Analysis of mousy off-flavor compound 2-Acetyl-tetrahydropyridine using Liquid Chromatography Mass Spectrometry with Electrospray Ionization in sour beer

Abstract

Mousy off-flavor describes N-heterocycles compounds related to spoilage in the brewing industry. It has also been identified in sour beers through sensory analysis. Therefore, preventing spoilage N-heterocycles development is essential to preserve end-products and obviate economic losses. To this day, no methods or protocols have been reported to identifying mousy off-flavor compounds in a beer matrix. The main objective of this work was to develop a standardized quantification method for 2-acetyl-3,4,5,6-tetrahydropyridine (ATHP) in beer matrix, by Liquid Chromatography Mass Spectrometry with Electrospray Ionization (LC-MS-ESI). Extraction of ATHP in the samples was performed using QuEChERS (quick, easy, cheap, effective, rugged, and safe) technique. Over a dozen different potentially mousy cask-aged sour beers including other spontaneously fermented beverages were provided, based on sensory analysis, to determine the variation in ATHP levels. Results indicated ATHP was found in all the samples, ranging from 1.64 ± 0.06 to 57.96 ± 2.15 µg L⁻¹. Herein, we described our detection method of mousy-off flavor compounds which enables future research to mitigate the occurrence of such defects in fermented beverages matrix.

• •ATHP content in samples varied from 1.64 ± 0.06 to 57.96 ± 2.15 µg L⁻¹.
• •The recovery range of ATHP using LC-MS-ESI varied from 71% to 97%.
• •Basified QuEChERS salting-out procedure is applicable for ATHP extraction from beer and other fermented beverages matrices.

Graphical abstract

Specifications table

Subject area:	Food Science
More specific subject area:	Beer analytic and sample pre-treatment
Name of your method:	Quantitative analysis of 2-Acetyl-tetrahydropyridine using Liquid Chromatography Mass Spectrometry with Electrospray Ionization
Name and reference of original method:	N.A.
Resource availability:	Equipment: •Agilent 1260 Infinity II mass spectrometer equipped with Agilent 6470 Triple Quad LC-MS (Agilent technoloigies, USA) and electrospray ionization Source (ESI), •Kinetex (5 µm EVO C18 100 Å 150 × 2.1 mm) column conected with 2.1 mm i.d. guard column packed with the same material (Phenomenex, Lane Cove, NSW, Australia), •Sonificator (TS – 6000 Digital Ultrasonic Cleaner, RayTech, USA), •pH-meter (Benchtop pH-mV Meter 0–14 pH range, Sper Scientific, USA). Software: •Offline Method Editor and Mass Hunter Optimizer (Agilent, USA), •Agilent MassHunter Workstation Data Aquisition, (Agilent, USA), •Agilent MassHunter Quantative analysis (for QQQ) (Agilent, USA). Reagents: •Ammonium acetate HPLC Grade crystalline form (Fisher Chemical, USA) methanol (Fisher Chemicals, USA), •Acetonitrile HPLC/UHPLC-grade (Fisher Chemicals, USA), •Ammonium hydroxide reagent grade (Fisher Science Education, USA), •Water for MS (Fisher Chemicals, USA), •2-Acetyl-3,4,5,6-tetrahydropyridine Hydrochloride (Technical Grade) (ATHP) standard (Toronto Research Chemicals, Canada)., •2-Acetyl-3,4,5,6-tetrahydropyridine-13C2 Hydrochloride (Technical Grade) (13C2 ATHP) internal standard (Toronto Research Chemicals, Canada). Standards were in crystalline form with hydrochloride, •QuEChERS salts, which consisted of 4 g magnesium sulfate, 1 g sodium chloride (Agilent Technologies, US).

