X-ray microanalysis of Rb⁺ entry into cricket Malpighian tubule cells via putative K⁺ channels

The initial stages of insect excretion occur in the secretory Malpighian,or renal, tubules and depend upon the near isosmotic transport of water into the tubule lumen. Water transport is driven by the transepithelial transport of K⁺, Na⁺ and Cl^–. These processes of ion transport in insect Malpighian tubules have been extensively reviewed(Maddrell and O'Donnell, 1992; Nicolson, 1993; Van Kerkhove, 1994; Beyenbach, 1995; Pannabecker, 1995; Dow and Davies, 2001; Beyenbach, 2003; Dow and Davies, 2003). It is clear that some variation in the processes involved occurs between different species. The entry of K⁺ into the tubule cells across the basolateral cell membrane appears to involve three possible mechanisms. The presence of Na⁺/K⁺-ATPase in Malpighian tubules(Fogg et al., 1991) has suggested a possible means of K⁺ entry (e.g. Weihart et al., 2003a; Torrie et al., 2004), and a Na⁺/K⁺/2Cl^– co-transporter has also been implicated (e.g. Wiehart et al.,2003a; Ianowski and O'Donnell,2004; Scott et al.,2004). The third possible mechanism for K⁺ entry is via Ba²⁺ sensitive K⁺ channels (e.g. Leyssens et al., 1993; Leyssens et al., 1994; Wiehart et al., 2003b; Scott et al., 2004). The role of the latter in basolateral transport processes in vertebrate epithelia has been recently reviewed (Cotton,2000; Warth,2003).

Secretory processes in the Malpighian tubules of the black field cricket Teleogryllus oceanicus have been extensively investigated(Marshall et al., 1993; Marshall and Xu, 1999; Xu and Marshall, 1999a; Xu and Marshall, 1999b; Xu and Marshall, 1999c; Xu and Marshall, 2000). The Malpighian tubules consist of three segments: short proximal and distal segments and a long main segment. Xu and Marshall have shown that Na⁺/K⁺-ATPase has only a minor role in K⁺secretion in the main segment and therefore transport across the basolateral membrane (Xu and Marshall,1999a). These authors also showed that a Na⁺/K⁺/2Cl^– co-transporter was involved in K⁺ entry into the cells.

Using Rb⁺ as a tracer for K⁺, Marshall and Xu showed that the main segment secreted fluid at a slightly higher rate when Rb⁺ was substituted for K⁺ in the incubating saline and that Rb⁺ replaced much of the intracellular and secreted K⁺(Marshall and Xu,1999). Because ouabain, a potent inhibitor of Na⁺/K⁺-ATPase, resulted in an increase in intracellular Na but did not change the concentration of intracellular Rb, it was concluded that Rb⁺ did not enter the cell viaNa⁺/K⁺-ATPase. It was considered that a major route for Rb⁺ entry was possibly via Ba²⁺ sensitive K⁺ channels since Xu and Marshall showed that incubating saline containing Ba²⁺ reduced fluid secretion by the main segment by 40%(Xu and Marshall, 1999a).

Rubidium has been commonly used as a tracer for K⁺ in the study of K⁺ channels (e.g. Niki et al., 1990; Kupriyanov et al.,1995). Using Rb⁺ as a tracer for K⁺, in combination with quantitative X-ray microanalysis and elemental imaging(Marshall and Xu, 1998), we have followed the entry of Rb⁺ into the Malpighian tubule cells. The aim of the present investigation was to determine whether Ba²⁺blocked Rb⁺ entry into the main segment cells of the tubule thereby indicating whether Ba²⁺ sensitive K⁺ channels are a major route for K⁺ entry into Malpighian tubule cells in the cricket.

Experimental animals were adult male black field crickets (Teleogryllus oceanicus L.) that had been raised in the Zoology Department at La Trobe University (Melbourne, VIC, Australia) and had ecdysed four weeks prior to the experiment. The colony was reared at 27°C with a relative humidity of approximately 70% and a photoperiodic regime of 12 h:12 h light:dark reverse cycle. They were fed ad libitum on fresh lettuce and Barastoc mouse breeder ration (Ridley Agri Products, Pakenham, Victoria, Australia) and had access to water. Experiments were started during the animal's dark phase,which is their active period. All crickets were killed by decapitation.

Various artificial treatment salines were prepared for different experiments. These were modifications of Coast's saline(Coast, 1988) and the chemical composition of these is given in Table 1. All chemicals were supplied by Sigma-Aldrich (St Louis, MO,USA). The salines were adjusted to a pH of 7.1 with 1 mol l^–1NaOH. The osmotic concentration of each saline was determined using a vapour pressure osmometer (Wescor 5500, Logan, UT, USA) and ranged between 330 and 350 mosmol kg^–1. The salines were used at room temperature(22°C) and air was bubbled through 10 min before use. The treatment salines were kept at 4°C for no more than three days.

