SEARCH

SEARCH BY CITATION

Keywords:

  • apoptosis;
  • cell proliferation;
  • mitochondria-rich cells;
  • paracellular channel;
  • pillar cell system atrophy

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGEMENTS
  7. LITERATURE CITED

The gill structure of the Amazonian fish Arapaima gigas, an obligatory air breather, was investigated during its transition from water breathing to the obligatory air breathing modes of respiration. The gill structure of A. gigas larvae is similar to that of most teleost fish; however, the morphology of the gills changes as the fish grow. The main morphological changes in the gill structure of a growing fish include the following: (1) intense cell proliferation in the filaments and lamellae, resulting in increasing epithelial thickness and decreasing interlamellar distance; (2) pillar cell system atrophy, which reduces the blood circulation through the lamellae; (3) the generation of long cytoplasmic processes from the epithelial cells into the intercellular space, resulting in continuous and sinuous paracellular channels between the epithelial cells of the filament and lamella that may be involved in gas, ion, and nutrient transport to epithelial cells; and (4) intense mitochondria-rich cell (MRC) proliferation in the lamellar epithelium. All of these morphological changes in the gills contribute to a low increase of the respiratory surface area for gas exchange and an increase in the water–blood diffusion distance increasing their dependence on air-breathing as fish developed. The increased proliferation of MRCs may contribute to increased ion uptake, which favors the regulation of ion content and pH equilibrium. Anat Rec, 296:1664–1675, 2013. © 2013 Wiley Periodicals, Inc.

The gill is the main respiratory organ in most fishes; however, some species obtain O2 directly from atmospheric air by using modified structures in the head or digestive tract (Graham, 1997). The gills of all fishes are similar in their basic structure. Morphometric differences are usually related to the number and size of gill elements (filament and lamella) that affect the efficiency of O2 uptake and are associated with the oxygen needs and/or their habitats (Gray, 1954; Hughes, 1984; Fernandes, 1996; Fernandes et al., 2007; Duncan and Fernandes, 2010). Active and/or hypoxia-tolerant fishes are characterized by a large number of long filaments, a great number of lamellae per millimeter of filament length and, consequently, a large respiratory surface area (Santos et al., 1994; Fernandes, 1996; Mazon et al., 1998; Fernandes et al., 2007; Chapman, 2007). Conversely, the gills of sluggish or benthic fishes have low numbers of lamellae per millimeter of filament length and thus display a small respiratory surface area (Gray, 1954; Hughes, 1984). Recently, the plasticity of the gill dimensions and the reversible remodeling of gill morphology have been studied in fish exposed to pollutants and transferred to clean water (Cerqueira and Fernandes, 2002; Fernandes and Mazon, 2003; Nilsson et al., in press), in response to hypoxia and temperature (Sollid et al., 2003; Sollid and Nilsson, 2006; Perry et al., 2012) and during defense against parasite infections (Nilsson et al., 2012).

In air-breathing fish, the gills exhibit irreversible changes in morphology depending on the relative dependence on water or air for gas exchange (Low et al., 1988; Mattias et al., 1996, 1998; Perna and Fernandes, 1996; Takasusuki et al., 1998; Wilson et al., 1999; Graham et al., 2007). The smallest gill surface area has been documented for the South American lungfish, Lepidosiren paradoxa (Lepidosirenidae), which has irregularly arranged short papillar gill lamellae. These gills are unable to take up the necessary amount of O2 to support oxidative metabolism (Wright, 1974; Moraes et al., 2005) which is provided by the well-developed lungs, evolved during exposure to chronic hypoxia along evolutionary time scale (Johansen and Lenfant, 1967; Bassi et al., 2005; Moraes et al., 2005). Among the Teleostei, the development of a swim bladder modified for O2 uptake from atmospheric air in the pirarucu, Arapaima gigas, allowed changes in the gill morphology as fish developed. Fish is exclusively water-breathing until nine days post-hatch (∼1 g, 1.8 cm length) (Lüling, 1964) and, during development, the fish become dependent on atmospheric air (Sawaya, 1946; Graham, 1997). These fish are then known as obligate air-breathing fish. A. gigas is endemic to the Amazon Basin and is one of largest freshwater fish: it grows up to 4,000–5,000 g (∼80 cm length) in its first year of life (Val and Almeida-Val, 1995) and reaches up to 250,000 g (2–3 m length) over the course of its lifespan (Saint-Paul, 1986; Salvo-Souza and Val, 1990). Without access to air, a 10 g (∼12 cm; 1 month old) fish can survive twice as long as a 1,000 g (∼50 cm; 5–7 months old) fish can (Brauner et al., 2004). In fish lower than 1,000 g the gills are responsible for 23–30% of the whole-body O2 uptake (Gonzales et al 2010) but, in fish between 1,000 and 2,000 g fish, only 20–25% of O2 are took up by the gills (Sawaya, 1946; Stevens and Holeton, 1978; Brauner and Val, 1996). Conversely, CO2 is primarily excreted by the gills (63–78%) and, to a lesser extent, by the swim bladder (15%) and kidney (6%) (Hulbert et al., 1978; Stevens and Holeton, 1978; Brauner and Val, 1996). At low magnification, scanning electron microscopy (SEM) revealed well-developed lamellae in the gill filaments of 10–100 g fish (∼22–25 cm; 2–3 months old) but not in those of 724 g and 1,000 g fish (Brauner et al., 2004; Gonzalez et al., 2010). The lamellar organization of a 100 g A. gigas is similar to that of other teleosts, and its respiratory surface area is similar to those of facultative air-breathing fish of a similar size (Costa et al., 2007); the average respiratory surface area of the swim bladder exceeded that of the gills by a factor of 2.8 indicating the importance of the swim bladder for respiration, even in juvenile fish (Fernandes et al., 2012). Data concerning ion regulation has demonstrated that the rate of diffusive Na+ loss is higher in 724 g fish (∼45 cm; 5 months old) than in 67 g fish (∼22 cm, 2 months old) (Gonzalez et al., 2010) and Brauner et al. (2004) identified high density of mitochondria-rich cells (MRCs) along the outer cell layer of the filament epithelium of 1,000 g fish. However, the changes in the lamellar structure and the morphological processes involved in the gill remodeling of A. gigas during development are not yet known.

