Fentanyl Analog Screening using LC-TIMS-TOF MS/MS

Summary

A method for the screening of fentanyl analogs based on their retention time, mobility, and mass spectrometry fragmentation pattern.

Abstract

The use of fentanyl and the emergence of fentanyl analogs over recent decades has become an increasing concern to the community at large. Fentanyl and its analogs are the major contributors to fatal and nonfatal overdoses in the United States. Most recent cases of fentanyl-related overdose are linked to illicitly manufactured fentanyl and its associated extreme potency. In the present work, we describe a high-throughput analytical protocol for the screening of fentanyl analogs. The use of complementary liquid chromatography, trapped ion mobility spectrometry, and tandem mass spectrometry allow for the separation and assignment of hundreds of fentanyl analogs from a single sample in a single scan. The described approach takes advantage of the recent development of data-dependent acquisition and data-independent acquisition using parallel accumulation in the mobility trap followed by sequential fragmentation using collision-induced dissociation. The fentanyl analogs are confidently assigned based on their retention time, mobility, and MS fragmentation pattern.

Introduction

Fentanyl and its analogs are the major contributors to fatal and nonfatal overdoses in the United States^1,2. The Center for Disease Control and Prevention (CDC) reported that the number of synthetic opioid-related overdose deaths from 2013 through 2021 was over 258,000. In 2021 alone, over 68,000 overdose deaths could be attributed to synthetic opioids, totaling 82% of all overdose-related deaths in the nation³. Since 2013, hundreds of fentanyl analogs have been identified, with varying potency⁴. With the emergence of illicitly manufactured fentanyl analogs, the schedule II synthetic opioid itself remains the most popular synthetic opioid available in the United States³. According to the Center for Forensic Science Research and Education (CFSRE), the top reported fentanyl analog in 2022 was fluorofentanyl, with additional non-fentanyl related synthetic opioids now being introduced into the volatile drug market supply at a rapid rate⁵.

Due to the overwhelming volume of fentanyl and fentanyl-related analogs circulating in the drug market, the DEA has implemented a fentanyl signature profiling program entitled Operation Death Dragon, in order to track the methodology used to synthesize these compounds with hopes of linking drug seizures back to their origin⁶. In 2018, 94% of the drug seizures were identified as being synthesized by the Janssen method, while the remaining 6% were synthesized with the Siegfried method⁶. The primary difference between the two methods is the presence of the fentanyl analog, benzyl fentanyl, detected as an impurity in the Janssen method, while the presence of despropionyl fentanyl (4-ANPP), a metabolite/precursor of fentanyl, is an impurity detected when synthesizing with the Siegfried method⁷.

The use of gas and liquid chromatography in tandem with mass spectrometry (GC-MS and LC-MS, respectively) for the targeted screening and quantification of synthetic opioids is regularly implemented in toxicology laboratories. GC-MS has been considered the gold standard for the detection of drugs of abuse in biological specimens. Access to publicly available mass spectral libraries⁸ and instrumentation marketed as plug and play systems⁹ are some reasons why GC-MS has remained an integral part in laboratories for both comprehensive screening as well as targeted quantification^9,10. However, current GC-MS quantification methods in the literature tend to have a limited scope of analytes¹¹ and quickly become outdated and not applicable to current casework. More importantly, the limits of detection and quantification are not comparable to LC-MS methods (< 1 ng/mL)¹², therefore increasing the potential for false-negative results. One such comparison between GC-MS and LC-MS which looked at postmortem specimens, noted that out of 134 positively identified cases of carfentanil, one of the most potent synthetic opioids to date, 104 of those cases screened negative for carfentanil utilizing GC-MS¹³. In drug analysis laboratories, GC-MS is more frequently used and amendable for synthetic opioids due to the high concentration of the analyzed samples. Yet, GC-MS is still utilized in conjunction with additional techniques such as Fourier transform-infrared spectroscopy (FT-IR) and scanning electron microscopy (SEM) for the confirmation of these compounds¹⁴. The application of GC-MS to the analysis of biological specimens in forensic toxicology requires sample preparation methods that include the extraction of these compounds using either liquid-liquid extraction (LLE) or solid-phase extraction (SPE)¹⁵. LLE can be conducted with a variety of solvents, however in a high-production lab, LLE may not be cost-efficient or timely. LLE consumes large quantities of solvents as well as sample volume, while the alternative, SPE, can be automated and requires minimum sample volume¹⁶. A recent GC-MS study reported the separation of 20 different isomeric fentanyl analogs using three separate GC thermal programs¹⁷. While baseline separation among isomers was successful, the applicability of this method to relevant forensic casework and workflows is limited.

