Comprehensive profiling of ceramides in human serum by liquid chromatography coupled to tandem mass spectrometry combining data independent/dependent acquisition modes.

Ceramides are sphingolipids with a structural function in the cell membrane and are involved in cell differentiation, proliferation and apoptosis. Recently, these chemical species have been pointed out as potential biomarkers in different diseases, due to their abnormal levels in blood. In this research, we present an overall strategy combining data-independent and dependent acquisitions (DIA and DDA, respectively) for identification, confirmation, and quantitative determination of ceramides in human serum. By application of liquid chromatography-tandem mass spectrometry (LC-MS/MS) method in DIA mode we identified 49 ceramides including d18:1, d18:0, d18:2, d16:1, d17:1 and t18:0 species. Complementary, quantitative determination of ceramides was based on a high-throughput and fully automated method consisting of solid-phase extraction on-line coupled to LC-MS/MS in DDA to improve analytical features avoiding the errors associated to sample processing. Quantitation limits were at pg mL-1 level, the intra-day and between-days variability were below 20 and 25 %, respectively; and the accuracy, expressed as bias, was always within ±25 %. The proposed method was tested with the CORDIOPREV cohort in order to obtain a qualitative and quantitative profiling of ceramides in human serum. This characterization allowed identifying d18:1 ceramides as the most concentrated with 70.8% of total concentration followed by d18:2 and d18:0 with 13.0 % and 8.8 %, respectively. Less concentrated ceramides, d16:1, d17:1 and t18:0, reported a 7.1 % of the total content. Combination of DIA and DDA LC-MS/MS analysis enabled to profile qualitative and quantitatively ceramides in human serum.

Combining data acquisition modes in liquid-chromatography-tandem mass spectrometry for comprehensive determination of acylcarnitines in human serum.

Acylcarnitines (ACs) are metabolites involved in fatty acid ß-oxidation and organic acid metabolism. Metabolic disorders associated to these two processes can be evaluated by determining the complete profile of ACs. In this research, we present an overall strategy for identification, confirmation, and quantitative determination of acylcarnitines in human serum. By this strategy we identified the presence of 47 ACs from C2 to C24 with detection of the unsaturation degree by application of a data-independent acquisition (DIA) liquid chromatography-tandem mass spectrometry (LC-MS/MS) method. Complementary, quantitative determination of ACs is based on a high-throughput and fully automated method consisting of solid-phase extraction on-line coupled to LC-MS/MS in data-dependent acquisition (DDA) to improve analytical features avoiding the errors associated to sample processing. Quantitation limits were at pg mL-1 level, the intra-day and between-day variability were below 15-20%, respectively; and the accuracy, expressed as bias, was always within ± 25%. The proposed method was tested with 40 human volunteers to determine the relative concentration of ACs in serum and identify predominant forms. Significant differences were detected by comparing the ACs profile of obese versus non-obese individuals.

Vitamin D3 levels in women and factors contributing to explain metabolic variations.

The elucidated metabolism of vitamin D3 in humans has been the support to explain the high involvement of this liposoluble vitamin in physiological functions. Clinical studies have associated levels of vitamin D3 metabolites with several disorders. Despite this knowledge, there is a controversy regarding the estimation of deficiency and the physiological and supraphysiological levels of vitamin D3 metabolites. The association between serum concentrations of vitamin D3 metabolites and several potentially influential factors (namely, age and anthropometric, seasonal, spatial and metabolic factors) is analyzed in this study. For this purpose, 558 women were recruited and interviewed in several Spanish provinces before blood sampling. Serum vitamin D3 and its metabolites were determined using an SPE-LC-MS/MS platform. The concentration range for vitamin D3 was 1.7-21.1 nmol/L and was influenced by body mass index (BMI), waist-to-hip ratio (WHR) and seasonal period. 25-hydroxyvitamin D3 levels were within 4.8-147.2 nmol/L and were related to WHR, season, latitude and calcium intake. The range of 24,25-dihydroxyvitamin D3, 0.3-15.0 nmol/L, was associated to BMI, WHR, season, latitude and calcium intake. Finally, energy intake influenced the vitamin D 25-hydroxylase through the 25-hydroxyvitamin D3/vitamin D3 ratio, which regulates the synthesis of the circulating form. According to these results, it is worth emphasizing the relevance of all these factors to explain the variability in serum levels of vitamin D3 and its metabolites. All these factors should be considered in future studies assessing the alteration of vitamin D3 metabolism.

