Characterization and toxicity of a phosphate-binding exobiopolymer produced by Acinetobacter haemolyticus MG606

• Acinetobacter haemolyticus MG606 produces a 50 kDa heteropolysaccharide;

• EBP exhibits in vitro toxicity and free radical production at high concentrations;

• Mitochondrial and lysosomal damage is apparent at high EBP concentrations;

• EBP exhibits dose-dependent toxicity in mice;

• LD₅₀ of EBP following intraperitoneal administration in mice is 92.31 mg/kg.

Bioremediation of anthropogenic pollutants using natural polymers has received immense interest over the past decades. Several polymers of microbial and plant origin have been recognized as potential bioremediants for adsorptive removal of heavy metals and organic contaminants (Nguyen et al. 2012; More et al. 2014). However, in striking contrast to research on bioremediants for highly toxic compounds, bioremediation methods for inorganic anions are still in infancy (More et al. 2014).

Phosphate is one of the most important anionic pollutants which has attracted attention in recent years. Phosphate is generally present in small amounts in natural waters. However, its ever increasing agricultural usage coupled with escalating discharges from household and poultry wastes has resulted in augmentation of phosphate levels in freshwater bodies. The accumulation of phosphate in lakes and estuaries is a prerequisite for and a cause of algal blooms, culminating in eutrophication (Rockstrom et al. 2009; UNEP 2012). Few reports exist on phosphate bioremediation using microbial exobiopolymers (EBP) and plant-derived polymers (Yang et al. 2011; Riahi et al. 2014). However, the practical applicability of these biopolymers is compromised by two important constraints: first, phosphate-binding capacity of these polymers is in the range of 2–10 mg phosphate/g polymer which is too low for commercially viable applications. The low adsorption efficiency of pristine polymers could be augmented by functionalization with metals but this approach is limited by high production costs and leeching of metal ions during sorbent recovery (Nguyen et al. 2015). The second constraint is the toxicity/biocompatibility of these polymers is largely unknown. Biopolymers are generally considered safe due to their polysaccharide nature and this caveat forms the basis for a dearth of systematic toxicity studies (Ramberg et al. 2010). Nonetheless, this generalization is void since biopolymer structure and composition exhibit significant interstrain differences even in members of the same species. These structural and compositional differences transform into divergence in physicochemical properties and toxicological profiles (Ruas-Madiedo et al. 2010). A precise knowledge of toxicological profile is important to determine human and environmental safety, safe exposure levels during manufacturing, methods of disposal and countermeasures to be undertaken during accidental leakage.

Our group has earlier reported the high affinity binding of phosphate with Acinetobacter EBP with possible applications in environmental monitoring (Kaur et al. 2013a, 2013b). In a recent report, we have also reported phosphate removal in batch equilibrium studies and proposed a mechanism of phosphate binding (Kaur & Ghosh 2015). In the current study, we report further characterization of EBP and its toxicity in a macrophage cell line and mouse lethality assay.

All chemicals and media components used were purchased either from HiMedia (India) or other suppliers and were of the highest purity available.

Acinetobacter haemolyticus MG606 was grown in nutrient broth for routine subculturing. The strain was transferred to EBP production medium for extraction of EBP as described earlier (Kaur & Ghosh 2015).

Molecular weight was determined by gel permeation chromatography using Sephacryl S-300 column and eluted with 0.1 M NaCl. The eluting fractions were manually collected and retention volume calculated. The molecular weight was calculated from retention volume by comparison with retention volume of dextran standards of different molecular weights (Badrinathan et al. 2012).

Monosaccharide composition was determined by aldol-acetate method. Briefly, EBP was hydrolyzed with 2 M trifluoroacetic acid and the liberated monosaccharides were converted to aldol-acetate by sequential treatment with sodium borohydride and pyridine/acetic anhydride. The derivatized monosaccharides were analyzed by gas chromatography mass spectrometry (GCMS). Linkage analysis was performed by methylation analysis (Bales et al. 2013). The partially methylated aldol acetates were determined by GCMS as described above.

Fourier transform infrared (FTIR) spectrum was recorded by the potassium bromide pellet method (Kaur & Ghosh 2015). FTIR spectrum was deconvoluted using PeakFit software.

Mouse macrophage cell line (RAW 264.7) was maintained according to standard methods (Singh et al. 2013). Briefly, RAW 264.7 cells (National Centre for Cell Sciences, Pune, India) were maintained in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS). The flasks were incubated at 37 °C until confluent and then trypsinized. The cell suspension was seeded in T75 flasks for subculturing. All culture components were obtained from HiMedia, India.

Confluent cultures were trypsinized and cell number was counted on a haemocytometer. The cell suspension was adjusted to 5 × 10⁴ cells/mL with DMEM containing 10% FBS and 0.2 mL of cell suspension was seeded in 96 well plates. The plates were incubated overnight for cell attachment and culture medium was then replaced with culture medium containing EBP (100–500 μg/mL) or culture medium (untreated control). Following 24 hours incubation, culture medium was replaced with 0.2 mL DMEM containing MTT (0.5 mg/mL). Cell viability was calculated from the amount of formazon produced and expressed as percentage of untreated control (Singh et al. 2013).

