Non-layered bioprinting via dynamic optical projection and uses thereof
阅读说明：本技术 经由动态光学投影的无分层生物打印及其用途 (Non-layered bioprinting via dynamic optical projection and uses thereof ) 是由 彼得·钟 瞿鑫 张阿平 陈绍琛 于 2015-05-20 设计创作，主要内容包括：本申请涉及经由动态光学投影的无分层生物打印及其用途。一种用于3D微制造的系统和方法,该用于3D微制造的系统和方法将能够引发光聚合的光朝向一个空间光调制器投影,该空间光调制器调制响应于对应于结构的层的数字掩模的光。投影光学器件将该调制光聚焦在支持于一个载物台上的一种可光聚合材料内的一个光学平面上。一种计算机控制器使得该空间光调制器投影出对应于这些数字掩模的一种图像序列,同时协调该载物台的移动以移动该光学平面在该可光聚合材料内的位置以依序投影出该序列的每个图像从而通过逐渐地光聚合该可光聚合材料产生该结构。(The present application relates to non-layered bio-printing via dynamic optical projection and uses thereof. A system and method for 3D micro-fabrication projects light capable of initiating photo-polymerization toward a spatial light modulator that modulates light responsive to a digital mask corresponding to a layer of a structure. Projection optics focus the modulated light onto an optical plane within a photopolymerizable material supported on a stage. A computer controller causes the spatial light modulator to project a sequence of images corresponding to the digital masks while coordinating movement of the stage to move the position of the optical plane within the photopolymerizable material to project each image of the sequence in sequence to produce the structure by progressively photopolymerizing the photopolymerizable material.)
1. A system for 3D microfabrication of a structure, comprising:
a light source configured to project light within an optical path, the light source emitting light at a wavelength configured to initiate photopolymerization;
a spatial light modulator disposed within the optical path and configured to modulate light from the light source in response to a set of digital masks corresponding to layers of the structure;
projection optics configured to focus the modulated light to an optical plane;
a stage configured to support a substrate in contact with photopolymerizable material in the optical plane, wherein the stage is configured to move along at least one axis; and
a computer processor operable to:
controlling the spatial light modulator to project a sequence of images corresponding to the set of digital masks; and
coordinating movement of the stage to move the position of the optical plane within the photopolymerizable material to project each image of the sequence in sequence to produce the structure by progressively photopolymerizing the photopolymerizable material along the at least one axis.
2. The system of claim 1, wherein the spatial light modulator comprises a digital micromirror device.
3. The system of claim 1, wherein the photopolymerizable material is a prepolymer solution contained within a container, and the system further comprises a transparent window configured to be in contact with the prepolymer solution, wherein an interface between the window and the prepolymer solution is substantially coincident with the optical plane.
4. The system of claim 3, wherein the stage is moved toward the optical plane such that successive layers of the structure protrude through previously polymerized layers.
5. The system of claim 3, wherein the window is sealed to an end of a tube to be submerged in the prepolymer solution and the stage is moved away from the optical plane such that successive layers polymerize on top of previously polymerized layers.
6. The system of claim 3, wherein the window is treated to inhibit adhesion of the photopolymerizable material.
7. The system of claim 3, wherein the window is formed of a material selected from the group consisting of: silicon dioxide, sapphire, Polydimethylsiloxane (PDMS), transparent ceramics, and transparent plastics.
8. The system of claim 3, wherein the container comprises a well within a multi-well plate.
9. The system of claim 8, wherein the multi-well plate comprises a plurality of containers, wherein at least a portion of the plurality of containers contain different prepolymer solutions.
10. The system of claim 1, wherein the substrate comprises an electrode or a multi-electrode array.
Technical Field
The present invention relates to a three-dimensional structure for microfabricating photopolymers and a method for fabricating bioscaffolds with smooth (non-delaminated) contours.
Background
Traditional cell culture is usually performed in culture dishes, and the more commonly used cell biotechnology relies on the use of cultured cells on a 2D platform. It is well known that the mechanical properties of the cell substrate influence the differentiation, growth and motility of cells. Therefore, the mechanical properties of the material are important considerations when designing the cell scaffold. In order to simulate the actual cellular environment, it is necessary to create a 3D functional tissue model using a 3D bio-fabrication method rather than a 2D patterning method. Furthermore, the manufacturing process and/or materials must be cell-friendly. Current manufacturing systems for biological stents utilize methods such as bio-inkjet printing and raster laser projection, which are limited in scalability and speed, and do not create complex structures and scalable tissue constructs. The casting method can achieve basic geometries, but cannot replicate complex 3D geometries with complex biomimetic features and aspect ratios. Laser micro-photocuring rapid prototyping (laser- μ SL) technologies have become popular because of their ability to create 3D scaffolds [ horse skin force (Mapili) et al, 2005; gansel et al, 2009. However, stent fabrication using laser- μ SL is typically slow due to the point-by-point laser scanning. Many existing layer-by-layer photocuring rapid prototyping methods are not characterized by dynamically changing masks and concurrent stage movement, and therefore produce undesirable layer artifacts (artifacts) that can alter cellular responses and do not reflect true natural physiology.
One method that has been commercialized by Carbon3D corporation (Carbon3D, Inc.) (Redwood City, CA) for immersion printing of polymer nanostructures involves projecting an image through a transparent floor of a reservoir of liquid resin while gradually lifting the immersed substrate off the floor as the resin cures on the bottom of the substrate. The oxygen permeable transparent floor of the cell allows "dead zones" of dissolved oxygen to inhibit polymerization at the cell floor. Techniques used in carbon3D systems are described in a number of U.S. patents (both issued and pending) to Joseph de simon et al, including number 8,263,129, which is incorporated herein by reference. The reservoir and substrate of the carbon3D system remain stationary with only the submerged substrate moving in the z-axis to lift the printed object off the floor as the printing progresses, which limits the variety of shapes that can be created. Furthermore, the dependence of the technology on immersion limits the types and combinations of materials that can be used.
Conventional stereolithographic methods are not well suited for high throughput fabrication of complex cell-supporting 3D microstructures, particularly within substrates such as multi-well plates commonly used in life sciences. These drawbacks severely limited the widespread adoption of 3D printed cell culture methods, as researchers often relied on products configured as commonly used laboratory instrument interfaces that are integrated with established experimental workflows. Multi-well cell culture plates (except for microscope slides, petri dishes, cell culture flasks, etc.) are often used as de facto standards on which to design specialized culture environments. Within multi-well plates, many 3D hydrogel cell culture platforms are characterized by biocompatible and/or biologically derived materials that have been polymerized or otherwise fabricated, but typically as unpatterned bulk structures. Similarly, high throughput cell culture systems with integrated microfluidic or multi-electrode arrays have also been developed on multi-well or standard microscope slide formats.
