MURDOCH RESEARCH REPOSITORY

This is the author’s final version of the work, as accepted for publication following peer review but without the publisher’s layout or pagination.

The definitive version is available at http://dx.doi.org/10.1093/jmicro/dfp010

Bartels, H., Schmiedl, A., Rosenbruch, J. and Potter, I.C. (2009) Exposure of the gill epithelial cells of larval lampreys to an ion-deficient environment: a stereological study. Journal of Electron

Microscopy, 58 (4). pp. 253-260.

http://researchrepository.murdoch.edu.au/7510/

Copyright: © The Author 2009.

It is posted here for your personal use. No further distribution is permitted.

Exposure of the gill epithelial cells of larval lampreys to an ion-

deficient environment: a stereological study

Helmut Bartels1, Andreas Schmiedl2, Johannes Rosenbruch1 and Ian C. Potter3

1Anatomische Anstalt, Ludwig-Maximilians-Universität München, 80336 München

2Zentrum Anatomie, Medizinische Hochschule Hannover, 30625 Hannover, Germany

3Centre for Fish and Fisheries Research, School of Biological Sciences and

Biotechnology, Murdoch University, Murdoch 6150, Western Australia, Australia

Abstract

Three kinds of epithelial cells comprise the surfaces of the gill filaments and lamellae of

larval lampreys (ammocoetes): ammocoete mitochondria-rich cells (AMRCs), intercalated

mitochondria-rich cells (IMRCs) and pavement cells. Selected characteristics of these cell

types in ammocoetes of Geotria australis held in distilled water and in 10% sea water were

compared using an ultrastructural stereological approach to determine which of those cell

type(s) respond to exposure to an ion-deficient environment in a manner that indicates that

they are involved in ion uptake. Particular focus was placed on the enigmatic AMRC, which

comprises ca 60% of the cells and contains numerous mitochondria. The mean percentage

contributions of both AMRCs and pavement cells to the total number of the three cell types in

the two experimental groups were not significantly different, whereas that of IMRCs was

>7% in distilled water and <1% in 10% sea water (P < 0.001). Furthermore, the mean apical

surface areas of neither AMRCs nor pavement cells differed significantly between the two

experimental groups, whereas that of IMRCs was nearly 3-fold greater in distilled water than

in 10% sea water. The volume densities and size of mitochondria in AMRCs did not differ

between the two exposure regimes. The above comparisons provide no indications that the

uptake of Na+ and Cl− in the gill epithelium of ammocoetes involves either the AMRC or

pavement cell but, when considered in conjunction with data on ion-transporting cells in

other vertebrates, they are consistent with the conclusion that the IMRC plays a crucial role in

this process.

Key words: mitochondria-rich cells; gill epithelium; lampreys; ionic regulation; fresh water

Introduction

Lampreys (Petromyzontiformes), together with the hagfishes (Myxiniformes), are the sole

survivors of the early agnathan (jawless) stage in vertebrate evolution [1]. All species of

lampreys have a protracted freshwater larval phase that culminates in a radical

metamorphosis [2]. Some of these species then migrate to the sea, where they feed

predominantly on teleost fish, before returning to fresh water to breed and thereby complete

their anadromous migration [3].

In contrast to fully metamorphosed individuals of the anadromous species of lampreys, which

can be readily acclimated to full-strength sea water [4–7], the larvae (ammocoetes) of the

lamprey Petromyzon marinus, whose serum osmolarity is ∼225 mosmol l−1, are unable to

osmoregulate in hypertonic environments [7,8]. They thus die when exposed to osmolarities

>350 mosmol l−1, a value that corresponds to approximately one-third of full-strength sea

water [9]. This is consistent with the fact that ammocoetes of P. marinus can survive direct

transfer to a salinity of 10‰ but not 20‰ [7].