Lyophilization as pre-processing for sample storage in the determination of vitamin D3 and metabolites in serum and plasma.

Determination of vitamin D levels in human biological specimens has gained a high relevance over the last decades, essentially because low levels have been associated with several biological disorders. In fact, vitamin D deficiency has become a worldwide health concern covering all ages and genders. The storage of biofluids has to be considered for determination of vitamin D and metabolites in order to fully preserve matrices status. This study attempts to evaluate lyophilization of serum and plasma as a pre-processing step for sample storage prior to quantitative analysis of vitamin D3 and its main hydroxylated metabolites -25(OH)D3, 24,25(OH)2D3 and 1,25(OH)2D3. The protocol including sample lyophilization was characterized in terms of analytical features and compared to the same method, based on SPE-LC-MS/MS, without lyophilization. Sensitivity, precision and accuracy were not affected when we operated with lyophilized serum and plasma and results provided by a set of twenty-four serum samples from DEQAS (Vitamin D External Quality Assessment Scheme) were in agreement with reported concentrations for 25(OH)D3 and 1,25(OH)2D3. A stability study programmed for 9 months allowed ensuring that the concentration of vitamin D3 and metabolites in lyophilized serum and plasma stored at room temperature was not affected during this period. This research has demonstrated that the quantitation of target metabolites is not under the influence of lyophilization. Therefore, including lyophilization prior to analysis could reduce shipment and storage costs, avoid delays of sample processing, and increase the stability of the target analytes due to an effective quenching process.

Optimization of a MALDI-Imaging protocol for studying adipose tissue-associated disorders.

Matrix-assisted laser desorption ionization (MALDI) imaging mass spectrometry (IMS) is increasingly recognized for its potential in the discovery of novel biomarkers directly from tissue sections. However, there are no MALDI IMS studies as yet on the adipose tissue, a lipid-enriched tissue that plays a pivotal role in the development of obesity-associated disorders. Herein, we aimed at developing an optimized method for analyzing adipose tissue lipid composition under both physiological and pathological conditions by MALDI IMS. Our studies showed an exacerbated lipid delocalization from adipose tissue sections when conventional strategies were applied. However, our optimized method using conductive-tape sampling and 2,5-dihydroxybenzoic acid (DHB) as a matrix, preserved the anatomical organization and minimized lipid diffusion from sample sections. This method enabled the identification of a total of 625 down-regulated and 328 up-regulated m/z values in the adipose tissue from a rat model of extreme obesity as compared to lean animals. Combination of MALDI IMS and liquid chromatography (LC)-MS/MS data identified 44 differentially expressed lipid species between lean and obese animals, including phospholipids and sphingomyelins. Among the lipids identified, SM(d18:0_18:2), PE(P-16:0_20:0), and PC(O-16:0_16:1) showed a differential spatial distribution in the adipose tissue of lean vs. obese animals. In sum, our method provides a valuable new tool for research on adipose tissue that may pave the way for the identification of novel biomarkers of obesity and metabolic disease.

Dry sweat as sample for metabolomics analysis.

Sweat is gaining popularity in clinical metabolomics as this biofluid is non-invasively sampled and its composition is modified by several pathologies. There is a lack of standardized strategies for collection of human sweat. Most studies have been carried out with fresh sweat collected after stimulation. A promising and simple alternative is sampling dry sweat by a solid support impregnated with a suited solvent. This research was aimed at comparing the metabolomics coverage provided by dry sweat collected by two solid supports (gauzes and filter papers) impregnated with different solvents. The dissolved dry sweat was analyzed by a dual approach: gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-tandem mass spectrometry (LC-MS/MS). Among the tested sampling strategies, filter paper impregnated with 1:1 (v/v) ethanol‒phosphate buffer resulted the combination providing the highest metabolomics coverage (tentative identification of one hundred seventy-five compounds). Dry and fresh sweat were compared by using pools from the same individuals to evaluate compositional differences. Families of metabolites such as carnitines, sphingolipids and N-acyl-amino acids, among others, were exclusively identified in dry sweat. Comparison of both samples allowed concluding that dry sweat is better for analysis of low polar metabolites and fresh sweat is more suited for polar compounds. As most of the identified metabolites are involved in key biochemical pathways, this study opens interesting possibilities to the use of dry sweat as a source of metabolite markers for specific disorders. Sampling of dry sweat could provide a standardized approach for collection of this biofluid, thus overcoming the variability limitations of fresh sweat.

