Metformin hydrochloride

Biodegradation in water and sediment: simulation tests

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

biodegradation in water: simulation testing on ultimate degradation in surface water

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2008-09-08 to 2009-10-05

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 308 (Aerobic and Anaerobic Transformation in Aquatic Sediment Systems)

Version / remarks:

24 April 2002

Deviations:

no

GLP compliance:

yes (incl. QA statement)

Radiolabelling:

yes

Remarks:

The label is at the carbon of the two imino-groups C=NH

Oxygen conditions:

aerobic

Inoculum or test system:

natural water / sediment

Details on source and properties of surface water:

The water/sediment systems were sampled from a river (Rhine River, Mumpf, AG, Switzerland) and from a pond (Fröschweiher pond, Möhlin, AG, Switzerland) and consisted of natural water filtered through a 0.2 mm sieve and the upper 10 cm layer of sediment sieved through a 2 mm mesh.


Test System Origin Date of Sampling
River River Rhine, Mumpf, AG, Switzerland 04-Feb-2009
Pond Fröschweiher pond, Möhlin, AG, Switzerland 04-Feb-2009

The water was sampled at a depth of about 30 cm and the sediment was sampled from the top 10 cm layer. The sampling locations were not in areas that received effluent discharges. They were also located far from human activity. The sampling sites were located 1 m from firm land. The aquatic systems were transported to Harlan Laboratories Ltd. in sealed containers.

Details on source and properties of sediment:

The water/sediment systems were sampled from a river (Rhine River, Mumpf, AG, Switzerland) and from a pond (Fröschweiher pond, Möhlin, AG, Switzerland) and consisted of natural water filtered through a 0.2 mm sieve and the upper 10 cm layer of sediment sieved through a 2 mm mesh.


Test System Origin Date of Sampling
River River Rhine, Mumpf, AG, Switzerland 04-Feb-2009
Pond Fröschweiher pond, Möhlin, AG, Switzerland 04-Feb-2009

The water was sampled at a depth of about 30 cm and the sediment was sampled from the top 10 cm layer. The sampling locations were not in areas that received effluent discharges. They were also located far from human activity. The sampling sites were located 1 m from firm land. The aquatic systems were transported to Harlan Laboratories Ltd. in sealed containers.

Details on inoculum:

The water-sediment systems were stored at 5 ± 1 °C before use during one week. The sediment was passed through a 2.0 mm sieve and the water through a 0.2 mm sieve. Then, the sediments were added to the flasks to a depth of 2.1 cm (river) and 2.3 cm (pond), corresponding to a wet weight of 200 g (river) and 180 g (pond). This corresponds to 120 g dry weight for the river and 81.5 g dry weight for the pond system. A water volume of 600 mL was added to the sediment to reach a depth of 7.0 cm in both systems. The sediment / water volume ratio was about 1:3. After filling into the metabolism flasks, the sediments were allowed to settle down and acclimatisation under aerobic conditions was started.

Duration of test (contact time):

>= 56 - <= 100 d

Initial conc.:

0.021 mg/L

Based on:

test mat.

Parameter followed for biodegradation estimation:

CO2 evolution

radiochem. meas.

Details on study design:

Preparation of the Aquatic Systems
The water-sediment systems were stored at 5 ± 1 °C before use during one week. The sediment was passed through a 2.0 mm sieve and the water through a 0.2 mm sieve. Then, the sediments were added to the flasks to a depth of 2.1 cm (river) and 2.3 cm (pond), corresponding to a wet weight of 200 g (river) and 180 g (pond). This corresponds to 120 g dry weight for the river and 81.5 g dry weight for the pond system. A water volume of 600 mL was added to the sediment to reach a depth of 7.0 cm in both systems. The sediment / water volume ratio was about 1:3. After filling into the metabolism flasks, the sediments were allowed to settle down and acclimatisation under aerobic conditions was started.

Experimental Design
The experiment was performed in open gas-flow-systems in 1000 mL glass metabolism flasks (inner diameter: approximately 10.6 cm, surface area: approximately 88.2 cm2, see Scheme 1). The flasks were incubated under continuous ventilation with moistened air at 20 ± 2 °C in the dark. The water phases were slightly agitated from the top without disturbing the sediment. The flasks were connected to a series of two volatile traps, the first trap containing 50 mL ethylene glycol and the second trap 60 mL 2N NaOH to trap organic volatiles and 14CO2, respectively.
Samples were uniquely identified by project number, sediment type etc. according to Harlan Laboratories Ltd.'s identification system as stated in the SOP.

Acclimation
After a visual settling of colloidal particles, the water/sediment systems were equilibrated under aerobic conditions for approximately two weeks at 20 ± 2 °C in the dark before treatment. During this period, the pH, oxygen content, redox potential of the water and, in addition, the redox potential of the sediment was measured a least twice a week in control duplicates.

Monitoring of Physico-Chemical Parameters
At the start, during and/or at the end of the incubation period, the following parameters were measured:
Water: pH, redox potential, oxygen content and TOC (not during).
Sediment: pH, redox potential, biomass (start and end only) and TOC (start and end only).

The day before application (day -1), the equilibrium obtained was considered to be sufficient to apply the test item. On this day, the water phases of all samples to be treated and the two control samples were analysed for their oxygen concentration, pH and redox potential. In addition, the redox potentials in the sediments were measured.

These parameters were also measured at each sampling interval in the treated samples to be worked-up and in two control samples.

Determination of Microbial Biomass in the Sediments
The microbial biomass of the sediments was determined by using a modification of the respiratory method described by Anderson and Domsch. The sieved sediments were mixed with sea sand to facilitate aeration. Thereafter, subsamples were amended with increasing amounts of glucose and submitted for respiratory measurements, which were used to determine the microbial biomass of the sediment. The determination of the microbial biomass was performed at 20 ± 2 °C in the dark.
Aliquots of the sediment/sand mixture were packed into all-glass columns and measured semi-continuously by means of an IR-gas-analyser (X STREAM R, Emerson Process Management) for a minimum of 3 hours. The total volume of CO2, evolved during approximately one hour, was calculated. For the calculations, the values, which corresponded to a maximum initial and constant CO2-production, were used. The value obtained was extrapolated to 100 g dry sediment.

