The drug development process is very expensive and time consuming and the attrition rate remains high partly due to the lack of efficacy or unexpected toxicity seen when drugs are tested in clinical trials. The failure rate is often attributed to the preclinical testing, which is done in in vitro disease models that are likely not very predictive of in vivo outcomes and in animal models that are not physiologically relevant to humans. Hence, there is a growing need to find models that are morphologically and physiologically relevant to homeostasis and disease states in humans. This article summarizes some of the key considerations brought up in a recent Biocompare webinar titled Developing High-throughput 3D Cellular Models and Co-cultures

3D disease models

Recent developments in 3D cellular models that mimic cellular microenvironments and aim to close the translational gap—referred to as the “valley of death” in drug development—are showing a lot of promise. They are carefully designed, maintained, and used in a way that recapitulates the key cellular interactions, functions, and morphology of either the normal or disease state being studied. Here are some key considerations when determining what type of model is relevant and how it’s going to be used.

Context of use

What do you want to model? What is the biological question?

  • Determine what it is that needs to be modeled
  • It’s important to first start by generating a model that closely represents the homeostatic or normal state and later use triggers or conditions that will induce the disease state.
  • Use data from literature to understand the disease pathophysiology and figure out what can be used for disease initiation and progression

What level of complexity do you need to generate the model?

  • Tissues in the body consist of various cell types, vasculature etc. It’s important to identify the target cell and then figure out what other cell types need to be incorporated in the 3D model to mimic the cellular environment or state. The more complex the model, the harder it is to maintain, use, and analyze.
  • Test and validate the types of cells that will be grown together, and the experimental conditions needed to generate a viable, reliable co-culture model

Establishing a disease model

Here are some questions that need to be answered before the 3D model can be considered suitable for studying a given disease:

  • Is the disease pathophysiology known and how well can it be mimicked in a 3D model?
  • Can the functional outcomes of the disease be measured using the model?
  • What trigger/s can be used to initiate the disease progression in the 3D model?
  • Does the trigger itself cause any changes in the model?
  • What are the parameters to be studied to validate the disease model?
  • What types of assay readouts are available to monitor the parameters, and can they meet the throughput requirements?

Sourcing the right cells

There are pros and cons associated with choosing the cells that go into creating a 3D co-culture model. Primary cells are the most physiologically relevant, but they are harder to work with and fewer in number. Cells derived from blood or immortalized cell lines are not the most relevant but can be obtained easily and in higher numbers, which is helpful for high-throughput or long-term studies. Hence, its necessary to:

  • Determine the types and numbers of cells needed to generate the 3D models and to perform the various studies
  • Find out if the cells can be used as is or if they will need to undergo polarization or differentiation
  • Test and optimize experimental conditions and media with individual cells before incorporating them into the 3D co-culture model
  • Ensure that the cell itself does not trigger any phenotypic or functional changes in the model

Functional design & optimization

The model should be functionally validated before use and it’s also important to know what outcomes it can be used to measure. Experimental data obtained from different models and studies should be used to set up the protocols and determine markers for use. It’s critical to cross-check the data obtained using the 3D model with that in the literature, keeping in mind that some of the characterization done previously may have been in 2D models. While the data from 2D and 3D models may not match exactly, they should trend in the same direction (although there can be exceptions). Here are some criteria for optimizing functionality:

    • Check compatibility (viability, maturity) of the cells in different media and under various conditions before optimizing the protocols. Set up adequate controls.
    • Use markers to ensure that the cells being added to the co-culture are not affecting the 3D model, the processes and functional outcomes
    • Figure out which assays, protocols and markers can adequately monitor the functional outcomes and under what conditions

When building a good in vitro model, it’s always important to start by asking all the relevant questions. Knowing all the limitations and conditions under which the model is built and applied will help determine the efficacy of the model and the predictivity of the results obtained. A good understanding of the biological question will ensure that the model is fit-for-purpose. Sometimes a 2D model is all you need. When modeling a disease, knowing how much of the pathophysiology can be truly recapitulated in vitro is critical. When testing potential drug candidates, making sure that the model can show reversal of the disease phenotype is a must. No model can provide all the answers, but finding one that can offer the best results is imperative for drug developers.