Therefore, the main goal of this study was to investigate the main processes involved in the gill remodeling of A. gigas fish during development, including apoptosis and cell proliferation. Furthermore, we investigated the degree of change in the lamellae structure and the distribution of pavement and MRCs during the transition from the water-breathing to air-breathing modes of respiration as the fish grows. We focused on the gill remodeling and its possible implications for respiratory or ion transport functions.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGEMENTS
  7. LITERATURE CITED

Specimens of wild A. gigas (Cuvier, 1817) [N = 8, body mass (MB) ranged from 1.6 to 5,020 g, body length (LB) ranged from 6 to 81 cm)] were collected in the Jacaré Lake system (S03° 31′537″ W60° 40′181″) in the Amazon River near Manacapuru city (100 km from Manaus). After capture, the fish were euthanized by immersion in water containing 0.5% Benzocaine®. The gills were immediately removed and fixed by immersion in 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4, 300 mOsmol) at 4°C.

Morphological and Morphometric Analyses

The gill arches were individually separated, the rakers and bones were removed, and the gill filaments of each gill arch from the left or right side (chosen at random) were placed in sequence with the lateral (opercular) side down. Random samples from all arches, including the anterior and posterior hemibranchs, were obtained by superimposing a square-lattice grid onto the gill filaments. The sample sites were marked by a pin prick. The samples were used for lamellar morphometry (Fig. 1) and for the evaluation of tissue changes during the transition from water-breathing to obligate air-breathing using immunohistochemistry, transmission electron microscopy (TEM), and SEM.

image

Figure 1. Diagram indicating the measurements of the gill variables. A: Total lamellar height; (B) protruded lamellar height (potential functional lamellae); (C) filament epithelium thickness; (D) interlamellar distance; (E) lamellar epithelium thickness; (F) Total lamellar thickness.

Download figure to PowerPoint

Lamellar morphometry

The gill samples were dehydrated in a graded ethanol series, embedded in Historesin (Leica) and systematically sectioned through the filament (3 µm thickness; longitudinal sections). Every 10th section throughout each gill sample from anterior to posterior hemibranchs was collected. The sections were stained with toluidine blue and acid fuchsin. The total and potential respiratory lamellar height (protruded lamellae), lamellar thickness, epithelial filament height and distance between the lamellae were measured as shown in Fig. 1. These measurements were performed in 20 non-contiguous random field samples from 5 filaments that were systematically generated by the CAST System software 2.0 (Olympus, Denmark) in each histological section. Each field sample was composed of ten consecutive lamellae.

Immunohistochemistry analyses

For immunohistochemistry, sections (8 µm in thickness) were processed according to routine protocols for analyses to be performed (cell proliferation, apoptosis, or Na+/K+-ATPase identification) and mounted onto poly-l-lysine-coated slides (Sigma, St Louis, MO) (Dang et al., 2000, Mazon et al., 2004). Cell proliferation was detected using proliferating cell nuclear antigen (PCNA) immunohistochemistry. The sections were cleared in xylene and placed into a solution of 1% H2O2 in methanol to remove endogenous peroxidase activity, re-hydrated and incubated with 1% ZnSO4 at 45°C followed by a wash in distilled water at 4°C. This procedure was repeated twice to break the nuclear membrane. After this step, the sections were washed in 50 mM Tris-buffered saline with 150 mM NaCl and 0.3% Triton X-100, pH 7.4 (TBS-T) and incubated overnight at room temperature in a humidified chamber with a 1:100 dilution of the primary antibody PC10 (Oncogene Research Products, Calbiochem). The next day, the sections were washed in TBS-T and incubated with the secondary antibody GAM (Sigma) diluted 1:150 in TBS. Then, the sections were washed with TBS-T and incubated with anti-peroxidase (PAP; produced in mouse, Sigma) diluted 1:800 in TBS. The proliferating cells were visualized by staining the sections with 3,3′diaminobenzidine (DAB)-Ni (0.5 mg mL−1) and 2.5 mg mL−1 ammonium nickel(II) sulfate hexahydrate in TB that contained 0.0125% H2O2. The negative controls were obtained by omitting either the first or the second antibody and were incubated and stained as described above.

Apoptotic cells were detected using the terminal transferase dUTP nick end labeling (TUNEL) assay (ApopTag Plus Peroxidase in situ, Chemicon). The sections were cleared in xylene, re-hydrated, and treated with proteinase K to remove nucleases and other proteins binding to the DNA fragments. Then, they were washed with a solution of 3% H2O2 in phosphate buffered saline (PBS) to remove endogenous peroxidase activity. Nick end labeling was then carried out according to the manufacturer's instructions. To visualize the nuclei of apoptotic cells, the sections were stained with DAB-Ni (3,3′-diaminobenzidine tetrahydrochloride, Sigma). The negative controls were obtained by omitting the TdT enzyme and applying proteinase K to control the incorporation of nonspecific nucleotides, followed by incubation and staining as described above.