Development of a qualitative/quantitative strategy for comprehensive determination of polar lipids by LC-MS/MS in human plasma.

Polar lipids, especially glycerophospholipids, constitute the main components of cell membranes and are precursors of signaling molecules in many cellular and physiological processes. For this reason, the development of methods with high capability for detection of polar lipids in biological samples is required. In this research, the objective was to develop a method for comprehensive qualitative/quantitative determination of polar lipids in plasma by a combination of acquisition methods with a triple quadrupole mass analyzer. The strategy was optimized in two steps: (a) a first step for detection of lipids by monitoring selective fragmentation patterns representative of each lipid family and (b) a second step for confirmation of lipid species by detection and identification of product ions associated with the conjugated fatty acids. The acquisition list was divided into two multiple reaction monitoring (MRM) methods to ensure the detection of all transitions with suited instrumental sensitivity according to chromatographic retention time and relative abundance in plasma. The combination of the two MRM methods allowed the detection of 398 polar lipids in plasma in 64 min. Precision, estimated as within-day variability, was below 6.8% for all determined lipid families, while between-day variability was below 24.0%. This strategy has been applied to a cohort formed by 384 individuals in order to obtain a qualitative and quantitative distribution of polar lipids in human plasma. The most concentrated lipid families in relative terms were lysophospholipids, plasmalogens, and phosphatydilcholines, with mean relative concentration of 58.0, 17.1, and 8.3%, respectively. Then, sphingomyelins and phosphatidylethanolamines reported a relative concentration of 2.0%, followed by phosphatidylserines, with 1.1%. Graphical abstract.

Determination of glycerophospholipids in vegetable edible oils: Proof of concept to discriminate olive oil categories.

Glycerophospholipids (GPLs) constitute a chemical family within the saponifiable fraction of vegetable oils. GPLs have been scarcely studied in edible oils owing to the lack of sensitive and selective analytical methods. We have developed a method for identification, confirmation and relative quantitation of GPLs in vegetable oils. The method is based on solid-phase extraction (SPE) for isolation of GPLs and determination by liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS). As proof of concept, the approach has been applied to characterize GPLs in different olive oil categories, thus revealing compositional changes, which could be explained by factors such as the quality of fruits and the extraction process. Families such as glycerophosphatidic acids and phosphatidylglycerides are remarkable because of their capability to discriminate virgin olive oils from the rest of categories. These results open a door to additional studies targeted at the identification of olive oil quality by monitoring these lipids.

Comprehensive analysis of pig feces metabolome by chromatographic techniques coupled to mass spectrometry in high resolution mode: Influence of sample preparation on the identification coverage.

Pig feces is an interesting biological sample to be implemented in metabolomics experiments by virtue of the information that can be deduced from the interaction between host and microbiome. However, pig fecal samples have received scant attention, especially in untargeted metabolomic studies. In this research, an analytical strategy was planned to maximize the identification coverage of metabolites found in pig fecal samples. For this purpose, two complementary platforms such as LC-QTOF MS/MS and GC-TOF/MS were used. Concerning sample preparation six extractant solvents with different polarity grade were tested to evaluate the extraction performance and, in the particular case of GC-MS, two derivatization protocols were compared. A total number of 303 compounds by combination of all the extractants and analytical platforms were tentatively identified. The main identified families were amino acids, fatty acids and derivatives, carbohydrates and carboxylic acids. For GC-TOF/MS analysis, the recommended extractant is methanol, while methoxymation was required in the derivatization protocol since this step allows detecting the α-keto acids, which are direct markers of the microbiome status. Concerning LC-QTOF MS/MS analysis, a dual extraction approach with methanol (MeOH) or MeOH/water and ethyl acetate is proposed to enhance the detection of polar and non-polar metabolites.

Evaluation of short-term storage prior to analysis of vitamin D3 and metabolites in human serum by liquid chromatography coupled to tandem mass spectrometry.

