2017 Progress Report: Organotypic Culture Model to Analyze DevelopmentalLimbMalformationsResulting from Toxicant/Teratogen Exposure

EPA Grant Number: R835736C002
Subproject: this is subproject number 002 , established and managed by the Center Director under grant R835736
(EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).

Center: Mickey Leland National Urban Air Toxics Research Center (NUATRC)
Center Director: Beskid, Craig
Title: Organotypic Culture Model to Analyze DevelopmentalLimbMalformationsResulting from Toxicant/Teratogen Exposure
Investigators: Tuan, Rocky
Institution: University of Pittsburgh , Vanderbilt University
EPA Project Officer: Aja, Hayley
Project Period: December 1, 2014 through November 30, 2018 (Extended to November 30, 2019)
Project Period Covered by this Report: December 1, 2016 through November 30,2017
RFA: Organotypic Culture Models for Predictive Toxicology Center (2013) RFA Text |  Recipients Lists
Research Category: Chemical Safety for Sustainability


Our goal is to develop robust in vitro three-dimensional (3D) organotypic culture models (OCMs) based on human mesenchymal stem cells (MSCs) to first examine critical phenomena of embryonic limb development that are prime targets of limb teratogenesis, and then examine their susceptibility to perturbation by known and candidate teratogens and environmental toxicants.

Progress Summary:

The Limb Development team is building and testing three distinct organotypic culture models for chondrogenesis (OCM1), hypertrophy (OCM2), and segmentation (OCM3) for each one of their milestones.

Increasing morphological consistency in OCM1 and 2 (chondrogenesis and hypertrophy). The morphologic inconsistencies amongst the cultures were due to poor adherence of the micromass to the culture surface over the course of the culture period (17 days). We tested both additional surface coating modifications and gel overlays. We settled on the use of a gel overlay comprised of 5% w:v photocrosslinkable gelatin (PXL mGL) (Fig. 1A), which resulted in the cultures remaining planar and adherent to the surface and significantly increased consistency between cultures and experiments (Fig. 1B). When we compared chondrogenic differentiation between normal-uncovered cultures and those with gel-overlay, we found that the gel overlay increased chondrogenic gene expression in the underlying mesenchymal cells (Fig. 1C).

The increased utility of the gel overlay is evident in experiments involving epi-fluorescent microscopic analysis. This year, Project 2 has been developing several stage-specific differentiation GFP promoter-reporters. Work with the collagen type 2 GFP promoter reporter in uncovered and gel overlay cultures in Day 7, 10, and 14 of chondrogenesis showed that promoter activity increased in both types of cultures overtime but in the gel-overlay cultures, we can reliably count the numbers of GFP positive cells and their intensity (Fig. 2A). We compared the fold change in Col2 gene expression assayed by RT-PCR with the corrected total cell fluorescence (CTCF) fold change and found excellent agreement (Fig. 2B).

This assay will be the model for validation of reporter-promoter constructs we will be using in the coming year. In summary, the gel overlay resulted in morphologically consistent cultures that could be quantitatively analyzed by microscopy. The gel overlay is incorporated into our final OCM 1 and 2 (version 1) designs.

Developing a chondrogenic micromass sensitive to thalidomide. Another issue with our model from Year 2 was an apparent insensitivity of human adult MSC cultures to thalidomide, which caused us to reconsider our model in light of known and putative mechanisms of thalidomide teratogenesis. Specifically, it is thought that thalidomide may prevent angiogenesis during limb formation and that the mechanism of action is via disruption of normal cereblon-mediated protein turnover and concomitant changes in gene expression. Cartilage and vasculature have an antagonistic relationship. In light of this, we wanted to incorporate vasculogenesis into our culture by using the gel overlay technique to include endothelial cells, specifically GFP-labeled HUVECs.

