MSACL 2023 Abstract

Introduction:
Analytical sensitivity is a prerequisite for robust and reliable assays and is itself dependent on method parameters used by mass spectrometers (i.e., ion source and compound-dependent settings). Despite this importance, optimal settings are often considered de facto for a particular mode of instrument operation instead of being determined empirically. Care must be taken when developing new or transferring established assays onto alternate mass spectrometers as particular settings can be sub-optimal depending on the make, model, and application. Herein, we demonstrate an injection-based routine to screen method parameters on seemingly related mass spectrometers, noting key differences in optimal settings to achieve the best assay performance.

Methods:
Samples were analyzed by LC-MS/MS on an ARIA™ TLX-4 system (Thermo Scientific) coupled to either a Triple Quad™ 5000, 5500, or 7500 mass spectrometer (SCIEX). Both the 5000 and 5500 models used a Turbo V™ ion source equipped with an electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI) probe. Similarly, the 7500 system used an OptiFlow™ Pro ion source configured to operate in either ESI or APCI mode, with the latter requiring the removal of the E Lens™ Technology.

Clinically-relevant assays utilizing selected reaction monitoring (SRM) were chosen as representatives for positive ESI (thyroglobulin – Tg), negative ESI (Reverse T3 – rT3), and positive APCI (Testosterone). The LC setup was different between assays; however, separation was generally performed using reversed-phase chromatography at ≥ 1 mL/min with a total cycle time ≤ 4 min.

The relative performance (i.e., peak area and signal-to-noise ratio) for SRM transitions specific to each analyte were compared across replicate injections of both reference standards and extracted specimens. Method parameters were screened using a custom AutoIt program to create acquisition methods with the desired range of values and step size for any combination of settings. Source variables considered were nebulizer gas (GS1), heater gas (GS2), curtain gas (CUR), collision gas (CAD), IonSpray™ voltage (ISV), and ion source temperature (TEM), with nebulizer current (NC) optimized for APCI only. Transition-specific variables included entrance potential (EP), collision energy (CE), and collision cell exit potential (CXP). Additionally, declustering potential (DP) was ramped for 5000 and 5500 systems, while the analogous Q0 dissociation in simple mode (Q0DS) was specific only to 7500 models.

Results:
On 5000/5500 systems, positive ESI preferred high GS1 (≥ 40 psi) and ISV (≥ 4000 V) settings for the tryptic peptide specific for Tg protein. Negative ESI for the iodine-containing, small molecule rT3 was less dictated by GS1 (≥ 20 psi) and instead dependent on ISV (≤ -4000 V). Conversely, both ESI modes on the 7500 model were optimized with low settings for GS1 (10 to 40 psi) and ISV (positive: 1500 to 2500 V, negative: -1500 to -3000 V), which is attributed to the E Lens™ Technology improving desolvation by increasing the field strength experienced by ESI droplets. Testosterone by positive APCI mode had markedly similar optima for GS1 (≥ 50 psi) and NC (≥ 3 µA) on all systems.

Higher values for TEM generally increased analyte peak area, along with an especially noticeable increase in background noise on the 7500 system. GS2 is only used with ESI modes and produced minor differences on all systems regardless of the values tested (20 to 90 psi). As expected, increasing curtain gas decreased signal intensity for all analytes and primarily served as a protective barrier from contamination of the source and ion optics.

The CAD parameter specific to a method was ramped along with transition-specific CE to determine the best conditions for analyte fragmentation, which optimized to similar values regardless of the instrument model. Additionally, values for EP and CXP did not improve assay performance regardless of the settings used. Unique to the 7500 systems, Q0DS was found to attenuate signal intensity with increasing values (i.e., in-source fragmentation), similar to DP on 5000/5500 models. With the optimal Q0DS value enabled, both ESI modes demonstrated similar response compared to acquisition with this parameter disabled. Uniquely, positive APCI showed a four-fold enhancement in signal with Q0DS enabled (40 V), likely due to enhanced ion declustering.

Conclusion:
Injection-based optimization of method parameters used by mass spectrometers provided a simple means to ensure assay performance for routine clinical analysis.