Elimination of the antimicrobial action of the organoarsenical cancer therapeutic, 4-(N-(S-glutathionylacetyl)amino) phenylarsonous acid, before finished product sterility testing

Abstract

Objectives

Arsenical compounds have been used therapeutically for over 2000 years finding particular relevance as antimicrobials. After being replaced by more selective and consequently less toxic antibiotics in the last century, arsenicals have recently made a resurgence as anticancer drugs (specifically arsenic trioxide and its derivatives). Arsenical parenteral formulations require post-manufacture sterility testing; however, their intrinsic antimicrobial activity must be neutralised before testing to eliminate the possibility of false (no-growth) test results.

Methods

A range of thiol-containing compounds was screened to establish a suitable deactivation agent for the novel organoarsenical compound, 4-(N-(S-glutathionylacetyl)amino) phenylarsonous acid (GSAO). Dimercatopropanol (DMP) was found to successful deactivate GSAO and was validated according to pharmacopoeial sterility test guidelines (specifically the method suitability test/sterility validation test).

Key findings

DMP is an effective way of deactivating GSAO before sterility testing and can be used for pharmacopoeial sterility tests. Our results affirm previous research highlighting the sensitivity of Staphylococcus aureus to arsenical compounds

Conclusions

A method of deactivating the arsenical drug GSAO before the post-manufacture sterility test was established and validated. DMP is a commonly used chelator/deactivation agent so this work may have implications for other inorganic therapeutic agents.

Introduction

Arsenicals have been employed as therapeutics for over 2000 years. Nobel laureate Paul Ehrlich, originator of the ‘magic bullet’ hypothesis, developed arsphenamine in 1909, and it remained the ‘gold standard’ treatment for syphilis for almost 40 years – being considered as the world first ‘blockbuster’.^[1] With advances in medicinal chemistry and pharmacology during the mid-20th century, drug candidates with enhanced efficacy and lower toxicity were developed, and unsurprisingly the therapeutic use of arsenical compounds waned. However, a renaissance for arsenic-based drugs came in the 1980s, culminating in the approval of As₂O₃ (Trisenox, Cephalon/Teva, North Wales, PA, USA) for the treatment of acute promyelocytic leukaemia by the US Food and Drug Administration in 2000.^[2]

The organoarsenical compound 4-(N-(S-glutathionylacetyl)amino) phenylarsonous acid (GSAO) is a novel anticancer arsenical drug candidate (Figure 1). As with previous arsenicals, GSAO's therapeutic response depends upon its reactivity with closely spaced cysteine thiols in select proteins.^[3,4] GSAO is a pro-drug that is activated by γ-glutamyl transpeptidase at the cell surface to produce GCAO.^[5] GCAO enters the cell via an organic ion transporter and is further processed by dipeptidases to 4-(N-(S-cysteinylacetyl)amino) phenylarsonous acid (CAO) in the cytosol. CAO enters the mitochondrial matrix where the arsenical moiety cross-links Cys57 and Cys257 on the matrix face of adenine nucleotide translocase (ANT),^[6] which inactivates the transporter leading to partial uncoupling of oxidative phosphorylation, increase in superoxide production and arrest of proliferation of the cell.^[3] CAO reacts with ANT only when cells are proliferating.

Figure 1.

Structure of 4-(N-(S-glutathionylacetyl)amino) phenylarsonous acid (GSAO).

GSAO has recently been administered clinically as part of an intravenous phase I clinical trial sponsored by Cancer Research UK. The investigational medicinal product was lyophilised and manufactured aseptically and therefore was required to comply with the pharmacopoeial test for sterility.^[7] The sterility test was performed using a filtration technique that washes the drug solution through a microbe-retaining (0.45 μm) membrane to minimise any effects the active compound may have on inhibition of microbiological growth. Potent antimicrobial agents may require deactivation before this procedure as at low residual concentrations they may still affect the ability of the medium to support cell growth. Akers et al.^[8] specifies deactivation agents for antibiotics, phenols and sulphonamides. A non-pharmacopoeial protocol for the deactivation and sterility testing of neoarsephenamine and sulpharsephenamine was reported nearly 60 years ago.^[9] However, this protocol required custom-prepared media and is inadequate by modern standards as it was shown to retain some inhibition of Staphylococcus aureus growth and preceded the availability of filtration-based sterility testing methods. Despite this, it remains that the only literature reference that reports agents, or procedures, for the sterility deactivation of drugs based upon arsenic. (Additionally, there are no equivalent reports for other drugs based upon metal or metalloid elements.) This study describes procedures for deactivating an arsenical-based drug that are fully compliant with modern requirements.

