Osmoregulation in the Parasitic Protozoan Tritrichomonas foetus

Abstract

Tritrichomonas foetus was shown to undergo a regulatory volume increase (RVI) when it was subjected to hyperosmotic challenge, but there was no regulatory volume decrease after hypoosmotic challenge, as determined by using both light-scattering methods and measurement of intracellular water space to monitor cell volume. An investigation of T. foetus intracellular amino acids revealed a pool size (65 mM) that was similar to that of Trichomonas vaginalis but was considerably smaller than those of Giardia intestinalis and Crithidia luciliae. Changes in amino acid concentrations in response to hyperosmotic challenge were found to account for only 18% of the T. foetus RVI. The T. foetus intracellular sodium and potassium concentrations were determined to be 35 and 119 mM, respectively. The intracellular K⁺ concentration was found to increase considerably during exposure to hyperosmotic stress, and, assuming that there was a monovalent accompanying anion, this increase was estimated to account for 87% of the RVI. By using light scattering it was determined that the T. foetus RVI was enhanced by elevated external K⁺ concentrations and was inhibited when K⁺ and/or Cl⁻ was absent from the medium. The results suggested that the well-documented Na⁺-K⁺-2Cl⁻ cotransport system was responsible for the K⁺ influx activated during the RVI. However, inhibitors of Na⁺-K⁺-2Cl⁻ cotransport in other systems, such as quinine, ouabain, furosemide, and bumetanide, had no effect on the RVI or K⁺ influx in T. foetus.

Tritrichomonas foetus is an anaerobic, flagellate protozoan that parasitizes the urogenital tract of cattle, causing the disease trichomoniasis, which results in fetal loss and transient infertility (23). A member of the trichomonad group, T. foetus is closely related to the human pathogen Trichomonas vaginalis (12). How these parasites maintain cell volume and survive the changes in external fluid associated with their environment and during sexual transmission is an intriguing problem. They lack the contractile vacuole that is present in some other protozoans, such as Crithidia (7), while cysts, such as those formed by Giardia, have not been reported. Efficient osmoregulatory mechanisms therefore appear to be essential for the survival of these parasites.

Despite evidence that protozoan parasites can survive exposure to a much wider range of osmolalities than most vertebrate cells, only a small number of studies have been conducted on osmoregulation in parasites. These investigations, which focused principally on hypoosmotic challenge in diplomonads (2, 20, 21) and trypanosomatids (6, 8, 24), revealed that organic osmolytes, particularly alanine, are centrally involved in regulatory volume processes in these organisms. More recently, a role for inorganic ions in hypoosmotic responses has also been established for Giardia intestinalis (19), Hexamita inflata (2), and Leishmania major (17).

Maintenance of cell volume is a fundamental cellular homeostatic mechanism that must have arisen very early in the evolution of cells. As T. foetus occupies a critical phylogenetic position in the eukaryotic line of descent (12), studies of osmoregulation in this protozoan could provide important information about the development of osmoregulatory mechanisms in organisms lacking a cell wall and the commonality of such mechanisms in protozoans. In this paper we attempt to broaden this important yet neglected area of protozoan cell physiology. We describe the effects of medium anisosmolality on cell volume and intracellular amino acid and ion compositions in T. foetus and report that in contrast to other protozoans studied previously, T. foetus exhibits a regulatory volume increase (RVI) in response to hyperosmotic challenge but surprisingly does not exhibit a regulatory volume decrease (RVD) after hypoosmotic cell swelling.

MATERIALS AND METHODS

Parasite growth and preparation.

T. foetus strain KV1 was cultured axenically in modified TYM medium (9) that did not contain agar, contained 14 mM glucose instead of 14 mM maltose, and was supplemented with 10% heat-inactivated newborn calf serum. Cells were harvested during the mid-log phase of growth by centrifugation at 700 × g for 5 min, washed, resuspended in phosphate-buffered saline (PBS) (1.8 mM KH₂PO₄, 5 mM K₂HPO₄, 150 mM NaCl; pH 7.4) or spent culture medium, and kept at 25°C. Cells were enumerated with a hemocytometer, and protein content was determined by the method of Lowry et al. (18).

