MeV-Ion Microprobe Analyses of Whole Drosophila Suggest That Zinc and Copper Accumulation is Regulated Storage not Deposit Excretion

Heavy metals can be essential as well as toxic. For example, while high dosages of zinc are usually toxic, zinc is a component of at least 300 enzymes and also plays an important role in gene regulation because a major category of transcription factor – zinc finger proteins – chelate zinc in a way that promotes a protein structure that binds DNA (Coleman, 1992). The mechanisms of zinc sequestration and regulation in arthropods are unknown: metallothioneins in Drosophila melanogaster bind copper and cadmium; zinc, however, neither binds to nor induces the synthesis of metallothioneins (Maroni and Watson, 1985). In addition to its well-known roles in proteins, zinc is used in large quantities by many arthropods and annelids in structures such as claws and mouthparts. The form of these accumulations is unknown and of interest because the volume fractions of zinc are thought to be too low to affect the mechanical properties as zinc–biominerals and too high for metalloproteins (measured concentrations reach 25 % of dry mass; Schofield, 1997). It is also not known how such large quantities of zinc are made available for incorporation into the cuticle during molting. Considering that heavy metals are employed in many essential roles, at times in large quantities, the assumption by most previous investigators that the accumulation of heavy metals is deposit excretion – a detoxification mechanism – now seems questionable. Here, we suggest that accumulation plays a more complex role as part of a system for regulating the biological availability of heavy metals.

Zinc, iron and copper were first reported in insects in certain types of storage vacuoles in the Malpighian tubules (the excretory organs of insects) of the housefly Musca domestica (Sohal et al. 1976). It has been suggested that these metals are accumulated in intracellular storage vacuoles as a detoxification mechanism, because they could be reabsorbed into the hemolymph if they were excreted from the Malpighian tubule into the gut (Maddrell, 1977). It was reasoned that metal ions are small in comparison with the molecules of organic toxicants and would therefore diffuse more readily through the cells of the hindgut. Recently, zinc has been identified as a prominent component of storage vacuoles in cells of Malpighian tubules in Drosophila hydei larvae (Zierold and Wessing, 1990). Drosophila have four Malpighian tubules, two anterior and two posterior. Each anterior tubule is joined with a posterior tubule at their proximal ends and connected from this ‘Y’ junction to the gut by a short duct, the ureter. In the experiments of Zierold and Wessing (1990), cross sections of tubules from near the ureters were analyzed using an electron microprobe. Zinc concentrations averaged approximately 20 000 p.p.m. (wet mass) in storage vacuoles in cells of the anterior tubules. In contrast, measured concentrations were more than an order of magnitude lower in storage vacuoles located in the posterior tubule cells.

Maddrell (1977) suggested a comparison of consumed and accumulated quantities of metals as a test of the hypothesis that metals are sequestered in Malpighian tubule cells as a detoxification mechanism. Experiments reported here make this comparison and test some predictions of the detoxification hypothesis. Our results lead us to believe that zinc and copper accumulations are not simply a detoxification mechanism, and we suggest that storage aids in the regulation of the biological availability of these metals. We use a new MeV ion microprobe technique (Schofield and Lefevre, 1993) to determine the zinc content and distribution in whole Drosophila and in dissected organs. This penetrative technique does not require sectioning, in contrast to the electron probe techniques used in the earlier experiments (Zierold and Wessing, 1990). We are not aware of other techniques that could have been used for the measurements of element quantities that are reported here. The proton probe techniques employed here are also more sensitive than electron probe techniques by 2–3 orders of magnitude (Legge et al. 1988), with sensitivity decreasing for thicker samples. The distribution of zinc that we observed in the Malpighian tubules did not completely coincide with the distribution previously reported from electron probe microanalysis (Zierold and Wessing, 1990). Instead of an asymmetric distribution, we found that both the anterior and posterior Malpighian tubules contain comparable quantities of zinc in larval D. hydei. However, we found that zinc concentrations varied unpredictably by an order of magnitude along the tubules, and we think that the previous investigators may have inadvertently taken sections of the anterior tubules from high-concentration regions and sections of the posterior tubules from low-concentration regions.