Our initial studies showed that even after filtration, residual GSAO inhibited the growth of inoculated bacteria (specifically S. aureus), thereby failing to comply with the validation procedures of the sterility test. We did, however, observe occasional S. aureus growth, as well as suitable growth with Candida albicans and Clostridium sporogenes. The search for a suitable GSAO deactivation agent centred upon using sterilisable thiol compounds that would be compatible with the growth of inoculated bacteria. We investigated simple thiol-containing compounds and media, as well as reactive dithiols suggested by the literature.^[10]

Materials and Methods

GSAO was produced by Dr Reddy's Laboratories (Hyderabad, India) to a modified version of the published synthesis^[3] and thereafter prepared as a lyophilised pharmaceutical, comprising a 2-ml fill in 10-ml vials of 50 mg/ml GSAO in 0.1 m glycine buffer pH 7 according to the procedures described in Elliott et al.^[11] Cysteine, sodium thioglycollate, bovine foetal calf serum, 2-mercaptoethanol and 2,3 dimercatopropanol (DMP) were all purchased from Sigma Aldrich (Poole, UK). Nonsterile chemicals were filtered aseptically through a 0.2-μm filter before use. Staphylococcus aureus (ATCC 6538), C. albicans (ATCC 10231), C. sporogenes (ATCC19404), Aspergillus niger (ATCC 16404), Bacillus subtillis (ATCC 6633) and Pseudomonas aeruginoisa (ATCC 9027) were prepared from certified Quanti-cults (Oxoid, Basingstoke, UK) to 1000 colony-forming units (CFU)/ml. Sterility filtration test units (Steritest EZ Device) were purchased from Millipore (Watford, UK) and consisted of two canisters capable of being filled simultaneously (as would be done with test solutions) and independently (for different growth media). Washing buffer A was purchased from bioMerieux (Basingstoke, UK), and fluid thioglycollate medium (FTM) and trypic soya broth (TSB) were purchased from Cherwell (Bicester, UK). All broths were certified as fertile by the supplier and positive fertility control experiments were performed in house. (TSB is equivalent to the soya-bean casein digest specified in the pharmacopoeial procedure.) Water for irrigation (WFI) was purchased from Baxter's Healthcare Ltd. (Norfolk, UK).

In the experiments described, TSB was incubated at 20–25°C and FTM at 30–35°C for 3 days in the case of bacterial inoculation and 5 days in the case of fungal inoculation. Unless otherwise specified, all results reported are based upon single incubations.

Where appropriate, statistical runs tests^[12] were performed to establish the probability of a growth/no-growth result series being generated at random.

Assessment experiments with fetal calf serum and sodium thioglycollate

Experiments to assess the ability of fetal calf serum and sodium thioglycollate to deactivate GSAO were conducted with Steritest units, according to the pharmacopoeial guidelines^[7] using 10 vials of GSAO with WFI as the reconstitution solution. After filtration, washing with 3 × 200 ml of washing buffer A, adding broth and inoculum, both containers were then incubated at the prescribed temperatures and checked for visible signs of microbial growth.

In the first experiment, TSB containing 5% v/v fetal calf serum was added to both filtration containers. One container was inoculated with 100 CFU of S. aureus, and the second container with 100 CFU of C. albicans.

In the second experiment, the washing buffer A solution contained 0.05% v/v sodium thioglycollate. FTM was added to one container and inoculated with 100 CFU of C. sporogenes, and TSB was added to the second container and inoculated with 100 CFU of S. aureus.

GSAO inhibitory titration experiments

To determine the inhibitory concentration of GSAO in TSB and FTM, aliquots of 10 ml of TSB in sterile Sterilin containers (Sterilin Ltd., Newport, UK) were aseptically prepared with GSAO concentrations of 4.6, 2.3, 1.1, 0.6 and 0.3 mm and inoculated with 100 CFU of S. aureus before incubated at the prescribed temperature and checked for visible signs of microbial growth. Equivalent experiments were performed with FTM inoculated in the same manner with S. aureus and P. aeruginoisa.