Treatment and Sampling
The target was to apply 14C-Metformin HCl to the aerobic aquatic systems at a quantity corresponding to about 11.0 μg 14C-Metformin HCl per sample. This rate was also considered sufficient to enable analysis of the parent compound and its degradation products. The concentration was far below its water solubility.

Preparation of the Stock Solution
The test item 14C-Metformin was supplied by the Sponsor, dissolved in about 1 ml ethanol/water (4:1; v/v) as free base.

Preparation of the Application Solution
An application solution was prepared by diluting the entire volume of the stock solution into 40 mL of 0.17 mM HCl (equimolar concentration) to yield a neutral solution. Its radioactivity content was measured by Liquid Scintillation Counting (LSC). This solution amounted to 1’695’153’200 dpm in 41 ml 0.17 mM HCl. Based on the mean radioactivity measured and the known specific activity of the test item (27.79 MBq/mg), the amount of the test item in the application solution was calculated to be 1.0 mg 14C-Metformin HCl in 41 ml 0.17 mM HCl. An aliquot of 500 μL of the application solution was calculated to be applied to each sample to reach the target concentration.

Treatment
The purity and stability of the test item were determined in the application solution prior to and after the treatment.
A volume of 500 μL of the application solution (or 0.16 mM HCl solution for the control samples) was applied dropwise and evenly onto the water surface of each sample using a Hamilton syringe. The actual amount of 14C-Metformin HCl applied was accurately determined as follows: Prior to and after treatment of the test systems, the same application volume was separately applied to 200 mL acetonitrile/water (4:1; v/v). Triplicate aliquots of 1 mL were measured by LSC. The mean radioactivity of an application aliquot taken before and after treatment amounted to 20’757’200 dpm (corresponding to 12.45 μg 14C-Metformin HCl per sample) and was taken as the 100% value. After treatment, the samples were connected to the flow-through system and incubated at 20 ± 2 °C in the dark. During incubation, the water was gently agitated without disturbing the sediment.

Sampling - Water and Sediment
Duplicate samples for the river system were taken for analyses immediately after treatment (time 0) and after 1, 7, 12, 21, 30 and 56 days of incubation. For the pond system duplicate samples were taken immediately after treatment (time 0) and after 1, 7, 12, 21, 56 and 100 days of incubation.

Sampling - Volatiles
Ethylene glycol and sodium hydroxide (NaOH) traps were monitored for radioactivity by LSC at each sampling interval and were additionally exchanged in the river at day 12, 20, 33, 48 and 55. The actual volume of ethylene glycol and sodium hydroxide solutions was determined and their radioactivity content was measured separately for each trap. The verification of the nature of the radioactivity in the sodium hydroxide traps of the the river test system was performed on a pool of sodium hydroxide solutions from different intervals. For this purpose, 0.5 mL of the alkaline solution was diluted with 3 mL of bi-distilled water and the precipitation was induced by addition of 3 to 6.5 mL of a saturated barium hydroxide solution. The suspension was centrifuged for 5 to 10 minutes and the supernatant tested for quantitative precipitation by adding another one to two drops of the saturated Ba(OH)2 solution. If no turbidity developed upon the second addition of Ba(OH)2, the supernatant was counted by LSC measurement. If turbidity was observed, LSC measurement was performed after another precipitation step. The absence of radioactivity in the supernatant after precipitation was taken as proof that only 14CO2 was present in the NaOH solutions.

Extraction Procedure and Sample Preparation - Water Samples
The water phase was removed from the sediment with a pipette without disturbing the underlying sediment. Practically the total amount of water was removed. The small quantity of residual sediment pore water was treated as sediment in further processing as well as in material balance calculations. The water phase was first submitted to LSC measurement for determination of its radioactivity content before being analysed by HPLC and 1D-TLC after the water phases were concentrated under reduced pressure at 38 °C using a rotary evaporator. The resulting solution was re-measured by LSC to determine the work-up recovery. The water from day 0 was analysed directly by HPLC without concentration. A characterisation of the radioactive fractions was carried out with the reference item supplied by the Sponsor.
In the case of high radioactive carbon dioxide formation in the river system from day 7 onwards the water was additionally treated as follows:
After removing the water phase from the sediment, the radioactivity in the water was determined indirectly by the difference between the initial radioactivity in the water and the remaining activity after the residual 14CO2 was stripped off after acidifying the samples to pH 2-3 using HCl. An aliquot of 1 mL water was used for this purpose. Inorganic carbon will thus be removed and the residual activity measured, derived from organic material.

Extraction Procedure and Sample Preparation - Sediments
After removing the water phase from the test system, the sediment was submitted to an extraction step using acetonitrile/water (4:1; v/v) at room temperature. From day 1 onwards the latter extractant was replaced by acetonitrile/0.1 M HCl (1:1; v/v) which was used up to four extraction steps. Extractions at room temperature were performed in a shaker at about 200-250 strokes per minute each for about 30 minutes. The radioactivity in the individual extracts was quantified by LSC (duplicate aliquots). The amount of solvent used was in general about 1 mL/g sediment (wet weight basis). Soxhlet extraction using acetonitrile/water (4:1; v/v) for 4 hours was additionally performed on the extracted sediments from day 1 onwards. This extraction method was performed when more than 10% of the applied radioactivity remained non-extractable. Due to the high radioactive carbon dioxide formation in the river system, the sediment was additionally treated on day 30 and 56 as follows:
After removing the water phase from the test system, the radioactive carbon dioxide in the sediment was determined directly using NaOH traps after the addition of the extractant acetonitrile/0.1 M HCl (1:1; v/v). The carbon dioxide was released from the sediment by repeated shaking over the course of one hour and under continuous ventilation with air. The first NaOH trap was exchanged once during this period. Afterwards, the sediment was extracted as described above. All extracts containing more than 2% of the radioactivity applied were combined and concentrated in a rotary evaporator at about 30 °C. The concentrated extracts were measured by LSC for recovery and submitted to HPLC and/ or 1D-TLC analysis. The remaining non-extractables in the sediments after the extraction procedure were quantified by LSC after combustion of aliquots up to 1.0 g of air-dried and homogenised sediment. Reflux extraction with acetonitrile/0.1 M HCl (1:1; v/v) for at least four hours was conducted followed additionally by organic matter fractionation for one interval of the river and pond test system (day 56, duplicate determination). The radioactivity content in the reflux extracts was determined by LSC. The reflux extracts were analysed by HPLC.

