2D vs 3D Cell Culture: Choosing the Right Model for Disease Research

In our last article, we compared 2D organ-on-a-chip devices and 3D organoids with respect to their use in drug discovery, highlighting their importance in modeling diseases and evaluating efficacy and safety during drug discovery and development . We also presented the major catch-22 challenge in disease modeling; disease biology is complex and developing a ‘good’ disease model requires a deep understanding of disease mechanisms and pathogenesis, both of which come through gathering Real-World Evidence and modeling the disease. In this article, we take a closer look at some of the most important considerations for choosing an approach to model disease, focusing specifically on 2D vs 3D models with a few notable examples.

While an animal model offers the obvious advantage of studying disease in a living whole-organism context, developing and working with animals is not only tedious, time-consuming and costly but increasingly challenged by evolving guidance on their use in drug discovery and development. Of note, the FDA’s Modernization Act 2.0, enacted in 2022, and the proposed FDA Modernization Act 3.0 (April 2025), specifically promote the use of human-relevant cell-based assays as an alternative to animals (1, 2). Meanwhile, cell-based methods are faster and much more cost-effective to set up and run, with greater versatility and flexibility in choice of experimental endpoints. These advantages are strengthened when human cells are used, which is increasingly becoming easier thanks to advances in induced pluripotent stem cell (iPSC) technology. 

Cell Source Considerations

When it comes to cell-based assays to study disease (or any biological process, really), two important decisions will need to be made regarding cell source. Firstly, which species to work with and secondly, whether to use primary cells, immortalized cell lines (such as HepG2 and HEK293), or stem cell-derived cells such as iPSCs. Unlike immortalized cell lines that can proliferate indefinitely, iPSCs maintain their stemness and differentiation potential through self-renewal mechanisms, but they are not immortalized in the traditional sense; they require specific culture conditions and have passage limits to maintain their pluripotent state and genomic stability, as detailed in our previous article.

Let’s look at these decisions in a bit more detail:

Firstly, why choose human cells over non-human cells?

Despite similarities in the way that humans and non-human animals experience certain diseases, there will always be some level of species-specific variation in pathobiology owing to possible sequence and structural variations in the disease-causing protein(s), as well as how the immune system and metabolic pathways react to and are affected by the disease. 

Going a step further to consider the modeling of a new drug for any given disease, the way that drug interacts with the target (e.g., signal transduction pathways, receptor-activated pathways, immune reactions/biomarker expression patterns, and cell-based phenotypic processes such as trogocytosis) will also differ between humans and other animals due to species variations in metabolism that impact distribution, bioavailability, excretion and other pharmacokinetic parameters. Nowhere else is this lack of translatability between animals and humans more widely documented than in the Alzheimer’s disease field. According to a comprehensive review published in 2022, 98 unique compounds failed in Phase II and III Alzheimer’s clinical trials between 2004-2021, despite many showing promise in preclinical animal studies (3). As one notable example, Semagacestat, which was under clinical development by Eli Lilly, actually made disease symptoms worse in humans in clinical trials despite marked reduction of amyloid beta plaques in animal studies. 

Why iPSC-derived vs primary human cells?

iPSC-derived cells, discussed in detail in a previous article, offer many in vitro experimental advantages over primary human cells, including: 

• Consistency and reproducibility. Because iPSCs are capable of indefinite proliferation in culture (when maintained under appropriate pluripotent conditions), it is possible to scale them for large studies without compromising reproducibility or consistency. This is in contrast to primary cells, where cell numbers are often limited and the cultures are short-lived, making it impossible to repeat experiments with exactly the same cell type.
• iPSCs are often easier to work with in culture than primary cells. This can be the case for several reasons; some human primary cells do not attach well to culture vessels and require the presence of specialized substrates to do so, and/or do not survive for longer than ~48-72 hours in culture. Additionally, when using specialized assay plates (such as those for impedance measurements), compatibility with different coating materials can be limited, making it challenging to provide the optimal attachment conditions that some primary cells require. One relevant example is human iPSC-derived Kupffer cells, which are much easier to culture than primary human Kupffer cells isolated from the human liver. Importantly, human iPSC-derived Kupffer cells can be matured as long as the ‘appropriate’ cell culture conditions and media formulations are provided.
• iPSCs offer the possibility to work in a traceable donor-pure (i.e., no donor mixing) genetic background, for example by allowing patient-specific disease modeling by comparing wild-type and diseased patient-derived iPSCs. Advances in precision gene-editing technologies also offer the possibility to create isogenic cell lines, that is pairwise mutation vs. control iPSCs. 

