Q: Why the sudden surge in interest for using 3D cell cultures?

A: Many scientists, like Dr. Mina Bissell, Dr. Joan Brugge, and others, have been advocating the move to 3D cell cultures for a long time now, but developing 3D cultures is complicated. Selecting the right methodology to generate the appropriate 3D in vitro cell culture systems, and having the technology to correctly interpret the data obtained using these culture systems, is more complicated. In the early days, these complexities were somewhat of a deterrent. However, after big pharma spent billions of dollars on 2D monolayer cell culture models that showed promise in preclinical drug development but didn’t translate to the clinic, it became very apparent that, at least for cancer research, we were using the wrong models.

Spheroid Morphologies of various cancer cell lines grown on matrix free 3D plates. Panc-1 (pancreatic cancer), MDA-MB-231 (triple negative breast cancer) HT29 (Colon) MiaPaCa2 (pancreatic cancer) TC71 (Ewings Sarcoma)When grown on non-adherent surfaces, cancer cells have an inherent tendency to migrate and form clusters that turn into 3D multicellular tumor spheroids. Initially, these spheroids have a rather loosely organized architecture, but in time the cells secrete an extracellular matrix that results in a compact spheroid having a hypoxic inner core with physiological characteristics resembling what is often seen in the 3D tumor microenvironment. Thus, with 3D models we can replicate in a laboratory some important properties that we see in a tumor. Another advantage with 3D systems is that we can carry out co-culture studies. For instance, we can grow tumor cells with fibroblasts and better understand the cross communication between cells, another important feature of the multicellular tumor microenvironment. 3D cultures certainly resemble the tumor architecture and offer huge benefits, but one has to keep in mind the cost and the added patience needed to optimize these systems in order to maximize the benefits.

Q: When did you start using 3D cultures in your lab?

A: I was always interested in setting up 3D cell cultures, but when you are running a high-throughput screening (HTS) facility you also have to take into account the costs. A single plate used in a 3D screen can cost you up to $100. So when you run a genome-wide screen you can end up spending approximately $50,000 just for the plates. However, I was convinced that we had to move from 2D to 3D, so I started out by doing very small screens. We saw, and it’s also shown by others, that when we compare the expression profiles of a selected panel of proteins important in a particular cell line, there is significant overlap in the expression profiles between 3D cultures and tumor tissues. However, the 2D expression profile is a complete outlier. Hence, for identifying targets for drug development and for drug screening purposes we are moving more toward using 3D cell culture systems.

However, when using 3D screens, you also have to develop technologies that will help you get reliable and trustworthy data. For instance, we helped develop a 3D scanner that can accurately measure parameters like volume, area, and cell viability to get good readouts. We also have an IN Cell Analyzer 6000 Imager that enables us to capture detailed images of the spheroid morphology and identify relevant treatment-induced alterations of these morphologies. If one spends the time to develop a system and uses it correctly, then the data that you obtain from these 3D screens tend to have significant clinical relevance.

Q: How much of your screening is now in 3D cultures?

A: Right now 30 to 40 percent of our screening is in 3D cultures. We still offer both 2D and 3D models for screening, as some investigators are skeptical about 3D screening. We have completed three screens in 3D and we have three manuscripts coming out soon. At the present time we have two ongoing screens using 3D multicellular tumor spheroid models. Once these manuscripts get published and the 3D systems prove their value, then I expect more folks to start using them. The other factor is the cost. The cost of doing a screen with 3D cultures is at least two to three times more expensive. However, a lot of companies are now coming out with new plates and reagents for 3D culture, and as more laboratories use them I expect the prices will start coming down in a couple of years.

Q: How did you go about evaluating what 3D models to use for your studies?

A: There are many techniques that one can use today to generate 3D cell culture models. These include matrix- and non-matrix-based systems, the spinning flask and the hanging drop methods. As our objective is to use 3D systems for both HT RNAi screens and small molecule screens, it was imperative that we use non-matrix-based systems. We looked around for almost a year and came across a company, SCIVAX Life Sciences Inc. that manufactured cell culture plates of non-matrix transparent cycloolefin resinous sheets comprising nanoscale indented patterns. Since this plate is a non-matrix-based system we can transfect the cells with siRNA and seed cells on this plate, which then form spheroids and, using viability or morphological readouts, determine the biological significance.