Extraction Procedure and Sample Preparation - Organic Matter Fractionation
After the room temperature and reflux extractions, the non-extractable residues in the sediments from one interval of both the river and pond test system (day 56, duplicate determination) were subsequently submitted to organic matter fractionation to investigate the nature of the remaining non-extractable radioactivity. Fractionation of the sediment organic matter was performed in order to determine the percentage of the applied radioactivity bound to the humic and fulvic acids as well as the humin fraction of the sediment. A procedure based on Stevenson was used. Basically, this fractionation consists of the extraction of sediment with 0.5 M NaOH solution, and subsequent precipitation of the humic acids by reducing to pH 1. Centrifugation permits separation of the fulvic acids, which remain in the liquid-phase. The humin fraction remains non-dissolved together with the clay minerals and aluminium oxides.

Compartment:

natural water / sediment

% Recovery:

94.5

St. dev.:

6.2

Remarks on result:

other: total recovery of river system

Compartment:

natural water / sediment: freshwater

% Recovery:

1.3

Remarks on result:

other: recovery in river water after 56 days

Compartment:

natural water / sediment

% Recovery:

97

St. dev.:

3.3

Remarks on result:

other: total recovery in pond system

Compartment:

natural water / sediment: freshwater

% Recovery:

11.5

Remarks on result:

other: recovery in pond water after 56 days

Compartment:

natural water / sediment: freshwater

% Recovery:

8.2

Remarks on result:

other: recovery in pond water after 100 days

Key result

% Degr.:

74.9

Parameter:

radiochem. meas.

Sampling time:

56 d

Remarks on result:

other: river

Key result

% Degr.:

2.1

Parameter:

radiochem. meas.

Sampling time:

100 d

Remarks on result:

other: pond

Key result

Compartment:

water

DT50:

7.5 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: river

Key result

Compartment:

entire system

DT50:

14.2 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: river

Key result

Compartment:

water

DT50:

4.2 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: pond

Key result

Compartment:

entire system

DT50:

152.3 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: pond

Transformation products:

yes

Remarks:

only at very low amounts (<1.2% at river and

No.:

#1

Details on transformation products:

A few minor metabolites were formed in the entire systems, but none individually exceeded 1.2% of the applied radioactivity in either aquatic system.

Evaporation of parent compound:

no

Volatile metabolites:

no

Residues:

no

Details on results:

TEST CONDITIONS
- Aerobicity (or anaerobicity), moisture, temperature and other experimental conditions maintained throughout the study: Yes
- Anomalies or problems encountered: No

MAJOR TRANSFORMATION PRODUCTS in Warter
No major transformation product identified

MINOR TRANSFORMATION PRODUCTS in Water
- Range of maximum concentrations in % of the applied amount and day(s) of incubation when observed:
River: Metabolite M1(1.2%), Metabolite M4 (0.4%)
Pond: no transformation product identified in water
- Range of maximum concentrations in % of the applied amount at end of study period: No transformation product identified in freshwater from river (day 56) and pond (day 100).

TOTAL UNIDENTIFIED RADIOACTIVITY (RANGE) OF APPLIED AMOUNT:
3 metabolites detected in very low amounts.

MINERALISATION
- % of applied radioactivity present as CO2 at end of study: 74.9% (river) and 2.1% (pond)

VOLATILIZATION
- % of the applied radioactivity present as volatile organics at end of study: <0.1% (river and pond)

The parent substance dissipated very rapidly from the water phase of both systems. The test item in the river system decreased from initial levels of 101.3% of the applied radioactivity to less than 8.6% of applied. Three metabolites were detected in the river water phase in minor amounts, none exceeding 1.2% of applied radioactivity. The test item in the pond system decreased from initial levels of 101.1% of the applied radioactivity to 8.2% of applied. Only one metabolite was detected to a level which was below the limit of quantification. Details are provided in the attached table.

Validity criteria fulfilled:

yes

Conclusions:

In aerobic aquatic systems, 14C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

Executive summary:

The route and rate of degradation of 14C-Metformin HCl, i.e. N,N-Dimethylbiguanide hydrochloride, was investigated in two aquatic systems (river and pond) incubated under aerobic conditions at 20 °C in the dark.

The water/sediment systems were sampled from a river (Rhine River, Mumpf, AG, Switzerland) and from a pond (Fröschweiher pond, Möhlin, AG, Switzerland) and consisted of natural water filtered through a 0.2 mm sieve and the upper 10 cm layer of sediment sieved through a 2 mm mesh.

The test systems were first acclimated under aerobic conditions in the dark for about two weeks prior to treatment. During this time, the measured values for pH, oxygen concentration and redox potential in water, and redox potential in sediment had reached relatively constant values.

14C-Metformin HCl was then applied to the water surface of each aquatic sample at a dose of about 0.021 mg test item per litre water. During the entire incubation period, the water phase was aerated by a stream of air through the samples whilst the sediment remained undisturbed. The exiting air passed through sodium hydroxide and ethylene glycol solutions to trap 14CO2 and organic volatiles, respectively.