Commercial 3D bioprinters capable of printing directly into multi-well plates are commercially available from a number of companies, including the BioAssemblyBot from Advanced Solutions, Inc. (Louisville, Kyoki (KY), USA (US))TMBioprinters, from Cyfuse biomedical corporation (Tokyo, Japan (JP))
A highly defined 3D cell culture microenvironment can be used in a wide range of physiological contexts, including neuronal culture in vitro. The goal of studying isolated neural cultures is to test and explore simple in vitro systems that can represent physiologically relevant models. Isolated neural cultures are the cornerstone of neuroscience research, but their utility reflecting natural physiology is limited due to their inherent uncertain connectivity. Although a large amount of neurophysiological data detailing the function of single-layer neurons is widely available, the conditions used for these cultures are significantly different from those present in native tissues. Thus, it can be reasonably assumed that the behavior of 2D cultures is not a good representation of the complex system of neurophysiology in vivo.
Recent work demonstrated that neurons grown in 3D scaffolds with incorporated glia provided significant morphological and electrophysiological differences compared to neural networks grown in 2D cultures. This is due to the fact that: the 3D neural scaffold more closely resembles the complex neural environment existing in the body.
In vitro replication of neural circuits can help elucidate a substantial part of the neural environment and facilitate testing of connectivity models that have an impact on higher levels of processes. Furthermore, when patient-specific induced pluripotent stem cells (ipscs) are employed for in vitro disease models, controlling the functional arrangement of the neural population may be critical to the reconstruction of normal neural circuits relative to pathological neural circuits. Conventional devices for neuron patterning are often limited to 2D scenarios and use substrates that do not reflect native mechanochemical properties. On the other hand, methods of creating soft 3D stents can be expensive, time consuming, or limited to simple geometric features. In view of the limitations of current platforms, there is a need to establish a high throughput apparatus for deterministically controlling and systematically investigating network dynamics of the underlying neural circuits in a soft 3D environment that is more physiologically representative.
Engineering simplified neural circuits with adaptability for systematically increasing complexity provides an attractive model for studying neural wiring (wiring) and functional connectivity. High density recording and stimulation of current via surface microelectrode arrays in vitro provides rich information about the state of the network, and the utility of in vitro models extends also to drug screening in view of its high-throughput nature. Recent innovations in culturing neurons and directing nerve growth provide some control over network connectivity, cell density, and neural phenotype in order to achieve some ordering and simplicity of the network. Notably, the advent of 3D culture in which neurons are grown within hydrogel scaffolds (having a thickness of at least 10 cell diameters) has been used to address the limitations of 2D models by: 1) maintain more relevant cell-cell/cell-matrix interactions, 2) protect neurons from changes in medium pH, 3) use mechanically softer interfaces that are more reflective of natural physiology than the hard substrate in 2D culture, and 4) provide high surface area for growth and migration. Therefore, the development of 3D constructs for isolated neural networks and the development of physiologically relevant "brain-on-a-chip" systems is essential. New advances in this area demonstrate the possibility of maintaining a neuron population in a 3D scaffold while directing nerve growth in a controlled geometric pattern. However, challenges still remain in: 1) achieving high spatial resolution and high density in terms of recording and stimulation, 2) simplifying expensive and laborious manufacturing techniques, and 3) creating heterogeneous co-cultures of neurons and glia, which have proven to be critical for proper functioning of neurons.
Brief summary of the invention
In one exemplary embodiment, a method and system for rapidly producing polymer scaffolds with highly specified 3D geometries is provided. The bio-fabrication method of the present invention is referred to herein as "non-layered 3D printing via dynamic optical projection" or "L3 PDOP. The L3PDOP 3D printing system and method of the present invention may receive an input comprising a 3D computer generated model and generate structures of photopolymerizable materials with high resolution from the model. The 3D printing system of the present invention uses a dynamically controlled DMD (digital micromirror device) to reflect a high resolution pattern of UV light on a photopolymerizable substrate. Integration with custom computer software allows for simultaneous and continuous control of the projected image and the linear stage that controls the position of the substrate relative to the focal plane of the projected light. Due to the continuous movement of the stage, the 3D structure is manufactured with a smooth (i.e. no delamination) profile in the Z (vertical) direction. By controlling the composition of the prepolymer solution, e.g. by containing naturally derived and/or synthetic biomaterials, bioactive molecules, heterogeneous cell populations, 3D structures simulating the natural tissue environment can be generated within seconds. In addition, the rapid control of the micromirror array provides micrometer and submicrometer scale variation of local material properties (such as porosity and material stiffness) via spatio-temporal patterning of the instantaneous UV exposure throughout the polymer.
According to embodiments of the present invention, a system and a method for 3D micro-fabrication are provided by: light capable of initiating photopolymerization is projected toward a spatial light modulator that modulates light responsive to a digital mask corresponding to a layer of the structure. Projection optics focus the modulated light onto an optical plane within a photopolymerizable material supported on a stage. A computer controller causes the spatial light modulator to project a sequence of images corresponding to the digital masks while coordinating movement of the stage to move the position of the optical plane within the photopolymerizable material to project each image of the sequence in sequence to produce the structure by progressively photopolymerizing the photopolymerizable material.
The method and system of the present invention provides a novel photocuring rapid prototyping process via dynamic continuous control of a micromirror array (DMD) and a linear stage. Specifically, the device achieves improvements over the state of the art today due to its speed, scalability and non-layered Z resolution. By providing a non-layered resolution, the fabricated structure does not exhibit the planar artifacts created using conventional layer-by-layer fabrication methods involving discrete movement of the linear stage to a new height position. Furthermore, the L3PDOP 3D printing platform of the present invention provides for fast and scalable manufacturing of highly specified biomimetic structures. This method provides the ability in terms of speed and scalability not well achieved using prior art techniques employing raster-based printing methods or soft lithography techniques. Furthermore, because the L3PDOP system can accept almost any set of high-definition images, this platform provides a method for generating models that are not only specific in terms of cell type, but also have overall tissue morphology. The adaptability of the platform allows modular addition and subsequent decoupling of different components of complex 3D constructs, providing a means for determining individual contributions of material types, co-culture populations, spatial cell arrangements, and biomimetic geometries. Modes of operation for the device include 3D micron-scale bioprinting of biological substrates via dynamic photocuring rapid prototyping, and optogenetic control of neurons with high spatiotemporal resolution, with concomitant integration of electrophysiological and fluorescence imaging. Photocured rapid prototyping 3D bioprinting allows photopolymerization of a variety of biocompatible materials (e.g., PEGDA, MeHA, GelMA) and facilitates cell encapsulation within these structures. Furthermore, these materials-and their incorporated cell types-can be used heterogeneously in the same scaffold to provide 3D patterned co-culture substrates, facilitating the study of interactions between multiple cell phenotypes, such as neuronal/glial interactions.
Because individual pixels of the micromirror array can be turned on and off rapidly throughout the fabrication process, highly localized (e.g., sub-micron) changes in the polymer structure in terms of porosity or material stiffness can be achieved by spatio-temporal control of UV exposure at the pixel level. This platform addresses the limitations in the art by: allowing simultaneous control of stent composition, microstructure, and larger scale 3D geometries as well as providing improvements in speed and scalability.