The MRCs were detected by immunohistochemistry staining for Na+/K+-ATPase; the paraffin was removed using xylene, and the sections were rehydrated, washed in 0.01 M TBS-T, and incubated with 20% normal goat serum and 0.1% Triton X-100 to inhibit endogenous peroxidase activity. The primary antibody, avian monoclonal antibody α5 (Developmental Studies Hybridoma Bank, Iowa City, IA), was diluted 1:300 in PBS and incubated overnight on the sections at room temperature in a humidified chamber. Next, the sections were washed in TBS-T and incubated for 1 h with the secondary antibody anti-mouse IgG (GAM, Sigma) diluted 1:150. The sections were then incubated with the antibody peroxide anti-peroxidase (PAP/mouse, Sigma) diluted 1:800 in TB for 1 h. The antibody complex was visualized by staining the sections at room temperature with DAB-Ni in TB to which 0.0125% H2O2 was added immediately prior to use. Negative controls were obtained by omitting either the first or the second antibody and were incubated and stained as described above.

The proliferating, apoptotic, and MRCs were quantified on 20 non-contiguous fields that included filament and lamellar epithelia. The images were obtained by systematic random sampling using an Olympus BX51 light microscope with a video camera and CAST System software 2.0 (Olympus, Denmark).

Scanning and transmission electron microscopy

For SEM, gill arch samples were dehydrated in an ethanol series ending with pure ethanol, washed twice in 1,1,1,3,3,3-hexamethyldisilazane (HMDS, Aldrich) and air-dried at room temperature. The filament pairs were glued onto the specimen stub using silver paint, coated with gold in a vacuum sputter (FCD 004 BAUSER) and examined under a DSM 940 ZEISS Scanning Electron Microscope at 25 kV. The frequency of MRCs in contact with the external medium and the MRC fractional areas (MRCFAs) were determined using random SEM photographs of the epithelial surface as described by Bindon et al. (1994) at a magnification of ×3,000 (five non-contiguous fields). The apical surface area of MRCs was defined by tracing the apical perimeter of all MRCs (whole cell and part of the cell) visible in the picture, that is, contacting the external medium, by calibrated computer screening using a morphometric software program (Sigma Scan Pro 3.0, Jandel Scientific Inc).

For TEM, samples of individual filaments (∼1 mm long) fixed in 2.5% glutaraldehyde buffered to pH 7.3 with 0.1 M phosphate buffer at 4°C were post-fixed in 1% osmium tetroxide buffered to pH 7.3 with the same buffer. Then, the samples were washed with a 0.9% NaCl solution, stained in bloc with uranyl acetate, washed with a 0.9% NaCl solution, dehydrated in a graded acetone series, and embedded in 812 SPURR embedding resin (Electron Microscopy Sciences, UK). Semi-thin (0.5 µm) sections were stained with toluidine blue and examined under an Olympus-Micronal photomicroscope. Ultra-thin (60 nm) sections were stained with uranyl acetate and lead citrate and examined with a JEOL JEM1200-EXII transmission electron microscope at 80 kV at a magnification of ×2,000 to ×15,000 depending on structure size.

Statistical Analyses

The data were expressed as the mean ± S.E.M, and the statistical significance of the differences between the morphological variables of the gills of different fish was determined using an analysis of variance (ANOVA). The Tukey test, with 95% confidence limits, was applied to compare the means whenever there was a significant difference (GraphPad InStat Software, San Diego, CA). Regression lines were calculated to determine the grow rate and the correlation coefficient (r2) was estimated to determine the goodness of fit line.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGEMENTS
  7. LITERATURE CITED

General Filament and Lamellar Morphology

The major morphological changes that occurred in the gills of A. gigas as the fish developed were intense cell proliferation of filament and lamellar epithelia, atrophy of the pillar cell system, enlargement of paracellular channels in the filament, and lamellar epithelia and changes in the distribution pattern of the MRCs.

In the 1–200 g fish (7 days to 3 months old), the internal filament structure contains well-defined primary afferent and efferent arteries, a central venous sinus, small nutritive blood vessels, cartilage, and a small amount of connective tissue surrounding the gill vessels and the cartilage. Fish that were 1,000 g or larger had a large number of small blood vessels in addition to those reported for small fish and a large amount of connective tissue in the core of each filament. The filament epithelium has 5–6 cell layers in the 1–200 g fish and reaches 20 or more cell layers during the transition to air-breathing (in fish larger than 500 g, 4–5 months old).

The well-defined lamellae (protruded lamellae), which are widely spaced from each other, are perpendicular to the horizontal axis of the filament (20–25 lamellae mm−1 on one side of the filament) in fish up to 200 g (Fig. 2A–D). In fish larger than 200 g, the protruded lamellae became shorter as they became encased within the filament epithelium (Fig. 2E–J). At low magnification, the filament surface of larger fish (> 1,000 g) seems smooth but, at high magnification, the filament exhibits parallel or disorganized folds (Fig. 2G,I), which were identified as the lamellar structure in light and transmission electron microscope sections (Figs. 2H,J and 3A–C). The number of lamellar structures decreases significantly in larger fish (from 17 to 14 lamellae mm−1 on one side of the filament in fish from 500 to 5,000 g, respectively).

image

Figure 2. Filament and lamellae of A. gigas. A,B: 46 g fish, (C,D) 105 g fish, (E,F) 575 g fish, (G,H) 1,343 g fish, and (I,J) 4,995 g fish. Scanning electron micrographs (A, C, E, G, and I) showing well-defined protruded lamellae in (A, C, E) but not in (F, H); light micrographs (B, D, F, H, and J) showing lamellar structures on gill filaments. Note the height and thickness of the epithelium of the filament (F) and the lamella (L) and the atrophy of the pillar cell system (*) as the fish grow. MRCs (arrowhead) are distributed in the filaments of fish up to 100 g and in the lamellar epithelium in fish larger than 500 g. F,J: Mucous cells between lamellae (arrow). v, vessels. Scale bar is in µm.

