Extraction of Non-Protein Amino Acids from Cyanobacteria for Liquid Chromatography-Tandem Mass Spectrometry Analysis

The extraction protocol outlined here for the analysis of NPAAs applies to analyzing any amino acids in biological samples. For guides on isolating and culturing cyanobacterial strains, one can refer to the methods presented in the study by Violi et al.³⁸. The first step in the protocol takes the sample to a point where normalization between samples against the dry weight can be achieved. The second step is cell lysis to release the analytes and can be performed using an array of techniques, including mechanical disruptions/lyses such as probe sonication as described in this protocol, freeze/thaw cycles, grinding and bead milling, and non-mechanical disruptions such as enzymatic, detergent and/or chemical lysis. Mechanical disruption is known to be advantageous over non-mechanical as it allows for a greater capacity of the sample to lyse while allowing the intracellular bonds and proteins to remain intact⁴¹, although the sample matrix may dictate the optimal method for cell lysing.

Protein precipitation is the critical third step of this protocol when extracting amino acids for analysis. TCA is the most commonly used solvent; however, perchloric acid, acetone, methyl tert-butyl ether (MTBE), methanol, and/or acetonitrile^8,42 have also been used, where each extraction solvent is purposed to extract and precipitate different substrates. The model described here extracts the NPAA BMAA and its isomers from a cyanobacterial matrix, and although a range of different solvents have been utilized, the two most common are 10% TCA in water (aqueous) and 10% TCA in acetone. In general, protein precipitation with TCA is commonly used to extract amino acids to fractionate free amino acids from the protein-bound amino acids. Additionally, TCA extraction allows for determining the total protein content, reducing contaminants for the free fraction, and reducing the activity of proteases with minimal protein degradation⁴³. The TCA extraction of the free fraction works by initially rinsing out organic-soluble substances, leaving behind proteins and insoluble compounds such as cell wall remnants in the precipitate, which is then followed by thermal hydrolysis extraction of the protein-bound amino acids (bound fraction) using a strong acid.

The fractionation of the free fraction with 10% aqueous TCA plus sonication produces the most extensive protein precipitation compared to other organic solvents⁴⁴ and the best amino acid recoveries⁴². Some studies opt to use an acid combined with an organic solvent (i.e., 10%-20% TCA in acetone⁴³) to precipitate more molecules, including smaller peptides/biomolecules, minimize protein degradation, and reduce contaminants such as salts⁴³. Another advantage of 10% TCA in acetone is its faster drying rate in the preparation of the pellet for hydrolysis, minimizing moisture residue to prevent amino acid modification. However, 10% aqueous TCA plus sonication has better extraction efficiencies of free amino acids when compared to 10% TCA in acetone⁴⁴ alone. Additionally, 10% aqueous TCA precipitation has limitations such as a long drying time, not being able to precipitate all proteins or small peptides/biomolecules, and depending on the analyte of interest (i.e., proteins or amino acids), requiring additives and reducing agents to prevent oxidization and degradation^45,46.

This protocol uses a combination of 10% aqueous TCA and 10% TCA in acetone to increase protein precipitation, ensure smaller peptides/biomolecules are precipitated, allow for faster pellet drying times, and increase extraction efficiencies, taking advantage of both solvent extraction properties. However, the combination of 10% aqueous TCA and 10% TCA in acetone can precipitate small peptides/biomolecules in the free fraction after the 10% TCA in acetone supernatant is combined with the aqueous free fraction. In this case, the precipitates should be transferred to the bound fraction for hydrolysis.

The second half of this protocol involves the release of amino acids (i.e., BMAA) from the protein pellet via acid-vapor hydrolysis at elevated temperatures in an oxygen-free environment. The predominately limiting factor to the hydrolysis step is that it is time-consuming and laborious to prepare. The overnight incubation prevents fast and time-efficient sample preparation and analysis. In addition, the quantitative transfer of the pellet from a centrifuge tube to a shell vial (step 3.4) is an arduous process that requires diligence and patience to ensure an accurate point of normalization to the dry weight. Possible issues the user may face are the incomplete transfer of the pellet, wet precipitated proteins sticking to pipette tips and the original tube, and solid particles of the pellet blocking the pipette tip. A practical suggestion in aiding an easier transfer of the pellet is to remove about 0.3-0.5 mm off the end of the pipette tip with a pair of scissors, allowing larger pellet particles to be drawn and released into the shell vial. This transfer method is common practice for protein extraction for all amino acid analyses. However, modifications to improve the efficiency and accuracy of quantitatively transferring the pellet into the shell vial should be investigated.

