For many years, cells cultured on flat monolayers have represented the reference in vitro models to grow and study cells and to evaluate the cellular mechanism in response to biophysical and biochemical stimulations. However, it has been described that two-dimensional (2D) culture conditions do not faithfully reflect in vivo conditions given that proper tissue architecture and cell–cell interactions are partially lacking in monolayers [1
]. Three-dimensional (3D) cell cultures have therefore been introduced with the objective to better mimic in vivo systems compared to cells cultured as monolayers on plastic [2
]. Recently, it was demonstrated that the gene expression profiles [6
] as well as the responses to drug treatment in 3D tumor models resemble the in vivo situation more closely [7
]. Therefore, 3D cell models, including the well-described production of 3D microtissues, are increasingly being employed as biologically relevant systems for drug development and preclinical drug testing [8
]. Particularly, in the field of anticancer drug development, the use of 3D microtissues for drug testing appears appropriate. It is well known that tumor cells cultured in vitro with traditional models in conventional static 2D culture in multiwell plates poorly mimic the pathological environment associated with avascular tumors. In this respect, 3D microtissues are far better at reproducing inherent O2
, nutrient, and metabolite gradients, providing a more accurate prediction of drug toxicity and efficacy.
Numerous 3D culture models currently exist for cancer-related research, such as liquid overlay-based culture [9
], non-adhesive surfaces [10
], stirred-tank culture strategies [11
], and hanging drops [12
]. However, many of these culture systems are characterized by a high labor cost or a limited capacity to mimic the extracellular matrix (ECM) [13
]. In this respect, the 3D microtissues are often used in suspension (i.e., not attached to a substrate), making experiments in flow conditions difficult (i.e., to simulate blood flow). More recently, microfluidics has also been explored to obtain 3D tumor microtissues, often called spheroids [14
]. Compared to traditional 3D cell culture methods, microfluidics possesses many advantages such as high-throughput and low-cost for drug screening. In addition, microfluidics also offers the possibility to work in flowing conditions for high-fidelity microtissue models and to integrate hydrogels to create biological microstructures [16
With respect to the production of systems including ECM, recent studies have been performed using hydrogels to mimic native ECM [17
] and as cell-instructive materials [18
]. Hydrogels obtained with natural biopolymers have been employed to develop 3D cell culture systems and to create scaffolds for tissue engineering to mimic the in vivo microenvironments and cellular support matrices [19
]. Hydrogels have also been integrated in microfluidic systems for tissue engineering application [21
Biomaterials for 3D cell cultures are generally expensive and therefore not always usable within large screening platforms; they include collagen type I [22
], hyaluronic acid [23
], and Matrigel [24
]. In this respect, gelatin could represent an interesting alternative to these costly biomaterials. Gelatin is a heterogeneous mixture of water-soluble proteins of high average molecular masses, present in collagen; it has been routinely used to modify substrates/matrices that are inert to cells [25
]. Interestingly, gelatin forms thermoreversible gels, but to guarantee its use in biomedicine, chemical or physical cross-linking procedures are often needed [26
]. Unfortunately, the chemical crosslinking protocol reactions (i.e., the residues of unreacted crosslinkers) can result in adverse cellular effects [27
]. To cite some examples, Anseth et al. [28
] and Khademhosseini et al. [29
] proposed photo-crosslinkable gelatin hydrogels to promote 3D cell cluster formation of aortic valvular interstitial and HepG2 liver cancer cells, respectively. Both research groups prepared gelatin modified with methacrylamide (named GelMA) to create covalently stabilized gelatin hydrogels for in situ cell encapsulation and 3D cell culture. The GelMA hydrogel approach was also used by other authors to prepare stable hydrogels for modularly engineering biomimetic osteon and epidermal tissue engineering, respectively [30
]. The majority of methods for 3D microtissue production with gelatin involve the use of rather complex processes and chemical crosslinkers, potentially resulting in toxic effects on the seeded cells.
An alternative cross-linking method for gelatin hydrogels is represented by controlled dehydration, as reported by Yannas et al. The authors demonstrated that gelatin becomes covalently cross-linked when the water content falls below a critical trace level of about 0.2 g/100 g protein [32
]. Recently, this method was re-introduced to fabricate cross-linked gelatin molds with variable size (depending on the dehydration of the gelatin dispersion) to produce microfluidic chip platforms [33
]. In addition, the gelatin cross-linking by dehydration was also employed to obtain the reversible bonding of different materials currently employed in microfabrication, such as poly(methyl methacrylate) (PMMA), PDMS, and glass (paper submitted). Interestingly, up to now, no articles have appeared, to the best of our knowledge, describing the production of 3D microtissues by gelatin substrates cross-linked by dehydration (i.e., without the use of chemical cross-linkers).
