|
Here’s something fairly uncomfortable to consider if you regularly culture cells with fetal bovine serum: there is a whole lot of cow protein and genetic material in that golden elixir, and the implications of this scientific reality are still being worked through.
Let’s zoom in super niche for a moment: FBS contains bovine extracellular vesicles, loaded with bovine-derived miRNAs. When you culture cells in FBS-containing media, those vesicles can be taken up by your cells. In some systems, their miRNA cargo has been shown to alter gene expression. Here’s why: miRNAs are potent post-transcriptional regulators. A single miRNA can suppress a whole network of targets. Many bovine and human miRNA sequences are similar in the specific stretch of sequence that determines which genes they bind to. As a result, bovine miRNAs have been shown in some systems to engage human gene regulatory pathways after uptake from serum-derived vesicles. How strong or consequential that effect is in your specific model is almost never validated when changing batches of FBS. For most routine cell culture, this is arguably background. But there are contexts where it may matter more: RNA biology experiments, EV research, gene expression studies, or any work where you’re trying to attribute a transcriptional phenotype to a specific treatment. In these cases, you may have an additional source of regulatory RNA entering your cells from your media. Unfortunately, it's one rarely flagged in the methods section. The extracellular vesicle (EV) research community has been aware of this issue for some time; EV-depleted serum became common practice in that field for good reasons, even though depletion methods don’t fully eliminate bovine EVs. The implications may extend beyond EV workflows, and yet they’re almost never discussed outside that context. This is one dimension of what “FBS batch variability” can mean biologically. Scientists usually focus on growth factor concentrations or adhesion proteins varying between lots. But variability extends to differences like this as well - differences most scientists have never even thought to consider, much less one quantified across batches. Just something to think about, particularly if you’re validating a new lot of FBS and running sensitive transcriptional experiments. Exosome-depleted FBS can help reduce this variable, or a chemically defined alternative can eliminate it altogether. Anyway. One more reason "the same experiment, same results" thing doesn't always pan out 🙃. As always, literature below if you want to dive deeper! Wei et al. demonstrates RNA present in fetal bovine serum contaminates extracellular RNA analyses and can be misattributed to cultured cells. https://www.nature.com/articles/srep31175? Beninson & Fleshner show experimental evidence that exosomes present in FBS can influence cell behaviour in vitro, specifically demonstrating suppression of macrophage inflammatory responses. https://www.sciencedirect.com/science/article/abs/pii/S0165247814002387?via%3Dihub Urzì et al. is a great review detailing how FBS-derived extracellular vesicles and RNA complicate EV studies. https://pubmed.ncbi.nlm.nih.gov/36214482/
0 Comments
Something that annoys me to an arguably irrational level about the “serum-free media” space is how confusing the vocabulary gets once you start digging.
I’ve lived and breathed this stuff for the better part of a decade, and it still sometimes takes me a beat to understand a serum-free supplement. You see these words everywhere: Serum-free. Animal-free. Chemically defined. Reproducible alternative to FBS. A quick Google and you’d be forgiven for thinking there are a wide variety of equivalent, high-quality serum replacements out there. But these words can be layered together in ways that blur real scientific differences. First, some formulations simply replace FBS with another complex biological supplement. This is a reasonable strategy; for example, replacing FBS with human platelet lysate in cell therapy manufacturing to reduce cross-species risk. But you still have batch-to-batch variation. ”Animal-free” removes animal origin risk but can include other undefined extracts. Love this for ethics or regulatory positioning. It's less useful for having control over what’s in your media or batch reproducibility. “Reproducible alternative to FBS” has given me the ick several times. I’ve seen this wording, dug into the supplement, then realized it isn’t chemically defined in the slightest. There may be process consistency or tighter QC than raw serum but if the inputs are undefined, reproducibility has a ceiling. “Chemically defined” is the gold standard. Every component and concentration is known. This one tends to be safer, but there are levels. You’ll sometimes find purified bovine serum albumin or similar components, despite the fact that albumin is serum-derived and carries residual batch variability depending on purification and lipid loading. To be clear, this is how the category evolved rather than a criticism of any specific product, and perfect shouldn't be the enemy of good. People come to me because they’re over FBS for a lot of different reasons: Ethics. Reproducibility. Regulatory risk. Cost. Most existing serum replacements were designed to tackle one or two of these, not eliminate all of them simultaneously - fair. Still, when we built our replacement, we were deliberate about the vocabulary: fully chemically defined and fully animal-free, designed for reproducibility. I am regularly challenged on these terms by people who’ve been burned before - by products that were "reproducible" but not defined, or animal-free at the product level but not across the supply chain. And frankly, I love it, because I know our language matches the technical reality. The goal has always been to provide an alternative that can realistically compete with FBS on price, sustainability, regulatory AND experimental control/reproducibility across time, labs, and geographies. This vocabulary, and how we use it, matters a lot if we as scientists want to achieve that. To help prospective customers quickly understand the breadth of what FRS Pioneer can do, we've put together this helpful diagram. As always, please get in touch if you have any questions about using FRS Pioneer in your laboratory.  There’s a hidden modelling assumption baked into many AI- or ML-enabled biological datasets: If your models are trained on cell culture data, your media formulation is part of the model.
Here’s how a default choice I’ve seen teams make early on, accidentally becomes expensive to unwind later: When companies using cell biology to train predictive models are first established, cell culture processes are often given relatively minimal thought. Many companies treat this as basic infrastructure, while the critical part of the model sits downstream in protein folding, functional phenotypes, or biological performance metrics. Unfortunately, by starting from standard cell culture processes, many of these companies begin using processes that rely on fetal bovine serum or other undefined, variable cell culture media inputs. These model datasets are intended to become moats, but they’re only as robust as the conditions under which they were generated. Serum and other animal-derived nutritional supplements influence growth rate, shape stress responses, and impact metabolism, signaling, phenotypic baselines, and every facet of cell biology. From a modelling perspective, this means your training data can encode variability that isn’t obvious until conditions change. Unsurprisingly, this shows up in the data set as unexplained variability or performance drift. The major challenge is that this risk is easy to miss early on, when your model data is fairly limited, e.g. it’s been collected in one laboratory, using one batch of FBS, under “pretty constant” conditions. Usually, this risk rears its ugly head when datasets grow. Suddenly, you’ve used up your batch of FBS and you’re switching to a new one, or you’re transferring your findings to a different lab/company/site and the model’s results aren’t holding up. At that point, performance drift is sometimes blamed on “biology” or “model issues,” when it’s at least partially due to the “basic” cell culture processes that were popped in place three years ago. Reality is, if your experimental system isn’t controlled, your training data isn’t either. This will have implications for your model. My suggestion here is simple; think about this early! Media formulation should be a deliberate modelling decision, not a background reagent choice that we just roll into because “oh yeah, the literature says DMEM + 10% FBS so let’s go for it.” Teams that commit early to chemically defined, stable culture conditions are less likely to face costly or time-consuming surprises when models are applied, transferred, or scaled. Stable inputs in, more reliable models out, and a healthier data moat over time. That tends to keep everyone happy! |
What's been happening?Sharing the Media City journey has been important to us because we want to encourage the next generation of scientists to establish companies that will advance scientific research. Check back regularly for the "building in public" updates on what it looks like to establish a scientific company. Archives
March 2026
Categories |
RSS Feed