The use of iPSCs from a known source (e.g., a known patient) also eliminates donor variability and unknown exposure history. However, donor selection remains an important consideration for many disease-related therapeutic development programs. Cells from a diverse source of donors (including different ethnicities, ages, and other demographic factors) have proven useful for nonclinical efficacy and safety studies during drug development and can provide valuable insights for clinical trial design. For example, for rare pediatric cancers, the United States FDA requires a diverse panel of ethnicities/donors’ samples to be tested (4, 5).

2D vs 3D cell culture considerations? How to choose?

Two-dimensional (2D) cell cultures are routinely used in disease modeling and compound/drug screening. These consist of cells grown as single layers on flat surfaces such as a culture dish or microplate. Depending on the experimental question, the cells are then exposed to small molecule drugs, siRNAs or other agents and the resulting phenotypes are evaluated using a diverse range of assays. While 2D cultures are popular due to ease of handling, availability of well-established protocols, scalability and cost effectiveness, they have certain limitations that can cause a disconnect between in vitro research and the in vivo situation. These include changes in cell shape and function as a result of being forced to grow on a flat surface, the lack of a physiologically-relevant microenvironment as well as the absence of cell-cell and cell-matrix interactions that occur in vivo (4 and references within). This difference is very pronounced in hepatocytes, which express markedly different cytochrome P450 (CYP) profiles in 2D vs 3D cultures.

In contrast to 2D cultures, 3D formats are designed to recapitulate, albeit to varying extents, the spatial architecture and cellular complexity of in vivo tissues. 3D formats include organoids (self-assembling multicellular structures derived from iPSCs or patient tissues), spheroids (simple 3D cell aggregates), and organ-on-a-chip systems (microfluidic platforms with controlled microenvironments). It is well-accepted that 3D models better mimic in vivo conditions through enhanced cell-cell interactions, tissue-like organization, physiologically-relevant microenvironments, and the ability to incorporate multiple cell types with diverse morphologies. However, while offering greater biological relevance and predictivity for drug testing and disease modeling compared to 2D cultures, 3D models demand greater technical expertise, specialized equipment, longer experimental timelines, better/sophisticated software (including AI-driven analysis software for large data sets), and are more expensive.

While 2D and 3D models each have their own pros and cons, the following may help you to decide which way to go, depending on the experimental setup and goal. 

Cell Density Considerations

Cell density requirements differ significantly between 2D and 3D culture systems. In 2D cultures, density is calculated in a straightforward way by surface area (number of cells per cm² of culture flask), with specific requirements depending on the assay type, e.g., imaging or electrophysiology experiments typically need fewer cells for visualization and measurement, while biochemical assays such as ELISA or Western blotting require higher cell densities to generate enough protein for analysis. 

In contrast, 3D culture density considerations focus on establishing the appropriate cell types and ratios needed to generate functional organoids or tissue models. The specific cellular composition depends on both the biological system being modeled and the experimental requirements, including assay duration, as longer culture periods may require different initial seeding ratios to maintain tissue architecture and function over time. To optimize cell density, we recommend that you clearly define assay goals and experimental limitations. For example, if you are looking for secreted factors from microglia in an organoid containing CNS cell types, you should aim to maximize the % of the microglia relative to the other cell types. Otherwise, you will need >8 organoids to run a single ELISA dosage/condition and that will not be cost-effective.

What aspects of the disease do you want to model? 

The exact disease features to be recapitulated will be defined by the disease of interest, but generally includes some of the following: phenotypic changes, gene and protein expression signatures, biomarker expression, disease progression over time and secondary consequences, inflammatory/immune responses (if relevant), and activation of key pathways involved in human disease, e.g., fibrosis, inflammation, cell death.

When selecting a disease model, consider whether you want to study the root cause of disease (if known) versus the consequences of disease. You may also want to study a phenotypic aspect of the disease, including both acute and chronic effects. Critical questions should include: Is the disease in the model caused in the same way as in humans, i.e., does the model contain the same mutation(s)? If studying disease progression and secondary consequences, can the model demonstrate these changes within a reasonable culture timeframe?

Example: Modeling Alzheimer’s Disease

It’s important to recognize that no single 2D or 3D culture system or individual assay can faithfully model all aspects of a disease. Therefore, the most comprehensive approach involves combining different cell-based models and complementary assays to capture the full spectrum of disease features. 

To model Alzheimer’s disease, for instance, one would plan assays to model the key known aspects of the disease. Bear in mind that for Alzheimer’s and many other neurodegenerative diseases, no consensus exists regarding the mechanisms of disease.

