Inflammation in the central nervous system (CNS) has been found to resemble the inflammatory processes taking place in other tissues, and the CNS is therefore no longer considered an immunologically privileged organ (Barker and Widner 2004). Although inflammation is a natural response to damage in the CNS the associated processes unavoidably result in often permanent morphological and physiological changes and are not seldom succeeded by functional disturbances (Whitney et al. 2009). Microglia are considered the resident immune cells of the CNS and the immune response is believed to be ignited by a combined action of microglia, microglia derivated macrophages and invading monocytes/macrophages. The activated microglia, i.e. microglia derivated macrophages, secrete a variety of proinflammatory and neurotoxic factors that apart from igniting the inflammatory cascade are believed to induce and/or exacerbate neurodegeneration (Stoll et al. 1998; Liu and Hong 2003; Mhatre et al. 2004; Chew et al. 2006). Products of the reactive microglia derivated macrophages mediate both the activation of astrocytes, recruiting of lymphocytes and granulocytes and the neuronal injury (Giulian et al. 1993).

It is well known that neuroinflammation regardless of cause, i.e. stroke, traumatic brain injury or neurodegenerative maladies like Alzheimer’s disease and multiple sclerosis (MS), includes loss of neural tissue (for details see Back and Schüler 2004; Pantano et al. 2006; Aktas et al. 2007). In MS, microglia are activated by auto-reactive T cells that peak during the acute disease causing a collapse of the blood–brain barrier and introducing an invasion of all the members of a full-scale immune response (Rasmussen et al. 2007).

Metallic gold plates were used to reconstruct skull defects over 3,000 years ago (Habal 1979), and in the beginning of the nineteenth century gold salts were claimed to be effective in the treatment of depression, epilepsy, migraine, alcoholism, and even impotence. Gold salts got a reputation of being a “nervine” pharmacological element (Richards et al. 2002). However, gold salts were first introduced to modern medicine in 1890 when Robert Koch, Nobel Laureate from 1905, discovered that Mycobacterium tuberculosis could not live in the presence of gold ions in an in vitro setup. This started a not very successful use of gold compounds as a remedy for tuberculosis, but in the slipstream of its use as an antibiotic, clinical knowledge leading to recognition of the immuno-modulatory effects of gold ions accumulated. Hence, since 1927 gold salts have been used in the treatment of rheumatic arthritis, and when more recent medication fails gold thio salts are still used (see also Hashimoto et al. 1992; Yamaguchi et al. 2001).

Although the underlying mechanisms of gold ions are still not fully clarified it has been recognized that gold ions are powerful inhibitors of macrophages and polymorphonuclear leucocytes (Vernon-Roberts 1979; Fleming et al. 1996; Hostýnek 1997). In the 1960s it was demonstrated that gold inhibits the lysosomal enzymes of phagocytotic cells in the inflamed synovial tissue (Persellin and Ziff 1966; Yanni et al. 1994), and gold ions were found to inhibit antigen processing, to suppress NF-κ B-binding activity and I-κ B-kinase activation, and in turn to reduce production of proinflammatory cytokines (Yang et al. 1995; Traber et al. 1999; Yoshida et al. 1999). How gold ions induce their antinociceptive effects is, however, still not known.

The use of gold compounds in medicine has been limited because of their adverse effects. Both parenterally and perorally administered gold compounds can give rise to pronounced nephrotoxicity, and careful monitoring is therefore needed when administering the traditional gold compounds (Hashimoto et al. 1992; Tozman and Gottlieb 1987; Felson et al. 1990; Merchant 1998). Gold thio salts have only occasionally been suggested to cause neurotoxicity, most likely because the blood–brain barrier seems to exclude gold ions from the CNS. However, gold thioglucose, a little used gold compound in the twentieth century, has been associated with significant tissue loss in brain and medulla oblongata (Debons et al. 1982a, b; McGirr et al. 1984). These toxic effects of gold ions relate to the massive whole body exposure that results from treatment with gold compounds in order to obtain a therapeutic level at the inflamed joint(s). This problem can, however, be by-passed by making metallic gold particles/implants the source of gold ions, i.e. by placing a local depot of gold in the inflamed tissue. This “auromedication” is practicable because metallic gold is not, as it has been thought, an inert metal, but is slowly dissolved in the organism by a process coined dissolucytosis (Danscher 2002; Larsen et al. 2007).