Duplicate samples for the river system were taken for analyses immediately after treatment (time 0) and after 1, 7, 12, 21, 30 and 56 days of incubation. The river part of the study was terminated after 56 rather than the standard 100 days, since 74.9% of 14C-Metformin was mineralised to radioactive carbon dioxide within 56 days. Additionally, 12.8% of the applied radioactivity remained non-extractable and the level of the radioactivity in the water and extractables was 3.7% of applied. For the pond system duplicate samples were taken immediately after treatment (time 0) and after 1, 7, 12, 21, 56 and 100 days of incubation.

The water and sediment phases were separated and the sediment exhaustively extracted with acetonitrile/0.1 M HCl (4:1; v/v). Soxhlet extraction using acetonitrile/water (4:1; v/v) was additionally performed on the sediments from day 1 onwards. All samples were measured by LSC for determination of their radioactivity content. The water phases and sediment extracts were then submitted to chromatographic analysis using HPLC for quantification of the radioactive fractions present. A total radioactivity balance and the distribution of radioactivity were established for each interval.

Total recoveries of the applied radioactivity averaged 94.5% ± 6.2% and 97.0% ± 3.3% in the river and pond systems, respectively.

The level of radioactivity decreased steadily in the water phase of both aquatic systems. In the river system the level of radioactivity decreased further to levels of 1.3% after 56 days of incubation. An amount of 8.2% of the applied radioactivity was determined in the water phase of the pond system after 100 days of incubation.

The amount of total radioactivity in the sediment increased with time, from an initial level of 1.5% of the applied radioactivity for the river system to 46.1% within 7 days of incubation. In the river system the radioactivity decreased from day 7 onwards to 15.2% after 56 days of incubation due to degradation of 14C-Metformin HCl. The amount of radioactivity in the pond sediment increased during the entire incubation period. At day 100 the majority of the radioactivity (81.5% of applied) was found in the pond sediment.

Non-extractable residues accounting for up to 13.9% of applied radioactivity were observed for the river and 35.6% for the pond system throughout the study. Organic matter fractionation of the bound residues on day 56 showed that the majority of the radioactivity was associated with the insoluble fraction (humin).

In the entire river system, the amount of parent substance declined after 56 days of incubation from an initial level of 101.3% to 2.2% of the applied radioactivity. The initial amount of 101.1% of the item, observed in the entire pond system, declined to 56.6% of the applied radioactivity until the end of the study (day 100).

Besides the parent compound, several minor degradation products were detected, none individually exceeding 1.2% of the applied radioactivity in both systems.

The rates of dissipation (DT50, DT75, DT90) of 14C-Metformin HCl from the water phase and the entire system were calculated by using the Origin calculation software.

River - Water                DT50: 7.5d       DT75: 14.9d       DT90: 24.8d       

River - Total System      DT50: 14.2d       DT75: 28.5d       DT90: 47.3d

Pond - Water                DT50: 4.2d       DT75: 25.1d       DT90: 62.7d       

Pond - Total System      DT50: 152.3d       DT75: >1y       DT90: >1y

A very high rate of mineralisation of the test item was observed in the river system throughout the study. Radioactive carbon dioxide accounted for 74.9% of the applied radioactivity by the end of the incubation period of 56 days. In the pond system, the production of 14CO2 was comparatively low accounting for 2.1% of the applied radioactivity after 100 days of incubation. No organic volatile compounds were formed at any time point during the study (<0.1% of applied).

In aerobic aquatic systems, 14C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

Conclusion

In aerobic aquatic systems,¹⁴C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

Reference 2

Endpoint:

biodegradation in water: sediment simulation testing

Type of information:

experimental study

Adequacy of study:

key study

Study period:

2008-09-08 to 2009-10-05

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

according to guideline

Guideline:

OECD Guideline 308 (Aerobic and Anaerobic Transformation in Aquatic Sediment Systems)

Version / remarks:

24 April 2002

Deviations:

no

GLP compliance:

yes (incl. QA statement)

Radiolabelling:

yes

Remarks:

The label is at the carbon of the two imino-groups C=NH

Oxygen conditions:

aerobic

Inoculum or test system:

natural water / sediment

Details on source and properties of surface water:

The water/sediment systems were sampled from a river (Rhine River, Mumpf, AG, Switzerland) and from a pond (Fröschweiher pond, Möhlin, AG, Switzerland) and consisted of natural water filtered through a 0.2 mm sieve and the upper 10 cm layer of sediment sieved through a 2 mm mesh.


Test System Origin Date of Sampling
River River Rhine, Mumpf, AG, Switzerland 04-Feb-2009
Pond Fröschweiher pond, Möhlin, AG, Switzerland 04-Feb-2009

The water was sampled at a depth of about 30 cm and the sediment was sampled from the top 10 cm layer. The sampling locations were not in areas that received effluent discharges. They were also located far from human activity. The sampling sites were located 1 m from firm land. The aquatic systems were transported to Harlan Laboratories Ltd. in sealed containers.

Details on source and properties of sediment:

The water/sediment systems were sampled from a river (Rhine River, Mumpf, AG, Switzerland) and from a pond (Fröschweiher pond, Möhlin, AG, Switzerland) and consisted of natural water filtered through a 0.2 mm sieve and the upper 10 cm layer of sediment sieved through a 2 mm mesh.


Test System Origin Date of Sampling
River River Rhine, Mumpf, AG, Switzerland 04-Feb-2009
Pond Fröschweiher pond, Möhlin, AG, Switzerland 04-Feb-2009

The water was sampled at a depth of about 30 cm and the sediment was sampled from the top 10 cm layer. The sampling locations were not in areas that received effluent discharges. They were also located far from human activity. The sampling sites were located 1 m from firm land. The aquatic systems were transported to Harlan Laboratories Ltd. in sealed containers.