In one aspect of the invention, a prepolymer solution containing a radical absorber for inhibiting photo-activated radical polymerization/crosslinking is used to limit the depth of cure. A surface treatment may be applied to the window, which may be composed of sapphire or glass (or any other sufficiently rigid optically transparent material), to prevent the structure from polymerizing/sticking to the submerged window. This is in contrast to the carbon3D approach, where printing using the platform of the present invention occurs directly on the floor of the bath, as opposed to directly on the immersion probe. This allows 3D structures to be printed quickly within each well of a multi-well plate as the plate is translated in the X-Y axis to perform well-to-well movement. By not relying on immersion of a single large tank, each well may contain a different polymer composition and may ultimately maintain a different 3D structure.
In another aspect of the invention, the method and the apparatus are operable to record network activity over a plurality of domains. For example: 1) multiple Electrode Arrays (MEAs) -printing 3D neuronal network structures (or any electrically active cell type) on top of it-allowing orthogonal recording of electrical activity coincident with optogenetic stimulation; and 2) fluorescence imaging of voltage sensitive dyes and proteins (which provides improved single cell resolution in network activities) can be performed simultaneously by employing the same or similar light sources used for bioprinting.
In one application of the 3D printing method and system of the present invention, a brain chip is created by integrating 3D heterogeneous neural cultures, optogenetically-activated high resolution stimulation, and concurrent electrical and image-based recording of neural activity. The fabrication methods utilized in the present platform provide control over the morphological complexity and cellular and material composition of the neural environment, allowing the possibility of increased complexity of system control.
In one aspect of the invention, a system for 3D microfabrication of a structure includes a light source configured to project light within an optical path, the light source emitting light at a wavelength configured to initiate photopolymerization; a spatial light modulator disposed within the optical path and configured to modulate light from the light source in response to a set of digital masks corresponding to layers of the structure; projection optics configured to focus the modulated light to an optical plane; a stage configured to support a substrate in contact with photopolymerizable material in the optical plane, wherein the stage is configured to move along at least one axis; and a computer processor operable to: controlling the spatial light modulator to project a sequence of images corresponding to the set of digital masks; and coordinating movement of the stage to move the position of the optical plane within the photopolymerizable material to project each image of the sequence in sequence to produce the structure by progressively photopolymerizing the photopolymerizable material along the at least one axis. In some embodiments, the spatial light modulator may be a digital micromirror device. The photopolymerizable material may be a prepolymer solution contained within a container, and the system further comprises a transparent window configured to be in contact with the prepolymer solution, wherein an interface between the window and the prepolymer solution is substantially coincident with the optical plane. In one embodiment, the stage is moved towards the optical plane such that successive layers of the structure protrude through the previously polymerized layer. In another embodiment, the window is sealed to an end of a tube to be immersed in the prepolymer solution and the stage is moved away from the optical plane so that successive layers polymerize onto previously polymerized layers. The window may be treated to inhibit adhesion of the photopolymerizable material and may be formed of a material selected from the group consisting of: silicon dioxide, sapphire, Polydimethylsiloxane (PDMS), transparent ceramics, and transparent plastics. In some embodiments, the container comprises a well within a multi-well plate. The multiwell plate may comprise a plurality of containers, wherein at least a portion of the containers contain different prepolymer solutions.
In other embodiments, the substrate comprises an electrode or a multi-electrode array. In these embodiments, the photopolymerizable material may be a conductive polymer and at least a portion of the sequence of images corresponds to interconnect structures aligned with one or more electrodes.
The photopolymerizable material may be a photo-crosslinkable hydrogel selected from the group consisting of: gelatin methacrylate [ GelMA ], methacrylated hyaluronic acid [ MeHA ] and polyethylene glycol diacrylate [ PEGDA ].
The system can include a second light source disposed within the optical path, the second light source emitting light at a wavelength configured to stimulate an optically active biological material. The system may also include an image acquisition device disposed within the optical path.
In some embodiments, at least a portion of the digital masks may be configured to cause the spatial light modulator to generate sub-patterns within selected layers of the structure.
In some embodiments, the stage is movable in X, Y and the Z axis.
In some embodiments, the light source emits light within an ultraviolet region and is selected from the group consisting of: a laser, a light emitting diode or an array of light emitting diodes, and arc lamps.
In another aspect of the invention, a method for 3D microfabrication of a structure includes controlling a spatial light modulator to project a sequence of patterns of light from at least one light source having an optical path and to emit light at a wavelength configured to induce photopolymerization on a photopolymerizable substrate, wherein the substrate is mounted on a stage configured to move along at least one axis, and wherein the sequence of patterns corresponds to layers of the structure; and simultaneously and continuously controlling the sequence of patterns and the stage movement to produce the structure by gradually photopolymerizing the photopolymerizable substrate along the at least one axis. The sequence of patterns is preferably generated by a set of digital masks transmitted from a computer controller to the spatial light modulator.
In some embodiments, the photopolymerizable substrate comprises a prepolymer solution contained within a container, and the method further comprises positioning a transparent window in contact with the prepolymer solution to define a polymerization plane. In one embodiment, the stage is moved towards the photopolymerization plane so that successive layers of the structure protrude through the previously polymerized layer. In another embodiment, the window is sealed to an end of a tube to be submerged in the prepolymer solution and the stage is moved away from the polymerization plane so that successive layers polymerize onto previously polymerized layers. Different stage movement directions may be combined, with some configurations constructed from a combination of top-down and bottom-up stage movement.
In some embodiments, the container may be a well within a multi-well plate, which may comprise a plurality of containers, wherein at least a portion of the plurality of containers contain different prepolymer solutions. In other embodiments, the substrate may be an electrode or a multi-electrode array. The photopolymerizable substrate may be a conductive polymer and at least a portion of the sequence of patterns corresponds to the interconnect structure aligned with one or more of the electrodes.
The photopolymerizable substrate may be a photocrosslinkable hydrogel selected from the group consisting of: gelatin methacrylate [ GelMA ], methacrylated hyaluronic acid [ MeHA ] and polyethylene glycol diacrylate [ PEGDA ].
In some embodiments, the method further comprises disposing a second light source to emit light coincident with the optical path, wherein the second light source emits light at a wavelength configured to stimulate a photoactive biological material.
In some embodiments, the method further comprises disposing an image acquisition device within the optical path; and collecting an image of the substrate or the structure.
In some embodiments, at least a portion of the digital masks are configured to cause the spatial light modulator to generate sub-patterns within selected layers of the structure.
In some embodiments, the stage is configured for movement in X, Y and the Z axis.
In some embodiments, the light source emits light within an ultraviolet region and is selected from the group consisting of: a laser, a light emitting diode or an array of light emitting diodes, and arc lamps.
Drawings
FIG. 1A is a schematic diagram of one embodiment of an L3PDOP 3D printing system; fig. 1B is a set of SEM images of a complex geometry created using an L3PDOP printer.
