Chosen Design

Our design approach involves the use of a clear, small tank that holds media and has small openings for the mechanical loading interface. The tank will allow for temporary maintenance of internal conditions and sterility as it is removed from the incubator for imaging. Media will be changed regularly and fed into a glass reservoir which feeds into Teflon tubing which is hooked up to the tank. Imaging will be done using an epifluorescence microscope and vital dyes. Mechanical loading will be done using a T-shaped interface that compresses the growth plate by pressing against metaphyseal bone. Compression will be done using a compressive spring although the interface can be attached to an actuator in the future. Thickness of the growth plate will be constrained by two clear plates on either side of the growth plate that sit at a fixed distance apart from each other. A nutrient gradient will be induced by cutting small holes into the bone that sits on either side of the growth plate longitudinally, allowing for media transfer that mimics metaphyseal and epiphyseal arterial invasion.

This design is a more developed iteration of the horizontal tank design, this time with a vertical, sandwich-shaped tank and impermeable boundaries on the lateral edges of the physis that ensure a parabolic oxygen gradient within the slice that is mainly governed by Fick’s 2^nd Law. Media leakage is less of an issue since the physis samples will be loaded from the top end of the bioreactor although incubation and imaging will have to be done in a sterile area since the top of the tank will be open. Imaging of vertical slices was accomplished via the use of a prism that redirects line of sight from a microscope in a horizontal fashion.

Alternative Design 1: Mechanical Loading via the Use of Grips

This design theoretically allows for compressive, tensile, torque and shear loading and having a constant media flow means that the media doesn’t have to be changed. However, the serrated grips are more likely to shear the bone that they grip than load the growth plate cartilage as intended. Sterility is also more of an issue due to media constantly flowing in and out as well as the growth plate being exposed to the open environment during imaging. Physiological gradients aren’t accounted for and in order for loading to be successfully implemented, the grips need to be attached to an actuator that is capable of three degrees of freedom, which is extremely difficult to implement.

One alternative design we considered was the use of serrated grips for mechanical loading. The grips would clamp onto either side of the growth plate (epiphyseal/metaphyseal sides) and load the growth plate with compression, tension, torque and shear forces. Compression would be done by pushing the two clamps together, tension would be done by pulling the two clamps apart, torque would be done by having the two clamps rotate in opposite directions or having one twist while the other one is held in place, and shear would be done by having the two plates move laterally in opposite directions. Flow into tank would be mediated by a pump which would circulate media in and out of the tank at a constant flow based on the rate at which the growth plate consumes media. Thickness of growth plate would be constrained by two clear plates held a fixed distance apart from one another. Growth plate would be submerged in media and lifted out with one of the constraints popping off for imaging purposes if necessary. Imaging would more than likely be done using a standard light microscope.

Alternative Design 2: Two-Plate Loading System

This design also theoretically allows for compressive, tensile, torque and shear loading and similar to Alternative Design 1, having a constant media flow means that the media doesn’t have to be changed. However, having a permeable adhesive would restrain longitudinal and lateral growth, there would be no way to constrain thickness, and physiological gradient wouldn’t be accounted for. In addition, having a large number of moving parts as well as a constant flow would make integration and implementation incredibly difficult.

Another design alternative that we considered was the use of a two-plate loading system. To accurately place a static, or cyclic load of known amount onto calf ulna physeal tissue, the bioreactor would house 2 plates with permeable adhesives on their surfaces that are able to move dynamically when required. The plates would twist along their centerline to induce torsional force on the tissue, pull apart to stretch the tissue for tensile testing, move along the line of contact in opposite directions to see how the tissue responds to shear forces, and to sandwich the growth plate in between them for compression testing. The design would contain a centralized source of liquid nutrients and growth factors with a constant flow underneath a plate. The flow will slowly release liquid onto a permeable membrane with a wiper system attached to it. The wiper would then push the pooling liquid out to the far edges of the membrane radially. The liquid would permeate through the membrane over time and provide growth factors and nutrients to the surface, where the growth place would grow. The flow system to deliver the nutrients would be comprised of one tube that would cycle through the system to help recycle unused substrates, and 2 delivery tubes (one per plate) connected to the main tube that would release the fluid onto the wiper mechanism discussed above. A pump would keep the fluid cycling through the system. To view and image the growth plate, a high resolution camera would be attached to a mechanism that will allow it to swivel and move. The mechanism would contain a swivel, a hemisphere, and a grid. The swivel allows the camera to move around from a singular vantage point so that more of the tissue can be captured from one location. The hemisphere allows for rotation of the camera around a fixed point of focus, which allows for different angles of imaging the tissue. The grid allows for the translational movement of the camera so it can move around and see all parts of the growth plate. A 10X microscope and the camera’s zoom would be used for magnification.

Alternative Design 3: Horizontal tank with spring-loaded interface

The tank and interface components of the bioreactor should be easy to replicate, which will allow for utilization of multiple devices at once, which will allow for an n=3 sample size. Small holes in the bone will closely mimic physiological blood flow, translating into more significant experimental results. An epifluorescence microscope and incubator are both readily available in the lab, which will allow for cell tracking with vital dyes and easy CO₂ transport/pH control within the bioreactor. A compressive spring is also the easiest verifiable means of loading the growth plate, as the load is dependent on the displacement of the spring. With the current design, the growth plate is kept within a sterile environment and the tank/constraints are clear, so imaging is not much of a concern. It is also easy to remove from the bioreactor and image. On the other hand, it is currently only capable of compressive loading and sterility of the loading device might pose a challenge since the outer end of the interface will more than likely depress into the bioreactor as it compresses the growth plate. Preventing media leakage from the loading opening could also pose an issue, although this could be resolved by orienting the tank itself vertically so that all openings/valves are located on the top.