Download figure to PowerPoint

image

Figure 3. Lamellar structure of A. gigas. A: 105 g fish, (B) 575 g fish, and (C) 1,343 g fish. Note the atrophy of the pillar cell system in (C) and the paracellular channels between epithelial cells that exhibit numerous cytoplasmic processes (arrows) in (A, B and C). Cytoplasmic processes (arrows) from pillar cells (PC) projected into reduced blood spaces (*) are shown in (C). BL, basal lamina; MRC, mitochondria-rich cells. Scale bars are in µm.

Download figure to PowerPoint

The internal structure of the lamellae is similar to that of other teleost fish; it is characterized by a pillar cell system (pillar cells and blood spaces), which is outlined by endothelial cells that constitute the lower and upper cells of the proximal and marginal channels, respectively. The pillar cell system is covered by the basal lamina and the lamellar epithelium, which is continuous with the filament epithelium (Figs. 2 and 3). The pillar cell system is well developed in fish up to 200–500 g, but the blood spaces constituted by the flanges of pillar cells gradually become reduced in fish larger than 500–600 g, although they are still present in 5,000 g fish (Figs. 2 and 3). The total height of the lamella increased as the fish grew (b = 0.16) more than the potentially functional lamella did (protruded lamella in contact with environmental water) (b = 0.10) (Fig. 4A) and became almost vestigial in 1,000–5,000 g fish. This change is due to the increasing thickness of the filament and lamellar epithelium, which, in turn, reduces the distance of the interlamellar space (Figs. 2 and 4B,C). The interlamellar space is large in 2 g fish, increases in fish from 2 to 100 g (6 to 25–30 µm, respectively) but decreases significantly in animals ranging from 1,000 to 5,000 g (15–8 µm, respectively). In contrast, the lamellar thickness continuously increases (Figs. 2 and 4C). The filament and epithelial changes during the transition from water-breathing to air-breathing are characterized by intense epithelial cell proliferation until fish reach 500–600 g (330 ± 3 × 102 PCNA-positive cells mm−2), and this is followed by a decrease in larger fish (175 ± 2 × 102 PCNA-positive cells mm−2) (Figs. 5A,C,E,G,I and 6). No significant increase in apoptosis occurs during fish growth (mean apoptotic cells: 16 ± 1 × 102 mm−2) (Figs. 5B,D,F,H,J and 6).

image

Figure 4. Relationship between gill variables and body mass as A. gigas grows. A: Total and potential lamellar height, (B) Filament and lamellar thickness, (C) Lamellar thickness and interlamellar distance.

Download figure to PowerPoint

image

Figure 5. Cell proliferation [PCNA-positive cells in (A, C, E, G, I) are identified by black nuclei, arrows] and apoptotic cells [TUNEL-positive cells in (B, D, F, H, J) are identified by black nuclei, arrows] in the gills of A. gigas. A, B: 34 g fish; C, D: 212 g fish; E, F: 575 g fish; G, H: 1,343 g fish; I, J: 4,995 g fish. The number of PCNA-positive and TUNEL-positive cells was numerous in the filament epithelium. F, filament; L, lamella. Scale bars are in µm.

Download figure to PowerPoint

image

Figure 6. Cell proliferation (PCNA-positive cells) and apoptotic cells (TUNEL-positive cells) in the gills of A. gigas. (*) indicates a significant difference from the next smallest fish size (P < 0.05).

Download figure to PowerPoint

Cell Types in the Filament and Lamellar Epithelium

The main cell types in the filament epithelium of A. gigas are pavement to cuboidal epithelial cells in the outermost cell layer. Mitochondria-rich, mucous, and rodlet cells are distributed among pavement or cuboidal cells and generally have reduced contact with the external environment. The lamellar epithelium (2–3 cell layers) is characterized by cuboidal cells in the outermost epithelial cell layer in fish up to 200 g (Fig. 3A); in fish larger than 500 g, the epithelium consists of 4–5 cell layers and has numerous elongated MRCs occupying the lamellar epithelium (Fig. 3B).

The surface architecture of the pavement cells consists of short microridges distributed at random throughout the apical surface in the interbranchial septum and filament. This architecture changes from long microridges defining the cell boundary (Fig. 7A) to the absence of these microridges in areas close to the onset of lamellae (Fig. 7B) in fish up to 200 g. A high frequency of short microridges is found on the surface of epithelial cells close to the lamellae and in the interlamellar region of the filaments in fish larger than 500 g (Fig. 7C,D). The junctions between PVCs or between PVCs and MRCs are characterized by tight junctions at the most apical part of the lateral surface of the cells, short zonula adherens and, generally, one macula adherens (Fig. 8).

image

Figure 7. SEM micrographs of gill epithelium of A. gigas. A, B: Pavement cells with short microridges (arrowhead) throughout the cell surface and long microridges (arrow) defining the cellular limits in the interbranchial septum and filament (A) and short microridges throughout apical surface of PVCs, close to the onset of lamella (B) in a 2 g fish. C, D: Pavement cells in larger fish (C: 575 g fish; D: 5,000 g fish). Note the MRC (arrow) with short microvilli on the cell surface between the pavement cells (PVCs). Scale bars are in µm.

Download figure to PowerPoint

image

Figure 8. TEM micrographs of the gill epithelium of A. gigas. A, B: MRCs. Note the shape of the MRCs in the filament and the microvilli on the apical surface of the MRCs (arrow) in (A). B: Depicts the junctional complex (arrowheads) having one macula adherens (arrows) at the apical lateral surface of cells sealing the paracellular channels (*). PVC, pavement cell. Scale bars are in µm.