Fig. 2A-2D illustrate an exemplary production flow for producing a cell-filled 3D scaffold using an L3PDOP printing method.
Fig. 3 is a series of bright field and confocal fluorescence micrographs illustrating scaffolds prepared using an L3PDOP printer and cell interactions with the scaffolds.
Fig. 4A-4C are a series of confocal fluorescence micrographs illustrating the cell interaction in the fabricated scaffold.
FIG. 5 is a set of fluorescence micrographs (A-D) showing different 3D scaffolds.
Fig. 6A is a scanning electron micrograph of a log stake holder; figure 6B is an optical micrograph of a vessel microstructure using an embodiment of the printing platform of the present invention.
Fig. 7 is a diagram of a brain chip platform that will integrate 3D heterogeneous neural cultures, optogenetically-activated high resolution stimulation, and concurrent electrical and image-based recordings of neural activity.
Fig. 8A-8C illustrate an alternative embodiment of the 3D printing platform of the present invention, wherein fig. 8A graphically illustrates printing of polymer scaffolds using different prepolymers in a multi-well plate, fig. 8B provides a detailed view of the optical assembly of fig. 8A, and fig. 8C graphically illustrates an exemplary production flow for non-layered printing of a 3D scaffold.
Fig. 9A and 9B illustrate an example of using the alternative printing platform of fig. 8A-8C, where fig. 9A is a screen shot of CAD software for designing and rendering two different 3D models, and fig. 9B provides a photomicrograph of the resulting printed hydrogel.
FIG. 10A shows a theoretical rendering at different magnifications of the intrinsic structure of a simple cube produced by different light exposure sub-patterns; FIG. 10B shows an example of how an exposure pattern can produce different intrinsic morphologies.
Fig. 11 provides a photomicrograph of a patterning algorithm applied to different checkerboards of complex shapes.
Fig. 12 provides a photomicrograph of a patterning algorithm applied to different checkerboards of complex shapes.
Fig. 13A-13C show CAD designs of rat brain and possible sub-patterning arrangements under different "layers" or "slices", respectively.
Fig. 14 shows a possible sub-patterning arrangement for a human lower limb model.
Figure 15 is a photomicrograph showing a GelMA holder with culture wells, interconnect pathways/channels, and reverse grooves for electrode sites aligned with one MEA.
Figure 16 is a set of photomicrographs showing NPCs and elongated neurites that have accumulated in these wells after stent printing with different GelMA concentrations relative to the control.
FIG. 17 is a set of photomicrographs of a neural cell culture in a scaffold of the invention labeled to assess cell morphology, function, and synaptic connectivity after one month relative to a glass control.
Figure 18 is a set of photomicrographs of neurons cultured in scaffolds of the invention labeled to assess functional connectivity relative to a glass control.
FIG. 19 is a set of photomicrographs of neurons cultured in scaffolds of the invention labeled to assess synaptoprotein expression relative to a glass control.
Fig. 20A is a photomicrograph of a Multiple Electrode Array (MEA) using nerve cluster imprinting according to an embodiment of the present invention, with the inset being a magnified image of the center of the well cluster. Fig. 20B is a graph of the peak activity measured at
Fig. 21A and 21B are diagrammatic illustrations, respectively, of one exemplary MEA using standard flat plate electrodes versus conductive polymer electrodes using different sized 3D printing.
22A-22B illustrate an exemplary biological neural network created using the platform of the present invention and configured to implement an AND logic gate; FIG. 22C illustrates an exemplary training sequence and test sequence as a function of time; FIG. 22D shows a truth table for an AND logic gate; FIG. 22E illustrates an exemplary training program.
23A-23C illustrate an exemplary biological neural network created using the platform of the present invention, configured to implement an XOR logic gate, where FIGS. 23A and 23B illustrate an exemplary training sequence over time; FIG. 23C shows a view of an MEA; FIG. 23D shows a truth table for an XOR logic gate; and figure 23E is a set of theoretical ROC curves generated during this neural mesh test.
Figures 24A and 24B illustrate an exemplary neural network or sensor structure and corresponding MEA sheets using neuron-filled hydrogel imprints for implementing the neural network or sensor structure according to one embodiment of the present invention.
Figures 25A and 25B are diagrammatic views of a neural network computing system constructed in accordance with an embodiment of the present invention.
Fig. 26A and 26B graphically illustrate an example of customized MEA patterning capability enabled by the 3D printing platform of the present invention for use in high throughput screening of parallel samples.
Detailed description of exemplary embodiments
The basic elements of an L3PDOP
A suitable UV
As an alternative to the DMD chip, a galvanometer optical scanner or a polygon scanning mirror may be used. Both techniques are commercially available and their use in fast scanning confocal microscopy is known. The selection of an appropriate scanning mechanism for use in conjunction with the systems and methods of the present invention is within the level of skill in the art.
An example of a structure produced using the printing system of the present invention is shown in fig. 1B, which is a series of Scanning Electron Microscope (SEM) images of PEG microwells with complex geometries, including (B) hierarchical, (c) helical, (d) embryonic-like, and (e) patterned. FIG. 1B (a) is a combination of an array of structures shown in (b) - (e). Images (f) - (i) of fig. 1B are the inverses of the micropores of (B) - (e), demonstrating the versatility of the L3PDOP printing method.
This
Fig. 2A-2C illustrate an exemplary process sequence for manufacturing a 3D stent using the
In one embodiment of the invention, one or more image acquisition devices may be included within the optical path such that the incident or projected image is translated through the projection optics such that the focal plane of the image coincides with the focal plane of the image acquisition device. The image acquisition device may be used to capture the incident light, the projected image, or a pattern of light emitted, reflected, transmitted, or otherwise converted by the substrate. Suitable image acquisition devices include CMOS cameras, CCD cameras or microscopes, for example fluorescence microscopes.
Prototypes of L3PDOP micro photocuring rapid prototyping ("μ SL") systems were constructed to fabricate 3D scaffolds such as tubes, catheters, round stakes and vascular structures. This system has been used to create structures from different biopolymers including polyethylene glycol diacrylate (PEGDA, functionalized with fibronectin for cell adhesion), methacrylated hyaluronic acid (MeHA) and gelatin methacrylate (GelMA) for 3D scaffolds. These precisely engineered biomimetic scaffolds provide a unique platform for the study of cell-microenvironment interactions. The encapsulated cells showed good cell viability both on the surface of the scaffold and in all geometries inside these structures. The cells respond individually and collectively to geometric cues throughout the larger-sized pattern.