Larvae of Drosophila hydei and larvae and adults of Drosophila melanogaster were raised on either standard cornmeal, molasses and yeast medium or on instant Drosophila medium (Carolina Biological Supply; Beaverton, Oregon, USA). Samples of the media were assayed by atomic absorption analysis. The standard medium contained 11 p.p.m. iron, less than 0.3 p.p.m. copper and 140 p.p.m. zinc. The instant medium contained 36 p.p.m. iron, 0.2 p.p.m. copper and 300 p.p.m. zinc (all by dry mass). Water in the media is thought to have been the source of the high zinc content. Different batches of medium were prepared using water from the same source. Typical iron, copper and zinc concentrations in fruit are in the tens of parts per million (by dry mass; Holland et al. 1996). A number of adults were placed on instant medium prepared with 5 mmol l⁻¹ CuSO₄ (approximately 800 p.p.m., dry mass, copper in the medium) or with 5 mmol l⁻¹ ZnCl₂ (approximately 800 p.p.m. zinc, dry mass) for 1 week and then placed on unsupplemented instant medium for 1 day prior to examination.

The approximate rate of food consumption was calculated in order to compare the quantity of metal consumed with the quantity accumulated. The time required to exchange the contents of the gut was measured following the method of Maroni and Watson (1985). Adult flies were placed on medium containing powdered charcoal. At a range of times thereafter, flies were dissected to determine the distance that the dark food had progressed through the gut. By 65 min after being placed on charcoal food, 55 % of the flies dissected contained black food throughout the gut. Maroni and Watson (1985) reported similar results (1 h) for larvae. In order to calculate the rate of consumption, the volume of food in the gut was estimated by measuring the length and diameter of the food column visible within the gut. We estimated that an average of approximately 10⁻⁴ cm³ of food was consumed per hour, or approximately 4×10⁻⁵ g dry mass. While this rate is a rough approximation, the actual rate of consumption could be less than this value by more than an order of magnitude without affecting our conclusion that only a small fraction of the zinc and copper consumed was stored.

Adults and third-instar larvae were either examined whole or dissected for clarity. Whole flies were freeze-substituted with acetone followed by vacuum evaporation (Fujita et al. 1987). Dissections were arranged on 6 μm thick lexan plastic and allowed to dry in air.

We investigated the possibility that freeze-substitution with acetone caused metals to be leached from or redistributed within the specimens. To investigate leaching of metals from the flies, a number of flies were raised together and then either dried whole in air or freeze-substituted with acetone. The iron, copper and zinc contents of flies subjected to the two different treatments are compared in Table 1. Although the metal content of the flies is quite variable and the sample size small, we can conclude that, on average, less than half of the content of any of these metals could have been leached by the acetone treatment (95 % confidence level, two-sample t-test). Wilcoxon’s two-sample rank test (Snedecor and Cochran, 1989) gives the same confidence level for copper and zinc, although not for iron.

To determine whether freeze-substitution redistributed metals within the flies, the metal distribution in dissected dried flies was compared with the distribution in whole freeze-substituted flies. No differences in the distribution of iron, copper or zinc were observed.

Specimens were analyzed using a scanning ion microprobe using protons accelerated across a voltage drop of 4 MV (Lefevre et al. 1983). Two ion beam techniques were used: PIXE (proton induced X-ray emission) and STIM (scanning transmission ion microscopy). In combination, these two techniques can be used to measure the depth-averaged concentrations of a number of elements simultaneously in biological specimens (Schofield and Lefevre, 1993). PIXE data yield the quantity of an element in the sampled volume, while STIM data give a value for the total quantity of material in this volume.