Cysteine assessment experiment

An experiment to assess the ability of cysteine to deactivate GSAO at various concentration was conducted by repeating the GSAO inhibitory titration experiment above with 0.1 mg/ml cysteine in all TSB solutions and inoculated with 100 CFU of S. aureus before incubated at the prescribed temperature and checked for visible signs of microbial growth.

Mercaptoethanol assessment experiment

An experiment to assess the ability of mercaptoethanol to deactivate GSAO was conducted by aseptically preparing, in sterile Sterilin containers, 10 ml aliquots of TSB with 14 mm mercatopethanol together with 1.1 mm GSAO concentration. These were inoculated with 100 CFU S. aureus before being incubated at the prescribed temperatures and checked for visible signs of microbial growth.

DMP assessment/titration experiments

Experiments to assess the ability of DMP to deactivate GSAO, the concentration at which this may occur and to determine if DMP would inhibit microbial growth were performed as follows. Aliquots of 10 ml of TSB in sterile Sterilin containers were aseptically prepared with a GSAO concentration of 1.1 mm and spiked with DMP to an equivalent of 10, 20, 30 40, 50, 100 and 200 μm. Samples were inoculated with 100 CFU of S. aureus. The procedure above was repeated using FTM, with C. sporogenes as the inoculum. Controls were tested with 100 μm DMP. All containers were incubated at the prescribed temperatures and checked for visible signs of microbial growth.

Assessment of DMP-mediated deactivation with Steritest units

To assess the suitability of DMP to deactivate GSAO when the latter is being sterility tested with Steritest membrane filtration units, DMP was injected into GSAO vials before reconstitution as follows. One hundred microlitres of 10 mm DMP solution (a five-molar excess with respect to GSAO) was aseptically injected into the GSAO vials immediately before reconstitution with washing buffer A. After filtration and washing with two 200-ml volumes of washing buffer A, the Steritest containers are filled with TSB or FTM and inoculated with S. aureus or C. sporogenes, respectively, and incubated at the prescribed temperatures. The S. aureus inoculation was performed in TSB because the inhibitory titration experiments had shown the organism to be more sensitive to GSAO in that medium.

Method suitability test/sterility validation testing

Method suitability (or validation) testing of the inactivation procedure was performed according to the instructions specified in the European Pharmacopoeia.^[7] Specifically, 100 μl of a 10-mm DMP solution was aseptically injected into ten GSAO vials immediately before reconstitution with washing buffer A. After filtration and washing with two 200-ml volumes of washing buffer A, the Steritest containers were filled with TSB or FTM. Enough containers were prepared to determine the growth of each validation organism separately. TSB media were inoculated with C. albicans, B. subtilis, P. aeruginoisa and A. niger before being incubated at the prescribed temperature. FTM were inoculated with S. aureus and C. sporogenes before being incubated at the prescribed temperature.

Long-term DMP-mediated deactivation

The method suitability test/sterility validation testing procedure indicated above was repeated without immediate (post-filtration) microbial inoculations. After 14 days of incubation (the duration required by the pharmacopoeial test) at the prescribed temperature, both TSB and FTM Steritest containers media were found to show no growth and were subsequently inoculated with S. aureus and C. sporogenes, respectively, before being incubated at the prescribed temperature.

Results

For both the assessment experiments using fetal calf serum and sodium thioglycollate, the growth of S. aureus remained inhibited. C. albicans grew normally in TSB with 5% fetal calf serum, as did C. sporogenes in FTM after washing with 0.05% sodium thioglycollate.

The results of the GSAO inhibitory titration experiments are shown in Table 1 and demonstrate that the inhibition of S. aureus growth by GSAO is concentration dependent, with growth occurring when the GSAO concentration is at or below 0.6 mm. The experiments with S. aureus and P. aeruginoisa in FTM showed no inhibition of P. aeruginoisa growth at any concentration tested and normal S. aureus growth at GSAO concentrations less than 1.1 mm. Delayed S. aureus growth, observable only on day 3, occurred at 2.3 mm, with no observable growth at 4.6 mm. Runs tests for the three S. aureus experiments, where a combination of (+) and (−) results were observed, indicated that the data series was non-random, with probabilities of <0.05 of the null hypotheses being correct (i.e. the run patterns being random) in all cases.

Table 1.