As in teleost fishes, the gills of lampreys are the main extrarenal organs of ionic regulation

and osmoregulation [10,11]. The surfaces of the gill filaments and lamellae of ammocoetes

comprise three types of epithelial cells, namely the pavement cells and two types of

mitochondria-rich (MR) cells [12] (see Figs. 1 and 2). One of the MR cell types resembles,

both ultrastructurally and through its possession of carbonic anhydrase and vacuolar (V)-type

H+ATPase, the intercalated MR cell (IMRC) that is found in other ion-transporting epithelia,

such as the toad and turtle urinary bladders, renal collecting duct and amphibian epidermis. In

those ion-transporting epithelia, the IMRC actively secretes protons and plays a role in acid–

base regulation and/or ion uptake [7,11–13]. The other type of MR cell in the ammocoete gill,

the mitochondria of which are characterized by a distinct electron-dense matrix, has no

morphological counterpart in the ion-transporting epithelia of other vertebrates. Since it

disappears during metamorphosis and does not reappear during any subsequent stage in adult

life [12,14], it was consequently termed the ammocoete mitochondria-rich cell (AMRC) [12].

This cell type corresponds to the MR cell of Morris and Pickering [15], the MR platelet cell

of Youson and Freeman [16] and the ion-uptake cell of Mallatt and Ridgway [17]. Because

this cell type is by far the most abundant in the ammocoete gill epithelium and because its

large number of mitochondria could provide the energy for the active transport of ions, the

above-mentioned authors [15–17] concluded that it is responsible for the uptake of Na+ and

Cl− from the hypotonic environment in which larval lampreys live.

Although the post-metamorphic freshwater stages of anadromous lampreys are faced, during

both their downstream and upstream migrations, with the same osmotic problems as their

ammocoetes, they do not likewise possess a cell type with the ultrastructural characteristics of

the AMRC. Such migrating lampreys possess only the IMRCs and pavement cells in their gill

epithelium and yet osmoregulate effectively in fresh water [11,12]. It has thus been argued

that, in upstream-migrating lampreys, the IMRC and/or pavement cell must be responsible for

ion uptake and that, in larval lampreys, the AMRC may not be involved in taking up Na+ and

Cl− from the environment [11,12].

Previous studies have shown that the stimulation of ion-transport processes in various

epithelia is accompanied by quantitative changes in some of their morphological

characteristics, including, for example, the frequency of certain cell types, the density of the

apical surface area and the volume density of mitochondria [18–21]. The present study

explores whether comparable changes occur in the gill epithelium of ammocoetes when ion-

uptake mechanisms are stimulated, placing particular emphasis on the response of the

enigmatic AMRC. For this purpose, an ultrastructural stereological approach was used to

compare the characteristics of the gill epithelial cells of ammocoetes in distilled water with

those of ammocoetes held in 10% sea water. The latter environment has a higher osmolarity

than fresh water but, at ca 100 osmol l−1, is closer to the osmolarity of the amocoete serum. It

was readily tolerated by ammocoetes and the osmotic gradient between it and the blood was

reduced to the point that the uptake of Na+ and Cl− was expected to be limited.

Materials and Methods

Ammocoetes of Geotria australis were caught in southwestern Australian rivers and

transported in river water to the laboratory, where they were kept unfed at least for 1 week

before and during the experiment in aerated aquaria containing a natural substrate into which

they readily burrowed. The aquaria were located in a constant-temperature room maintained

at 18 ± 1°C. One group of four ammocoetes was later transferred to an aquarium containing

distilled water, while another group of four was placed in an aquarium containing 10% sea

water and could thus serve as a ‘control’, as they were exposed to limited ionic or

hypoosmotic stress. Both aquaria possessed a clean sand substrate, with water aerated and

filtered, and were kept at 18°C as stated above. After 14 days, the ammocoetes were

anaesthetized with 0.01% benzocaine and decapitated.

