Roundtable on Opportunities and Challenges in Developing 3-Dimensional Human Tissue Models to Study Musculoskeletal and Skin Physiology and Pathophysiology

November 24, 2014

Background

The goal of this roundtable was to discuss the "Opportunities and Challenges in Developing 3-Dimensional Human Tissue Models to Study Musculoskeletal and Skin Physiology and Pathophysiology." Current biomedical research is frequently carried out in a 2-dimensional (2-D) tissue culture environment, despite the fact that human tissues are 3-dimensional (3-D) structures that require the interactions of multiple cell types with one another and with the environment to maintain their shape and function. Most primary cells in 2-D cultures quickly lose in vivo properties such as tissue-specific gene expression, cell polarity, and cell-cell and cell-matrix contacts.

Increasing knowledge about cell biology, materials science, and microfabrication technologies is enabling scientists to engineer functional units of human tissues (i.e., 3-D human tissue models). These highly controllable in vitro model systems would closely mimic functions of human tissues or organs. They could be used to study normal developmental biology or disease pathogenesis, or for drug screening. Patient-specific models could be generated to study rare diseases or diseases where animal models are not available. They would go beyond the 2-D "flat biology" to increase the complexity and diversity of in vitro models and assays, and bridge the gap between simple cell cultures and the full complexity of animal models. They also have the potential to minimize or replace animal testing.

Prototypic studies of 3-D tissue models have begun to underscore their importance in basic and translational research. For example, bone-forming cells are able to bridge gaps in broken bones when the spaces are filled with implants that approximate natural bone architecture, but good test beds to study bone remodeling are needed to advance this work beyond the empirical stage. Although researchers have used 3-D models of skin for some time, these models are still relatively simple and involve only a few cell types. They do not include hair, sweat glands, or other complex skin appendages, nor do they maintain homeostasis. Creation of these models will require broad-based interdisciplinary science, and has potential applications across multiple NIAMS mission areas.

In 2011, the NIH, collaborating with the U.S. Food and Drug Administration (FDA) and Defense Advanced Research Projects Agency (DARPA), launched a "Tissue Chip for Drug Screening" program as part of the NIH Common Fund (http://www.nih.gov/news/health/sep2011/od-16.htm). Through the program, researchers will develop 3-D human tissue chips that accurately model the structure and function of human organs (http://www.ncats.nih.gov/research/reengineering/tissue-chip/tissue-chip.html). The goal is to use the chips "to predict whether a candidate drug, vaccine or biologic agent is safe or toxic in humans in a faster and more cost-effective way than current methods." Three of the projects funded under this Common Fund initiative are relevant to the NIAMS mission (http://www.ncats.nih.gov/research/reengineering/tissue-chip/projects/awards-2014.html), indicating a growing interest in generating complex in vitro models.

In advance of the meeting, participants were encouraged to consult with colleagues on several questions:

  • What specific NIAMS questions can be best answered by 3-D human tissue models?
  • Are there appropriate in vitro 3-D tissue models for common and rare diseases of interest to NIAMS and are any of these models ready for
    • studies of disease pathogenesis?
    • functional studies of gene variants?
    • high-throughput drug screening?
    • drug toxicity studies?
    • testing of therapies?
  • What are the three greatest challenges in developing these models? What are the potential options for overcoming these challenges?
  • What innovative, creative scientific approaches are needed to advance research in developing these models?

These topics and responses collected by the participants from their research communities in advance served as the basis for the discussion. A brief summary of the common and crosscutting themes in the responses is listed in the textbox. Although not all responses were discussed at the roundtable, NIAMS leadership and the appropriate program staff read each comment. The NIAMS greatly appreciates the community’s input on these questions.

Common/Crosscutting Themes

  • Cell sourcing. Noting that primary cells are in limited supply and present diversity and reproducibility challenges, participants noted the need for:
    • Uniform primary cell isolation protocols and characterization standards.
    • Cell expansion protocols.
    • Cell banks representing population diversity.
    • Patient-specific cell banks.
    • Standardized differentiation protocols to generate many types of cells from induced pluripotent stem cell (iPSC) lines.
  • Need to increase the complexity of 3-D human tissue models.
    • Cellular complexity could be improved with the inclusion of:
      • Immune cells.
      • Nervous system.
      • Vasculature.
    • The following elements should be considered when engineering the microenvironment:
      • Cellular and extracellular matrix components.
      • Stem cell niche.
    • The capability of cells to create their own microenvironment if given an appropriate overall scaffold/structure.
  • Need to improve the long-term survival of 3-D human tissue models.
  • Need to incorporate systemic effects and integration of musculoskeletal and skin tissues with other organs.
    • Metabolism and elimination of drugs by other organs (e.g., liver).
    • Need a common culture medium that supports all organs.
  • Modeling chronic diseases in the NIAMS mission areas is difficult.
    • Need mature tissue phenotypes or even aged 3-D human tissue models.
    • May need patient-specific cell sources.
    • May need systemic and environmental inputs.
  • Outcome measures and readouts for 3-D human tissue models:
    • Need to be validated against in vivo.
    • Should be physiologically/functionally relevant.
    • Real time, non-destructive monitoring is desirable.
  • Genomic and epigenomic editing tools are essential and may be particularly useful for monongenic disease modeling. Tools are needed for:
    • Functional genomic studies.
    • Gene variant studies.
  • Need multi- and interdisciplinary teams to build and improve 3-D human tissue models.
  • Need standards for all aspects of 3-D human tissue modeling.

