Radioisotopes of gold (Au) as shown in Table 1 have attractive nuclear properties that make them ideal for imaging and therapy. Two radioisotopes of gold stand out; one of which is ¹⁹⁸Au, which has a moderate-energy beta particle (0.96 keV β⁻ max), making it a good candidate for therapy, and more than 90% emission of a 412-keV gamma emission that allows for in vivo tracking and dosimetry calculation. These properties led to its initial use as a brachytherapy agent for prostate cancer. A second radioisotope, ¹⁹⁹Au, has a low-energy beta maximum (0.46 MeV) and emits a 158-keV gamma with 36% yield that is easily detected by SPECT cameras. Additionally, ¹⁹⁹Au can be produced carrier free by neutron irradiation of ¹⁹⁸Pt to produce ¹⁹⁹Pt that decays in 3.2 min to ¹⁹⁹Au.

Gold nanoparticles (AuNPs) have been around for quite some time, but their harsh production methods with toxic chemicals have prevented them from being attached easily to biomolecules for evaluation in vivo. A majority of metal nanoparticles are made with toxic chemicals that are not conducive for in vivo applications. Early on we evaluated safe, green methods to produce gold nanoparticles in aqueous solutions that would be amenable to use in biomolecular applications.

Our first synthesis consisted of using THPAL, a trimeric phosphino alanine, to reduce gold salts in water containing green organics, which would coat the surface of the gold nanoparticles and form 12–15 nm-sized gold nanoparticles with hydrodynamic diameter of 60–85 nm (Kannan et al. 2006). These methods have been shown to be robust and nontoxic in the nonradioactive form. The methods have been altered to allow for formation of both ¹⁹⁸Au and ¹⁹⁹Au nanoparticles and also have been shown amenable to formation with other coin metals such as palladium and silver. We have developed a library of such conjugates, three of which have been evaluated in induced tumor models and two are being translated into large animals for imaging and treatment of prostate cancer.

Gum arabic-coated gold nanoparticles (GA-¹⁹⁸AuNPs) were the first and most studied to date (Chanda et al. 2010). The reaction consists of heating water containing gum arabic, adding gold either as the NaAuCl₄ salt or the H¹⁹⁸AuCl₄ acid along with the THPAL, heating this solution briefly, which converts from a pale-yellow solution to a red-burgundy one. Quality control has shown that this method results in 99% conversion of the radioactive gold to the nanoparticle form, which then can be brought to neutral pH and remains stable in the radioactive form for over 2 weeks. This formulation was initially evaluated for therapeutic efficacy in SCID mice bearing induced human prostate tumors. Different modes of injection were tried, with intratumoral injection resulting in the highest uptake and retention. These studies prompted us to evaluate their efficacy in tumor stabilization in the same tumor models. Tumors received a 70 Gy dose by administering 405 μCi. An overall 82% reduction in tumor volume was noted between the mice receiving GA ¹⁹⁸AuNP compared with controls receiving saline injections. Biodistributions at the end showed a reduction of gold nanoparticles in the tumor from 70% at 24 h to 19.9 ± 4.2% ID at 30 days post injection. Clearance was predominantly through the urinary track, and minimum uptake was observed in other organs as can be seen in Table 2. Based on these results, these nanoparticles are being evaluated for their therapeutic efficacy in canines with spontaneous prostate cancer. Prostate cancer in dogs has been shown to mimic that observed in humans on the functional level as well as in metastatic rate and occurrence. Studies to date have shown that dogs treated with the nanoconstructs tolerate the treatment well and that the disease is stabilized.