LC-MS has gained popularity in forensic testing, especially due to the limitations of GC-MS when it comes to non-volatile and heat-sensitive compounds^18,19. LC-MS screening utilizing triple quadrupoles (QQQ) and ion trap instrumentation have been successful in detecting synthetic opioids at low concentrations (<1 ng/mL)^{12,20,21,22,23,24}. Typically, these LC-MS methods are used as secondary confirmation methods to complement immunoassay and/or GC-MS results. In 2017, Shoff et al. at the Miami-Dade County medical examiner department (MDME) toxicology laboratory developed a comprehensive screening method for 44 opioid-related compounds using ultra high pressure liquid chromatography (UHPLC)-ion trap-MS^{n 13}. Similar to targeted MRM methods, this ion trap method utilized a scheduled precursor list (SPL) containing retention times, precursor target ions, as well as primary daughter ions for MS³ spectral fragmentation when possible. What separates this screening method from methods developed on triple quadrupoles and linear quadrupole ion traps is the additional detail provided in the spectral data. What this screening method cannot provide is quantitative analysis, which is rarely developed on ion trap instrumentation, as well as the identification of unknowns¹³. LC-QQQ methods for synthetic opioids can simultaneously screen and quantify a predefined target list. The use of multiple reaction monitoring (MRM) transitions for the identification and quantification of compounds is a trusted data acquisition technique and is the most common technique utilized in the literature for the detection of synthetic opioids^12,20. Reported linear ranges for the quantification of an array of synthetic opioids span from 0.01-100 ng/mL, with more potent synthetic opioids, such as carfentanil, detected at the sub-ng/mL range^{11,24,25,26,27}.

The separation of isomeric synthetic opioids has been addressed in LC-MS methods. One such method separated 174 isomeric fentanyl analogs within a 16 min runtime using a biphenyl column²⁸. In addition, fluctuations in LC-dependent parameters such as column efficiency, mobile phase pH, and pressure changes create retention time shifts that must be accounted for in targeted assays with consistent monitoring and larger collection windows (>0.4 min), which may result in potential overlap of these once resolved isomers. Additional chromatographic techniques for the separation of isomers have been explored, including the use of 2-dimensional liquid chromatography (2D-LC)²⁹; and while this technique does have the ability to provide orthogonal separation of compounds, the disadvantages outweigh the advantages including excessive run-times, cost, difficulty in method development, and usefulness compared to alternative separations³⁰.

High resolution mass spectrometry (HRMS) is becoming increasingly reliable for the identification of synthetic opioids. Several instrumentation companies have marketed targeted methods developed for the purposes of identifying a broad range of synthetic opioids in complex biological matrices^19,28,31. These methods, unlike previously discussed methods, can store analytical data for retroactive analysis. Thus, specimens that were previously analyzed with undetermined results can be revisited later for discovery of newly identified compounds. What separates HRMS from QQQ is accurate mass identification. While both MS techniques can accurately quantitate at low concentrations, HRMS has been proven to be more effective for initial screening and discovery of unknown compounds^32,33,34. HRMS, specifically with time-of-flight (TOF) MS analyzers, has been at the forefront of NPS discovery, allowing forensic testing laboratories to deliver time-sensitive data on new compounds to both law enforcement and the scientific community to scale the spread of some of these compounds while increasing awareness and education³³. The use of TOF for the detection of drugs of abuse has become extremely comprehensive, with methods containing an upwards of 600 compounds separated within a 10 min chromatographic program. Previous studies have been reported the advantages of trapped ion mobility spectroscopy (TIMS) coupled to TOF for detection and separation of isomeric opioids³⁵.

Taking advantage of the orthogonality between liquid chromatography, trapped ion mobility spectrometry and mass spectrometry, the presented method provides a broad characterization of fentanyl analogs based on the retention time, isotopic pattern, mobility, and fragmentation pattern.

Protocol

1. Sample preparation

1. Store the fentanyl analog screening kit at -20 °C once they have been received. Resuspend each sample to a final concentration of 400 µg/mL by the addition of 500 µL of pure LC/MS grade methanol.
2. Mix using a plate mixer or vortexer at 400 rpm for 1 h or vortex at medium speed for 15 min. After resuspension store the vials at -20 °C.
3. Dilute each standard to a final concentration of 1 ng/mL using pure LC/MS grade methanol.
4. Group the standards into groups, such that each group does not contain any isomeric standards. See Supplementary Table 1 for the breakdown of the 14 groups of fentanyl standards.

