*2.8. Statistical Analysis*

Results were collected from at least six independent experiments. All data are expressed as mean ± standard error (SE). Data were analyzed by Student's *t* test.

**Figure 3.** Flowchart of process parameter optimization. Every parameter consists of the values tested and, on the right, the corresponding measurable variable for new cell cultures. Fabrication parameters are in the left column and design parameters in the right.

#### **3. Results: Scaffolds Production**

Following the method developed, experimental work was first carried out to find the best way to produce scaffolds which can sustain cell cultures. Sequential work was done to set scaffold design features and manufacturing process parameters.

#### *3.1. Optimization of Process Parameters*

Processing parameters were optimized to achieve high quality scaffold printing for cell culture application. Thus, different physical scaffold variables were measured to ensure the correct fit between the computer design and the printed scaffold. The processing parameters included both fabrication and design parameters as shown in the "Experimental Setup" section (Table 1). Processing parameters were chosen according to the literature and the state-of-art [9,11,16,17]. However, the process optimization methodology explained here, based on a sequential flowchart (Figure 3), is both innovative and unique.

Experiments were initially carried out with a generic scaffold design (see Section 2.4 "Methods") to set the fabrication parameters and then adjusted to the design parameters required to produce the scaffolds.

Fabrication parameters (extruder and bed temperature, deposition velocity, and layer height) were introduced with Slic3r software. These parameters are related to the characteristics of the polymeric material (mainly PCL) and the printing process. However, different values were tested for the parameters (by checking the measurable variable mentioned in Table 1) in order to meet scaffold manufacturing requirements.

Once the polymeric material and its fabrication parameters had been characterized and set, design features were subsequently established using the SolidWorks 3D software. Parameters, such as filament diameter, distance between filaments, and deposition angle, were tested. These are related to the three-dimensional design of the scaffold and the effect they have on the cancer cell culture.

First, to determine the optimal fabrication parameters, a fixed scaffold design was established as a control pattern: 90◦ deposition angle, 0.4 mm in diameter filament and 1 mm distance between filaments. This enabled us to do printings with the same design, but different fabrication parameters, to find the optimal ones. Later, as the design parameters were optimized, they were replaced.

Following the flowchart defined in Figure 3, all the parameters were characterized and selected sequentially to obtain the appropriate setup for producing 3D-printed scaffolds. The optimization of each process parameter is described in the following sections.

#### *3.2. Extruder Temperature*

Poly(ε-caprolactone) was chosen as the polymer to work with because of its compatibility with cell cultures. PCL has a low melting point (60 ◦C). To achieve enough malleability and considering there is some heat dissipation, higher temperatures were also tested to find the optimal value (Table 1). A fixed scaffold design described in the Methods section was printed. Then, *the printed filament diameter* was measured as a physical variable. Low extruder temperatures (65–80 ◦C) could not melt the material enough, thus the amount of the extruded material was low. As a consequence, the printed filament diameter was smaller than the one designed (0.4 mm). High temperatures (90 ◦C) melt the polymer excessively and also increase the diameter of the filament due to flattening and some blobs being produced. Therefore, the optimal extruder temperature was established at 85 ◦C. The printed filament diameter was 0.39 ± 0.05 mm.

#### *3.3. Bed Temperature*

To set the optimal bed temperature, a generic geometrical scaffold design was printed, and two different measurable variables were evaluated. Material adhesion was assessed as a binomial variable (yes/no), firstly testing the lowest temperature (25 ◦C, Table 1). If the printed material had not adhered enough to the surface (no), another printing was performed, this time with a higher bed temperature. Once the material had adhered to the surface (yes), the first layer height was then measured. Bed temperatures ranging from 25 to 33 ◦C gave a non-adherent first layer scaffold. In addition, much higher temperatures (37 ◦C) melt the material excessively, flattening the filament and decreasing the height of the first layer (lower than the 0.4 mm designed one). A 35 ◦C bed temperature was considered optimal as this allowed first layer adhesion and the filaments were not flattened. Their first layer height was 0.37 ± 0.07 mm.

