Use of Charged Aerosol Detection as an Orthogonal Quantification Technique for Drug Metabolites in Safety Testing (MIST)

Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientific, Bannockburn, IL, USA; 2Thermo Fisher Scientific, Chelmsford, MA, USA

2 Use of Charged Aerosol Detection as an Orthogonal Quantification Technique for Drug Metaboites in Safety Testing (MIST)

Use of Charged Aerosol Detection as an Orthogonal Quantifi cation Technique for Drug Metabolites in Safety Testing (MIST) Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientifi c, Bannockburn, IL, USA; 2Thermo Fisher Scientifi c, Chelmsford, MA, USA

AbstractThe U.S. FDA guidelines on drug metabolites in safety testing (MIST) published in 2008 requires that the relative quantitation of human metabolites be obtained as soon as feasible in the drug development process. The goal is to ensure that at least one of the animal models used is not only producing the same metabolites found in humans, but at the same relative abundance. Identifi cation of drug metabolites is typically accomplished using various LC-MS approaches. However, metabolite quantifi cation is often diffi cult to accomplish with MS techniques alone. In some cases, the addition of UV detection is suffi cient for metabolite quantifi cation. But this requires that both parent drug and its metabolites possess similar and suffi ciently active chromophores, which is not always the case. Charged aerosol detection can detect any nonvolatile compound, typically with low ng sensitivity and similar response independent of chemical structure.

Presented here are the results from two test cases used to evaluate the usefulness of charged aerosol detection for MIST: buspirone (which has a strong UV chromophore) and erythromycin (which has a weak UV chromophore). These drugs were analyzed directly following incubation with human liver microsomes (drugs at ≤100 µM). The fi ve major metabolites for buspirone and the four major metabolites for erythromycin were identifi ed using an a Thermo Scientifi c LTQ Orbitrap mass spectrometer and quantifi ed by charged aerosol detection. The range, linearity, and sensitivity of this approach are discussed. Charged aerosol detection provides a complementary technique to those already being employed in industries capable of extending the range of in vitro drug metabolites that can be monitored during the drug development stage.

IntroductionInterest in metabolite and trace impurity analysis by the pharmaceutical industry is intensifying due to concerns with mass balance studies, regulatory commitments in reporting active pharmaceutical ingredient (API) impurities, MIST, and cleaning validation of manufacturing equipment. Most often an analytical requirement for accurately reporting the level of metabolites or impurities is to obtain reference standards or use a radiolabeled drug approach. Because many of these standards are unavailable and the radiochemical approaches are time consuming and expensive, quantifi cation of drug metabolites can often be diffi cult in early development stages. The situation is further exacerbated becase several types of HPLC detectors, such as UV or evaporative light scattering detection (ELSD), either do not provide uniform response across the target analytes or lack the sensitivity to detect these compounds. Although electrospray ionization (ESI) MS techniques are extremely sensitive and powerful approaches, ionization effi ciency can vary between metabolites and lead to quantifi cation issues.

The Thermo Scientifi c Dionex Corona™ ultra™ charged aerosol detector is mass sensitive and can be added to the traditional HPLC-UV or LC-MS platform. This detector provides the most consistent response for all nonvolatile and some semivolatile analytes of all HPLC detection techniques.1 The detector works by charging particles (see Figure 1) and is not dependent on light scattering which can vary between analytes. The work presented here examines the combination of UV and MS detection already employed in this fi eld, with charged aerosol detection for quantifi cation of drug metabolites in two specifi c cases. This work was completed in collaboration with Bristol-Myers Squibb (BMS), using a mix of HPLC equipment and method conditions.2 Although one analytical system and condition is emphasized, a variety of analytical equipment was used throughout the study.

All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries.

This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others.

FIGURE 1. Charged aerosol detector fl ow path schematic.

FIGURE 2. Schematic of one of the LC-charged aerosol dectector-MS systems used for analysis, with a total fl ow of 1 mL/min split ~5.7/1 to the charged aerosol dectector.

LPN 2953

FIGURE 3. Structure for buspirone and fi ve of its known metabolites.3

FIGURE 4. Chromatographic results for 60 µM buspirone before and after 120 min incubation in HLM.

FIGURE 5. Response as peak area for injection of equal concentrations of buspirone and the fi ve metabolites. Response deviation of each detection technique is calculated as percent RSD.

FIGURE 6. Linear response curve for buspirone with charged aerosol detection from ~4 to 124 ng on column.

FIGURE 7. Results for charged aerosol detection pre- and postincubation and UV at 210 nm preincubation the 100 µM erythromycin sample.

FIGURE 8. Structure (top) of erythromycin. Plot of relative area response of the four metabolites and the parent peak with charged aerosol detection detection (bottom).

DiscussionThe results discussed here where obtained in collaboration with BMS. The optimization of the system and development of the methods were ongoing throughout the work. Several components were identifi ed as crucial to maintaining quality data on both the charged aerosol detector and mass spectrometer. The column diameter and load volume was important, as peak splitting and shifting was observed with injection volumes greater than 30 µL. The confi guration of the fl ow splitting when running the LC-charged aerosol detector-MS system was determined to be the most crucial factor. The fl ow splitter must be positioned close to the mass spectrometer inlet using 0.004 inch or smaller i.d. tubing from the splitter to the mass spectrometer (Figure 2). The optimal fl ow-split range is estimated between 100 to 200 µL/min fl owing to the mass spectrometer, and the remaining fl ow to the charged aerosol detector. Because of the levels of sensitivity required for these analyses, a minimum fl ow rate of 500 µL/min is recommend for the charged aerosol detector.

The fi rst example, buspirone, was chosen because it is a well-characterized example, has a strong UV chromophore, and the standard material is readily available. The post HLM incubation of buspirone for both the charged aerosol detector and UV (Figure 4) show a similar metabolite profi le. The fi ve major metabolites shown in Figure 3 were quantifi ed as relative peak areas from the initial 60 µM buspirone sample. A sixth, minor metabolite was a volatile fragment of the buspirone and was not detected using the charged aerosol detector. The standards for the buspirone and the fi ve metabolites were prepared and analyzed at concentrations of 15 µM. The results for the two UV wavelengths (charged aerosol detection and MS) are shown in Figure 5. The greatest variability was found in the UV trace at 254 nm with an RSD of 53% followed by the MS at 37%. The charged aerosol detector and low wavelength UV had similar deviations of 20 and 17%, respectively.

The charged aerosol detector is typically linear from its limit of quantifi cation to ~500 ng on column, although not a linear detector over the full dynamic range. In Figure 6, the buspirone standard was analyzed using a UHPLC method from 3.9 to 124 ng on column. The six-point calibration curve exhibited excellent linearity over this range with a correlation coeffi cient of 0.9994. The concentration range of interest needed for the testing of drug metabolites typically falls in this mass-on-column range, which is linear for the charged aerosol detector independent of analyte.

MethodsSample PreparationThe buspirone and erythromycin standards were incubated at 1, 30, 60, and 100 μM substrate concentrations of human liver microsome (HLM). At the end of the incubation period, the samples were quenched with an equal part of acetonitrile. The fi nal analytical concentration of the metabolite samples discussed in this work is therefore one half of the concentration listed as the sample name (i.e., 60 µM pre-equals 30 µM postincubation).

Analytical Conditions Column: SB C18, 1.8 µm, 4.6 × 150 mmMobile Phases: A: 0.1% Formic acid in water, B: 0.1% Formic acid in acetonitrileFlow Rate: 1 mL/minTotal Run Time: 15 min

Detection Conditions Thermo Scientifi c Dionex UltiMate™ 3000 Diode Array Detector (Thermo Scientifi c Accela PDA detector used in some examples): UV Wavelength 1: 220 or 210 nm depending on example (values listed) UV Wavelength 2: 254 nm

Dionex Corona ultra:

Filter: High Gas: 35 psi Nitrogen Range: 100 pA full scale

LTQ Orbitrap high-resolution mass spectrometer Full scan MS with 1 × DDS

Flow Split

Valco three-way splitter Volumes fi xed by backpressure of different tubing diameters and lengths Volume measured at charged aerosol detector inlet.

ResultsBuspirone

Erythromycin

The second example, erythomycin, was chosen because it is representative of an array of drug candidates that fall into the nonvolatile analyte category with weak or no UV chromophores. As shown in Figure 7, the preincubation 100 µM standard has virtually no response with low wavelength UV. After incubation, the charged aerosol detector was able to detect the four major metabolites observed by MS. The relative area response for those four metabolites and the parent peak were evaluated at the 100 and 60 µM levels (Figure 8). As expected, the relative intensities of the metabolites increased at the lower concentration over the same incubation time.

The level of detection on the charged aerosol detector for the buspirone was determined to be ~1 ng on column with the short run UHPLC chemistry. This is typical for nonvolatile analytes under optimized conditions. Depending on molecular weight of the compound and the injection volume used, the molar concentration limit of detection can vary. While more sensitive than other nebulizer-based detection techniques, charged aerosol detection still may not provide suffi cient sensitivity to meet the detection levels needed for in vivo studies. The MetPro practice (advocated by J. Joseph in 2009), of normalizing the MS metabolite ratios at a single, higher-level point with UV data then transferring that to the MS results at the low levels has been employed in previous work.1 This was examined with the charged aerosol detector results during the calibration and is further described in Cai, et al, (2010).2 This offers a method to reduce the analytical deviations due to ionization variability, while keeping the sensitivity offered by the mass specrometer.

Conclusion The UHPLC-UV-charged aerosol detector-MS system described in this study was able to provide crucial information needed to meet requirements of the 2008 FDA MIST initiative. The use of charged aerosol detection as confi rmation of low-wavelength UV quantifi cation for chromophoric compounds can provide additional confi dence to the results. In areas where more specifi c wavelengths are used, such as the buspirone 254 nm, the charged aerosol detector can highlight large under- or overestimation as shown with metabolite.3 The area where this technology is most applicable is for parent or drug compounds that do not contain a strong UV chromophore. In these cases, laboratories are often left with no other option but to accept the MS quantifi cation during the early drug discovery stages. This can lead to major issues in later drug development stages. The charged aerosol detector—while not a stand-alone solution— can provide those additional pieces of data to make accurate interpretations of in vitro data without excessive cost or time requirements.