Figure 3 provides a photomicrograph of an example of cellular response to complex 3D geometric cues showing dynamic interactions with the scaffolds to remodel the position and shape of the structures. The upper panel (a) shows GelMA (gelatin methacrylate) scaffolds with encapsulated NIH/
FIGS. 4A-4C provide micrographs of cells still encapsulated within the GelMA scaffolds, showing 3D cell spreading while maintaining viable cell-material interactions.3D reconstruction (FIG. 4A) and confocal fluorescence micrographs looking at different Z-planes throughout the scaffold (FIG. 4B) reveal that NIH/3T3 fibroblasts on the surface of the scaffold exhibit a morphology different from cells inside the structures.in FIG. 4B, cells still embedded within the GelMA scaffold four days after encapsulation show pseudopodia protrusion preferentially towards the surface of the encapsulating structure.F-actin (red) and nuclei (blue) of the cells are stained.in FIG. 4C, 10T1/2 cells encapsulated within the GelMA scaffold express a smooth muscle cell phenotype, as shown via staining for α -actin, and 8 days after encapsulation maintain cell-material interactions, as shown by the pseudopodia protrusion of 3D. all scales are 50 μm.
Microfabrication of highly complex 3D biomimetic structures characterized by the reconstitution of cellular and biomaterial components of natural physiology can yield countless applications in clinical settings as well as within the basic development effort. Providing tissue constructs based on patient-specific anatomy at a scalable and high throughput level can address the rapidly growing market in regenerative medicine. Complex scaffold geometries can be used for cell seeding and/or cell encapsulation to provide optimal conditions for directing stem cell differentiation and/or maintaining their pluripotency. Commercial services can be established to provide customized high-throughput multi-cell scaffold arrays for studying cellular responses within complex 3D geometries and interfaces. Commercial applications are also more widespread to provide micron resolution components in seconds using any photopolymerizable material. This provides a significant improvement over current rapid prototyping systems that still have limitations in manufacturing time, feature resolution, and layer artifacts.
Fig. 5 is a set of fluorescence micrographs (a-D) showing examples of different 3D scaffolds constructed from hyaluronic acid using the 3D printing platform of the present invention. Fig. 6A is a scanning electron micrograph of a log stake scaffold and fig. 6B is an optical micrograph of a vascular microstructure prepared from GelMA using the L3PDOP printing platform of the present invention.
The CAD model used to generate the L3PDOP printed mask for the 3D scaffold is typically composed of a series of digital image slices. These images may also be derived from MRI or CT scans. Using a control computer and appropriate software, the images are automatically and sequentially loaded into the DMD chip one by one, and then projected into the photopolymerizable materials to form 3D structures by a digital, sequential polymerization process. The adaptability of the platform allows modular addition and subsequent decoupling of different components of complex 3D constructs, providing a means for determining individual contributions of material types, co-culture populations, spatial cell arrangements, and biomimetic geometries. Because individual pixels of the micromirror array can be turned on and off rapidly throughout the fabrication process, highly localized (e.g., sub-micron) changes in the polymer structure in terms of porosity or material stiffness can be achieved by spatio-temporal control of UV exposure at the pixel level.
The instrumentation used in the system of the present invention (i.e., DMD array, UV lamp, linear stage) may be controlled via software written in Visual C + + or Go (golang. org) or similar object-oriented programming languages. The software may provide a graphical user interface that may be 1) implemented in a stand-alone computer application that executes locally and is controlled, or 2) implemented as a Web-based application framework hosted by a server connected to the instrument and accessed by remote clients via the internet, which provides centralized control over each component parameter, allowing the user to specify the sequence of images projected, exposure time/image, intensity of UV source (aperture setting), initial stage position, and overall height of the structure. In addition to image sequences that can be imported as bitmaps (. BMP) or PNGs (or equivalent), the software can accept text-based data files (. DAT files) formatted in a custom syntax that specifies the mirror state (on versus off) of each micromirror in the array for each voxel (i.e., x, y, z coordinates) of a 3D structure. Thus, the 3D structure may be specified not only via a sequence of bitmap images but also via an algorithmic/mathematical function. In addition, the ability to transform a topographical bitmap representation of a 3D structure into a DAT file characterized by the appropriate syntax can be included. Furthermore, the software may import any other standardized format that specifies 3D structure information, such as STL files. Ultimately, the software will control the stage preferentially along the X and Y axes, which can provide the ability to produce multiple geometries in large array formats in high throughput, batch.
To match the mechanical properties of biological tissues, naturally derived gelatin methacrylate (GelMA) hydrogels can be used as matrix materials. (see, e.g., J.W. Nicol et al, "Cell-filled microengineered gelatin methacrylate hydrogels," Biomaterials 31, 5536-5544 (2010)), the addition of methacrylate moieties to the pendant groups of native gels enables photopolymerization or photocrosslinking of the hydrogels, allowing complex structures to be engineered that support Cell adhesion and growth.
In one experiment to synthesize GelMA, pigskin gelatin (Sigma Aldrich) was mixed at 10% (w/v) into phosphate buffered saline (PBS; Gibco) and stirred at 60 ℃ until completely dissolved. Methacrylic acid anhydride (MA; sigma) was added to the solution at a rate of 0.5ml/min until an 8% (v/v) concentration of MA was obtained in the gelatin solution. The solution was stirred at 50 ℃ for 1 hour, then diluted 2x with warm PBS and dialyzed using a 12-14kDa cut-off dialysis bag (Spectrum Laboratories) in distilled water at 40 ℃ for one week to remove unreacted groups from the solution. The GelMA solution was frozen overnight at-80 ℃ and lyophilized in a lyophilizer (Labonco) for one week. The freeze-dried GelMA foam was stored at-80 ℃ until further use. To prepare the hydrogel prepolymer, freeze-dried GelMA macromer was mixed into PBS at 10% or 15% concentration and stirred at 60 ℃ until completely dissolved. Photoinitiators (1% (w/v), Irgacure 2959, steam polishing (CIBASpecialty Chemicals)), UV absorbers (0.1% (w/v), HMBS, (2-hydroxy-4-methoxy-benzophenone (benzphenone) -5-sulfonic acid), sigma), and radical quenchers (0.01% (w/v), TEMPO, sigma) were added to the solution to allow photopolymerization and provide effective cure depth and optimal pattern resolution. To assess the degree of cure, FTIR spectra of GelMA macromers were measured for comparison with crosslinked GelMA to assess the extent of spectral change.
The biological manufacturing method of the invention can be used to create a biomimetic three-layer helical stent. Scaffold parameters (spacing, height, etc.) can be optimized to facilitate delivery of nutrients, oxygen, and other factors. (a) Different combinations of GelMA solution concentration, (b) UV exposure time, and (c) UV intensity can be used to control these scaffold parameters. It is also possible to combine different biochemical factors (drugs/growth factors or biomolecules) and 3D biological structural arrangements. This functional scaffold allows for the temporary release of entrapped biomolecules to facilitate prolonged and sequential signaling to optimize cellular function. The 3D biomimetic scaffold can be fabricated on a glass slide or a PDMS substrate. A flexible PDMS chip can be used to evaluate how mechanical stress affects the scaffold and the encapsulated biomaterial.
To reduce the possibility of UV damage, a recently discovered photoinitiator, lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP), may be used. (see, e.g., B.D. Ferbanks (Fairbanks) et al, "photo-initiated polymerization of PEG-diacrylate with lithium phenyl-2,4,6-trimethylbenzoylphosphinate: polymerization rate and cytocompatibility," biomaterials 30, 6702-6707 (2009) ") LAPs provide greater water solubility, higher polymerization efficiency with 365nm light sources, and minimal cytotoxicity. Furthermore, LAP has a significant absorption above 400nm, which allows for efficient polymerization using visible light. In other implementations, photoinitiators with absorption across other wavelengths in the visible spectrum may be employed.
