(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Low-cost, versatile, and highly reproducible microfabrication pipeline to generate 3D-printed customised cell culture devices with complex designs [1]

['Cathleen Hagemann', 'United Kingdom Dementia Research Institute Centre', 'Institute Of Psychiatry', 'Psychology', 'Neuroscience', 'King S College London', 'Maurice Wohl Clinical Neuroscience Institute', 'London', 'United Kingdom', 'The Francis Crick Institute']

Date: 2024-03

Cell culture devices, such as microwells and microfluidic chips, are designed to increase the complexity of cell-based models while retaining control over culture conditions and have become indispensable platforms for biological systems modelling. From microtopography, microwells, plating devices, and microfluidic systems to larger constructs such as live imaging chamber slides, a wide variety of culture devices with different geometries have become indispensable in biology laboratories. However, while their application in biological projects is increasing exponentially, due to a combination of the techniques, equipment and tools required for their manufacture, and the expertise necessary, biological and biomedical labs tend more often to rely on already made devices. Indeed, commercially developed devices are available for a variety of applications but are often costly and, importantly, lack the potential for customisation by each individual lab. The last point is quite crucial, as often experiments in wet labs are adapted to whichever design is already available rather than designing and fabricating custom systems that perfectly fit the biological question. This combination of factors still restricts widespread application of microfabricated custom devices in most biological wet labs. Capitalising on recent advances in bioengineering and microfabrication aimed at solving these issues, and taking advantage of low-cost, high-resolution desktop resin 3D printers combined with PDMS soft lithography, we have developed an optimised a low-cost and highly reproducible microfabrication pipeline. This is thought specifically for biomedical and biological wet labs with not prior experience in the field, which will enable them to generate a wide variety of customisable devices for cell culture and tissue engineering in an easy, fast reproducible way for a fraction of the cost of conventional microfabrication or commercial alternatives. This protocol is designed specifically to be a resource for biological labs with limited expertise in those techniques and enables the manufacture of complex devices across the μm to cm scale. We provide a ready-to-go pipeline for the efficient treatment of resin-based 3D-printed constructs for PDMS curing, using a combination of polymerisation steps, washes, and surface treatments. Together with the extensive characterisation of the fabrication pipeline, we show the utilisation of this system to a variety of applications and use cases relevant to biological experiments, ranging from micro topographies for cell alignments to complex multipart hydrogel culturing systems. This methodology can be easily adopted by any wet lab, irrespective of prior expertise or resource availability and will enable the wide adoption of tailored microfabricated devices across many fields of biology.

Funding: The Serio lab acknowledge support of the UK Biotechnology and Biological Sciences Research Council (BBSRC) [BB/T014318/1] [BB/W006561/1] and of the Dementia Research Institute (UKDRI). F.S.T and A.S. are part of the Horizon Europe “MAGIC” consortium (101080690; www.magic-horizon.eu ); this work is funded by UK Research and Innovation (UKRI) under the UK government’s Horizon Europe funding guarantee grant numbers 10080927, 10079726, 10082354 and 10078461. Work in the Tedesco lab was also supported by the European Research Council (759108), AFM-Telethon (21687), BBSRC (BB/M009513/1), CureCMD (576031), Muscular Dystrophy UK and the NIHR (the views expressed are those of the authors and not necessarily those of the National Health Service, the NIHR, or the Department of Health). This research was funded in whole, or in part, by the Wellcome Trust. We would like to acknowledge the Making Lab facility, a Science Technology Platform at the Francis Crick Institute; A.S, A.I and F.S.T acknowledge support by the Francis Crick Institute, which receives its core funding from Cancer Research UK, the UK Medical Research Council (MRC) and the Wellcome Trust (CC0102). K.S.S and C.D.S acknowledge funding from the Leverhulme Trust (RPG-2022-174). C.D.S acknowledges generous support through a Wellcome Trust Career Development Award (225257/Z/22/Z). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability: All data and procedures are available within the paper and additional information and designs are available as part of this publication and associated files, as well as available to download from the corresponding GitHub repository ( https://github.com/SerioLab/SOL3D ).

To demonstrate the applicability of this method to several different biological experiments and provide an effective ready-to-use pipeline for other labs that do not have expertise in microfabrication, we demonstrated its use to develop customisable culture devices ranging from μm to mm and cm scale, with complex 3D shapes or and micro topographies. Together with the detailed protocols, we also provide the designs for each device showcased, which can be customised to fit different experimental needs.

Driven initially by our own experience with adapting microfabrication techniques to hard biological questions and inspired by the many recent technical advancements by different groups, we aimed to create an optimised and universally effective pipelines that would include the production and post processing protocol for so ft l ithography on 3D vat polymerised moulds ( SOL3D ), using a low-cost commercially available printer and materials.

