(692g) A 3D Printed Microfluidic Bioreactor to Engineer Biphasic Construct

INTRODUCTION: Tissue engineered
constructs coupled with high-throughput bioreactors can be used as innovative in
vitro pre-clinical models of tissue functions and disease pathogenesis for
drug screening and toxicity assessment [1]. Such in vitro models are
ideally suited to study tissue-tissue interactions in health and disease in
composite tissue units such as those found in the musculoskeletal system. Among
these, the osteochondral junction is of particular clinical importance as it
has been identified as a locus of degenerative joint diseases such as
osteoarthritis [2]. To generate in vitro model of adjacent tissues, we
have developed an innovative 3D printed bioreactor for biphasic constructs,
which can host cells in a small volume and allow optical access for direct
monitoring of cellular responses during culture and differentiation in 3
dimensions [3]. The aim of this work is to obtain an osteochondral junction
model from adult human mesenchymal stem cells (hMSCs), after delivery of
spatially controlled local stimuli.

METHODS: A combined
modelling/experimental approach was employed to design a viable bioreactor
fluidics. Models of the flow path were created using SolidWorks and tested
using ANSYS Fluid Flow (CFX), with an inlet volume flow rate of 1 ml/day, and
outlet open to the environment. The central chamber that will host the
constructs was considered as filled with methacrylated gelatin (gelMA) with a
permeability of 1*10^-16 m² and a porosity of 0.8 [4]. Velocities
were measured in CFX Post and the design was altered by changing the diameter
and height of the channel to achieve flow rates in the constructs not damaging
for cells. Bioreactors were designed using SolidWorks and printed with a Viper
SLA System at high definition with a layer resolution of 50μm using the Somos® WaterClear
Ultra 10122 resin (DSM N.V., The Netherlands). O-rings (McMaster-Carr),
glass coverslips (Deckglaser), and adhesive films (Professional Label) were
used to complete the system. Bone marrow derived hMSCs were seeded in the
bioreactors chambers in gelMA (10⁶ cell/ml) and photocrosslinked in
situ. The hydraulic circuit hosted the incubator, consists of a pump, a
10ml syringe filled with expansion medium, a valve to remove air bubbles, tubes
and a collection bag. A dynamic Live Assays was performed to assess cellular
viability over time, adding calcein-AM to the medium in the syringe without
compromising the overall setup sterility. An Alamar Blue test was also
performed within the bioreactor to confirm cell viability. Finally, osteogenic
(DMEM, 10% FBS, 5% PSF, 10 nM dexamethasone, 0.1 mM L-ascorbic acid
2-phosphate, and 10 mM beta-glycerophosphate) and chondrogenic (DMEM, 5% PSF,
10 ng/ml recombinant human TGF-β3 (PeproTech), 1% ITS+, 50 μM
L-ascorbic acid 2-phosphate, 55 μM sodium pyruvate, and 23 μM
L-proline) media were fluxed in the bioreactor for 21 days to induce
differentiation which was assessed by histology (Alcian Blue for cartilage,
Alizarin Red for bone). Finally, 5x10³ hMSCs were plated in a 96
well-plate and incubated with a lentiviral Col2 reporter added at an MOI = 40
(multiplicity of infection = transducing units / #cell) in fresh growth medium
without PSF and containing 6 ug/mL of SureEntry transduction reagent. 20h
later, medium containing the lentiviral particles was removed and substituted
with fresh complete growth medium. Cells were harvested after 72h and added to
the non-transfected hMSCs to act as sentinel cells to monitor differentiation.

RESULTS SECTION: Two
single-inlet bioreactor designs were first developed with higher and lower
perfusion velocities in the chamber hosting the constructs (Fig. 1). The
dynamic live assays in the bioreactors produced a successful staining of the
cells that can be monitored optically without the need to break sterility in
the bioreactor system (Fig. 2). The dynamic Alamar blue assay was also
effective in confirming the possibility of non-destructive monitoring within
the system. The bioreactor with higher perfusion velocities appeared to be the
most effective with the gelMA constructs and that design was adopted for the
two inlets, biphasic model. Experimental fluidic tests with the two
inlets/outlets biphasic bioreactor confirmed modelling predictions, showing minimal
to no mixing between the flowing media (Fig. 3). Differentiation of hMSCs
towards a chondrogenic and osteogenic phenotype in the two sides of the
constructs was confirmed by histological staining and optically by fluorescence
of the Col2 transfected sentinel cells in the half of the construct exposed to
chondrogenic medium (data not shown).

DISCUSSION: Development of
bioreactors is essential for the generation of biphasic constructs that require
specific media for each tissue component and cannot be cultured in regular
multiwell plates. The successful dynamic staining of the cells in the
constructs within the bioreactors show the potential of the system to perform a
continuous, real-time monitoring of constructs maintaining sterility. This is
particularly relevant when using sentinel cells with reporter genes that can signal
differentiation as well as response to stress and cross-talk between tissues,
depending on the specific reporters adopted.

SIGNIFICANCE: Disease modeling
and drug screening must account for interactions between adjacent tissues. Once
scaled for high throughput, our biphasic bioreactor will be an essential step
in developing new strategies to prevent and treat joint diseases.

REFERENCES: [1] Lozito et al.
(2013). Stem Cell Research & Therapy, 4(Suppl 1):S6. [2] Alexander et al.
(2014). Exp Biol Med, 239(9):1080-95. [3] Nichols et al. (2018) Biomed Microdev
20: 18. [4] Iannetti et al. (2016). PLoS One, 11(9):e0162774.

ACKNOWLEDGEMENTS: Ri.MED
Foundation, CASIS (grant no. GA-2016-236), Commonwealth of Pennsylvania, US EPA
– Science to Achieve Results (STAR) program #835-73601, NIH (1UG3 TR002136-01)