The third and fourth gill pouches of each ammocoete were excised and immediately placed in

a solution of 2.5% glutaraldehyde in 0.1 M Na–cacodylate–HCl buffer, pH 7.4. Small pieces

of gill tissue were removed under a dissection microscope and, after thorough rinsing in the

buffer, postfixed in 2% OsO4 in 0.1 M Na–cacodylate–HCl buffer, dehydrated in ethanol and

embedded in Epon 812. Ultrathin sections, cut almost perpendicular to the long axis of the

gill lamellae, were stained with lead citrate and uranyl acetate and examined under a Philips

CM 10 or Zeiss EM 900 electron microscope at 80 kV.

Stereology

The following parameters were determined in at least five tissue blocks of the gills obtained

from each of the four ammocoetes maintained in distilled water and in 10% sea water: (i) the

frequency of occurrence of the three cell types (AMRC, IMRC and pavement cell) at the

surface of the gill lamellae and filaments, (ii) the density of the apical surface, i.e. surface of

the apical membrane related to a unit volume, and the mean apical surface area per unit cell

volume in the AMRC, IMRC and pavement cell and (iii) the volume density and outer

surface/volume ratio of mitochondria in the AMRC. In addition, the mean volume weighted

volume of mitochondria in AMRC and the volume density (VV) of the AMRC related to a

defined reference space independent from the cell size were determined to exclude cellular

and mitochondrial swelling, respectively [22–24]. Since these two parameters did not differ

between the two experimental groups (P > 0.05), it was concluded that the experimental

design induced neither cellular nor mitochondrial swelling.

The frequency of occurrence was determined by examining 298, 396, 398 and 646 cells

(mean 435 cells) in each of the four ammocoetes held in 10% sea water and 357, 386, 467

and 751 cells (mean 490 cells) in each of the four animals maintained in distilled water. Thus,

1738 and 1961 cells of the above two groups, respectively, were examined. The other

stereological parameters were determined by examining between 17 and 20 AMRCs and

pavement cells and 8 and 10 IMRCs per animal in each of the two groups.

The test fields were selected using the ‘systematic quadrats subsampling’ method, which

combines a random start outside the section (e.g. in the upper-right corner) and a subsequent

systematic, meandering investigation of the section in fixed distances in the x- and y-

directions [25]. Because this enables the test fields to be randomly distributed and all regions

of the section to be examined, all areas of the section have the same opportunity to be

evaluated [26]. Up to 70 test fields in each section were examined.

The electron microscopic image was transferred to a computer screen using a CCD camera. A

rectangular test field frame (23.15 × 15.25 cm), which defined the area to be analysed, was

placed on the screen. The cells at the surface of the gill filaments and lamellae were identified

as either an AMRC, IMRC or pavement cell using cytological criteria [11]. The number of

each cell type within or cut by the upper or left side of the frame was recorded and expressed

as a percentage of the total number of cells counted.

The surface densities and volume densities were determined by point and intersection

counting [26], using a point raster with arched test lines (cycloids) placed on the computer

screen to guarantee intersections with the surface independent of the direction of sectioning.

The two extreme points of each test line served as test points. The number of test lines and

test points were 18 and 36, respectively. The surface densities (SV), the volume densities (VV)

and the volume to surface ratio (VS ratio) were evaluated using the following formulae

[22,26]:

mean volume weighted volume = π/3 × L15 × 1000/3 × magnification ×L03 (μm3)

where IS = intersections with cell or mitochondrial surface, LT = length of the test

line, Pmito = test points falling on mitochondria, Pcell = test points falling on the cell, L15 =

size classification of the rules used and L0 = length of the ruler.

The Mann–Whitney–Wilcoxon rank sum test was used to test for differences (P < 0.05)

between the values for each of the characters recorded for the two experimental groups.

Results and discussion

In ammocoetes held in 10% sea water, the three cell types at the gill surface were distributed

in the same manner as those in ammocoetes maintained in river water [12]. Thus, the AMRCs

and pavement cells were arranged in large groups; the AMRCs covered most of the lamellar

surface (except the tips) and were also located on the filaments in the regions between the

lamellae and at the bases of the filaments, while the pavement cells lied at the tips of the

lamellae. IMRCs were rare and singly intercalated between the AMRCs (Figs. 1 and 2).

