Many drugs have looked like blockbusters in the cell-culture dish—easily infiltrating target cells and carrying out their tasks—only to flop in animals or people. The problem is simple: unlike those culture dishes, “we’re not flat,” says Shuichi Takayama of the University of Michigan in Ann Arbor.
One issue with traditional 2-D culture is that cells in a single layer attached to glass or plastic have unfettered access to the media above them. They grow unusually quickly, as they imbibe a steady stream of nutrients. When drugs are added, the cells absorb them just as easily. But when those same drugs come up against the complex vasculature and cellular barriers in a living organism, they may fail to even reach their targets.
In addition, 2-D culture requires cells to interface with an unnatural material. “The interactions [cells] have with the plastic or glass, it’s not the same as the cell-cell interactions they have in vivo,” says Glauco Souza of Nano3D Biosciences, Inc., in Houston. Even when researchers layer a culture dish with proteins such as those found in the extracellular matrix, “that’s a very artificial substrate.”
After more than a century of mostly flat cultures, cells in the laboratory can now experience 3-D environments. “This area has really heated up because people have realized the benefits of 3-D,” says Jeffrey Morgan of Brown University in Providence, Rhode Island. “There is a lot of innovation going on.”
Adding the third dimension makes a world of difference. It changes cells’ gene expression, their protein production, and their very shapes. It can influence how cells differentiate, proliferate, interact, and survive in culture. For example, liver tissue grown in 2-D cannot metabolize drugs.
And in a recent study, John March of Cornell University in Ithaca, New York, discovered that a 3-D intestinal culture (see below) absorbs drugs in a manner more akin to the natural intestine, compared to 2-D culture. In real intestines, the blood pressure medication antenolol enters at the base of intestinal villi. But it does not easily cross 2-D intestine models, because they only represent the villi tips (Biotechnol Bioeng, 109:2173-8, 2012).
In selecting a 3-D method, it’s important to consider the matrix surrounding your cells. The matrix allows cells to grow in an environment more closely resembling the tissue from which they are derived, and in some cases helps them retain some of their unique morphology. Many researchers use water-based gels or solid scaffolds to support cells. While some of these are based on biological materials, such as collagen, others are wholly artificial, such as polystyrene. The “gold standard,” Souza says, is BD Bioscience’s Matrigel—“a soup of extracellular matrix proteins from mouse sarcoma.”
Another method involves suspending cells in solution so they have only each other to cling to. To make culture parallel in vivo biology even more closely, a handful of devices pump media around cells in 2-D or 3-D culture to imitate fluid flow in a living organism.
3-D culture has a variety of uses. By seeding more than one cell type, researchers can better simulate the in vivo milieu. Some bioengineers are starting to assemble stem cells into entire tissues and organs. Cancer biologists, long stymied by tumors’ minimal adhesion to flat dishes, are embracing the technology as well.
Drug screens are also a key potential use for 3-D cultures, but not all techniques work for high-throughput experiments. Scientists performing large screens may wish to focus on devices that easily incorporate dozens or hundreds of individual cultures and are amenable to automation.
Here, The Scientist profiles five methods to raise your cells off the culture dish floor.
SCALE THE SCAFFOLD
3D Biotek, LLC, of North Brunswick, New Jersey, has engineered polymer fibers into a precise meshwork. Just drop the mesh in a dish to give cells a scaffold to grow on. The solid structure has fully interconnected channels, improving media flow compared to random-layout scaffolds, says Carlos Caicedo-Carvajal, manager of research and development at the company. The scaffolds are made of either polystyrene or biodegradable polycaprolactone (J Tissue Eng, doi:10.4061/2011/362326, 2011).
3D Biotek has tested a variety of cell types. Many scientists use the scaffolds to support stem cells during tissue engineering, Caicedo-Carvajal says. Stephen Suh, a user at the Hackensack University Medical Center in New Jersey, cultures patients’ tumor cells on the scaffolds so he can test for drugs that will kill the cancer.
• Because nutrients flow easily through the evenly spaced pores and channels, 3D Biotek scaffolds can be thicker than random versions: up to 600 µm thick, compared to the industry standard of 200 µm, Caicedo-Carvajal says. Thicker cultures can hold more cells and grow longer, he adds.