Osmolality was changed by dilution of cell suspensions with appropriate combinations of PBS or culture media and distilled water, 2.5 M NaCl, 2.5 M choline chloride, or 2.5 M NaNO₃.

Light scattering and intracellular water space.

Light scattering was utilized to monitor anisosmotic cell volume changes, as described previously for G. intestinalis and H. inflata (2, 21). Optical density at 550 nm was recorded by using a Beckman DU 7500 spectrophotometer with a microcomputer system for T. foetus suspensions in PBS or culture medium. Up to four samples were read simultaneously against a blank every 5 or 20 s for a set period of time. To examine the effects of external ions on T. foetus RVI, the light scattering of cells suspended in a series of buffers with different ion contents was monitored. The following buffers were used: all-K⁺ buffer (1.8 mM KH₂PO₄, 5 mM K₂HPO₄, 150 mM KCl; pH 7.4), K⁺-free buffer (1.8 mM NaH₂PO₄, 5 mM Na₂HPO₄, 150 mM NaCl; pH 7.4), and Cl⁻-free buffer (1.8 mM KH₂PO₄, 5 mM K₂HPO₄, 150 mM NaNO₃; pH 7.4). For experiments that involved studying the effects of inhibitors on T. foetus RVI, cells were preincubated with the inhibitor for 3 min before light scattering was monitored and hyperosmotic challenge was initiated.

The intracellular water space of normal and anisosmotic cells was determined by using an inulin exclusion method adapted for protozoans (15). Aliquots of cells resuspended in PBS were added to equal volumes of inulin [¹⁴C]carboxylic acid (2 μCi ml⁻¹), ³H₂O (10 μCi ml⁻¹), and PBS and incubated at room temperature for 1 min before centrifugation through oil. The supernatant was removed, the pellet was processed, and the radioactivity was counted with a Packard Tri-Carb 1900 liquid scintillation system, as previously described (15). For the time course of intracellular water space during volume recovery, cells were suspended in spent culture medium and d-[1-¹⁴C]mannitol (2 μCi ml⁻¹) was used as the marker for extracellular water space.

Amino acid and intracellular ion analysis.

The amino acids of control and osmotically stressed T. foetus cells were analyzed as described previously (14). Briefly, cells suspended in spent culture medium were centrifuged through oil into sulfosalicylic acid. The supernatant and oil layers were aspirated, which left the cell pellet to be processed and analyzed with a Beckman 6300 amino acid analyzer.

Intracellular ion contents were determined by layering aliquots of normal and hyperosmotic cells in spent culture medium (1 ml) over 300 μl of an oil mixture (4 parts of dibutyl phthalate plus 1 part of isooctyl phthalate; final density, 1.03 g ml⁻¹) and centrifuging the preparations at 10,000 × g for 1 min. Each supernatant was removed, and the oil layer was washed with deionized water before it was aspirated. The pellets were resuspended in 0.1% Triton X-100 (200 μl) and left at −20°C for 1 h. Samples were then heated in a water bath at 90°C for 5 min, vortexed, and centrifuged at 10,000 × g for 5 min. Supernatant samples (150 μl) were taken and recentrifuged before analysis with a Radiometer Copenhagen FLM3 flame photometer.

The extracellular water space of the cell pellets after centrifugation through oil was determined by using d-[1-¹⁴C]mannitol (2 μCi ml⁻¹) as described above and was used to calculate values for the contaminating extracellular amino acids or ions. These values were used as corrections in determining the intracellular amino acid and ion concentrations.

RESULTS

Hypoosmotic cell volume change in T. foetus.