For PIXE, X-rays of characteristic energy are counted for each element of interest. The possible extremes of X-ray attenuation within the specimen and of variations in the ratio between the number of X-rays produced and the quantity of the element present (due to slowing of the beam protons) are calculated. These calculated extremes are used in determining the PIXE measurement uncertainty. In cases where more precise measurements of local concentrations and element quantities are desired, tomography can be employed to reduce these uncertainties greatly (Schofield, 1995). When there was substantial overlap between the zinc and copper peaks in the X-ray data, the X-ray tallies were decreased by the appropriate fraction.

For STIM, a detector is used to measure the residual energy of single protons that have passed through the specimen. The projected density is calculated from these data. The resolution currently available with the STIM–PIXE technique is of the order of several micrometers.

Whole freeze-substituted flies were examined to ascertain the general distribution of zinc in the abdomen of a fly. Fig. 1A is a STIM image showing a sagittal view of the abdomen of an adult male D. melanogaster. Fig. 1B was made using PIXE and shows that zinc X-rays appear to originate mainly from tubule-shaped structures. Fig. 1C is also a PIXE image of this region, but in iron X-rays. While this field contains similar quantities (within a factor of approximately 2) of zinc and iron, the iron is much more generally distributed than is the zinc.

Fig. 2 shows the same fly as in Fig. 1 but in a dorsal–ventral view. Once again, the zinc appears to be localized in tubules (Fig. 2B) but the iron is more generally distributed (Fig. 2C).

The tubules extend along the dorsal and posterior sides of the abdomen and exhibit an approximate bilateral symmetry. Their location is consistent with that of the Malpighian tubules (Miller, 1965).

To determine whether Malpighian tubules were in fact the organs that accumulated zinc, we dissected out the internal organs from the abdomens of several adult and one larval D. melanogaster. These organs were placed directly onto lexan plastic discs, and the location of each organ was sketched. The organs were allowed to dry in air onto the plastic and were examined without further treatment. The zinc content of the Malpighian tubules greatly exceeded that found in any other organs, including the internal reproductive apparatus of both males and females, the entire digestive tract, the fat body, the epidermis and the muscles. Fig. 3 shows the Malpighian tubules and a portion of the midgut – other organs also produced dramatically smaller zinc X-ray signals similar to those of the gut.

Images of several undissected adult D. melanogaster, similar to Figs 1B and 2B but including images of the head and thorax, showed no other zinc accumulations comparable with those in the Malpighian tubules. A whole pupa in its puparium was also examined. No concentrations comparable with those in the Malpighian tubules of the larva were found, although the general concentration of zinc in the pupa was higher than in the organs other than the Malphigian tubules of the adults examined. Apparently, zinc was reabsorbed from the larval Malpighian tubules and dispersed. The possibility that zinc was redistributed deserves further examination since it may indicate that the stored zinc plays a role in metamorphosis. Since we had shown that the mouthparts of many arthropods contain large quantities of zinc (Schofield, 1997), we scanned the cephalopharyngeal skeleton of a D. melanogaster larva, but found no accumulation of zinc comparable with that in the Malpighian tubules.

A dissected set of Malpighian tubules from an adult D. melanogaster which had been reared on standard medium was examined to determine the distribution of zinc and calcium within this organ (Fig. 3). The position of the dissected tubules on the holder were sketched (Fig. 3A) for comparison with the STIM image (Fig. 3B). Examination of the tubules in calcium X-rays showed that calcium is accumulated in the distal (initial) segments at the end of each of the tubules (Fig. 3C). This confirms earlier work that showed that calcium is present in concretions that fill the lumen of the distal segments of Drosophila Malpighian tubules (Sohal et al. 1976; Wessing and Eichelberg, 1978). It is likely that the calcium accumulation shown here is located in these concretions.

Fig. 3D shows the distribution of zinc in these tubules. Very little zinc was found in the ureters, the region between the gut and the splitting of the tubule pair. This is consistent with the lack of intracellular storage vacuoles in the proximal region of the ureter (Sohal et al. 1976; Wessing and Eichelberg, 1978; Zierold and Wessing, 1990). The distal segments of the tubules also contain less zinc than the main segments of the tubules, although the difference is less pronounced than the difference between the main segments and the ureters. The region of low zinc content in the distal segments overlaps with the region of high calcium content, suggesting cellular specialization along the tubule (contrast Fig. 3C,D).