Tensile testing can be performed on the 3D helical stent to determine failure strain, ultimate tensile strength, and strain energy density. Each bioscaffold (2cm x 1cm x 0.5cm) can be placed in an Instron 5542 machine equipped with grips and a 50N load cell. Mechanical testing of three strain cycles in the range of 0.05 and 0.2 can be applied to the sample at a strain rate of 2mm/min until failure. The slow rate may be selected to minimize viscoelastic effects. The results of these measurements can be compared to natural tissues from the literature and guide the design of these scaffolds.
Since GelMA is derived from the extracellular matrix (ECM) of mammals, this biomaterial should exhibit low toxicity and good biocompatibility. However, if these GelMA scaffolds do not achieve the mechanical properties desired for the intended application, Hyaluronic Acid (HA), naturally derived non-sulfated glycosaminoglycans, and components of the extracellular matrix may be used in combination with GelMA. Fibroblasts can also be incorporated in the same or alternate layers to create a multicellular spatially distributed environment.
An important object of the printing system and method of the present invention is to provide the creation of a bio-fabrication platform to develop a micro-tissue chip with integrated biosensors using, for example, mouse cardiomyocytes and human induced pluripotent stem cell (hiPSC) derived cardiomyocytes. The printing platform of the present invention allows for the fast and scalable manufacturing of 3D, highly specified biomimetic structures. Furthermore, the adaptability of this platform allows modular addition and subsequent decoupling of different components of complex 3D constructs, providing a means for determining material types, spatial cell arrangements, and individual contributions to biomimetic geometry for recapitulation of natural tissue physiology. The interaction of cells with their microenvironment-consisting of extracellular matrix (ECM), aqueous environment, soluble factors and neighboring cells-is the basis for cellular processes such as migration, differentiation, lineage specificity and tissue morphogenesis. The integrated sensing capability will allow manipulation of the cellular environment with high precision and molecular coordinated visualization inside the cells in response to their environmental cues.
Unlike currently available 3D printers, the dynamic optical projection printer (i.e., the L3PDOP printer) according to embodiments of the present invention allows for matte, disposable, in-hole printing of mechanically soft biomaterial hydrogels with fully defined complex 3D geometries at near nanometer resolution. For most material types and geometries (with a total print frame size of 3mmx 5mm x 1 mm), the printing process for a single well can be completed in 15 to 60 seconds, allowing printing of one complete 24 well plate in 5 to 15 minutes. Furthermore, each hole within the plate may be pre-filled with a polymerizable solution having a unique material composition and/or concentration independent of the geometry to be fabricated. By multiplexing these different parameters over different wells within one culture plate, a high throughput 3D cell screening platform for drug discovery, clinical diagnostics, and basic life science research can be manufactured on demand in an efficient time scale using user instructions.
In one embodiment of the printing platform of the present invention, a tool is created to study isolated 3D neural networks using stereolithography, which allows fine control of the composition and structure of the 3D neural environment while achieving high resolution stimulation and recording. The utility of this approach achieves systematic control over the complexity of the neural network and facilitates the study of complex interactions between the neural environment and the functional scaffold. Finally, recent work comparing the neurophysiology of 3D cultures with 2D cultures has shown that 3D cultures exhibit complex discharge patterns with significantly reduced synchronized activity, similar to the discharge patterns found in vivo networks and in contrast to the activity exhibited by 2D cultures. These findings indicate that the classical and well-studied behavior of isolated neural networks can be completely altered in a 3D environment. Thus, the platform of the present invention represents a significant advance in the current understanding of the neurocomputational aspects of discrete neural networks and their subsequent utility for studying the nervous system. These advantages impact basic neuroscience research into normal and pathological neurological states, while also facilitating clinical translation of neural network models by providing disease models that accelerate progression and significantly reduce the cost of drug discovery for neurodegenerative and psychiatric diseases in hundreds of individuals.
In one embodiment of the 3D printing platform, an in-hole high throughput printing method is performed as illustrated in fig. 8A-8C. In this method, the prepolymer solution is not completely isolated within the target substrate, as opposed to the "inversion" method illustrated in FIGS. 2A-2C and described above with reference to FIGS. 2A-2C. Instead, the image is projected such that it is focused on the distal end of the optical assembly 82 (in this embodiment, the focusing
Referring to FIG. 8C, this FIG. 8C graphically illustrates time t during printing1、t2And t3When the
Under this printing scheme, the projected light is not transmitted through the previously printed layer to print subsequent layers as is done in the "upside down" method described above. In addition to producing structures with fully enclosed features, this method allows the printing of much higher structures. This is in contrast to attempts to print closed cavities or channels using the "inversion" method, which often produces unwanted polymerization in the "void" areas because the light must pass through these areas to print subsequent layers. However, this "inversion" approach may provide increased utility when combined with this immersion-based approach, as one approach may be used to fabricate subassemblies of a complete stent using one material type, while another approach may be subsequently used to print different materials within the same stent.
The initial results using the non-inverted method are shown in fig. 9A and 9B, where fig. 9A is a screen shot of CAD software for designing and rendering two different 3D models, one open-sided cube (left) and one open-sided cylinder (right). The printing was performed on a standard microscopic slide with a PDMS pad placed over the slide to define a well containing the prepolymer solution. The final printed hydrogel (formed from PEGDA) is shown in fig. 9B, clearly reflecting the desired topography as determined by the CAD design shown in fig. 9A.
To ensure preferred adhesion of the cured scaffold to the substrate (as opposed to the optical window), the target glass substrate may be chemically modified to facilitate cross-linking covalent bonding of the printed structures to the slide or aperture plate of the glass base. In one embodiment, methacrylation is used. In addition, the optical window may be surface treated to render it hydrophobic. The problem with window adhesion can also be partially alleviated by its own continuous manufacturing technique, which has the following advantages: the substrate does not reside stationary at the printing interface, which occurs with conventional layer-by-layer printing.
The continuous motion of the stage, synchronized with the dynamic projection mechanism, not only allows for "no-layering" printing, but also provides the ability for layered nano-and/or micro-patterning of printed materials. Specifically, sub-regions within a large scale structure can be "sub-patterned" using different sequences of spatio-temporal light modulation. As a simple example, a flat or cubic structure such as in fig. 10A and 10B can be printed by continuous exposure of the same projected square produced by steady-state constant light intensity or by intermittent pulsed light intensity. When the light is pulsed or flashed, the photon flux available to initiate free radical polymerization at a particular pixel location can be variable. This translates into a non-linear difference in polymerization kinetics on neighboring voxels. When the stage is continuously moving while the light is modulated in this manner, the crosslinking or polymerization efficiency can be varied with high spatial and temporal resolution, thereby achieving micron or submicron patterning of material properties within a larger simple monolithic structure.
