the respiratory and Orgeig 2001). The surfactant isolated from N. forsteri con- units themselves may confer stability by substantially reducing tains large amounts of sphingomyelin and PE and only traces the collapse pressures that are present in the much smaller of PG (Orgeig and Daniels 1995). The amount of sphingo- alveoli of mammals (Daniels et al. 1998a). In these lungs or myelin is even greater in the goldfish swim bladder than in the

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Surfactant in Air-Breathing Organs of Fish 745

Australian lungfish (Daniels and Skinner 1994). The most concentrations (10 mg/mL, compared with 1 mg/mL PL) to abundant minor PL in fish surfactant is also sphingomyelin, measure the surface activity of fish surfactant. The CBS is cur- yet the surfactant essentially lacks PG. The significance of the rently regarded as the most sophisticated and accurate device variation in the minor PLs, in particular PG, is unknown. for measuring the biophysical behavior of pulmonary surfactant In terms of surfactant lipid composition, the strongest evi- mixtures (Veldhuizen et al. 1998; Schu¨rch et al. 2001). The dence that the surfactant system had a single evolutionary origin other methodologies are less reliable, particularly for highly comes from the similarity in composition between the Austra- unsaturated, highly fluid surfactant mixtures, as they tend to lian lungfish (N. forsteri), a primitive sarcopterygian closely suffer from leakage of PLs out of the surfactant film (Veldhuizen related to the stem ancestor of the tetrapods, and the primitive et al. 1998; Schu¨rch et al. 2001). Hence, the poor surface activity actinopterygian air-breathing fish (Ginglymodi and Polypteri- previously reported for fish surfactant appears to be a true formes). Hence, the similarity in composition between prim- reflection of its biophysical properties and is not an artifact of itive members of the two branches (Actinopterygii and Sar- low PL concentrations or inferior (leaky) measurement devices copterygii) suggests that the surfactant system evolved once in (Table 3). a common ancestor. Furthermore, the difference in the sur- It is likely that the surface-active properties observed for fish factant between these primitive representatives of the two surfactant directly reflect the differences in lipid and possibly branches on the one hand and the more derived sarcopterygians protein composition that we have described. For example, the and tetrapods on the other hand supports our suggestion of a concentration of DSP in the tarpon swim bladder is only 4% protosurfactant. or less of total PL, which is even lower than that of other air- breathing fishes (9%–14%; Table 2; Smits et al. 1994) and vir- tually nonexistent compared with that of reptiles, birds, and Surface Activity mammals (30%–50%; Veldhuizen et al. 1998; Daniels and Or- The surface activity of tarpon surfactant was marginally better geig 2001). Given that DSP, and in particular DPPC, is the Њ Њ at 22 C than at 37 C, as reflected by a slightly lower STmin and major PL responsible for achieving low surface tensions upon a reduced %SAcomp (Table 3). This presumably reflects an ad- compression (Veldhuizen et al. 1998), it is not surprising that aptation to function at temperatures that are more similar to fish surfactant is unable to generate low surface tensions. A the body temperature experienced by the animal. It has pre- further characteristic of fish surfactant that mitigates against viously been demonstrated that surfactant isolated from cold- low surface tensions is the extremely high concentration of acclimated animals, such as marsupials and bats, which are cholesterol (Chol/PL: 0.2–0.3; Chol/DSP: 1.0–5.6), compared capable of undergoing torpor, functions more effectively at with that in mammals (Chol/PL: ∼0.08; Chol/DSP: ∼0.12; Dan- room temperature (22Њ–24ЊC) than at 37ЊC. Conversely, if the iels et al. 1995a), reptiles (Chol/PL: ∼0.03–0.08; Chol/DSP: surfactant is isolated from the active animals, it is more active ∼0.1–0.3; Daniels et al. 1996; Daniels and Orgeig 2001), and at 37ЊC than at room temperature (Lopatko et al. 1998; Codd amphibians (Chol/PL: ∼0.03–0.1; Chol/DSP: ∼0.2–0.3; Daniels et al. 2002). Whether fish are capable of adjusting their sur- et al. 1994; Daniels and Orgeig 2001). While cholesterol (of the factant composition and function to match changes in body order of 5%–10% of total PL) is essential in surfactant to pro- temperatures is unknown. The activity of tarpon surfactant mote the spreading of the saturated PL film upon inspiration certainly appears sensitive to temperature. (Notter et al. 1980; Fleming and Keough 1988), high concen- The surface activity of tarpon surfactant is extremely poor, trations of cholesterol tend to inhibit surfactant surface activity compared with that of mammalian surfactant measured under (Notter et al. 1980). Yet another characteristic of surfactant the same conditions with the same device. Mammalian sur- from the tarpon and pirarucu swim bladders is the very low ! factant typically experiences an STmin of 1 mN/m, an STeq of concentration of SP-B (2, 17, and 44 pg/mg protein in the three ∼ ! 25 mN/m, and a %SAcomp of 20% (Schu¨rch et al. 1992; Table fish tested), compared with 2 ng/mg protein in mice, determined 3). Moreover, the surface activity was also poor relative to that by the same method (Tokieda et al. 1999). As SP-B is closely measured in lizards on the Wilhelmy balance (Daniels et al. associated with the surfactant lipids and has a crucial role in 1998a). Lizard surfactant demonstrated intermediate surface promoting PL adsorption to the surfactant film (Haagsman activity with an STmin of 13–14 mN/m, an STeq of 25 mN/m, and Diemel 2001), the up to 100-fold lower concentration in and a %SAcomp of 30%. However, the surface activity of tarpon lavage of this essential protein no doubt also contributes to the surfactant was very similar to that previously reported for other poor surface-active properties of fish surfactant. ray-finned fishes (Teleostei and Ginglymodi) and lungfishes We have previously argued that nonmammals, due to their