AcknowledgmentsThe authors would like to thank Hong Cai and colleagues at Bristoyl-Myers Squibb for their work on this collaboration. Generation of the data used would not have been possible without their help.

References1. Dependence of Response on Chemical Structure, ESA—A Dionex Company,

Application Brief 70-8913 Rev B.

2. Cai, et. al. A Novel Detection Technology: Charged Aerosol Detection (CAD) Coupled with HPLC, UV, and LTQ-Orbitrap MS for Semi-Quantitation of Metabolites in Drug Discovery Metabolism Studies, ASMS oral presentation, Salt Lake City, UT, May, 2010.

3. Zhu. et al. Drug Metab. Dispos. 2005, 33, 500–507.

101. Liquid eluent enters from the HPLC system2. Pneumatic nebulization occurs 3. Small droplets enter the drying tube4. Large drops exit to drain 5. Dried particles enter the mixing chamber6. Gas stream passes over corona needle7. Charged gas collides with particles and transfers charge 8. High mobility species are removed9. Remaining charged particles measured with a very sensitive electrometer10. Signal is transferred to chromatographic data software

1

2

4

5 6

7

8

9

26802

28120

0.850 µL

DualGradient

Pump

DAD

TCC

0.150 µL

ultra

Autosampler

28121

HO

HO

N

O

O N NN

NMet1

Met4

Met2

Met5

Met3O

N N

N

N

N

O

Chemical Formula: C21H31N5O2

Exact Mass: 385.24778

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

NO

O

OH

NNN

N

O

HN

O

O

N NN

N

NO

O N+

-ON

N

N

N

O

O

NNN

N

Buspirone

28122

-5

-5

20Buspirone 60 µM t = 0 by CAD

pA

-5

15Buspirone 60 µM t = 120 by CAD

pA

15Matrix Blank t = 120 by CAD

pA

-2

12Buspirone 60 µM t = 0 by UV at 254 nm

mAU

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5Minutes

11.0 11.5-1.5

8.0Buspirone 60 µM t = 120 by UV at 254 nm

mAU

28123

UV 220

0.E+00

1.E+05

2.E+05

3.E+05

4.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

UV 254

0.E+005.E+041.E+052.E+052.E+053.E+053.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 53%

CAD

00.5

11.5

22.5

33.5

BUSP Met1 Met2 Met3 Met4 Met5

PA/n

g

RSD 20%RSD 17%

HRMS

0.E+005.E+081.E+092.E+092.E+093.E+09

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 37%

28124

R2 = 0.9994

0

50,000

100,000

150,000

200,000

250,000

300,000

0 20 40 60 80 100 120 140

Mass on Column (ng)

Res

pons

e (p

eak

area

)

28125

-5

45Erythromycin 100 µM t = 0

pA

-5

45Erythromycin 100 µM t = 120

pA

7.88 8.5 9.0 9.5 10.0 10.5 11.025

100Erythromycin 100 µM t = 0

mAU

Minutes

28126

O

OCH3

H3C

OH

OH

CH3

HO

H3C

CH3O

CH3

H3C

O

O

OCH3

CH3

OHCH3

OO

HON

CH3

H3C

CH3

Chemical Formula: C37H67NO13

Exact Mass: 733.46124

Erythromycin

0

10

20

30

40

50

60

70

80

90

8.7 8.9 9.2 10.1 10.2Retention Time

Rel

ativ

e P

erce

nt A

rea

60 µM100 µM

+O

+O

+O

EYR

-CH2

-CH2

3Thermo Scientific Poster Note • LPN2953-01_e 11/11SV

Use of Charged Aerosol Detection as an Orthogonal Quantifi cation Technique for Drug Metabolites in Safety Testing (MIST) Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientifi c, Bannockburn, IL, USA; 2Thermo Fisher Scientifi c, Chelmsford, MA, USA

AbstractThe U.S. FDA guidelines on drug metabolites in safety testing (MIST) published in 2008 requires that the relative quantitation of human metabolites be obtained as soon as feasible in the drug development process. The goal is to ensure that at least one of the animal models used is not only producing the same metabolites found in humans, but at the same relative abundance. Identifi cation of drug metabolites is typically accomplished using various LC-MS approaches. However, metabolite quantifi cation is often diffi cult to accomplish with MS techniques alone. In some cases, the addition of UV detection is suffi cient for metabolite quantifi cation. But this requires that both parent drug and its metabolites possess similar and suffi ciently active chromophores, which is not always the case. Charged aerosol detection can detect any nonvolatile compound, typically with low ng sensitivity and similar response independent of chemical structure.

Presented here are the results from two test cases used to evaluate the usefulness of charged aerosol detection for MIST: buspirone (which has a strong UV chromophore) and erythromycin (which has a weak UV chromophore). These drugs were analyzed directly following incubation with human liver microsomes (drugs at ≤100 µM). The fi ve major metabolites for buspirone and the four major metabolites for erythromycin were identifi ed using an a Thermo Scientifi c LTQ Orbitrap mass spectrometer and quantifi ed by charged aerosol detection. The range, linearity, and sensitivity of this approach are discussed. Charged aerosol detection provides a complementary technique to those already being employed in industries capable of extending the range of in vitro drug metabolites that can be monitored during the drug development stage.

IntroductionInterest in metabolite and trace impurity analysis by the pharmaceutical industry is intensifying due to concerns with mass balance studies, regulatory commitments in reporting active pharmaceutical ingredient (API) impurities, MIST, and cleaning validation of manufacturing equipment. Most often an analytical requirement for accurately reporting the level of metabolites or impurities is to obtain reference standards or use a radiolabeled drug approach. Because many of these standards are unavailable and the radiochemical approaches are time consuming and expensive, quantifi cation of drug metabolites can often be diffi cult in early development stages. The situation is further exacerbated becase several types of HPLC detectors, such as UV or evaporative light scattering detection (ELSD), either do not provide uniform response across the target analytes or lack the sensitivity to detect these compounds. Although electrospray ionization (ESI) MS techniques are extremely sensitive and powerful approaches, ionization effi ciency can vary between metabolites and lead to quantifi cation issues.

The Thermo Scientifi c Dionex Corona™ ultra™ charged aerosol detector is mass sensitive and can be added to the traditional HPLC-UV or LC-MS platform. This detector provides the most consistent response for all nonvolatile and some semivolatile analytes of all HPLC detection techniques.1 The detector works by charging particles (see Figure 1) and is not dependent on light scattering which can vary between analytes. The work presented here examines the combination of UV and MS detection already employed in this fi eld, with charged aerosol detection for quantifi cation of drug metabolites in two specifi c cases. This work was completed in collaboration with Bristol-Myers Squibb (BMS), using a mix of HPLC equipment and method conditions.2 Although one analytical system and condition is emphasized, a variety of analytical equipment was used throughout the study.

All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries.

This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others.

FIGURE 1. Charged aerosol detector fl ow path schematic.

FIGURE 2. Schematic of one of the LC-charged aerosol dectector-MS systems used for analysis, with a total fl ow of 1 mL/min split ~5.7/1 to the charged aerosol dectector.

LPN 2953

FIGURE 3. Structure for buspirone and fi ve of its known metabolites.3

FIGURE 4. Chromatographic results for 60 µM buspirone before and after 120 min incubation in HLM.

FIGURE 5. Response as peak area for injection of equal concentrations of buspirone and the fi ve metabolites. Response deviation of each detection technique is calculated as percent RSD.

FIGURE 6. Linear response curve for buspirone with charged aerosol detection from ~4 to 124 ng on column.

FIGURE 7. Results for charged aerosol detection pre- and postincubation and UV at 210 nm preincubation the 100 µM erythromycin sample.

FIGURE 8. Structure (top) of erythromycin. Plot of relative area response of the four metabolites and the parent peak with charged aerosol detection detection (bottom).

DiscussionThe results discussed here where obtained in collaboration with BMS. The optimization of the system and development of the methods were ongoing throughout the work. Several components were identifi ed as crucial to maintaining quality data on both the charged aerosol detector and mass spectrometer. The column diameter and load volume was important, as peak splitting and shifting was observed with injection volumes greater than 30 µL. The confi guration of the fl ow splitting when running the LC-charged aerosol detector-MS system was determined to be the most crucial factor. The fl ow splitter must be positioned close to the mass spectrometer inlet using 0.004 inch or smaller i.d. tubing from the splitter to the mass spectrometer (Figure 2). The optimal fl ow-split range is estimated between 100 to 200 µL/min fl owing to the mass spectrometer, and the remaining fl ow to the charged aerosol detector. Because of the levels of sensitivity required for these analyses, a minimum fl ow rate of 500 µL/min is recommend for the charged aerosol detector.

The fi rst example, buspirone, was chosen because it is a well-characterized example, has a strong UV chromophore, and the standard material is readily available. The post HLM incubation of buspirone for both the charged aerosol detector and UV (Figure 4) show a similar metabolite profi le. The fi ve major metabolites shown in Figure 3 were quantifi ed as relative peak areas from the initial 60 µM buspirone sample. A sixth, minor metabolite was a volatile fragment of the buspirone and was not detected using the charged aerosol detector. The standards for the buspirone and the fi ve metabolites were prepared and analyzed at concentrations of 15 µM. The results for the two UV wavelengths (charged aerosol detection and MS) are shown in Figure 5. The greatest variability was found in the UV trace at 254 nm with an RSD of 53% followed by the MS at 37%. The charged aerosol detector and low wavelength UV had similar deviations of 20 and 17%, respectively.

The charged aerosol detector is typically linear from its limit of quantifi cation to ~500 ng on column, although not a linear detector over the full dynamic range. In Figure 6, the buspirone standard was analyzed using a UHPLC method from 3.9 to 124 ng on column. The six-point calibration curve exhibited excellent linearity over this range with a correlation coeffi cient of 0.9994. The concentration range of interest needed for the testing of drug metabolites typically falls in this mass-on-column range, which is linear for the charged aerosol detector independent of analyte.

MethodsSample PreparationThe buspirone and erythromycin standards were incubated at 1, 30, 60, and 100 μM substrate concentrations of human liver microsome (HLM). At the end of the incubation period, the samples were quenched with an equal part of acetonitrile. The fi nal analytical concentration of the metabolite samples discussed in this work is therefore one half of the concentration listed as the sample name (i.e., 60 µM pre-equals 30 µM postincubation).