To compare our data with previous reports (Zierold and Wessing, 1990), the zinc concentration and distribution were determined in the proximal portion of the Malpighian tubules of a third-instar D. hydei larva (Fig. 4). Zinc was found both in the anterior pair of tubules and in the posterior pair. The ureter contained comparatively little zinc, especially in the proximal portion.

To make quantitative comparisons of zinc concentrations, we examined each pixel of the 64×64 pixel PIXE image. The areal (projected) density of zinc obtained from the PIXE data was divided by the total areal density calculated from the STIM data. This zinc concentration is a mean value for the volume traversed by the beam and so is depth-averaged. The zinc concentration varied considerably in the main portions of the tubules. Over the field of this figure, the zinc concentration in the anterior tubules varied from 28 000±7000 p.p.m., dry mass, to 2000±1000 p.p.m. (the ranges delineate a 68 % confidence interval). Each of the measured concentration values is for a volume of the specimen that projects onto a single pixel of Fig. 4B. These volumes are large in comparison with the size of intracellular zinc-containing vacuoles, which were measured individually by Zierold and Wessing (1990) using electron probe microanalysis and compared with the surrounding cytoplasm; the sampled volumes for PIXE in the current experiments are approximately the size of an entire Malpighian tubule cell.

Mean zinc concentrations in the field of this image were calculated in order to compare the anterior and posterior tubules and the gut. The mean concentration for the main segment of the two anterior tubules was 12 000±2000 p.p.m., dry mass; for the two posterior tubules, it was 8000±2000 p.p.m. The mean concentration of zinc in the portion of the gut shown in Fig. 4 was found to be 280±50 p.p.m., dry mass, which was comparable to the zinc concentration in the food. The relatively few zinc X-rays originating in the gut are not indicated with the gray scale used for Fig. 4B. Zinc concentrations decreased across the ureter from the high values of the main segments to the low values of the gut.

Several other dissections were examined to test the generality of these observations. Results for all of the dissections are included in Table 2. Mean concentrations were calculated from data at ten or more locations spaced at even intervals along the portion of the organ that was scanned. The 68 % confidence intervals for the absolute concentration in the pooled sampled volumes are given, as an indication of the uncertainty associated with the technique, and not the standard deviation for ten individual measurements. When concentrations were not calculated, observational data were included in the table. Main segments of all tubules contained elevated concentrations of zinc. In contrast, the zinc concentration in all the ureters and distal tips of the tubules examined was substantially lower.

For a quantitative comparison between zinc accumulation and copper accumulation, a number of adults were placed on food containing 5 mmol l⁻¹ ZnCl₂ or 5 mmol l⁻¹ CuSO₄ (approximately 800 p.p.m., dry mass, added metal). After 1 week, flies were placed on unsupplemented food for 1 day and then freeze-substituted with acetone. In Table 3, the abdominal contents of these metals are compared for these flies as well as for flies raised on unsupplemented medium.

Maroni and Watson (1985) found low-molecular-mass proteins, thought to be metallothioneins, that bound cadmium and copper in D. melanogaster. Zinc, however, did not bind to or induce a similar protein. Our observations support these earlier results because we found no correlation between zinc and sulfur distributions. Metallothioneins are characterized by a ratio of sulfur (in cysteine) to metal of approximately 3:1. Nevertheless, Table 3 shows that the fly abdomens contained more zinc than copper. The flies must employ an effective sequestration mechanism other than the metallothionein mechanism.