Qualitatively, the ultrastructure of the AMRCs and pavement cells of ammocoetes maintained

in distilled water was indistinguishable from that of the same cell types in animals held in

10% sea water (Fig. 2). The IMRC of ammocoetes held in 10% sea water possessed a

relatively smooth surface with few microplicae (Fig. 1a and c). However, when ammocoetes

were kept in distilled water, their IMRCs became enlarged at their apical surfaces through the

production of long, slender microplicae (Fig. 1b and d). Under both conditions, IMRCs

contained membranous tubulovesicles in the apical cytoplasm and between the mitochondria.

These tubulovesicles appeared more abundant in cells of ammocoetes in 10% sea water than

in distilled water (Fig. 1c and d).

Irrespective of whether ammocoetes were held in 10% sea water or distilled water, the

percentage contributions made by the number of the three cell types to the total number of

cells in the upper layer of the gill epithelium followed the same sequence, i.e. the AMRC was

more abundant than the pavement cell, which was more abundant than the IMRC (Fig. 3).

The ranges in the percentage contributions of AMRCs in 10% sea water (56–75%) and

distilled water (52–61%) overlapped and the corresponding means of 65.2 and 58.0%,

respectively, were not significantly different (P> 0.05). The mean percentage contributions of

the pavement cells in 10% sea water (34.0%) and distilled water (34.5%) were virtually

identical (P > 0.05; Fig. 3). In contrast to the situation with the AMRC and pavement cell, the

range in percentage contributions made by the IMRC to the number of cells in the surface

layer of the gill epithelium in the 10% sea water group (0–1%) did not overlap that for the

distilled water group (5–10%) and the respective means of 0.8 and 7.5% were thus

significantly different (P < 0.001; Fig. 3).

The mean apical surface densities of the AMRCs of ammocoetes held in 10% sea water

(0.100 μm2 μm−3) and distilled water (0.108 μm2 μm−3) were not significantly different (P >

0.05; Fig. 4) nor were the corresponding values of 0.099 and 0.110 μm2 μm−3 for the

pavement cells (P > 0.05; Fig. 4). In contrast, the mean apical surface density of 0.312

μm2 μm−3 for the IMRCs in ammocoetes held in distilled water was nearly three times greater

than the 0.119 μm2 μm−3 recorded for animals maintained in 10% sea water (P < 0.001,

Fig. 4). Furthermore, the ranges of the values for the two groups did not overlap. The trends

exhibited by the apical surface densities for the three cell types in the two groups of

ammocoetes were paralleled by those for the apical surface area to cellular volume. Thus, the

values of the apical surface area to cellular volume determined for IMRCs of ammocoetes in

distilled water increased by nearly a factor of 3 compared to those in 10% sea water (0.872

versus 0.301 μm2μm−3), while those for AMRCs and pavement cells did not differ (0.231

versus 0.215 μm2 μm−3 and 0.242 versus 0.267 μm2 μm−3).

Given that stimulating ion transport leads to an increase in the frequency of the ion-

transporting cells and/or their apical surface density (see the ‘Introduction’ section), the lack

of differences between these two parameter in AMRCs (and pavement cells) of ammocoetes

held in distilled water and 10% sea water, as determined in the present study, suggests that

neither the AMRC nor the pavement cell plays a significant role in ion uptake in larval

lampreys. This conclusion is supported by a recent immunohistochemical study on the

ammocoete gill epithelium, which used an antiserum for labelling carbonic anhydrase IIb [7].

That study failed to detect carbonic anhydrase immunoreactivity in AMRCs and thus did not

confirm the results of an earlier histochemical study by Conley and Mallatt [27]. Since

Na+ and Cl− are taken up in exchange for H+ or HCO3−, respectively [28,29], which are

generated by the activity of carbonic anhydrase, the absence of this enzyme in AMRCs [7]

implies that these cells do not contain a basic requirement for the uptake of Na+ and Cl−and

thus cannot be involved in this process.

