Heterogenous hESC aggregates formed by the traditional method of scraping colonies of hESC off the culture medium surface. The heterogeneity of the clumps makes for inconsistent differentiation.
Courtesy of Mark Ungrin, University of Toronto, and PLoS ONE

As research with stem cells and induced pluripotent stem cells (iPSCs) forges ahead, a growing number of "stem cell beginners" are joining a field long-dominated by a small cabal of experts. Techniques for inducing pluripotency are rapidly evolving, but researchers working with already-pluripotent cell lines must navigate the tricky waters of maintaining culture conditions for growth and differentiation.

"There's no point where these cells become routine," says Meri Firpo, a stem cell biologist from the University of Minnesota. "You really have to evaluate them all the time." Whether you're working with human embryonic stem cells (hESCs) or human iPS cells (hiPSCs), you'll need to plate and replate the cells for months to keep them from spontaneously differentiating.

For new researchers receiving their first batch of cells frozen on the brink of differentiation, "you have to set aside all of your pre-formed notions about doing mammalian cell culture," says Travis Berggren, director of the stem cell core facility at the Salk Institute for Biological Sciences. Many shortcuts possible with standard cell culturing don't apply to stem cells. Knowing which culture media and materials help best in tending your cells in an undifferentiated state, and even characterizing them, can be difficult since the exact cell mechanisms underlying pluripotency are not known. And keeping cultures free of pathogens requires constant testing of incoming reagents.

The Scientist asked researchers to describe challenges they'd surmounted in culturing and characterizing hESCs. Here's what they said:
The cells themselves

User: Mark Ungrin, postdoc in Peter Zandstra's lab at the University of Toronto

Project: Improving consistency and yield in hESC differentiation by creating aggregates of hESCs from single cell suspensions.

Problem: Techniques for differentiating hESCs and hiPSCs require the formation of cell aggregates. But the clumps are usually heterogeneous in size and shape, leading to an inconsistent differentiation yield. To achieve consistency Ungrin tried a published protocol for creating aggregates from single cell suspensions. Initially, he was able to create stable aggregates but then they stopped surviving. "The aggregate size just wasn't consistent from experiment to experiment," he recalls. "Sometimes I'd get multiple aggregates within a single well."

Solution: Ungrin spent six months troubleshooting the culture media to no avail. Then, he turned to the cells themselves, finding that two factors seemed to help: To his surprise, older hESCs, which had been passaged many times, were more likely to form stable aggregates. And cells in the predifferentiation stage seem more likely to survive reduction to single-cell suspension, so Ungrin made sure to harvest the population before all of the cells had differentiated. He also found that a small molecule inhibitor of Rho-associated kinase, shown in previously published studies to promote survival, worked well with his cells (PLoS ONE, 3:e1565, 2008).

Considerations: "One of the bugbears of working with human ES cells, is that things that you do in one lab can be difficult to reproduce for reasons that are not obvious," he says. Even differences between how two different lab technicians pipette the cells from one plate to another can affect survival. When troubleshooting, Ungrin suggests remembering that the age and state of your cells, in addition to the culture media, can affect survival.

Marker manhunt
hESCs stained positive for SSEA4 (stage-specific embryonic antigen).
Courtesy of Sheena Abraham, Virginia Commonwealth University

User: Raj Rao, stem cell bioengineer at Virginia Commonwealth University, Richmond

Project: Characterizing and monitoring human embryonic stem cells using cell surface markers.

Problem: There are no set standards for assessing the pluripotency of hESCs and how the cells change as they differentiate. Commonly used cell surface markers include stage-specific embryonic antigen-3 and 4 (SSEA-3 and SSEA-4). But different cell lines vary in their expression of SSEA-3 and SSEA-4. "It makes sense to have a battery of markers," says Rao. So he set out to find more.

Solution: Rao chose to use lectins, proteins that bind and modify cell surface sugars throughout embryonic development. According to the literature on mouse ESC characterization, lectin receptors are displayed on cell surfaces at preimplantation and implantation stages of development. The group tested 14 different lectins, looking for ones expressed with SSEA-4. They pulled out a few lectins, such as Ricinus Communis agglutinin and Maackia amurensis, that bound tightly to the hESCs, predicting pluripotency.

Considerations: The more markers you test, the more certain you can be about the pluripotency of your cells, Rao says. For those who want to try characterizing based on lectins, it's a relatively straightforward and inexpensive process; you'll need a cohort of antibodies and access to a table-top flow cytometer. Remember though, expression of lectin and other cell surface markers such as SSEA-4 varies between cell lines and even colonies from the same line. One way to increase your confidence in the results is to increase the number of samples you analyze.

Mycoplasma mania
hESCs growing as a well defined colony in chemically defined media not requiring feeder cells.
Travis Berggren / Salk Institute for Biological Studies

User: Travis Berggren, stem cell core facility director, Salk Institute for Biological Studies, La Jolla, Calif.

Project: Studying the mechanisms controlling hESC self-renewal mediated by fibroblast feeder layers.

