Transforma)ons of Illicit Drugs in Bodies and Sewers
Pros and Cons of Biomarker Consolida)on in Sewage Epidemiology 1 1 2 Kevin J. Bisceglia , Amanda R. Stashin , Katrice A. Lippa 1Department of Chemistry, Hofstra University, Hempstead, NY 11549 2Chemical Sciences Division, NIST, Gaithersburg, MD 20899
Introduc)on:
Methodology:
ā¢āÆ Sewage epidemiology (SE) aSempts to monitor community-Āālevel drug use by tracking the occurrence of relevant biomarkers in municipal sewage. SE may represent a valuable public health surveillance tool, but ļ¬rst its reliability must be established.
Biomarker stability in wastewater
ā¢āÆ Consensus prac<ce in SE is to use only the dominant drug metabolite as biomarker. When doing so, sources of uncertainty that depend on biomarker iden<ty are o^en two-Āāto-Āāļ¬ve <mes larger than those inļ¬uenced by study area1.
ā¢
ā¢
ā¢āÆ This occurs because catchment-Āāspeciļ¬c sources of uncertainty can be minimized by careful study design, while biomarker-Āāspeciļ¬c uncertain<es are both more poorly characterized and more diļ¬cult to control.
ā¢
ā¢āÆ Here, we evaluate the poten<al of composite measurands ā which capture most or all major metabolites of a target compound ā to reduce two biomarker-Āāspeciļ¬c sources of uncertainty:
Drug and (for cocaine only) metabolite transforma<on was inves<gated in sewage inļ¬uent at 9, 23, and 31 Ā°C and circumneutral pH, under condi<ons designed to s<mulate growth of suspended aerobic bacteria. Reactors (in triplicate) were spiked simultaneously with all compounds at nM concentra<ons, shaken (180 rpm) in the dark, and sampled repeatedly over 24 h. Samples were analyzed by direct injec<on liquid chromatography isotopic dilu<on mass spectrometry (LC-ĀāIDMS/MS)2. Cociaine and its metabolites were ļ¬t to an ester hydrolysis model using nonlinear least squares regression. Other drugs were ļ¬t using nonspeciļ¬c pseudo-Āāļ¬rst order kine<cs.
Analysis of metabolic excre)on proļ¬les
ā¢āÆ Mean excre<on frac<ons for composite measurands were computed as the sum of their components; uncertain<es were computed as the root sum of the squares (RSS) of component SDs.
Independent urine samples: ā¢
Rela<ve abundances of drug and drug metabolites in independently collected, but concurrently analyzed, urine samples was used as another means of evalua<ng excre<on-Āārelated variability.
ā¢
Uncertainty was evaluated as the SD among all urine samples, and from an analysis of variance/covariance (ANCOVA) among measurements within each sample. Normal distribu<ons were assumed, and evaluated when possible (KāS test).
Var X + Y = Var X + Var ! + 2Cov X, Y
Controlled dose studies:
(1)āÆ variability in the metabolic excre<on proļ¬les of drugs of abuse, which are thought to arise from gene<c and lifestyle diļ¬erences within the popula<on
ā¢
(2)āÆ variability in biomarker transforma<ons that may occur during transport municipal sewer systems.
1 Var X = nā1
For each drug, subject-Āāweighted, mean excre<on frac<ons were computed from controlled dose studies. Only studies that administered radio-Āālabeled compound and/or monitored > 80% of known metabolites were included. Standard devia<ons (SD) were computed as the root mean square error (RMSE) from an analysis of variance (ANOVA)
!
X! ā X !!!
!
1 Cov X, Y = nā1
!
X! ā X Y! ā Y !!!
Results: Cocaine
Stability of cocaine & metabolites in municipal sewage
Frac)on of a cocaine dose (and associated uncertainty) captured by diļ¬erent biomarkers
Poten)al metabolic transforma)on pathways for cocaine
20
70
Cocaine
Concentra)on (nM)
8
Controlled dose data are from a radio-Āālabeled study (n = 14)3; urine data are from
Smith (n = 30)4. Uncertainty is expressed as SD (error bars) and rela<ve SD (%RSD). Absolute excre<on of COCtot could not be es<mated from independent urine samples. Note that RSD for BE, the concensus SE biomarker for cocaine, is 5-Āā10 )mes larger than for composite measurands.