To overcome these challenges and facilitate the production of complex 3D constructs suitable for cell culture, several successful post processing and coating approaches have been attempted [ 33 – 35 ]. However, these protocols generally involve either long heat and detergent treatments [ 36 ], which often cause print deformation, or expensive techniques [ 37 ], not accessible to every lab. Others have circumvented this issue by using custom-made resins [ 38 ]. One example of the latter, coating of UV resin vat polymerised prints with parylene is effective in creating usable moulds and is sufficient to overcome curing inhibition of PDMS [ 39 ], but requires the use of specialised equipment and adds another potentially technically challenging step to optimise.

One possible solution to these problems would be to combine PDMS with UV resin vat polymerised 3D-printed moulds and effectively employ 3D printing in lieu of photolithography in conventional pipelines. However, curing of PDMS on vat polymerisation resin prints can be challenging as constituents of most commercially available resins inhibit PDMS polymerisation [ 28 – 31 ]. Furthermore, this curing inhibition makes demoulding difficult and can result in leaching of cytotoxic uncured PDMS monomers into the cell culture medium of even successfully demoulded designs [ 32 ], making the devices unusable for cell culture applications.

Unfortunately, most commercially available resins for UV vat polymerisation are cytotoxic and cannot be used for cell culture applications [ 24 ]. Additionally, the composition of these resins is often proprietary, and conversion or production of biocompatible resins requires skills limited to dedicated chemistry laboratories [ 25 ]. Some biocompatible resins are commercially available; however, they tend to be sold at a much higher cost than even high-resolution resin, and more than the actual printers in some cases (e.g., Phrozen sonic mini 4K printer = GBP 365 [ 26 ], 1L Zortrax Raydent Crown and Bridge resin = GBP 392 [ 27 ]—prices at time of writing, for illustration only), undermining the applicability of 3D SLA printing for cell culture purposes ( S1 Fig ). Moreover, unlike PDMS and other silicon-based materials used for soft-lithography resins do not have tuneable stiffness and are generally not optically clear, factors that prevent their ability to act as a suitable substrate for cell culture or microscopy.

There are a variety of different 3D printing techniques, ranging from filament deposition to vat polymerisation of resins. Most of these techniques are commercially available but more complex and intricate methods have been developed over time, such as two-photon-based microfabrication [ 20 – 23 ]. Although these custom-built, high-resolution setups are significant technical advancements, we focus here on vat polymerisation as it represents the most accessible and cost-effective form of 3D printing with a μm-scale resolution and provide further details on the other techniques as a complete overview in Supplementary guide 1 ( S1 Text ). Vat polymerisation is using a layer-by-layer construction of complex volumes using a UV curable resin as a material. Features are built though shining UV light on a thin volume of resin with a build plate next to it. The resin cures and attaches to the build plate. The plate then moves and gives space for a new layer of resin that is polymerised on to the layer of resin from before. The final construct is printed by using patterned UV light that resembles the end design layer by layer.

3D printing technology has emerged as an accessible and adaptable tool for fast prototyping and fabrication of small objects. Alongside their increasing availability, rapid technological advancements in 3D printers have led to the development of several open-source projects aiming to enable any wet lab to create and adopt critical and innovative modelling strategies [ 14 – 19 ]. This is particularly important when considering the challenges experienced by laboratories in less developed countries in sourcing equipment or specific consumables, or the sometimes-steep practical barrier that some labs encounter when venturing into cell culture and biology from a different field.

In short, conventional microfabrication generates micron and millimetre-scaled features with a combination of photolithography and soft lithography methods. Particularly photolithography requires specialised facilities and expertise, and generally allows to only generate features within 1 scale at the time, as it is based on serial deposition of photo sensible polymers of defined thickness ranging from 1 μm to generally 500 microns maximum. While photolithography is essential for creating many popular cell culture devices, the specialised equipment, relatively long timescales, and expertise needed, limits wider adoption of custom microdevices [ 11 ].

We present here a detailed guide on how to use this pipeline, intended for biological and biomedical wet labs, together with the necessary information and context on the techniques involved across the fields of bioengineering and microfabrication and several application examples. What follows is a short contextualisation of the relevant technologies, in abridged form, and we provide an in-depth explanation and description of the available systems within the Supplementary guide ( S1 Text ).

Capitalising on recent important advancements in bioengineering and microfabrication aimed at solving these issues, and taking advantage of low-cost, high-resolution methods, we have developed an optimised a low-cost and highly reproducible microfabrication pipeline, thought specifically for biomedical and biological wet labs with not prior experience in the field, which will enable them to generate a wide variety of customisable devices for cell culture and tissue engineering with features that vary from 20/50 μm to several centimetres using 3D vat polymerisation.

