Experimental Design for Untargeted Metabolomics
The design of experiments holds significant importance in metabolomics research, especially when different individuals handle sample collection and analysis. Prior to initiating the experiment, it is crucial to engage in detailed discussions and meticulous planning. Several considerations come into play, such as choosing the appropriate metabolomics methods, understanding their respective pros and cons, including the selection between targeted and untargeted metabolomics. Factors like detection throughput, sensitivity, metabolite degradation, sample stability, and sample size play a pivotal role in selecting the right analytical platforms and detection methods. Moreover, when working with animal or human samples, strict adherence to ethical review standards is essential.
A well-thought-out experimental design serves as a guarantee for obtaining reliable conclusions from experimental data and prevents data that may fall short of meeting research objectives. It's worth noting that biological samples are often influenced by various factors like age, gender, race, health conditions, and diet, all of which can impact human samples. In the case of cell experiments, growth conditions, encompassing the type of culture medium, temperature, environment, and genetic modifications, can affect changes in metabolites within samples. Consequently, these relationships between experimental conditions and results should be thoroughly considered during the experiment. Factor analysis can systematically explore different factors in laboratory research, and the use of stratified random grouping ensures balance among multiple factors.
In a study, the types and intensities of the effects of different factors or treatments on individuals are unknown. Therefore, sample size is a critical consideration in untargeted metabolomics. Estimating sample size requires taking into account aspects such as experimental design, variable detection, minimum expected differences, and data analysis methods. Typically, sample size estimation is based on two factors: 1. Preliminary data from small-sample pilot experiments; 2. Experimental platforms and downstream data analysis. Additionally, factors like experimental costs and requirements for participant selection need to be considered.
At present, there is no universal method for the experimental design of untargeted metabolomics. The design must address specific biological questions while considering sample characteristics (sample number, sample size, balance between groups and sample sizes, cell culture, and clinical sample collection plans) and the specific conditions of the metabolomics laboratory (experimental platform, analysis time, and experimental costs, etc.).
Untargeted metabolomics methods to analyze blood-derived samples (Dudzik et al., 2021)
Untargeted Metabolomics Sample Collection
In many studies, after sample collection, biological experiments and untargeted metabolomics analysis are conducted separately. Therefore, it is necessary to establish a unified sample collection process to provide reliable samples for both analyses simultaneously. The choice of methods during sample collection can impact the reliability of results. Since metabolites are not uniformly distributed in the body, it is essential to select appropriate sample types based on the experimental purpose. Factors such as circadian rhythms and diet can also influence metabolites, making it advisable to collect samples at the same time each day. Additionally, randomization of experimental groups and collection order should be considered during the collection process. When dealing with animal experiments, the time for animals to acclimate to the environment should be taken into account.
The collected sample volume needs to be adjusted based on the specific analysis methods. Data related to the samples, such as urine osmolarity, biomass, protein concentration, etc., should also be collected to normalize the results to the initial sample quantity. After sample collection, it is crucial to aliquot the samples according to the subsequent analysis needs to avoid unnecessary repeated freeze-thaw cycles. Choosing suitable container materials ensures that metabolites will not be adsorbed or lost during the sample storage process. When collecting blood samples, selecting appropriate anticoagulants (citrate, heparin, or EDTA) is necessary. For unstable compounds that require stability in the oxidation-reduction state during storage, adding antioxidants can be considered.
Reports indicate that the conditions and time of sample storage are crucial factors introducing interference, especially when collected samples are used for analysis by different research groups or when there is a time gap between sample collection and data acquisition. Therefore, these issues should be taken into consideration during the sample collection process. For cell samples, the above factors are usually minimized to reduce their impact on the analysis results. However, for experiments with a longer duration, attention should still be paid because different culture media, collection methods, storage times, and conditions are factors causing changes in metabolites.
Untargeted Metabolomics Sample Preprocessing
Metabolites' inherent reactions and enzymatic activities make many metabolites highly unstable. Therefore, in the metabolomics workflow, sample preprocessing is a critical step. Certain compound families, such as those involved in energy and redox metabolism, require specialized processing methods to ensure the accuracy of quantitative analysis. The main objectives of sample preprocessing include the rapid termination of metabolic reactions and the extraction of low-molecular-weight compounds from biological samples.
For cell samples, a gentle wash can be performed before quenching and extraction to remove the interference of the culture medium. However, washing can stimulate cells, altering their metabolic state, and may lead to cell membrane or cell wall damage, causing the leakage of intracellular metabolites. Currently, there is no uniform method or recommendation in the literature for cell washing solutions. Some studies use PBS or ammonium acetate buffer to remove residual culture medium, while others suggest terminating metabolism immediately after cell extraction, avoiding any washing steps.
