Matthew J. Sikora, PhD
Department of Pathology, University of Colorado, School of Medicine

Models to study estrogen require strict control of steroid levels in cell culture.

My lab focuses on the action of steroid hormones, e.g. estrogen, in breast cancer. Models to study estrogen require strict control of the levels of steroids in tissue culture, and the best way to accomplish this is through the use of dextran-treated charcoal stripped fetal bovine serum (CSS). Treating serum with charcoal removes lipophilic compounds, including steroids, and depletion of steroids from tissue culture allows them to then be added back in known amounts. Importantly, residual steroids leftover from charcoal stripping in CSS can wreak havoc on experiments, as experiments often rely on the presumption that steroids are absent in control experiments. A major source of variability in CSS that influences the likelihood of CSS containing residual steroids post-treatment is the broad variability of starting material used by commercial sources. In an attempt to circumvent this variability, we rigorously test all incoming lots of CSS prior to purchase to validate that there are in fact no residual steroids present. ‘Failure’ of test lots has proven to be a major issue that can negatively affect a lab’s productivity, cause stress, waste funds, and can ultimately cause issues with reproducibility.

A single round of CSS testing required roughly 8 hours of benchwork.

During my postdoctoral work, I was in charge of maintaining our sera supplies, and had scheduled testing of new CSS lots months in advance of demand to assure no lapses in supply of validated CSS. In October 2015, with about 3 months of CSS remaining at the current usage rate, I requested samples from commercial suppliers. A single round of CSS testing required roughly:

  • About 8 hours of benchwork, over the span of nearly two weeks
  • 10-12 96 well tissue culture plates, associated consumables/plasticware
  • Cell proliferation assay reagents

In late 2015, I initiated a round of this testing. The first test lot failed testing, i.e. testing identified that the lot contained residual estrogenic steroids. The company sent another lot, which also failed – only to find out after testing was complete that this was the same lot as the first sample, an error which cost us an additional 3 weeks in communication and testing. A new lot of CSS was sent, but this also failed. Since the testing takes roughly 2 weeks, compounded with the error in the first replacement, I had already performed 3 rounds of testing (~24 hours of bench work) over 7-8 weeks. At this point, it was December, and delays in getting a third lot was going to push additional testing in to early 2016.

At this point, I started asking the lab to monitor their CSS use (which I was already tracking), to be conscious that our supply was limited, and that I was having difficulties finding a new lot. I estimated that we had enough serum to last through January.

Continued failed lots of FBS required rationing of CSS usage.

In early January, two new lots arrived from our regular supplier and an additional supplier. I tested these immediately, but to my great dismay, both lots failed. At this point, I had to sound alarm bells in the lab – we had a number of experiments and projects that required serum continuity, and I could not allow the supply to be fully exhausted. We had meetings about ‘rationing’ CSS, and I had to put our remaining supply on lockdown, reserved only for critical needs and long-term ongoing studies. We had a large group using CSS – about 10 trainees and technicians – and many studies ground to a halt.

In desperation, untested FBS was used which resulted in several weeks of lost effort on experiments.

A fourth lot was en route, but production delays bit us again, and I was forced to buy a small supply of CSS without validating it. I warned lab members of the issues with using these in our studies, but several anxious graduate students forged ahead with the untested CSS (which ended up being a prior failed lot). Unfortunately, at least a couple of them lost a couple weeks of effort on experiments that were incompatible with the presence of residual steroids.

Finally, in early February, a final lot arrived…

Case study PDF download


Download the full case study

About the Author

Matthew Sikora, PhD from the University of Colorado, Anschutz Medical Center. Dr. Sikora shares his experience in the struggle to find a consistent source of high-quality FBS in this case study.

To save yourself 5-months of benchwork and $45,000 in experiments, contact us.

Get exclusive updates direct to your inbox.