Several bioengineering and fabrication strategies have been developed to create custom-engineered culture environments that direct the cell’s environment [ 10 , 11 ] and they have enabled countless new biological insights. Most of these strategies include 2 key parts: a suitable material than can be made biocompatible and a method for shaping it into the desired forms. PDMS (polydimethylsiloxane) is a biocompatible, optically clear silicon-based elastomer with tuneable stiffness (800 kPa—10 MPa), compatible with multiple chemical modifications suitable for cell culture, and represents the most widely used material for fabrication of microfluidic devices and countless other culture platforms for investigating complex cellular interactions in vitro, while the biocompatibility also depends on the used cell type and surface modification [ 12 , 13 ]. Most of these cell culture devices and platforms have features ranging from micrometres to several millimetres, depending on the size of the cells, the required volume of the culture medium, and experimental paradigm. Being able to fabricate custom devices with both micron, millimetre and in some case centimetre scale opens countless possibilities for biological experiments but this often requires the combination of multiple techniques, and in some case, a variety of different expertise and equipment or resources that are not common in biologically focused labs. Consequently, it remains challenging for wet labs to perform rapid prototyping of user-handleable macroscale devices with microscale features for cell culture applications.

Stem cell-based models are an invaluable resource, which allow the study of nearly any cell type in vitro [ 1 – 4 ]. The advent of cellular reprogramming and subsequent access to patient-derived stem cell models have also galvanised their position as an ideal tool to investigate cellular processes in health and disease [ 5 – 9 ]. While stem cell models offer significant control over the identity of cultured cell types, the conventional culture systems used for them typically lack the ability to control key parameters of the culture itself, which greatly influence the analysed biological processes. These parameters include the relative position of the cultured cells, grouping, cell–cell and cell–material interactions and many others, depending on the biological questions asked. Although a plethora of commercially devices allow some degree of control over culture conditions, they are often non-customisable and require adaptation of the biological experimental parameters to the specific device characteristics, rather than the more desirable opposite.