Quenching can halt the activity of enzymes in biological samples. Common quenching methods include the use of organic solvents, freezing, heating, or a combination of methods. Using liquid nitrogen ensures the complete termination of all enzymatic or non-enzymatic reactions. Cold methanol or boiling ethanol is commonly used in the literature, and while concerns exist about heating and thermal degradation, boiling ethanol is an effective method for complete denaturation of all existing enzymes. For metabolomics experiments with adherent cells, it is recommended to avoid using trypsin digestion and instead directly scrape the cells.
Typically, samples should be rapidly quenched using liquid nitrogen or by adding a cooling quenching/extraction solvent. Water-organic solvent mixtures are the most commonly used quenching solvents as they reduce water content (slowing non-catalytic hydrolysis processes), denature proteins (preventing enzymatic reactions), and provide a sufficiently polar environment for the extraction and dissolution of various metabolites. Since most metabolites are water-soluble, compound extraction for liquid samples can usually be achieved by adding 3-4 times the volume of organic solvent. In addition to organic solvents, certain additives can enhance sample stability. For example, EDTA can effectively chelate trivalent iron ions, neutralizing their catalytic activity. The addition of low concentrations of chloroform or acid can also improve the stability of extracted metabolites.
When processing tissue samples, they can be frozen first and then processed through grinding or solvent extraction. After centrifugation to remove the supernatant, the pellet is resuspended for LC-MS analysis. Due to differences in the solubility of metabolites in the extraction solvent and the resolubilization solvent, it is advisable to remove insoluble compounds through high-speed centrifugation before LC-MS analysis. Another approach is to extract metabolites using different solvents based on compound polarity (lipid-soluble and water-soluble) and then choose the most suitable LC-MS method for analysis.
Select Services
Untargeted Metabolomics QC Samples
The purpose of quality control (QC) is to control and minimize the impact of all factors affecting the untargeted metabolomics workflow. QC samples are typically injected at the beginning and end of the LC-MS analysis sequence, as well as at regular intervals between analysis samples, to assess the stability and repeatability of the analytical method. In metabolomics analysis, there are different types of QC samples, including 1. a mix of equal amounts of all samples; 2. a mix of equal amounts of some samples; 3. surrogate biological samples (samples with the same matrix as actual samples); 4. commercial biological samples; 5. combined extracts after extraction; 6. a set of specific mixtures. Among these, only methods 1 and 2 are commonly used for preparing QC samples in metabolomics research. When preparing a mixed QC sample, equal amounts of all samples in the sequence are combined, and the final QC sample volume needs to be sufficient for the entire sequence analysis. In large-sample metabolomics analysis, it may not be possible to obtain all samples for mixing before the experiment begins. In such cases, existing samples can be mixed to prepare QC samples (method 2), and these two types of QC samples contain the same matrix as the detection samples, making them commonly used in experiments. Additionally, alternative samples with the same matrix as the detection samples (method 4) and commercial biological samples, such as NIST1950 plasma samples, can be selected as QC samples.
The main purposes of using QC samples include:
1. At the beginning of the sequence, stabilize and balance the LC-MS system by injecting QC samples according to the different detected samples, usually 5-15 injections.
2. Examine the stability and repeatability of the analytical method. It is typically recommended to inject QC samples at regular intervals throughout the entire sequence, with the interval determined by factors such as analysis time and total sample number.
3. Use a normalization strategy to correct retention time differences to ensure the repeatability of the analysis method.
4. Correct differences in peak intensity due to systematic errors in the analysis platform, such as the commonly used LOESS regression method.
5. Provide a reference to distinguish noise signals. Since metabolite peak intensity shows a linear response to concentration changes, this standard can be used to discard unreliable peaks by evaluating their response to diluted samples. For example, analyzing QC samples and diluted QC samples, observing peak intensity to remove interference peaks. However, this method may result in the loss of information for compounds with originally low concentrations.
For large-sample studies (hundreds to thousands of samples), the studied samples need to be analyzed in batches. For small sample sizes, all samples can be analyzed in one run. The sample order for detection needs to be randomized before analysis to avoid the influence of an uneven distribution and clustering on sample grouping. QC samples need to be injected every few samples according to the number of samples. QC samples can be placed in different sample vials to avoid solvent evaporation errors between consecutive injections from the same sample vial. Additionally, a set of QC samples must be run before the start of the sequence to balance the analysis system.
Reference
- Dudzik, Danuta, and Antonia García. "Untargeted Metabolomics Methods to Analyze Blood-Derived Samples." Metabolomics (2021): 173-187.
You may be interested in:
