Do initial concentration and activated sludge seasonality affect pharmaceutical biotransformation rate constants?

Abstract Pharmaceuticals find their way to the aquatic environment via wastewater treatment plants (WWTPs). Biotransformation plays an important role in mitigating environmental risks; however, a mechanistic understanding of involved processes is limited. The aim of this study was to evaluate potential relationships between first-order biotransformation rate constants (kb) of nine pharmaceuticals and initial concentration of the selected compounds, and sampling season of the used activated sludge inocula. Four-day bottle experiments were performed with activated sludge from WWTP Groesbeek (The Netherlands) of two different seasons, summer and winter, spiked with two environmentally relevant concentrations (3 and 30 nM) of pharmaceuticals. Concentrations of the compounds were measured by LC–MS/MS, microbial community composition was assessed by 16S rRNA gene amplicon sequencing, and kb values were calculated. The biodegradable pharmaceuticals were acetaminophen, metformin, metoprolol, terbutaline, and phenazone (ranked from high to low biotransformation rates). Carbamazepine, diatrizoic acid, diclofenac, and fluoxetine were not converted. Summer and winter inocula did not show significant differences in microbial community composition, but resulted in a slightly different kb for some pharmaceuticals. Likely microbial activity was responsible instead of community composition. In the same inoculum, different kb values were measured, depending on initial concentration. In general, biodegradable compounds had a higher kb when the initial concentration was higher. This demonstrates that Michealis-Menten kinetic theory has shortcomings for some pharmaceuticals at low, environmentally relevant concentrations and that the pharmaceutical concentration should be taken into account when measuring the kb in order to reliably predict the fate of pharmaceuticals in the WWTP. Key points • Biotransformation and sorption of pharmaceuticals were assessed in activated sludge. • Higher initial concentrations resulted in higher biotransformation rate constants for biodegradable pharmaceuticals. • Summer and winter inocula produced slightly different biotransformation rate constants although microbial community composition did not significantly change. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00253-021-11475-9.


Chemicals and reagents
Pharmaceutical standards were purchased in solid form from Sigma Aldrich/Fluka Analytical (Zwijndrecht, the Netherlands) and Merck Group (Darmstadt, Germany). The deuterated internal standards were acquired from Toronto Research Chemicals (North York, Canada).

Pharmaceutical extraction and analysis
For the analytical determination of micropollutant concentrations in wastewater, an optimized method was developed in collaboration with RadboudUMC. Samples (2.0 mL) were taken at time-points 0, 4, 8, 24, 48 and 96 hours after the start of incubation. Wastewater samples were centrifuged and the supernatant was stored at -20 °C until chemical analysis. To recovery the analytes present in the wastewater, solid phase extraction (SPE) was performed using Oasis HLB 3cc SPE cartridges (sorbent bed of 60 mg from Waters Corporation (Milford, USA)).
First an acidic buffer solution with a of pH ~2.2 was prepared, by dissolving 10 mM NH4HCO2, and 3.25 mL FA in 250.0 mL Milli-Q water. Subsequently the SPE cartridges were conditioned with 2.0 mL of MeOH and flushed with 2.0 mL of acidic buffer solution prior to sample loading. SPE extraction was applied on 1.0 mL supernatant, spiked with 0.1 mL of deuterated internal standards (50 µg/L in methanol), diluted in 4.0 mL acidic buffer.
Afterwards, 5.0 mL of MeOH was used to elute the retained OMPs. Vacuum was applied on each step. Next, the extracts were evaporated under a moderate stream of nitrogen using a sample concentrator set at 40 oC. Sample extracts were reconstituted in 0.1 % v/v FA (mobile phase A). To extract metformin, a liquid-liquid extraction was performed (as in Yoshida and Akane 1999) (SI 3). 0.1 mL Sodium dodecyl sulfate (2.0 mM) was added to 0.5 mL of wastewater including 1.0 mL ACN. The tubes were placed in the freezer at -20 °C for 30 min and the upper organic layer was taken for analysis. SPE and LLE recovery data was calculated dividing the measured yield concentration by the measured original concentration, see Equation 1 ) Furthermore, 1/X weighting was used to minimise the mathematical error for the for the lowest calibration points (Almeida et al. 2002). This weighing was selected because it gave the lowest standard errors for all the points in the calibration curve compared to 1/X 2 or not weighing. Quadratic curve fitting with the same 1/X weighting (y = ax2 + bx + c) was applied for the quantification of metformin. According to (Liu et al. 2019), the nonlinearity of metformin calibration curve is produced by an interplay between the deuterated internal standard and the original compound. Optimised LC-MS/MS parameters are provided in Figure S1, Table S1, and S2.

SI 2. Biotransformation rate constants
Acetaminophen was either not transformed (low concentration treatment in the summer experiment) or completely transformed within 24 hours. After 24 hours, the concentration increased again in some replicates, which might be the case due to acetaminophen contamination that occurred during the solid phase extraction method, making it difficult to obtain precise measurements at low concentrations. Therefore, only part of the curve was used, excluding these replicates/timepoints to calculate k b values. Furthermore, in order to estimate kb values for acetaminophen with an exponential decay models (according to Equation 1), the first timepoint was replaced by the theoretical concentration of the spiked treatment (i.e. background concentration + 3 or 30 nM). This was because the first measurement of acetaminophen (t = 0 h, < 20 min after addition) in the spiked activated sludge treatments were much lower than the concentration in the inactivated sludge treatment or even < LOD, likely due to fast uptake processes. One outlier was removed in order to calculate the kb for metformin in the winter experiment: a value of > 300 nM metformin in the AS3 treatment at time = 8 h, which was over a hundred times higher than the theoretically added concentration (3 nM), matching the other two technical replicates. This value was also outside the interquartile range and this was likely caused by technical sampling error.