C. A. HUGHEY (1), C. M. McMinn (1), J. Phung (1);
(1) James Madison University, Harrisonburg, VA, U.S.A.
Technical Session 8 - Methods of Analysis
Tuesday, June 16
1:30–3:15 p.m.
Fiesta 3,4,6,8
A robust analytical platform for untargeted differential analysis of beer samples by liquid chromatography mass spectrometry (LC/MS) was developed for the identification of compounds unique to a particular year of production or hop. The presented methodology could also be used to identify differentially expressed compounds that result from a change in a raw ingredient, a change in the brewing process, or prolonged storage. Twenty-four single-hop India pale ales (Mikkeller brewery, Copenhagen, Denmark) from two different production years were analyzed using an Agilent 6530 q-TOF-MS in positive- and negative-ion modes. A mixture of target compounds, a QC beer and an internal standard were used to monitor changes in instrument response, retention time (RT), reproducibility, and mass error. Unique molecular features (~3,000/sample) were extracted and aligned in “Mass Profiler Professional” for both the QC and Mikkeller beers. Differential analysis of the QC beer identified days that were “out of spec.” Analysis of the Mikkeller beers afforded differentiation by hop and year of production; the latter yielded the greatest compositional differences. Features unique to each production year underwent MS/MS. A METLIN and literature search afforded tentative identifications. The QC beer was run in triplicate at the beginning and end of every day/work list. Target compounds were quantified in the QC beer to determine the intraday and interday RT RSDs, response RSDs, and mass error. Over the course of 13 months, the RT deviation was 3.0 s ±1.3 s, the mass error was routinely <5 ppm, and the interday response RSD was <15%. To demonstrate that the target compounds, most of which were flavonoids, adequately represented the response trends of compounds in the beer, target compound responses were compared to the summed response of features found in all QC samples (n = 54 for each polarity) above a given abundance threshold. This included 418 and 124 molecular features in positive- and negative-ion modes, respectively. Indeed, the two responses tracked quite well, thus demonstrating that unknown molecular features in a sample QC may be used to monitor instrument response. The use of external standards found in the QC, however, offers additional assurance that the QC beer does not change over time. Differential analyses of randomized Mikkeller IPAs run in triplicate were conducted based on hop, year of production, and hop year. Hierarchical clustering revealed that beers strongly clustered by year and not hop for both polarities, as IPAs produced with the same hop in different years did not cluster together. In an attempt to identify features responsible for this clustering, features unique to each production year were identified and targeted for MS/MS analysis. Exact masses and MS/MS spectra (when available) were searched against the METLIN Database. Tentative assignments were also made by comparison to the literature. Compounds such as 5-methylthioadenosine have been reported in aged beer (Food Chemistry 135:1284-1289, 2012) and were identified in the earlier production beer.
Christine (Chrisi) Hughey is an associate professor of chemistry at James Madison University, a predominately undergraduate institution, located in Harrisonburg, VA. She received her Ph.D. degree in analytical chemistry from Florida State University in 2002. Since 2006 she has used mass spectrometry to characterize complex food samples. In 2011, she began “beeromics” (untargeted differential analysis) out of her love for craft beer. Over time, she and her research group have developed a robust platform that can be used to answer a variety of questions of interest to brewers. At the ASBC Annual Meeting Christine aims to present this methodology in hopes of fostering industry collaborations.
View Presentation