Download figure to PowerPoint

Long cytoplasmic processes between adjacent cells result in continuous and sinuous paracellular channels among the cells of the filament and lamellar epithelium (Figs. 3 and 8). The paracellular channels are narrow between the cells of the outermost epithelial cell layer and are sealed by shallow junctional complexes at the apical lateral surface of cells. In contrast, they are wide near the basement membrane. Numerous long microvillus-like cytoplasmic processes project from the epithelial cells into the paracellular channel and almost occlude the channel lumen in large fish (up to 1,000 g). Inside the channel, flocculent precipitates and cell debris may be found. Similar cytoplasmic processes project from the pillar cells into the blood spaces in larger fish.

The apical surface of the MRCs contains numerous microvilli, some of which are slightly depressed relative to the surface of the pavement cells (Fig. 7C,D). They are elongated in shape and have a basal nucleus, numerous mitochondria, and a well-developed tubular system. The cytoplasm of MRCs contains areas of different electron densities (Fig. 8A,B). The MRCs are distributed mainly in the interlamellar epithelium close to base of the lamella in fish ranging in size from 200 to 600 g (Figs. 2B,D,E and 7). In fish weighing ∼1,000 g or more, the MRCs are distributed throughout the filament and the lamellar epithelium (Figs. 2H,J and 7). MRCs show different degrees of immunoreactivity against Na+/K+-ATPase (Fig. 9). Light MRCs (MRCL) and dark MRCs (MRCD) are present at a low frequency in fish under 600 g (5 × 102 MRC mm−2) and a high frequency in fish larger than 1,000 g (100 × 102 MRC mm−2) in the lamellae (P < 0.001), while in the filament epithelium, the frequency of light and dark MRCs continuously increase as the fish grow (Figs. 9 and 10A,B). The MRC frequency at the epithelial surface increases as the fish grow (Fig. 10C), and the fractional surface area of the MRC ranges from 3 to 8–10% of the total area of the filament for small (1.7 g) to large (5,000 g) fish, respectively.

image

Figure 9. Na+/K+-ATPase immunostaining (dark gray or black cytoplasm) in MRCs in the gill epithelia of A. gigas. A: 34 g fish; (B) 105 g fish; (C) 575 g fish; (D) 1,343 g fish; and (E) 4,995 g fish. The frequency of MRCs was lower in fish up to 600 g than in fish of 1,000 g or more. A–E: Show MRCs in the lamellae and interlamellar region of the filaments, and (F) shows MRCs in the filament epithelium that lacks a lamellar structure from a 1,343 g fish. Scale bars are in µm.

Download figure to PowerPoint

image

Figure 10. A, B: Relationship between dark and light MRC frequency and body mass in the lamellar (A) and filament (B) epithelia as A. gigas grows. C: Relationship between chloride cells at the filament epithelial surface (apical MRC) and body mass as A. gigas grows.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGEMENTS
  7. LITERATURE CITED

The gill changes during the transition from water-breathing to air-breathing modes of respiration are unique in A. gigas. The irreversible gill remodeling that occurs as A. gigas grows results in a very low increase of protruded lamella, reduction of the interlamellar space, increase in the water–blood distance, atrophy of the pillar cell system restricting the blood flow in the respiratory lamellar units, enlargement of the paracellular channel between epithelial cells, and an increase in the MRC frequency. All of these morphological changes suggest that the transfer of respiratory gases (O2 and CO2) through the gills of fish larger than ∼500–1000 g is more difficult than in smaller ones due to the increase of physical barrier for gases diffusion but, the increase of MRC frequency favor ion uptake from the water, as pointed out by Brauner et al. (2004).

The morphological remodeling in the gills of A. gigas is characterized by intense interlamellar cell proliferation and low apoptotic cells in the filaments and lamellae. Cell proliferation and apoptosis are the main processes involved in the reversible lamellar insertion in the filament of the water-breathers Carassius carassius and C. auratus during exposure to normoxia and hypoxia (Sollid et al., 2003) and acclimation in low and high temperatures (Sollid et al., 2005; Sollid and Nilsson, 2006, Smith et al., 2012). In the case of A. gigas, the irreversible burying of the lamellae into the filament epithelium in fish larger than 500–600 g is due to cell proliferation in the filament epithelium, which has been termed the interlamellar cell mass (ILCM) in Carassius by Sollid et al. (2003). These changes are likely related to the development of a swim bladder for air breathing. The gills of A. gigas from 1,000 to 5,000 g exhibit some similarities to the gill lamella of the air-breathing mudskipper Periophthalmodon schlosseri (Low et al., 1988; Wilson et al. 1999, Graham et al., 2007) in terms of the lamellar epithelium thickness, the presence of MRCs in the lamella, and a very narrow interlamellar space, including partial or complete interlamellar fusions.

The gill changes resulted from cell proliferation in the filament and lamellar epithelium is expected to affect, at least to some degree, the gill function. In general, short protruded lamella and the increasing water–blood distance decrease the morphological diffusion capacity for respiratory gas exchanges (Bindon et al., 1994; Sollid et al., 2005; Tzaneva et al., 2011). The reduction of interlamellar distance for water-ventilated lamellae increase the gill resistance, which favors the bypass of water flow around the lamellae between the adjacent filaments or through physiological dead space during the aquatic respiratory cycle reducing the O2 transference from water to the blood (Piiper, 1998). Thus, the morphological changes in the gills of A. gigas may reduce the capacity of O2 to diffuse through the lamella. Previous data showed that the percentage of O2 uptake by the gills of fish weighing between 2,000 and 3,000 was only 23% (Stevens and Holeton, 1978); however, some physiological adjustments for compensating the gill morphological restrictions may occurs in fish smaller than 1,000 g, as Gonzalez et al. (2010) reported higher O2 uptake percentage in ∼724.2 g fish (30%) than in ∼67 g fish (24%).