As T. foetus is normally harvested at the medium osmolality, 300 mosmol kg⁻¹, this osmolality was defined as isotonic. Under isotonic conditions, the optical density at 550 nm (OD₅₅₀) of T. foetus suspensions remained constant for up to 15 min before a gradual decrease was detected, which was attributable to cell settling. When the osmolality was decreased by diluting the isotonic PBS with appropriate amounts of distilled water, a rapid decrease in the OD₅₅₀ was observed, suggesting that there was cell swelling. An osmolality below 120 mosmol kg⁻¹ was not used because at lower osmolalities cell lysis was observed by light microscopy. By using a mathematical approach developed for G. intestinalis (21), the OD₅₅₀ data were converted into relative cell volumes. Figure 1 shows the responses of T. foetus to a range of osmolalities from 300 to 120 mosmol kg⁻¹. The cells reached a maximum volume by 5 min and remained swollen, and the swelling was proportional to the decrease in osmolality. No RVD was evident after any osmotic shift. Several alterations were made to the experimental conditions, such as keeping the cells on ice before osmotic shock, addition of 5 mM glucose, suspending the cells in growth medium, and conducting the experiment at 37°C, but no RVD was observed in T. foetus.

FIG. 1.

Time course of relative cell volume changes in response to hypoosmotic challenge. T. foetus suspensions in PBS at 25°C were subjected to a range of hypotonic conditions by dilution with appropriate amounts of distilled H₂O, and the OD₅₅₀ was recorded. The data were converted into relative cell volumes by using a previously described mathematical approach (21). Symbols: ▪, 120 mosmol kg⁻¹; ○, 150 mosmol kg⁻¹; ♦, 180 mosmol kg⁻¹; ▵, 210 mosmol kg⁻¹; •, 240 mosmol kg⁻¹; □, 270 mosmol kg⁻¹; ▴, 300 mosmol kg⁻¹.

The results of light-scattering experiments were confirmed by measurements of intracellular water space. Using inulin [¹⁴C]carboxylic acid and ³H₂O as markers for extracellular and intracellular water space, respectively, we determined that under isotonic conditions 1 mg of total cell protein (1.98 × 10⁷ cells) was equivalent to a volume of 4.04 ± 0.49 μl (n = 6) or 20.4 ± 2.5 μl per 10⁸ cells or 0.20 pl per cell. The intracellular water space measured 5 min after a range of hypoosmotic shifts again showed that there was cell swelling proportional to the osmotic shift, with a 50% decrease in the medium osmolality (to 150 mosmol kg⁻¹) resulting in exactly a twofold increase in the water space. For a cell that behaves as a perfect osmometer it is predicted that V_iO_i = V_fO_f, where V_i and O_i are the initial water space and osmolality and V_f is the new water space after an osmotic shift to a new osmolality (O_f). As shown in Fig. 2 there was a linear relationship between V_i/V_f and O_f, as predicted by this equation. The relative increases in intracellular water space were larger than the relative increases in cell volume (Fig. 1) for the same osmotic shift. This was expected since the water space is considered to be a more accurate measure of the osmotically active component of the cell (which has been reported to be 80% of total cell volume for H. inflata [2]).

FIG. 2.

Relationship between volume change and osmotic shift. The cellular water space of T. foetus (Vf) was measured 5 min after hypoosmotic challenge (•) and 2 min after hyperosmotic challenge (▴) by using inulin [¹⁴C]carboxylic acid and ³H₂O in PBS at 25°C. The ratio of the initial volume (Vi) to the final volume (Vi/Vf) is plotted versus osmolality. The values are means ± standard errors from five separate experiments. Where error bars are not shown, the standard error is smaller than the symbol.

Hyperosmotic cell volume change in T. foetus.