(Daniels et al. 1998a; Table 3). The STmin values measured with larger respiratory units, less complex lungs, and greater natural alternative instruments, that is, either the Wilhelmy balance or lung distensibility, do not require a surfactant mixture capable the Enho¨rning Bubble Surfactometer, ranged from 17 to 26 of the extremely low surface tensions typical of mammalian mN/m (Daniels et al. 1998a). This study is the ﬁrst to use the surfactants. A more detergent-like surfactant (demonstrating modern, state-of-the-art CBS in combination with high PL higher STmin and less variation) would be perfectly adequate

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 746 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

(Daniels et al. 1998a). Surfactant of nonmammals is not The lipid composition is conserved, and homology of SP-A, primarily responsible for maintaining alveolar stability and in- SP-B, and SP-D demonstrates a single evolutionary origin for creasing lung compliance but rather is required as an anti- the system. adhesive and an antiedema agent (Daniels et al. 1998a). Nev- ertheless, fish surfactant, due to its composition and function, Acknowledgments appears to be in a class of its own—the protosurfactant. The extraordinary differences between fish surfactant and that of This study was supported by an Australian Research Council mammals, reptiles, and amphibians suggest that fish surfactant (ARC) grant to C.B.D., an ARC research fellowship to S.O., a has functions that have not yet been elucidated. These may postgraduate research scholarship to L.C.S., an ARC grant to include functioning as an innate immune system or as an an- M.B.B., the Alberta Heritage Foundation for Medical Research, tioxidant system or even simply the provision of a medium for the Canadian Institute of Health Research, the Silva Casa Foun- dissolving other antioxidant, antiviral, or antimicrobial agents. dation (S.S.), the Natural Sciences and Engineering Research Council (Canada; C.J.B.), and Conselho Nacional de Desen- Conclusion volvimento Cientifico e Tecnolo´gico (Brazil; A.L.V.). We thank Helen Blacker for assistance with lipid analyses, Stanley Cheng It appears that the evolution of the surfactant system predated for the surface activity measurements, and Mark Mano, who the evolution of lungs and swim bladders. However, the func- assisted with the cholesterol analysis. tional evolution of the surfactant system is closely coupled with that of lungs or swim bladders in the fish and lungs in the tetrapods. It also appears that there are two different types of Literature Cited surfactant, one in the actinopterygian fishes that is high in cholesterol and USPs and one in the advanced sarcopterygian Bishop I.R. and G.E.H. Foxon. 1968. The mechanism of fishes and tetrapods that is relatively low in Chol and USPs breathing in the South American lungfish, Lepidosiren par- and high in DSPs. Fish surfactant is likely to be highly spread- adoxa: a radiological study. J Zool (Lond) 154:263–271. able but is not very surface active. The tetrapod surfactant is Bourbon J.R. and B. Chailley-Heu. 2001. Surfactant proteins much more surface active and may, therefore, allow the de- in the digestive tract, mesentery, and other organs: evolu- velopment of more complex lungs with smaller respiratory tionary significance. Comp Biochem Physiol A 129:151–161. units and a greater total respiratory surface area, paving the Brainerd E.L., K.F. Liem, and C.T. Samper. 1989. Air ventilation way for the evolution of high-performance lungs in amniotes. by recoil aspiration in polypterid fishes. Science 246:1593– It is possible that fish surfactant is an archaic or “proto-” sur- 1595. factant that evolved into tetrapod surfactants but is also im- Brooks R.E. 1970. Ultrastructure of the physostomatous swim- portant as a lipid lining for the swim bladders in the modern bladder of rainbow trout (Salmo gairdneri). Z Zellforsch Mik- fish. rosk Anat 106:473–483. Many reptilian, amphibian, and fish lungs are essentially bag- Codd J.R., S. Schu¨rch, C.B. Daniels, and S. Orgeig. 2002. like, with a large central airspace and lacking a bronchial tree, Torpor-associated fluctuations in surfactant activity in and with few exceptions they can collapse completely (Bishop Gould’s wattled bat. Biochim Biophys Acta 1580:57–66. and Foxon 1968; Guimond and Hutchison 1976; Hughes and Copeland D.E. 1969. Fine structural study of gas secretion in Vergara 1978; Martin and Hutchison 1979; Stark-Vancs et al. the physoclistous swimbladder of Fundulus heteroclitous and 1984; Brainerd et al. 1989; Frappell and Daniels 1991a, 1991b). Gadus callarias and in the euphysoclistous swimbladder of A very important function of surfactant in these lung types is Opsanus tau. Z Zellforsch Mikrosk Anat 38:305–311. to prevent the epithelial surfaces from sticking together (Daniels Creasey J.M., R.E. Pattle, and C. Schock. 1974. Ultrastructure et al. 1995a). Nonmammalian surfactant does not significantly of inclusion bodies in type II cells of lung, human and sub- influence inflation compliance. Given the highly derived nature simian. J Physiol 237:35–37. of the mammalian lung, it is likely that regulation of surface Daniels C.B., O.V. Lopatko, and S. Orgeig. 1998a. Evolution of tension during nonatelectatic breathing is a late development surface activity related functions of vertebrate pulmonary in the evolution of surfactant. Based on our studies of reptilian, surfactant. Clin Exp Pharmacol Physiol 25:716–721. amphibian, and piscine surfactants, we conclude that while Daniels C.B. and S. Orgeig. 2001. The comparative biology of acting as an antiadhesive may have been the original function pulmonary surfactant: past, present and future. Comp of surfactant, it led naturally to the alveolar stability and com- Biochem Physiol A 129:9–36. pliance roles that dominate mammalian surfactant function. Daniels C.B., S. Orgeig, and A.W. Smits. 1995a. Invited per- The surfactant system has been highly conserved, morpho- spective: the evolution of the vertebrate pulmonary surfac- logically and biochemically, throughout (and despite) the enor- tant system. Physiol Zool 68:539–566. mous radiation of air-breathing organs among the vertebrates. Daniels C.B., S. Orgeig, A.W. Smits, and J.D. Miller. 1996. The

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Surfactant in Air-Breathing Organs of Fish 747