Details on inoculum:

The water-sediment systems were stored at 5 ± 1 °C before use during one week. The sediment was passed through a 2.0 mm sieve and the water through a 0.2 mm sieve. Then, the sediments were added to the flasks to a depth of 2.1 cm (river) and 2.3 cm (pond), corresponding to a wet weight of 200 g (river) and 180 g (pond). This corresponds to 120 g dry weight for the river and 81.5 g dry weight for the pond system. A water volume of 600 mL was added to the sediment to reach a depth of 7.0 cm in both systems. The sediment / water volume ratio was about 1:3. After filling into the metabolism flasks, the sediments were allowed to settle down and acclimatisation under aerobic conditions was started.

Duration of test (contact time):

>= 56 - <= 100 d

Initial conc.:

0.021 mg/L

Based on:

test mat.

Parameter followed for biodegradation estimation:

CO2 evolution

radiochem. meas.

Details on study design:

Preparation of the Aquatic Systems
The water-sediment systems were stored at 5 ± 1 °C before use during one week. The sediment was passed through a 2.0 mm sieve and the water through a 0.2 mm sieve. Then, the sediments were added to the flasks to a depth of 2.1 cm (river) and 2.3 cm (pond), corresponding to a wet weight of 200 g (river) and 180 g (pond). This corresponds to 120 g dry weight for the river and 81.5 g dry weight for the pond system. A water volume of 600 mL was added to the sediment to reach a depth of 7.0 cm in both systems. The sediment / water volume ratio was about 1:3. After filling into the metabolism flasks, the sediments were allowed to settle down and acclimatisation under aerobic conditions was started.

Experimental Design
The experiment was performed in open gas-flow-systems in 1000 mL glass metabolism flasks (inner diameter: approximately 10.6 cm, surface area: approximately 88.2 cm2, see Scheme 1). The flasks were incubated under continuous ventilation with moistened air at 20 ± 2 °C in the dark. The water phases were slightly agitated from the top without disturbing the sediment. The flasks were connected to a series of two volatile traps, the first trap containing 50 mL ethylene glycol and the second trap 60 mL 2N NaOH to trap organic volatiles and 14CO2, respectively.
Samples were uniquely identified by project number, sediment type etc. according to Harlan Laboratories Ltd.'s identification system as stated in the SOP.

Acclimation
After a visual settling of colloidal particles, the water/sediment systems were equilibrated under aerobic conditions for approximately two weeks at 20 ± 2 °C in the dark before treatment. During this period, the pH, oxygen content, redox potential of the water and, in addition, the redox potential of the sediment was measured a least twice a week in control duplicates.

Monitoring of Physico-Chemical Parameters
At the start, during and/or at the end of the incubation period, the following parameters were measured:
Water: pH, redox potential, oxygen content and TOC (not during).
Sediment: pH, redox potential, biomass (start and end only) and TOC (start and end only).

The day before application (day -1), the equilibrium obtained was considered to be sufficient to apply the test item. On this day, the water phases of all samples to be treated and the two control samples were analysed for their oxygen concentration, pH and redox potential. In addition, the redox potentials in the sediments were measured.

These parameters were also measured at each sampling interval in the treated samples to be worked-up and in two control samples.

Determination of Microbial Biomass in the Sediments
The microbial biomass of the sediments was determined by using a modification of the respiratory method described by Anderson and Domsch. The sieved sediments were mixed with sea sand to facilitate aeration. Thereafter, subsamples were amended with increasing amounts of glucose and submitted for respiratory measurements, which were used to determine the microbial biomass of the sediment. The determination of the microbial biomass was performed at 20 ± 2 °C in the dark.
Aliquots of the sediment/sand mixture were packed into all-glass columns and measured semi-continuously by means of an IR-gas-analyser (X STREAM R, Emerson Process Management) for a minimum of 3 hours. The total volume of CO2, evolved during approximately one hour, was calculated. For the calculations, the values, which corresponded to a maximum initial and constant CO2-production, were used. The value obtained was extrapolated to 100 g dry sediment.

Treatment and Sampling
The target was to apply 14C-Metformin HCl to the aerobic aquatic systems at a quantity corresponding to about 11.0 μg 14C-Metformin HCl per sample. This rate was also considered sufficient to enable analysis of the parent compound and its degradation products. The concentration was far below its water solubility.

Preparation of the Stock Solution
The test item 14C-Metformin was supplied by the Sponsor, dissolved in about 1 ml ethanol/water (4:1; v/v) as free base.

Preparation of the Application Solution
An application solution was prepared by diluting the entire volume of the stock solution into 40 mL of 0.17 mM HCl (equimolar concentration) to yield a neutral solution. Its radioactivity content was measured by Liquid Scintillation Counting (LSC). This solution amounted to 1’695’153’200 dpm in 41 ml 0.17 mM HCl. Based on the mean radioactivity measured and the known specific activity of the test item (27.79 MBq/mg), the amount of the test item in the application solution was calculated to be 1.0 mg 14C-Metformin HCl in 41 ml 0.17 mM HCl. An aliquot of 500 μL of the application solution was calculated to be applied to each sample to reach the target concentration.

Treatment
The purity and stability of the test item were determined in the application solution prior to and after the treatment.
A volume of 500 μL of the application solution (or 0.16 mM HCl solution for the control samples) was applied dropwise and evenly onto the water surface of each sample using a Hamilton syringe. The actual amount of 14C-Metformin HCl applied was accurately determined as follows: Prior to and after treatment of the test systems, the same application volume was separately applied to 200 mL acetonitrile/water (4:1; v/v). Triplicate aliquots of 1 mL were measured by LSC. The mean radioactivity of an application aliquot taken before and after treatment amounted to 20’757’200 dpm (corresponding to 12.45 μg 14C-Metformin HCl per sample) and was taken as the 100% value. After treatment, the samples were connected to the flow-through system and incubated at 20 ± 2 °C in the dark. During incubation, the water was gently agitated without disturbing the sediment.

Sampling - Water and Sediment
Duplicate samples for the river system were taken for analyses immediately after treatment (time 0) and after 1, 7, 12, 21, 30 and 56 days of incubation. For the pond system duplicate samples were taken immediately after treatment (time 0) and after 1, 7, 12, 21, 56 and 100 days of incubation.