In contrast to the situation with the AMRCs (and pavement cells), the frequency and the

apical surface density of IMRCs were significantly greater in the distilled water group than in

the 10% sea water group. Our results are consistent with the view that the IMRCs play a

central role in the uptake of ions from the hypotonic riverine environment in which larval

lampreys live [11]. This conclusion is also consistent with the recent demonstration that V-

type H+-ATPase and carbonic anhydrase are present in IMRCs [7]. On the basis of their

ultrastructure and the presence of H+-ATPase and carbonic anhydrase, the lamprey IMRC is a

type of epithelial cell that also occurs, for example, in the amphibian epidermis, the toad and

turtle urinary bladders and renal collecting duct [7,11,13,30,31]. This cell type comprises

three subtypes: one (A) containing the H+-ATPase in its apical membrane and an

HCO3−/Cl− exchanger in its basolateral membrane, another (B) exhibiting the opposite

locations of the H+ pump and the anion exchanger and a third (C) in which the H+ pump and

the anion exchanger are both located in the apical membrane [32–34]. Whereas the subtypes

A and B are involved in acid–base regulation, secreting H+ or HCO3−, respectively, the

subtype C has been held responsible for the uptake of Cl− [32–34]. In the epithelium of the

turtle bladder, the apical surface area of subtype A cells becomes enlarged under acidotic

conditions through exocytosis of cytoplasmic membranous tubules and vesicles containing

the H+-ATPase [19,35]. The almost 3-fold amplification of the apical membrane of the

IMRCs of ammocoetes in an ion-deficient environment thus suggests that this cell type is

stimulated to actively secrete H+ under these conditions. We have previously suggested that

the IMRCs in the ammocoete gills belong to the subtype A or C [11], which cannot be

distinguished in the presence of H-ATPase and rod-shaped particles in their apical membrane

alone. The view that many of the IMRCs in the ammocoete gill epithelium belong to the

subtype C is supported by the observation of Reis-Santos and co-workers [7] that many of

these cells show a strong cytoplasmic immunoreactivity for H+-ATPase, which is typical for

this subtype [13,34]. Thus, an active H+ secretion through the subtype A and C cells would

not only provide the driving force for the uptake of Na+ but also favour the uptake of

Cl− through subtype C cells as the actively secreted H+ can bind to the HCO3− as it leaves the

cell through the apical membrane and thereby establish an HCO3− gradient across this

membrane which would drive the uptake of Cl− [13].

Although the above data do not support the view that the AMRCs are involved in ion uptake

in ammocoetes, these cells must have some important function given that they constitute

∼60% of the cells at the gill surface and that mitochondria make up ∼30% of their cell

volume [36]. The possibility thus remains that the AMRCs may still be involved in the uptake

of Na+ through an epithelial Na+ channel (ENaC) that is energized by the Na+/K+-ATPase

present in their basolateral membrane as discussed by Reis-Santos et al. [7]. Thus, one might

propose that AMRCs are present in such excessive numbers that there is a reserve of these

cells, still taking up sufficient Na+ from an ion-deficient environment to compensate passive

and renal losses that they do not need increase in number. Still, any stimulation of Na+ uptake

mechanisms through the AMRC would be expected to be reflected by an increase in energy

consumption [37]. Therefore, to test whether oxygen consumption is increased in AMRCs,

when ammocoetes are exposed to an ion-deficient environment, we explored whether the

mitochondrial mean volume weighted volume, volume density and/or outer surface/volume

ratio of the AMRC in ammocoetes in distilled water were significantly greater than those in

10% sea water. The appropriateness of such an approach is supported by the finding that

prolonged stimulation of Na+ uptake in the mammalian renal distal tubule results in an

increase in mitochondrial volume density [18,20].

The mean volume density of the mitochondria in the AMRC in the distilled water group, i.e.