Analytical Conditions Column: SB C18, 1.8 µm, 4.6 × 150 mmMobile Phases: A: 0.1% Formic acid in water, B: 0.1% Formic acid in acetonitrileFlow Rate: 1 mL/minTotal Run Time: 15 min

Detection Conditions Thermo Scientifi c Dionex UltiMate™ 3000 Diode Array Detector (Thermo Scientifi c Accela PDA detector used in some examples): UV Wavelength 1: 220 or 210 nm depending on example (values listed) UV Wavelength 2: 254 nm

Dionex Corona ultra:

Filter: High Gas: 35 psi Nitrogen Range: 100 pA full scale

LTQ Orbitrap high-resolution mass spectrometer Full scan MS with 1 × DDS

Flow Split

Valco three-way splitter Volumes fi xed by backpressure of different tubing diameters and lengths Volume measured at charged aerosol detector inlet.

ResultsBuspirone

Erythromycin

The second example, erythomycin, was chosen because it is representative of an array of drug candidates that fall into the nonvolatile analyte category with weak or no UV chromophores. As shown in Figure 7, the preincubation 100 µM standard has virtually no response with low wavelength UV. After incubation, the charged aerosol detector was able to detect the four major metabolites observed by MS. The relative area response for those four metabolites and the parent peak were evaluated at the 100 and 60 µM levels (Figure 8). As expected, the relative intensities of the metabolites increased at the lower concentration over the same incubation time.

The level of detection on the charged aerosol detector for the buspirone was determined to be ~1 ng on column with the short run UHPLC chemistry. This is typical for nonvolatile analytes under optimized conditions. Depending on molecular weight of the compound and the injection volume used, the molar concentration limit of detection can vary. While more sensitive than other nebulizer-based detection techniques, charged aerosol detection still may not provide suffi cient sensitivity to meet the detection levels needed for in vivo studies. The MetPro practice (advocated by J. Joseph in 2009), of normalizing the MS metabolite ratios at a single, higher-level point with UV data then transferring that to the MS results at the low levels has been employed in previous work.1 This was examined with the charged aerosol detector results during the calibration and is further described in Cai, et al, (2010).2 This offers a method to reduce the analytical deviations due to ionization variability, while keeping the sensitivity offered by the mass specrometer.

Conclusion The UHPLC-UV-charged aerosol detector-MS system described in this study was able to provide crucial information needed to meet requirements of the 2008 FDA MIST initiative. The use of charged aerosol detection as confi rmation of low-wavelength UV quantifi cation for chromophoric compounds can provide additional confi dence to the results. In areas where more specifi c wavelengths are used, such as the buspirone 254 nm, the charged aerosol detector can highlight large under- or overestimation as shown with metabolite.3 The area where this technology is most applicable is for parent or drug compounds that do not contain a strong UV chromophore. In these cases, laboratories are often left with no other option but to accept the MS quantifi cation during the early drug discovery stages. This can lead to major issues in later drug development stages. The charged aerosol detector—while not a stand-alone solution— can provide those additional pieces of data to make accurate interpretations of in vitro data without excessive cost or time requirements.

AcknowledgmentsThe authors would like to thank Hong Cai and colleagues at Bristoyl-Myers Squibb for their work on this collaboration. Generation of the data used would not have been possible without their help.

References1. Dependence of Response on Chemical Structure, ESA—A Dionex Company,

Application Brief 70-8913 Rev B.

2. Cai, et. al. A Novel Detection Technology: Charged Aerosol Detection (CAD) Coupled with HPLC, UV, and LTQ-Orbitrap MS for Semi-Quantitation of Metabolites in Drug Discovery Metabolism Studies, ASMS oral presentation, Salt Lake City, UT, May, 2010.

3. Zhu. et al. Drug Metab. Dispos. 2005, 33, 500–507.

101. Liquid eluent enters from the HPLC system2. Pneumatic nebulization occurs 3. Small droplets enter the drying tube4. Large drops exit to drain 5. Dried particles enter the mixing chamber6. Gas stream passes over corona needle7. Charged gas collides with particles and transfers charge 8. High mobility species are removed9. Remaining charged particles measured with a very sensitive electrometer10. Signal is transferred to chromatographic data software

1

2

4

5 6

7

8

9

26802

28120

0.850 µL

DualGradient

Pump

DAD

TCC

0.150 µL

ultra

Autosampler

28121

HO

HO

N

O

O N NN

NMet1

Met4

Met2

Met5

Met3O

N N

N

N

N

O

Chemical Formula: C21H31N5O2

Exact Mass: 385.24778

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

NO

O

OH

NNN

N

O

HN

O

O

N NN

N

NO

O N+

-ON

N

N

N

O

O

NNN

N

Buspirone

28122

-5

-5

20Buspirone 60 µM t = 0 by CAD

pA

-5

15Buspirone 60 µM t = 120 by CAD

pA

15Matrix Blank t = 120 by CAD

pA

-2

12Buspirone 60 µM t = 0 by UV at 254 nm

mAU

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5Minutes

11.0 11.5-1.5

8.0Buspirone 60 µM t = 120 by UV at 254 nm

mAU

28123

UV 220

0.E+00

1.E+05

2.E+05

3.E+05

4.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

UV 254

0.E+005.E+041.E+052.E+052.E+053.E+053.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 53%

CAD

00.5

11.5

22.5

33.5

BUSP Met1 Met2 Met3 Met4 Met5

PA/n

g

RSD 20%RSD 17%

HRMS

0.E+005.E+081.E+092.E+092.E+093.E+09

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 37%

28124

R2 = 0.9994

0

50,000

100,000

150,000

200,000

250,000

300,000

0 20 40 60 80 100 120 140

Mass on Column (ng)

Res

pons

e (p

eak

area

)

28125

-5

45Erythromycin 100 µM t = 0

pA

-5

45Erythromycin 100 µM t = 120

pA

7.88 8.5 9.0 9.5 10.0 10.5 11.025

100Erythromycin 100 µM t = 0

mAU

Minutes

28126

O

OCH3

H3C

OH

OH

CH3

HO

H3C

CH3O

CH3

H3C

O

O

OCH3

CH3

OHCH3

OO

HON

CH3

H3C

CH3

Chemical Formula: C37H67NO13

Exact Mass: 733.46124

Erythromycin

0

10

20

30

40

50

60

70

80

90

8.7 8.9 9.2 10.1 10.2Retention Time

Rel

ativ

e P

erce

nt A

rea

60 µM100 µM

+O

+O

+O

EYR

-CH2

-CH2

4 Use of Charged Aerosol Detection as an Orthogonal Quantification Technique for Drug Metaboites in Safety Testing (MIST)

Use of Charged Aerosol Detection as an Orthogonal Quantifi cation Technique for Drug Metabolites in Safety Testing (MIST) Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientifi c, Bannockburn, IL, USA; 2Thermo Fisher Scientifi c, Chelmsford, MA, USA

AbstractThe U.S. FDA guidelines on drug metabolites in safety testing (MIST) published in 2008 requires that the relative quantitation of human metabolites be obtained as soon as feasible in the drug development process. The goal is to ensure that at least one of the animal models used is not only producing the same metabolites found in humans, but at the same relative abundance. Identifi cation of drug metabolites is typically accomplished using various LC-MS approaches. However, metabolite quantifi cation is often diffi cult to accomplish with MS techniques alone. In some cases, the addition of UV detection is suffi cient for metabolite quantifi cation. But this requires that both parent drug and its metabolites possess similar and suffi ciently active chromophores, which is not always the case. Charged aerosol detection can detect any nonvolatile compound, typically with low ng sensitivity and similar response independent of chemical structure.

Presented here are the results from two test cases used to evaluate the usefulness of charged aerosol detection for MIST: buspirone (which has a strong UV chromophore) and erythromycin (which has a weak UV chromophore). These drugs were analyzed directly following incubation with human liver microsomes (drugs at ≤100 µM). The fi ve major metabolites for buspirone and the four major metabolites for erythromycin were identifi ed using an a Thermo Scientifi c LTQ Orbitrap mass spectrometer and quantifi ed by charged aerosol detection. The range, linearity, and sensitivity of this approach are discussed. Charged aerosol detection provides a complementary technique to those already being employed in industries capable of extending the range of in vitro drug metabolites that can be monitored during the drug development stage.

IntroductionInterest in metabolite and trace impurity analysis by the pharmaceutical industry is intensifying due to concerns with mass balance studies, regulatory commitments in reporting active pharmaceutical ingredient (API) impurities, MIST, and cleaning validation of manufacturing equipment. Most often an analytical requirement for accurately reporting the level of metabolites or impurities is to obtain reference standards or use a radiolabeled drug approach. Because many of these standards are unavailable and the radiochemical approaches are time consuming and expensive, quantifi cation of drug metabolites can often be diffi cult in early development stages. The situation is further exacerbated becase several types of HPLC detectors, such as UV or evaporative light scattering detection (ELSD), either do not provide uniform response across the target analytes or lack the sensitivity to detect these compounds. Although electrospray ionization (ESI) MS techniques are extremely sensitive and powerful approaches, ionization effi ciency can vary between metabolites and lead to quantifi cation issues.

The Thermo Scientifi c Dionex Corona™ ultra™ charged aerosol detector is mass sensitive and can be added to the traditional HPLC-UV or LC-MS platform. This detector provides the most consistent response for all nonvolatile and some semivolatile analytes of all HPLC detection techniques.1 The detector works by charging particles (see Figure 1) and is not dependent on light scattering which can vary between analytes. The work presented here examines the combination of UV and MS detection already employed in this fi eld, with charged aerosol detection for quantifi cation of drug metabolites in two specifi c cases. This work was completed in collaboration with Bristol-Myers Squibb (BMS), using a mix of HPLC equipment and method conditions.2 Although one analytical system and condition is emphasized, a variety of analytical equipment was used throughout the study.

All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries.

This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others.

FIGURE 1. Charged aerosol detector fl ow path schematic.

FIGURE 2. Schematic of one of the LC-charged aerosol dectector-MS systems used for analysis, with a total fl ow of 1 mL/min split ~5.7/1 to the charged aerosol dectector.

LPN 2953

FIGURE 3. Structure for buspirone and fi ve of its known metabolites.3

FIGURE 4. Chromatographic results for 60 µM buspirone before and after 120 min incubation in HLM.