• Compatible with most 2-D assays
• Cells interact with an artificial, albeit biocompatible, surface.
• Extra effort is needed to pull the cells out of the scaffold, such as longer exposures to trypsin, compared to flat cultures. That makes post-culture experiments a bit more tedious, Suh says.
The scaffolds suffer a problem common to 3-D culture: it’s difficult for nutrients to reach cells in the center. To keep nutrient supplies high, Suh changes the media every 10–12 hours. Another option is to purchase a bioreactor to force fluid flow.
Scale & Throughput
The scaffolds are sized to fit growth vessels from single 100-mm dishes to 96-well plates. Automated high-throughput experiments such as drug screens are plausible, Caicedo-Carvajal says, although most users to date have stuck to small-scale experiments.
Polystyrene scaffolds cost $79 per pack, and the biodegradable version goes for $199 per pack (www.3dbiotekstore.com). The number of scaffolds per pack varies depending on the size; for example, the six-well size comes three to a pack.
Many scientists prefer to forego the scaffold and grow cells suspended in solution, forcing them to self-assemble into spheroids. “The cells are attached to each other and not to a hard plastic”—a more physiological configuration, Takayama says. Cells in a spheroid build their own, natural extracellular matrix. A popular method is to pipet liquid droplets onto the underside of a petri dish lid, so when the lid is placed atop the dish the cells sink to the bottom of the drop. The hanging-drop technique is inconvenient, though, because every time you go to change the media you have to flip the lid to access the drops. Plus, the drops can easily spread and merge, evaporate, or fall if jarred.
To get around the disadvantages of flipping the lid, Takayama developed a special plate with holes in the top (Analyst, 136:473-78, 2011). Each droplet hangs beneath a hole, and researchers can access the drops by pipetting into and out of the holes. The Perfecta3D Hanging Drop Plates designed by Takayama are available through 3D Biomatrix in Ann Arbor.
“Just about any cell type will form spheroids under the right conditions,” Takayama says, but globes of cells are particularly well suited to mimic tumors. User Gary Luker, also at the University of Michigan, studies chemokines in cancer cells using the Perfecta plates (Biochem J, 442:433-42, 2012).
• The plate includes a reservoir of water to limit evaporation from the drops.
• Because of the holes in the lid, you can observe the spheroids without disturbing them.
• A lip around the underside of each hole surrounds each droplet and minimizes spreading or merging with neighboring drops.
• It isn’t feasible to make spheroids much larger than 500 microns in diameter because the cells in the center starve and die.
• The drops can fall if jostled, for example by a slammed incubator door or jerky robot.
“Be patient with it the first time,” Luker says; it takes some practice to line a pipette up with the tiny holes above each droplet.
Scale & Throughput
The plates come in 96-drop and 384-drop formats. The plates are also compatible with pipetting robots and automated plate readers.
$206 for eight plates (www.3DBiomatrix.com)
Instead of using gravity in hanging drops, Nano3D’s Bio-Assembler system relies on magnetism to generate spheroids and other shapes (Nat Nanotech, 5:291-96, 2010). Researchers first feed the cells a suspension of iron oxide nanoparticles called Nanoshuttle. Then, placing a magnet atop the culture dish pulls the cells away from the bottom. “The magnetic field functions as an invisible scaffold,” says Souza, cofounder and chief scientific officer of the company.
Not only can the technique be used to create spheroids from a variety of cell types, it can also make macroscopic structures such as sheets of cells. For example, Nano3D is developing a wound-healing assay with sheets of kidney cells. User Mikhail Kolonin, of the University of Texas in Houston, relies on the Bio-Assembler to grow adipocytes that don’t form properly in 2-D culture.
• The system forces all cell types together, a big advantage in cases where certain types would otherwise separate from a co-culture, Kolonin says.
• For a more organized co-culture, you can use a magnetic probe to pick up and layer tissues like a lasagna.
• Since the only extra equipment is the magnet, using the Bio-Assembler is no more complicated than 2-D culture. To change the media, just moving the magnet to the dish’s bottom will hold the cells down.
• Spheroids will start to suffer necrosis in the center once they reach about 1 mm across, Kolonin says. Researchers can minimize this effect by shortening the culture time or combining several small tissues into a large one on an as-needed basis, Souza says.