Micromass cultures were prepared with both encapsulated and non-encapsulated HUVEC overlays and induced to undergo chondrogenesis for 14 days. In cultures with non-encapsulated HUVECs, individual GFP positive cells are evident throughout the culture. In cultures with gel-encapsulated HUVECs, we see evidence of tubulogenesis around the micromass cultures quantified by node number, branch number, and total branch length (Fig. 3A and D). When the vehicle for thalidomide, DMSO 1:300 v:v, was applied there was a reduction in tubulogenesis (Fig. 3B and D), while the addition of thalidomide resulted in no tubule formation by Day 14 (Fig. 3C and D). We then examined the effects of thalidomide on chondrogenesis in the underlying MSC cultures (Fig. 3E). Three trends were evident: (1) Thalidomide has no impact on MSC only cultures chondrogenic gene expression. (2) The addition of HUVECS to the culture increased chondrogenesis (as shown by Col2 and Agg expression). (3) Thalidomide causes a very significant decrease in chondrogenesis in micromass cultures containing gel overlayed HUVECs. Next, we determined the mechanism of these decreases and whether it is similar to known mechanisms of thalidomide action. Thalidomide promotes the interaction of the transcription factor Ikaros with Cereblon, which results in proteolytic degeneration of the Ikaros/Cereblon complex and changes in gene expression. Via Western blot, we determined that our cultures express Cereblon and Ikaros. We hypothesize that thalidomide will decrease Cereblon and Ikaros in both MSC and HUVECs, angiogenesis will be completely inhibited and MSC chondrogenesis will be decreased. Incorporation of endothelial cells into a micromass system should prove to be a novel and efficient way to study thalidomide teratogenesis.

Finalizing design of OCM 3 (joint segmentation). In Year 3 in collaboration with Project 5, we successfully generated rods of MSC-laden hydrogels using the following methodology: (1) loading MSCs at 40 x 106 cells/mL in PXL mGL hydrogels; (2) loading the suspension into a 1ml insulin syringe; (3) extruding the liquid cell-laden hydrogel into tubing with an appropriate inner diameter; (4) crosslinking the cell-laden hydrogel within the tubing; and (5) extruding the pre-cartilaginous MSC-laden rod into medium (Fig. 4). 24 hours after extrusion, the MSCs in these tubes are viable and undergo normal chondrogenic and hypertrophic differentiation. During differentiation, the morphology of the tubes changed from linear to wavy and the tubes could even curl up on themselves (data not shown). This would be a problem for mechanical stimulation with the actuator as well as microscopy. For this reason, the lab refocused efforts on working with the concentric ring model.

We decided to test the use of cell-laden (40 x 106 cells/mL) PXL hydrogels in the trough of the concentric ring model. After 7 and 14 days of chondrogenic differentiation, gel constructs in the troughs remained as straight linear bars or sheets. With this new procedure, the seeding of the constructs is vastly simplified. The hydrophobic walls and viscosity of the hydrogel keep the constructs in place. We also found that the gel did not break with repeated 90-degree flexion (Fig. 5). As a proof-of-concept we initiated an experiment with cell-laden hydrogels (40 x 106 cells/mL) and manual activation in place of a mechanical actuator. Each culture was subjected to 1hr of manual flexing, in the hood, in HEPES-containing chondrogenic medium at room temperature. did this for 7 days and consequently fixed the cultures. In the static cultures chondrogenesis appeared to progress normally as shown by Alcian Blue staining (Fig. 6B). In flexed cultures, we saw reductions in Alcian blue staining at the site of flexion as compared to the static ends of the culture (Figure 6A). This provides evidence of reduced chondrogenesis at the point of flexion. We also assayed GDF-5 production by IHC. We found increased GDF5 production in the areas of flexion as compared to the unflexed ends (Fig. 6C). Increases in GDF5 were tempo-spatially coincident with decreases in cartilage matrix production. Together this is a modest indication that interzone like changes are occurring in MSC-based flexed chondrogenic cultures. These results helped us to decide to move forward with the concentric ring model for the joint segmentation OCM.