influence of temperature, phylogeny, and lung structure on giant salamanders of North America. Pp. 313–338 in G.M the lipid composition of reptilian pulmonary surfactant. Exp Hughes, ed. Respiration of Amphibious Vertebrates. 1st ed. Lung Res 22:267–281. Academic Press, London. Daniels C.B., S. Orgeig, J. Wilsen, and T.E. Nicholas. 1994. Haagsman H.P. and R.V. Diemel. 2001. Surfactant-associated Pulmonary-type surfactants in the lungs of terrestrial and proteins: functions and structural variation. Comp Biochem aquatic amphibians. Respir Physiol 95:249–258. Physiol A 129:91–108. Daniels C.B., S. Orgeig, P.G. Wood, L.C. Sullivan, O.V. Lopatko, Hickman C.P.J., L.S. Roberts, and A. Larson. 2001. Fishes. Pp. and A.W. Smits. 1998b. The changing state of surfactant lipids: 507–537 in M.J. Kemp, ed. Integrated Principles of Zoology. new insights from ancient animals. Am Zool 38:305–320. 11th ed. McGraw-Hill, New York. Daniels C.B. and C.H. Skinner. 1994. The composition and Hughes G.M. 1967. Evolution between air and water. Pp. 64– function of surface active lipids in the goldfish swim bladder. 80 in A.S. De Reuck and R. Porter, eds. Development of the Physiol Zool 67:1230–1256. Lung. 1st ed. Churchill, London. Daniels C.B., A.W. Smits, and S. Orgeig. 1995b. Pulmonary ———. 1970. Ultrastructure of the air-breathing organs of surfactant lipids in the faveolar and saccular lung regions of some lower vertebrates. Pp. 599–600 in P. Favard, ed. Pro- snakes. Physiol Zool 68:812–830. ceedings of the Seventh International Congress of Electron Daniels C.B., P.G. Wood, O.V. Lopatko, J.R. Codd, S.D. John- Microscopy. Vol. 3. Socie´te´ Française de Microscopie E´ lec- ston, and S. Orgeig. 1999. Surfactant in the gas mantle of tronique, Paris. the snail Helix aspersa. Physiol Biochem Zool 72:691–698. ———. 1973. Ultrastructure of the lung of Neoceratodus and DeGroodt M., A. Lagasse, and M. Sebruyns. 1960. Elektronmi- Lepidosiren in relation to the lung of other vertebrates. Folia kroskopische Morphologie der Lungenalveolen des Protopterus Morphol (Prague) 21:155–161. und Ambystoma. Pp. 418–421 in W. Bargmann, D. Peters, and Hughes G.M. and J.S.D. Munshi. 1973. Nature of the air- C. Wolpers, eds. Proceedings of the Fourth International Con- breathing organs of the Indian fishes Channa, Amphipnous, ference on Electron Microscopy. Springer, Berlin. Clarias, and Saccobranchus as shown by electron microscopy. Dehadrai P.V. and S.D. Tripathi. 1976. Environment and ecol- J Zool (Lond) 170:245–270. ogy of freshwater air-breathing teleosts. Pp. 39–72 in G.M. Hughes G.M. and G.A. Vergara. 1978. Static pressure-volume Hughes, ed. Respiration of Amphibious Vertebrates. 1st ed. curves for the lung of the frog (Rana pipiens). J Exp Biol 76: Academic Press, London. 