Introduction

Mousy off-flavor is a bouquet of N-heterocycles: 2-ethyl-1,3,4,5,6-tetrahydropyridine (enamine form); 2-ethyl-3,4,5,6-tetrahydropyridine (imine form) (ETHP), 2-acetyl-1,4,5,6-tetrahydropyridine (enamine form); 2-acetyl-3,4,5,6-tetrahydropyridine (imine form), acetyl-1-pyrroline (imine form) and 2-acetyl-2-pyrroline (enamine form) (APY) [1] which are related to the spoilage of beverage and commonly obtained in wild fermentation end-products, e.g., wine without sulfites, sour beer, ciders, kombucha and others [2,3]. Sensorial descriptions of mousy off-flavor include corny (sometimes referred to corn-based cheese flavor chips odor), cereal, fresh popcorn, sour bread odor, mouse urine and cracky off-flavor [4], [5], [6]. The threshold in water of ATHP, ETHP and APY is 1.6 µg kg⁻¹ 150 µg kg⁻¹ and 0.1 – 0.06 µg L⁻¹, respectively [7].In roasted barley tea, APY threshold was 0.053 µg kg⁻¹ [8]. Although APY alone is described as having nutty and buttery flavor, in complex with ATHP and ETHP envelopes mousy off-flavor in wines [1,8,9]. There is little research reporting N-heterocycles using Gas Chromatography Mass Spectrometry (GS-MS) and Liquid Chromatography Mass Spectrometry (LC-MS) techniques [1,3]. However, identification using LC-MS is applicable for ATHP only. A previous study described an ATHP method via LC-MS using a tandem with atmospheric pressure ionization source (APCI) [3]. But analytical laboratories commonly have only an electrospray ionization source (ESI). ESI source covers a broader range of analytical compounds, reflecting the economic reasons why analytical laboratories choose to have it instead of APCI source [10].

The beer matrix is a complex mix of micro- and macro-components from grains and hops, which includes organic acids, carbonyl compounds, and over six hundred flavor compounds [11,12]. Also, over 160 beer styles are described in the American Brewers Guidelines, which refers to plenty of possible modifications of the beer brewing process [13]. Lager and ales are historically the popular styles, but sour beers, which are fermented in the cask and can involve wild environmental microbes inoculation, are gaining popularity among consumers. Nowadays people have a tendency to support local craft breweries, which encourages experimental craft brews as a point of differentiation. However, sour beer have a higher risk of developing mousy off -flavor resulting in decreased sales, which sometimes are crucial for smaller breweries.

Sour beers have an even more complicated matrix than regular beer due to long fermentation in the cask, microbial activity during fermentation, and their metabolite levels, especially increased organic acids and salt content [14], [15], [16]. Nonetheless, other grain-based spontaneously fermented beverages or kombucha, which have high microbial variation and population counts, may be applicable for detecting mousy-off flavor either [17,18].

The QuEChERS method is a liquid-liquid extraction procedure that directly analyses the solution extract in samples from different matrices such as food, effluents, and biological samples [19]. Main advantages of QuEChERS include its environmentally friendliness, low cost, high accuracy and rapidity [20]. Due to intermediate polarity acetonitrile can simultaneously remove ATHP from different beer samples. The basified (pH 9–10.5) salt-assisted extraction process (salting out) initiates the protonation of ATHP imine to enamine, which is more volatile and more detectable. In addition to promoting the separation of the organic phase, the resulting acetonitrile layer is available for direct analysis by LC, without the need to filter or dry the extract. The application of this method has already been effectively demonstrated in detecting pesticides, mycotoxins and bitter compounds in hops and beer [21,22]. However, the QuEChERS application for extracting ATHP is demonstrated here, along with LC-MS-ESI for quantitation, for the first time.

Method details

Beverage samples

Traditional commercially available lager and ale beers without ATHP and sixteen suspected mousy off-flavor sour beers were selected for developing the method and its validation. All suspected mousy beers were fermented in the cask. Three samples out of sixteen were more than five years old (samples 11, 13, and 14). Additionally, three samples were selected outside of the sour beer category. Hard kombucha sample (sample 18), spontaneously fermented non-alcoholic grain beverage (sample 19) and tea beer (sample 16). Kombucha and non-alcoholic grain beverage samples were fermented with a mixed-culture microbial population and suspected to have mousy off-flavor.

Standards and mobile phase preparation

Beer matrix involves complex compounds which affects mousy off-flavor compounds detection using spectrometer analysis²³. To minimize the matrix affect an internal standard of ATHP was applied. Standards of main mousy off-flavor compound with analytical data are shown in Table 1. For LC-MS-ESI two standards: 2-acetyl-3,4,5,6-tetrahydropyridine (ATHP) and 2-acetyl-3,4,5,6-tetrahydropyridine ¹³C₂ (¹³C₂ ATHP) were selected to optimize the method in sour beer samples. Also, ATHP derivative 2-ethyl-3,4,5,6-tetrahydropyridine (ETHP) was evaluated through LC-MS-ESI. However, the abundance was low compared to ATHP and thus not taken for further investigations in the beer matrix. APY was not selected because it is a volatile compound and requires the use of another instrument (i.e. GC)Moreover, APY without ATHP and ETHP do create favorable aroma [1,8,9].