Verifying chondrogenesis and hypertrophy of adult human MSC micromass cultures under fluid flow (and defining those parameters). Another major focus was developing a fluidic modality for our cultures. As a proof of concept, we tested chondro-genic and hypertrophic micro-masses under flow in the commercially available Kiyatec system which uses a polyether-sulfone (PES) porous membrane that separates the unit into chambers with two different medium flows (upper and lower). We had to change this membrane as medium would not bead on the PES membrane and micromass cultures would not form. We tested glass and custom-made solid and porous polycarbonate membranes, the latter maintained a bead that would permit micromass formation. Despite good flow conditions we faced two additional problems: (1) the large volume above and below the culture and (2) the distance between the culture surface and the top and bottom of the bioreactor was too large even for our long-working distance objectives making non-invasive ELISAs on conditioned media and epiflourescent microscopic observation difficult or impossible. As a result, we 3D-printed our own prototype for proof-of-concept experiments with constant medium flow using a polycarbonate membrane separating upper and lower chambers similar to the Kiyatec system but with a low profile for microscopy and small chambers in conservation of reagents and downstream ELISA assays. The upper and lower chambers are sealed with glass coverslips using silicone (Fig. 7). Although printable in any dimension, we are generating reactors with an OD equal to wells of a 12-well plate and ID chambers equal to the well of a 48 well plate. These dimensions are the current dimensions of Version 1 of OCM 1 and 2 already in use. These models have the capacity for microfluidics both as independent units or in limited series. Project 2 will move forward with this design to begin differentiation and toxicant tests.

Future Activities:

Project 2 will optimize flow conditions for chondrogenesis and hypertrophy in the microfluidic enabled versions of our OCMs as well as optimize non-invasive medium sampling and microscopic observation for the evaluation of culture development. With these parameters we can work together with Project 5 to engineer optimal reactors for this component of Project 2. An optimized microfluidic bioreactor will also allow connections to other OCMs in the VPROMPT group. The evaluation of known and candidate teratogens on chondrogenesis and hypertrophy was performed in previous years on 10 μL (200,000 cells) micromass cultures to provide dose response curves. This data will be helpful in future QAPP approved experimentation on 2ul cultures. In preparation for connection of the Limb OCMs to the liver device, we have performed chondrogenic differentiation assays using liver-conditioned medium (LCM) and noted a reduction in chondrogenesis in LCM medium (with TGFB3) in contrast to unconditioned medium (with TGFB3). This finding indicates we will have to restore “normal” nutrient levels in the LCM before putting it on our cultures. We are currently testing the addition of 10X DMEM to the LCM which would restore missing nutrients but only dilute potential liver by products of toxicants by 10%. Finally, we will show how Limb OCMs composed of rat cells differ from OCMS with human cells-especially poignant in the case of thalidomide.

Journal Articles:

No journal articles submitted with this report: View all 6 publications for this subproject

Supplemental Keywords:

Micromass culture, adult human mesenchymal stem cell chondrogenesis, hypertrophy, joint segmentation  

Relevant Websites:

Center for Cellular and Molecular Engineering (CCME) Exit   Vanderbilt-Pittsburgh Resource for Organotypic Models for Predictive Toxicology (VPROMPT) Exit

Progress and Final Reports:

Original Abstract
  • 2015 Progress Report
  • 2016 Progress Report
  • 2018
  • Final

  • Main Center Abstract and Reports:

    R835736    Mickey Leland National Urban Air Toxics Research Center (NUATRC)

    Subprojects under this Center: (EPA does not fund or establish subprojects; EPA awards and manages the overall grant for this center).
    R835736C001 Mammosphere Bioreactor For Life-Stage Specific Toxicology
    R835736C002 Organotypic Culture Model to Analyze DevelopmentalLimbMalformationsResulting from Toxicant/Teratogen Exposure
    R835736C003 Validating a fetal membrane on a chip model for characterizing reproductive toxicant exposure risks
    R835736C004 Organotypic Liver Model for Predictive Human Toxicology and Metabolism
    R835736C005 Systems Engineering & Analysis for Organotypic Culture Models