Cellular activation was determined by monitoring stimulation of reactive oxygen species (ROS) and nitrite production by dichlorofluorescein (DCF) method and Greiss reaction, respectively. Mitochondrial and lysosomal integrity was determined by safranin O and neutral red uptake, respectively (Singh et al. 2013).

Female Swiss mice weighing 20–30 g and aged 8–12 weeks old were acclimatized to 22 ± 3 °C; 30–70% relative humidity and 12 hour light/dark cycle for 7 days before commencing experimentation. These laboratory conditions were also maintained during the course of experimentation. Food and water were provided ad libitum during the entire course of study. All animal experimentation was performed at Venus Medicine Research Centre, Baddi, India.

Precalculated amounts of EBP were separately dissolved in pyrogen-free saline to administer doses ranging from 60 to 140 mg/kg EBP. The EBP solutions of different concentrations were injected intraperitoneally to administer indicated doses. The dose range was selected on the basis of a preliminary sighting study performed with a wider dose range (data not shown). Body weight, rectal temperature, clinical signs of toxicity and mortality were observed for 7 days after administration of a single dose of EBP. Gross necropsy was performed on the same day in animals that died during study and on day 7 in all surviving animals.

The dye absorbance (MTT formazon, safranin O and neutral red) and fluorescence intensity (DCF) in untreated cells (control) was considered 100% and absorbance in EBP-treated cells was converted to percentage of control. The data for all in vitro and in vivo parameters were converted to mean ± SD. The intergroup differences were compared by one-way analysis of variance (ANOVA) followed by Tukey's test.

Size separation of polysaccharide fraction revealed a single polysaccharide component with a molecular weight of approximately 50 kDa. Monomer composition revealed the heteropolysaccharide nature of EBP with relative monosaccharide proportion of glucose:xylose:arabinose:ribose:galactose:allose:lyxose 3:2:2:2:2:2:1. Additionally, small amounts of mannose (approximately 5%) were also detected. Monosaccharide composition of A. haemolyticus MG606 EBP determined by aldol-acetate method is in agreement with previously reported composition determined by silylation method (Kaur & Ghosh 2015).

GCMS of methyl glycosides revealed the presence of 2-, 3-, 4/5- and 6-linked hexoses as well as 2,4-, 2,6-, 3,6-, 4,6-linked hexoses. Similarly, 2-, 3- and 4/5-linked pentoses were also detected along with 2,4-, 3,4- and 3,5-linked hexoses. It is worth mentioning here that pyranose and furanose forms isomerize during acid methanolysis and hence, a distinction between 4- and 5-linked sugars was not possible. Further, high proportions (>25%) of tri-linked (1,2,4-, 1,2,3-, 1,2,5-, 1,2,6-, etc.) and tetra-linked (1,2,3,4-, 1,2,4,6-, etc.) along with unmethylated monosaccharides (<5%) were detected. Based on existing literature on microbial EBPs, it is highly unlikely that tetra-linked monosaccharides could be present in such high proportions in EBP. The detection of unmethylated along with tetra-linked monosaccharides suggest the possibility of undermethylation of EBP, presumably due to low solubility in solvent (dimethylsulfoxide) (Laine et al. 2002). This precluded a reliable quantitative estimation of different linkages in EBP and hence, the results obtained are not shown. Nonetheless, despite these shortcomings, a qualitative analysis of different linkages was still pragmatic and validated the presence of (1,3)- and (1,6)-linked sugars as determined from FTIR studies.

Acinetobacter genus comprises of 27 species (Peleg et al. 2012), however, EBP produced by only three named species (A. baumanni, A. calcoaceticus and A. junii) and one unnamed species (Acinetobacter sp. 12S) have been characterized (Pirog et al. 2009; Bales et al. 2013; Sen et al. 2014). Recently, we have reported that A. haemolyticus produces EBP with chemical characteristics significantly different from other Acinetobacter EBPs (Kaur & Ghosh 2015). The most distinctive feature of A. haemolyticus MG606 was the presence of pentose sugars in very high proportions which were either not detected or were present in minuscule amounts in EBPs produced by other Acinetobacteria. The linkage pattern of hexoses in A. haemolyticus MG606 EBP is similar to that observed in other Acinetobacter sp. (Satpute et al. 2010; Bales et al. 2013).

