Bismuth Autometallography: Protocol,Specificity,and Differentiation

Abstract

We provide a detailed protocol of the autometallographic bismuth technique and evaluate the specificity of the technique. We show by the multi-element technique “proton-induced X-ray microanalysis” (PIXE) that the autometallographic grains contain silver, bismuth, and sulfur, proving that autometallography can be used for specific tracing of bismuth bound as bismuth sulfide clusters in tissue sections from Bi-exposed animals or humans. In sections from animals exposed concurrently to selenium and bismuth, the autometallographic grains also contain selenium. This demonstrates that, if present in excess in the organisms, selenium will bind to exogenous bismuth, creating bismuth selenide clusters. As a further possible control for specificity and as a tool for differentiating among autometallographically detectable metals in sections containing more than one, we describe how bismuth sulfide clusters can be removed from Epon-embedded tissue sections by potassium cyanide.

The necessary observations for performing autometallography (AMG) were accomplished in the 19th century by an unknown photographer. He found that an exposed photographic plate, from which the silver bromide crystals had been removed, could not be developed by a normal chemical developer. However, if he added silver nitrate to the developer the picture emerged. Such a silver ion-enriched chemical developer was called a “physical developer” because it was believed that the developing process differed from that of the chemical developer. For no reason whatever, it seems, it was assumed that if the silver ions were added to the developer they were reduced to silver atoms near the catalyst before moving “physically” to the surface of the catalyst, as opposed to the “normal” chemical development in which the silver ions (released from silver bromide crystals in the film) were believed to adhere to the catalyst first and then to be reduced in situ to metallic silver atoms.

Liesegang (1911) introduced the technique in histology as a way of silver-enhancing silver in tissue sections treated with silver salts. His intention was to apply Ramon y Cajal's en bloc staining to tissue sections (see also Danscher 1982). Since that time, there has been an increasing use of this special silver-containing photographic developer in histology (e.g., Roberts 1935; Timm 1958; Haug 1967, 1973; Brunk et al. 1968; Hacker et al. 1988; Lormée et al. 1989; Stierhof et al. 1992; Ross et al. 1996; Danscher et al. 1997b; Hainfeld et al. 1999). The term “autometallography” has been proposed as the formal name for the technique when it is performed on tissue sections. This term refers to the fact that certain metal-containing clusters graph their own presence in silver if exposed to an AMG developer (Danscher 1984). It is believed that electrons released from hydroquinone molecules cohering to an AMG cluster cause a reduction of silver ions, which have connected to the catalytic crystal lattices, to silver atoms. Because silver atoms are AMG catalytic, the process will continue until the original nano-sized catalytic crystallites have been amplified to visible silver crystals (Figure 1). The AMG technique allows clusters containing only a few atoms to be observed at both the EM and the LM level. Detailed protocols have been worked out for each AMG metal and they include ways of differentiating among the individual metals (Danscher et al. 1997a).

Figure 1

A “theoretical scheme” of the AMG process. Bismuth–sulfide–selenide molecules (Bi) create a crystallite/cluster that catalyzes the reduction of adhering silver ions (Ag⁺) to metallic silver atoms (Ag°). The electrons originate from reducing molecules, e.g., hydroquinone, that adhere to the catalyst and later to the growing silver shell.

That sections from bismuth-treated mice contained catalytic specks that could be traced with AMG was shown by Ross et al. (1996). These authors found AMG grains in sections from brain, kidney, and liver of mice that had been exposed IP to huge amounts of bismuth subnitrate for several weeks.

Bismuth was established as a separate element by Claud J. Geoffroy in 1753. Whereas its soluble compounds are poisonous, its insoluble compounds have been used for centuries as a remedy for several intestinal ailments and disorders, e.g., for treatment of syphilis, as a diuretic, and as an anti-hypertensive drug. Even today such pharmaceutics are used in developed and in third world countries. Relatively recent applications are the use of a combination of antibiotics and bismuth-containing drugs (Hopkins 1997) for curing peptic ulcers and the use of bismuth in a paste used by neurosurgeons to enhance growth of granulation tissue (Sharma et al. 1994). Bismuth compounds reduce the toxicity of some forms of cancer therapy and are used as additives to dental fillers or bone implants and in catheters and surgery threads to make them visible to X-rays and scanners. They are also used to alleviate diaper rash and help colostomized patients by virtue of their deodorizing capacity. Bismuth-containing products are furthermore used as antiseptic powders, ointments, and burn bandages to improve wound healing.