Problem: When Berggren worked as a staff scientist at the WiCell Institute in 2005, some of his hESCs started detaching from their cell culture dishes and dying. Many of the cells in the lab were contaminated with mycoplasma, a type of bacteria that is antibiotic resistant and typically not visible.

Solution: The group spent two weeks purging their cells and culture reagents, autoclaving every incubator, and bleaching and testing the facility. But they lost more than just two weeks of work. Since mycoplasma starts growing slowly and doesn't initially interfere with differentiation, it's likely the researchers unknowingly worked with contaminated cells for longer than a month. Even after detecting mycoplasma, some coworkers still thought their cells looked viable, even though they tested positive, Berggren says. "That tells you that you really need to test early on."

Berggren's group established a testing routine which they still follow religiously three years later. They test new cultures weekly and older cultures every 2-4 weeks, by testing for DNA or enzymes specific to mycoplasma. All new culture materials, such as mouse fibroblasts used in the feeder layer, are put in a quarantined incubator. They don't enter the lab's main stream of culture materials until they've tested clean at least twice.

Considerations: Mycoplasma is especially troublesome for labs that move from work in mouse ESCs to work in human ESCs because human stem cells suffer much more severe consequences as a result of exposure to it, by losing their full differentiation potential and/or dying. Labs working with hiPSCs must also watch out that the cells used to deliver the pluripotency factors are not contaminated.

Dynamics of differentiation
Neural progenitor cells (nestin+, red) and neurons (TuJ1+, green) derived from hESCs.
Courtesy of Wu Ma, ATCC Stem Cell Center

User: Wu Ma, developmental biologist, ATCC Stem Cell Center, Manassas, Va.

Project: Maintaining hESCs for later conversion into neural progenitor cells.

Problem: Ma knew the extra-cellular matrix (ECM) was critical for stem cell renewal and differentiation, but he wanted to determine which elements of the ECM helped stem cells form neural progenitor cells.

Solution: Ma's group tested two different hESC lines, plating each onto five different surfaces: poly-D-Lysine, fibronectin, laminin, type 1 collagen, and Matrigel (an ECM product sold by BD Biosciences). He found that cells cultured in laminin were best at differentiating into neurons. That effect was partially mediated by the stem cells's integrin receptor: blocking the a6b1 receptor slowed differentiation.

Considerations: Laminin worked best for Ma, but there's no one-size-fits-all answer to which medium to choose. Different culture media are constantly being developed and the efficacy of one culture can even vary from cell line to cell line, Ma says.

Ma advises when choosing culture media and other materials, to read up not only on how past studies stem cell studies have addressed the issue of microenvironment, but also on factors known to be involved in normal in-vivo development. If you're differentiating to neurons, for example, familiarize yourself with the literature on factors involved in brain development. But know the limits of the literature, says Ma. "Under the same culture condition some human ESC lines produce more neurons than others" because each line has its own genetic background.

Tips on getting the most from your cultures

Get your training now
The National Institutes of Health's T15 program which provides training courses on hESC culture and maintenance is in its last year of funding. The NIH Stem Cell Task Force is still discussing the future of the program, but in the meantime, a few courses in Virginia, Minnesota, and Maryland are slated for early 2009. In general, stem cell courses "come in all different flavors," says Tenneille Ludwig of the WiCell Institute, adding that you'll have to spend some time researching which course will fit your timeline, interest level, and pocketbook.

Start by shadowing.
Even better than a short course is training in a veteran's lab, says Meri Firpo, an assistant professor at the University of Minnesota's Stem Cell Institute, who has hosted students and postdocs in her lab for months at a time. She sends her students back with enough culture media to get started, which is important because the quality of materials can vary from lot to lot.

Test your control cells first.
Mahendra Rao, vice president of research, stem cells, and regenerative medicine at Invitrogen in Carlsbad, Calif., recommends two inexpensive and readily available cell lines: the BG01V line, a karyotypically abnormal variant of a federally approved ESC line, and the NT-2 or N tera-2, an embryonic carcinoma line. Both are available from ATCC, a private nonprofit biological resource center. "Those are two simple controls," Rao says. "If your media is going to kill those cells, then your media's going to kill your ES cells."

Don't change anything.
When you receive a cell line, follow the exact instructions of the cell supplier. Do this at first, Firpo says, "until you have them frozen down and in enough quantities to change the conditions to suit your own needs. At least by then, you'll have enough cells."

Read up.
If you don't have an experienced technician to peer at your cells, recent books and review chapters with pictures are the next best thing, says Mark Ungrin, who coauthored a recent review on phenotypic analysis of hESCs. Methods books cover the fundamentals in exhaustive detail: Mahendra Rao recommends Human Embryonic Stem Cell Protocols by Kursad Turksen and Human Stem Cell Manual: A Laboratory Guide by Jeanne Loring.

Establish a routine.
Start a good quality-control program early and stick to it, says Ludwig, which includes checking the temperature, humidity, and carbon dioxide levels in cell incubators. That should include all reagents. "Anything that contains a protein is going to have a high level of biological variability," she adds.