FIGURE 5. Response as peak area for injection of equal concentrations of buspirone and the fi ve metabolites. Response deviation of each detection technique is calculated as percent RSD.

FIGURE 6. Linear response curve for buspirone with charged aerosol detection from ~4 to 124 ng on column.

FIGURE 7. Results for charged aerosol detection pre- and postincubation and UV at 210 nm preincubation the 100 µM erythromycin sample.

FIGURE 8. Structure (top) of erythromycin. Plot of relative area response of the four metabolites and the parent peak with charged aerosol detection detection (bottom).

DiscussionThe results discussed here where obtained in collaboration with BMS. The optimization of the system and development of the methods were ongoing throughout the work. Several components were identifi ed as crucial to maintaining quality data on both the charged aerosol detector and mass spectrometer. The column diameter and load volume was important, as peak splitting and shifting was observed with injection volumes greater than 30 µL. The confi guration of the fl ow splitting when running the LC-charged aerosol detector-MS system was determined to be the most crucial factor. The fl ow splitter must be positioned close to the mass spectrometer inlet using 0.004 inch or smaller i.d. tubing from the splitter to the mass spectrometer (Figure 2). The optimal fl ow-split range is estimated between 100 to 200 µL/min fl owing to the mass spectrometer, and the remaining fl ow to the charged aerosol detector. Because of the levels of sensitivity required for these analyses, a minimum fl ow rate of 500 µL/min is recommend for the charged aerosol detector.

The fi rst example, buspirone, was chosen because it is a well-characterized example, has a strong UV chromophore, and the standard material is readily available. The post HLM incubation of buspirone for both the charged aerosol detector and UV (Figure 4) show a similar metabolite profi le. The fi ve major metabolites shown in Figure 3 were quantifi ed as relative peak areas from the initial 60 µM buspirone sample. A sixth, minor metabolite was a volatile fragment of the buspirone and was not detected using the charged aerosol detector. The standards for the buspirone and the fi ve metabolites were prepared and analyzed at concentrations of 15 µM. The results for the two UV wavelengths (charged aerosol detection and MS) are shown in Figure 5. The greatest variability was found in the UV trace at 254 nm with an RSD of 53% followed by the MS at 37%. The charged aerosol detector and low wavelength UV had similar deviations of 20 and 17%, respectively.

The charged aerosol detector is typically linear from its limit of quantifi cation to ~500 ng on column, although not a linear detector over the full dynamic range. In Figure 6, the buspirone standard was analyzed using a UHPLC method from 3.9 to 124 ng on column. The six-point calibration curve exhibited excellent linearity over this range with a correlation coeffi cient of 0.9994. The concentration range of interest needed for the testing of drug metabolites typically falls in this mass-on-column range, which is linear for the charged aerosol detector independent of analyte.

MethodsSample PreparationThe buspirone and erythromycin standards were incubated at 1, 30, 60, and 100 μM substrate concentrations of human liver microsome (HLM). At the end of the incubation period, the samples were quenched with an equal part of acetonitrile. The fi nal analytical concentration of the metabolite samples discussed in this work is therefore one half of the concentration listed as the sample name (i.e., 60 µM pre-equals 30 µM postincubation).

Analytical Conditions Column: SB C18, 1.8 µm, 4.6 × 150 mmMobile Phases: A: 0.1% Formic acid in water, B: 0.1% Formic acid in acetonitrileFlow Rate: 1 mL/minTotal Run Time: 15 min

Detection Conditions Thermo Scientifi c Dionex UltiMate™ 3000 Diode Array Detector (Thermo Scientifi c Accela PDA detector used in some examples): UV Wavelength 1: 220 or 210 nm depending on example (values listed) UV Wavelength 2: 254 nm

Dionex Corona ultra:

Filter: High Gas: 35 psi Nitrogen Range: 100 pA full scale

LTQ Orbitrap high-resolution mass spectrometer Full scan MS with 1 × DDS

Flow Split

Valco three-way splitter Volumes fi xed by backpressure of different tubing diameters and lengths Volume measured at charged aerosol detector inlet.

ResultsBuspirone

Erythromycin

The second example, erythomycin, was chosen because it is representative of an array of drug candidates that fall into the nonvolatile analyte category with weak or no UV chromophores. As shown in Figure 7, the preincubation 100 µM standard has virtually no response with low wavelength UV. After incubation, the charged aerosol detector was able to detect the four major metabolites observed by MS. The relative area response for those four metabolites and the parent peak were evaluated at the 100 and 60 µM levels (Figure 8). As expected, the relative intensities of the metabolites increased at the lower concentration over the same incubation time.

The level of detection on the charged aerosol detector for the buspirone was determined to be ~1 ng on column with the short run UHPLC chemistry. This is typical for nonvolatile analytes under optimized conditions. Depending on molecular weight of the compound and the injection volume used, the molar concentration limit of detection can vary. While more sensitive than other nebulizer-based detection techniques, charged aerosol detection still may not provide suffi cient sensitivity to meet the detection levels needed for in vivo studies. The MetPro practice (advocated by J. Joseph in 2009), of normalizing the MS metabolite ratios at a single, higher-level point with UV data then transferring that to the MS results at the low levels has been employed in previous work.1 This was examined with the charged aerosol detector results during the calibration and is further described in Cai, et al, (2010).2 This offers a method to reduce the analytical deviations due to ionization variability, while keeping the sensitivity offered by the mass specrometer.

Conclusion The UHPLC-UV-charged aerosol detector-MS system described in this study was able to provide crucial information needed to meet requirements of the 2008 FDA MIST initiative. The use of charged aerosol detection as confi rmation of low-wavelength UV quantifi cation for chromophoric compounds can provide additional confi dence to the results. In areas where more specifi c wavelengths are used, such as the buspirone 254 nm, the charged aerosol detector can highlight large under- or overestimation as shown with metabolite.3 The area where this technology is most applicable is for parent or drug compounds that do not contain a strong UV chromophore. In these cases, laboratories are often left with no other option but to accept the MS quantifi cation during the early drug discovery stages. This can lead to major issues in later drug development stages. The charged aerosol detector—while not a stand-alone solution— can provide those additional pieces of data to make accurate interpretations of in vitro data without excessive cost or time requirements.

AcknowledgmentsThe authors would like to thank Hong Cai and colleagues at Bristoyl-Myers Squibb for their work on this collaboration. Generation of the data used would not have been possible without their help.

References1. Dependence of Response on Chemical Structure, ESA—A Dionex Company,

Application Brief 70-8913 Rev B.

2. Cai, et. al. A Novel Detection Technology: Charged Aerosol Detection (CAD) Coupled with HPLC, UV, and LTQ-Orbitrap MS for Semi-Quantitation of Metabolites in Drug Discovery Metabolism Studies, ASMS oral presentation, Salt Lake City, UT, May, 2010.

3. Zhu. et al. Drug Metab. Dispos. 2005, 33, 500–507.

101. Liquid eluent enters from the HPLC system2. Pneumatic nebulization occurs 3. Small droplets enter the drying tube4. Large drops exit to drain 5. Dried particles enter the mixing chamber6. Gas stream passes over corona needle7. Charged gas collides with particles and transfers charge 8. High mobility species are removed9. Remaining charged particles measured with a very sensitive electrometer10. Signal is transferred to chromatographic data software

1

2

4

5 6

7

8

9

26802

28120

0.850 µL

DualGradient

Pump

DAD

TCC

0.150 µL

ultra

Autosampler

28121

HO

HO

N

O

O N NN

NMet1

Met4

Met2

Met5

Met3O

N N

N

N

N

O

Chemical Formula: C21H31N5O2

Exact Mass: 385.24778

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

NO

O

OH

NNN

N

O

HN

O

O

N NN

N

NO

O N+

-ON

N

N

N

O

O

NNN

N

Buspirone

28122

-5

-5

20Buspirone 60 µM t = 0 by CAD

pA

-5

15Buspirone 60 µM t = 120 by CAD

pA

15Matrix Blank t = 120 by CAD

pA

-2

12Buspirone 60 µM t = 0 by UV at 254 nm

mAU

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5Minutes

11.0 11.5-1.5

8.0Buspirone 60 µM t = 120 by UV at 254 nm

mAU

28123

UV 220

0.E+00

1.E+05

2.E+05

3.E+05

4.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

UV 254

0.E+005.E+041.E+052.E+052.E+053.E+053.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 53%

CAD

00.5

11.5

22.5

33.5

BUSP Met1 Met2 Met3 Met4 Met5

PA/n

g

RSD 20%RSD 17%

HRMS

0.E+005.E+081.E+092.E+092.E+093.E+09

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 37%

28124

R2 = 0.9994

0

50,000

100,000

150,000

200,000

250,000

300,000

0 20 40 60 80 100 120 140

Mass on Column (ng)

Res

pons

e (p

eak

area

)

28125

-5

45Erythromycin 100 µM t = 0

pA

-5

45Erythromycin 100 µM t = 120

pA

7.88 8.5 9.0 9.5 10.0 10.5 11.025

100Erythromycin 100 µM t = 0

mAU

Minutes

28126

O

OCH3

H3C

OH

OH

CH3

HO

H3C

CH3O

CH3

H3C

O

O

OCH3

CH3

OHCH3

OO

HON

CH3

H3C

CH3

Chemical Formula: C37H67NO13

Exact Mass: 733.46124

Erythromycin

0

10

20

30

40

50

60

70

80

90

8.7 8.9 9.2 10.1 10.2Retention Time

Rel

ativ

e P

erce

nt A

rea

60 µM100 µM

+O

+O

+O

EYR

-CH2

-CH2

5Thermo Scientific Poster Note • LPN2953-01_e 11/11SV

Use of Charged Aerosol Detection as an Orthogonal Quantifi cation Technique for Drug Metabolites in Safety Testing (MIST) Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientifi c, Bannockburn, IL, USA; 2Thermo Fisher Scientifi c, Chelmsford, MA, USA

AbstractThe U.S. FDA guidelines on drug metabolites in safety testing (MIST) published in 2008 requires that the relative quantitation of human metabolites be obtained as soon as feasible in the drug development process. The goal is to ensure that at least one of the animal models used is not only producing the same metabolites found in humans, but at the same relative abundance. Identifi cation of drug metabolites is typically accomplished using various LC-MS approaches. However, metabolite quantifi cation is often diffi cult to accomplish with MS techniques alone. In some cases, the addition of UV detection is suffi cient for metabolite quantifi cation. But this requires that both parent drug and its metabolites possess similar and suffi ciently active chromophores, which is not always the case. Charged aerosol detection can detect any nonvolatile compound, typically with low ng sensitivity and similar response independent of chemical structure.