• Nano3D does not know the long-term implications of the Nanoshuttle particles for cell health and viability, although they have seen no short-term problems. The particles are excreted into the extracellular matrix within days, while the cells remain suspended by the magnetic matrix.
Souza advises titrating your cells to achieve the desired tissue size and checking to ensure the nanoparticles are innocuous in your lines.
Scale & Throughput
The magnetic lids come in single, six- and 24-well formats; a 96-well version is forthcoming. The company has not tried to automate the culture but it should be feasible, Souza says.
$500 for a 24-well kit, including the magnetic lid and Nanoshuttle (www.n3Dbio.com)
The Microfluidic Cell Array (MiCA) “mimics a small piece of tissue with a blood flow past it,” says Philip Lee, cofounder and director of research and development at CellASIC Corporation (which was recently acquired by EMD Millipore) in Hayward, California. Each flow unit is built into three adjacent wells of a 96-well plate. The cells go in the middle well, supported by a gel and surrounded by a silicone membrane that acts as a “fake epithelial barrier,” Lee says. Flanking the cell well, and connected to it by microchannels, are a slightly raised inlet well for fresh media and a lower outlet well for used media. Gravity pulls media through the system at a rate of 100 µL per day (Biomed Microdevices, 13:753-58, 2011).
CellASIC has experimented with several cell types. Thus far, the most common use is to study tumor biology, Lee says, and scientists have also examined cell migration with the system. At the Gladstone Institutes in San Francisco, Faith Kreitzer uses the apparatus to culture stem cells and differentiate them into germ layers.
• No pumps required. “It’s just a plate that you can throw in the incubator,” Lee says. “If you have a pipette, you can run it.”
• The plate’s bottom is a glass slide, so it offers “really good imaging quality,” Kreitzer says.
• It takes just a few µL of gel per chamber, a significant savings given the high cost of some gels, Kreitzer says.
• CellASIC is still working to compare MiCA to regular cultures without media flowing past the cells, and to validate it as a culture method.
• Kreitzer would like to be able to pluck out just a subset of the heterogeneous cells in her cultures—for example, to sequence a specific type—but currently she can only harvest them all at once.
Kreitzer recommends a hands-on demo to learn important skills, such as distinguishing the inlet and outlet wells and changing the media without creating bubbles that could interfere with the fluid flow. Lee advises users to get all of their experimental parameters, such as output measures, settled using nonflow culture first. That way, you’re not perfecting multiple variables once you start with the MiCA plates.
Scale & Throughput
The 96-well plate holds 32 units. Automation is possible.
One disposable plate costs $150 (www.cellasic.com/MiCA.html).
TALL & SKINNY
Cornell University’s John March doesn’t want to grow cells in just any 3-D shape, but to mimic the specific fingers of intestinal villi. Tall, thin structures are difficult to mold in hydrogel scaffolds because they fall apart when tugged out of a mold. The technique March devised involves a succession of molds. First, he carves villi-shaped holes into plastic using an ultraviolet or carbon dioxide laser, then uses this mold to make silicone villi. Next, he pours alginate over the silicone villi to form an alginate mold, which he fills with the hydrogel he wants to grow his cells on. Finally, he dissolves the alginate with EDTA, leaving the villus-shaped hydrogel. Intestinal cell types will grow over, and into, these structures (Lab Chip, 11:389-92, 2011).
This technique could replicate any long, narrow tissue, such as taste buds or muscle cells. March is studying how intestinal tissue interacts with its resident bacteria, which occupy different niches, such as the tips of villi or the valleys in between them.
• Dissolving the alginate protects the hydrogel from shearing.
• Once fashioned, the inserts simply drop into a well, so it’s easy to use them and to visualize the cells under a microscope.
• The inserts are not yet commercially available.
• Epithelial cells don’t attach well to hydrogels—it’s not known why—so cultures may not last more than a couple of weeks before the cells lift off.
You’ll need to find a hydrogel that supports your cell type and obtain precise physical parameters for a mold that simulates the in vivo structure, March says.
Scale & Throughput
The researchers fashion each insert for a six-well dish by hand; March is working to make larger, intestine-sized versions that might eventually be used for transplants.
The materials cost 50 cents, “tops,” per well insert, March estimates.