In the current paper, the use of the gelatin dehydration process is described to produce ultra-low attachment microfluidic channels for 3D microtissues with human colon adenocarcinoma (HT-29). Notably, the gelatin hydrogels were fabricated into a PMMA microfluidic chip and, as control on 6-well plates. To determine cell viability and proliferation rate, the generated 3D microtissues were analyzed using live/dead assay and live cells counting. Finally, to morphologically characterize the 3D microtissue in the proposed novel substrate, confocal microscopy analyzes were performed.
2. Materials and Methods
2.1. Materials and Equipment
The poly(methyl methacrylate) (PMMA) slabs used in this study, thickness 1.2 mm, were from a single batch purchased from Goodfellow Cambridge Ltd., (Huntingdon, UK). Gelatin from porcine skin type A was obtained from Sigma-Aldrich (St. Louis, MO, USA). The microfabrication process was performed by a micromilling, Minitech Machinery Corporation (Norcross, GA, USA), equipped with microtools (flute end mills diameter of 889 μm) from Performance Micro Tool (Janesville, WI, USA). Silicone peroxide tubing/60 Shore i.d. 0.75 mm obtained from IDEX Health & Science Gmbh (Oak Harbor, WA, USA) and Hypo Needles 18 AWG purchased from Warner Instruments LLC (Hamden, CT, USA) were used to test the microfluidic chip sealing.
2.2. Fabrication of Microfluidic Chips and Gelatin Coating
Microfluidic chips were designed and fabricated using PMMA slabs (thickness 1.2 mm) by micromilling. For the micromachining, a CNC micromilling machine was employed. During micromilling, spindle speed, feed speed, and plunge rate per pass were set at 10,000 rpm, 15 mm∙s−1
, and 20, respectively. The produced slabs have the following characteristics: (A) a top slab (75 mm × 20 mm, length/width) containing a linear 35 mm length microchannel (880/880 µm, width/depth) and (B) a bottom slab acting as sealing part for the microchip. The deposition of gelatin coating and chip sealing, named GEL-D bonding method (paper under revision), were performed following the general scheme reported in Figure 1
. A gelatin dispersion in water (15% w/v), degassed under vacuum for 10 min to eliminate bubbles, was heated at 70 °C for 10 min. Typically, 500–800 µl of the warm (45–50 °C) gelatin dispersion in form of a droplet was deposited at the center of both the top and bottom PMMA parts of the chip; thereafter, both slabs were spun at high speed (1500 rpm for 20 s) using a spin coater (Laurell, WS-650 Series, North Wales, PA, USA) or, alternatively, a low-cost modified computer cooling fan, as recently reported [34
]. The top and bottom parts of the chip were assembled, clamped and the sealed chip was incubated at 4 °C for 5–10 min and maintained at room temperature (24 °C) for further 24–48 h before use. The control uncoated chips were bonded by a solvent (ethanol) evaporation method [35
2.3. Cell Seeding and Characterization
The in vitro cell culture experiments on gelatin-coated chips were performed using the HT-29 and HepG2 cell lines. Cells were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA) and routinely cultivated in RPMI 1640 and DMEM media, respectively, supplemented with penicillin-streptomycin solution (10,000 U/mL) and 10% Fetal bovine serum (FBS). Before cell seeding, to prevent contamination, the microfluidic chips were treated with a 10% penicillin-streptomycin solution in Phosphate-Buffered Saline (PBS) for 12 h at 4°C. Cells were seeded in the microfluidic chips (50 µL, 5 × 104 cells/mL) by a micropipette. The seeded microfluidic chip was maintained in a 5% humidified CO2 incubator at 37°C. Cells growth was monitored using inverted light microscope for up to 10 days. Pictures of the channel were taken after the seeding and at days 1, 2, 5, 7, 8, 9, and 10. Quantitative analyses of 3D tumor microtissue distribution and sizes (on chip) were performed using the ImageJ software (NIH, Bethesda, MD, USA). The surface area of the 3D microtissues was determined by tracing the contours with the freehand function in ImageJ and measured using the measure function. A surface area of 9 mm2 was analyzed for each experiment.
The live/dead staining was performed using Calcein AM and Sytox orange nucleic acid stains at concentrations of 1 μM and 0.250 μM, respectively, and cell death visualized with an epifluorescence microscope [36
]. Hoechst 33342 (1 µL/mL) was used for nuclei staining. The 3D structure analyses were performed using a confocal microscope Leica equipped with 4× and 10× objectives. The cells were fixed using a solution of 2% Glutaraldehyde in 0.1M Phosphate Buffer, pH 7.3. After 20 min, the fixative agent was removed and the samples were washed by flowing three times PBS in the microchannel. After permeabilization with 0.5% Triton-X100, the cells were incubated with blocking buffer (1% bovine serum albumin (BSA) in PBS) and stained with Phalloidin for 1 h at room temperature and 4′,6-diamidino-2-phenylindole (DAPI) (1 µL/mL)