Presented here are the results from two test cases used to evaluate the usefulness of charged aerosol detection for MIST: buspirone (which has a strong UV chromophore) and erythromycin (which has a weak UV chromophore). These drugs were analyzed directly following incubation with human liver microsomes (drugs at ≤100 µM). The fi ve major metabolites for buspirone and the four major metabolites for erythromycin were identifi ed using an a Thermo Scientifi c LTQ Orbitrap mass spectrometer and quantifi ed by charged aerosol detection. The range, linearity, and sensitivity of this approach are discussed. Charged aerosol detection provides a complementary technique to those already being employed in industries capable of extending the range of in vitro drug metabolites that can be monitored during the drug development stage.

IntroductionInterest in metabolite and trace impurity analysis by the pharmaceutical industry is intensifying due to concerns with mass balance studies, regulatory commitments in reporting active pharmaceutical ingredient (API) impurities, MIST, and cleaning validation of manufacturing equipment. Most often an analytical requirement for accurately reporting the level of metabolites or impurities is to obtain reference standards or use a radiolabeled drug approach. Because many of these standards are unavailable and the radiochemical approaches are time consuming and expensive, quantifi cation of drug metabolites can often be diffi cult in early development stages. The situation is further exacerbated becase several types of HPLC detectors, such as UV or evaporative light scattering detection (ELSD), either do not provide uniform response across the target analytes or lack the sensitivity to detect these compounds. Although electrospray ionization (ESI) MS techniques are extremely sensitive and powerful approaches, ionization effi ciency can vary between metabolites and lead to quantifi cation issues.

The Thermo Scientifi c Dionex Corona™ ultra™ charged aerosol detector is mass sensitive and can be added to the traditional HPLC-UV or LC-MS platform. This detector provides the most consistent response for all nonvolatile and some semivolatile analytes of all HPLC detection techniques.1 The detector works by charging particles (see Figure 1) and is not dependent on light scattering which can vary between analytes. The work presented here examines the combination of UV and MS detection already employed in this fi eld, with charged aerosol detection for quantifi cation of drug metabolites in two specifi c cases. This work was completed in collaboration with Bristol-Myers Squibb (BMS), using a mix of HPLC equipment and method conditions.2 Although one analytical system and condition is emphasized, a variety of analytical equipment was used throughout the study.

All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries.

This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others.

FIGURE 1. Charged aerosol detector fl ow path schematic.

FIGURE 2. Schematic of one of the LC-charged aerosol dectector-MS systems used for analysis, with a total fl ow of 1 mL/min split ~5.7/1 to the charged aerosol dectector.

LPN 2953

FIGURE 3. Structure for buspirone and fi ve of its known metabolites.3

FIGURE 4. Chromatographic results for 60 µM buspirone before and after 120 min incubation in HLM.

FIGURE 5. Response as peak area for injection of equal concentrations of buspirone and the fi ve metabolites. Response deviation of each detection technique is calculated as percent RSD.

FIGURE 6. Linear response curve for buspirone with charged aerosol detection from ~4 to 124 ng on column.

FIGURE 7. Results for charged aerosol detection pre- and postincubation and UV at 210 nm preincubation the 100 µM erythromycin sample.

FIGURE 8. Structure (top) of erythromycin. Plot of relative area response of the four metabolites and the parent peak with charged aerosol detection detection (bottom).

DiscussionThe results discussed here where obtained in collaboration with BMS. The optimization of the system and development of the methods were ongoing throughout the work. Several components were identifi ed as crucial to maintaining quality data on both the charged aerosol detector and mass spectrometer. The column diameter and load volume was important, as peak splitting and shifting was observed with injection volumes greater than 30 µL. The confi guration of the fl ow splitting when running the LC-charged aerosol detector-MS system was determined to be the most crucial factor. The fl ow splitter must be positioned close to the mass spectrometer inlet using 0.004 inch or smaller i.d. tubing from the splitter to the mass spectrometer (Figure 2). The optimal fl ow-split range is estimated between 100 to 200 µL/min fl owing to the mass spectrometer, and the remaining fl ow to the charged aerosol detector. Because of the levels of sensitivity required for these analyses, a minimum fl ow rate of 500 µL/min is recommend for the charged aerosol detector.

The fi rst example, buspirone, was chosen because it is a well-characterized example, has a strong UV chromophore, and the standard material is readily available. The post HLM incubation of buspirone for both the charged aerosol detector and UV (Figure 4) show a similar metabolite profi le. The fi ve major metabolites shown in Figure 3 were quantifi ed as relative peak areas from the initial 60 µM buspirone sample. A sixth, minor metabolite was a volatile fragment of the buspirone and was not detected using the charged aerosol detector. The standards for the buspirone and the fi ve metabolites were prepared and analyzed at concentrations of 15 µM. The results for the two UV wavelengths (charged aerosol detection and MS) are shown in Figure 5. The greatest variability was found in the UV trace at 254 nm with an RSD of 53% followed by the MS at 37%. The charged aerosol detector and low wavelength UV had similar deviations of 20 and 17%, respectively.

The charged aerosol detector is typically linear from its limit of quantifi cation to ~500 ng on column, although not a linear detector over the full dynamic range. In Figure 6, the buspirone standard was analyzed using a UHPLC method from 3.9 to 124 ng on column. The six-point calibration curve exhibited excellent linearity over this range with a correlation coeffi cient of 0.9994. The concentration range of interest needed for the testing of drug metabolites typically falls in this mass-on-column range, which is linear for the charged aerosol detector independent of analyte.

MethodsSample PreparationThe buspirone and erythromycin standards were incubated at 1, 30, 60, and 100 μM substrate concentrations of human liver microsome (HLM). At the end of the incubation period, the samples were quenched with an equal part of acetonitrile. The fi nal analytical concentration of the metabolite samples discussed in this work is therefore one half of the concentration listed as the sample name (i.e., 60 µM pre-equals 30 µM postincubation).

Analytical Conditions Column: SB C18, 1.8 µm, 4.6 × 150 mmMobile Phases: A: 0.1% Formic acid in water, B: 0.1% Formic acid in acetonitrileFlow Rate: 1 mL/minTotal Run Time: 15 min

Detection Conditions Thermo Scientifi c Dionex UltiMate™ 3000 Diode Array Detector (Thermo Scientifi c Accela PDA detector used in some examples): UV Wavelength 1: 220 or 210 nm depending on example (values listed) UV Wavelength 2: 254 nm

Dionex Corona ultra:

Filter: High Gas: 35 psi Nitrogen Range: 100 pA full scale

LTQ Orbitrap high-resolution mass spectrometer Full scan MS with 1 × DDS

Flow Split

Valco three-way splitter Volumes fi xed by backpressure of different tubing diameters and lengths Volume measured at charged aerosol detector inlet.

ResultsBuspirone

Erythromycin

The second example, erythomycin, was chosen because it is representative of an array of drug candidates that fall into the nonvolatile analyte category with weak or no UV chromophores. As shown in Figure 7, the preincubation 100 µM standard has virtually no response with low wavelength UV. After incubation, the charged aerosol detector was able to detect the four major metabolites observed by MS. The relative area response for those four metabolites and the parent peak were evaluated at the 100 and 60 µM levels (Figure 8). As expected, the relative intensities of the metabolites increased at the lower concentration over the same incubation time.

The level of detection on the charged aerosol detector for the buspirone was determined to be ~1 ng on column with the short run UHPLC chemistry. This is typical for nonvolatile analytes under optimized conditions. Depending on molecular weight of the compound and the injection volume used, the molar concentration limit of detection can vary. While more sensitive than other nebulizer-based detection techniques, charged aerosol detection still may not provide suffi cient sensitivity to meet the detection levels needed for in vivo studies. The MetPro practice (advocated by J. Joseph in 2009), of normalizing the MS metabolite ratios at a single, higher-level point with UV data then transferring that to the MS results at the low levels has been employed in previous work.1 This was examined with the charged aerosol detector results during the calibration and is further described in Cai, et al, (2010).2 This offers a method to reduce the analytical deviations due to ionization variability, while keeping the sensitivity offered by the mass specrometer.

Conclusion The UHPLC-UV-charged aerosol detector-MS system described in this study was able to provide crucial information needed to meet requirements of the 2008 FDA MIST initiative. The use of charged aerosol detection as confi rmation of low-wavelength UV quantifi cation for chromophoric compounds can provide additional confi dence to the results. In areas where more specifi c wavelengths are used, such as the buspirone 254 nm, the charged aerosol detector can highlight large under- or overestimation as shown with metabolite.3 The area where this technology is most applicable is for parent or drug compounds that do not contain a strong UV chromophore. In these cases, laboratories are often left with no other option but to accept the MS quantifi cation during the early drug discovery stages. This can lead to major issues in later drug development stages. The charged aerosol detector—while not a stand-alone solution— can provide those additional pieces of data to make accurate interpretations of in vitro data without excessive cost or time requirements.

AcknowledgmentsThe authors would like to thank Hong Cai and colleagues at Bristoyl-Myers Squibb for their work on this collaboration. Generation of the data used would not have been possible without their help.

References1. Dependence of Response on Chemical Structure, ESA—A Dionex Company,

Application Brief 70-8913 Rev B.

2. Cai, et. al. A Novel Detection Technology: Charged Aerosol Detection (CAD) Coupled with HPLC, UV, and LTQ-Orbitrap MS for Semi-Quantitation of Metabolites in Drug Discovery Metabolism Studies, ASMS oral presentation, Salt Lake City, UT, May, 2010.