Vitamin D has been widely determined in clinical trials to elucidate its biochemical involvement in a great number of pathologies. The analysis of vitamin D and its hydroxymetabolites in biofluids such as serum or plasma is a challenging task due to limitations associated to the low concentrations of some metabolites (typically, dihydroxymetabolites), methodological interferences, and the low stability of the compounds. Among these limitations, efforts have been targeted at optimizing instrumental improvements to develop more sensitive and selective methods, while the stability of vitamin D and metabolites has not been exhaustively evaluated. In this research, several aspects regarding the short-term storage conditions of human serum have been studied to evaluate their influence on the determination of vitamin D3 and metabolites. An experimental plan has been applied to assess the influence of two relevant parameters: the storage temperature for a period of two months and the number of freeze-thaw cycles. The storage temperature affected in a different manner to vitamin D3 and its metabolites, being vitamin D3 and 1,25-dihydroxyvitamin D3 the two analytes more affected by this parameter. Concerning the freeze-thaw cycles, this variable must be limited to two cycles owing to its significant influence on the variability for quantitation of dihydroxymetabolites in human serum. Finally, lyophilization was also tested to check if serum concentrations of vitamin D3 and its metabolites were affected by this preprocessing step. The results revealed that only vitamin D3 experienced a decrease in serum concentration after two months, which does not constitute a real problem as vitamin D3 is not currently a crucial parameter to be determined in clinical trials due to its scant biological activity.

Study of sample preparation for determination of endocannabinoids and analogous compounds in human serum by LC-MS/MS in MRM mode.

Endocannabinoids are lipids with a key role in physiological processes such as the immune response or the metabolism. This involvement explains their association to pathologies such as cancer, obesity or multiple sclerosis. The determination of endocannabinoids constitutes a challenge for clinical laboratories due to the variety of biological matrices and the wide range of concentrations at which they can be found. This research deals with the comparison of three sample preparation strategies (viz., on-line SPE, off-line SPE for interferents removal, and protein precipitation) for subsequent LC-MS/MS analysis of 14 endocannabinoids and analogous compounds in serum. As a result, the on-line coupling between SPE and LC-MS/MS is proposed as the best approach for this determination. The proposed method allows full automation of the overall process, shortening of the analysis time, and avoidance of errors associated with sample preparation steps. The improvement in sensitivity and selectivity thus achieved allows obtaining quantification limits at the pg mL-1 level, which makes possible the application of the method for clinical studies.

Metabolomics analysis of human sweat collected after moderate exercise.

Sweat is a promising biofluid scarcely used in clinical analysis despite its non-invasive sampling. A more frequent clinical use of sweat requires to know its whole composition, especially concerning to non-polar compounds, and the development of analytical strategies for its characterization. The aim of the present study was to compare different sample preparation strategies to maximize the detection of metabolites in sweat from humans collected after practicing moderate exercise. Special emphasis was put on non-polar compounds as they have received scant attention in previous studies dealing with this biofluid. Sample preparation by liquid-liquid extraction (LLE) using extractants with different polarity index was compared to deproteination. Then, derivatization by methoxymation with subsequent silylation was compared to direct analysis of sweat extracts to check the influence of derivatization on the subsequent determination of volatile organic compounds (VOCs). 135 compounds were tentatively identified by combining spectral and retention time information after analysis by gas chromatography coupled to mass spectrometry in high resolution mode (GC-TOF/MS). Lipids, VOCs, benzenoids and other interesting metabolites such as alkaloids and ethanolamines were identified. Among the tested protocols, methyoxiamination plus silylation after LLE with dichloromethane was the best option to obtain a representative snapshot of sweat metabolome collected from different body parts after moderate exercise. Passive and active sweat pools from a cohort of volunteers (n = 6) were compared to detect compositional differences which can be explained by the sampling process and sweating induction. As most of the identified compounds are metabolites involved in key biochemical pathways, this study opens new opportunities to extend the applicability of human sweat as a source of metabolite biomarkers of pathologies or specific processes such as dehydration or nutritional unbalance.

Influence of sample preparation on lipidomics analysis of polar lipids in adipose tissue.

The main limitations of lipidomics analysis are the chemical complexity of the lipids, the range of concentrations at which they exist, and the variety of samples usually analyzed. These limitations particularly affect the characterization of polar lipids owing to the interference of neutral lipids, essentially acylglycerides, which are at high concentration and suppress ionization of low concentrated lipids in mass spectrometry detection. The influence of sample preparation on lipidomics analysis of polar lipids in adipose tissue by LC-MS/MS was the aim of this research. Two common extractants used for lipids isolation, methanol:chloroform (MeOH:CHCl3) and methyl tert-butyl ether (MTBE), were qualitatively and quantitatively compared for the extraction of the main families of lipids. The obtained results showed that each family of lipids is influenced differently by the extractant used. However, as a general trend, the use of MTBE as extractant led to higher extraction efficiency for unsaturated fatty acids, glycerophospholipids and ceramides, while MeOH:CHCl3 favored the isolation of saturated fatty acids and plasmalogens. The implementation of a solid-phase extraction (SPE) step for selective isolation of glycerophospholipids prior to LC-MS/MS analysis was assayed to evaluate its influence on lipids detection coverage as compared to direct analysis. This step was critical to enhance the detection coverage of glycerophospholipids by removal of ionization suppression effects caused by acylglycerides.