2. HPLC mobile phases preparation

1. Prepare mobile phase A (MPA) using 5 mM/L ammonium formate (NH₄HCO₂) in H₂O with 0.05% formic acid (85%).
   1. Weigh out 0.078 g of NH₄HCO₂, pour into a 250 mL volumetric flask and fill with water to about 2/3^rd. Swirl to mix well.
   2. Once fully dissolved, add 0.125 mL of formic acid (85%). Fill with water to the fill line (250 mL) and mix.
2. Prepare mobile phase B (MPB) using 0.05% formic acid in a 1:1 methanol acetonitrile mixture.
   1. Pour 125 mL of acetonitrile into a volumetric flask and add methanol until it reaches the fill line.
   2. Swirl to mix well. Add 0.125 mL of 0.05% formic acid to the flask and mix to homogenize.

3. HPLC method development

1. Open the LC software to the home screen. In the sample table, located at the center of the window, create a new protocol.
   1. Annotate the vial column with the sample location. Annotate the sample ID column using the format YYYYMMDD_NAME (representative descriptor). Annotate the volume column with the desired the injection volume (15 µL).
2. Annotate the data path column with the location where the data will be saved to.
3. HPLC method input
   1. Select New from the bar in the sample table located in the middle of the screen. Under the separation method, select the Drop-down Arrow and click New Method in the pop-up window.
   2. A new window will pop up labeled separation method, here edit the acquisition time or the instrument control framework (ICF) system method.
   3. Double click the Edit Method button for the ICF system to open up the popup to edit the binary gradient, as seen in Figure 1.
   4. Under the binary gradient tab, a visual representation of the gradient is on the left and the breakdown of the gradient is on the right, set the time stamps, the flow rate and the MPA and MPB concentrations.
   5. Ensure the stop time is set to 18 min, the flow rate to 0.400 mL/min, and pressure limit min and max at 0 psi and 6000 psi.
   6. Set the time stamp concentrations as following:
      At 0.00 min, set B concentration to 20.0.
      At 1.50 min, set B concentration to 25.0.
      At 3.00 min, set B concentration to 27.0.
      At 6.00 min, set B concentration to 27.0.
      At 6.50 min, set B concentration to 30.0.
      At 7.00 min, set B concentration to 95.0.
      At 16.00 min, set B concentration to 95.0.
      At 16.50 min, set B concentration to 20.0.

4. Initialization of the HPLC

1. In the HPLC column section, insert the LC column (monolithic C18 HPLC column 100 mm x 4.6 mm) and column guards (guard column 5 mm x 4.6 mm). Pay attention to the minimum and maximum column pressure.
2. Run a sample blank (same buffer solution) to test for any leaks and create a baseline. Look for leaks. Leaks could be observed by column pressure fluctuations. If there are leaks determine where they might be coming from and tighten the connections.
3. Run a sample blank (same buffer solution) using the desired LC program to precondition the column.