Some copper may have been sequestered in a manner similar to zinc. Copper has been shown to accumulate in storage vacuoles in the epithelium of the larval midgut (Filshie et al. 1971; Tapp and Hockaday, 1977). Two of several hundred storage vacuoles examined using an electron microprobe contained high concentrations of zinc instead of copper (Tapp and Hockaday, 1977). The zinc-containing granules were indistinguishable in appearance from the copper-containing granules. The storage vacuoles containing copper (and occasionally zinc) in the midgut epithelium (Filshie et al. 1971; Tapp and Hockaday, 1977) and the vacuoles containing zinc in the Malpighian tubules (Zierold and Wessing, 1990) are similar in ultrastructure. These similarities suggest the possibility that the mechanisms of copper accumulation in the midgut and of zinc accumulation in the Malpighian tubules are also similar. Maroni and Watson (1985) noted the presence of copper in the particulate fraction of homogenates and suggested that this fraction of the copper (15–25 %) was not associated with metallothionein. Perhaps this fraction of copper is sequestered in a manner similar to zinc.

The different distribution patterns of copper and zinc are demonstrated in Fig. 5, which shows an adult D. melanogaster fed food containing 5 mmol l⁻¹ CuSO₄ (approximately 800 p.p.m., dry mass, copper) and approximately 300 p.p.m. zinc (not supplemented) and chased with unsupplemented food containing less than 1 p.p.m. copper. The PIXE images in the top row (Fig. 5A–C) show sagittal views in iron, copper and zinc X-rays, respectively. The bottom row of PIXE images (Fig. 5D–F) are dorsal–ventral views, again in iron, copper and zinc X-rays, respectively. The copper images have been corrected for overlap of copper and zinc X-ray peaks. Copper contrasts with zinc in that it is distributed generally and not only in the Malpighian tubules. The general distribution of copper in copper-fed flies, including its appearance in the Malpighian tubules, has been noted previously (Marchal-Segault et al. 1990). The localized zinc distribution pattern suggests specialized regulation of zinc and indicates that sequestration mechanisms for copper and iron discriminate against zinc.

Maddrell (1977) has suggested that the amount of metals consumed by flies should be compared with the accumulated quantity as a test of the hypothesis that Malpighian tubules store metals as a detoxification mechanism. Our blackened food studies indicate that 8-day-old adults fed on food containing 300 p.p.m., dry mass, zinc have consumed roughly 2.3×10⁻⁶ g of zinc since eclosion. The fly shown in Fig. 1 (fed on 300 p.p.m. zinc) had accumulated only 2.1×10⁻⁸±0.8×10⁻⁸ g of zinc in its abdomen; this is less than 1 % of the estimated quantity consumed.

In his detoxification hypothesis, Maddrell (1977) suggested that heavy metals are permanently stored in, rather than excreted from, the Malpighian tubules because they would be reabsorbed from the hindgut if excreted and, presumably, would return to the Malpighian tubules. For this cycling of metals to be problematic, a substantial proportion of the metals passing through the gut would have to be absorbed. Thus, the detoxification hypothesis predicts that a substantial proportion of the consumed zinc would be absorbed and stored. Our finding that less than 1 % of the consumed zinc was stored is thus inconsistent with the detoxification hypothesis.

Since a food concentration of 300 p.p.m. of zinc is high relative to typical concentrations in fruit (tens of parts per million; Holland et al. 1996), it is possible that the storage mechanism was saturated. However, if the storage mechanism were saturated, then it is unlikely that storage is a very important detoxification mechanism because the flies fed on this undoped medium appeared normal and healthy, and zinc is fatally toxic only at considerably higher dosages (larvae, Maroni and Watson, 1985; adults, R. M. S. Schofield, J. H. Postlethwait and H. W. Lefevre, unpublished observations).

Our observations of other flies also did not support the hypothesis that storage is necessary to prevent re-uptake of toxic elements. The flies reported in Table 3 had all accumulated less than 1 % of the estimated quantities of zinc consumed. As with zinc, less than 1 % of the consumed copper was accumulated in flies fed food containing approximately 800 p.p.m., dry mass, copper. However, approximately three times the estimated quantity of copper consumed since eclosion was found in flies fed food containing approximately 0.2 p.p.m. copper, suggesting that copper had been retained from the larval stage.