Covariance matrix (n =(n 3=0) the rela)ve abundance Variance/Covariance matrix 30)for for the relative abundance (unitless) of cocaine and its metabolites in urine (unitless) of cocaine and its major metabolits in urine a
a
COC CE EME EEE BE mOHBE EC
āH2Oā denotes hydrolysis; āOxā is oxida<on; āEtOHā is trans-Āāesteriļ¬ca<on from ethanol co-Āāinges<on; āpyrolysisā metabolites form during smoking crack cocaine
a
CE
EME
EEE
BE
mOHBE
EC
3.1 9.4 8.4 -10.3 -0.2 -11.6
133.2 28.7 -43.8 -0.3 -138.1
31.8 -32.8 -0.5 -35.0
341.7 0.1 -182.7
0.6 0.7
424.2
4 0 0
6
12
18
24
30
60
200
50
150
40
100
0
8
130
6
115
4
100
2
85
Ecgonine0 Methyl0Ester
0 0
6
12
ā¢āÆ N-Āādemethyla<on to amphetamine (AM) deriva<ves is common and may represent a substan<al por<on of an original MA load. However, AM has legal uses and mul<ple precursors; its inclusion is unlikely to reduce uncertainty.
70 0
5 10 15 20 25 30
6
12
Simulated in-Āāsewer Simulated In-Sewer Accumulation accumula)on of BE of Benzoylecgonine
Methamphetamine (MA)
S
15.0# 10.0# 5.0# 0.0# 5#
10#
15#
20#
Time#(hrs)#
25#
30#
1.6#
6.0# 4.0# 2.0# 0.0# 10#
15#
20#
25#
18
24
30
24
30
Cocaine and metabolites possessing an alkyl ester are readily hydrolyzed at pH 7.2 and 23 Ā°C. W h i l e b e n z o y l e c g o n i n e appears stable, in-Āāsewer hydrolysis of cocaine and cocaethylene can cause reasonable accumula)on (~10-Āā20%) during transport in sewer systems.
HO HO
O O OH
O
HMMA-glucuronide
Darker arrows represent dominant metabolic pathways in humans. A^er demethyla<on and cleavage of the methylenedioxy ring, a majority of the MDMA dose ubdergoes conjuga<on with glucuronides and sulfates. T h e r e d d a s h e d l i n e e n c l o s e s metabolites included in the Ī£DHMA composite measurand, while those inside of the black dashed line are represented by Ī£HMMA.
Stability of MDMA in municipal sewage
8.0#
5#
12
Frac)on of a MDMA dose (and associated uncertainty) captured by diļ¬erent biomarkers
AM#
0#
H N
HO
HO
Concentra3on#(nM)#
Concentra3on#(nM)#
Concentra0on#(nM)#
20.0#
HMMA-sulfate
O
H N
glucuronide & sulfate conjugates
10.0#
MA#
H N
S
HO
DHMA-3-sulfate
25.0#
Conclusions:
O
4-hydroxyamphetamine (OHMA)
glucuronide & sulfate conjugates
O
4-Hydroxy-3-methoxymethamphetamine (HMMA)
3,4-Dihydroxymethamphetamine (DHMA)
O
Stability of methamphetamine and amphetamine in municipal sewage
0#
H N
O
Frac)on of a methamphetamine dose (and associated uncertainty) captured by diļ¬erent biomarkers
4-Hydroxy-3-methoxyamphetamine (HMA)
HO
NH2
4-hydroxymethamphetamine (OHMA)
NH 2
O
HO
HN
O
HO
O
HO
3,4-Methylenedioxymethamphetamine (MDMA) HO
NH 2
3,4-Dihydroxyamphetamine (DHA)
H N
HO
O
6
O
Amphetamine (AM)
0
Note that the sum of all metabolites (as represented by the mass balance) remains constant, however.
HO
3,4-Methylenedioxyamphetamine (MDA)
O
18
Time (h)
HO
NH 2
O
H N
30
Poten)al metabolic transforma)on pathways for MDMA O
NH2
HN
ā¢āÆ Finally, substan<al frac<ons of MA and AM form nonspeciļ¬c compounds (e.g., hydroxybenzoic acid) that are ill-Āāsuited for SE.
Only one controlled dose study (n = 24)6 and one urine study (n = 25)7 has considered conjugate forma<on. Inclusion of hydroxylated MA metabolites reduces uncertainty, but the eļ¬ect is not large.
MDMA
24
S t a b i l i t y h a s i m p o r t a n t implica<ons for the selec<on of metabolites for monitoring drug abuse.