5. timsTOF MS/MS method development

1. Open the timsControl application. On the left side of the window are the MS and TIMS Settings.
   1. Under MS settings set the scan beginning and ending to 50 m/z and 1800 m/z, respectively. Select positive mode for the ion polarity and select parallel accumulation-serial fragmentation for the scan mode.
   2. Under TIMS Settings set the Mode to custom, 1/K0 start to 0.40 Vs/cm², 1/K0 end to 1.85 Vs/cm², ramp time set to 150.0 ms, and MS averaging set to 1.
2. In the lower selection of tabs under Source perform the following changes in the two setting boxes.
   1. In the source, set the end plate offset to 500 V, capillary to 4500 V, nebulizer to 3.0 bar, dry gas to 10.0 L/min, and dry temp to 200 °C.
   2. In the syringe pump settings, ensure the syringe is Hamilton 1 mL, active is enabled, and the flow rate set to 80.0 µL/h.
3. In the Tune tab perform the following changes in the setting tabs for General, Processing, and TIMS.
   1. Under general settings tab, setup the following sections: transfer, collision cell, quadrupole, focus pre TOF, and detection.
      1. Under the transfer settings, set the deflection 1 delta to 70.0 V, funnel 1 RF to 341.0 Vpp, CID energy to 0.0 eV, funnel 2 RF to 300.0 Vpp, multipole RF to 300.0 Vpp.
      2. Under the collision cell settings, ensure that the collision energy is set to 6.0 eV and the collision RF to 1200.0 Vpp.
      3. Under the quadrupole settings, ensure ion energy is set to 6.0 eV and low mass to 250.00 m/z.
      4. Under the focus pre TOF settings, set transfer time to 75.0 µs and pre-pulse storage to 5.0 µs.
   2. Under the processing settings tab, set up the following sections: mass spectra peak detection and mobility peak detection.
      1. In the mass spectra peak settings, unselect sum of intensities (area) and set the absolute threshold to 667.
   3. Under the TIMS tab, set up the following sections: offsets, ion change control, and advanced parameters.
      1. Under the offsets setting set Δt1 to -20.0 V, Δt2 to -100.0 V, Δt3 to 40.0 V, Δt4 to 80.0 V, Δt5 to 0.0 V, Δt6 to 120.0, and collision cell in to 250.0 V.
      2. Under the ion charge control settings, click the box to enable and set the target intensity to 5.00 M.
      3. Under the advanced parameters, enable the lock accumulation to mobility range.
4. In the MS/MS tab set the Scan mode to parallel accumulation-serial fragmentation and set up the following sections: precursor ions, scheduling, active exclusion, collision energy settings, isolation width settings, and TIMS stepping.
   1. Under precursor ions, set the number of parallel accumulation-serial fragmentation ramps to 8, charge minimum to 1, and charge maximum to 5.
   2. Under scheduling settings, enable the precursor repetitions. Under active exclusion, check the box to enable and set the release to 0.40 min after.
   3. Do not adjust the collision energy settings and isolation width settings. Click the TIMS Stepping Box to enable.

6. Mobility and mass calibration

1. Perform calibration for both m/z and mobility domains. Under the m/z calibration settings, select one of the preloaded tuning mix profiles in the reference list box that appear in the drop-down menu.
2. To calibrate, load the syringe for the TOF with a tuning mix solution. On the right side of the window, use the settings titled calibration mode for the various calibration types to be set to achieve the highest score.
3. Ensure that the score is as close to 100% as possible. Switch between linear, quadradic and enhanced quadratic to achieve the best score.
4. Calibrate for mobility following steps 6.1-6.3 used for calibrating m/z.
5. Once calibrated, save the MS method by selecting the Method tab on the top bar. In the drop-down menu, select Save As to generate a new MS method file.

7. Creating data independent acquisition (dia) parallel accumulation-serial fragmentation method

1. Once a data set has been collected in data dependent acquisition (dda) parallel accumulation-serial fragmentation, run the experiments in dia. Open the ion mobility software application and load the dda method saved in step 6.5.
2. Leave all the settings the same expect for the MS setting, change this from parallel accumulation-serial fragmentation to dia- parallel accumulation-serial fragmentation.
3. On the bottom of the panel, select the MSMS tab and then click on Window Editor to open the dia window popup shown in Figure 2.
4. Load the previously saved dda data set (.m file) using the open analysis button located at the top of the popup window.
5. A heat map appears at the bottom left displaying windows with a polygon of windows running diagonally across the graph. Click and drag to resize the polygon so that it fits the data in the heat map (corners are selectable by double clicking on the lateral lines a point will appear that can be dragged and resized).
6. To the right of the windows are the window settings. Set the mass width (50 is go to) and set the horizontal overlap, both in Da.
7. Set the number of mobility windows in per mass width and the vertical overlap.
8. Click Calculate Windows on the bottom right of the popup to see the windows displayed with the new settings. Once suitable, click Apply dia-PASEF Windows to Method. This will close the popup bringing it back to the home screen.