Table 1.

Analytical data for ATHP and ¹³C₂ ATHP standards.

	Name	Abbreviation	Chemical formula	Molecular weight (g/mol)	Specific ions (m/z)	Comment
1.	2-Acetyl-3,4,5,6-tetrahydropyridine	ATHP	C₇H₁₁NO	125	56/84/98	standard
2.	2-Acetyl-3,4,5,6-tetrahydropyridine	13C2 ATHP	C₇H₁₁NO	127	56/84/100	internal standard (IS)

*Note: Standards were in crystalline form with hydrochloride. The hydrochloride dissociates in the solvent and is not included in the molecular weight formula.

Stock solutions of 0.1 mg mL⁻¹ standard and internal standard was prepared in 100% methanol and stored in amber glass vial at -15 °C before using for the analysis. On the same day of analysis working solutions were prepared using UltraP water. Serial dilution technique was applied to reach a desired concentration. To compare the ATHP stability in water and methanol-based stock solutions same concentrations were prepared, and after a period analytical data was collected to evaluate peak area abundance. ATHP in water had lower stability than ATHP in methanol-based stock solutions (Fig. 1). In addition, experimental water-based stock solution was kept in the 4 °C and ATHP stability decreased 1.3-fold in three days and 1.7-fold after 7 days. In comparison, methanol-based stock solutions maintained stable ATHP and ¹³C₂ ATHP after 2-month period, which were kept in -15 °C.

Fig. 1.

Changes in peak areas of 2-acetyl-3,4,5,6-tetrahydropyridine ATHP and 2-acetyl-3,4,5,6-tetrahydropyridine ¹³C₂ (¹³C₂ ATHP), as internal standard stability, in water and methanol-based stock solutions over a period.

Initially, for the mobile phase ammonium acetate 10 mM was used as solvent A. The pH correction to 9.0 was made with 20% NH₄OH. Methanol was selected as Solvent B. ATHP is polar compound and adaptation of ‘like dissolves like’ principle was implied. Overview of stock, working and mobile phase solvents are shown in Fig. 2.

Fig. 2.

Methanol-based stock solution and water-based working solution preparation and mobile phase solvent A and B.

0.1mg mL⁻¹ ATHP standard stock solution preparation: weighted 1 mg of ATHP was dissolved up to 10 mL with 100% methanol.

10µg mL⁻¹ ATHP standard working solution preparation: 1 mL of 0.1 mg mL⁻¹ ATHP of stock solution was diluted up to 10 mL with UltraP water.

0.1mg mL^{−1 13}C₂ ATHP standard stock solution preparation: weighted 1 mg of ¹³C₂ ATHP was dissolved up to 10 mL with 100% methanol.

10µg mL⁻¹³C₂ ATHP standard working solution preparation: 1 mL of 0.1 mg mL^{−1 13}C₂ ATHP of stock solution was diluted up to 10 mL with UltraP water.

10mM ammonium acetate: weighted 0.77 g of ammonium acetate was dissolved with UltraP water up to 800 mL, the pH correction to 9.0 was made with 20% NH₄OH. The solution was diluted up to 1 L.

Stock solutions can be kept for at least 2 months (at -15 °C). Mobile phases and calibration curve can be made up in advance. However, on the day of analysis, it is important to make fresh internal standard and working solutions.

Equipment and conditions

LC-MS-ESI

Agilent 1260 Infinity II mass spectrometer equipped with Agilent 6470 Triple Quad LC/MS (Agilent technoloigies, USA) and electrospray ionization Source consisting of a binary pump, autosampler and column oven was used. Data aquisition and processing were performed using Agilent MassHunter software.

LC separation

A 100 µl aliquot of sample was injected and chromatographed on a Kinetex (5 µm EVO C18 100 Å 150 × 2.1 mm) column connected with 2.1 mm i.d. guard column packed with the same material (Phenomenex, Lane Cove, NSW, Australia). 10 mM ammonium acetate (pH 9.0) was prepared as solvent A, and 100% methanol as solvent B. 20% ammonium hydroxide was used for pH corrections. For separation isocratic conditions were implemented: solvent A was selected at 40%, solvent B at 60%. A flow rate of 500 µl mL⁻¹ and run time was set for 3 min for lager type beer and 6 min for ale type beer.