3. Zhu. et al. Drug Metab. Dispos. 2005, 33, 500–507.

101. Liquid eluent enters from the HPLC system2. Pneumatic nebulization occurs 3. Small droplets enter the drying tube4. Large drops exit to drain 5. Dried particles enter the mixing chamber6. Gas stream passes over corona needle7. Charged gas collides with particles and transfers charge 8. High mobility species are removed9. Remaining charged particles measured with a very sensitive electrometer10. Signal is transferred to chromatographic data software

1

2

4

5 6

7

8

9

26802

28120

0.850 µL

DualGradient

Pump

DAD

TCC

0.150 µL

ultra

Autosampler

28121

HO

HO

N

O

O N NN

NMet1

Met4

Met2

Met5

Met3O

N N

N

N

N

O

Chemical Formula: C21H31N5O2

Exact Mass: 385.24778

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

NO

O

OH

NNN

N

O

HN

O

O

N NN

N

NO

O N+

-ON

N

N

N

O

O

NNN

N

Buspirone

28122

-5

-5

20Buspirone 60 µM t = 0 by CAD

pA

-5

15Buspirone 60 µM t = 120 by CAD

pA

15Matrix Blank t = 120 by CAD

pA

-2

12Buspirone 60 µM t = 0 by UV at 254 nm

mAU

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5Minutes

11.0 11.5-1.5

8.0Buspirone 60 µM t = 120 by UV at 254 nm

mAU

28123

UV 220

0.E+00

1.E+05

2.E+05

3.E+05

4.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

UV 254

0.E+005.E+041.E+052.E+052.E+053.E+053.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PARSD 53%

CAD

00.5

11.5

22.5

33.5

BUSP Met1 Met2 Met3 Met4 Met5

PA/n

g

RSD 20%RSD 17%

HRMS

0.E+005.E+081.E+092.E+092.E+093.E+09

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 37%

28124

R2 = 0.9994

0

50,000

100,000

150,000

200,000

250,000

300,000

0 20 40 60 80 100 120 140

Mass on Column (ng)

Res

pons

e (p

eak

area

)

28125

-5

45Erythromycin 100 µM t = 0

pA

-5

45Erythromycin 100 µM t = 120

pA

7.88 8.5 9.0 9.5 10.0 10.5 11.025

100Erythromycin 100 µM t = 0

mAU

Minutes

28126

O

OCH3

H3C

OH

OH

CH3

HO

H3C

CH3O

CH3

H3C

O

O

OCH3

CH3

OHCH3

OO

HON

CH3

H3C

CH3

Chemical Formula: C37H67NO13

Exact Mass: 733.46124

Erythromycin

0

10

20

30

40

50

60

70

80

90

8.7 8.9 9.2 10.1 10.2Retention Time

Rel

ativ

e P

erce

nt A

rea

60 µM100 µM

+O

+O

+O

EYR

-CH2

-CH2

6 Use of Charged Aerosol Detection as an Orthogonal Quantification Technique for Drug Metaboites in Safety Testing (MIST)

Use of Charged Aerosol Detection as an Orthogonal Quantifi cation Technique for Drug Metabolites in Safety Testing (MIST) Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientifi c, Bannockburn, IL, USA; 2Thermo Fisher Scientifi c, Chelmsford, MA, USA

AbstractThe U.S. FDA guidelines on drug metabolites in safety testing (MIST) published in 2008 requires that the relative quantitation of human metabolites be obtained as soon as feasible in the drug development process. The goal is to ensure that at least one of the animal models used is not only producing the same metabolites found in humans, but at the same relative abundance. Identifi cation of drug metabolites is typically accomplished using various LC-MS approaches. However, metabolite quantifi cation is often diffi cult to accomplish with MS techniques alone. In some cases, the addition of UV detection is suffi cient for metabolite quantifi cation. But this requires that both parent drug and its metabolites possess similar and suffi ciently active chromophores, which is not always the case. Charged aerosol detection can detect any nonvolatile compound, typically with low ng sensitivity and similar response independent of chemical structure.

Presented here are the results from two test cases used to evaluate the usefulness of charged aerosol detection for MIST: buspirone (which has a strong UV chromophore) and erythromycin (which has a weak UV chromophore). These drugs were analyzed directly following incubation with human liver microsomes (drugs at ≤100 µM). The fi ve major metabolites for buspirone and the four major metabolites for erythromycin were identifi ed using an a Thermo Scientifi c LTQ Orbitrap mass spectrometer and quantifi ed by charged aerosol detection. The range, linearity, and sensitivity of this approach are discussed. Charged aerosol detection provides a complementary technique to those already being employed in industries capable of extending the range of in vitro drug metabolites that can be monitored during the drug development stage.

IntroductionInterest in metabolite and trace impurity analysis by the pharmaceutical industry is intensifying due to concerns with mass balance studies, regulatory commitments in reporting active pharmaceutical ingredient (API) impurities, MIST, and cleaning validation of manufacturing equipment. Most often an analytical requirement for accurately reporting the level of metabolites or impurities is to obtain reference standards or use a radiolabeled drug approach. Because many of these standards are unavailable and the radiochemical approaches are time consuming and expensive, quantifi cation of drug metabolites can often be diffi cult in early development stages. The situation is further exacerbated becase several types of HPLC detectors, such as UV or evaporative light scattering detection (ELSD), either do not provide uniform response across the target analytes or lack the sensitivity to detect these compounds. Although electrospray ionization (ESI) MS techniques are extremely sensitive and powerful approaches, ionization effi ciency can vary between metabolites and lead to quantifi cation issues.

The Thermo Scientifi c Dionex Corona™ ultra™ charged aerosol detector is mass sensitive and can be added to the traditional HPLC-UV or LC-MS platform. This detector provides the most consistent response for all nonvolatile and some semivolatile analytes of all HPLC detection techniques.1 The detector works by charging particles (see Figure 1) and is not dependent on light scattering which can vary between analytes. The work presented here examines the combination of UV and MS detection already employed in this fi eld, with charged aerosol detection for quantifi cation of drug metabolites in two specifi c cases. This work was completed in collaboration with Bristol-Myers Squibb (BMS), using a mix of HPLC equipment and method conditions.2 Although one analytical system and condition is emphasized, a variety of analytical equipment was used throughout the study.

All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries.

This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others.

FIGURE 1. Charged aerosol detector fl ow path schematic.

FIGURE 2. Schematic of one of the LC-charged aerosol dectector-MS systems used for analysis, with a total fl ow of 1 mL/min split ~5.7/1 to the charged aerosol dectector.

LPN 2953

FIGURE 3. Structure for buspirone and fi ve of its known metabolites.3

FIGURE 4. Chromatographic results for 60 µM buspirone before and after 120 min incubation in HLM.

FIGURE 5. Response as peak area for injection of equal concentrations of buspirone and the fi ve metabolites. Response deviation of each detection technique is calculated as percent RSD.

FIGURE 6. Linear response curve for buspirone with charged aerosol detection from ~4 to 124 ng on column.

FIGURE 7. Results for charged aerosol detection pre- and postincubation and UV at 210 nm preincubation the 100 µM erythromycin sample.

FIGURE 8. Structure (top) of erythromycin. Plot of relative area response of the four metabolites and the parent peak with charged aerosol detection detection (bottom).

DiscussionThe results discussed here where obtained in collaboration with BMS. The optimization of the system and development of the methods were ongoing throughout the work. Several components were identifi ed as crucial to maintaining quality data on both the charged aerosol detector and mass spectrometer. The column diameter and load volume was important, as peak splitting and shifting was observed with injection volumes greater than 30 µL. The confi guration of the fl ow splitting when running the LC-charged aerosol detector-MS system was determined to be the most crucial factor. The fl ow splitter must be positioned close to the mass spectrometer inlet using 0.004 inch or smaller i.d. tubing from the splitter to the mass spectrometer (Figure 2). The optimal fl ow-split range is estimated between 100 to 200 µL/min fl owing to the mass spectrometer, and the remaining fl ow to the charged aerosol detector. Because of the levels of sensitivity required for these analyses, a minimum fl ow rate of 500 µL/min is recommend for the charged aerosol detector.

The fi rst example, buspirone, was chosen because it is a well-characterized example, has a strong UV chromophore, and the standard material is readily available. The post HLM incubation of buspirone for both the charged aerosol detector and UV (Figure 4) show a similar metabolite profi le. The fi ve major metabolites shown in Figure 3 were quantifi ed as relative peak areas from the initial 60 µM buspirone sample. A sixth, minor metabolite was a volatile fragment of the buspirone and was not detected using the charged aerosol detector. The standards for the buspirone and the fi ve metabolites were prepared and analyzed at concentrations of 15 µM. The results for the two UV wavelengths (charged aerosol detection and MS) are shown in Figure 5. The greatest variability was found in the UV trace at 254 nm with an RSD of 53% followed by the MS at 37%. The charged aerosol detector and low wavelength UV had similar deviations of 20 and 17%, respectively.

The charged aerosol detector is typically linear from its limit of quantifi cation to ~500 ng on column, although not a linear detector over the full dynamic range. In Figure 6, the buspirone standard was analyzed using a UHPLC method from 3.9 to 124 ng on column. The six-point calibration curve exhibited excellent linearity over this range with a correlation coeffi cient of 0.9994. The concentration range of interest needed for the testing of drug metabolites typically falls in this mass-on-column range, which is linear for the charged aerosol detector independent of analyte.

MethodsSample PreparationThe buspirone and erythromycin standards were incubated at 1, 30, 60, and 100 μM substrate concentrations of human liver microsome (HLM). At the end of the incubation period, the samples were quenched with an equal part of acetonitrile. The fi nal analytical concentration of the metabolite samples discussed in this work is therefore one half of the concentration listed as the sample name (i.e., 60 µM pre-equals 30 µM postincubation).

Analytical Conditions Column: SB C18, 1.8 µm, 4.6 × 150 mmMobile Phases: A: 0.1% Formic acid in water, B: 0.1% Formic acid in acetonitrileFlow Rate: 1 mL/minTotal Run Time: 15 min

Detection Conditions Thermo Scientifi c Dionex UltiMate™ 3000 Diode Array Detector (Thermo Scientifi c Accela PDA detector used in some examples): UV Wavelength 1: 220 or 210 nm depending on example (values listed) UV Wavelength 2: 254 nm

Dionex Corona ultra:

Filter: High Gas: 35 psi Nitrogen Range: 100 pA full scale

LTQ Orbitrap high-resolution mass spectrometer Full scan MS with 1 × DDS

Flow Split

Valco three-way splitter Volumes fi xed by backpressure of different tubing diameters and lengths Volume measured at charged aerosol detector inlet.