Exhaled breath condensate to discriminate individuals with different smoking habits by GC-TOF/MS.

Smoking is a crucial factor in respiratory diseases and lung inflammation, which are the reasons for high mortality worldwide. Despite the negative impact that tobacco consumption causes on health, few metabolomics studies have compared the composition of biofluids from smoker and non-smoker individuals. Exhaled breath condensate (EBC) is one of the biofluids less employed for clinical studies despite its non-invasive sampling and the foreseeable relationship between its composition and respiratory diseases. EBC was used in this research as clinical sample to compare three groups of individuals: current smokers (CS), former smokers (FS) and never smokers (NS). Special attention was paid to the cumulative consumption expressed as smoked pack-year. The levels of 12 metabolites found statistically significant among the three groups of individuals were discussed to find an explanation to their altered levels. Significant compounds included monoacylglycerol derivatives, terpenes and other compounds, the presence of which could be associated to the influence of smoking on the qualitative and quantitative composition of the microbiome.

Prostate Cancer Patients-Negative Biopsy Controls Discrimination by Untargeted Metabolomics Analysis of Urine by LC-QTOF: Upstream Information on Other Omics.

The existing clinical biomarkers for prostate cancer (PCa) diagnosis are far from ideal (e.g., the prostate specific antigen (PSA) serum level suffers from lack of specificity, providing frequent false positives leading to over-diagnosis). A key step in the search for minimum invasive tests to complement or replace PSA should be supported on the changes experienced by the biochemical pathways in PCa patients as compared to negative biopsy control individuals. In this research a comprehensive global analysis by LC-QTOF was applied to urine from 62 patients with a clinically significant PCa and 42 healthy individuals, both groups confirmed by biopsy. An unpaired t-test (p-value < 0.05) provided 28 significant metabolites tentatively identified in urine, used to develop a partial least squares discriminant analysis (PLS-DA) model characterized by 88.4 and 92.9% of sensitivity and specificity, respectively. Among the 28 significant metabolites 27 were present at lower concentrations in PCa patients than in control individuals, while only one reported higher concentrations in PCa patients. The connection among the biochemical pathways in which they are involved (DNA methylation, epigenetic marks on histones and RNA cap methylation) could explain the concentration changes with PCa and supports, once again, the role of metabolomics in upstream processes.

Identification of metabolomics panels for potential lung cancer screening by analysis of exhaled breath condensate.

Exhaled breath condensate (EBC) is one of the less employed biofluids when searching for clinical markers, despite its non-invasive sampling and the potential relationship between its composition and respiratory disease phenotypes such as lung cancer. The advanced stage at which lung cancer is usually detected is the main reason for the high mortality rate of this carcinogenic disease. In this preliminary research, EBC was used as clinical sample to develop a screening tool for lung cancer discrimination from two control groups (with and without risk factor). Three panels of metabolites were configured using the PanelomiX tool to minimize false negatives (specificity) and false positives (sensitivity). The combination of five metabolites led to three panels providing a sensitivity above 77.9%, specificity above 67.5% and the area under the curve (AUC) above 77.5% for the three panels. An additional study was developed as a first approach to study the statistical significance of metabolites at different stages of lung cancer.

Metabolomics analysis of exhaled breath condensate for discrimination between lung cancer patients and risk factor individuals.

The search for new clinical tests aimed at diagnosing chronic respiratory diseases is a current research line motivated by the lack of efficient screening tools and the severity of some of these pathologies. Alternative biological samples can open the door to new screening tools. A promising biofluid that is rarely used for diagnostic purposes is exhaled breath condensate (EBC), the composition of which has been inadequately studied. In this research, untargeted analysis of EBC using gas chromatography time-of-flight mass spectrometry has been applied to a cohort of patients with lung cancer (n = 48), risk factor individuals (active smokers and ex-smokers, n = 130) and control healthy individuals (non-smokers without respiratory diseases, n = 61). An identical protocol was applied to the two EBC fractions provided by the sampling device (upper and central airways and distal airway) from each individual, which allowed the compositional differences between the two EBC fractions to be detected. Tentative compounds that contribute to discrimination between the three groups were identified, and a relevant role for lipids such as monoacylglycerols and squalene was found. These results could support the ability of metabolomics to go inside the study of lung cancer.