Here are a few examples of assays that could be used to model Alzheimer’s; this list is not exhaustive (6-13):

• Amyloid pathology. Aβ42/Aβ40 ratio measurements and amyloid aggregation assays (e.g., Thioflavin T) to study plaque formation and amyloid processing
• Tau pathology. Tau phosphorylation assays using AT8- and PHF-1 binding antibodies to model neurofibrillary tangles and tau spreading between neurons
• Neuroinflammation. Cytokine profiling (e.g., IL-1β, TNF-α, IL-6) and microglial activation assays to assess anticipated inflammatory responses and immune dysfunction
• Synaptic dysfunction. Synaptic marker expression and electrophysiological recordings to measure synaptic loss and network activity
• Neuronal death. Cell viability assays (e.g., caspase-3, lactate dehydrogenase (LDH) release assay and neurite outgrowth measurements to study neurodegeneration and damage to axons
• Mitochondrial dysfunction. Autophagy markers such as LC3 and P62 can be measured to assess protein clearance mechanisms and mitochondrial toxicity assays (e.g., ATP and ROS production, and oxygen consumption rate) can be used to shed light on how the mitochondria are responding to cellular stress

Which format will you choose? 

As highlighted in this article, choosing between 2D and 3D cell culture models depends largely on your experimental goals. While 2D formats are ideal for high-throughput screening and mechanistic studies, 3D models better recapitulate disease biology for physiologically relevant studies. Human iPSC-derived models offer significant advantages over animal models, and align with recent FDA guidelines promoting human-relevant, non-animal testing methods.

Are you finding it difficult to choose the right approach for your project? Please feel free to contact our expert tech support team here.

References:

1. Han JJ. FDA Modernization Act 2.0 allows for alternatives to animal testing. Artif Organs. 2023 Mar;47(3):449-450. doi: 10.1111/aor.14503. Epub 2023 Feb 10. 
2. Carratt SA, Zuch de Zafra CL, Oziolor E, et al. An industry perspective on the FDA Modernization Act 2.0/3.0: potential next steps for sponsors to reduce animal use in drug development. Toxicol Sci. 2025 Jan 1;203(1):28-34. 
3. Kim CK, Lee YR, Ong L, et al. Alzheimer’s Disease: Key Insights from Two Decades of Clinical Trial Failures. J Alzheimers Dis. 2022;87(1):83-100.
4. U.S. Food and Drug Administration. Enhancing the Diversity of Clinical Trial Populations – Eligibility Criteria, Enrollment Practices, and Trial Designs: Guidance for Industry. November 2020. 
5. Russell ES, Aubrun E, Moga DC, et al. FDA draft guidance to improve clinical trial diversity: Opportunities for pharmacoepidemiology. J Clin Transl Sci. 2023 May 2;7(1):e101. 
6. Juarez-Moreno K, Chávez-García D, Hirata G, Vazquez-Duhalt R. Monolayer (2D) or spheroids (3D) cell cultures for nanotoxicological studies? Comparison of cytotoxicity and cell internalization of nanoparticles. Toxicol In Vitro. 2022 Dec;85:105461. 
7. Gade Malmos K, Blancas-Mejia LM, Weber B, et al. ThT 101: a primer on the use of thioflavin T to investigate amyloid formation. Amyloid. 2017 Mar;24(1):1-16.
8. Kwak SS, Washicosky KJ, Brand E, et al. Amyloid-β42/40 ratio drives tau pathology in 3D human neural cell culture models of Alzheimer’s disease. Nat Commun. 2020 Mar 13;11(1):1377. 
9. Strang KH, Goodwin MS, Riffe C, et al. Generation and characterization of new monoclonal antibodies targeting the PHF1 and AT8 epitopes on human tau. Acta Neuropathol Commun. 2017 Jul 31;5(1):58. 
10. Gao C, Jiang J, Tan Y, Chen S. Microglia in neurodegenerative diseases: mechanism and potential therapeutic targets. Signal Transduct Target Ther. 2023 Sep 22;8(1):359. 
11. Shankar GM, Walsh DM. Alzheimer’s disease: synaptic dysfunction and Aβ. Mol Neurodegener. 2009 Nov 23;4:48. 
12. Linsley JW, Reisine T, Finkbeiner S. Cell death assays for neurodegenerative disease drug discovery. Expert Opin Drug Discov. 2019 Sep;14(9):901-913.
13. Zhang J, Zhang Y, Wang J, et al. Recent advances in Alzheimer’s disease: mechanisms, clinical trials and new drug development strategies. Signal Transduct Target Ther. 2024 Aug 21;9:211.

Karen O’Hanlon Cohrt is an independent Science Writer with a PhD in biotechnology from Maynooth University, Ireland (2011). After her PhD, Karen relocated to Denmark where she held postdoctoral positions in mycology and later in human cell cycle regulation, before moving to the world of drug discovery. Karen has been a full-time science writer since 2017, and has since then held numerous contract roles in science communication and editing spanning diverse topics including diagnostics, molecular biology, and gene therapy. Her broad research background provides the technical know-how to support scientists in diverse areas, and this in combination with her passion for learning helps her to keep abreast of exciting research developments as they unfold. Karen is currently based in Ireland, and you can follow her on Linkedin here.