Sampling - Volatiles
Ethylene glycol and sodium hydroxide (NaOH) traps were monitored for radioactivity by LSC at each sampling interval and were additionally exchanged in the river at day 12, 20, 33, 48 and 55. The actual volume of ethylene glycol and sodium hydroxide solutions was determined and their radioactivity content was measured separately for each trap. The verification of the nature of the radioactivity in the sodium hydroxide traps of the the river test system was performed on a pool of sodium hydroxide solutions from different intervals. For this purpose, 0.5 mL of the alkaline solution was diluted with 3 mL of bi-distilled water and the precipitation was induced by addition of 3 to 6.5 mL of a saturated barium hydroxide solution. The suspension was centrifuged for 5 to 10 minutes and the supernatant tested for quantitative precipitation by adding another one to two drops of the saturated Ba(OH)2 solution. If no turbidity developed upon the second addition of Ba(OH)2, the supernatant was counted by LSC measurement. If turbidity was observed, LSC measurement was performed after another precipitation step. The absence of radioactivity in the supernatant after precipitation was taken as proof that only 14CO2 was present in the NaOH solutions.

Extraction Procedure and Sample Preparation - Water Samples
The water phase was removed from the sediment with a pipette without disturbing the underlying sediment. Practically the total amount of water was removed. The small quantity of residual sediment pore water was treated as sediment in further processing as well as in material balance calculations. The water phase was first submitted to LSC measurement for determination of its radioactivity content before being analysed by HPLC and 1D-TLC after the water phases were concentrated under reduced pressure at 38 °C using a rotary evaporator. The resulting solution was re-measured by LSC to determine the work-up recovery. The water from day 0 was analysed directly by HPLC without concentration. A characterisation of the radioactive fractions was carried out with the reference item supplied by the Sponsor.
In the case of high radioactive carbon dioxide formation in the river system from day 7 onwards the water was additionally treated as follows:
After removing the water phase from the sediment, the radioactivity in the water was determined indirectly by the difference between the initial radioactivity in the water and the remaining activity after the residual 14CO2 was stripped off after acidifying the samples to pH 2-3 using HCl. An aliquot of 1 mL water was used for this purpose. Inorganic carbon will thus be removed and the residual activity measured, derived from organic material.

Extraction Procedure and Sample Preparation - Sediments
After removing the water phase from the test system, the sediment was submitted to an extraction step using acetonitrile/water (4:1; v/v) at room temperature. From day 1 onwards the latter extractant was replaced by acetonitrile/0.1 M HCl (1:1; v/v) which was used up to four extraction steps. Extractions at room temperature were performed in a shaker at about 200-250 strokes per minute each for about 30 minutes. The radioactivity in the individual extracts was quantified by LSC (duplicate aliquots). The amount of solvent used was in general about 1 mL/g sediment (wet weight basis). Soxhlet extraction using acetonitrile/water (4:1; v/v) for 4 hours was additionally performed on the extracted sediments from day 1 onwards. This extraction method was performed when more than 10% of the applied radioactivity remained non-extractable. Due to the high radioactive carbon dioxide formation in the river system, the sediment was additionally treated on day 30 and 56 as follows:
After removing the water phase from the test system, the radioactive carbon dioxide in the sediment was determined directly using NaOH traps after the addition of the extractant acetonitrile/0.1 M HCl (1:1; v/v). The carbon dioxide was released from the sediment by repeated shaking over the course of one hour and under continuous ventilation with air. The first NaOH trap was exchanged once during this period. Afterwards, the sediment was extracted as described above. All extracts containing more than 2% of the radioactivity applied were combined and concentrated in a rotary evaporator at about 30 °C. The concentrated extracts were measured by LSC for recovery and submitted to HPLC and/ or 1D-TLC analysis. The remaining non-extractables in the sediments after the extraction procedure were quantified by LSC after combustion of aliquots up to 1.0 g of air-dried and homogenised sediment. Reflux extraction with acetonitrile/0.1 M HCl (1:1; v/v) for at least four hours was conducted followed additionally by organic matter fractionation for one interval of the river and pond test system (day 56, duplicate determination). The radioactivity content in the reflux extracts was determined by LSC. The reflux extracts were analysed by HPLC.

Extraction Procedure and Sample Preparation - Organic Matter Fractionation
After the room temperature and reflux extractions, the non-extractable residues in the sediments from one interval of both the river and pond test system (day 56, duplicate determination) were subsequently submitted to organic matter fractionation to investigate the nature of the remaining non-extractable radioactivity. Fractionation of the sediment organic matter was performed in order to determine the percentage of the applied radioactivity bound to the humic and fulvic acids as well as the humin fraction of the sediment. A procedure based on Stevenson was used. Basically, this fractionation consists of the extraction of sediment with 0.5 M NaOH solution, and subsequent precipitation of the humic acids by reducing to pH 1. Centrifugation permits separation of the fulvic acids, which remain in the liquid-phase. The humin fraction remains non-dissolved together with the clay minerals and aluminium oxides.

Compartment:

natural water / sediment

% Recovery:

94.5

St. dev.:

6.2

Remarks on result:

other: total recovery in river system

Compartment:

natural water / sediment

% Recovery:

97

St. dev.:

3.3

Remarks on result:

other: total recovery in pond system

Compartment:

natural sediment

% Recovery:

15.2

Remarks on result:

other: recovery in river sediment after 56 days

Compartment:

natural sediment

% Recovery:

81.4

Remarks on result:

other: recovery in pond sediment after 56 days

Compartment:

natural sediment

% Recovery:

81.5

Remarks on result:

other: recovery in pond sediment after 100 days

Key result

% Degr.:

74.9

Parameter:

radiochem. meas.

Sampling time:

56 d

Remarks on result:

other: river

Key result

% Degr.:

2.1

Parameter:

radiochem. meas.