The response of T. foetus to hyperosmotic conditions was also assessed by both light scattering and measurement of intracellular water space. Cell suspensions were subjected to various hyperosmotic conditions by addition of appropriate volumes of 2.5 M NaCl. The upper osmolality limit was 525 mosmol kg⁻¹ because motility, as observed by light microscopy, was lost within 5 min at higher osmolalities. When light scattering was used, the cells were observed to initially undergo volume decreases that were proportional to the increase in the medium osmolality, with the maximal volume changes occurring approximately 2 to 3 min after hypertonic challenge (Fig. 3) In contrast to the results for hypoosmotic challenge, volume recovery was subsequently observed. When the hyperosmotic challenge was carried out in culture medium, the RVI resulted in cells that attained more than 95% of the original volume after 30 min (Fig. 3). However, when similar experiments were performed in PBS buffer, the cells subjected to the smallest osmotic shift (375 mosmol kg⁻¹) required approximately 60 min to recover their volume, and the recovery period was even longer for cells exposed larger osmotic shifts (450 and 525 mosmol kg⁻¹) (data not shown). These results suggest that accumulation of osmolytes that were present in the culture medium but were absent from the PBS buffer was important in the volume recovery process.

FIG. 3.

Time course of relative cell volume changes in response to hyperosmotic challenge. T. foetus suspensions in culture medium at 25°C were subjected to hypertonic conditions by addition of appropriate volumes of 2.5 M NaCl, and the OD₅₅₀ was recorded. The data were converted into relative cell volumes by using a previously described mathematical approach (21). Symbols: ▴, 300 mosmol kg⁻¹; □, 375 mosmol kg⁻¹; ♦, 450 mosmol kg⁻¹; ○, 525 mosmol kg⁻¹.

The time course for cell shrinkage and RVI was confirmed by measuring intracellular water space. Figure 4 shows the intracellular water space values determined at various times during exposure of T. foetus (suspended in culture medium) to 495 mosmol kg⁻¹. The volume after 30 min, 20.05 μl per 10⁸ cells, was not significantly different from the initial value prior to hyperosmotic challenge. Intracellular water space values were also recorded 2 min after different hyperosmotic shifts (Fig. 2), and the results confirmed that the initial cell shrinkage was indeed proportional to the increase in osmolality. A 50% increase in osmolality (to 450 mosmol kg⁻¹) resulted in a 36% initial decrease in the intracellular water space, which agrees well with the predicted decrease, 33%. Overall, the results confirmed the validity of the light-scattering methodology for monitoring hyperosmotic cell shrinkage (and RVI) in T. foetus, as well as hypoosmotic cell swelling in T. foetus and other protozoans.

FIG. 4.

Time course of cellular water space changes in response to hyperosmotic challenge (495 mosmol kg⁻¹), as determined with d-[1-¹⁴C]mannitol and ³H₂O in culture medium at 25°C. The values are means ± standard errors from five separate experiments.

Rapid reversible volume changes.

Having determined that T. foetus could endure a wide range of osmolalities (from 120 to 525 mosmol kg⁻¹), we investigated the ability of cells to respond to multiple osmotic shifts. Cells were initially swollen at 120 mosmol kg⁻¹ for 4 min, and then sufficient 2.5 M NaCl was added to restore the osmolality to 300 mosmol kg⁻¹. As shown in Fig. 5 the cells were able to recover their original volume. Furthermore, subsequent addition of 2.5 M NaCl to obtain an osmolality of 450 mosmol kg⁻¹ resulted in the same cell shrinkage that was exhibited by control cells that had not been subjected to the hypoosmotic shift and restoration of isosmotic conditions. The results are consistent with the hypothesis that no loss of intracellular osmolytes and minimal cell damage arose from the osmotic shifts. In similar experiments, cells were subjected to a hyperosmotic challenge for 2 min and then restored to isotonic conditions prior to any RVI. They returned to the original cell volume, and the swelling induced by a subsequent hypoosmotic shift was identical to that of controls (data not shown).

FIG. 5.