FIG. 10A shows a theoretical rendering of the intrinsic structure of a simple cube produced by different light exposure sub-patterns. In this case, the light is modulated in a checkerboard fashion as a function of the voxel's distance from the surface of the cube: larger checkerboards are used for more interior portions of the cube, while smaller sizes are used for regions of the cube closer to the surface. In fig. 10A, the magnified image in the lower graph below the full view also shows how this sub-patterning of light can interact with pre-polymers having different properties in terms of material properties (such as polymerization efficiency, macromer concentration, and molecular weight). This may result in more precise detailed structures or alternatively more diffuse internal patterning. As shown in fig. 10B, the same exposure pattern (upper image) can theoretically produce different internal topographies (shown in the center image and lower image of fig. 10B), depending on the material used. Fig. 11 and 12 provide examples of the same patterning algorithm, e.g., different checkerboards, applied to more complex shapes, e.g., circular rings. This sub-patterning can be applied to complex, large scale structures to locally modulate porosity and diffusivity. One approach may be to increase the pore size of the more deep interior regions that block low diffusion rates, thereby "normalizing" the diffusivity of the overall bulk structure. Similarly, local differences in mechanical properties such as elastic modulus can be patterned to affect the death and phenotype of cells encapsulated within these printed constructs. To provide another example, this sub-patterning technique can be applied as shown in the biomimetic example presented in fig. 13A-13C, showing the rat brain (fig. 13A-CAD design) and the sub-patterning arrangement of possibilities at different "layers" or "slices" (fig. 13B and 13C) and the human lower limb model (fig. 14).
Example 1: brain chip
The L3PDOP platform described above creates two powerful modes of operation, which are illustrated diagrammatically in fig. 7. The
Importantly, the platform of the present invention can not only function as a manufacturing and optogenetic stimulation modality but also be used to record network activity across multiple domains: 1) multi-electrode array (MEA) -printing 3D neural network structures on it-allowing orthogonal recording of electrical activity coincident with optogenetic stimulation; and 2) fluorescence imaging of voltage sensitive dyes and proteins can be performed simultaneously by using the same or similar light sources used for bioprinting, thereby increasing the resolution of a single neuron of the network activity.
In this example, the goal was to create a brain chip platform that integrates 3D heterogeneous neural cultures, optogenetically-activated high-resolution stimulation, and concurrent electrical and image-based recording of neural activity. Furthermore, the fabrication methods utilized in the present platform provide control over the topographic complexity and cellular and material composition of the neural environment, allowing the possibility of increased complexity of system control.
The goal of studying isolated neural cultures is to test and explore simple in vitro systems that can represent physiologically relevant models. Although a large amount of neurophysiological data detailing the function of single-layer neurons is widely available, the conditions used for these cultures are significantly different from those present in native tissues. Given the large body of literature showing the effects of the environment on neural function, it is reasonable to assume that the behavior of 2D cultures is not a good representation of the complex system of neurophysiology in vivo. Recent work demonstrated that neurons grown in 3D scaffolds with incorporated glia provided significant morphological and electrophysiological differences compared to neural networks grown in 2D cultures. This is due to the fact that: the 3D neural scaffold more closely resembles the complex neural environment existing in the body.
Example 2: neural clusters from ipscs
The L3PDOP platform of the present invention has myriad applications in the following areas: basic neuroscience research, clinical diagnostics, screening of new drug candidates, and comparative research of artificial versus biological neural networks in conjunction with neuromorphic hardware (e.g., "brain chip" devices). The printing technique of the present invention enables easy and rapid manufacturing of 3D hydrogels that support culture of iPSC-derived neurons, with the following capabilities: 1) 3D patterning of hydrogel microenvironments via dynamic optical photo-curing rapid prototyping for high-throughput scaffold fabrication, co-registered with the MEA with high precision; 2) directional arrangement of nerve clusters and modulation of cell arrangement, growth and connectivity via hydrogel composition, geometry and size; 3) support for functional connectivity, as demonstrated by positive expression of relevant biomolecular markers, and 4) spontaneous electrical activity during long-term culture of neural populations, corresponding to the spatial arrangement of these neural populations on the 3D printed hydrogel-MEA substrate.
The photocured rapid prototyping 3D bioprinting described herein allows for the photopolymerization of a variety of biocompatible materials. For example, but not by way of limitation, such materials include gelatin methacrylate (GelMA), methacrylated hyaluronic acid (MeHA), polyethylene glycol diacrylate (PEGDA)). The printing method also facilitates cell encapsulation within these structures. Other classes of specialized photo-crosslinkable polymers may include conductive polymers such as PANI-PAAMPSA [ polyaniline-poly (2-acrylamido-2-methylpropanesulfonic acid) ]. (see, e.g., Liu (Yoo) J.E. et al, "Directly patternable, highly conductive polymers for branched applications in organic electronics)", Proc. Natl.Acad.Sci.USA (PNAS), (2010)107(13), 5712-. Furthermore, these materials-as well as the different cell populations incorporated-can be used heterogeneously in the same scaffold to provide multi-component 3D patterned co-culture substrates, facilitating the study of complex cell-cell interactions and cell-ECM interactions.
For 3D patterning of Neural Progenitor Cells (NPCs), hydrogels were prepared using photopolymerizable macromer solutions of 5% to 7.5% gelatin methacrylate (GelMA) in DPBS. Arrays of interconnected microwells with different radii (5 to 300 μm), depths (100 to 500 μm) and channel widths (5 to 50 μm) were printed via the photocuring rapid prototyping technique described above. GelMA hydrogel scaffolds were used to culture and pattern iPSC-derived NPCs. UV light mediates polymerization that occurs within 30 to 60 seconds of patterned exposure to produce scaffolds 5mm x 3mm x 100 to 500 μm. The transparent MEA is aligned with the image projected from the DMD by translating the linear stage using a microscopic camera mounted under the projection stage. The 3D GelMA mount with pre-defined culture wells, interconnecting pathways/channels, and reverse slots for electrode sites was precisely aligned with the MEA to allow direct two-sided printing of hydrogels, as shown in figure 15.
Upon inoculation, NPCs aggregate within the wells and cause neurites to elongate to varying degrees along the guide channels, depending on the water content of the material the percent concentration of GelMA (5% versus 7%) has a substantial effect on cell spreading, differentiation, alignment and connectivity (fig. 16) after one month of culture, scaffolds were fixed in 4% paraformaldehyde and immunofluorescent labeled with antibodies to synapsin, VGlut1, MAP2, GFAP and β -III tubulin to assess cell morphology, function and synaptic connectivity as shown in fig. 17, markers of functional connectivity exhibited differences based on substrate type and patterned dimensions, with higher levels of synapsin expression observed in the culture on the GelMA substrate when compared to the glass control, sparse expression of VGlut1 was also observed, indicating early commitment of some neurons to glutamatergic phenotype.