Depending on your purpose you have to be very careful in selecting the right technology to generate your spheroids. In a hanging drop method, cells accumulate at the bottom of the drop by gravity and form spheroids. Using this model you get uniformly sized spheroids, whereas in the plate-based system that we use, you get multiple spheroids but they are not all of uniform size. We use the ultra-low-attachment round-bottom plates from Corning, where cells form large single spheroids, just as in the hanging drop method. We are currently, in collaboration with n3D Biosciences Inc., developing an advanced 3D model using polylysine-coated nanoshells and a bio-assembler system. Using this model, we are able to generate a tumor in the laboratory utilizing patient biopsy tissue. Our goal is to optimize this model for it to be used in the new therapeutic approach of personalized medicine. So depending on what system you are using, you can generate 3D structures of different morphology to be used to address very specific and relevant biological questions.

Q: What changes did you have to make in your lab or in the protocols when using 3D cultures?

A: There were no major changes made, except that the plates used were different. The other difference is that when we use 96-well plates for 2D cultures we normally seed between 5,000 and 10,000 cells per well. For 3D cultures, we use 10,000-40,000 cells per well for seeding, because the number of cells is very important in determining the integrity of the 3D structure and is dependent on the cell line used. So once you have identified which technique you are going to use to generate the 3D spheroids, then you have to test a range of seeding densities using your cell line in order to identify the concentration that gives spheroids with the best integrity. The other aspect to consider is that not all cell lines will form spheroids. For example, some lung cancer cell lines will form spheroids two to three days after plating, but then they break up. So you have to be very careful in selecting the right cell line for your study.

There are other simple but really important things to keep in mind. For instance, when washing the cells in a 2D adherent cell line you can just suck out the media and replace it with new media. However, spheroids are not attached and they can get sucked out while exchanging the media. So we had to develop a new method for changing the media to enable their continuous growth. At the same time, with spheroids we can get away with changing the media every three to five days, instead of every day.

Q: What are some of the things that need to be optimized for 3D cultures?

A: When doing siRNA screens in 3D, we were using higher amounts of cells as compared with screens in 2D, so we scaled up the siRNA and lipid concentrations accordingly. However, higher concentrations of siRNA and lipids lead to increased toxicity to the cells, and so we had to come up with a fine balance. In the case of small molecule screens, the time when the drug is added to the plate is an important consideration. Some cell lines tend to form tight spheroids and thus, if the drugs are added after the formation of the spheroids, one has to consider the effects associated with drug penetration. Other considerations are drug stability in media and duration of drug treatment before terminating the study. So there are a number of factors that need to be worked out, and we learned this the hard way as we better understood our 3D systems. We are constantly tweaking protocols and talking to companies as they put out new reagents and assay technologies that are specific for 3D culture use. We have developed our technologies over time, through trial and error, and we have fairly good standard operating procedures to get robust and reproducible data for our siRNA and drug screens and also for mechanism-of-action studies. We now have data to show that 3D systems are superior to 2D culture systems, at least for cancer biology.

Dr. Geoffrey Bartholomeusz is an associate professor and director of the siRNA Screening Services in the Department of Experimental Therapeutics at MD Anderson Cancer Center in Texas. He has worked extensively as a cancer biologist, with expertise in both molecular biology and drug development. A goal of his research is to utilize high-throughput siRNA screens in a 3D cell culture assay to identify novel targets regulating the tumor architecture. His hypothesis-driven study proposes that altering the tumor architecture will lower the levels of hypoxia within solid tumors, sensitizing these tumors to irradiation and/or chemotherapy. In an attempt to confirm this hypothesis, his team has developed a 3D spheroid cell culture model that has similarities to hypoxic regions of solid tumors. They have performed a high-throughput siRNA screen and identified and validated potential targets whose silencing reduced the levels of hypoxia within the spheroid and inhibited HIF1 activity. Studies are currently ongoing to confirm the hypothesis and test small molecules developed against these targets as potential anticancer agents. As a member of a multi-investigator group within MD Anderson Cancer Center, Dr. Bartholomeusz is also involved in developing a technology that will permit them to regenerate tumors in the lab utilizing biopsy tissue, with the goal of developing cancer-patient-specific drug cocktails.