Systemic administration (i.p.) of EBP in mice (n = 4/group) resulted in dose-dependent mortality. No mortality was observed at the lowest dose (60 mg/kg) while 25, 75, 75 and 100% mortality was observed at dose levels of 80, 100, 120 and 140 mg/kg, respectively, within 2 hours of EBP administration. Gross necropsy of dead animals revealed petechial to ecchymotic hemorrhages in heart, kidney and liver. No further mortalities were observed over a period of 7 days and gross necropsy revealed no pathological changes in animals that survived until study completion (7 days post-administration). The value of LD₅₀ was determined to be 92.31 mg/kg. The LD₅₀ of EBP was comparable to LD₅₀ of high molecular weight dextran sulphates (0.1–0.5 mg/kg), neutral mannan present in baker's yeast (75 mg/kg) and carrageenan (5,000–9,150 mg/kg) (Nagase et al. 1984; CIR 2012; Necas & Bartosikova 2013). The proximity of LD₅₀ values of EBP with polysaccharides used in pharmaceuticals (dextrans) and present in food (yeast neutral mannan) support the biocompatability of EBP. The low in vivo toxicity is also in consonance with the observed low in vitro toxicity in RAW 264.7 cells, thereby validating high biocompatibility of EBP.

Mortality and changes in organ appearance (gross necropsy) represent extreme toxicological outcomes whereby the toxicity is significant enough to manifest itself as organ damage, culminating in death. Although suitable for preliminary investigations, these endpoints have generally been criticized due to their low sensitivity and non-specificity. Another drawback of these endpoints is that they provide little or no information on the mechanism of toxicity. Therefore, additional endpoints are generally recommended which are more sensitive and provide subtle clues about possible mechanisms of toxicity. These endpoints include changes in body weight, body temperature and clinical signs of toxicity (Iezhitsa et al. 2002).

Daily body weights were recorded from the day of EBP injection until study termination (day 7). Comparative analysis of body weights between all observation days revealed no significant (p > 0.05; F value 0.149) changes in body weights when treated with 60 mg/kg of EBP while a significant difference in body weight was observed at 80 mg/kg of EBP treatment on the 6th and 7th day (p < 0.05; F value 3.048). The average body weight was reduced by 1.5 and 10.1% on day 7 compared to initial body weights (day 0) in 60 and 80 mg/kg treatment groups, respectively. Similarly, the average body weight was reduced by 14.5 and 13.8% in 100 and 120 mg/kg groups, respectively, while body weights could not be recorded in the 140 mg/kg group due to treatment-related mortality in all animals. Further, statistical comparison of body weights could not be performed in the 100, 120 and 140 mg/kg groups since the sample size was too small (n < 2) due to treatment-related mortalities within 2 hours of EBP administration (Table 1).

Rectal temperatures were recorded at study initiation (0 min) and then at 30, 60, 90, 120, 180, 240 and 480 min. The range of rectal temperature at 0 min was 35.8 ± 1 °C. No change in rectal temperature was observed at the lowest dose (60 mg/kg) during the 480 min observation period. One animal in the 80 mg/kg EBP group exhibited a 2 °C fall in body temperature after 30 min and subsequently died with 1 hour of treatment. No other animal in this group exhibited hypothermia. All animals (n = 4) in the next higher dose group (100 mg/kg) exhibited marked hypothermia (>1.9 °C) after 30–60 min post-administration. Three animals of this group died within 30 min of induction of hypothermia while one animal recovered from hypothermia and survived until study termination (day 7). Similarly, one animal died of hypothermia after 60 min while one animal recovered from hypothermia and survived until study termination. Two additional animals in the 120 mg/kg group and all animals in the 140 mg/kg group died within 30 min of EBP administration and hence, rectal temperatures could not be recorded at 30 min. Nonetheless, it is plausible to assume that these animals which died within 30 min of EBP administration could have also exhibited hypothermia considering the general trend in the lower dose group (100 mg/kg EBP) and the observed hypothermia in two animals of the 120 mg/kg group. A similar hypothermic response has also been reported following the administration of toxins and toxic chemicals in rodents (Ismail et al. 1990).

Clinical signs of toxicity were recorded through the duration of study. A complete list of all parameters is provided in Table 2. The frequency and intensity of adverse effects exhibited a dose-dependent pattern. Apart from hypothermia and death as discussed above, other signs of toxicity observed were tremors, decreased grasp/hold, pilo-erection, weak grip strength, dull furcoat and hunched posture with shrunken flanks. The clinical signs of toxicity observed in animals that died during the study and those which survived are shown in Table 3.

EBP was identified as a 50 kDa heteropolysaccharide composed of pentose and hexose sugars. The in vitro toxicity of EBP was apparent at high concentration and comparable to several commercially important polysaccharides, thus indicating the biocompatible nature of EBP. The LD₅₀ of EBP in mice was determined to be 92.31 mg/kg which is in close agreement with other, commercially important polysaccharides. The in vivo results complemented observations in in vitro studies and confirmed the high biocompatibility of EBP. The high biocompatibility supports the application of EBP as a safe biosorbent for phosphate remediation.

This work was supported by a grant from University Grants Commission (UGC), India. TK received Maulana Azad National Fellowship from UGC.