Examination of amino acid hydrogen isotope measurements of scalp hair for region-of-origin studies

• NCJ Number: 306999
• Journal: Rapid Communications in Mass Spectrometry Volume: 37 Issue: 4 Dated: 2023
• Author(s): Christy J. Mancuso; James R. Ehleringer; Seth D. Newsome
• Date Published: February 2023
• Grant Number(s): 15PNIJ-21-GG-04157-SLFO
• Agencies: NIJ-Sponsored
• Publication Type: Research (Applied/Empirical)
• Award: Amino Acid Specific Stable Isotope Ratios in Human Keratin Tissue for Region-of-Origin and Residency Determination

Annotation

This article presents research into the use of hydrogen isotope analyses of ammino acids in diet and tap water to improve geolocation models for human and wildlife forensics.

Abstract

Hydrogen isotope (δ²H) analysis of keratinaceous bulk tissues has been used in forensic science to reconstruct an individual's travel history or determine their region-of-origin. Here, the authors use a compound-specific approach to examine patterns of individual amino acid δ²H values in relation to those of local tap water, bulk scalp hair tissues, and region-of-origin. The authors measured δ²H values of amino acids in anonymously collected scalp hair (n = 67) and tap water from 28 locations in the United States. Samples were hydrolyzed into their constituent amino acids, derivatized alongside in-house reference materials, and analyzed in triplicate using a GC-C-IRMS system. Non-essential amino acid (AA_NESS) δ²H values and their corresponding tap water samples varied systematically across continental regions. Hydrogen isotope values of alanine, glutamic acid, and glycine were significantly correlated with tap water and an estimated 42%–51% of the hydrogen atoms in these AA_NESS were derived from tap water. The authors used linear discriminate analysis (LDA) to explore regional patterns in scalp hair bulk tissue and amino acid δ²H values. For the model that included AA_NESS data, 87% of the variance was explained by the first linear discriminant axis (LD1), and was driven by bulk hair tissue, alanine, and proline. This model had an overall 72% successful reclassification with samples from the south and northwest regions reclassifying correctly 92% and 78% of the time, respectively. For the model that included AA_ESS data, LD1 explained 81% of the variation and was driven bulk hair, threonine, valine, phenylalanine, and isoleucine. The overall reclassification rate for the model that included AA_ESS was 70%. The authors’ findings suggest that δ²H analyses of AA_NESS and AA_ESS could help improve geolocation models for human and wildlife forensics by simultaneously providing information about both dietary and tap water inputs of hydrogen to tissue synthesis. (Published Abstract Provided)