GSAO titration experiments with different media, inoculates and in the presence/absence of cysteine

GSAO (mm)	TSB/S. aureus	TSB/S. aureus + Cysteine	FTM/S. aureus	FTM/P. aeruginosa
0.3	+	+	+	+
0.6	+	+	+	+
1.1	−	−	+	+
2.3	−	−	+ (delayed)	+
4.6	−	−	−	+

FTM, fluid thioglycollate medium; GSAO, 4-(N-(S-glutathionylacetyl)amino) phenylarsonous acid; TSB, trypic soya broth. Note that control experiments (inoculum and broth) are not shown.

The difference between the inhibitory concentration of GSAO in TSB and FTM is consistent with the inactivation of GSAO by thiol groups in the latter media. From these experiments, the GSAO minimum inhibitory concentration for S. Aureus in TSB was considered to be 1.1 mm, and subsequent experiments (with mercaptoethanol and DMP) were designed accordingly. The observation that the microbiological inhibition is GSAO concentration dependent is commensurate with our developmental sterility filtration tests (highlighted in the introduction) where it was observed that even without deactivation S. aureus growth can occasionally occur.

The cysteine inhibitory assessment experiment (shown in the second column of Table 1) demonstrated that at a concentration of 0.1 mg/ml, the presence of cysteine did not alter the GSAO concentration/S. aureus growth relationship shown by the control containers (shown in the first column of Table 1).

The inhibition of S. aureus growth was observed in the mercaptoethanol assessment experiment, where at 1.1 mm of GSAO 14 mm of mercaptoethanol failed to neutralise the GSAO sufficiently to allow acceptable S. aureus growth in TSB.

The DMP titration experiments shown in Table 2 demonstrated that DMP, at all concentrations tested (10, 20, 30 40, 50, 100 and 200 μm), deactivated the antimicrobial action of GSAO (at 1.1 mm) on S. aureus in TSB. Furthermore, the DMP alone did not inhibit the growth of S. aureus or C. sporogenes in TSB and FTM, respectively, and that the DMP as prepared under the conditions described did not generate growth in either non-inoculated TSB or FTM. As expected, DMP/GSAO and DMP-alone containers showed C. sporogenes growth in FTM.

Table 2.

The effect of DMP at various concentrations with different media and inoculates

GSAO conc (mm)	DMP conc in the broth (μm)	Broth/inocculum	Results (all containers)
None	10–200	TSB/S. aureus	+
None	10–200	TSB/no inoculum	−
1.1	10–200	TSB/S. aureus	+
1.1	100	TSB/no inoculum	−
None	100	FTM/C. sporogenes	+
1.1	10–200	FTM/C. sporogenes	+
1.1	100	FTM/no inocculum	−

DMP, dimercatopropanol; FTM, fluid thioglycollate medium; GSAO, 4-(N-(S-glutathionylacetyl)amino) phenylarsonous acid; TSB, trypic soya broth. Note that 1.1 mm is considered the inhibitory concentration of GSAO, and where a DMP concentration range is given, the specific values are 10, 20, 30 40, 50, 100 and 200 μm.

The tests of the DMP-deactivation procedure with the Steritest units showed growth in all the inoculated Steritest containers, with both S. aureus and C. sporogenes after the prescribed incubation time.

The method suitability (or validation) with the DMP-deactivation procedure showed growth in all the inoculated Steritest containers, C. albicans, B. subtilis, P. aeruginoisa and A. niger in TSB, and, S. aureus and C. sporogenes in FTM. These results demonstrate that the DMP-based GSAO deactivation methodology is suitable for full pharmacopoeial sterility testing.

Inoculation of Steritest containers reported as ‘sterile’ (no growth after 14 days incubation) by S. aureus, for TSB, and C. sporogenes, for FTM, showed growth within the prescribed 3-day incubation period. This particular test was repeated several times with the same result. Note that S. aureus was used to test the TSB medium because our results demonstrated enhanced sensitivity of this organism in this medium.