Apart from its use in medicine, bismuth is an important element in atomic plants (Weast 1975–1976) as an additive to plastics, conveying flame-retardant and smoke-inhibiting qualities to the products. Bismuth catalysts are widely used in industrial organic chemistry, and bismuth compounds are increasingly used in coloration of glass and as a component of metallic paints, as an additive to cosmetics, and in bismuth shotgun pellets (Pamphlett et al. 2000). This vast and quickly growing utilization of bismuth for a multitude of different purposes indicates that this heavy metal is very much a part of our environment and that research into how it is metabolized and what possible toxic effects it may hide seems worthwhile. The present technique, with its ability to localize bismuth–sulfur–selenium clusters in tissue sections at the LM and EM levels, makes it a valuable tool in such efforts.

Materials and Methods

Twenty mice of the BALB/c strain and 25 Wistar rats (Møllegaards Breeding Center; Ejby, Denmark) were used. The animals were housed in plastic cages under the following conditions: 12-hr light/dark cycle, 22C, and 50% humidity. They were fed Altromin No 1314 from Brogaarden (Gentofte, Denmark) ad libitum and had free access to tapwater. The study was undertaken in accordance with the Danish and University of Aarhus guidelines for animal welfare. Before perfusion the animals were anesthetized with sodium pentobarbital (50 mg/kg bw) and sacrificed by transcardial perfusion with 3% glutaraldehyde in a 0.1 M phosphate buffer or with 1% paraformaldehyde and 2% glutaraldehyde in a 0.1 M phosphate buffer.

Figure 2

Micrograph of 30-μm cryostat AMG-developed section from rat cerebellum. The animal had 15 bismuth gunshot pellets IP for 1 month before sacrifice by transcardial perfusion. AMG-developed for 60 min and counterstained with toluidine blue. Note the intense AMG staining of the glia and granular cell layer. Note also the staining of bismuth accumulations in the capillary wall, a phenomenon seen most often in the cerebellum. Bar = 27 μm.

Figure 3

Cryostat section (30 μm) from rat spinal cord. The animal had 15 bismuth shotgun pellets (Eley 12) placed in its peritoneal cavity for 3 weeks. AMG-developed for 60 min and counterstained with toluidine blue. Abundant AMG staining is seen in the motor neurons. Bar = 23 μm.

Figure 4

Photograph of a 30-μm cryostat section of rat kidney. The animal had a 1000 mg/kg IP injection of bismuth subnitrate 3 months before sacrifice by transcardial perfusion with glutaraldehyde. AMG-developed for 60 min and counterstained with toluidine blue. The intense staining of the proximal tubules is in contrast to the complete renal corpuscle devoid of AMG grains. Bar = 27 μm.

Figure 5

Micrograph of 3-μm Epon section from rat kidney. The animal had bismuth shotgun pellets implanted for 1 month before it was anesthetized and sacrificed by transcardial perfusion with buffered glutaraldehyde. AMG-developed for 60 min and counterstained with toluidine blue. AMG grains are seen in both proximal and distal convoluted tubule cells. Bar = 27 μm.

Figure 6

Electron micrograph of a bismuth-loaded neuron in the subiculum of a rat brain. The animal was treated with 500 mg/kg bismuth subnitrate IP and sacrificed 2 months later. AMG-developed for 60 min and counterstained with lead citrate and uranyl acetate. The AMG-traced bismuth clusters are located in lysosomes. Bar = 950 nm.

Figure 7

EM picture from rat cerebellum of a rat treated with 500 mg/kg bismuth subnitrate IP and allowed to survive for 2 weeks. AMG-developed for 60 min and counterstained with lead citrate and uranyl acetate. The capillary is located in the molecular layer. Note that the AMG grains are located in the basal lamina of the vessel. A major bismuth-containing organelle, most likely a phagocyte, is seen close to the capillary. Bar = 550 nm.

