Analysis of Cereulide in Reconstituted Infant Formula Using Triple Quadrupole LC-MS/MS

Significance of the topic

Reliable detection and quantification of cereulide, a heat-stable emetic toxin produced by Bacillus cereus, is critical for food safety—particularly for infant formula where small exposures can cause acute illness. The European Food Safety Authority (EFSA) has set an acute reference dose (ARfD) for infants and identified 0.054 µg/L in reconstituted formula as a concentration that may exceed safety thresholds, driving the need for analytical methods capable of accurate measurement at sub-µg/L levels in complex, lipid- and protein-rich matrices.

Objectives and overview of the study

The study compared two sample-preparation strategies for quantifying cereulide in reconstituted infant formula using triple quadrupole LC-MS/MS: a procedure based on ISO 18465 and a QuEChERS-based extraction. Objectives were to evaluate sensitivity, recovery, reproducibility, and matrix effects at concentrations relevant to EFSA guidance, and to demonstrate suitability of the workflow for routine monitoring.

Methodology

Samples: Commercial infant formula reconstituted per manufacturer instructions.

Sample preparation approaches:

• ISO 18465-based method: 2.5 mL sample mixed with 30 mL acetonitrile, shaken 1 hour, centrifuged (3,000 rpm, 5 min), supernatant filtered (PTFE 0.45 µm) prior to LC-MS/MS.
• QuEChERS method: 2 mL sample mixed with 8 mL water, 150 µL formic acid and 10 mL acetonitrile; shake 10 s, add extraction salt packet, hand-shake 10 s, shake 5 min, centrifuge (3,000 rpm, 5 min) and use supernatant for analysis. This approach is faster and uses dedicated extraction salts to partition matrix components.

Calibration and quantification:

• External calibration curve from 0.002 to 0.1 µg/L (linear, R² ≈ 0.9997); calibration accuracies 98.1–105.1%.
• Quantifier transition used was m/z 1170.70 → 314.15 (other qualifiers: 172.05, 357.30, 499.10).

Used Instrumentation

• UHPLC: Nexera X3 (Shimadzu).
• LC column: Shim-pack Scepter C18-120 (50 mm × 2.1 mm I.D., 1.9 µm).
• Mass spectrometer: LCMS-8060RX triple quadrupole (Shimadzu), ESI positive mode, MRM acquisition.
• Key LC conditions: mobile phase A = 10 mM ammonium formate/0.1% formic acid in water; mobile phase B = acetonitrile; flow 0.4 mL/min; gradient 80→95% B (0–6 min), 95% (6–10 min), revert to 80% (10.01–13 min); column 40°C; injection 5 µL; flow switching valve to introduce flow to MS between 3–8 min.
• Key MS settings: interface +1 kV, focus +2 kV, nebulizing 2.0 L/min, drying 5.0 L/min, heating 15.0 L/min, interface and DL at 300°C, heat block 500°C.

Main results and discussion

Spike-and-recovery tests were performed for a target concentration equivalent to 0.054 µg/L in reconstituted formula (EFSA relevance). Key findings:

• Recovery (n=3): ISO method = 90.6% (%RSD 9.7); QuEChERS = 100.1% (%RSD 2.3).
• Post-spike (matrix effect) recoveries were 108.7–109.3% for both methods, indicating modest ionization enhancement from the matrix.
• Final prepared vial concentrations after extraction for the spiked samples were 0.0045 µg/L (ISO method) and 0.0108 µg/L (QuEChERS), reflecting dilution/cleanup factors; both were quantifiable against the calibration range.
• Calibration linearity, low limits of quantification, and acceptable accuracy demonstrate that both workflows can reliably quantify cereulide at or below the EFSA-derived threshold without using an internal standard in this study.

Discussion points:

• The QuEChERS approach provided higher apparent recovery and markedly better precision, likely due to more efficient extraction/partitioning and cleaner extracts for this lipid- and protein-rich matrix.
• Matrix-induced ion enhancement observed in post-spike tests argues for routine assessment of matrix effects and, for robust routine use, inclusion of an isotopically labelled internal standard to correct for ionization variability.
• ISO 18465 provides a validated framework tailored to emetic toxin analysis; QuEChERS delivers a pragmatic alternative for higher throughput with comparable or improved analytical performance in this matrix.

Benefits and practical applications

• Both methods enable sensitive quantification of cereulide at concentrations relevant to infant safety (EFSA criteria).
• QuEChERS is faster and produced better precision in this comparison, making it attractive for routine screening and monitoring laboratories handling many samples.
• The LC-MS/MS setup (triple quadrupole MRM) provides specificity and sensitivity necessary for regulatory surveillance and product safety testing of infant formula and other challenging food matrices.

Future trends and potential uses

• Adoption of isotopically labelled cereulide standards to improve quantification accuracy and correct matrix effects in routine QA/QC workflows.
• Further method harmonization and inter-laboratory validation (including ring trials) to align QuEChERS adaptations with ISO 18465 performance criteria.
• Automation and miniaturization of clean-up steps (e.g., dispersive SPE variants) to increase throughput while maintaining or improving recovery in lipid-rich matrices.
• High-resolution MS as a complementary tool for confirmatory identification in complex cases, and development of certified reference materials for method calibration and validation.

Conclusion

This study demonstrates that both an ISO 18465-based extraction and a QuEChERS-based workflow coupled to triple quadrupole LC-MS/MS can accurately quantify cereulide in reconstituted infant formula at concentrations relevant to EFSA risk thresholds. The QuEChERS method achieved higher recovery and superior reproducibility in this comparison, but routine implementation should include matrix-effect controls and preferably an isotopically labelled internal standard to ensure robust quantitative performance across diverse sample lots.

References

1. European Food Safety Authority (EFSA), Eskes C, et al. Rapid risk assessment on acute reference dose (ARfD) of cereulide in infants and information on acute consumption of infant formulae. EFSA Journal, 2026.
2. ISO 18465. Microbiology of the food chain - Quantitative determination of emetic toxin (cereulide) using LC-MS/MS. 2017.