Sampling time:

100 d

Remarks on result:

other: pond

Key result

Compartment:

water

DT50:

7.5 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: river

Key result

Compartment:

entire system

DT50:

14.2 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: river

Key result

Compartment:

water

DT50:

4.2 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: pond

Key result

Compartment:

entire system

DT50:

152.3 d

Type:

(pseudo-)first order (= half-life)

Temp.:

20 °C

Remarks on result:

other: pond

Transformation products:

yes

No.:

#1

Details on transformation products:

A few minor metabolites were formed in the entire systems, but none individually exceeded 1.2% of the applied radioactivity in either aquatic system.

Evaporation of parent compound:

no

Volatile metabolites:

no

Residues:

no

Details on results:

RADIOACTIVITY IN SEDIMENT (see attached document "Study B68275 - Mass Balance in Water - Sediment")
The amount of total radioactivity in the sediment increased with time, from an initial level of 1.5% of the applied radioactivity for the river system to 46.1% within 7 days of incubation. In the river system the radioactivity decreased from day 7 onwards to 15.2% after 56 days of incubation due to degradation of 14C-Metformin HCl. The amount of radioactivity in the pond sediment increased during the entire incubation period. At day 100 the majority of the radioactivity (81.5% of applied) was found in the pond sediment.
Immediately after application, only 0.1% of the applied radioactivity was extractable in both systems. The total radioactivity was found in the water. The amount of total extractables in the river system reached its maximum on day 7, with 39.0% of the applied radioactivity. This amount decreased to 2.4% in the river system on day 56. The amount of total extractables in the pond system increased and reached its maximum on day 100, with 49.6% of the applied radioactivity. Soxhlet extractions, performed from day 7 onwards, recovered up to 18.1% (river) 37.6% (pond) of the applied radioactivity.
The amount of non-extractable radioactivity in the sediments increased in both test systems, reaching maximum mean values of 13.9% and 35.6% of the applied radioactivity in the river (day 30) and pond (day 56) sediments, respectively.

HARSH EXTRACTION AND ORGANIC MATTER FRACTIONATION (see attached document "Study B68275 - Distribution of radioactivity from Metformin HCl in sediment")
Acidic harsh extraction under reflux followed by organic matter fraction was conducted on the non-extracted residues from the day 56 interval.
Reflux extraction released mean amounts of 1.8% (river) and 15.4% (pond) of the applied radioactivity. The acidic reflux extract of the pond system, analysed by HPLC, consisted mainly of 14C-Metformin HCl and two other metabolites. The majority of the radioactivity found in the sediment remained still bound, with non-extractable residues accounting for 11.0% of applied radioactivity for the river and 20.2% for the pond system after reflux.
Subsequent organic matter fractionation indicated that the remaining non-extractable radioactivity was mainly bound to the humic acids and insoluble humin fraction amounting to 4.1% and 12.4% of the applied radioactivity, respectively, for the pond system.
Amounts of 0.1% and 8.5% were determined for the river system (Table 11). The corresponding range for the fulvic acids was low, i.e. 2.4% (river) and 3.7% (pond).

Validity criteria fulfilled:

yes

Conclusions:

The route and rate of degradation of 14C-Metformin HCl in two sediments (river and pond) under aerobic conditions were investigated at 20 °C in the dark.
Radiolabelled Metformin HCl dissipated rapidly from the water phase of both systems into the sediment layers. DT50 values of 7.5 and 4.2 days were calculated via single first-order (river) and dual first-order (pond) kinetics, respectively. Degradation half-lives for the total system were 14.2 days for the river and 152.3 days for the pond system.
Once in the river sediment, 74.9% of 14C-Metformin HCl was mineralised to radioactive carbon dioxide without the formation of significant quantities of metabolites and bound residues. In contrast to the river, degradation and mineralisation of 14C-Metformin HCl in the pond sediment proceeded slowly. Instead, the formation of a higher amount of bound residues was observed. The acidic reflux extract of the pond system, analysed by HPLC, consisted mainly of 14C-Metformin HCl (and two other metabolites at concentrations of 0.4 and 0.7%, respectively). Non-extractable residues accounted for up to 35.6% of applied radioactivity for the pond system. A maximum amount of 13.9% of applied was detected in the river system. Organic matter fractionation of the bound residues on day 56 showed that the majority of the radioactivity was associated with the insoluble fraction (humin) in both systems.
A very high rate of mineralisation of the test item was observed in the river system throughout the study. Radioactive carbon dioxide accounted for 74.9% of the applied radioactivity by the end of the incubation period of 56 days. In the pond system, the production of 14CO2 was comparatively low accounting for 2.1% of the applied radioactivity after 100 days of incubation. No other organic volatile compounds were formed at any time point during the study (<0.1% of applied).
A few minor metabolites were formed in the entire systems, but none individually exceeded 1.2% of the applied radioactivity in either aquatic system.
In aerobic aquatic systems, 14C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

Executive summary:

The route and rate of degradation of 14C-Metformin HCl, i.e. N,N-Dimethylbiguanide hydrochloride, was investigated in two aquatic systems (river and pond) incubated under aerobic conditions at 20 °C in the dark.

The water/sediment systems were sampled from a river (Rhine River, Mumpf, AG, Switzerland) and from a pond (Fröschweiher pond, Möhlin, AG, Switzerland) and consisted of natural water filtered through a 0.2 mm sieve and the upper 10 cm layer of sediment sieved through a 2 mm mesh.

The test systems were first acclimated under aerobic conditions in the dark for about two weeks prior to treatment. During this time, the measured values for pH, oxygen concentration and redox potential in water, and redox potential in sediment had reached relatively constant values.

14C-Metformin HCl was then applied to the water surface of each aquatic sample at a dose of about 0.021 mg test item per litre water. During the entire incubation period, the water phase was aerated by a stream of air through the samples whilst the sediment remained undisturbed. The exiting air passed through sodium hydroxide and ethylene glycol solutions to trap 14CO2 and organic volatiles, respectively.