ResultsBuspirone

Erythromycin

The second example, erythomycin, was chosen because it is representative of an array of drug candidates that fall into the nonvolatile analyte category with weak or no UV chromophores. As shown in Figure 7, the preincubation 100 µM standard has virtually no response with low wavelength UV. After incubation, the charged aerosol detector was able to detect the four major metabolites observed by MS. The relative area response for those four metabolites and the parent peak were evaluated at the 100 and 60 µM levels (Figure 8). As expected, the relative intensities of the metabolites increased at the lower concentration over the same incubation time.

The level of detection on the charged aerosol detector for the buspirone was determined to be ~1 ng on column with the short run UHPLC chemistry. This is typical for nonvolatile analytes under optimized conditions. Depending on molecular weight of the compound and the injection volume used, the molar concentration limit of detection can vary. While more sensitive than other nebulizer-based detection techniques, charged aerosol detection still may not provide suffi cient sensitivity to meet the detection levels needed for in vivo studies. The MetPro practice (advocated by J. Joseph in 2009), of normalizing the MS metabolite ratios at a single, higher-level point with UV data then transferring that to the MS results at the low levels has been employed in previous work.1 This was examined with the charged aerosol detector results during the calibration and is further described in Cai, et al, (2010).2 This offers a method to reduce the analytical deviations due to ionization variability, while keeping the sensitivity offered by the mass specrometer.

Conclusion The UHPLC-UV-charged aerosol detector-MS system described in this study was able to provide crucial information needed to meet requirements of the 2008 FDA MIST initiative. The use of charged aerosol detection as confi rmation of low-wavelength UV quantifi cation for chromophoric compounds can provide additional confi dence to the results. In areas where more specifi c wavelengths are used, such as the buspirone 254 nm, the charged aerosol detector can highlight large under- or overestimation as shown with metabolite.3 The area where this technology is most applicable is for parent or drug compounds that do not contain a strong UV chromophore. In these cases, laboratories are often left with no other option but to accept the MS quantifi cation during the early drug discovery stages. This can lead to major issues in later drug development stages. The charged aerosol detector—while not a stand-alone solution— can provide those additional pieces of data to make accurate interpretations of in vitro data without excessive cost or time requirements.

AcknowledgmentsThe authors would like to thank Hong Cai and colleagues at Bristoyl-Myers Squibb for their work on this collaboration. Generation of the data used would not have been possible without their help.

References1. Dependence of Response on Chemical Structure, ESA—A Dionex Company,

Application Brief 70-8913 Rev B.

2. Cai, et. al. A Novel Detection Technology: Charged Aerosol Detection (CAD) Coupled with HPLC, UV, and LTQ-Orbitrap MS for Semi-Quantitation of Metabolites in Drug Discovery Metabolism Studies, ASMS oral presentation, Salt Lake City, UT, May, 2010.

3. Zhu. et al. Drug Metab. Dispos. 2005, 33, 500–507.

101. Liquid eluent enters from the HPLC system2. Pneumatic nebulization occurs 3. Small droplets enter the drying tube4. Large drops exit to drain 5. Dried particles enter the mixing chamber6. Gas stream passes over corona needle7. Charged gas collides with particles and transfers charge 8. High mobility species are removed9. Remaining charged particles measured with a very sensitive electrometer10. Signal is transferred to chromatographic data software

1

2

4

5 6

7

8

9

26802

28120

0.850 µL

DualGradient

Pump

DAD

TCC

0.150 µL

ultra

Autosampler

28121

HO

HO

N

O

O N NN

NMet1

Met4

Met2

Met5

Met3O

N N

N

N

N

O

Chemical Formula: C21H31N5O2

Exact Mass: 385.24778

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

NO

O

OH

NNN

N

O

HN

O

O

N NN

N

NO

O N+

-ON

N

N

N

O

O

NNN

N

Buspirone

28122

-5

-5

20Buspirone 60 µM t = 0 by CAD

pA

-5

15Buspirone 60 µM t = 120 by CAD

pA

15Matrix Blank t = 120 by CAD

pA

-2

12Buspirone 60 µM t = 0 by UV at 254 nm

mAU

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5Minutes

11.0 11.5-1.5

8.0Buspirone 60 µM t = 120 by UV at 254 nm

mAU

28123

UV 220

0.E+00

1.E+05

2.E+05

3.E+05

4.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

UV 254

0.E+005.E+041.E+052.E+052.E+053.E+053.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 53%

CAD

00.5

11.5

22.5

33.5

BUSP Met1 Met2 Met3 Met4 Met5

PA/n

g

RSD 20%RSD 17%

HRMS

0.E+005.E+081.E+092.E+092.E+093.E+09

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 37%

28124

R2 = 0.9994

0

50,000

100,000

150,000

200,000

250,000

300,000

0 20 40 60 80 100 120 140

Mass on Column (ng)

Res

pons

e (p

eak

area

)

28125

-5

45Erythromycin 100 µM t = 0

pA

-5

45Erythromycin 100 µM t = 120

pA

7.88 8.5 9.0 9.5 10.0 10.5 11.025

100Erythromycin 100 µM t = 0

mAU

Minutes

28126

O

OCH3

H3C

OH

OH

CH3

HO

H3C

CH3O

CH3

H3C

O

O

OCH3

CH3

OHCH3

OO

HON

CH3

H3C

CH3

Chemical Formula: C37H67NO13

Exact Mass: 733.46124

Erythromycin

0

10

20

30

40

50

60

70

80

90

8.7 8.9 9.2 10.1 10.2Retention Time

Rel

ativ

e P

erce

nt A

rea

60 µM100 µM

+O

+O

+O

EYR

-CH2

-CH2

7Thermo Scientific Poster Note • LPN2953-01_e 11/11SV

Use of Charged Aerosol Detection as an Orthogonal Quantifi cation Technique for Drug Metabolites in Safety Testing (MIST) Gregory Malek,1 Chris Crafts,2 Marc Plante,2 Matthew Neely,1 and Bruce Bailey2

1Thermo Fisher Scientifi c, Bannockburn, IL, USA; 2Thermo Fisher Scientifi c, Chelmsford, MA, USA

AbstractThe U.S. FDA guidelines on drug metabolites in safety testing (MIST) published in 2008 requires that the relative quantitation of human metabolites be obtained as soon as feasible in the drug development process. The goal is to ensure that at least one of the animal models used is not only producing the same metabolites found in humans, but at the same relative abundance. Identifi cation of drug metabolites is typically accomplished using various LC-MS approaches. However, metabolite quantifi cation is often diffi cult to accomplish with MS techniques alone. In some cases, the addition of UV detection is suffi cient for metabolite quantifi cation. But this requires that both parent drug and its metabolites possess similar and suffi ciently active chromophores, which is not always the case. Charged aerosol detection can detect any nonvolatile compound, typically with low ng sensitivity and similar response independent of chemical structure.

Presented here are the results from two test cases used to evaluate the usefulness of charged aerosol detection for MIST: buspirone (which has a strong UV chromophore) and erythromycin (which has a weak UV chromophore). These drugs were analyzed directly following incubation with human liver microsomes (drugs at ≤100 µM). The fi ve major metabolites for buspirone and the four major metabolites for erythromycin were identifi ed using an a Thermo Scientifi c LTQ Orbitrap mass spectrometer and quantifi ed by charged aerosol detection. The range, linearity, and sensitivity of this approach are discussed. Charged aerosol detection provides a complementary technique to those already being employed in industries capable of extending the range of in vitro drug metabolites that can be monitored during the drug development stage.

IntroductionInterest in metabolite and trace impurity analysis by the pharmaceutical industry is intensifying due to concerns with mass balance studies, regulatory commitments in reporting active pharmaceutical ingredient (API) impurities, MIST, and cleaning validation of manufacturing equipment. Most often an analytical requirement for accurately reporting the level of metabolites or impurities is to obtain reference standards or use a radiolabeled drug approach. Because many of these standards are unavailable and the radiochemical approaches are time consuming and expensive, quantifi cation of drug metabolites can often be diffi cult in early development stages. The situation is further exacerbated becase several types of HPLC detectors, such as UV or evaporative light scattering detection (ELSD), either do not provide uniform response across the target analytes or lack the sensitivity to detect these compounds. Although electrospray ionization (ESI) MS techniques are extremely sensitive and powerful approaches, ionization effi ciency can vary between metabolites and lead to quantifi cation issues.

The Thermo Scientifi c Dionex Corona™ ultra™ charged aerosol detector is mass sensitive and can be added to the traditional HPLC-UV or LC-MS platform. This detector provides the most consistent response for all nonvolatile and some semivolatile analytes of all HPLC detection techniques.1 The detector works by charging particles (see Figure 1) and is not dependent on light scattering which can vary between analytes. The work presented here examines the combination of UV and MS detection already employed in this fi eld, with charged aerosol detection for quantifi cation of drug metabolites in two specifi c cases. This work was completed in collaboration with Bristol-Myers Squibb (BMS), using a mix of HPLC equipment and method conditions.2 Although one analytical system and condition is emphasized, a variety of analytical equipment was used throughout the study.

All trademarks are the property of Thermo Fisher Scientifi c Inc. and its subsidiaries.

This information is not intended to encourage use of these products in any manners that might infringe the intellectual property rights of others.

FIGURE 1. Charged aerosol detector fl ow path schematic.

FIGURE 2. Schematic of one of the LC-charged aerosol dectector-MS systems used for analysis, with a total fl ow of 1 mL/min split ~5.7/1 to the charged aerosol dectector.

LPN 2953

FIGURE 3. Structure for buspirone and fi ve of its known metabolites.3

FIGURE 4. Chromatographic results for 60 µM buspirone before and after 120 min incubation in HLM.

FIGURE 5. Response as peak area for injection of equal concentrations of buspirone and the fi ve metabolites. Response deviation of each detection technique is calculated as percent RSD.

FIGURE 6. Linear response curve for buspirone with charged aerosol detection from ~4 to 124 ng on column.

FIGURE 7. Results for charged aerosol detection pre- and postincubation and UV at 210 nm preincubation the 100 µM erythromycin sample.

FIGURE 8. Structure (top) of erythromycin. Plot of relative area response of the four metabolites and the parent peak with charged aerosol detection detection (bottom).