149–165. Denison R.H. 1941. The soft anatomy of Bothriolepis. J Pa- Hughes G.M. and E.R. Weibel. 1976. Morphometry of fish leontol 15:553–561. lungs. Pp. 213–232 in G.M. Hughes, ed. Respiration of Am- Dorn E. 1961. U¨ ber den Feinbau der Schwimmblase von An- phibious Vertebrates. 1st ed. Academic Press, London. guilla vulgaris L. Z Zellforsch Mikrosk Anat 55:849–912. ———. 1978. Visualization of layers lining the lung of the Engle M.J. and D.H. Alpers. 2001. Surfactant-like particles me- South American lungfish (Lepidosiren paradoxa) and a com- diate tissue-specific functions in epithelial cells. Comp parison with the frog and rat. Tissue Cell 10:343–353. Biochem Physiol A 129:163–171. Jasinski A. 1973. Air-blood barrier in the respiratory intestine of Fahlen G. 1967. Morphology of gas bladder of Coregonus la- the pond loach Misgurnus fossilis L. Acta Anat 86:376–393. varetus. L. Acta Univ Lund Sect II Med Math Sci Rerum Nat Jasinski A. and W. Kilarski. 1964. The gas gland in the swim- 28:1–37. bladder of the burbot (Lota lota L.) and stone perch (Acerina Fange R. 1983. Gas exchange in fish swim bladder. Rev Physiol cernua L.), its macro- and microscopic structure based on Biochem Pharmacol 97:111–158. observations of electron microcopy. Acta Biolog Cracov Ser Fleming B.D. and K.M.W. Keough. 1988. Surface respreading Zool 7:111–125. after collapse of monolayers containing lipids of pulmonary Johansen K. 1968. Air-breathing fishes. Sci Am 219:102–111. surfactant. Chem Phys Lipids 49:81–86. Klika E. and V. Janout. 1967. The visualization of the lining film Frappell P.B. and C.B. Daniels. 1991a. Temperature effects on of the lung alveolus with the use of Maillet’s modification of ventilation and metabolism in the lizard, Ctenophorus nu- Champy’s method. Folia Morphol (Prague) 15:318–329. chalis. Respir Physiol 86:257–270. Klika E. and A. Lelek. 1967. A contribution to the study of ———. 1991b. Ventilation and oxygen consumption in agamid lungs of the Protopterus annectens and Polypterus senegalensis. lizards. Physiol Zool 64:985–1001. Folia Morphol (Prague) 15:168–175. Graham J.B. 1976. Respiratory adaptations of marine air- Lesur O., R.A. Veldhuizen, J.A. Whitsett, W.M. Hull, F. Poss- breathing fishes. Pp. 165–187 in G.M. Hughes, ed. Respiration mayer, A. Cantin, and R. Begin. 1993. Surfactant-associated of Amphibious Vertebrates. 1st ed. Academic Press, London. proteins (SP-A, SP-B) are increased proportionally to alve- ———. 1997. Air-Breathing Fishes: Evolution, Diversity and olar phospholipids in sheep silicosis. Lung 171:63–74. Adaptation. Academic Press, San Diego, Calif. Lopatko O.V., S. Orgeig, C.B. Daniels, and D. Palmer. 1998. Guimond R.W. and V.H. Hutchison. 1976. Gas exchange of the Alterations in the surface properties of lung surfactant in the