2. Results

SOL3D fabrication allows the generation of complex 3D-shaped stencils for precise control of cell positioning and grouping within open wells 3D-printed stencil-aided dry plating devices to control cell location and number within standard well plates. Conventional open well culture systems generally do not allow control over cell position, grouping, and numbers in an easy and reproducible fashion, limiting the complexity of in vitro modelling experiments. Several techniques are available to overcome these limitations and to create precise arrangements of cells within culture vessels—from microfluidic devices to cell bioprinting—; however, most rely on creating compartmentalised structures that limit the manipulation of the cells granted by the open well systems. A different approach that allows both to increase complexity within conventional culture vessels and to maintain an open well system are stencil-like plating devices [40,41]. These systems are temporary structures that guide the organisation of cells within an open well, although at present they suffer from the same fabrication limitations as the abovementioned strategies. Moreover, this method is especially affected by the technical limitations of feature sizes and aspect ratios dictated by photolithography, resulting in thin devices that are difficult to handle and have limited customisation possibilities (S8 Fig). We decided to model these types of devices, using human-induced pluripotent stem cell-derived MNs as a cell model system, for an initial proof-of-principle of our optimised SOL3D protocol, based on an engineered platform we recently developed for MN cultures using a micropatterned substrate to facilitate axonal elongation [42]. We combined and optimised this platform with our SOL3D moulding protocol to create a tailored plating strategy for investigating hiPSC-derived MN behaviour with control over cell location and orientation. We designed moulds for casting PDMS stencil-well devices, rectangular extruded features with funnel-shaped media reservoirs as complex 3D features to ease cell seeding. This optimised design permits rapid and facile manual seeding as cells can settle into micro-sized wells in a suitable volume of medium to avoid excessive evaporation and cell death (Fig 2A and 2B). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. SLA 3D printing enables control over cell location and number in an open well. (A) Schematic overview of the investigation dry plating strategy to combine PDMS casts from 3D-printed moulds with PDMS microgroove substrate to control cell body location and number. Well sizes range from 600 μm × 1,000 μm to 50 μm × 1,000 μm in 50 μm intervals. (B) Schematic overview of funnel shape well for easy manual seeding in microwells. (C) Representative images of stencil devices filled with pre stained (Silicon Rhodamine-tubulin) MN progenitors with device still in place (top), after device removal (middle), and axonal β-III Tubulin following fixation after 7 days of culture (bottom). (D) Comparison of seeded cell area after stencil removal to CAD specified values by fold change. (E) Representative images of 3D aggregoid with 2D axon elongation stained with DAPI (first image), axonal β-III Tubulin (second image), and Dendritic MAP-2 (third image) simultaneously (fourth image). https://doi.org/10.1371/journal.pbio.3002503.g002 PDMS stencils from these 3D moulds allow seeding by “dry plating,” whereby a stencil is placed in a dry conventional tissue culture plastic vessel and cells in suspension are manually pipetted in the stencil device, isolating the cell bodies from the residual well and allowing them to adhere at these specific positions. With the cell bodies secured, the stencil device can be removed and the whole well filled with culture medium, while the adhered cells remain in their specified position. For this “dry plating” process, a strong fluidic seal surrounding the PDMS stencil wells is necessary, requiring a flat surface between the stencil and substrate below. Without specific steps to adjust the surface roughness of prints, PDMS casts from 3D-printed moulds are inherently rougher than those from micropatterned silicon wafers used for casting microfabricated PDMS devices (S7 and S9 Figs). We, therefore, implemented an additional clamping step before PDMS curing, using a silanised glass slide (see M&M) to cover the PDMS surface, which is in contact with air, taking advantage of the flat surface provided by the glass (S10A Fig). We evaluated the efficiency of clamp-cured stencil fluid seals when placed on a PDMS micropatterned surface with 10 × 10 μm-grooves using a blue dye. An effective seal was achieved in all stencils cured using the additional clamping, denoted by dye reaching the microgroove substrate in the well area only and spreading within these specific grooves. Stencils cast without clamping showed uncontrolled dye spreading throughout the devices, verifying a lack of fluid seal (S10B Fig). We were then able to use the optimised stencil devices to answer a biological question and investigate the minimum number of iPSC-MNs required to form a self-organised 3D neural aggregate on microgrooves for axonal elongation, a process determined by chemotaxis and topography. To achieve this, we used the above-described stencils with a funnel-shaped reservoir and rectangular wells, varying in Y-dimension to reduce the stencil well size and control cell amount. The well dimensions were homogenous and faithful to CAD specifications throughout the print sizes down to 50 μm in Y (S11 Fig). These PDMS stencil-well devices were placed on the extra cellular matrix (ECM) coated and dried micropatterned surface with axonal guidance grooves [42], and the iPSC-derived MN cell suspension was manually pipetted into the dry wells of the device. To avoid potential air pockets in the smaller wells, as it is common for non-functionalised PDMS, we performed oxygen plasma treatment on stencils prior to cell “dry plating” (S12 Fig). Compact rectangular “aggregoids” (i.e., 3D cell clusters generated by reaggregating single cells from a culture) with decreasing size were achieved during seeding and were maintained following device removal. Staining with β-III-tubulin after 7 days in differentiation medium revealed that wells with a size of down to 150 μm provide suitable cell numbers for aggregate formation. However, the 2 smallest well sizes did not provide the environment for aggregate formation and cells migrated across the topography (Fig 2C and 2D). Subsequent staining with compartment-specific markers showed a clear separation between dendrites and axons in the open well devices of compact aggregoids (Fig 2E). In summary, stencil-well devices cast with SOL3D can be used to control cell location in an open well, facilitate control over different cell numbers in the same device, and enable cell compartment-specific investigations.