 

Opportunities to Study Musculoskeletal and Skin Physiology and Pathophysiology Using 3-D Human Tissue Models

When discussing "What specific NIAMS questions can be best answered by 3-D human tissue models?" participants noted that 3-D models of mineralized bone that include osteoblasts, osteoclasts, osteocytes, and the bone marrow compartment and respond to various physiological stimuli could advance the development and testing of potential osteoporosis therapies. Functional skeletal muscle-bone models would be important for aging, exercise, and muscle-disease-related research. The NIAMS scientific community is also well-positioned to make 3-D constructs of human skin for studying skin biology or monogenic skin diseases and for screening potential therapeutic compounds to treat skin diseases. Development of sophisticated skin models that incorporate multiple skin components such as hair follicles, nerves, sweat glands, immune cells, and fat would enhance these and related research opportunities.

Many diseases and disorders within the NIAMS mission involve inflammation and immune responses. For example, 3-D models could provide powerful platforms to study innate and adaptive responses in the presence and absence of different cell types or triggers. Models could be developed to explore how specific components of the immune response lead to tissue damage that characterizes autoimmune diseases. 3 D models of joints that recapitulate inflammatory processes would enable studies into the complex etiology and pathology of osteoarthritis and rheumatic joint diseases. 3-D cultures also could advance research on paracrine signaling between keratinocytes and immune cells during inflammation. Comments collected before the meeting raised the potential that 3-D models could be used to study host-microbe interactions in health and disease. Furthermore, the shared interest in inflammation and immunity among different research communities may provide a foundation for collaborations among groups working in seemingly unrelated areas (e.g., skin and cartilage).

Because the tissue microenvironment plays a critical role in the epigenetic regulation of gene expression, the growth of human cells in a 3-D environment provides an in vitro platform from which investigators can elucidate gene regulatory networks and signaling pathways that modulate stem cell behavior in vivo during development and tissue repair. In addition to serving as model systems for confirmatory studies of previously identified genes or genetic variants, human tissue models can provide ideal platforms for discovery of basic cellular, transcriptional, and metabolic processes. As one example, participants noted that 3-D cultures could provide an environment for studying how and when muscle protein isoforms change.

As the complexity of developmental models increases, investigators could study how the availability of growth factors or other molecules in the cellular environment influence normal development and congenital disorders within the NIAMS mission. 3-D human tissue models could also be used to screen drug candidates and to test new therapies for toxicity and efficacy.

Challenges in Developing 3-D Human Tissue Models to Study Musculoskeletal and Skin Physiology and Pathophysiology

Selecting appropriate cell sources (adult stem cells, iPSCs, embryonic stem cells) for various models was cited as one of the major challenges. Heterogeneity within and among cells from different sources creates challenges when screening compounds for toxicity or potential benefit or when modeling diseases, and also provides an opportunity to understand basic biology and cell behavior. For example, repeated studies using cultures and models developed with cells collected from multiple male and female patients could provide insights into sex differences in disease susceptibility, progression, and response to treatment. Variation also could be leveraged to develop robust 3-D tissue cultures by combining cells from different donors to build different tissue compartments (e.g., building a 3-D skin culture from cells that make good dermis, cells that allow skin to heal instead of scarring, cells that produce a mature keratin layer).

Heterogeneity in cell populations is important for normal development, homeostasis, and regeneration, but the role of subpopulations is frequently poorly understood. Lack of knowledge about the factors that influence variations among cells was also cited as one of the major challenges. Cells that look indistinguishable in culture may display different gene expression patterns. Much variability might be attributed to differences in the cells’ immediate environments, or micro-niches, but little is known about these factors or the mechanisms by which they influence cell behavior.