Response to multiple changes in medium osmolality. T. foetus cells suspended in isotonic PBS at 25°C were exposed at time zero to hypoosmotic stress (120 mosmol kg⁻¹) (▪) or were maintained isotonically (300 mosmol kg⁻¹) (▴ and ○). At time A the hypoosmotic cells (▪) were returned to isotonic conditions by appropriate addition of 2.5 M NaCl, while control cell preparations (▴ and ○) were diluted with an equivalent volume of isotonic PBS. At time B, the test cells and one lot of control cells (○) were subjected to hyperosmotic stress (450 mosmol kg⁻¹) by addition of 2.5 M NaCl, while the other control cell preparation (▴) was diluted with an equivalent volume of isotonic PBS.

Intracellular amino acid pools.

The intracellular amino acid pools of T. foetus were monitored under hyperosmotic conditions to determine whether there were any movements that could account for the observed RVI (Table 1) The total concentration of the intracellular amino acid pool in isotonic controls was determined to be 65 mM. The amino acids that dominated the pool (namely, alanine, glutamate, leucine, phenylalanine, and valine) are also the amino acids present at the highest concentrations in the TYM culture medium (14).

TABLE 1.

Intracellular amino acid composition of T. foetus in response to hyperosmotic stress (495 mosmol kg⁻¹)^a

Amino acid	Concn (nmol 108 cells−1) at:
	Zero time	15 min	30 min
Alanine	153	166	190
Arginine	18	28	38
Aspartate	10	26	32
Citrulline	1.4	23	26
Cystine	1.7	13	13
Glutamate	102	189	219
Glutamine	24	25	30
Glycine	66	78	97
Histidine	16	31	34
Isoleucine	114	135	147
Leucine	297	330	328
Lysine	75	196	171
Methionine	39	59	60
Ornithine	2	2.6	2.9
Phenylalanine	108	143	140
Proline	43	55	64
Serine	24	70	88
Threonine	21	60	79
Tryptophan	18	18	18
Tyrosine	30	52	48
Valine	162	185	200
Total	1,324	1,885	2,022

^aThe values are the means of results from two separate experiments that differed by <5%. The total amino acid concentrations at zero time and 15 and 30 min were 65, 115, and 101 mM, respectively.

In response to a hyperosmotic shift to 495 mosmol kg⁻¹, increases in the pools of many amino acids were evident during the 30-min period of the RVI, with most of the changes occurring in the first 15 min. The total concentration of the amino acid pool increased by approximately 50% over the 30 min (from 1,324 to 2,022 nmol per 10⁸ cells). Increases in the concentrations of the large glutamate and lysine pools accounted for about one-third of the total increase, while the largest relative increases were those for serine and threonine (both approximately fourfold). However, the accumulation of amino acids resulted in only a 36 mM increase in the total intracellular concentration of these potential osmolytes at 30 min, at which time the volume recovery was complete. Assuming that 1 mM amino acid was equivalent to 1 mosmol, accumulation of amino acids could account for only 18% of the RVI after this osmotic shift. In similar experiments conducted under hypoosmotic conditions, no changes in the amino acid pool were observed.

Intracellular ion concentrations.

Flame photometry was used to determine the intracellular sodium and potassium concentrations of T. foetus before and after hyperosmotic challenge. Under isosmotic conditions the intracellular ion concentrations were 0.71 ± 0.1 μmol per 10⁸ cells (35 mM) for Na⁺ and 2.43 ± 0.04 μmol per 10⁸ cells (119 mM) for K⁺. Combining these concentrations (and assuming that there were monovalent counterions) gave an intracellular inorganic ion concentration of 308 mM. Combining this value with the total concentration of the intracellular amino acid pool (65 mM) gave a concentration of 373 mM, which was greater than the requirement for 300-mosmol kg⁻¹ intracellular osmolytes and suggested that bound and hence osmotically inactive ions and/or amino acids were present.