All foreign bodies are met by responses from the immune system and, if it is a gold particle or a gold implant, a disintegrating chemical attack will be orchestrated by macrophages that home in on the surface of the implant. Because gold ions have immunosuppressive characteristics that include involvement of macrophages, the bio-released gold ions will affect the dissolution causing macrophages and condition a dampening down of the inflammatory process. The level of suppression depends on the amount of released gold ions that again depends on (1) the size of the available gold surfaces, (2) the number of attacking macrophages, and (3) whether the macrophages are activated or not. Therefore, a metallic implant will serve as a “depot of gold ions” from which immuno-modulatory gold ions will be released at a speed depending on the immunological state of the tissue.

Treatment with gold compounds as e.g. aurothiomalate, i.e. Myocrisin®, is an approach very different from auromedication. When Myocrisin® is applied it causes a spread of gold ions to all parts of the body apart from the CNS and an exposure of all cells, but as the injected gold ions are taken up within a very short period of time by a rather broad spectrum of cells in the whole body, the therapeutic effect on the activated macrophages of the joint(s) suffering from autoimmune inflammation is short. Therefore, the treatment has to be repeated again and again in order to maintain a therapeutic level in the inflamed joint(s). The uptake of gold ions in the cells of the proximal tubules in the nephrons is an important element in the toxicity profile of gold ions (Ogura et al. 1996).

Auromedication on the other hand, based as it is on a relatively sparse bio-liberation of gold ions from a depot of metallic gold implants, releases gold ions locally on a permanent basis, i.e. establishes a local therapeutic level of gold ions. As only insignificant amounts of gold ions will find their way into lymph and blood vessels, no damage will occur to the epithelial cells of the nephrons in kidney (Danscher 2002). These qualities make auromedication a safe treatment from a toxicological point of view. The possibility of a local allergic reaction, i.e. contact allergy towards gold implants/particles made of 99.99% pure gold has to be studied in order to fence conceivable problems. Unwanted reactions of gold ions have been studied in patients that had evolved skin rashes in relation to treatment with gold thio compounds (Hashizume et al. 2008). In patients with gilded stents for stenosis of the cardiac vessels contact allergy has been observed (Svedman et al. 2009), while another study finds that gold-plated stents appear to produce fewer macroscopic and histopathologic changes in the aorta than other types of stents and concludes that gold is a useful intravascular material because it reacts only minimally with the vessel wall (Tanigawa et al. 1995). In summary it is fair to say that gold and gilded implants are intensively used in a multitude of surgery approaches with great success, including as a remedy for lagophthalmos (Choi and Driscoll 2004).

An important benefit from using the auromedication approach is that the release of gold ions seems to be dependent on the degree of inflammation adjacent to gold implants, i.e. whether the local macrophages are activated or not. While auromedicated tissues that are not inflamed will release only tiny amounts of gold ions, immunologically active tissue releases manifold more. The data pointing to such an inflammation depending bio-release of gold ions were gained from a pilot experiment involving transplantation of hearts in rats (Danscher 2010). More research to confirm these findings is underway in our laboratory. If confirmed, the consequences are that a given gold depot can control inflammation in the area of implantation for a very long time, as an activation of the local microglia and monocyte derivated macrophages will cause an increased release of gold ions that again will down regulate the inflammation making the autoimmune event short-lived because the inflammatory cascade might not unfold at all.

Auromedication in the brain of mice with cryo-lesions has ignited a hope of finding ways of suppressing inflammation in the brain (Larsen et al. 2008; Pedersen et al. 2009a, b). By inserting a local depot of metallic gold, that can be made removable, in an area of the brain prone to inflammation it might be possible, in a safe way, to secure a control of the immunological processes in that particular area as long as it is found needed.

Autometallographic (AMG) studies have revealed that gold ions originating from gold salts have their metabolic end station in the lysosomal compartment, and the excess accumulation of gold containing macromolecules causes rupture of the lysosomal membrane (Davies et al. 1971, Burkhardt et al. 1978; Danscher 1981, 1991; Brunk et al. 1997; Møller-Madsen et al. 1984). Gold ions resulting from gold salts have also been shown to be transported over the placental barrier and into the human embryo (Møller-Madsen et al. 1987) and to accumulate in the dorsal root ganglia (Schiønning et al. 1992).

