Assessing the impact of sample preparation techniques on CSIA-AA δ13C and δ15N results

Abstract:

Amino acids (AA) require derivatization in order to become volatile for gas chromatography, and cation exchange chromatography (CEC) and phosphate buffered saline (P-Buffer) are also often used to clean and prepare samples for GC-IRMS (Ueda et al. 1989). While these techniques are common practice in sample preparation for compound-specific isotope analysis of AA (CSIA-AA), their relative impact on CSIA-AA results remains unknown. In this study, three treatments were applied to an AA standard (i. derivatization, ii. derivatization and CEC, iii. derivatization, CEC and P-Buffer) to assess their relative impacts on δ¹³C and δ¹⁵N chromatography, permille values, and reproducibility – overall and between AA. Each treatment had three replicate samples and each sample was injected into a GC-IRMS (TRACE 1310 with a BPX5 column, Delta V) in triplicate. Instrument drift was accounted for by alternating injection triplicates by sample type, as well as by applying a linear drift correction. A correction was also applied to δ¹³C values to account for the addition of C in derivatization (Silfer et al. 1991). One-way and two-way ANOVA were used to assess differences using a 95% CI. The samples are in queue for injection in δ¹⁵N mode, but the δ¹³C injections yielded some notable results. Overall sample mean and reproducibility did not differ between treatments (P= .979 and P = .082, respectively), and though reproducibility was similar among AA and treatments (P = .635), significant differences were detected in the δ¹³C values among AA and treatments (P = .016). Several significant differences in chromatography between treatments are noted in Figure 1. The differences identified in this study highlight the importance of testing the assumptions that underly CSIA-AA methods (ex. H₀: results will be the same, regardless of sample preparation techniques) so that such differences might be accounted for in future studies.

Figure 1. Mean chromatography for each treatment in δ¹³C mode. Peak amplitude and area differed significantly between treatments (P < .001 in both cases), each decreasing in value with increased sample preparation. Peak background was significantly less for the solely derivatized samples (P = .013), relative to the more chemically processed samples. Run-times of threonine, serine, valine, leucine and isoleucine were similar between treatments; but run-times of lysine, tyrosine, phenylalanine, glutamic acid and glycine were significantly different (P < .001).