Selection of Serum, Plasma, and Anticoagulants for Metabolomics

Metabolomics, as one of the numerous omics disciplines, has become an indispensable tool in clinical research, ushering in a new era for clinical diagnosis, early disease monitoring, treatment prediction, and the monitoring of treatment efficacy.

Situated downstream of the gene regulation network and the protein interaction network, metabolism reflects the executed biological activities. The comprehensive analysis of metabolites provides a more direct characteristic label and a more sensitive detection method for human phenotypes.

In practical research, metabolites are susceptible to various factors. The significant challenge lies in the intricate diversity of biological specimens, making sample selection, collection, and preservation essential preprocessing steps in metabolomics.

However, many researchers, in the process of biomarker discovery, often focus primarily on the influence of instruments, frequently neglecting sample-related variables. Blood samples, a common specimen type in metabolomics, offer many advantages in clinical settings, being easily obtainable with minimal trauma to patients.

Nevertheless, the operational procedures for blood samples in clinical settings may not always be suitable for biomarker discovery. These samples may undergo different preprocessing steps, and some might be stored in various biobanks, with variations in aliquoting time, storage time, and conditions, all of which could significantly impact the stability of metabolites.

Selection of Serum and Plasma

Serum and plasma are widely used in metabolomics research, but there is still debate about which sample is more suitable for metabolomics studies.

Plasma is obtained by adding anticoagulants to whole blood and separating it through centrifugation, while serum is obtained without anticoagulants by separating the supernatant after coagulation and centrifugation to remove blood cells and fibrinogen.

Overall, there is not much difference in the types of metabolites between serum and plasma, but the concentrations of most metabolites differ significantly between the two sample types.

Plasma, separated from whole blood per unit volume, contains more, without platelets and post-coagulation protein fragments. However, serum has lower total protein content after separation, which may have more benefits for small molecule analysis, especially as an excess of proteins can lead to ion competition and reduced sensitivity.

Compared to serum, the advantage of plasma is its quick processing without the need to wait for blood clotting. The coagulation time (30-60 min) and room temperature during the handling introduce some variables, and these changes have been shown to affect the analysis of serum metabolite components.

For example, clotting time may increase the degradation of certain enzymes and metabolites, promoting the loss of metabolites. Activated platelets can release certain compounds into the serum during the clotting process, such as peptides, xanthine, hypoxanthine, and sphingosine-1-phosphate.

In contrast, plasma is more suitable for metabolite analysis as it does not undergo the blood clotting process, and the results are easier to replicate. Both serum and plasma can be used for metabolomics research, but sampling should be consistent—either use plasma for all samples or use serum for all samples.

If serum is chosen as the sample, strict adherence to Standard Operating Procedures (SOP) is essential. In the same study, to prevent changes in metabolites originating from activated platelets, it is necessary to standardize the clotting time for all samples.

When collecting serum, it is recommended to use additive-free collection tubes made of plastic or, preferably, glass. Plasma has better reproducibility, but when choosing plasma, the matrix effects caused by anticoagulation should also be considered.

Serum or plasma for metabolomics studies and beyond (Liu et al., 2018)

Table 1 Metabolite Differences Between Plasma and Serum

Anticoagulant Selection

The choice of anticoagulant is also a crucial factor affecting the quality of the metabolic profile. Common anticoagulants include sodium heparin, lithium heparin, sodium citrate, or EDTA salts.

Heparin is an anticoagulant that inhibits thrombin activation, while EDTA and citrate salts chelate calcium ions in the blood. EDTA not only inhibits blood clotting but also inhibits enzymes in red blood cells that depend on Mg2+, making plasma more suitable for downstream metabolomic analysis.

EDTA and heparin are in solid form in blood collection tubes, while sodium citrate is in liquid form. When using sodium citrate, a 3.8% solution is typically prepared in a vacuum blood collection tube, with a ratio of 1:9 for sodium citrate anticoagulant to blood. A significant reduction in the concentration of sodium citrate in whole blood can significantly affect the assessment of platelet function and increase variability.

Research has shown minimal differences in the results of metabolite analysis by LC-MS among three common anticoagulants (lithium heparin, sodium citrate, and EDTA dipotassium). Other studies have found that plasma samples anticoagulated with EDTA and heparin exhibit similar metabolic characteristics.

