3 Approaches for Metabolite Detection
In a series of research papers, the main focus was set on the detection and identification of new drug-related metabolites formed within plants after uptake of the
parent drug from water or soil. Thereby different strategies have been applied for
the structural identification of these metabolites. According to Schymanski et al.
[68], there are five levels of structure identification, ranging from level 1 (confirmed
structure), where comparison with a reference standard is obligatory, to level
5 where only the accurate mass of the potential metabolite is known. The possibility
to characterize drug-related metabolites according to level one is primarily restricted
by the commercial availability of appropriate standards or the possibility to synthesize these compounds in-house. An excellent example for a study where a high
percentage of the detected metabolites could be assigned to structures according to
level 1 following the Schymanski scheme has been published by Riemenschneider
et al. [38]. Here out of 17 carbamazepine-derived metabolites found in tomato plants
treated with a total of 7.5 mg carbamazepine per plant over a growing period of
89 days, 10 could be identified unambiguously by comparison of their highresolution mass spectra with those from commercially available standards. Nevertheless, it should be taken into account that (with a very few exceptions) the
availability of standards is limited to phase I metabolites. In most cases, only
tentative or probable structures (according to levels 2 and 3 of the Schymanski
scheme) could be presented for drug-related metabolites formed in the plant. These
are mainly based on measuring accurate mass, employing either (quadrupole) timeof-flight mass spectrometers ((Q)TOF/MS) or Orbitrap MS (mostly after chromatographic separation), together with the generation of MS/MS spectra allowing the
interpretation of molecule fragmentation. Frequently these data are combined with
information from the literature such as published pathways for the formation of
metabolites in mammals and/or published MS spectra for analogous compounds. A
few older papers propose structures on the basis of low-resolution MS, fragmentation spectra, and data from additional chemical experiments such as enzymatic
digestion or alkaline hydrolysis [36, 58, 69]. However, it should be stressed that
despite the possibility to measure accurate mass, probable structure for drug-related
transformation products in plants cannot always be proposed (level 5).
Appropriate standard substances are even more required, when, in addition to the
proposal of structures, quantitative information is desired. The large majority of
studies discussed in this chapter involve the use of MS detection employing an
electrospray source for ionization. Here the response factor for different substances
can vary substantially, allowing proper quantification only if standard substances are
available for calibration. One approach showing considerable potential to overcome
this obstacle is the methodology proposed by the Kruve group [70–72], showing a
way toward standard-free quantification. Thereby they studied the efficiency of
negative ion formation in the ESI source via deprotonation of substituted phenols
and benzoic acids [70]. The observed correlations between the ionization efficiencies
achieved and different molecular properties allowed the construction of a linear
Analytical Approaches for the Determination and Identification of Drug. . .
501
In a series of research papers, the main focus was set on the detection and identification of new drug-related metabolites formed within plants after uptake of the
parent drug from water or soil. Thereby different strategies have been applied for
the structural identification of these metabolites. According to Schymanski et al.
[68], there are five levels of structure identification, ranging from level 1 (confirmed
structure), where comparison with a reference standard is obligatory, to level
5 where only the accurate mass of the potential metabolite is known. The possibility
to characterize drug-related metabolites according to level one is primarily restricted
by the commercial availability of appropriate standards or the possibility to synthesize these compounds in-house. An excellent example for a study where a high
percentage of the detected metabolites could be assigned to structures according to
level 1 following the Schymanski scheme has been published by Riemenschneider
et al. [38]. Here out of 17 carbamazepine-derived metabolites found in tomato plants
treated with a total of 7.5 mg carbamazepine per plant over a growing period of
89 days, 10 could be identified unambiguously by comparison of their highresolution mass spectra with those from commercially available standards. Nevertheless, it should be taken into account that (with a very few exceptions) the
availability of standards is limited to phase I metabolites. In most cases, only
tentative or probable structures (according to levels 2 and 3 of the Schymanski
scheme) could be presented for drug-related metabolites formed in the plant. These
are mainly based on measuring accurate mass, employing either (quadrupole) timeof-flight mass spectrometers ((Q)TOF/MS) or Orbitrap MS (mostly after chromatographic separation), together with the generation of MS/MS spectra allowing the
interpretation of molecule fragmentation. Frequently these data are combined with
information from the literature such as published pathways for the formation of
metabolites in mammals and/or published MS spectra for analogous compounds. A
few older papers propose structures on the basis of low-resolution MS, fragmentation spectra, and data from additional chemical experiments such as enzymatic
digestion or alkaline hydrolysis [36, 58, 69]. However, it should be stressed that
despite the possibility to measure accurate mass, probable structure for drug-related
transformation products in plants cannot always be proposed (level 5).
Appropriate standard substances are even more required, when, in addition to the
proposal of structures, quantitative information is desired. The large majority of
studies discussed in this chapter involve the use of MS detection employing an
electrospray source for ionization. Here the response factor for different substances
can vary substantially, allowing proper quantification only if standard substances are
available for calibration. One approach showing considerable potential to overcome
this obstacle is the methodology proposed by the Kruve group [70–72], showing a
way toward standard-free quantification. Thereby they studied the efficiency of
negative ion formation in the ESI source via deprotonation of substituted phenols
and benzoic acids [70]. The observed correlations between the ionization efficiencies
achieved and different molecular properties allowed the construction of a linear
Analytical Approaches for the Determination and Identification of Drug. . .
501