If metals were permanently stored to remove them from the metabolic pool (deposit excretion), we would expect the quantity of metal in the fly to be roughly proportional to the quantity of metal consumed. If it is assumed that the metals diffuse through the gut wall into the body cavity, the rate of diffusion given by the diffusion equation is proportional to the concentration difference between the lumen of the gut and the hemolymph. If the concentration of metal in the hemolymph is low as a result of efficient sequestration of the metal or if the metal concentration in the hemolymph is approximately proportional to that in the gut, then the diffusion rate is approximately proportional to the concentration in the gut. Therefore, if metals diffused into the hemolymph, were sequestered and permanently stored, then the quantities of these metals accumulated in a given period would be proportional to the quantity consumed. Table 3 shows that this proportionality was not observed for either copper or zinc. The copper content of the abdomen was relatively invariant, varying by only a factor of 5, while the copper concentration in the medium changed by a factor of more than 2500, from approximately 0.2 p.p.m. (more than an order of magnitude lower than typical concentrations in fruit) to 800 p.p.m. The quantity of zinc accumulated was also relatively invariant, although the concentration of zinc in the food was only varied by a factor of approximately 4. The possibility that food intake was greatly reduced for zinc-or copper-supplemented food must be considered, but this seems unlikely because supplementation did not reduce the quantities of the unsupplemented metal accumulated (Table 3). The observed lack of proportionality between the medium concentrations and the quantities of metal accumulated is thus inconsistent with deposit excretion. Our observations are, instead, consistent with the hypothesis that flies actively regulate their zinc and copper contents, attempting to maintain stores that are independent of metal concentrations in their diets.

The ability to take up and locally to sequester large quantities of zinc is widespread among the arthropods. Zinc accumulations have been reported in the region immediately surrounding the midgut of an ant (Schofield et al. 1988). The zinc concentration in this region, which includes the Malpighian tubules, reached 5000 p.p.m., dry mass, (Schofield, 1990), comparable to the concentrations we found in fly Malpighian tubules here. It was suggested that this intestinal accumulation might be related to the high concentrations (40 000 p.p.m., dry mass) of zinc found at the cutting edge of the ant mandibles. Such cuticular accumulations of zinc have been found in the mandibles of insects (Hillerton and Vincent, 1982), the tarsal claws of insects (Schofield, 1990), the fangs of spiders (Schofield and Lefevre, 1989), the chelicera, pedipalpi, tarsal claws and stings of scorpions (in concentrations reaching 250 000 p.p.m., dry mass) and the mechanical structures of a number of other organisms (Schofield, 1997), although not in fruit flies. Even though cuticular accumulations of zinc are not present in fruit flies, the mechanisms of zinc uptake and storage may well be similar to those in these other arthropods.

While we are unaware of arthropods that use large quantities of copper, the relative invariance of stored quantities of copper with respect to food concentrations prompts the suggestion that storage of both copper and zinc is beneficial in regulating the biological availability of these metals.

In conclusion, when flies were fed on food containing several hundred parts per million of zinc or copper, a minor fraction of the metal consumed was stored. The site-specific storage of zinc in the Malpighian tubules contrasted markedly with the more dispersed accumulations of iron and copper. This specificity in metal accumulation is more suggestive of regulation for biological availability than of deposit excretion. More compelling, the failure to observe a proportionality between the quantities of zinc and copper estimated to have been consumed and the quantities accumulated is inconsistent with deposit excretion and is, instead, consistent with the hypothesis that storage is regulated, perhaps to control biological availability. The technique demonstrated here promises to be of further use in the study of the metabolism of minor element because, in contrast with alternative techniques, whole organisms can be imaged and element quantities in specified volumes or organs can be measured.

We would like to thank Mike Grimes and Dr Cao Hua for artwork and Ariane deLuca for summarizing German references.