Two forms of hydrolysis techniques can be employed to release amino acids from their protein-bound state: liquid-phase hydrolysis and acid-vapor hydrolysis. Liquid-phase hydrolysis, which involves the addition of 6 M HCl onto the sample, which is then placed in a 110 °C oven overnight, is an alternative to the acid-vapor hydrolysis described in this protocol. Extraction techniques that employ liquid-phase hydrolysis may require further processing, such as desalination, especially if AQC derivatization is chosen, as it requires basic conditions for tagging⁴⁰. Acid-vapor hydrolysis avoids desalination and allows the user the freedom to choose the derivatization technique upon reconstitution of the final pellet before filtration. Furthermore, independent of the chosen hydrolysis method, the samples may require further processing, such as SPE for matrix purification and concentration^29,47,48, depending on the choice of derivatization technique and/or analytical analysis (e.g., reverse phase LC-MS/MS vs. hydrophilic interaction liquid chromatography [HILIC] LC-MS/MS³⁷). The reconstitution of the final pellet in 20 mM HCl in this protocol is appropriate for derivatization using PCF reagents⁴⁸ via a commercially available amino acid hydrolysis kit³⁹ (see Table of Materials). Both AQC and PCF will derivatize all the amino acids, protein, and non-protein in origin present in the sample, and the selectivity of the analytes is obtained through the use of LC-MS/MS and its combination of retention time matching and careful selection of the quantifier and qualifier MRMs³⁸.

Current hydrolysis protocols are not optimized for the release of BMAA, 2,4-DAB, and AEG but for the 22 protein amino acids based on existing methods for protein hydrolysis³², where incubation for longer than 18 h results in the degradation and modification of certain amino acids, affecting LC-MS/MS analysis and, thus, the obtained concentrations³². Beach et al.⁴⁹ investigated hydrolysis over time (0.5-120 h) to optimize hydrolysis for analyzing BMAA and the proteinogenic amino acids. They found that although there was an early rapid release of BMAA during the first 0.5 h of the hydrolysis, the BMAA levels continued to increase as the hydrolysis time increased, without degradation, even after 5 days⁴⁹. There are also still different views and questions about the true nature of bound BMAA and its isomers, with some studies suggesting that rather than BMAA bonding^50,51 or misincorporation into proteins¹⁴, BMAA-protein association is superficial⁵². However, the current consensus in the analysis of "bound" BMAA requires the validated and widely accepted hydrolysis step that breaks apart peptides/proteins to release amino acids and, thus, BMAA and its isomers.

The recovery rates of 20 protein amino acids were determined in a study by Sedgwick et al.⁴² using TCA extraction, resulting in approximately 100% recovery. In the case of BMAA and its isomers from algal matrices, studies comparing the efficiency of BMAA extraction using various extraction solvents have found 10% TCA to be the most efficient for free BMAA²⁷. The recovery rates for protein-bound BMAA extracted using liquid-phase hydrolysis were established in studies by Glover et al.²⁵ and by Faassen et al.⁵³, where the accuracy was determined by spiked recovery methods, with an average BMAA recovery rate of 108.6% and 70%, respectively. Faassen et al. described procedures for spike-recovery experiments and illustrated how the recovery differs depending on the stage of the extraction process the spike was made⁵³. Faassen et al. additionally determined the efficiency of the acid-vapor protocol presented here, with recovery rates of 83.6% for the BMAA isomers spiked prior to extraction and 68.6% for those spiked prior to hydrolysis²⁴. These recoveries, methods of extraction, and analyses were further validated by Banack²⁶, with recovery rates of 94%-106% for BMAA in a cyanobacterial matrix and a %Relative Standard Deviation (%RSD) between 5.6% and 20%, meeting the FDA criteria for accuracy⁵⁴. It is recommended that each laboratory initially establish its recovery rates and accuracy before utilizing this protocol via spike recovery experiments. The recovery methods presented in Faassen et al.⁵³ and Glover et al.²⁵ may be followed as a guide.

Alternative forms of hydrolysis techniques can be utilized instead of overnight oven hydrolysis for faster protein-bound extraction of the analyte. For example, a microwave extractor could significantly reduce the hydrolysis time from 24 h to as little as 10 min⁵⁵. Microwave hydrolysis for amino acids uses the same principles as the conventional thermal method, using a glove box to prepare and assemble the microwave vessel in an oxygen-free environment and adding 6 M HCl. Once air-tight, the vessel containing the sample undergoes microwave radiation. Microwave hydrolysis has been used to extract amino acids from various sample matrices^56,57,58; however, microwave hydrolysis has yet to be developed and validated for the analysis of BMAA.

In conclusion, 10% aqueous TCA protein precipitation is the optimal method to extract free non-protein amino acids from cyanobacterial matrices²⁷, and thermal hydrolysis provides a reliable and reproducible extraction of the protein-bound fraction. This protocol to extract the NPAA BMAA and its isomers from cyanobacteria is validated and widely accepted. Future directions to further optimize the extraction of BMAA will focus on the hydrolysis step, which is performed according to the specifications of the 22 protein amino acids. However, the extraction protocol presented here should still be considered efficient and effective for analyzing BMAA and its isomers via LC-MS/MS.