Figure 8

Micrograph of 3-μm Epon section of the spinal cord from a control animal not bismuth-treated. The section was autometallographically developed for 60 min and counterstained with toluidine blue. The section was void of AMG grains. Bar = 27 μm.

Figure 9

Epon section (3 μm) from the same rat spinal cord as shown in Figure 3. The section was treated with 1% potassium cyanide for 10 min before AMG, then developed for 60 min and counterstained with toluidine blue. Note that the section is completely devoid of AMG grains. Bar = 27 μm.

The protocol given below represents an optimized AMG bismuth technique. Autopsy or biopsy specimens from animals that have been exposed to bismuth, e.g., by using bismuth-containing drugs or having bismuth shotgun pellets implanted, can either be frozen and cut on a cryostat or fixed in buffered formaldehyde- or glutaraldehyde-containing fixatives. Some animals were exposed concurrently to selenium and bismuth. Optimally fixed tissue blocks can be obtained by transcardial perfusion with glutaraldehyde and/or formaldehyde in a 0.1 M phosphate buffer for approximately 10 min, followed by immersion in the fixative for at least 2 hr. Seemingly, the fixatives have no damaging effects on bismuth–sulfide–selenide clusters. Tissue can be stored in the fixatives for at least 1 year, and most likely for many more years, as is the case with tissue that holds mercury–sulfide–selenide clusters (Danscher and Møller-Madsen 1985).

The fixed tissue blocks can either be frozen and cut on a cryostat or embedded in paraffin, methacrylate, or Epon. For ultrastructural studies it is advisable to cut 100-μm sections on a vibratome (see below).

The cryostat, paraffin-, or Epon-embedded sections are placed on Farmer-rinsed glass slides. In cases where the sections are cut from fixed tissue blocks, the glass slides should always be dipped in a 0.5% gelatin solution and allowed to dry before use.

After being placed on the glass slides and allowed to settle, all sections should be covered with a thin film of gelatin by being dipped in the above gelatin solution. The sections are now ready for AMG development.

AMG Development Protocols

The Silver Lactate AMG Developer (Danscher et al. 1997a)

Preparation of AMG Solutions

1. I. Protective colloid: Dissolve 1 kg of non-refined acacia resin (crude gum arabic, also used by the liquorice industry) in 2 liters deionized water by intermittent stirring over 5 days at room temperature (20C). Filter the solution through layers of gauze and freeze suitable portions of the resulting colloid in plastic jars. Such jars can be stored for at least 1 year.

2. II. Citrate buffer (pH 3.7): Dissolve 25.5 g citric acid monohydrate and 23.5 g sodium citrate dihydrate in deionized or distilled water to make 100 ml.

3. III. Lactate buffer (pH 3.8): Dissolve 31.5 ml of 50% sodium lactate and 6 ml of 90% lactic acid in deionized or distilled water to make 100 ml.

4. IV. Hydroquinone: Dissolve 0.85 g hydroquinone in 15 ml deionized or distilled water. Prepare just before use.

5. V. Silver ion supply: Dissolve 0.11 mg silver lactate in 15 ml deionized water. The solution should be protected from light by wrapping tinfoil around the vial.

6. VI. AMG stop bath: A 5% sodium thiosulfate solution.

7. VII. Farmer solutions: 10% (9 parts 10% sodium thiosulfate and 1 part 10% potassium ferricyanide) for cleaning of glassware; 1% (9 parts 1% sodium thiosulfate and 1 part 1% potassium ferricyanide) for cleaning of section surface.

Finally, mix Solution I (60 ml), Solution II or III (10 ml), and Solution IV (15 ml) carefully in a 100-ml Farmer-cleaned beaker. Add Solution V immediately before use.

AMG Silver Enhancement of Tissue Sections

Place the slides in a jar filled with 26C distilled water and place the jars in a water bath set at 26C for at least 10 min. Heat the mixed AMG developer to the same temperature and exchange water in the jars with developer. Now cover the whole set-up with a dark hood. During development, it is advisable to shake the jar gently, e.g., with an electric device, to make the presence of silver ions and hydroquinone in the developer more even.