In response to hyperosmotic challenge, a dramatic increase in the intracellular K⁺ concentration was apparent (Fig. 6). By using data in Fig. 4 to calculate the K⁺ concentration, a total increase of 85 mM was recorded after 30 min. There was no detectable change in the intracellular Na⁺ concentration over the same period. Assuming that there was a monovalent counterion(s), the increase in the K⁺ concentration could be calculated to contribute 87% of the RVI that T. foetus exhibited when it was exposed to 495 mosmol kg⁻¹.

FIG. 6.

Time course of intracellular potassium changes in response to hyperosmotic challenge (495 mosmol kg⁻¹), as determined by flame photometry at 25°C. The values are means ± standard errors from at least three separate experiments. Where error bars are not shown, the standard error is smaller than the symbol.

Requirement for external potassium for RVI.

In an attempt to characterize the requirement for inorganic ions during RVI, the effects of buffers with various ion compositions were investigated by using light scattering to monitor volume changes. Figure 7 shows that the cells suspended in the all-K⁺ buffer exhibited a more rapid RVI than the cells suspended in PBS, while no RVI was observed for the cells suspended in the K⁺-free buffer. If cells were exposed to a hyperosmotic shift in K⁺-free buffer for 5 min and then diluted with hyperosmotic all-K⁺ buffer, a subsequent rapid RVI was observed (data not shown). In addition, the RVI was inhibited when T. foetus cells were suspended in Cl⁻-free buffer (data not shown). Therefore, it appeared that extracellular potassium and chloride were both required for the cells to undergo an RVI.

FIG. 7.

Effect of external potassium on RVI. T. foetus cells suspended in K⁺-free buffer (▴), PBS (□), and all-K⁺ buffer (•) were subjected to hyperosmotic stress (495 mosmol kg⁻¹) by addition of 2.5 M choline chloride, and the OD₅₅₀ was recorded.

Effects of potential inhibitors on RVI.

A range of potential inhibitors of volume-activated ion channels were assessed to determine their effects on the RVI in T. foetus by using light scattering to monitor volume changes. Of all the inhibitors tested, only N-ethylmaleimide (0.1 mM) resulted in unequivocal inhibition of T. foetus RVI. Inhibition of RVI was also observed when p-hydroxymercuribenzoic acid (pHMB) was present at a concentration of 0.5 mM. However, preincubation of T. foetus with pHMB at the inhibitory concentration resulted in some cell swelling prior to the hyperosmotic shift. The potential inhibitors that had no effect on RVI included amiloride (0.3 mM), bumetanide (0.5 mM), furosemide (0.5 mM), 3-(N-maleimidopropinyl)biocytin (1.0 mM), niflumate (0.5 mM), ouabain (1.0 mM), p-chloromercuribenzene sulfonic acid (1.0 mM), quinine (1.0 mM), and vanadate (0.5 mM).

Anisosmotic effects in culture.

T. foetus cells were grown in culture medium (400 mosmol kg⁻¹) to which additional NaCl was added for 2 days, and the intracellular amino acid pools were analyzed. The trends observed in vitro in response to the hyperosmotic shift were also evident in culture. Substantial increases in the concentrations of most amino acids were observed, and the totals increased almost threefold compared to the pool concentrations for cells grown in the normal isotonic medium (data not shown). Since similar changes were observed both for acute hyperosmotic shifts in vitro and for long-term shifts in culture, it appeared that the increase in the concentration of the amino acid pool in response to hyperosmotic challenge was biologically significant.

The growth rate of T. foetus cells in normal 300-mosmol kg⁻¹ culture medium was compared to the growth rates of cells grown in hypoosmotic (200-mosmol kg⁻¹) and hyperosmotic (400-mosmol kg⁻¹) media. The growth curves showed that the growth rates were similar at 300 and 400 mosmol kg⁻¹, while the growth rate of cells in 200-mosmol kg⁻¹ medium was considerably lower (approximately 2.5-fold lower) (data not shown). These results are consistent with the apparent lack of RVD and the presence of an effective RVI in response to acute osmotic shifts in vitro.