CO2 excretion may also be constrained, although it is more soluble in water. Increased thickness in the lamellar epithelium in Oncorhynchus mykiss treated with cortisol or growth hormone (Bindon et al., 1994) and in C. auratus in which ILCM was induced by acclimation at low temperature (Tzaneva et al., 2011) increases the arterial blood partial pressure of CO2 (PCO2). Partitioning respiratory physiological data from 1,000 to 2,000 g A. gigas (Randall et al., 1978; Stevens and Holeton, 1978; Brauner and Val, 1996) showed that ∼70% of CO2 was excreted by the gills and, more recently, Gonzalez et al. (2010) reported ∼85% of CO2 excretion across the gills in different size of fish; however, the higher blood PCO2 and HCO3 concentrations as well as the lower blood pH in larger fish provide evidence that there are some constraints for CO2 excretion in large (∼724 g) fish.

The intense cell proliferation in the filament and lamellar epithelia as A. gigas developed might also affect the ionic regulation that acts as a barrier preventing ion loss. In C. auratus, exhibiting ILCM, branchial Na+ and Cl efflux and ammonia excretion decreases compared to those lacking an ILCM (Mitrovic and Perry, 2009; Bradshaw et al., 2012; Smith et al., 2012). However, in A. gigas, despite a thicker gill epithelium, Gonzales et al. (2010) found significantly higher influx and efflux rates of Na+ in the gills of ∼724 g fish than in smaller individuals. The shallow tight junctions between pavement and MRCs may contribute to higher ionic loss in larger fish. Although, the physiological significance of the paracellular channels is still unclear, in A. gigas, they may facilitate ion and/or gas diffusion across the epithelium during fish lateral migrations between water with different Na+ concentrations (black and white water) (Castello, 2008). Similar paracellular channels have been described in the gills of the lungfish, L. paradoxa (Wright, 1974) and in the Amazonian freshwater stingray, the cururu ray, Potamotrygon sp. (Duncan et al., 2010). These channels form a pathway through which movement of water and ions may occur. The similarity of the paracellular channels in the gills of A. gigas to those found in the toad bladder (Wade and Discala, 1971) and in frog skin epithelium (Martinez-Palomo et al., 1971) suggests a possible role for these channels in osmotic regulation and/or gas and nutrient transport to epithelial cells. Wright (1974) suggested the development of an osmotic gradient in the paracellular channels which favor ionic regulation in L. paradoxa. It should be considered for A. gigas when in black and white water from Amazon basin.

The increased MRC frequency and its fractional surface area in growing A. gigas may help the fish overcome the higher salt leak through the epithelium depending on Na+ concentration and the pH of the water. The high MRC frequency and large apical surface area are directly related to ion uptake (Perry et al., 1992) and the proportion of juvenile and mature MRCs (light and dark immunoreactivity against Na+/K+-ATPase, respectively) (Dang et al., 2000).

The water of the rivers of the Amazon basin are soft and contain low concentration of osmolites, and some of them have low pH such as the Rio Negro (Val and Almeida-Val, 1995; Duncan et al., 2011). It is therefore expected that fish living in soft and ion-poor environments would have MRCs distributed in the filament and lamella (Duncan et al., 2010; Duncan et al., 2011). The MRC measurements in the present study showed that small fish (up to 500 g) have fewer MRCs and lower MRCFAs in the epithelial surface than the larger fish do. Almost half of the MRCs are immature cells with low immunoreactivity against Na+/K+-ATPase (light MRCs) and are not in contact with the environment. This result was clearly evidenced by the lower MRC frequency at the epithelial surface compared to the total MRC frequency in the gill epithelium and agrees with the differences in the gill Na+/K+-ATPase activity in small and large fish reported by Gonzalez et al. (2010). In A. gigas, the kidney most likely plays an important role in ion and pH regulation (Hochachka et al.,1978; Gonzalez et al., 2010).

In conclusion, the morphological changes in the gills of A. gigas during development impose more restriction for gill gas exchange, which may, at least in part, be compensate by physiological adjustments. The increase in MRC frequency and MRCFAs in growing fish favors ion uptake, as demonstrated by physiological data (Gonzalez et al., 2010), although the enlargement of paracellular channels may favor ion loss when fish are in extremely ion-poor water.

ACKNOWLEDGEMENTS

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGEMENTS
  7. LITERATURE CITED

A. gigas specimens were collected with a wild animal license issued by the Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis (IBAMA/DIREC, Lic. n.15/2004, Proc. n. 02005.000449/04-14). The authors acknowledge Dr. Izeni P. Farias for help with catching and maintaining the fish. CA Ramos thanks CNPq for scholarships.