Bio-released gold ions behave predictably as gold ions originating from gold compounds and they likewise pile up in the lysosomes (Danscher 2002; Larsen et al. 2007). This uptake of gold in the lysosomal compartment after exposure to metallic gold implants or gold particles has been confirmed in tissues from rats and mice (Danscher 2002; Larsen et al. 2007, 2008). In the brain, gold ions released from a metallic depot, e.g. gold particles, are taken up by both glia cells and neurons and accumulate in lysosomes (Fig. 1). If gold particles are placed in the cerebral ventricles they are attacked by small macrophages most likely originating from nearby microglia cells (Fig. 2). We have termed this foreign body reaction ‘dissolucytosis’ defined as the extracellular liberation of metal ions from the surface of metallic implants and metal particles bigger than 20 μm, i.e. particles that cannot be phagocytosed by macrophages. The process is believed to be initiated by a macrophage-induced reorganization of the bio-film that envelops any foreign body that is introduced in the organism (Larsen et al. 2007; Sennerby et al. 1993; Roach et al. 1998). The elaborated dissolution membrane makes it operational for the macrophages to control the chemical events at the gold surface, and the dissolucytosis of gold ions involves most likely the capacity of macrophages to release cyanide ions and to alter the oxygen tension and the pH in their vicinity (Amatore et al. 2008). As mentioned, vide supra, the dissolucytotic process is limited by the size of the gold surface facing the tissue, the amount of dissolucytotic macrophages, and their state of activity. The relatively slow speed of the process results in a limited liberation of gold ions securing that they are taken up almost exclusively by cells close to the implant.

Fig. 1
figure 1

Micrographs of 25 μm gold particles in cryo-lesioned neocortex containing tiny autometallographic (AMG) grains in adjacent cells. The grains result from AMG enhancement of gold nanoparticles created by accumulation of the bio-released gold ions in lysosomes. a 30-μm-thick cryo section showing two AMG enhanced gold particles in the inflammatory cerebral tissue. Scale bar 50 μm. b A neighbour section to section in a, but at a higher magnification. Note the heavy load of gold particularly in glia cells. Scale bar 20 μm. c High magnification of the cellular uptake of bio-released gold ion in a semi-thin Epon section. Scale bar 10 μm. d Electron micrograph of AMG silver enhanced gold ions located in lysosomes (arrows). Scale bar 1 μm

Full size image
Fig. 2
figure 2

Micrograph of AMG stained section of a mouse brain showing gold particles with adhering dissolucytotic macrophages in the lateral ventricle. The particles were injected 14 days before the mouse was killed. chp Choroid plexus, lv lateral ventricle, arrows macrophages loaded with bio-released gold

Full size image

In a recent study we injected 20–45 μm gold particles into the neocortex of mice before generating a cryo-injury. We found that the released gold ions reduced microgliosis and neuronal apoptosis, inflicted a transient astrogliosis, and increased the neural stem cell (NSC) response (Larsen et al. 2008). Following unilateral treatment with gold particles, NSC stimulation with increased M-CSF expression was seen. The metallic gold treatment significantly increased the expression of the growth factors VEGF, FGF, LIF and neurotrophin-4 (Pedersen et al. 2009a, 2010). Furthermore metallic gold has been found to reduce TNFα expression, oxidative DNA damage and pro-apoptotic signals after experimental brain injury, while at the same time an increase in the expression of the neuroprotective proteins MTI + II was recorded (Pedersen et al. 2009b). As the bio-liberated gold ions have been shown to possess pronounced anti-inflammatory, neuroprotective and neurostimulatory capacities in the mouse brain a guarded conclusion would be that metallic gold might have clinical potentials.

Neuroprotection and regeneratory qualities of metallic gold were seen as a reduction of apoptotic cell death at both the early and the late phase, and NSE and TUNEL staining confirmed a gold dependant circumvention of apoptosis in neurons (Larsen et al. 2008).

The amount of gold injected into the neocortex of the mice was 5.4 mg pure gold and the amount of released gold ions is so low that only the most sensitive quantitative technique can trace such small amounts. The tool that has made it possible to observe the dissolucytotic activities leading to liberation of gold ions is the histochemical technique autometallography (AMG) (Danscher 1981; Danscher and Stoltenberg 2006) (Figs. 3, 4). As gold nannocrystals/quantum dots (QD) composed of only a few gold atoms are sufficient to initiate the AMG process, the technique is unique in its field and far more sensitive than any of the available laser and diffraction techniques. The capacity of tracing gold in intact tissue makes the technique the tool for establishing the exact morphological location of gold nanoparticles at all magnifications including the highest ultrastructural levels. With the extraordinary sensitivity follows, however, that the technique is rather sensitive to catalytic impurities in the chemicals used. It is imperative, therefore, that the utensils are cleaned carefully in 10% Farmer solution (see protocols) in order to obtain optimal results.