The Malpighian tubules of T. oceanicus number approximately 150 and drain into a common ampulla that, in turn, empties into the hindgut via a cuticle-lined ureter. From each of eight crickets, two half ampullas with Malpighian tubules were pre-incubated in normal Coast's saline(NS) for 60 min. One set of half ampullas was incubated for a further 60 min in NS as the control group, and the other half was changed to Coast's saline containing 3.5 mmol l^–1 Ba²⁺ (Ba)(Table 1) for a further 60 min incubation as the treated group. Tubules from a further two groups of eight crickets were treated similarly except that the treatment salines were modified Coast's saline containing 8.6 mmol l^–1Rb⁺ (Rb) and modified Coast's saline containing 8.6 mmol l^–1 Rb⁺ and 3.5 mmol l^–1Ba²⁺ (RbBa) (Table 1).

Following incubation the tubules were briefly dipped into fresh final incubation medium containing 20% (w/vol.) dialyzed bovine serum albumin (BSA),mounted on aluminium microtome pins in a humidity chamber and frozen in liquid propane at –190°C using a plunge freezing device (Leica KF80,Wetzlar, Germany). The frozen samples were planed with a diamond trimming knife in a cryo-ultramicrotome (Leica FC4E) at –120°C. Planed samples were mounted in a substage suitable for cryo-scanning electron microscopy (cryoSEM). This was then transferred directly from the microtome chamber without exposure to air, using the Oxford Instruments transfer device(Oxford Instruments CT1500, Oxfordshire, UK), to the cryo-preparation chamber mounted on a scanning electron microscope (SEM) (JEOL JSM840A, Tokyo, Japan). The sample was then coated with 20 nm Cr at 103 K and transferred to the custom built cryo-stage in the SEM. This cryostage was cooled to–171°C by thermal transfer via copper braid to a cold finger, through which liquid nitrogen circulated from a 10 l stainless steel dewar. Another liquid nitrogen dewar cooled a large copper anti-contamination plate to 81 K. The holding capacity of the stage dewar was 24 h and that of the anti-contaminator was sufficient for overnight operation.

The microscope was equipped with a Link eXL X-ray analyser and a Link LZ4 30 mm² detector having both an ultrathin window (UTW) and a beryllium (Be) window. The take-off angle was 40 deg., and the solid angle was 0.011 steradians (sr). Beam current was monitored by means of a Faraday cup inserted into the beam and was stable (±2%) over periods of 120 h continuous operation.

X-ray microanalysis was used to determine elemental distributions and concentrations in the cytoplasm of the planed, frozen-hydrated Malpighian tubules. Full details of this method are given in Marshall and Xu(Marshall and Xu, 1998). Briefly, analyses were made in the form of quantitative elemental images, or maps (64×64 pixels with a dwell time of 6 s per pixel), at an accelerating voltage of 15 kV and a beam current of 1×10^–9 A. Quantitative data were subsequently extracted from selected regions of these images (usually the entire cross section of cells in a tubule) using the program Featurescan (Oxford Instruments). The Link X-ray imaging software for quantitative mapping calculates peak integrals, using multiple least squares fitting of library peaks to the spectrum for each pixel, and apparent concentrations, i.e. concentrations that have not been corrected for matrix effects. Apparent concentration images were corrected arithmetically to true concentration images by applying matrix corrections that were derived from static beam analyses of the same samples. The Phi-Rho-Zed model developed by Bastin and colleagues(Bastin et al., 1986) as applied to biological samples (Marshall,1982) was used to obtain quantitative analyses. The standards used for elements Z>10 were polished microprobe standards (Bio-Rad,Hercules, CA, USA) and these were also used as primary standards for C, N and O. However, secondary concentration standards for C, N and O were also derived from frozen-hydrated solutions of 20% BSA.

All elements were analysed from their K-line X-ray emissions except for Rb,which had to be analysed using the L-line emission. Because quantitative analysis using L-lines is inherently less accurate, the accuracy of Rb analysis was tested using frozen-hydrated preparations of 100 mmol kg^–1 RbCl and 100 mmol l^–1 KCl in 20%(w/vol.) BSA.