Spatial, temporal, and morphological control over cell–cell interaction using tailored plating devices We next explored the potential to use SOL3D for more complex cultures, incorporating different cell types and plating time points. First, to plate multiple cell populations within the same devices, we sought to take advantage of PDMS natural hydrophobicity coupled with the large rectangular well design and funnel-shaped well profile showed in Fig 2A. The high contact angle between media and hydrophobic PDMS allows to achieve confinement of the different cell suspension droplets, which generates enabling complete fluidic separation between adjacent wells containing different cell populations, even with manual plating. We sought to utilise this strategy to simultaneously plate different iPSC-derived MNs populations within the same tissue culture well in different spatially separate pockets of the PDMS device. For this, we used fluorescent RFP+ and untransfected MNs, which we plated manually within adjacent pockets in the same device placed on an ECM-coated well of a 6-well plate. The different cell populations were left to adhere for 2 h before the plating device was removed. After removing the device and further cell culture for 72 h, all MNs were stained with a silicon rhodamine tubulin dye (here depicted in green for visualisation), to visualise all neurites and cell bodies. A line graph analysis across the whole device showed that all wells contained MNs (RFP+ and RFP-/Tubulin+) and in every second well, RFP+ cells were present (Fig 3A and 3B), demonstrating multi cell type seeding in confined predetermined spatial groups within a single device. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Plating devices enable spatiotemporal control of cell plating with different geometries for construction of complex neural circuits. (A) Schematic overview of alternate seeding of RFP and non-RFP+ motor neurons in the same device and the following live cell staining. (B) Representative line profile of stained across the well showing segregation of individual populations to their designated wells—RFP+ only in wells 1 and 3, but SiR-tubulin+ (here in green) in all wells (top). Representative fluorescence images of stained RFP+/− motor neurons (bottom). (C) Schematic overview of the multi-device protocol for constructing a neural circuit using 2 stencil devices and 3 different cell types (MNs, cortical, and astrocytes) with seeding performed at different time points. (D) Composite of the complete circuit after 19 days of culture. GFP transfected motor neurons = green, Glial Fibrillary Acidic Protein (GFAP) identifies astrocytes, β-III–Tubulin identifies cortical neurons and the tubulin in GFP+ motor neurons. Blue device well shapes overlaid for illustrative purposes. (E) Schematic overview of protocol for manipulating aggregate geometry in combination with existing microgroove. (F) Representative images of SiR-tubulin live-cell stained motor neuron aggregoids at day 2. (G) Boxplot of aggregoid aspect ratio fold change by shape between CAD (blue line) and day 2 of culture from β-III-Tubulin channel (top). (H) Boxplot of aggregoid area fold change by shape between CAD (blue line) and day 2 of culture from β-III Tubulin channel. Data points for G and H can be found in the files 3G-Data and 3H-Data in S1 Data. https://doi.org/10.1371/journal.pbio.3002503.g003 We then sought to further increase the complexity of our in vitro cultures by seeding multiple cell types at different time points within the same well, taking advantage of the efficient and reversible fluid seal between our devices and the culture plate. For this, we placed 2 rectangular plating devices, approximately 2 mm apart from each other on an ECM-coated micropatterned surface within a well of a 6-well plate, as described above. Initially, one device was used to dry plate iPSC-derived cortical neurons [43] and astrocytes [44] in a 1:1 ratio and removed after 24 h, while the other device was kept empty. The whole tissue culture well was then filled with differentiation medium and cultured for 9 days. During this time, cortical axons guided by the microtopography extended toward the empty device, which maintained its initial fluidic seal even surrounded by medium. On day 9, GFP+ MNs were seeded in the second device by first lowering the level of the medium within the well to be below the edge of the plating device, and then seeding the MNs in suspension directly within it. After allowing for cell attachment, the second device was also removed, the well refilled with fresh medium and cells cultured for further 9 days. The position of the different cell types was then verified using immunocytochemistry (ICC) for astrocytes (GFAP) and neurons (both MNs and cortical neurons, β-III-Tubulin), as cortical neurons could be identified by the overlap of GFP and β-III-Tubulin. Using these tailored removable SOL3D-generated plating devices, we were able to easily plate 3 different cell types at 2 different time points within the same culture well, creating a complex neural circuit and demonstrating true spatiotemporal control over cell seeding in a cost-effective and highly adaptable fashion (Figs 3C, 3D, and S13). Additionally, we demonstrated also multiple time point seedings within the same well using large format “nesting” plating devices that can be used to construct large-scale cell and tissue arrangements (S14 Fig). Next, we tested whether the ease of available prototyping using our optimised SOL3D protocol could enable investigation and manipulation of the fundamental behaviour of complex iPSC-derived MN cultures. It has been shown that aggregation of cells using different geometries has an influence on the signalling environment and patterning of aggregates [45]. We therefore designed and fabricated moulds for PDMS stencils with 3 different well shapes: rectangular, circular, and triangular, to create geometrically constrained neural aggregates. With the advantage of producing multiscale features simultaneously, we were able to preserve the funnel reservoir and straight well design from previous moulds (Fig 3E). Using these multi-shaped stencils, we seeded motor neuron progenitors (MNPs) as before on our micropatterned substrate and allowed axons to extend for 11 days (Fig 3E). Here, MN aggregates maintained faithful area and aspect ratios to CAD specifications on day 2 after stencil removal. After 11 days, the groups also retained their specific geometry although showed slight changes in the aspect ratio and area over time (Fig 3F–3H). Next, we used these PDMS stencils on nonpatterned and uncoated tissue culture plastics to avoid cell adherence, directing self-organisation of aggregate-like structures (S15A Fig). Here, we seeded cortical progenitors in Matrigel and were able to generate differently shaped aggregates demonstrated by SiR-tubulin live dye images after 24 h (S15B Fig). SOL3D fabrication can therefore be used as a valid method of fabricating constructs for controlling cellular interactions in complex cultures of multiple geometries for both 2.5 (i.e., partially tridimensional adherent cultures) and 3D non-adherent cultures (e.g., aggregates) depending on the seeding substrate.