Another formulation was a receptor targeted approach which was developed using the Food and Drug Administration (FDA)-approved photochemical epigallocatechin-gallate (EGCg) (Shukla et al. 2011). EGCg is a major phytochemical component of green tea and has been used extensively as a food supplement because of its strong antioxidant properties. This method proved much simpler for two reasons: the nanoparticles could be formed at room temperature, and EGCg not only achieved reduction of the gold but also served as a stabilizing agent, resulting in an EGCg-conjugated gold nanoparticle (EGCg-¹⁹⁸AuNPs) formulation in a few minutes at room temperature in water. Additionally EGCg targets the LAMININ receptor (Lam 67R) which is overexpressed on human prostate cancer cells (Tachibana et al. 2004). Lam 67R is a 67-kDa cell surface protein, first isolated as a non-integrin high-affinity laminin binding protein from murine cancer cells in 1983 by two independent research groups (Rao et al. 1983; Malinoff and Wicha 1983). EGCg has been known to bind to Lam 67R with excellent specificity and selectivity (Tachibana et al. 2004). The in vivo biodistribution of EGCg-¹⁹⁸AuNPs was evaluated in SCID mice bearing human prostate cancer tumors and was shown to have an initial higher retention and to clear with time. Based on these results, they were evaluated in therapy trials with SCID mice bearing human prostate cancer. Due to the higher initial uptake, only a third of the dose activity (136 versus 405 μCi for the gold nanoparticles formed with gum arabic) needed to be injected to give a 70 Gy dose. The EGCg-¹⁹⁸AuNPs were shown to be more stable over time, remaining in solution far longer than their gum arabic analogs, and results showed similar reduction of tumor volumes (80% compared with 82%) with higher clearance from normal tissues except for the liver and the tumor, as can be seen in Table 2. From the data it is clear that selective targeting of the EGCg to the Lam 67R receptor results in an overall higher uptake and retention in the prostate tumor cells than GA-¹⁹⁸AuNP-treated ones. Due to the similar results with a smaller dose, these nanoparticles will also be evaluated in canine patients with spontaneous prostate cancer.

In addition to glycoprotein- and phytochemical-conjugated gold nanoparticles, we have also evaluated selective targeting by conjugating the 14-amino-acid peptide bombesin to the gold nanoparticle surface (McLaughlin et al. 2011a; Chanda et al. 2010). Bombesin (BBN) targets the gastrin-releasing peptide (GRP) receptors that are upregulated in a variety of cancers, predominantly breast, prostate, pancreatic, and lung cancers (Smith et al. 2003). The hypothesis was that attaching BBN to gold nanoparticles would result in selective uptake in prostate cancer, particularly metastases. Different formulations of gold nanoparticles conjugated to BBN (AuNP-BBN) with varying ratios of BBN to gold nanoparticles of 1:1, 2:1, and 3:1 were prepared and evaluated for their target specificity. AuNP-BBN conjugates evaluated through receptor binding assays demonstrated that the 1:3 AuNP-BBN conjugate showed the highest affinity and uptake (Chanda et al. 2010). This conjugate (BBN¹⁹⁸AuNP) was evaluated in SCID mice bearing human prostate PC-3 cells and in a spontaneous model of prostate cancer, the TRAMP mouse. Although the initial studies in PC-3 hinted at selective uptake, more thorough evaluation in which the uptake was compared with ⁶⁴Cu-NOTA-8-Aoc-BBN(7-14)NH₂ showed only nonselective uptake. Starch gold nanoparticles (S¹⁹⁸AuNP) were used as starting material to conjugate the BBN peptide. S¹⁹⁸Au-NPs were conjugated to BBN by use of BBN 7-14 conjugated to thioctic acid. ⁶⁴Cu-NOTA-8-Aoc-BBN(7-14)NH₂ was prepared as previously described (Prasanphanich et al. 2007). We used SCID mice bearing a flank model of human prostate cancer derived from a subcutaneous implant of PC-3 cells for pharmacokinetic studies. ⁶⁴Cu-NOTA-8-Aoc-BBN(7-14)NH₂ injected intravenously (i.v.) exhibited high uptake in the pancreas and tumor, similar to what has been previously reported; this uptake was reduced by pretreatment with cold BBN. The intraperitoneal (i.p.) route of administration of the agent showed similar accumulation in tissues except the tumor, which was lower and was not blocked by the cold BBN. However, other receptor-rich tissues such as the pancreas showed reduced uptake (p ≤ 0.05). GA¹⁹⁸Au-NP when injected i.p. showed no tumor uptake and no differences when injected with the blocking dose of BBN. Starch ¹⁹⁸Au-NP showed no tumor uptake and no reduction in receptor-rich tissues when injected with cold BBN. The BBN-conjugated starch ¹⁹⁸Au-NP showed similar results with no uptake observed in the tumor and no significant difference in uptake in the receptor-rich pancreas.