The Promises and Challenges of 3D Models

As I dig in to learn more about 3D models, I find a growing mountain of evidence that the differences in biology between 2D and 3D cultures could lead to different, sometimes opposite conclusions in research, and that the 3D models often better represent how cells behave in vivo.

“You can argue that 3D is not absolutely physiologically relevant,” says Research Scientist Drew Niles, “But it’s a step forward. That’s what we’re always doing with models: Trying to get them closer to being representative. I think everyone recognizes that 3D is a step forward.”

As researchers increasingly turned to 3D, Promega scientists recognized the need to formulate assays that would be effective in these systems. The ECM coating in a 3D model often prevents reagents from penetrating to the center of the spheroid. For example, the CellTiter-Glo^® Cell Viability Assay is a bioluminescent assay based on ATP detection, so incomplete lysis meant that much of the ATP present in the system was not measured, producing inaccurate results.

“With the original CellTiter-Glo^® product, we noticed that as structures got larger and larger, there was a deviation between how much ATP we were extracting with acid compared to how much we would get with just following the original procedure,” says Terry. “That’s actually what led to reformulating the ATP detection range to make CellTiter-Glo^® 3D.”

In the case of CellTiter-Glo^®, Promega scientists discovered that the lytic reagent was failing to reach the cells in the center of the spheroid. They increased the lytic capacity of the reagent by adding more detergent and increasing the exposure time, resulting in deeper penetration and more complete lysis. The same logic may not apply to every assay, though. In some cases, no changes to the reagent composition are necessary, such as in the RealTime-Glo™ MT Cell Viability Assay.

A few days after my first visit, I’m back in the lab with Mike. He’s getting ready to assay the cells that I had previously examined. As I watch Mike remove the media from the two plates, I realize that this is the slowest I have ever seen someone transfer only 70µl of liquid. He’s using an electronic 8-channel pipetter, but it takes around 7 seconds to empty each row of wells.

“You have to be careful about how fast you pipette because the microtissues will move around,” Mike tells me. “We actually had to buy pipetters that can run at less than 10µl per second. We go as slowly as possible.”

It’s so slow that it’s almost painful to watch. I easily could have finished the same task in less time using only a single-channel manual pipetter, and I can’t imagine the tedium of working at that speed every day, but Mike insists that it’s safer this way.

These 3D systems are attractive to scientists working in drug discovery, because they will often produce a more accurate reflection of how the drug will affect cells in vivo. For example, Kota et al (2018) used a high-throughput screening approach with 3D spheroids of pancreatic epithelial tumor cells and found a potential drug candidate (Proscillaridin A) that would not have been identified as a selective hit in a traditional 2D approach. They describe differences in drug penetration, cell:cell interactions and hypoxia that they believe contribute to the contrasting results.

Outside of the drug discovery process, 3D models could revolutionize clinical approaches to cancer treatment. Every cancer is genetically distinct, and even single mutations could influence drug responses. Researchers are developing methods of using tumor samples from patients to create 3D organoids, called patient-derived organoids (PDOs). When a panel of compounds with known efficacy profiles is applied to these PDOs, the results have high predictability to the clinical response. These results could guide clinical decisions, and the genome sequences of the organoids could help researchers find correlations between specific mutations and responses to different drugs. All of this could someday streamline treatment regimen development and help patients gain quicker access to the drugs that could likely be most effective for their specific cancer.

As scientists continue improving 3D models, it will be crucial to address questions about standardization and replicability, as well as the financial challenges of widespread implementation. With the right conditions, perhaps these systems could spark a revolution in how we study and treat cancer. Every advance in recapitulating tumor biology in vivo represents massive potential for improving medicine and patient care, but from the researcher’s perspective, these advances are made one tiny spheroid and one excruciatingly slow pipet at a time.