MS conditions

A positive ion mode in electrospray ionization was used to record generated multi-reactant-monitoring (MRM) of the ATHP ion. Negative ion mode was attempted but no useful ion transitions were identified. After different variations gas temperature was set for 300 °C; gas flow 10 L min⁻¹; nebulizer 45 psi; shealth gas temparature 250 °C; shealth gas flow 1 L min⁻¹; capillary positive and negative voltage was set to 1500 V and nozzle positive and negative voltage was set to 500 V.

For quantification ATHP and ¹³C₂ ATHP were detected by multiple reaction monitoring (MRM). Fragmentation was set for 135 after tested coarse range from 20 to 200, and collision energy was set to 16 V. Mass transitions, m/z 126.1 → m/z 98.0 (qualifier), 84.0 (quantifier), and 56.0 (qualifier) for ATHP, and m/z 128.1→ m/z 100.0 (qualifier) and 84.0 (quantifier) and 56.0 (qualifier) for ¹³C₂ ATHP. For all transitions the cell accelerator voltage was set to 5 V.

Set of quatification and method detection limit

The limit of quantitation (LOQ) was defined as the lowest fortification level attempted, and the method detection limit (MDL) was calculated by the following equation: MDL=Student t-value x Standard Deviation [24].The 10 µg mL⁻¹ ATHP working solution was serially diluted to make 1.5, 0.15, 0.0015 and 0.00015 µg mL⁻¹ solutions. An internal standard spiking solution was prepared by taking a 100-µL aliquot of the 0.1 µg mL^{−1 13}C₂ ATHP working solution. LOQ were determined to be 0.0015 µg mL⁻¹ in UltraP water (Fig. 3). The result of MDL was calculated to be 0.5 µg L⁻¹.

Fig. 3.

ATHP of 0.0015 µg mL⁻¹ in UltraP water with internal standard 0.1 µg mL^{−1 13}C₂ ATHP.

Calibration curve

Calibration curve was prepared with standard addition containing ATHP at 0.0015, 0.005, 0.01, 0.05, 0.1, 0.3, 0.5 µg mL⁻¹ and ¹³C₂ ATHP at a constant concentration of 0.1 µg mL⁻¹ A calibration curve was constructed by fitting a linear regression line (no weighting) to a set of calibration data with the ion response ratio (peak ratio) of ATHP and ¹³C₂ ATHP against ATHP concentration (Fig. 4).

Fig. 4.

Calibration curve of ATHP in UltraP water.

Sample treatment and method validation

QuEChERS for extractions

Two decades ago, the QuEChERS methods were developed for analyzing pesticide residues and other chemical contaminants from agricultural materials, including food and beverages¹⁹. QuEChERS is a procedure based on principles involving intermediate polar solvent acetonitrile, contributing to salting-out extraction. Liquid-liquid extraction extracts metabolites and hydrophilic compounds to acetonitrile (ACN), which can be used for further sample clean-up or direct analysis [19,25]. Following the procedure, a freshly opened beer sample was sonicated with a de-gas function for 15 min. Further, 10 mL aliquot was basified with ammonium hydroxide to reach pH 9 (depending on the sample in the range between 100 and 175 µl). The sample was mixed with can and vortexed for 1 min. QuEChERS salts, consisting of 4 g magnesium sulfate and 1 g sodium chloride, were added to the beer and ACN solution. The mixture was vortexed for 1 min and centrifuged for 5 min at 10.000 rpm. The upper layer was collected for further ATHP analysis. 1 mL of upper layer extract was spiked with 0.1 µg mL⁻¹ ATHP internal standard, basified with 10 µl of 20% NH₄OH and diluted up to 4 mL with UltraP water. The preparation flow diagram is shown in Fig. 5.

Fig. 5.

Beer and other fermented beverages sample preparation using QuEChERS for ATHP analysis.

Method validation and effect of a beer matrix

A recovery test assessed the ‘ethod's accuracy and reproducibility using lager and ale-type beer without ATHP. The pH of lager and ale beer was approximately 4–4.5. Before the salting-out procedure, beer samples were fortified by 0.0015 (n = 7), 0.015 (n = 3) and 0.15 (n = 3) µg mL⁻¹ of ATHP and 0.1 µg mL^-1 internal standard. The following equation calculated recovery of the ATHP: Recovery (%) = (RR₂ x 100)/ RR₁, where RR₁ – relative response of 0.0015 µg mL⁻¹ standard from the calibration curve, RR₂ – relative response of spiked beer sample. Results are presented in Table 2. Recovery of lager beer was from 78.2 to 97.5% and ale beer from 66.3% to 84.7%.

Table 2.