Cell differentiation in culture is controlled by adding or subtracting nutrients and growth factors from cell culture media. Current cell culture conditions are generally optimized for growing one or two types of cells and are not effective for building integrated 3-D human tissue models. As tissues add and subtract circulating factors (e.g., nutrients, waste products, metabolites, and signaling molecules) in vivo, other tissues respond to these changes in a complex series of sensitive feedback loops. Therefore, the order in which tissues are linked in series in a multisystem model will affect levels of nutrients, growth factors, and waste products, likely influencing each tissue’s behavior. It is essential to develop a universal culture medium that mimics blood’s ability to transport a wide variety of peptides, proteins, and other molecules.

In addition to needing viable populations of progenitor cells and the right culture conditions, 3-D cultures of bone, joint, muscle, and skin tissues need the correct mechanical stimuli for proper development, maintenance, and repair. Some of the stimuli can come from the matrices on which the cells are grown; others can be added with loading apparatuses or pulse vacuums. The availability of miniaturized mechanical loading devices was mentioned as an unmet need.

Incorporating vasculature (including capillaries and the lymphatics) and nerves was viewed as another important aspect of 3-D human tissue models. For example, innervated and vascularized 3-D models of human muscle could allow researchers to better study muscle diseases. Advances in technology such as the 3-D printing technology may make it easier to add vasculature into cell-based models.

Another challenge is recreating the aging process in culture. Many of the bone, joint, muscle, and skin conditions within the NIAMS mission become more common with age, and the relevant biological changes that lead to these conditions occur over years or decades. However, researchers are creating immature tissues, many of which begin to degrade in culture long before they are fully mature. The ability to create age-appropriate 3-D tissues and maintain them in culture for extended timeframes would allow investigators to examine issues related to chronic diseases and aging, the effects of damage in one component on others, and how older tissues respond to therapies. Development of tissue-specific bioreactors may help with this issue.

3-D models are complementary to whole organism studies. The scientific community needs standardized definitions and measures that reflect the extent to which a model mimics the tissue or condition of interest. 3-D human tissue models are intended to be minimally functional models instead of recapitulating all features of a tissue in vivo. The definition of "minimally functional" varies with the question that the model is being asked to address. Disease models must display patient or disease-relevant phenotypes to be useful for the study of a disease or response to therapy. However, readouts that correlate with patient experiences such as pain are likely to be far in the future.

As in any field, investigators who are developing and studying 3-D human tissue models are creating the necessary tools and research methods. Many of these advances are being made in partnership with engineers and computational experts. Remaining needs, as identified by the participants, include:

  • Technologies to capture and quantify the molecular, cellular, and functional readouts described above. These would include nondestructive methods such as imaging for repeatedly assaying 3-D models.
  • Improved labeling of target cells in live tissues and other methods for cell identification and monitoring.
  • Development of a “universal medium” to culture multiple cell and tissue types in an integrated system.
  • Genomic and epigenomic editing approaches for studying gene variants and modeling monogenic diseases.
  • Strategies to manage and analyze large amounts of data.
  • An open access catalog of protocols, including strategies for overcoming issues related to scale-up or miniaturization.
  • Commercially available cell lines, reagents, bioreactors that recapitulate native microenvironments, and other resources that researchers currently have to produce and maintain in their individual labs.

Participants

AYEHUNIE, Seyoum, Ph.D., MatTek Corp
BRENNER, Michael B., M.D., Harvard Medical School/Brigham and Women's Hospital
BURDICK, Jason A., Ph.D., University of Pennsylvania
BUXTON, Denis, Ph.D., National Heart, Lung, and Blood Institute, National Institutes of Health
CHRISTIANO, Angela M., Ph.D., Columbia University
GARLICK, Jonathan, D.D.S., Ph.D., Tufts University
GETSIOS, Spiro, Ph.D., Northwestern University Feinberg School of Medicine
GUILAK, Farshid, Ph.D., Duke University (co-chair)
HUNZIKER, Rosemarie, Ph.D., National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health
LUMELSKY, Nadya, Ph.D., National Institute of Dental and Craniofacial Research, National Institutes of Health
ORO, Anthony (Tony), M.D., Ph.D., Stanford University School of Medicine
SILVER, Alex, M.B.A., The Jackson Gabriel Silver Foundation / EB Research Partnership
TAGLE, Danilo A., Ph.D., National Center for Advancing Translational Sciences, National Institutes of Health
TRUSKEY, George A., Ph.D., Duke University
TUAN, Rocky S., Ph.D., University of Pittsburgh School of Medicine

NIAMS

BAKER, Carl, M.D., Ph.D.
CARTER, Robert H., M.D.
DRUGAN, Jonelle K., Ph.D., M.P.H.
KATZ, Stephen I., M.D., Ph.D. (co-chair)
KESTER, Mary Beth, M.S.
LINDE, Anita M., M.P.P.
McGOWAN, Joan A., Ph.D.
MOEN, Laura K., Ph.D.
SERRATE-SZTEIN, Susana A., M.D.
WANG, Fei, Ph.D. (co-chair)