The development can, of course, also take place on the glass slides by covering the sections with the developer, or the sections can be developed floating in the developer. In that case, the dish temperature should likewise be kept at 26C. In cases where short development periods are optimal, one can observe the sections in the microscope while developing (Hacker et al. 1988).

EM Re-embedding Techniques

Semithin sections are placed on a glass slide and AMG developed. After thorough rinsing of the section surface, place a drop of unpolymerized resin on the semithin section to be studied in the electron microscope and place a blank resin block on top. After 24 hr at 60C, remove the block from the glass slide by placing the preparation on a 90C hotplate for about 30 sec. After trimming, cut ultrathin sections and place on a grid.

Ultrathin sections treated according to the procedures given above can be stained conventionally with lead citrate and uranyl acetate.

EM Vibratome Technique

Sections of 100 μm are cut from tissue blocks on a vibratome. The sections are AMG-developed and regions to be analyzed in the EM are dissected out and treated with osmium and uranyl acetate before being embedded in Epon. Ultrathin sections can be cut directly from this preparation or, if it seems appropriate, semithin sections can be cut, placed on glass slides, analyzed first by LM, and then reembedded on top of a blank Epon block for ultrathin cutting.

Post-AMG Procedures

Depending on the thickness and penetrability of the sections, development is stopped after 50–120 min by replacing the developer with a 5% sodium thiosulfate solution, the AMG stop bath solution. After 10 min in the stop bath, the jars are placed under gently running 40C tapwater for 20 min, whereby the gelatin film is removed from the sections together with AMG grains that have grown in the developer itself, i.e., autocatalytic silver crystals, and finally the jars are filled twice with distilled water.

Additional Cleaning of Section Surface

Every second slide is dipped for 10 sec in a 1% Farmer solution. The sections are rinsed twice in distilled water, counterstained, and eventually embedded in DEPEX mounting medium and covered with a coverglass.

Controls of Specificity

Sections from different organs of animals not treated with bismuth served as blank controls of specificity. Multi-element analysis (PIXE) of AMG grains from developed sections served to prove the hypothesis that these silver grains wrap bismuth bound chemically to sulfur and, if the animals had been exposed concomitantly to bismuth and selenium, to selenium as well. Bismuth sulfide-containing sections that were treated with 1% potassium cyanide for 10 min, rinsed in distilled water, and AMG-developed were completely void of staining.

PIXE Analysis

Kidney sections from bismuth-exposed animals were cut on the cryostat and developed floating in the AMG developer. After 60 min they were removed, rinsed in distilled water, and placed on an “aerosol quality” Nucleopore filter, i.e., a high-purity ∼1 mg/cm²-thick polycarbonate membrane. Some of the developed sections were placed in a jar and subjected to enzymes to remove the organic part of the sections. The sections were treated as follows: (a) the tissue was homogenized in Tris buffer with six to eight strokes in a Potter–Elvehjelm homogenizer; (b) the homogenate was spun down at 8000 rpm for 15 min and then incubated for 2 hr with 4% Triton X-100; (c) after repeated washings and resuspensions by centrifugation, the black-stained pellet was exposed for 12 hr at 37C to a solution of 5% papain (Merck Chemicals; Darmstadt, Germany) in Tris buffer containing 0.1 M CaCl₂ at pH 7.6; and (d) after three or four washings the black pellet was treated with 0.5% collagenase (Sigma, St Louis, MO; Type I) in the same buffer at 37C for 2 hr. Finally the pellet, now consisting of small black aggregates, was washed four or five times in double-distilled, demineralized water. The samples were then placed on Nucleopore filters and PIXE analyzed. Because of this treatment, only major AMG grains were isolated. Each sample was bombarded with both 2 and 3 MeV protons to obtain the best detection limits for a wide range of elements (Kemp and Danscher 1979). The PIXE analysis yielded only relative concentrations of the elements because the exact weight of the dry tissue was unknown. Since the analysis was conducted to determine the atomic composition of AMG silver grains only and the ratio between silver and bismuth–selenium–sulfur, we did not try to establish the total weight of the samples.

