Preparation of cell samples for metabolomics

Guidelines for the extraction of metabolites from cells grown in tissue culture are found below. The protocol and metabolomics sample submission form can be downloaded as a pdf.

Protocol for the extraction of metabolites from cells grown in tissue culture

  1. Prepare the required volume of ice cold methanol (MS grade if possible, otherwise HPLC grade) solution with milli-Q water.
  2. Remove cells from the incubator and pour off the media from a confluent 6cm dish or T25 flask or similar of cells. Label and keep 1mL of media from each dish in a sealed tube if analysis of media is required.
  3. Quickly add 5mL of PBS buffer at 37oC, swirl and pour off to waste to remove excess media. Repeat once more.
  4. Snap freeze cells by adding ~ 10 mL of liquid nitrogen, enough to cover all cells (note this is an optional step – it is the quickest and most effective way to arrest metabolism, however, it is also common to simply add the ice cold MeOH directly (step 5) to the cells without the prior addition of liquid nitrogen).
  5. Add 500uL of methanol to cover the cells in the dish (500uL per 10cm dish or less volume can be used with a smaller dish). Do this before all the liquid nitrogen has boiled away if using the liquid nitrogen (step 4) (probably less than 2 mins). Swirl gently making sure all the cells are covered by the methanol, do this for at least 3 minutes. Note: from here on the process is less time critical as metabolism has been arrested, metabolites have been extracted into the methanol solution and enzymes largely denatured.
  6. Use an appropriate cell scraper to remove adhered cells material into suspension. Make sure this is done thoroughly as no trypsin step is used so adherent cells need to be mechanically displaced. The aim is to remove >90% of cells material and take into methanol suspension.
  7. On ice: Pipette the resulting lysate solution including cell debris into a 1.5mL Eppendorf tube (or equivalent for centrifugation). It is useful to draw the solvent in the cell dish up and down several times to ensure cells are suspended in solution and no left behind.
  8. Check that all cells have been removed from the dish using a microscope. If >10% of cells are still adhered, pipette the methanol suspension back into the dish and re-scrape. It’s important to get at least 90% of cells into suspension in the Eppendorf tube to ensure accurate results.
  9. Keep the Eppendorf tube ice cold and spin down the mixture in a centrifuge at max rpm (>13,000 rpm) for 30mins (ideally a cooled centrifuge but this is not essential).
  10. Measure the DNA context of the supernatant: Use our separate DNA normalisation protocol to perform this step.
  11. If there is a >20% variability in the DNA context across the samples the next step is to normalise each sample by adding a volume of MeOH to each sample so that the total volume of each sample is proportional to the amount of DNA present (total protein is sometimes used for this step and is optional but be have found DNA to be more accurate) (If there is <20% variability this step can be ignored and move on to step 13 below).
  12. How to normalise samples: For example if you have 3 samples and you measure the total DNA content for each as 100ng/µL, 75ng/µL and 50ng/µL respectively. You should normalise to the 50 ng/µL sample by diluting the others with MeOH to bring them to the same concentration (I.e. double the volume of the first sample (100ng/50ng/uL) and add a third of the volume of the 75ng/µL sample)**. Note: The DNA content for each sample is expected to be in the region of 50-100 ng/uL (from the extraction of a T25 flask or 10cm tissue culture dish of cells (~1-2 million cells). If you are getting DNA values <30 ng/uL the cell extraction efficiency or original starting number of cells is too low and metabolite coverage will be poor. In this case you might want to consider repeating the experiment. Don’t normalise to samples with DNA content <30ng/uL. If the sample with the lowest total DNA content is <30ng/uL or you have a few outliers near this value with the lowest DNA content – ignore these samples and normalise to the next highest total DNA (or lowest in the more representative group in case of clear outliers. Please report the total DNA content (or protein content or other normalisation measure used) for each sample on the sample submission form.
  13. 500uL of the normalised sample should next be filtered using a 10 kD molecular weight cut-off filter* (10 kD MWCO). This removes soluble protein from the metabolite solution (Amicon Ultra -0.5mL centrifugal filters UFC501024 recommended). Make sure you follow filter preparation instructions to remove glycerol (washing with milliQ water prior to use for at least 30mins in centrifuge).
  14. Transfer the filtered and normalised sample extract into a Waters Total Recovery autosampler vial*** with pre-slit ptfe cap. Please make sure this vial type is used.
  15. Put the sample in a vial in a -80oC freezer until the day of MS analysis. Ensure the sample vial is stored upright in a suitable tray to avoid loss as samples can remain in liquid state. Avoid freeze-thaw cycles. Transport on dry ice to the MS lab when submitting samples.
  16. The sample is ready for LC-MS/MS analysis using method 1 or 2 (see page 1). For method 3 (amino acids) an additional derivatisation step is required which can be performed in the McCullagh lab when you deliver the samples (please arrange this with Elisabete or Kourosh prior to sample delivery).
  17. A quality control (QC) sample should be made to monitor analytical reproducibility (do this before freezing all the samples for the first time). A QC sample contains an equal volume from each sample combined into a single vial. To make a QC sample add together an equal volume of each sample into a fresh total recovery vial to make at least 250 uL in total. The volume taken from each sample depends on the number of samples you have but typically this would be 10-20uL per sample. Label this ‘QC’ and submit along with your other samples.
  18. Optional: You may wish to also keep the media removed from each cell sample prior to metabolite extraction and submit for analysis along aside the cell extracts. This can be interesting as it allows changes in the use or excretion of metabolites to be monitored which can be useful depending on the metabolic system and experimental aims. Comparing fresh and spent media enables a comparison of what compounds the cells have utilised and excreted. All media samples must be normalised by the same ratio as the cells from each cell dish (I.e. according to the step 10) and filtered using a 10 kD MWCO filter as for cells above.
  19. Submitting samples for analysis: Once you have prepared your samples please contact the researcher in the McCullagh lab who you are working with on the project (usually Elisabete Pires or David Hauton) and arrange to deliver the samples and to make up any additional samples (for example derivatisation for amino acid normalisation) and create the unique sample sequence for your samples. These will all be done in the McCullagh lab. Next please complete the sample submission form below and submit, by email as well as paper copy at the time you bring your samples to the MS lab.

Note: For this protocol no cell counting is performed on flask harvested for MS analysis but additional flasks of cells should be gown and harvested for each experimental condition in order to determine the number of cells present. We make the fundamental assumption that there are a similar number of cells in the same flasks from the same experimental conditions. We recommend using a nanodrop or similar to determine the amount of DNA in each sample after cell lysis and normalise samples using methanol accordingly. Please record the normalisation factor you calculated and supply this in a separate excel sheet along with the completed sample submission form.