DISCUSSION

This study revealed that like several other protozoans, T. foetus could tolerate a large range of osmotic conditions. By observing their motility over a short time, it was determined that the cells were viable at osmolalities from 120 to 525 mosmol kg⁻¹. This resilient behavior towards anisosmotic challenge was also demonstrated by the cells' ability to grow in hypotonic and hypertonic culture media (200 and 400 mosmol kg⁻¹), although growth was slower at the lower osmolality.

Under hypoosmotic conditions, cells behaved as perfect osmometers, showing no resistance to water influx by swelling in proportion to the decrease in the medium osmolality. Surprisingly, no RVD such as those reported in the protozoan parasites L. major (8), G. intestinalis (21), H. inflata (2), and Crithidia luciliae (6) was observed. In contrast, we observed an RVI in response to a hyperosmotic shift. The cells initially shrank, with minimal volumes observed about 2 min after the hyperosmotic shift, and a subsequent RVI was complete in 30 min for cells suspended in culture medium and in 60 min for cells suspended in PBS. An RVI has not been reported previously for an amitochondriate protozoan. When exposed to an equivalent hyperosmotic shock in culture medium, the trypanosomatid Herpetomonas samuelpessoai has been shown to undergo a more rapid RVI, and the normal cell volume is restored in 15 min; however, this organism was unable to completely restore its volume when it was subjected to hyperosmotic stress in buffer (1).

The previously unreported intracellular water space of T. foetus (20.4 ± 2.5 μl per 10⁸ cells) is considerably smaller than the value reported for T. vaginalis (43 ± 9.4 μl per 10⁸ cells) but is quite similar to the value reported for G. intestinalis (19 ± 3.4 μl per 10⁸ cells) (14). Amino acid analysis revealed that T. foetus had a relatively small intracellular amino acid pool. The total concentration was only 65 mM, compared to 116 mM in G. intestinalis (14), 95 mM in H. inflata (2), and 148 mM in C. luciliae (14); it was, however, similar to the concentration in T. vaginalis, which has been reported to be 57 mM (14). It also appears that T. foetus does not have the high alanine concentrations that dominate the amino acid pools of the organisms mentioned above. When T. foetus was exposed to hypoosmotic conditions, no movement of amino acid pools was detected. These observations set T. foetus apart from parasites such as G. intestinalis (20), H. inflata (2), L. major (8), and C. luciliae (6), which exhibit rapid RVD due to the efflux of large portions of their intracellular amino acid pools, primarily alanine, in response to hypoosmotic stress. In contrast, a substantial movement of amino acids was recorded for T. foetus subjected to hyperosmotic conditions. However, as a consequence of the relatively small pool concentration, the amino acids were found to contribute only 18% of the RVI. A previous study of the free-living protozoan Acanthamoeba castellanii revealed that amino acids contributed 24% of an RVI and that alanine accounted for a major part of the increase in the pool concentration (11). Increases in the alanine concentration in response to hyperosmotic stress have also been observed and studied in L. major (5). While an increase in the alanine content was observed in T. foetus, the change in alanine content was a minor component of the changes in the amino acid pool.

The intracellular concentrations of Na⁺ and K⁺ (35 and 119 mM, respectively) fall between the corresponding values for a typical mammalian cell (5 to 15 and 140 mM). The ratio of K⁺ to Na⁺ (3.4) is similar to the corresponding value for Leishmania mexicana (4.08) (10). In response to hyperosmotic stress the amount of K⁺ was observed to increase steadily, while no change was observed in the Na⁺ levels. The overall increase in the K⁺ concentration (85 mM) contributed 87% of the RVI, suggesting that K⁺ and amino acids together account for the volume recovery process that T. foetus exhibits when it is exposed to 495 mosmol kg⁻¹; the combined contributions total 105%. The results from the H. samuelpessoai study bear the closest resemblance to those which we obtained for T. foetus; there is a dependence on the extracellular K⁺ concentration for RVI, together with an increase in the extracellular K⁺ concentration after hypoosmotic shock, suggesting that there is transport of K⁺ ions across the membrane of this parasite during osmoregulation (1).