Results

The above-delineated technique results in a detailed pattern of bismuth-specific AMG grains that can be observed at LM and EM levels in sections from different organs of bismuth-exposed experimental animals, e.g., brain, spinal cord, and kidney (Figures 2–7). Comparable sections from animals that were not exposed to bismuth were blank (Figure 8), as were bismuth-containing sections that had been treated with potassium cyanide (Figure 9).

Figure 10

PIXE spectra showing the difference between a tissue sample and the separated grains. Each element is represented by two to three peaks in the spectra, as indicated on the figure. Using the well-known ratios between the peaks from a single element solves the problem with interference between single peaks from different elements. It can be deduced from the spectra that Se and Bi are present in the AMG grains but that the ratio to the Ag content is lower in the grains than in the tissue.

PIXE analysis of AMG-developed sections from kidneys of bismuth-exposed rats revealed the presence of silver, bismuth, and sulfur atoms, while sections from animals that had been treated simultaneously with selenium contained selenium as well. Isolated AMG grains from the two sources contained Ag, Bi, S and Ag, Bi, S, Se, respectively. The yields in the PIXE spectra for Bi and Se relative to Ag were, however, much lower in spectra from the separated grains than from the developed sections (Figure 10). The reason is probably that a size fractionation of the AMG grains takes place during the centrifugation. It is mainly the coarse grains (around 0.5 μm) that are isolated. These grains have a relatively high Ag content (cf. Figure 1). Further Bi and Se are embedded in the center of the grains, which leads to attenuation of the X-ray signals from Bi and Se. The PIXE data are in accordance with these hypotheses. No conclusions can be drawn with respect to sulfur because the sulfur content associated with Bi in all cases is much lower than sulfur of other origin.

Discussion

As mentioned in the introduction, AMG–silver enhancement of accumulations of bismuth in mice exposed to bismuth subnitrate was introduced by Ross et al. in 1994. Their controls were blank sections from untreated animals, and they did not try to further evaluate the specificity of the technique or the chemical nature of the catalytic bismuth accumulations. In the present study we describe protocols on techniques available for performance of optimal AMG studies at LM and EM levels, and data suggesting that the catalytic bismuth compounds are bismuth–sulfide–selenide clusters. Based on the multi-element analysis (PIXE) of sections and AMG grain samples, it can be deduced with a reasonable degree of certainty that the AMG catalysts in question are small clusters, i.e., crystal lattices, of Bi₂Se₃ and Bi₂S₃. Because the AMG grain isolation technique results in a fraction of grains approximately 0.5 μm in size, the amounts of Bi, S, and Se relative to silver were low. In contrast, PIXE analysis of AMG-developed tissue sections containing all sizes of AMG grains revealed high relative levels of the three elements. The possibility of the existence of a non-sulfide–selenide-bound bismuth pool in the sections cannot be excluded from the present study but will be further analyzed.

Because the AMG technique is a tool for revealing several endogenous and exogenous metals in tissue sections, providing that they are present either as sulfide–selenide clusters (Ag, Hg, and Bi) or as pure metal clusters (Au and Ag), it is of some importance to be able to differentiate among the different metals if they are present in the same tissue. Protocols to this end have previously been worked out for Au, Zn, Ag, and Hg (Danscher et al. 1997a,b).

The only metal sulfide–selenide cluster that is known to be resistant to cyanide is mercury (Danscher et al. 1997a). We have found that bismuth sulfide clusters can be removed by 1% KCN for 10 min.

Despite these chemical possibilities for removing the different kinds of catalytic clusters from Epon sections, it should be stressed that use of a multi-element analysis is crucial for a precise and secure knowledge about the nature of the AMG metals present in a section for which the composition is not known in advance (Danscher et al. 1997a).

To summarize, the above protocols are based on the most recent AMG technology and describe the optimal procedures for LM and EM analysis of tissues containing bismuth sulfide clusters. It is shown that AMG grains in tissue sections from animals that were simultaneously exposed to bismuth and selenium cause creation of bismuth selenide clusters or bismuth–sulfide–selenide clusters. Ways to control specificity and to differentiate bismuth from other AMG metals in tissue sections are presented. The AMG bismuth technique should be of value to researchers in the field.