Duplicate samples for the river system were taken for analyses immediately after treatment (time 0) and after 1, 7, 12, 21, 30 and 56 days of incubation. The river part of the study was terminated after 56 rather than the standard 100 days, since 74.9% of 14C-Metformin was mineralised to radioactive carbon dioxide within 56 days. Additionally, 12.8% of the applied radioactivity remained non-extractable and the level of the radioactivity in the water and extractables was 3.7% of applied. For the pond system duplicate samples were taken immediately after treatment (time 0) and after 1, 7, 12, 21, 56 and 100 days of incubation.

The water and sediment phases were separated and the sediment exhaustively extracted with acetonitrile/0.1 M HCl (4:1; v/v). Soxhlet extraction using acetonitrile/water (4:1; v/v) was additionally performed on the sediments from day 1 onwards. All samples were measured by LSC for determination of their radioactivity content. The water phases and sediment extracts were then submitted to chromatographic analysis using HPLC for quantification of the radioactive fractions present. A total radioactivity balance and the distribution of radioactivity were established for each interval.

Total recoveries of the applied radioactivity averaged 94.5% ± 6.2% and 97.0% ± 3.3% in the river and pond systems, respectively.

The level of radioactivity decreased steadily in the water phase of both aquatic systems. In the river system the level of radioactivity decreased further to levels of 1.3% after 56 days of incubation. An amount of 8.2% of the applied radioactivity was determined in the water phase of the pond system after 100 days of incubation.

The amount of total radioactivity in the sediment increased with time, from an initial level of 1.5% of the applied radioactivity for the river system to 46.1% within 7 days of incubation. In the river system the radioactivity decreased from day 7 onwards to 15.2% after 56 days of incubation due to degradation of 14C-Metformin HCl. The amount of radioactivity in the pond sediment increased during the entire incubation period. At day 100 the majority of the radioactivity (81.5% of applied) was found in the pond sediment.

Non-extractable residues accounting for up to 13.9% of applied radioactivity were observed for the river and 35.6% for the pond system throughout the study. Organic matter fractionation of the bound residues on day 56 showed that the majority of the radioactivity was associated with the insoluble fraction (humin).

In the entire river system, the amount of parent substance declined after 56 days of incubation from an initial level of 101.3% to 2.2% of the applied radioactivity. The initial amount of 101.1% of the item, observed in the entire pond system, declined to 56.6% of the applied radioactivity until the end of the study (day 100).

Besides the parent compound, several minor degradation products were detected, none individually exceeding 1.2% of the applied radioactivity in both systems.

The rates of dissipation (DT50, DT75, DT90) of 14C-Metformin HCl from the water phase and the entire system were calculated by using the Origin calculation software.

River - Water                DT50: 7.5d       DT75: 14.9d       DT90: 24.8d       

River - Total System      DT50: 14.2d       DT75: 28.5d       DT90: 47.3d

Pond - Water                DT50: 4.2d       DT75: 25.1d       DT90: 62.7d       

Pond - Total System      DT50: 152.3d       DT75: >1y       DT90: >1y

A very high rate of mineralisation of the test item was observed in the river system throughout the study. Radioactive carbon dioxide accounted for 74.9% of the applied radioactivity by the end of the incubation period of 56 days. In the pond system, the production of 14CO2 was comparatively low accounting for 2.1% of the applied radioactivity after 100 days of incubation. No organic volatile compounds were formed at any time point during the study (<0.1% of applied).

In aerobic aquatic systems, 14C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

Conclusion

In aerobic aquatic systems,¹⁴C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

Description of key information

The route and rate of degradation of 14C-metformin hydrochloride in water and sediment was investigated in two systems (river and pond) in a Guideline conform OECD 308 study under GLP regulation.

Radiolabelled Metformin HCl dissipated rapidly from the water phase of both systems into the sediment layers. DT50 values of 7.5 and 4.2 days were calculated via single first-order (river) and dual first-order (pond) kinetics, respectively. Degradation half-lives for the total system were 14.2 days for the river and 152.3 days for the pond system. Once in the river sediment, 74.9% of 14C-Metformin HCl was mineralised to radioactive carbon dioxide without the formation of significant quantities of metabolites and bound residues. In contrast to the river, degradation and mineralisation of 14C-Metformin HCl in the pond sediment proceeded slowly. Instead, the formation of a higher amount of bound residues was observed. The acidic reflux extract of the pond system, analysed by HPLC, consisted mainly of 14C-Metformin HCl and two other metabolites. Non-extractable residues accounted for up to 35.6% of applied radioactivity for the pond system. A maximum amount of 13.9% of applied was detected in the river system. Organic matter fractionation of the bound residues on day 56 showed that the majority of the radioactivity was associated with the insoluble fraction (humin) in both systems.

A very high rate of mineralisation of the test item was observed in the river system throughout the study. Radioactive carbon dioxide accounted for 74.9% of the applied radioactivity by the end of the incubation period of 56 days. In the pond system, the production of 14CO2 was comparatively low accounting for 2.1% of the applied radioactivity after 100 days of incubation. No other organic volatile compounds were formed at any time point during the study (<0.1% of applied).

A few minor metabolites were formed in the entire systems, but none individually exceeded 1.2% of the applied radioactivity in either aquatic system.

In aerobic aquatic systems, 14C-Metformin HCl dissipates steadily from the water phase mainly via degradation and adsorption to the sediment.

River - Water                DT50: 7.5d       DT75: 14.9d       DT90: 24.8d       

River - Total System      DT50: 14.2d       DT75: 28.5d       DT90: 47.3d

Pond - Water                DT50: 4.2d       DT75: 25.1d       DT90: 62.7d       

Pond - Total System      DT50: 152.3d       DT75: >1y       DT90: >1y

Key value for chemical safety assessment

Half-life in freshwater:

7.5 d

at the temperature of:

20 °C

Half-life in freshwater sediment:

22 d

at the temperature of:

20 °C