2.2 SOL3D fabrication allows the generation of micro topographies In the initial experiments presented in Figs 2 and 3, we employed SOL3D-fabricated devices on top of microgrooves generated with conventional photolithography. As this method is not available to all labs, we then focused on achieving the same level of organisation within the culture but on a platform purely based on SOL3D-manufactured devices. While the 3D vat polymerisation printers do not have the same resolution as photolithographic equipment, we could not simply recreate the 10 μm grooves. Moreover, with current LCD-based illumination, repetitive patterns in close proximity to each other and close to the minimal resolution pose a challenge for 3D vat polymerisation printing, due to the illumination pattern and diffraction of the light (see supplementary guide page 15 in S1 Text). However, the key parameter is the biological organisation rather than the material geometry, and we, therefore, focused on obtaining a design that can both be printed with SOL3D and achieves the same neurite orientation. We developed a design with different groove geometry and dimensions (Fig 4A), with a groove depth of 200 μm and width of 100 μm. Optical profiling of the 3D prints reveals that the grooves are shallower than the CAD design, around 60 μm, and less wide, as expected (S16 Fig). Despite these limitations, we tested if the topography is sufficient to align axonal elongation. We plated MN neural aggregates [42] in a SOL3D-manufactured stencil device on SOL3D-manufactured grooves following the “dry plating” protocol (Fig 4B). Visualisation of axons (β- III-Tubulin) after 7 days of culture revealed an alignment with the topography throughout the axonal length (Fig 4C and 4D). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. SOLID fabricated grooves guide axonal elongation and alignment of muscle fibres. (A) CAD design of triangular grooves 100 μm wide and 200 μm deep with 75° triangular spacing in between. (B) Schematic overview of “dry” neural aggregate seeding in PDMS stencil devices on the 3D-printed grooves and axonal elongation. (C) Representative overview image of axons (β-III-Tubulin) on the SOLID fabricated grooves with magnifications of axons aligning with the topography in the proximal (blue) and distal (orange) compartment. (D) Representative 3D reconstruction of axons (β-III-Tubulin) on grooves in the proximal (left) and distal (right) compartment. Across compartment axonal alignment to the topography is given. (E) Schematic representation of myoblast seeding and differentiation on SOLID grooves. (F) At day 3 of differentiation myotubes (SiR-Tubulin) align to the given topography (left) compared to cells cultured on a nonpatterned surface (right). https://doi.org/10.1371/journal.pbio.3002503.g004 Another process benefitting from alignment to topographies is the formation of myotubes [46]. We seeded myoblasts on the SOL3D grooves and a flat control surface (Fig 4E). Cells were differentiated for 3 days and then visualised using SiR-Tubulin. Myoblasts cultured on the patterned surface show alignment to the topography, compared to the control conditions where cells are randomly positioned (Fig 4F). While the cells follow the given topography, the differences in size and shape of the SOL3D grooves compared to the photolithography pattern, might evoke different interaction and biological responses, making the 2 different groups not entirely comparable, but Sol3D manufactured grooves provide a suitable alternative to microfabricated grooves and provide guidance and cell alignment in some circumstances. As a resource, the SOL3D grooves provide a suitable alternative to microfabricated grooves and provide guidance and cell alignment. Taken together with the SOL3D stencil-like devices, a whole on-chip platform can be generated using SOL3D.

2.3 Customisable SOL3D fabrication as a tailored alternative to standardised commercially available culture platform The ability to create customisable devices and substrates suitable for cell culture or other biological experiments, with μm to cm sized features, in a fast, reliable, and cost-efficient manner would be particularly useful in any wet lab, granting independence from high costs, delivery times, and availability of the equivalent commercial products, while enabling substantial customisation. For example, most cell culture vessel layouts are standardised and not tailored to the need of an individual laboratory or a specific cell type, causing higher costs and potential compromises in experimental setups. We therefore aimed to test whether our optimised SOL3D mould protocol could be used to reproduce and further customise relevant features from popular commercially available cell culture products. These constructs can be customised in dimensions and/or shapes for individual experimental aims, while remaining cost-effective, highlighting the versatility and accessibility of our system to enhance biological investigations.