At this time it can only be presumed that volume regulation and control of cytoplasmic ionic composition are regulated in protozoans by systems such as those found in many higher eukaryotic organisms. While the dependence of RVI on external K⁺ and Cl⁻ concentrations seems to indicate that the well-documented Na⁺-K⁺-2Cl⁻ cotransporter (reviewed in references 13, 16, and 22) is the means of K⁺ uptake in T. foetus during hyperosmotic stress, the lack of any effect of quinine, furosemide, bumetanide, and ouabain does not. The lack of an effect of amiloride also seems to rule out the presence of an Na⁺/H⁺ exchanger, which is another system activated in various cell types during RVI (13, 16, 22). Additionally, the fact that the inhibitors 3-(N-maleimidopropinyl)biocytin and pHMB caused the cells to swell is puzzling as these compounds are both largely membrane impermeant. RVI and K⁺ uptake in Herpetomonas and K⁺ uptake and release in Leishmania were also found to be ouabain insensitive (1, 3).

The fact that K⁺ plays the major role in T. foetus volume regulation is not surprising, as most cells on a short-term basis respond to hypertonicity by rapid uptake of inorganic salts accompanied by an osmotic influx of water (13, 16, 22). While previous studies (with the exception of studies of H. samuelpessoai) showed that amino acids and other organic osmolytes are of primary importance in protozoan osmoregulatory processes, more recently there have been investigations of the presence or properties of systems involved in regulation of ionic composition in protozoans. The first information provided on K⁺ fluxes in diplomonads revealed that the efflux of K⁺, together with its counteranion, contributed approximately one-half of the total RVD (2, 19). This contrasts with the situation in Leishmania, in which a major efflux of Na⁺ and Cl⁻ and, to a lesser extent, an efflux of K⁺ occur in response to hypoosmotic stress (3, 17).

Our findings show that the response of T. foetus to anisomotic challenge is different from the response observed in all other protozoan parasites to date. While an RVD under hypoosmotic conditions is common, a response to hyperosmotic stress is less common, and only T. foetus, H. samuelpessoai (1), and L. donovani (4) have been reported to undergo an RVI. T. foetus stands alone in its ability to undergo an RVI but not an RVD. The physiological and evolutionary significance of this observation remains to be elucidated. The ability to respond to hypoosmotic challenge via an RVD is essentially a feature of eukaryotic organisms that lack a rigid cell wall and cannot resist swelling. Since trichomonads are placed on a very early branch of the eukaryotic line of descent (12), it is possible that they predate the evolution of mechanisms for responding to swelling. Alternatively, such mechanisms may have been lost because cell shape and the cytoskeleton allow moderate swelling with no compromise of cell function, and T. foetus may avoid exposure to more dramatic hypoosmotic challenge in its natural milieu in the bovine urogenital tract. Histidine kinase signaling systems are a common feature of RVI in other organisms (25). Our recent discovery of a putative histidine kinase gene in T. foetus (GenBank accession number AF462151) suggests that T. foetus may have evolutionarily conserved mechanisms for hyperosmotically induced osmolyte uptake. However, both the ion channels involved in the RVI and the signaling systems that control them require further investigation. Overall, these observations suggest that there is not a common osmoregulatory mechanism for protozoan parasites and that although there may be some common features, protozoan parasites have developed individual responses to osmotic challenge, which presumably are related to their exposure in their normal environments and to the availability of ions and organic molecules that act as osmolytes.