Influence of the collection tube on metabolomic changes in serum and plasma.

Major threats in metabolomics clinical research are biases in sampling and preparation of biological samples. Bias in sample collection is a frequently forgotten aspect responsible for uncontrolled errors in metabolomics analysis. There is a great diversity of blood collection tubes for sampling serum or plasma, which are widely used in metabolomics analysis. Most of the existing studies dealing with the influence of blood collection on metabolomics analysis have been restricted to comparison between plasma and serum. However, polymeric gel tubes, which are frequently proposed to accelerate the separation of serum and plasma, have not been studied. In the present research, samples of serum or plasma collected in polymeric gel tubes were compared with those taken in conventional tubes from a metabolomics perspective using an untargeted GC-TOF/MS approach. The main differences between serum and plasma collected in conventional tubes affected to critical pathways such as the citric acid cycle, metabolism of amino acids, fructose and mannose metabolism and that of glycerolipids, and pentose and glucuronate interconversion. On the other hand, the polymeric gel only promoted differences at the metabolite level in serum since no critical differences were observed between plasma collected with EDTA tubes and polymeric gel tubes. Thus, the main changes were attributable to serum collected in gel and affected to the metabolism of amino acids such as alanine, proline and threonine, the glycerolipids metabolism, and two primary metabolites such as aconitic acid and lactic acid. Therefore, these metabolite changes should be taken into account in planning an experimental protocol for metabolomics analysis.

Development of a method for enhancing metabolomics coverage of human sweat by gas chromatography-mass spectrometry in high resolution mode.

Sweat has recently gained popularity as clinical sample in metabolomics analysis as it is a non-invasive biofluid the composition of which could be modified by certain pathologies, as is the case with cystic fibrosis that increases chloride levels in sweat. However, the whole composition of sweat is still unknown and there is a lack of analytical strategies for sweat analysis. The aim of the present study was to develop and validate a method for metabolomic analysis of human sweat by gas chromatography-time of flight/mass spectrometry (GC-TOF/MS) in high resolution mode. Thus, different sample preparation strategies were compared to check their effect on the profile of sweat metabolites. Sixty-six compounds were tentatively identified by the obtained MS information. Amino acids, dicarboxylic acids and other interesting metabolites such as myo-inositol and urocanic acid were identified. Among the tested protocols, methyoxiamination plus silylation after deproteinization was the most suited option to obtain a representative snapshot of sweat metabolome. The intra-day repeatability of the method ranged from 0.60 to 16.99% and the inter-day repeatability from 2.75 to 31.25%. As most of the identified metabolites are involved in key biochemical pathways, this study opens new possibilities to the use of sweat as a source of metabolite biomarkers of specific disorders.

Study of sample preparation for quantitative analysis of amino acids in human sweat by liquid chromatography-tandem mass spectrometry.

The determination of physiological levels of amino acids is important to aid in the diagnosis and treatment of several diseases and nutritional status of individuals. Amino acids are frequently determined in biofluids such as blood (serum or plasma) and urine; however, there are less common biofluids with different concentration profiles of amino acids that could be of interest. One of these biofluids is sweat that can be obtained in a non-invasive manner and is characterized by low complex composition. The analysis of amino acids in human sweat requires the development of sample preparation strategies according to the sample matrix and small collected volume. The influence of sample preparation on the quantitative analysis of amino acids in sweat by LC-MS/MS has been assessed through a comparison between two strategies: dilution of sweat and centrifugal microsolid-phase extraction (c-µSPE). In both cases, several dilution factors were assayed for in-depth knowledge of the matrix effects, and the use of c-µSPE provided the best results in terms of accuracy. The behavior of the target analytes was a function of the dilution factor, thus providing a pattern for sample preparation that depended on the amino acid to be determined. The concentration of amino acids in sweat ranges between 6.20 ng mL(-1) (for homocysteine) and 259.77 µg mL(-1) (for serine) with precision, expressed as relative standard deviation, within 1.1-21.4%.