DiscussionThe results discussed here where obtained in collaboration with BMS. The optimization of the system and development of the methods were ongoing throughout the work. Several components were identifi ed as crucial to maintaining quality data on both the charged aerosol detector and mass spectrometer. The column diameter and load volume was important, as peak splitting and shifting was observed with injection volumes greater than 30 µL. The confi guration of the fl ow splitting when running the LC-charged aerosol detector-MS system was determined to be the most crucial factor. The fl ow splitter must be positioned close to the mass spectrometer inlet using 0.004 inch or smaller i.d. tubing from the splitter to the mass spectrometer (Figure 2). The optimal fl ow-split range is estimated between 100 to 200 µL/min fl owing to the mass spectrometer, and the remaining fl ow to the charged aerosol detector. Because of the levels of sensitivity required for these analyses, a minimum fl ow rate of 500 µL/min is recommend for the charged aerosol detector.

The fi rst example, buspirone, was chosen because it is a well-characterized example, has a strong UV chromophore, and the standard material is readily available. The post HLM incubation of buspirone for both the charged aerosol detector and UV (Figure 4) show a similar metabolite profi le. The fi ve major metabolites shown in Figure 3 were quantifi ed as relative peak areas from the initial 60 µM buspirone sample. A sixth, minor metabolite was a volatile fragment of the buspirone and was not detected using the charged aerosol detector. The standards for the buspirone and the fi ve metabolites were prepared and analyzed at concentrations of 15 µM. The results for the two UV wavelengths (charged aerosol detection and MS) are shown in Figure 5. The greatest variability was found in the UV trace at 254 nm with an RSD of 53% followed by the MS at 37%. The charged aerosol detector and low wavelength UV had similar deviations of 20 and 17%, respectively.

The charged aerosol detector is typically linear from its limit of quantifi cation to ~500 ng on column, although not a linear detector over the full dynamic range. In Figure 6, the buspirone standard was analyzed using a UHPLC method from 3.9 to 124 ng on column. The six-point calibration curve exhibited excellent linearity over this range with a correlation coeffi cient of 0.9994. The concentration range of interest needed for the testing of drug metabolites typically falls in this mass-on-column range, which is linear for the charged aerosol detector independent of analyte.

MethodsSample PreparationThe buspirone and erythromycin standards were incubated at 1, 30, 60, and 100 μM substrate concentrations of human liver microsome (HLM). At the end of the incubation period, the samples were quenched with an equal part of acetonitrile. The fi nal analytical concentration of the metabolite samples discussed in this work is therefore one half of the concentration listed as the sample name (i.e., 60 µM pre-equals 30 µM postincubation).

Analytical Conditions Column: SB C18, 1.8 µm, 4.6 × 150 mmMobile Phases: A: 0.1% Formic acid in water, B: 0.1% Formic acid in acetonitrileFlow Rate: 1 mL/minTotal Run Time: 15 min

Detection Conditions Thermo Scientifi c Dionex UltiMate™ 3000 Diode Array Detector (Thermo Scientifi c Accela PDA detector used in some examples): UV Wavelength 1: 220 or 210 nm depending on example (values listed) UV Wavelength 2: 254 nm

Dionex Corona ultra:

Filter: High Gas: 35 psi Nitrogen Range: 100 pA full scale

LTQ Orbitrap high-resolution mass spectrometer Full scan MS with 1 × DDS

Flow Split

Valco three-way splitter Volumes fi xed by backpressure of different tubing diameters and lengths Volume measured at charged aerosol detector inlet.

ResultsBuspirone

Erythromycin

The second example, erythomycin, was chosen because it is representative of an array of drug candidates that fall into the nonvolatile analyte category with weak or no UV chromophores. As shown in Figure 7, the preincubation 100 µM standard has virtually no response with low wavelength UV. After incubation, the charged aerosol detector was able to detect the four major metabolites observed by MS. The relative area response for those four metabolites and the parent peak were evaluated at the 100 and 60 µM levels (Figure 8). As expected, the relative intensities of the metabolites increased at the lower concentration over the same incubation time.

The level of detection on the charged aerosol detector for the buspirone was determined to be ~1 ng on column with the short run UHPLC chemistry. This is typical for nonvolatile analytes under optimized conditions. Depending on molecular weight of the compound and the injection volume used, the molar concentration limit of detection can vary. While more sensitive than other nebulizer-based detection techniques, charged aerosol detection still may not provide suffi cient sensitivity to meet the detection levels needed for in vivo studies. The MetPro practice (advocated by J. Joseph in 2009), of normalizing the MS metabolite ratios at a single, higher-level point with UV data then transferring that to the MS results at the low levels has been employed in previous work.1 This was examined with the charged aerosol detector results during the calibration and is further described in Cai, et al, (2010).2 This offers a method to reduce the analytical deviations due to ionization variability, while keeping the sensitivity offered by the mass specrometer.

Conclusion The UHPLC-UV-charged aerosol detector-MS system described in this study was able to provide crucial information needed to meet requirements of the 2008 FDA MIST initiative. The use of charged aerosol detection as confi rmation of low-wavelength UV quantifi cation for chromophoric compounds can provide additional confi dence to the results. In areas where more specifi c wavelengths are used, such as the buspirone 254 nm, the charged aerosol detector can highlight large under- or overestimation as shown with metabolite.3 The area where this technology is most applicable is for parent or drug compounds that do not contain a strong UV chromophore. In these cases, laboratories are often left with no other option but to accept the MS quantifi cation during the early drug discovery stages. This can lead to major issues in later drug development stages. The charged aerosol detector—while not a stand-alone solution— can provide those additional pieces of data to make accurate interpretations of in vitro data without excessive cost or time requirements.

AcknowledgmentsThe authors would like to thank Hong Cai and colleagues at Bristoyl-Myers Squibb for their work on this collaboration. Generation of the data used would not have been possible without their help.

References1. Dependence of Response on Chemical Structure, ESA—A Dionex Company,

Application Brief 70-8913 Rev B.

2. Cai, et. al. A Novel Detection Technology: Charged Aerosol Detection (CAD) Coupled with HPLC, UV, and LTQ-Orbitrap MS for Semi-Quantitation of Metabolites in Drug Discovery Metabolism Studies, ASMS oral presentation, Salt Lake City, UT, May, 2010.

3. Zhu. et al. Drug Metab. Dispos. 2005, 33, 500–507.

101. Liquid eluent enters from the HPLC system2. Pneumatic nebulization occurs 3. Small droplets enter the drying tube4. Large drops exit to drain 5. Dried particles enter the mixing chamber6. Gas stream passes over corona needle7. Charged gas collides with particles and transfers charge 8. High mobility species are removed9. Remaining charged particles measured with a very sensitive electrometer10. Signal is transferred to chromatographic data software

1

2

4

5 6

7

8

9

26802

28120

0.850 µL

DualGradient

Pump

DAD

TCC

0.150 µL

ultra

Autosampler

28121

HO

HO

N

O

O N NN

NMet1

Met4

Met2

Met5

Met3O

N N

N

N

N

O

Chemical Formula: C21H31N5O2

Exact Mass: 385.24778

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

Chemical Formula: C21H31N5O3

Exact Mass: 401.24269

NO

O

OH

NNN

N

O

HN

O

O

N NN

N

NO

O N+

-ON

N

N

N

O

O

NNN

N

Buspirone

28122

-5

-5

20Buspirone 60 µM t = 0 by CAD

pA

-5

15Buspirone 60 µM t = 120 by CAD

pA

15Matrix Blank t = 120 by CAD

pA

-2

12Buspirone 60 µM t = 0 by UV at 254 nm

mAU

6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5Minutes

11.0 11.5-1.5

8.0Buspirone 60 µM t = 120 by UV at 254 nm

mAU

28123

UV 220

0.E+00

1.E+05

2.E+05

3.E+05

4.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

UV 254

0.E+005.E+041.E+052.E+052.E+053.E+053.E+05

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 53%

CAD

00.5

11.5

22.5

33.5

BUSP Met1 Met2 Met3 Met4 Met5

PA/n

g

RSD 20%RSD 17%

HRMS

0.E+005.E+081.E+092.E+092.E+093.E+09

BUSP Met1 Met2 Met3 Met4 Met5

PA

RSD 37%

28124

R2 = 0.9994

0

50,000

100,000

150,000

200,000

250,000

300,000

0 20 40 60 80 100 120 140

Mass on Column (ng)

Res

pons

e (p

eak

area

)

28125

-5

45Erythromycin 100 µM t = 0

pA

-5

45Erythromycin 100 µM t = 120

pA

7.88 8.5 9.0 9.5 10.0 10.5 11.025

100Erythromycin 100 µM t = 0

mAU

Minutes

28126

O

OCH3

H3C

OH

OH

CH3

HO

H3C

CH3O

CH3

H3C

O

O

OCH3

CH3

OHCH3

OO

HON

CH3

H3C

CH3

Chemical Formula: C37H67NO13

Exact Mass: 733.46124

Erythromycin

0

10

20

30

40

50

60

70

80

90

8.7 8.9 9.2 10.1 10.2Retention Time

Rel

ativ

e P

erce

nt A

rea

60 µM100 µM

+O

+O

+O

EYR

-CH2

-CH2

©2011 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo Fisher Scientific Inc. and its subsidiaries. Specifications, terms and pricing are subject to change. Not all products are available in all countries. Please consult your local sales representative for details.

Thermo Scientific Dionex products are designed, developed, and manufactured under an ISO 9001 Quality System.

U.S./Canada (847) 295 7500Brazil (55) 11 3731 5140Austria (43) 1 616 51 25 Benelux (31) 20 683 9768 (32) 3 353 42 94

Denmark (45) 36 36 90 90 France (33) 1 39 30 01 10 Germany (49) 6126 991 0 Ireland (353) 1 644 0064 Italy (39) 02 51 62 1267

Sweden (46) 8 473 3380Switzerland (41) 62 205 9966 United Kingdom (44) 1276 691722Australia (61) 2 9420 5233 China (852) 2428 3282

India (91) 22 2764 2735 Japan (81) 6 6885 1213 Korea (82) 2 2653 2580 Singapore (65) 6289 1190 Taiwan (886) 2 8751 6655

www.thermoscientific.com/dionex