LITERATURE CITED

  1. Top of page
  2. ABSTRACT
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. ACKNOWLEDGEMENTS
  7. LITERATURE CITED
  • Bassi M, Klein W, Fernandes MN, Perry SF, Glass ML. 2005. Pulmonary oxygen diffusing capacity of the South American lungfish Lepidosiren paradoxa: physiological values by the Bohr method. Physiol Biochem Zool 78:560569.
  • Bindon SD, Gilmour KM, Fenwick JC, Perry SF. 1994. The effects of branchial chloride cell proliferation on respiratory function in the rainbow trout Oncorhynchus mykiss. J Exp Biol 197:4763.
  • Bradshaw JC, Kumai Y, Perry SF. 2012. The effects of gill remodeling on transepithelial sodium fluxes and the distribution of presumptive sodium-transporting ionocytes in goldfish (Carassius auratus). J Comp Physiol B 182:351366.
  • Brauner CJ, Matey V, Wilson JM, Bernier NJ, Val AL. 2004. Transition in organ function during the evolution of air-breathing; insights from Arapaima gigas, an obligate air-breathing teleost from the Amazon. J Exp Biol 207:14331438.
  • Brauner CJ, Val AL. 1996. The interaction between O2 and CO2 exchange in the obligate air breather, Arapaima gigas, and the facultative air breather, Lipossarcus pardalis. In: Val AL, Almeida-Val VMF, Randall DJ, editors. Physiology and biochemistry of the fishes of the Amazon. Manaus: National Institute of Amazonian Research. p 101109.
  • Castello L. 2008. Lateral migration of Arapaima gigas in floodplains of the Amazon. Ecol Freshw Fish 17:3846.
  • Cerqueira CCC, Fernandes MN. 2002. Gill tissue recovery after copper exposure and blood parameter responses in the tropical fish Prochilodus scrofa. Ecotoxicol Environ Saf 52:8391.
  • Chapman LJ. 2007. Mopho-physiological divergence across aquatic oxygen gradients in fishes. In: Fernandes MN, Glass ML, Kapoor BG, editors. Fish respiration and environment. Enfield: Science Publishers. p 255288.
  • Costa OTF, Pedreti AC, Schmitz A, Perry SF, Fernandes MN. 2007. Stereological estimation of surface area and barrier thickness of fish gills in vertical sections. J Microscopy 225:19.
  • Dang Z, Lock RA, Flik G, Wendelaar Bonga SE. 2000. Na+/K+-ATPase immunoreactivity in branchial chloride cells of Oreochromis mossambicus exposed to copper. J Exp Biol 203:379387.
  • Duncan WP, da Costa OTF, Sakuragui MM, Fernandes MN. 2010. Functional morphology of the gill in amazonian freshwater stingrays (Chondrichthyes: Potamotrygonidae): implications for adaptation to freshwater. Physiol Biochem Zool 83:1932.
  • Duncan WP, Fernandes MN. 2010. Physicochemical characterization of the white, black, and clearwater rivers of the Amazon Basin and its implications on the distribution of freshwater stingrays (Chondrichthyes, Potamotrygonidae). Pan Am J Aquat Sci 5:454464.
  • Duncan WP, Silva NF, Fernandes MN. 2011. Mitochondrion-rich cells distribution, Na+/K+-ATPase activity and gill morphometry of the amazonian freshwater stingrays (Chondrichthyes: Potamotrygonidae). Fish Physiol Biochem 37:523531.
  • Fernandes MN. 1996. Morpho-functional adaptations of gills in tropical fish. In: Val AL, Almeida-Val VMF, Randall DJ, editors. Physiology and biochemistry of the fishes of the Amazon. Manaus: National Institute of Amazonian Research. p 181190.
  • Fernandes MN, Cruz L, Costa OTF, Perry SF. 2012. Morphometric partitioning of the respiratory surface area and diffusion capacity of the gills and swim bladder in juvenile Amazonian air-breathing fish, Arapaima gigas. Micron 43:961970.
  • Fernandes MN, Mazon AF. 2003. Environmental pollution and fish gill morphology. In: Val AL, Kapoor BG, editors. Fish adaptations. Enfield: Science Publishers. p 203231.
  • Fernandes MN, Moron SE, Sakuragui MM. 2007. Gill morphological adjustments to environment and the gas exchange function. In: Fernandes MN, Gloss ML, Rantin FT, Kapoor BG, editors. Fish respiration and environment. Enfield: Science Publisher. p 93120.
  • Gonzalez RJ, Brauner CJ, Wang YX, Richards JG, Patrick ML, Xi W, Matey V, Val AL. 2010. Impact of ontogenetic changes in branchial morphology on gill function in Arapaima gigas. Physiol Biochem Zool 83:322332.
  • Graham JB. 1997. Air-breathing fishes: evolution, diversity, and adaptation. 1st ed. San Diego: Academic Press.
  • Graham JB, Lee HJ, Wegner NC. 2007. Transition from water to land in an extant group of fishes: air breathing and the acquisition sequence of adaptations for amphibious life in oxudercine gobies. In: Fernandes MN, Glass ML, Kapoor BG, editors. Fish respiration and environment. Enfield: Science Publishers. p 255288.
  • Gray IE. 1954. Comparative study of the gill area of marine fishes. Biol Bull 107:219225.
  • Hochachka PW, Moon TW, Bailey J, Hulbert WC. 1978. The osteoglossid kidney: correlations of structure, function, and metabolism with transition to air breathing. Can J Zool 56:820832.
  • Hughes GM. 1984. General anatomy of the gills. In: Hoar WS, Randal DJ, editors. Fish physiology. 1st ed. San Diego: Academic Press. p 172.
  • Hulbert WC, Moon TW, Hochachka PW. 1978. The osteoglossid gill: correlations of structure, function, and metabolism with transition to air breathing. Can J Zool 56:801808.
  • Johansen K, Lenfant C. 1967. Respiratory function in the South American lungfish, Lepidosiren paradoxa (Fitz). J Exp Biol 46:205218.