There is no specific recommendation for the selection of anticoagulants in metabolomics, but once the various limitations of the chosen anticoagulant are thoroughly considered and found to be suitable for the entire study, a decision should be made.

Impact of Hemolysis on Metabolic Profiles

Hemolysis is the most common pre-analytical error that can significantly alter the levels of metabolites in blood. It occurs when the cell membrane is damaged, leading to the release of hemoglobin and other cellular components, including metabolic products (structural proteins, lipids, carbohydrates), and enzymes. Careful handling during procedures can help avoid hemolysis.

The upper reference values for free hemoglobin are 20 mg/L in plasma and 50 mg/L in serum. When the concentration of free hemoglobin is around 300 mg/L (18.8 mmol/L), serum or plasma may exhibit a pink to red color.

However, in samples with mild hemolysis, this color change may not be noticeable. Therefore, the most reliable method to detect mild hemolysis is to measure the concentration of free hemoglobin in the sample.

Interference caused by hemolysis is approximately linear with the concentration of free hemoglobin in the sample, directly or indirectly affecting various tests, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (Cr), creatine kinase (CK), iron, lactate dehydrogenase (LDH), lipase, magnesium, phosphorus, potassium, urea, albumin, alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), glucose, and sodium.

Studies have shown that hemolysis, compared to clear serum, can cause significant changes in the levels of at least 43 metabolites. Among these, phenylalanine, tyrosine, acetate, citrate, formate, and succinate increase in hemolyzed samples, while concentrations of lipids such as phosphatidylcholine and sphingomyelin decrease in hemolyzed samples.

In assessing potential pre-analytical variables, it is recommended to determine hemolysis levels in samples using spectrophotometry and hemolysis index.

Storage Conditions

For metabolomic samples, it is generally recommended to prioritize storage at -80°C. If clinical conditions are limited, temporary storage at -20°C can be considered before transferring to long-term storage at -80°C. Storing plasma at -20°C for more than a month can result in significant changes in metabolites such as glucose and proline.

Currently, the impact of long-term low-temperature storage and repeated freeze-thaw cycles on the overall stability of the metabolome remains an unresolved issue. Despite several studies in recent years indicating that storage at -80°C for over 4 years or undergoing three freeze-thaw cycles does not show significant differences in metabolome integrity and minimal changes in lipid components, it is still recommended to avoid repeated freeze-thaw cycles as much as possible within the same study.

For optimal stability, it is advised to freeze samples at -80°C promptly, and if feasible, avoid repeated freeze-thaw cycles, especially when dealing with samples from different study groups within the same research project.

Indicators of Sample Quality

The stability of metabolites after cell separation can still be influenced by enzymes or numerous other proteins remaining in serum and plasma. Therefore, sample processing is a crucial factor affecting the stability of the metabolome.

Research has reported a significant impact of pre-analytical variables during serum and plasma processing on a considerable number of metabolites. The time of blood processing, time from blood coagulation to centrifugation, and the time from freeze-thawing the sample to analysis should be minimized for optimal stability. Additionally, when aspirating plasma, care should be taken to avoid touching the buffy coat to prevent the aspiration of blood cells.

Metabolites sensitive to pre-analytical variables, including serotonin, hypoxanthine, taurine, glutamine, glutamic acid, glucose, lactate, sulfur-containing amino acids, succinic acid, maltose, S1P cholesterol metabolites, triglycerides (TG), PCS, LPCS, and catecholamine derivatives, exhibit high variability. Therefore, when considering these compounds as candidate biomarkers, a thorough evaluation should be conducted.

For reproducible and reliable analysis of catecholamine derivatives and similar compounds, sample preprocessing is a critical step and should be performed quickly at controlled temperatures. Altering blood processing steps that affect platelet metabolism and induce platelet activation is crucial for the repeat analysis of serotonin specifically present in platelets.

In conclusion, metabolites sensitive to pre-analytical conditions must be thoroughly validated for reliability, and results should be carefully interpreted. Rapid processing at 4°C helps maintain effective sample preservation.

Reference

1. Liu, Xinyu, et al. "Serum or plasma, what is the difference? Investigations to facilitate the sample material selection decision making process for metabolomics studies and beyond." Analytica Chimica Acta 1037 (2018): 293-300.

You may be interested in:

Metabolomics Service

Untargeted Metabolomics

Targeted Metabolomics

Serum Untargeted Metabolomics