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions 748 C. B. Daniels, S. Orgeig, L. C. Sullivan, N. Ling, M. B. Bennett, S. Schu¨rch, A. L. Val, and C. J. Brauner

torpid marsupial Sminthopsis crassicaudata. J Appl Physiol tein A (SP-A) is expressed by epithelial cells of small and 84:146–156. large intestine. J Biol Chem 270:12162–12169. Lowry O.H., N.J. Rosebrough, A.L. Farr, and R.J. Randall. 1951. Schoel W.M., S. Schu¨rch, and J. Goerke. 1994. The captive Protein measurement with Folin phenol reagent. J Biol Chem bubble method for the evaluation of pulmonary surfactant: 193:265–275. surface tension, area, and volume calculations. Biochim Bio- Marquet E., H.J. Sobel, and R. Schwarz. 1974. Ultracytochem- phys Acta 1200:281–290. istry of the lung of Polypterus ornatipinnis. Cell Tissue Res Schu¨rch S., H. Bachofen, J. Goerke, and F. Green. 1992. Surface 155:437–447. properties of rat pulmonary surfactant studied with the cap- Martin K.M. and V.H. Hutchison. 1979. Ventilatory activity in tive bubble method: adsorption, hysteresis, stability. Biochim Amphiuma tridactylum and Siren lacertina (Amphibia, Cau- Biophys Acta 1103:127–136. data). J Herpetol 13:427–434. Schu¨rch S., H. Bachofen, and F. Possmayer. 2001. Surface ac- Merrick J.R. and G.E. Schmida. 1984. Australian Freshwater tivity in situ, in vivo, and in the captive bubble surfactometer. Fishes: Biology and Management. Griffin, North Ryde. Comp Biochem Physiol A 129:195–207. Miller L.D., S.E. Wert, and J.A. Whitsett. 2001. Surfactant pro- Smits A.W., S. Orgeig, and C.B. Daniels. 1994. Surfactant com- teins and cell markers in the respiratory epithelium of the position and function in lungs of air-breathing fishes. Am J amphibian, Ambystoma mexicanum. Comp Biochem Physiol Physiol 266:R1309–R1313. A 129:141–149. Stark-Vancs V.I., P.B. Bell, and V.H. Hutchison. 1984. Mor- Morris S.M. and J.T. Albright. 1975. The ultrastructure of the phological and pharmacological basis for pulmonary venti- swimbladder of the toadfish Opsanus tau L. Cell Tissue Res lation in Amphiuma tridactylum: an ultrastructural study. 164:85–104. Cell Tissue Res 238:1–12. ———. 1977. Cytochemical study of the lamellar bodies in the Stevens E.D. and G.F. Holeton. 1978. The partitioning of ox- swimbladder of the toadfish Opsanus tau L. Cell Tissue Res ygen uptake from air and from water by the large obligate 185:77–87. air-breathing teleost pirarucu (Arapaima gigas). Can J Zool Munshi J.S.D. 1976. Gross and fine structure of the respiratory 56:974–976. organs of air-breathing fishes. Pp. 73–104 in G.M. Hughes, Sullivan L.C., C.B. Daniels, I.D. Phillips, S. Orgeig, and J.A. ed. Respiration of Amphibious Vertebrates. 1st ed. Academic Whitsett. 1998. Conservation of surfactant protein A: evi- Press, London. dence for a single origin for vertebrate pulmonary surfactant. Notter R.H., S.A. Tabak, and R.D. Mavis. 1980. Surface prop- J Mol Evol 46:131–138. erties of binary mixtures of some pulmonary surfactant com- Sullivan L.C., S. Orgeig, P.G. Wood, and C.B. Daniels. 