8. HPLC ion mobility TOF data processing

1. Open the data analysis software. At the top left corner click on the File tab and select Open from the drop down. In the new files window select the files of choice and click Open.
2. Check the calibration. Right click on the file name under the analysis box and select Properties. A window labeled file name_analysis properties will appear. Select Calibration Status.
3. From the drop-down box, select Instrument Calibration for the Mass Spectrometer. Confirm that the error is no greater than 1 µg/mL. Select Initial Mobility Calibration and confirm that the error is no greater than 1 µg/mL.
4. Prepare a chromatogram in the data analysis software as described below.
   1. Right click on Chromatogram under the file name, select Edit Chromatogram. In type, click the drop-down box and select Extracted Ion Chromatogram.
   2. Under filter, select All MS and with scan mode select All. This is to view the peaks for the molecule of interest rather than only its fragments.
   3. Below this, look for the two filter options for molecule selection: masses or formula highlighted in blue.
   4. If using masses to extract an ion, insert the theoretical m/z of the molecule of interest. For this case select (mass for representative result) m/z.
   5. If using formula to extract an ion, insert the formula for the molecule as well as the ion forms of interest for the chromatogram. In this case, select the protonated ion, [M+H]⁺.
   6. Set the polarity to positive mode.
5. Mass spectrum
   1. Right click at the baseline of the peak of the compound and drag to the other edge of the peak. This will create a mass spectrum of fragments within that retention time.
6. Generation of Mobilogram
   1. Right click on the left tab titled mobilogram and select Edit Mobilogram. A window titled edit mobilogram traces will appear. Follow steps 8.4.2-8.4.7 for editing the chromatogram.
   2. In the retention time input, add the retention time range of the peak of interest.
   3. Once the parameters are selected, click Add followed by Ok at the top right of the edit mobilogram traces window.
   4. After a short amount of time, the software will process the desired selections and output a chromatogram. Repeat steps 8.6.1-8.6.3 for all ions in the mixture.
   5. Generating the compound spectra
      1. At the bottom of the spectrum view window, select Profile MS and Fragment MS. This will allow to include ions from full scan and PASEF.
      2. At spectrum view, right click and select Copy Compound Spectra. With the spectrum data that appears at the right, find more information on the compound such as resolution resolving power, intensity, and signal to noise ratio (S/N).
   6. Data processing
      1. To process the data, manually integrate the chromatogram and mobility peaks to yield important information on the molecule of interest.
      2. Right click Find and select Integrate only chromatogram or mobilogram. Left click and drag to highlight the desired peak. The information will be displayed which includes retention time, area, S/N, and mobility.

Results

The fentanyl analog screening kit of 250 analogs standards were divided into 14 groups: 12 groups of 17 analogs and 2 groups of 16 analogs, to avoid m/z interferences. Each analog is further characterized by their m/z, retention time (RT), mobility (K), and MS/MS fragmentation pattern.

Examples of isomeric separations are shown in Figure 3 and Figure 4 for the C₂₂H₂₈N₂O₂ and...

Discussion

The analytical separation of biological samples containing high isomeric content can be analytically challenging. In this paper, the method described aims to characterize 29 isomer sets, for a total of 185 analogs from a 250 opioid standard kit. During the test group preparation, it is important to ensure that there are no two analogs with m/z that cannot be experimentally distinguished. The data described here uses standards not from a human-based matrix, if this method is to be used in a field experiment; additional tr...

Materials

Name	Company	Catalog Number	Comments
Ammonium formate for HPLC	Fluka	17843-50G
Eppendorf Snap-Cap Microcentrifuge Safe-Lock Tubes	Fisher	05-402-25
ESI-L Low Concentration Tuning mix	Agilent	G1969-85000
Fentanyl Analog Screening (FAS) Kit	Cayman Chemical	9003237, 9003286, 9003380, 9003381	kit of 250 snthetic opioids, 210 fentanyl analogs, broken up into one kit and emergent panel versions 1-4
Formic acid Optima LC/MS	Fisher	A117-50
Onyx guard column (5 x 4.6 mm)	Phenomenex	CHO-7649	guard column for C18 columns
Onyx monolithic C18 HPLC column (100 x 4.6 mm)	Phenomenex	CHO-7643	reverse phase C18 LC column
Optima grade acetonitrile	Fisher	A996-4
Optima grade methanol	Fisher	A454-4
Optima grade water	Fisher	W7-4
Pipette	Fisher	05-719-510	kit of 1-10 µL, 10-100 µL, and 100-1000 µL pipette
Pipette tips 10µL	Fisher	94060100
Pipette tips 1000µL	Fisher	94056710
Pipette tips 200µL	Fisher	94060310
Plate mixer	IKA	MS 3 D S1	IKA MS 3 digtital
Prominence LC-20 CE ultrafast liquid chromatograph	Shimadzu, Japan		equiped with DGU-20A5, LC-20AD, SIL-20AC, CTO-20A, SPD-M20A, CBM-20A, SPD-20A
timsTOF Pro	Bruker Daltonics Inc., Billerica, MA		timsTOF instrument with PASEF