As no structural details can be seen in fully hydrated, cryoplaned tubules,untreated tubules were also frozen by high pressure freezing (HPF) (Leica HPF)for structural imaging. These samples were fractured under liquid nitrogen by separating the sample preparation `freezer hats', briefly etched in a cryo-preparation chamber (Oxford Instruments CT 1500), coated with 2 nm Pt and viewed on a cryostage cooled to –140°C at 2 kV in a Field Emission SEM (JEOL JSM 6340).

Data were statistically analysed by analysis of variance and comparison of means using the Tukey–Kramer HSD test. Analyses were carried out with the aid of the JMP software package (SAS, Cary, NC, USA). All results are presented as means ± standard error of the mean (±s.e.m.).

Whilst freezing by high pressure freezing is superior to propane plunging for structural preservation, the same major structural features in the main segment of the Malpighian tubules are preserved and identifiable in fractured tubules that have been cryofixed in propane (not shown). From the HPF samples,it can be seen that the structure of the single cell epithelium is characterised by well-defined basal infoldings resting on a thick basal membrane, cytoplasm containing spherical inclusions and nuclei, and prominent microvilli at the apical side of the cells. These enclose the fluid-filled tubule lumen that frequently contains spherites, which are predominantly calcium phosphate (A.T.M., unpublished)(Fig. 1).

The accuracy of Rb detection and quantification by X-ray microanalysis using the Rb L-line was verified from analyses of model solutions containing RbCl. The accuracy was determined to be sufficient, with analyses of frozen-hydrated solutions of 100 mmol kg^–1 RbCl+100 mmol kg^–1 KCl in 20% BSA, giving values of 100.7±2.5 mmol kg^–1 Rb and 99.7±7.8 mmol kg^–1 K(N=3 preparations).

It should be noted that X-ray microanalysis is unable to distinguish between ionised and non-ionised elements. Cellular elements measured by X-ray microanalysis are represented, therefore, by elemental symbols without an indication of ionisation state. Additionally, because the planed surface of frozen-hydrated Malpighian tubules is featureless in the SEM, the location and arrangement of the tubules was determined from oxygen X-ray images(Fig. 2A–D) as described in Marshall and Xu (Marshall and Xu,1998).

Tubules incubated in saline in which Rb⁺ replaced K⁺,showed a marked reduction in cellular and luminal K(Fig. 2C) compared with tubules incubated in normal saline (Fig. 2A), and much of this K was replaced by Rb in both cells and lumen(Fig. 2C). Tubules incubated in saline containing Ba²⁺ showed a considerable reduction in cellular and luminal K (Fig. 2B)compared with control tubules incubated in normal saline(Fig. 2A). Similarly tubules incubated in saline containing Rb⁺ and Ba²⁺ showed a reduction in both cellular and luminal K(Fig. 2D) compared with control tubules incubated in saline containing K⁺(Fig. 2A). Replacement of K by Rb was also considerably reduced in these tubules(Fig. 2D) compared with tubules incubated in saline containing Rb⁺ only(Fig. 2C).

The results for cellular element concentrations extracted from X-ray images(e.g. Fig. 2A–D) are summarised in Fig. 3. There were no significant differences (P>0.05) in Cl concentration between treatments. Cytoplasmic K concentration fell significantly(P<0.05) in Ba²⁺ treated tubules to 44% of the concentration in control tubules, indicating that a substantial fraction of cytoplasmic K enters via Ba²⁺ sensitive K⁺channels. In tubules treated with Rb⁺, a significant(P<0.05) 69% of cellular K in control tubules was replaced by Rb while 19% of cellular K was not replaced. After treatment with both Rb⁺ and Ba²⁺, the amount of Rb was significantly(P<0.05) reduced compared with the amount of Rb in tubules treated with Rb⁺ only. However, the amount of cellular K was not significantly different (P>0.05), suggesting the presence of a small K pool that is neither replaced by Rb nor affected by Ba²⁺sensitive K⁺ channels. Consistent with this, the combined concentration of cellular K and Rb recorded after Rb⁺ and Ba²⁺ treatment, totalled 87% of the amount of cellular K seen in tubules after Ba⁺ treatment only.

Three major mechanisms for basolateral K⁺ entry into insect Malpighian tubule cells have been proposed. These are entry by a Na⁺/K⁺-ATPase (e.g. Torrie et al., 2004), a Na⁺/K⁺/2Cl^– co-transporter (e.g. Ianowski and O'Donnell, 2004)and Ba²⁺ sensitive K⁺ channels (e.g. Beyenbach and Masia, 2002). However, the relative importance of each of these mechanisms in K⁺transport appears to vary considerably between different groups of insects.