31.0%, did not differ significantly from the 36.4% recorded for the 10% sea water group (P >

0.05). As the mean volume weighted volume and the mean outer surface/volume ratios of the

mitochondria in AMRC in ammocoetes in distilled water (0.474 μm3 and 11.3 μm2 μm−3) did

not differ significantly (P > 0.05) from those of the corresponding cells in animals held in

10% sea water (0.519 μm3 and 11.5 μm2 μm−3), mitochondrial size was also not influenced by

experimental procedure. These results also indicate that no mitochondrial swelling was

induced by the experimental procedure [23].

The above data suggest that energy consumption by the AMRC does not increase when

ammocoetes are exposed to an ion-deficient environment. They are thus consistent with the

results involving measurements of the frequency of occurrence and the density of the apical

surface area of AMRC (and pavement cell) in that they also provide no direct evidence that

the AMRC (and pavement cell) is involved in ion uptake by ammocoetes. Yet, the high

prevalence and large mitochondrial component of the AMRC implies that this cell type plays

an important role in some aspect(s) of the larval phase in the lamprey life cycle. It is

conceivable that the recently described high residual activity of an ouabain-resistant ATPase

in ammocoete gill homogenates, which is neither an Na+/K+-ATPase nor an H+-ATPase [7],

is localized in the AMRCs and could thus account for their high energy demand. It might

therefore be relevant that, whereas larval lampreys feed on plant material and possess

AMRCs, upstream-migrating lampreys neither feed nor possess AMRCs, but are able to

osmoregulate very effectively in a hypotonic environment. This led us to suggest that the

AMRC might be involved in the excretion of waste products produced by ammocoetes from

their food [11,12]. Thus, a candidate for the so-far unidentified ATPase could be a multidrug

resistance (MDR) transporter or a transporter related to MDR proteins (MRP), both of which

belong to the family of ABC-transporters and are capable of an ATP-dependant secretion of

organic compounds [38]. Further studies are needed to investigate whether these transporters

are present particularly in the AMRC of the ammocoete gill epithelium, and whether the

morphological and biochemical characteristics of the AMRC are influenced by diet.

Concluding remarks

The data in the present study are consistent with our hypothesis that the AMRC, which is by

far the most abundant cell type in the gill epithelium of ammocoetes, has a function other

than ion uptake [11]. This conclusion contrasts with that proposed by a number of authors for

this mitochondria-rich cell [15–17]. Although our data also provided no support for an

involvement of the pavement cell in ion uptake by larval lampreys, they provide further

indications that the IMRC plays a major role in this process [11].

Funding

Financial support was provided by the Australian Research Grants Committee.

Acknowledgments

The authors express their gratitude to Dr Jon Mallatt for critical reading of the manuscript.

The expert technical assistance of Ursula Fazekas is gratefully acknowledged.

References

1. Hardisty M W. Lampreys and hagfishes: an analysis of cyclostome relationships. In: Hardisty M W, Potter I C, editors. The Biology of Lampreys. 4B.London: Academic Press; 1982. p. 165-259.

2. Hardisty M W, and Potter I C (1971) The general biology of adult lampreys. In: Hardisty M W, and Potter I C (eds), The Biology of Lampreys, vol 1, pp 127–206 (Academic Press, London).

3. Potter I C (1980) Ecology of larval and metamorphosing lampreys. Can. J. Fish. Aquat. Sci. 37: 1641–1657.

4. Potter I C, Huggins R J. Observations on the morphology, behaviour and salinity tolerance of downstream migrating river lamprey (Lampetra fluviatilis). J. Zool. London 1973;169:365-379.

5. Potter I C, Beamish F W H. The freshwater biology of adult anadromous sea lampreys, Petromyzon marinus. J. Zool. London 1977;181:113-130.

6. Potter I C, Hilliard R W, Bird D J. Metamorphosis in the Southern Hemisphere lamprey, Geotria australis. J. Zool. London 1980;190:405-430.