To further assess functional connectivity, neurons were transfected with plasmid DNA for Wheat Germ Agglutinin (WGA) and Green Fluorescent Protein (GFP) reporter genes (fig. 18). Neuronal staining positive for WGA and negative for GFP indicates that postsynaptic uptake of WGA occurs through functional connectivity between WGA-expressing and non-expressing populations. Furthermore, synaptophysin expression was additionally assessed via a plasmid directed to the co-expression of GFP localized with synaptophysin expression (fig. 19). As previously discussed, greater expression was seen in the 3D scaffold condition (lower panel) compared to the glass control (upper panel).
Patterned neural networks also exhibit spontaneous, spatially dependent electrical activity. The MEA electrode locations corresponding to the centers of these nerve clusters within the hydrogel pores exhibited the highest activity when mapped to the recording channel. As shown in fig. 20A-20D, channel 20 (
Example 3: integration with 3D printed conductive polymer electrodes
High density 3D recording of neural networks can be achieved by: a bio-conductive polymer, such as PANI-PAAMPSA, is incorporated in a precise spatial arrangement to transmit neural electrical signals to a multi-electrode array (MEA) substrate using the photocuring rapid prototyping techniques presented herein. Fig. 21A and 21B show one possible spatial arrangement and different electrode lengths, respectively. Fig. 21A shows a simple grid-like array in
Example 4: adaptive, reprogrammable bio-logic gates, sensors, and hybrid neural networks
Synthetic mimics for neural computation exhibit much higher performance in tasks such as visual recognition and speech recognition compared to conventional von neumann (von Neuman) architectures. Within the cognitive and computational sciences, software algorithms (e.g., artificial neural networks) and hardware implementations (e.g., neuromorphic hardware) of neural circuits have been developed to engineer a computational substrate with equivalent capabilities. These techniques have been widely applied to solve computationally intensive problems such as data mining, automated medical diagnosis, and self-learning game strategies, as well as to expose underlying mechanisms under different dimensions of human cognition. However, the degree to which the specific computational processes and learning algorithms of these synthetic systems truly represent their biological counterparts remains largely unknown. The integrated neural fabrication platform described herein can be used to bridge this gap and to develop new technologies for hybrid artificial/biological neural network devices that can generate computational power that has not been realized by synthetic or natural versions of these networks separated.
One embodiment of creating a biological neural network using the invention described herein is a logic gate consisting of two
The logic gate can then be retrained using a new truth table to transform the network to a different logic gate type, e.g., transform the logic gate from an and gate to an exclusive or (XOR) gate, where the output will be "1" if and only if an input is "1" (fig. 23D). The XOR gate training sequence is shown in fig. 23A (showing the MEA) and fig. 23B (showing the corresponding spike grating pattern and the associated time period of the light stimulus), using blue light ("B") and green/yellow light ("G/Y"). In this example, as shown in fig. 23C,
On a larger scale, the biological neural network can be constructed to contain a plurality of numbers of inputs, outputs, and interconnections, as well as a plurality of intermediate layer nodes that ultimately connect the input layer with the output layer. In this implementation, the biological neural network may resemble an artificial neural network model, such as a perceptron. An example of a
Using the platform, neural stimulation can be provided optically by projecting light onto the optogenetically modified neuron or electrically via the MEA interface. In the case of optogenetic activation, these neurons can be modified to generate populations that express different light-sensitive opsins that elicit different electrical responses, e.g., channel rhodopsin-2 (ChR2) for depolarization/excitation versus salt bacteriorhodopsin (NpHR) for hyperpolarization/inhibition. By incorporating heterogeneous populations of wild-type excitatory and inhibitory neurons, such as glutamatergic and gabaergic neurons, and any of a myriad of interneuron species found in natural tissues or synthetically engineered to exhibit more exotic or non-naturally implemented behavior within each node at different ratios, more complex models of network dynamics characterized by inter-neuron modulation can be achieved via excitation or inhibition. Furthermore, these neurons may be subject to optogenetic modification before or after incorporation into the 3D scaffold. In the latter case, the scaffold material may be functionalized via 3D patterning to allow local transfection of plasmids encoding different opsins according to one defined spatial arrangement. Finally, the stent, network, and stimulation parameters may be modulated in conjunction with other environmental changes, such as the optically activated degating of neurotransmitters and the administration of other mechanical, chemical, and/or biological stimuli that may affect the training efficiency, robustness, and overall dynamics of the network.
In another experimental example, assessing the long-term fidelity of the delivered biological neural network may provide information about its ability to act as a long-term biological substrate for memory. After being subjected to a training period as previously described, a multi-well MEA plate or multi-chambered MEA microscope slide consisting of several in vitro biological neural networks can be maintained in culture for an extended period of time according to standard cell culture protocols. After a defined point in time, e.g., 1 to 4 weeks, the fidelity of the output response of the network may be quantified to determine the durability or volatility of the output response, i.e., the durability of the embedded memory. This non-von neumann model of integrated memory and computation provides the basis of many theoretical frameworks for human cognition and has recently been implemented in the context of hybrid CMOS-memristor neural circuits.
By integrating this high-throughput porous biological neural network within one larger hybrid artificial/biological neural network system, a hierarchical distributed computing algorithm can be implemented to take advantage of the features of both neural architecture types or to effectively segment the larger complex work into perceptron-specific unit tasks. In one embodiment, each well 252a (see fig. 25B) of a
In another embodiment, due to the high speed capabilities of the projection device, the biological neural network may be delivered to respond to more dynamic stimuli, such as to respond to self-learned video game strategies, most often in such a way that: artificial neural networks have been designed, for example, for use
Example 5: system for controlling a power supplyIn vitro model of a typical neural circuit in neuroscience research
In vitro culture and interrogation of isolated neuronal populations forms a core platform that is widely utilized throughout neuroscience research. The ability to forward engineer (i.e., to deterministically construct patterned neuronal connections) is necessary for corroborative evidence of connected-set models derived via reverse engineering (i.e., in vitro recordings and brain imaging at different scales) of existing native neural circuits. The true reconstruction of anatomical features that contribute to the hypothetical canonical loop depends on the ability to systematically pattern the heterogeneous neuron population with the appropriate relative spatial arrangement. The unique methods described herein for constructing these 3D neural network structures have the potential to provide increased complexity of systemic control that is difficult to achieve using conventional isolated neural populations. Furthermore, high density 3D recording combined via integrated printing of conductive polymer electrodes enables measurement of network dynamics in a suitable spatial context while using soft biocompatible hydrogels characterized by mechanical properties closer to those of in vivo tissues. Using a heterogeneous population of neural cell types that reflect the constituents found in the native neural circuits, the 3D printing methods described herein can be applied to create a compartmentalized and/or systematically reduced representation of the typical in vivo nervous system. These in vitro constructs can then be used to assess the functional significance of different system components in a more readily available form to determine, for example, individual and collective contributions of neural sub-populations in higher levels of cognition and behavior.
- 上一篇：一种医用注射器针头装配设备
- 下一篇：用于增材制造三维物体的设备