In the isolated Malpighian tubules of the cricket T. oceanicus,incubation in Rb⁺ containing saline results in the main segment of the tubules functioning normally with fluid secretion rates maintained and with Rb⁺ almost entirely replacing K⁺ in the secreted fluid (Marshall and Xu, 1999). It is a reasonable assumption, therefore, that Rb⁺ is entering the cells via at least some of the processes that normally facilitate K⁺ entry. The results suggest that in the main segments of isolated Malpighian tubules of the black field cricket (a) approximately 19% (27 mmol kg^–1 wet mass) of cellular K is not exchangeable by Rb⁺, (b) 56% (79 mmol kg^–1 wet mass) of cellular K enters by Ba²⁺ sensitive K⁺ channels, (c) 70% (69 mmol kg^–1 wet mass) of cellular Rb enters by Ba²⁺sensitive K⁺ channels with 30% (29.4 mmol kg^–1 wet mass) of cellular Rb entering by other means, and (d) 18% (26 mmol kg^–1 wet mass) of cellular K enters by a separate mechanism and this is not available to Rb⁺. The latter is inferred as follows.

A small cellular pool of K in the main segment cells, which is principally a cytoplasmic pool that provides only a small fraction of secreted K⁺, was thought to be maintained by Na⁺/K⁺-ATPase activity(Xu and Marshall, 1999b). We suggest that this pool corresponds to the small amount of K remaining(18–19%) in the main segment cells after incubation in Rb⁺containing saline. This K fraction was also present after Ba²⁺treatment during Rb⁺ incubation. This is supported by the observations of Marshall and Xu who analysed freeze-dried sections of Malpighian tubule main segment cells(Marshall and Xu, 1999). These authors found that Rb⁺ did not completely replace cellular K after incubation in Rb⁺ containing saline. Furthermore, they showed that ouabain failed to reduce the concentration of Rb attained in the cells during incubation in Rb⁺ containing saline, indicating that Rb⁺is not transported into these cells by the Na⁺/K⁺-ATPase.

It is apparent from the analyses of main segment cells, in tubules treated with Ba²⁺ after incubation in normal (K⁺ containing)saline, that a large fraction (56%) of cellular K in the main segment cells enters via Ba²⁺ sensitive K⁺ channels. Of the remaining fraction we have already suggested that 18% represents a cytoplasmic pool maintained by Na⁺/K⁺-ATPase. The source of the remaining 26% of cellular K may be determined from a consideration of the analyses of main segment cells of tubules treated with Ba²⁺ during incubation in Rb⁺ containing saline.

Following Ba²⁺ treatment of tubules incubated in Rb⁺containing saline, a fraction (30%) of cellular Rb remained that did not enter via K⁺ channels or, as previously reasoned, viaNa⁺/K⁺-ATPase. Xu and Marshall have argued that K⁺ may also enter the main segment cells of TeleogyllusMalpighian tubules via a Na⁺/K⁺/2Cl^– co-transporter(Xu and Marshall, 1999b). They showed that treatment with furosemide, an inhibitor of the Na⁺/K⁺/2Cl^– co-transporter, resulted in a large decline in tubular fluid secretion rate and a marked change in composition of the fluid. There was, however, no change in intracellular K concentration. We suggest that the fraction of cellular Rb remaining after Ba²⁺ blockade is derived from entry of Rb⁺viathe Na⁺/K⁺/2Cl^– co-transporter and represents the amount of K⁺ entering the main segment cells by this route.

Barium is a nonselective blocker of K⁺ channels (e.g. Vergara et al., 1999; Cotton, 2000; Jiang and MacKinnon, 2000; Alagem et al., 2001) so it is not possible to speculate on the type of K⁺ channel that may be involved in basolateral K⁺ transport in the Malpighian tubules of the cricket. Numerous types of K⁺ channels have been implicated in epithelial transport processes (Cotton,2000; Warth, 2003)but no specific types have so far been identified in insect Malpighian tubules. However, the involvement of inward rectifier K⁺ channels has been suggested by Evans et al. (Evans et al., 2005) in Drosophila Malpighian tubules, and ATP activated K⁺ channels have been implicated in the tubules of Tenebrio molitor (Wiehart et al.,2003b).

To summarise, we suggest that K⁺ enters the cells of the main segment of unstimulated Teleogryllus Malpighian tubules primarily via Ba²⁺ sensitive K⁺ channels. This route accounts for 56% of cellular K. A further 26% of cellular K enters via a Na⁺/K⁺/2Cl^–co-transporter. The remaining cellular K (18%) enters viaNa⁺/K⁺-ATPase and forms a non-exchangeable cytoplasmic K pool that makes a minor contribution to secreted K⁺. It is a reasonable presumption that these transport processes contribute to the K⁺ in the secreted fluid in a similar ratio to their contributions to cellular K.