  • Low W, Lane D, Ip Y. 1988. A comparative study of terrestrial adaptations of the gills in three mudskippers: Periophthalmus chrysospilos, Boleophthalmus boddaerti, and Periophthalmodon schlosseri. Biol Bull 175:434438.
  • Lüling K. 1964. Zur biologte und ökologte von Arapaima gigas (Pisces, Osteoglossidae). Zoomorphol 54:436530.
  • Martinez-Palomo A, Erlij D, Bracho H. 1971. Localization of permeability barriers in the frog skin epithelium. J Cell Biol 50:277287.
  • Mattias AT, Perna SA, Moron SE, Rodrigues JAO, Fernandes MN. 1996. Comparação entre a estrutura morfológica e a morfometria das brânquias de cascudo, Hypostomus regani e Hypostomus plecostomus (Loricariidae). In: UFSCar PPGERN Anais do Seminário Regional de Ecologia. São Carlos, SP: Universidade Federal de São Carlos. p 223236.
  • Mattias AT, Rantin FT, Fernandes MN. 1998. Gill respiratory parameters during progressive hypoxia in the facultative air-breathing fish, Hypostomus regani (Loricariidae). Comp Biochem Physiol A 120:311315.
  • Mazon AF, Fernandes MN, Nolasco MA, Severi W. 1998. Functional morphology of gills and respiratory area of two active rheophilic fish species, Plagioscion squamosissimus and Prochilodus scrofa. J Fish Biol 52:5061.
  • Mazon AF, Nolan DT, Lock RAC, Fernandes MN, Wendelaar Bonga SE. 2004. A short-term in vitro gill culture system to study the effects of toxic (copper) and non-toxic (cortisol) stressors on the rainbow trout, Oncorhynchus mykiss (Walbaum). Toxicol Vitro 1:691701.
  • Mitrovic D, Perry SF. 2009. The effects of thermally induced gill remodeling on ionocyte distribution and branchial chloride fluxes in goldfish (Carassius auratus). J Exp Biol 212:843852.
  • Moraes MFPG, Höller S, Costa OTF, Glass ML, Fernandes MN, Perry SF. 2005. Morphometric comparison of the respiratory organs in the South American lungfish Lepidosiren paradoxa (Dipnoi). Physiol Biochem Zool 78:546559.
  • Nilsson GE, Dymowska A, Stecyk JAW. 2012. New insights into the plasticity of gill structure. Respir Physiol Neurobiol 184:214222.
  • Perna SA, Fernandes MN. 1996. Gill morphometry of the facultative air-breathing loricariid fish, Hypostomus plecostomus (Walbaum) with, special emphasis on aquatic respiration. Fish Physiol Biochem 15:213220.
  • Perry SF, Fletcher C, Bailey S, Ting J, Bradshaw J, Tzaneva V, Gilmour KM. 2012. The interactive effects of exercise and gill remodeling in goldfish (Carassius auratus). J Comp Physiol B 182:935945.
  • Perry SF, Goss GG, Laurent P. 1992. The interrelationships between gill chloride cell morphology and ionic uptake in four freshwater teleosts. Can J Zool 70:17751786.
  • Piiper J. 1998. Branchial gas transfer models. Comp Biochem Physiol A 119:125130.
  • Randall DJ, Farrell AP, Haswell MS. 1978. Carbon dioxide excretion in the pirarucu (Arapaima gigas), an obligate air-breathing fish. Can J Zool 56:977982.
  • Saint-Paul U. 1986. Potential for aquaculture of South American freshwater fishes: a review. Aquaculture 54:205240.
  • Salvo-Souza R, Val AL. 1990. O gigante das aguas amazônicas. Ciência Hoje. Rio de Janeiro: SBPC. p 912.
  • Santos CTC, Fernandes M, Severi W. 1994. Respiratory gill surface area of a facultative air-breathing loricariid fish, Rhinelepis strigosa. Can J Zool 72:20092015.
  • Sawaya P. 1946. Sobre a biologia de alguns peixes de respiração aérea (Lepidosiren paradoxa Fitzinger e Arapaima gigas Cuvier). Bol Fac Fil Ciênc Let USP 11:255286.
  • Smith AA, Zimmer AM, Wood CM. 2012. Branchial and extra-branchial ammonia excretion in goldfish (Carassius auratus) following thermally induced gill remodeling. Comp Biochem Physiol A 162:185192.
  • Sollid J, De Angelis P, Gundersen K, Nilsson GE. 2003. Hypoxia induces adaptive and reversible gross morphological changes in crucian carp gills. J Exp Biol 206:36673673.
  • Sollid J, Nilsson GE. 2006. Plasticity of respiratory structures-Adaptive remodeling of fish gills induced by ambient oxygen and temperature. Respir Physiol Neurobiol 154:241251.
  • Sollid J, Weber RE, Nilsson GE. 2005. Temperature alters the respiratory surface area of crucian carp Carassius carassius and goldfish Carassius auratus. J Exp Biol 208:11091116.
  • Stevens ED, Holeton GF. 1978. The partitioning of oxygen uptake from air and from water by the large obligate air-breathing teleost pirarucu (Arapaima gigas). Can J Zool 56:974976.
  • Takasusuki J, Fernandes MN, Severi W. 1998. The occurrence of aerial respiration in Rhinelepis strigosa during progressive hypoxia. J Fish Biol 52:369379.
  • Tzaneva V, Gilmour KM, Perry SF. 2011. Respiratory responses to hypoxia or hypercapnia in goldfish (Carassius auratus) experiencing gill remodelling. Respir Physiol Neurobiol 175:112120.
  • Val AL, Almeida-Val VMF. 1995. Fishes of the Amazon and their environment: physiological and biochemical aspect. 1st ed. New York: Springer-Verlag.
  • Wade JB, Discala VA. 1971. The effect of osmotic flow on the distribution of horseradish peroxidase within the intercellular spaces of toad bladder epithelium. J Cell Biol 51:553.
  • Wilson JM, Kok TW, Randall DJ, Vogl WA, Ip KY. 1999. Fine structure of the gill epithelium of the terrestrial mudskipper, Periophthalmodon schlosseri. Cell Tissue Res 298:345356.
  • Wright DE. 1974. Morphology of the gill epithelium of the lungfish, Lepidosiren paradoxa. Cell Tissue Res 153:365381.