Fig. 3
figure 3

“Camera lucida” representation of the autometallographic process. When the hydroquinone molecules (hexagonal molecules, being the reductor molecules of the AMG developer) adhere to the gold quantum dots (stained gold), electrons are released into the QD, and this event builds up energy in the valence cloud that forms the attraction forces of the nanocrystal. The higher energy level brought about by the surplus electrons increases the statistical probability that silver ions (stained blue) that have attached themselves to the nanocrystal lattices of the IQ dots catch an electron and are turned into a silver atom (stained silver in the drawing). The “new” silver atom shares valence electrons with the valence cloud of the original nanocrystals and is therefore a genuine part of the original gold quantum dot. When the AMG enhanced QD’s have reached the most optimal size for electron, respectively, light microscopic analysis the process is stopped by the developer being replaced by a 5% sodium thiosulphate solution that removes silver ions from the section. Proton-induced X-ray emission spectroscopic analysis (PIXE) of the tissue surrounding gold implants confirms that gold ions are liberated, i.e. that the AMG observations are valid (Danscher 2002; Danscher et al. 1984)

Full size image
Fig. 4
figure 4

Autometallographic enhancement of gold nanoparticles can be performed in two different ways. To the left an AMG solution is poured into jars containing glass slides with the tissue sections, or the AMG developer is dripped directly onto the individual section. To the right the slides are first dipped into an autoradiographic emulsion, allowed to dry for 10 min and then placed in jars filled with a common chemical developer. Both approaches result in the tissue sections being permeated with a solution of reducing molecules and silver ions. As the gold nanocrystals in the tissue sections catalyse the reduction of silver ions to metallic silver on their surfaces either of the techniques will result in the gold quantum dots/gold nanocrystals becoming encapsulated in metallic silver and thereby made visible at EM and LM levels, i.e. the gold nanoparticle has graphed its own position in the tissue section (modified from Danscher and Stoltenberg 2006)

Full size image

When it was published that gold nanocrystals can be enlarged by AMG (Danscher 1981; Danscher and Nørgaard 1985) the technique rapidly became industrialized since colloidal gold particles had been introduced as markers of antibodies for ultrastructural analysis (Faulk and Taylor 1971; Bendayan 1980). The latter technique, however, did not allow histopathological/immunohistochemical evaluation of survey sections at light microscopic levels, and the AMG technique therefore came as a welcome solution of this problem. The AMG enhanced gold particles are easily identified at LM levels thanks to the black/brown stain resulting from nanocrystals being encapsulated in silver.

The size of the AMG grains is determined by the original size of the gold particles, the length of development, and the temperature at which the process takes place. The AMG enhanced gold particles are easily identified at LM levels thanks to the black/brown stain of the AMG grains (Danscher and Nørgaard 1985). This quality facilitates screening of sections at low magnifications and promotes the speed by which the diagnosis can be made. Image analysis can contribute to increase the amount of information that can be extracted from the sections. The AMG emulsion technique accentuates that when autoradiography is used for tracing radioactive isotopes in tissue sections it is necessary that these are free of AMG catalytic nanocrystals.

AMG was simultaneously introduced as an enhancer of targeting colloidal gold particles in immunohistochemistry (Holgate et al. 1983) and enzyme histochemistry (Danscher and Nørgaard 1983), and soon thereafter several different AMG developers and a wide assortment of gold-tagged molecules became commercially available.

Although auromedication in some form may lay a decade or more ahead, intra-cerebral application of metallic gold as a pharmaceutical source of gold ions represents a new medical concept that bypasses the blood–brain barrier and allows a direct depot drug delivery to inflamed brain tissue. Loss of nervous tissue is primarily the result of neuroinflammation following traumatic injuries, cerebral thrombosis and haemorrhage, and diseases like Alzheimer’s and MS (Inglese et al. 2004), and the marked impact of gold ions on the inflammatory response in our model includes an increased glia fibrillary acidic protein, GFAP, activity in the early phase along with an over time stable reduction in the number of microglia/macrophages and a significant reduction of ramified macrophages at day 14 (Larsen et al. 2008).