Discussion

Our studies clearly show that the organoarsenical, GSAO, inhibits the growth of S. aureus and that that growth is dependent upon both GSAO concentration and media type. The addition of monothiols to TSB growth medium did not mitigate the inhibition of S. aureus growth; however, it seems likely that the presence of thioglycollate in FTM is responsible for the slight increase in the GSAO inhibitory concentration (from 1.1 mm GSAO in TSB to 2.3 mm in FTM). The dithiol, DMP, consistently demonstrated a complete deactivation of GSAO's antimicrobial effect and was used effectively in validating a pharmacopoeial method suitability/sterility test. Importantly, DMP, at the concentrations used in this study, had no antimicrobial effect of its own and could be effectively sterilised (thus not introducing microorganisms into the test containers). DMP, as a vicinal dithiol, neutralises arsenicals by a complexation reaction.^[10] It should be noted that in our studies with Steritest units, DMP was added to the reconstitution buffer and would be largely filtered away during the membrane washing steps. Furthermore, it indicates that the DMP complexes with, and deactivates, the arsenite rapidly. The inclusion of a post-14-day inoculation shows that the deactivation does not reverse over the required incubation period. This procedure was used on several batches of GSAO product manufactured, tested and released in our GMP facility, and subsequently used in the phase I clinical trial highlighted in the introduction.

FTM, with a sodium thioglycollate concentration of 0.05% (w/v), shows a slight mitigation of GSAO inhibitory activity; however, the same level of sodium thioglycollate in the washing buffer showed no effect. Sykes et al.^[9] neutralised neoarsphenamine using a far higher concentration of sodium thioglycollate (0.4% w/v), with a custom-made extract of heart muscle digest (to provide protein-based thiols). Neither of the monothiols tested, cysteine or mercaptoethanol, were effective GSAO neutralisation agents.

Fetal calf serum was used in our studies as a sterile source of bovine serum albumin: the latter being considered as a general proteineous scavenger. However, we found fetal calf serum at the 5% (v/v) concentration used to be ineffective at deactivating GSAO. Previous researchers have indicated that human serum albumin reduces the cytotoxicity of arsenical–glutathione conjugates.^[13]

DMP is well known as an arsenical chelator and has, historically, been used in clinical settings: during World War II, it became known as ‘British anti-Lewisite’ and was used as a treatment after exposure to the chemical weapon 2-chloroethenylarsonous dichloride. In addition to this, DMP has been used as a treatment for other heavy metal poisoning,^[14] although recently it has being replaced by less toxic agents. Nonetheless, the general widespread use of DMP as an arsenical chelator/deactivator implies that it could be equally effective in deactivating the antimicrobial properties of other arsenical drugs also undergoing similar post-manufacturing sterility tests. The resurgence of therapeutic interest in organoarsenicals indicates that the sterility test deactivation step will become increasingly relevant and important in preparing materials for clinical studies. Furthermore, the use of DMP as a clinical chelation treatment for lead and mercury poisoning^[15] suggests that the method described in this paper, or a similar sterility test deactivation method based upon DMP, may work for other drug compounds based upon, or formulated with, toxic metals.

The results presented in this paper indicate that GSAO specifically inhibits the growth of S. aureus. A similar selectivity towards S. aureus has been previously reported for arsenicals.^[16] While resistance to arsenic has also been reported in S. aureus,^[17] it was believed that this was mediated through the phosphate transport system and so inferred upon arsenate (V) species, not arsenite (III) species as in GSAO. The possibility that arsenic-based compounds specifically inhibit S. aureus growth opens up possibilities for new therapeutic opportunities owing to the resistance of these bacteria to conventional antibiotics.

Conclusion

GSAO is a novel, putative anticancer treatment, currently in phase I clinical trial. When formulated for intravenous administration, it requires a post-manufacture ‘sterility test’. However, this test is compromised by the antibacterial properties of GSAO, specifically towards S. aureus. GSAO is not active against C. sporogenes, P, aeruginoisa or C. albicans. Our studies have shown that DMP is a useful antimicrobial neutralisation agent for GSAO, and using the procedure highlighted above, it did not inhibit the growth of any required pharmacopoeial organisms. The use of DMP as a deactivation agent was validated to the full requirements of the European Pharmacopoeial sterility test (method suitability test). The previously published methodology^[9] required custom-prepared media and was reported to be only partially effective. The resurgent use of arsenic and its derivatives as anticancer therapeutic agents indicates that sterility test procedures and their associated deactivation neutralisation steps, such as the one outlined in this paper, will become more prevalent. DMP is known to chelate to other toxic heavy metals, and so the neutralisation principle reported here for arsenicals may apply to other species. Our results concur with the small body of evidence that arsenicals have a selective activity to S. aureus.

Declaration