2001. ponents. J Lipid Res 21:10–22. The ontogeny of pulmonary surfactant secretion in the em- Orgeig S. and C.B. Daniels. 1995. The evolutionary significance bryonic green sea turtle (Chelonia mydas). Physiol Biochem of pulmonary surfactant in lungfish (Dipnoi). Am J Respir Zool 74:493–501. Cell Mol Biol 13:161–166. Szebeni J. and K. Toth. 1986. Lipid peridoxidation in hemo- Pattle R.E. 1976. The lung surfactant in the evolutionary tree. globin-containing liposomes. Effects of membrane phos- Pp. 233–255 in G.M. Hughes, ed. Respiration of Amphibious pholipid composition and cholesterol content. Biochim Bio- Vertebrates. 1st ed. Academic Press, London. phys Acta 857:139–145. Pelster B. 2001. The generation of hyperbaric oxygen tensions Tokieda K., H.S. Iwamoto, C. Bachurski, S.E. Wert, W.M. Hull, in fish. News Physiol Sci 16:287–291. K. Ikeda, and J.A. Whitsett. 1999. Surfactant protein-B- Perry S.F., R.J. Wilson, C. Straus, M.B. Harris, and J.E. Rem- deficient mice are susceptible to hyperoxic lung injury. Am mers. 2001. Which came first, the lung or the breath? Comp J Respir Cell Mol Biol 21:463–472. Biochem Physiol A 129:37–47. Van Rozendaal B.A., L.M. van Golde, and H.P.Haagsman. 2001. Phleger C.F. and B.S. Saunders. 1978. Swim-bladder surfactants Localization and functions of SP-A and SP-D at mucosal of Amazonian air-breathing fishes. Can J Zool 56:946–952. surfaces. Pediatr Pathol Mol Med 20:319–339. Prem C., W. Salvenmoser, J. Wu¨rtz, and B. Pelster. 2000. Swim Veldhuizen R.A.W., K. Inchley, S.A. Hearn, J.F. Lewis, and F. bladder gas gland cells produce surfactant: in vivo and in Possmayer. 1993. Degradation of surfactant associated pro- culture. Am J Physiol 279:R2336–R2343. tein B (SP-B) during in vitro conversion of large to small Rubio S., B. Chailley-Heu, R. Ducroc, and J.R. Bourbon. 1996. surfactant aggregates. J Biochem 295:141–147. Antibody against pulmonary surfactant protein A recognizes Veldhuizen R.A.W., K. Nag, S. Orgeig, and F. Possmayer. 1998. proteins in intestine and swim bladder of the freshwater fish, The role of lipids in pulmonary surfactant. Biochim Biophys carp. Biochem Biophys Res Commun 225:901–906. Acta 1408:90–108. Rubio S., T. Lacaze-Masmonteil, B. Chailley-Heu, A. Kahn, J.R. Wood P.G., O.V. Lopatko, S. Orgeig, J.M. Joss, A.W. Smits, and Bourbon, and R. Ducroc. 1995. Pulmonary surfactant pro- C.B. Daniels. 2000. Control of pulmonary surfactant secre-

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions Surfactant in Air-Breathing Organs of Fish 749

tion: an evolutionary perspective. Am J Physiol 278:R611– scription factor-1, hepatocyte nuclear factor-3b and surfac- R619. tant protein A and B in the developing chick lung. J Anat Zeng X., K.E. Yutzey, and J.A. Whitsett. 1998. Thyroid tran- 193:399–408.

This content downloaded from 23.235.32.0 on Sun, 8 Nov 2015 22:31:46 PM All use subject to JSTOR Terms and Conditions