Abundances of EC, BE, EME, and COC are all highly nega)vely correlated in urine The black dashed line encloses metabolites in the composite measurand COCtot; The red samples. This causes uncertainty in composite measurands that consolidate these dashed line encloses those in the measurand Echyd. metabolites to be reduced, even in independent sample sets (above).
Poten)al m etabolic t ransforma)on Methamphetamine (MA) metabolism is complicated by several factors. Licit use & illicit pathways for precursors ā¢āÆ The frac<on excreted as MA can vary from 2-Āā76% depending on urinary methamphetamine pH5, which is seldom reported.
18
Ecgonine
See transformation diagram for abbreviations.
Methamphetamine
Mass Balance
16 12
COC 15.5 4.4 24.7 7.3 -16.9 -0.3 -32.6
250
Benzoylecgonine
30#
Time#(hrs)#
Composite biomarkers reduce metabolism-Āārelated uncertainty via a āfunneling eļ¬ectā wherein a group of interrelated metabolites coalesce into a single measurand. ā¢āÆ Eļ¬ects of consolida<on are par<cularly pronounced when dominant metabolic pathways involve readily hydrolysable func<onal groups (e.g., esters and phase II conjugates) and structural backbones that are stable in wastewater. ā¢āÆ Cocaine and MDMA provide two examples. Consolida<on reduces metabolism-Āārelated uncertainty for both to ā 10 % RSD, a value in keeping with uncertain<es that are not biomarker-Āāspeciļ¬c (e.g., ļ¬ow and loading variability). ā¢āÆ Cocaine and MDMA monitoring can be streamlined by pre-Āātrea<ng samples (via acid-Āā or base-Āācatalyzed hydrolysis) to reduce the number of analytes to one (EChyd) and three (MDMA, DHMAhyd, and HMMAhyd), respec<vly. ā¢āÆ Pre-Āātreatment may also minimize concern about environmental transforma<ons (e.g., in-Āāsewer produc<on of BE) ā¢āÆ Improvements in precision come at the cost of reduced informa<on about metabolite ra<os ā¢āÆ There is a dire need for addi<onal controlled dose studies for drugs of abuse, to help iden<fy other composite biomarkers and to beSer characterize metabolism-Āārelated uncertainty in general.
MDMA#
1.2# 0.8# 31#C# 23#C# 9#C#
0.4# 0.0# 0#
5#
10#
15#
20#
25#
30#
Time#(hrs)#
MDMA is stable in municipal wastewater under all condi<ons inves<gated.
Only one controlled dose study (n = 20)8 and one urine study (n = 25)9 has considered conjugate forma<on. As with MA, MDMA metabolism and excre<on are extremely sensi<ve to urinary pH, which was not controlled or recorded in these studies. RSD, though large, does decrease by a factor of 2-Āā3 a^er inclusion of conjugates, however, a ļ¬nding that is predicted by ANCOVA results (not shown). Conversely, addi<on of MDA does nou reduce variability.
Acknowledgements References We wish to thank A. Lynn Roberts (Johns Hopkins) for technical guidance, and Seth Guikema (Hopkins) and D a v e D u e w e r ( N I S T ) f o r a s s i s t a n c e w i t h environmental data distribu<ons. Funding for A. Stashin was provided by a Lister Endowed Fellowship in Chemistry Research, awarded by Dr. Bruce and Doris Lister through Hofstra University.
1.āÆ 2.āÆ 3.āÆ 4.āÆ 5.āÆ 6.āÆ 7.āÆ 8.āÆ 9.
Cas<glioni, S.; Bijlsma, L. et al. Environ. Sci. Technol. 2013, 47, 1452-Āā1460. Bisceglia, K. J.; Roberts, A. L. et al. Anal Bioanal Chem 2010, 398, 2701-Āā2712. Jeļ¬coat A. R.; Perez-ĀāReyes M. et al. Drug Metab Dispos 1989, 17, 153ā159 Smith, M. L.; Shimomura, E. et al. J Anal Toxicol 2010, 34, 57-Āā63. BeckeS, A. H.; Rowland, M. . H. Nature 1965, 206, 1260ā1261. Li, L.; Lopez, J. C. et al. Therap Drug Monit 2010, 32, 504-Āā507. Shima, N.; Kamata, H. T. et al. XenobioSca 2006, 36, 259-Āā267. Schwaninger, A. E.; Meyer, M. R. et al. Clinical chemistry 2011, 57, 1748-Āā1756. Shima, N.; Katagi, M. et al. XenobioSca 2008, 38, 314-Āā324.