ATHP recovery (%) for lager and ale type beer.

ATHP concentration, µg mL−1	0.15 (n = 3)	0.015 (n = 3)	0.0015 (n = 7)
Lager	78.2 ± 3.9	80.2 ± 8.1	97.5 ± 27.8
Ale	71.4 ± 2.6	66.25 ± 10.1	84.7 ± 12.6

The results from the developed methods for ATHP using LC-MS-ESI are presented in Fig. 6 and Table 3. The normality test shows that selected samples vary in normal distribution. Two samples, one beer (sample 2) and non-beer (sample 16) did not reach LOQ while other potentially mousy samples had measurable amounts of ATHP ranging from 1.76 to 57.96 µg L⁻¹ (samples 1–15, and 17). Samples (11, 13 and 14) with higher ATHP concentration >40 µg L⁻¹ were older than five years old and continuously changed chemically in casks or bottles. The parameters that can impact the development of ATHP include a wide range: sugars, dissolved oxygen, amino acids, in particular lysine, proline, ornithine content, and microbial activity [4,[26], [27], [28]]. Brettanomyces bruxelensis is the most prevalent in the last stages of sour beer production in the cask and can be found in beer bottles after five years [18,29,30]. B. bruxellensis can partly develop ATHP through metabolic pathways l-lysine [29]. However, lactic acid bacteria (LAB) and acetic acid bacteria (AAB) may also be responsible⁴. Compared to suspected mousy wines, the variation in results is similar. ATHP concentration in eight suspected mousy wine samples using LC-MC-APCI varied from 0.52 to 26.45 µg L⁻¹³.

Fig. 6.

Normality test (QQ plot) and ATHP concentration in tested samples. Different letters indicate significant differences (one-way ANOVA and Tukey's HSD test, p < 0.05).

Table 3.

ATHP concentration in suspected mousy samples.

Sample	ATHP conc., µg L−1	Sample	ATHP conc., µg L−1	Sample	ATHP conc., µg L−1	Sample	ATHP conc., µg L−1
1	13.25 ± 0.98d	6	11.16 ± 0.90d	11	57.96 ± 2.15 g	16	< LOQ
2	< LOQ	7	8.07 ± 0.76c	12	2.44 ± 0.17ab	17	1.64 ± 0.06a
3	1.75 ± 0.54a,b	8	8.68 ± 1.87c	13	52.96 ± 1.92f	18	2.44 ± 0.17a,b
4	3.48 ± 0.45a,b	9	7.55 ± 0.37c	14	41.02 ± 1.39e	19	3.70 ± 0.29b
5	2.93 ± 0.61a,b	10	2.42 ± 0.20a,b	15	2.44 ± 0.17a,b

* Different superscript letters indicate significant differences (one-way ANOVA and Tukey's HSD test, p < 0.05).

Samples 1- 15 and 17 are suspected mousy sour beer samples; samples 16 - tea beer; sample 18 - hard kombucha; sample 19 - spontaneously fermented non-alcoholic grain beverage.

Limitations and conclusions

Beer has over 160 different styles and the varying matrices can differently suppresses ATHP ¹³. Having a reliable and reproducible method to measure ATHP enables future research to track and mitigate its occurrence in fermented beverages matrix. Herein, we developed an extraction and quantification method of ATHP using LC-MS-ESI with a recovery range from 71% to 97%. The salting-out extraction showed promising application, and by incorporating additional clean-up stages, e.g., solid phase extraction (SPE), the recovery may be improved. The sample size of this work is modest, and more investigation may be needed for other spontaneously fermented beverage matrices. Another limitation involves the unknown sensory threshold of ATHP, ETHP and APY in the beer matrix (i.e. higher concentrations may spoil beer, smaller concentrations can participate in favorable pallet of sour beer flavor), a study incorporating sensory panels along with analytical chemistry would help elucidate this relationship.

Funding

This work was supported by the Lallemand Inc.

Please declare any financial interests/personal relationships which may be considered as potential competing interests here.

CRediT authorship contribution statement

Paulina Martusevice: Conceptualization, Methodology, Visualization, Writing – original draft. Xueqi Li: Software, Validation, Data curation, Conceptualization, Methodology, Project administration. Matt Hengel: Supervision, Software, Methodology, Validation, Data curation, Investigation, Resources. Selina C. Wang: Supervision, Writing – review & editing, Resources. Glen Fox: Supervision, Conceptualization, Resources, Writing – review & editing, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

• The authors do not have permission to share data.