7. Reis-Santos P, McCormick S D, Wilson J M. Ionoregulatory changes during metamorphosis and salinity exposure of juvenile sea lamprey (Petromyzon marinus L.). J. Exp. Biol. 2008;211:978-988.

8. Beamish F W H, Strachan P D, Thomas E. Osmotic and ionic performance of the anadromous sea lamprey, Petromyzon marinus. Comp. Biochem. Physiol. 1978;60A:435-443.

9. Beamish F W H. Osmoregulation in juvenile and adult lampreys. Can. J. Fish. Aquat. Sci. 1980;37:1739-1750.

10. Morris R. Osmoregulation. In: Hardisty M W, Potter I C, editors. The Biology of Lampreys. Vol. 2. London: Academic Press; 1972. p. 192-239.

11. Bartels H, Potter I C. Cellular composition and ultrastructure of the gill epithelium of larval and adult lampreys. Implications for osmoregulation in fresh and seawater. J. Exp. Biol. 2004;207:3447-3462.

12. Bartels H, Potter I C, Pirlich K, Mallatt J. Categorization of the mitochondria-rich cells in the gill epithelium of the freshwater phases in the life cycle of lampreys. Cell Tissue Res. 1998;291:337-349.

13. Brown D, Breton S. Mitochondria-rich, proton-secreting epithelial cells. J. Exp. Biol. 1996;199:2345-2358.

14. Peek W D, Youson J H. Transformation of the interlamellar epithelium of the gills of the anadromous sea lamprey, Petromyzon marinus L., during metamorphosis. Can. J. Zool. 1979;57:1318-1332.

15. Morris R, Pickering A D. Ultrastructure of presumed ion-transporting cells in the gills of ammocoete lampreys, Lampetra fluviatilis (L.) and Lampetra planeri (Bloch). Cell Tissue Res. 1975;163:327-341.

16. Youson J H, Freeman P A. Morphology of the gills of larval and parasitic adult sea lamprey, Petromyzon marinus L. J. Morphol. 1976;149:73-104.

17. Mallatt J, Ridgway R L. Ultrastructure of a complex epithelial system: the pharyngeal lining of the larval lamprey Petromyzon marinus. J. Morphol. 1984;180:271-296.

18. Wade J B, O’Neill R G, Pryor J L, Boulpaep E L. Modulation of cell membrane area in renal collecting tubules by corticosteroid hormones. J. Cell Biol. 1979;81:439-445.

19. Stetson D L, Steinmetz P R. Correlation between apical intramembrane particles and H+ secretion rates during CO2 stimulation in turtle bladder. Pflügers Arch. 1986;407:S80-S84.

20. Kaissling B, Stanton B. Adaptation of distal tubule and collecting duct to increased sodium delivery: I. Ultrastructure. Am. J. Physiol. 1988;255:F1256-F1268.

21. Katz U, Gabbay S. Mitochondria-rich cells and carbonic anhydrase content of toad skin epithelium. Cell Tissue Res. 1986;251:425-431.

22. Braendgaard H, Gundersen H J. The impact of recent stereological advances on quantitative studies of the nervous system. J. Neurosci. Methods 1986;18:39-78.

23. Schmiedl A, Schnabel P A, Mall G, Gebhard M M, Hunnemann D H, Richter J, Bretschneider H J. The surface to volume ratio of mitochondria, a suitable parameter for evaluating mitochondrial swelling. Correlations during the course of myocardial global ischemia. Virchows Arch. A Pathol. Anat. Histopathol 1990;416:305-315.

24. Fehrenbach H, Schmiedl A, Wahlers T, Hirt S W, Brasch F, Riemann D, Richter J. Morphometric characterisation of the fine structure of human type II pneumocytes. Anat. Rec. 1995;243:49-62.

25. Muller A E, Cruz-Orive L M, Gehr P, Weibel E R. Comparison of two subsampling methods for electron microscopic morphometry. J. Microsc. 1981;123:35-49.