PDMS bonding for chamber slide devices Chamber slide systems and other microscopy-ready hybrid culture devices are commercially available systems that allow cells to be cultured within neighbouring wells directly on cover slides for high-resolution imaging, providing small well sizes and imaging-compatible set-ups for high throughput and convenience. To ascertain if our SOL3D protocol could be applied to mimic these constructs, we designed a chamber slide system that can be permanently bonded to an imaging coverslip either using oxygen plasma treatment or a UV-sensitive resin adhesive. Importantly, the adaptation of our construct for use with UV resin makes this method accessible to labs without a plasma cleaning system (Fig 5A). Our design was fabricated using the SOL3D protocol and was size matched to a 60 mm × 24 mm microscopy coverslip with 12 circular wells with funnel shapes. As described above, we generated a fluidic seal by clamping the device with a glass slide during PDMS curing to isolate neighbouring wells. This extremely flat PDMS surface allows fusion of PDMS to the glass slide using oxygen plasma bonding, or the simple application of a UV adhesive. It is important to note that the UV adhesive is resin based and therefore cytotoxic and cannot be used on any medium-facing area. Astrocyte progenitors were then seeded into selected wells at different concentrations (Fig 5B). Staining with a live dye (SiR-Tubulin) revealed an intact fluidic seal in both devices, liquids maintained in the respective wells, and healthy astrocyte progenitor populations. We further demonstrated the high-resolution imaging compatibility (Fig 5C), as cells are seeded on a glass slide. Ostensibly, we have demonstrated that both oxygen plasma and UV resin are suitable for PDMS bonding of a chamber slide device and highlighted the capabilities of 3D printing for the fabrication of bespoke chamber slides in a fast and cost-effective way. With the recent release of the Phrozen Mini 8k, with increased resolution, we sought to test our protocol for also this printer and the provided resin. Microfluidic devices are expensive and single use only, with no customisation opportunity, while they require a flat surface and a high grade of detail. We used a triple chamber design (S17A Fig) with channels of 100 μm in width and only 10 μm in height, a z-layer thickness that is barley achievable with the 4k Mini, even with optimised parameters. To further prove the advancements in printing quality, we did not clamp a glass slide onto the print for a flat surface but directly used the printed surface. Indeed, the channels were 9.4 μm in height and around 110 μm in width (S17B and S17C Fig). After oxygen plasma treatment, the surface was sealed onto a glass slide and a fluorescein solution was flowed through using fitted tubing and a syringe. Time-lapse images clearly show that the channels are connected and the solution is washed in, without leakage (S17D–S17F Fig). Overall, we confirmed that our pipeline works with the next-generation printer and that with improved quality of the prints fine grade details can be achieved also in the z axis, as well as a smooth print that does not require glass clamping. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. 3D printing can create fully customisable imaging chambers with complex well geometry suitable for cell culture via 2 different methods of PDMS bonding. (A) Schematic overview of design and manufacture of chamber slide device with large wells with (1) UV glue or (2) oxygen plasma bonding to a glass coverslip to seal the wells. (B) As demonstration of the seal quality and viability for cell culture astrocyte progenitors were seeded in different densities in nonadjacent wells and cultured in chamber slide device. Representative images of SiR-tubulin live dye-stained astrocyte progenitors 1 day after seeding in chamber slides bonded with different methods. (C) High-resolution imaging of astrocytes (GFAP) and mitochondria (TOMM20) cultured in custom-made SOL3D PDMS chambers. https://doi.org/10.1371/journal.pbio.3002503.g005

Custom microwell arrays for embryoid body formation The first design we tested for this purpose was an array of pyramidal-shaped microwells (390 × 350 × 150 μm) that we fabricated using the optimised protocol with no coating step, as it is required for this small feature size (<500 μm) (Fig 6A and 6B). These microwells have become essential for induction of specific cell lineages from iPSCs and for aggregate research [47,48]. One of the most important functions of these wells is to ensure homogenous aggregate size for reproducible results, for example, generating embryoid bodies (EBs) of regular size and shape. We used our moulded microwell arrays to form EBs from an iPSCs suspension (Fig 6C) and quantified their size after 4 days of culture on the devices. In our microwells, iPSCs formed EBs with consistent diameters, verifying the suitability of our custom PDMS moulds to create small regular arrays of features (Fig 6D). The microwells generated by our protocol are therefore suitable for generation of homogenous EBs with the benefit of substantial customisation of well shape and size at a low cost. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. PDMS substrates cast from 3D-printed devices permit regular-sized embryoid body. (A) Schematic of design, manufacturing and seeding of IPSCs on microwells. (B) Representative optical profile of 3D-printed microwell device with well sizes of 390 μm length × 350 μm width × 150 μm height. (C) Representative SiR-tubulin live cell dye images of IPSCs before seeding in microwell mould cast. (D) Representative SiR-tubulin live cell dye images of IPSCs seeded on PDMS cast from 3D-printed microwell device compared to flat PDMS substrate after seeding (left) and 2 days culture (right). (E) Representative SiR-tubulin image of fused embryoid bodies on microwell PDMS mould prior to detachment after 4 days in culture (top) and embryoid bodies following washing off the PDMS microwell substrate (bottom left). Quantification of embryoid body diameter detached from microwell PDMS mould demonstrates homogenous size of embryoid bodies (bottom right). Data points can be found in the file 6E-Data in S1 Data. https://doi.org/10.1371/journal.pbio.3002503.g006