26. Weibel E R. Stereological Methods. New York: Academic Press; 1979.

27. Conley D M, Mallatt J. Histochemical localization of Na+-K+-ATPase and carbonic anhydrase activity in gills of 17 fish species. Can. J. Zool. 1988;66:2398-2405.

28. Krogh A. Osmotic Regulation in Aquatic Animals. London: Cambridge University Press; 1939.

29. Kirschner L B. Sodium chloride absorption across the body surface: frog skins and other epithelia. Am. J. Physiol. 1983;244:R429-R443.

30. Brown D, Breton S. H+V-ATPase-dependent luminal acidification in the kidney collecting duct and the epididymis/vas deferens: vesicle recycling and transcytotic pathways. J. Exp. Biol. 2000;203:137-145.

31. Choe K P, O’Brien S, Evans D H, Toop T, Edwards S L. Immunolocalization of Na+/K+-ATPase, carbonic anhydrase II, and vacuolar H+-ATPase in the gills of freshwater adult lampreys, Geotria australis. J. Exp. Zool. A 2004;301:654-665.

32. Brown D, Hirsch S, Gluck S. An H+-ATPase is present in opposite plasma membrane domains in subpopulations of kidney epithelial cells. Nature 1988;331:622-624.

33. Stetson D L, Steinmetz P R. α and β types of carbonic anhydrase-rich cells in turtle bladder. Am. J. Physiol. 1985;249:F553-F565.

34. Larsen E H, Willumsen N J, Christoffersen B C. Role of proton pump of mitochondria-rich cells for active transport of chloride ions in toad skin epithelium. J. Physiol. 1992;450:203-216.

35. Stetson D L, Steinmetz P R. Role of membrane fusion in CO2 stimulation of proton secretion by turtle bladder. Am. J. Physiol. 1983;245:C113-120.

36. Mallatt J, Bailey J F, Lampa S J, Evans M A, Tate W. Quantitative ultrastructure of gill epithelial cells in the larval lamprey Petromyzon marinus. Can. J. Fish Aquat. Sci. 1995;52:1150-1164.

37. Huf E. Versuche über den Zusammenhang zwischen Stoffwechsel, Potentialbildung und Funktion der Froschhaut. Pfluegers Arch. 1935;235:655-673.

38. Sarkadi B, Homolya L, Szakács G, Váradi A. Human multidrug resistance ABCB and ABCG transporters: participation in a chemoimmunity defense system. Physiol. Rev. 2006;86:1179-1236.

Fig. 1. (a and b) Gill epithelium in the interlamellar region of the filament and at the bases of

the lamellae, showing IMRCs (arrows) intercalated between AMRCs in ammocoetes

maintained in (a) 10% sea water and (b) distilled water. (c and d) IMRCs at higher

magnification. The apical surface of the IMRCs in 10% sea water possesses few microplicae

(c), while that of the IMRCs in distilled water is enlarged by numerous microplicae (d). Bars

in a and b = 5 μm; c and d = 1 μm.

Fig. 2. Gill epithelium at the tips of lamellae, showing AMRCs (arrowheads) and pavement

cells (arrows) in ammocoetes maintained in (a) 10% sea water and (d) distilled water.

AMRCs (b and e) and pavement cells (c and f) in 10% sea water (b and c) and distilled water

(e and f) at higher magnification. Bars in a and d = 5 μm; in b, c, e and f = 1 μm.

Fig. 3. Means (±SD) for the percentage contributions made by ammocoete mitochondria-rich

cells (AMRCs), pavement cells (PCs) and intercalated mitochondria-rich cells (IMRCs) in the

gill epithelium of ammocoetes in distilled water and 10% sea water.

Fig. 4. Means (±SD) for the apical surface densities of ammocoete mitochondria-rich cells

(AMRCs), pavement cells (PCs) and intercalated mitochondria-rich cells (IMRCs) in the gill

epithelium of ammocoetes in distilled water and 10% sea water.