Large tissue engineered devices with complex designs Generating complex devices for tissue engineering often combines relatively small features within large constructs and has so far proven challenging to implement in most laboratories, as construction processes are complex and time consuming, requiring dedicated expertise. Most devices of this kind are therefore sourced from commercially available suppliers, with limited possibility of customisation and at a high cost. For example, tissue-engineered 3D muscle constructs use a variety of devices for suspending large cell laden hydrogels during culture using thin suspension posts [49,50]. They are comprised of small pillars with complex shaped end feet, which serve to suspend the hydrogel construct and provide mechanical stiffness to aid differentiation. As these posts are difficult to manufacture and arrive pre-made of a single size and shape, no customisation is available, e.g., miniaturisation or altered substrate stiffness. Successful 3D adaptations have been implemented for smaller and less complex muscle post, however, with low ease of production for 3D SLA printing [51]. We used our SOL3D protocol to fabricate a device for suspended 3D muscle culture with customisable post size and overall dimensions. The challenge, in this case, stems from the fact that these devices do not have large flat faces, they present thin complex features and need ideally to be produced as a single component to avoid complex assembly steps that can introduce variability. A single mould system would in this case not be sufficient, as the lack of air in contact with the complex shapes would prohibit the successful demoulding of the structure. We created a two-part mould/injection system using SOL3D, which can easily be assembled by clamping for curing after PDMS is poured into the mould. Optimisation of the moulds showed that an unequal distribution of the design between the 2 parts (70/30) is beneficial for successful demoulding, resulting in a reproducible single device with the desired dimensions, in this case twice as large as the commercial alternative—a 2 cm muscle compatible with 12-well plates (Fig 7A). We compared our 12-well plate 3D posts to the commercially available 24-well adapted equivalent (see M&M, Muscle culture), using immortalised myoblasts. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. 3D sandwich moulds for PDMS casting to generate complex cell culture devices. (A) Schematic overview of design and PDMS casting strategy for a 3D sandwich mould. Immortalised myoblast hydrogels were then formed around PDMS posts, differentiated and cultured for 2 weeks. (B) Comparative contractility analysis with microstimulation at 20 mV with 0.5 Hz frequency of differentiated 3D muscle between crafted PDMS posts and commercially available posts. (C) Directionality analysis of fibre alignment in differentiated 3D muscle fibres between PDMS posts and commercially available posts after 2 weeks differentiation. (D) Representative images of myoblast differentiation (Titin) and developmental stage (MyHC) on PDMS posts and commercially available posts after 2 weeks differentiation. (E) Images of posts with different dimensions suited for 12-, 24-, or 48-well plate (from left to right). (F) Illustration of compressing PDMS to measure the modulus through a lateral and rotating movement. (G) The Storage and Loss Modulus of PDMS with different formulations (1:10 or 1:20) at changing angular frequency. https://doi.org/10.1371/journal.pbio.3002503.g007 Following our protocol to generate 3D bioengineered muscle [50,52], we first created a pouring mould by filling liquid agarose around a 3D-printed rectangular spacer that was removed after the agarose has set (S18 Fig). Subsequently, myoblasts were seeded in fibrin hydrogels within the agarose mould, and the SOL3D-fabricated posts (or commercially available devices [53] used in Maffioletti and colleagues) were inserted within the still-settling fibrin constructs [50]. After 2 weeks of differentiation, we performed electrical micro stimulation to measure muscle contractility—a hallmark of successful 3D muscle culture—on both constructs at 20 mV with 0.5 Hz frequency (Fig 7B), which showed periodic contractions for both SOL3D and control devices. Immunostaining of the muscle tissue showed the presence of terminally differentiated myosin heavy chain (MyHC) and titin positive multi-nucleated fibres in both constructs. Directionality analysis revealed that myofibres were preferentially aligned along the posts (Fig 7C and 7D). After demonstrating the suitability of our 3D posts for muscle cell culture, we focussed on miniaturisation of the posts, to enable obtainment of the same biological outcomes with fewer materials. For this purpose, we designed and manufactured insets for 24- and 48-well plates with the SOL3D protocol (Fig 7E). However, in addition to changing the dimensions, we also utilised the benefits of the tuneable stiffness of PDMS by changing the ratio between monomer and curing agent, while remaining optically clear and retaining most of its surface properties. Here, we used the 24-well plate format moulds and polymerised PDMS in its standard formulation (1:10) and in a lower ratio of the curing agent (1:20). The lower the ratio of curing agent to PDMS the softer the material will be, and to confirm that we can also use softer PDMS in our 3D Sol3D moulds, we measured the modulus of the muscle devices using compression (Fig 7F). As expected, the storage and loss modulus are higher for 1:10 PDMS compared to the 1:20 formulation (Fig 7G). In summary, our protocol allows for complex and scalable features to be easily moulded in PDMS with the additional benefit of customisation in all aspects of design for improved function of 3D engineered muscle tissues.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002503

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.