#### *3.4. Deposition Velocity*

The goal with this parameter was to find a high deposition velocity without forgetting the quality of the printed scaffold. The printed filament diameter was chosen as the tangible variable with which to analyze the impact this parameter has on the scaffold. The optimal deposition velocity was established as being 10 mm/s. The filament diameter was 0.42 ± 0.05 mm. When the speed was faster (20 and 30 mm/s) the material did not have enough time to deposit itself on the surface, resulting in smaller filament diameter or sometimes even discontinuous filament production.

#### *3.5. Filament Diameter*

At this point, the diameter of the printed filament deposited on the collector was analyzed. Extrusion and deposition velocity can exert a direct influence on fiber morphology. Therefore, once he manufacturing velocity had been fixed, the diameter of the extruded filament was evaluated next. Three different design filament diameters were tested: 0.175, 0.3, and 0.5 mm. To ensure the filaments remained tangent along the vertical axis, the printer's layer height was adjusted to each design filament diameter. Diameters that were too large caused the adhesion of two contiguous filaments, favored by their proximity and elevated temperature. For this reason, the first variable studied was the possible adhesion of contiguous filaments, such as a binomial variable (yes/no). Thus, only the values that did not cause the adhesion of two filaments in the same layer were selected to continue the analysis (0.175 and 0.3 mm). The second measured variable was the printed filament diameter. A design diameter of 0.175 mm caused erratic printing because the amount of material was too low to form a linear filament. The final value tested, 0.3 mm, was found to be optimal as it gave a printed filament diameter of 0.31 ± 0.02 mm. The established filament diameter value also determined the thickness of each layer. The scaffolds were manufactured with eight layers, so the final thickness of the scaffolds was 2.4 mm.

#### *3.6. Layer Height*

Layer height is defined as the distance between two connected layers along the *Z* axis. Since all layers are designed and printed on top of each other, this parameter was determined by the printed filament diameter. For this reason, the filament diameter, although being a design parameter, was established before finishing, optimizing the fabrication parameters (Figure 3). In some cases, the deposited material tends to flatten out and so the printed height is lower. At that point, two different values were analyzed: 0.3 mm (the whole filament diameter) and 0.28 mm (because of a certain flattening) and the quality of the printing recorded (absence of blobs). In this case, 0.3 mm was found to be the optimal layer height for our design as flattening, due to high temperatures, did not occur. When evaluating smaller established heights, the printing process produced blobs.

The absence of filament flattening may be attributed to the relatively low extruder temperature used, (85 ◦C, see Section 2.1) which can be considered low compared with other biocompatible polymers used in 3D printing, such as PLA [9,12].

#### *3.7. Distance between Filaments*

This is a key parameter because it affects the pore size of the scaffold [9]. This design parameter consists of the shortest distance between the axis of two filaments located within the same layer. We were interested in achieving small pore sizes, thus, we focused on the testing small distances (0.5,

0.7, 1 mm). Nevertheless, small distances between filaments may be problematic if two contiguous filaments join. For this reason, the real distance between filaments was measured to take into account whether this value matched that of the one expected (designed).

Distances of 0.7 and 1 mm gave no filament joining, so real distances were higher than 0. Within these values, the smallest value was chosen (0.7 mm). Taking into account this parameter and the optimal filament diameter previously established, the distance between the outer parts of two contiguous filaments was, consequently, 0.4 mm (Figure 4).

**Figure 4.** Distances between two contiguous filaments. Axis (0.7 mm) and outer distance (0.4 mm) are represented. Filament diameter was fixed at 0.3 mm.

#### *3.8. Deposition Angle*

Once all previous parameters were optimized, three different scaffolds with different deposition angles were designed and manufactured, thus obtaining different pore characteristics, which may influence cell attachment and growth (Table 2). As high-quality printings for all three designs were achieved, it was agreed to test the adequacy for 3D cell culture with all three designs.

An MCF-7 breast carcinoma cell line was used to preliminarily evaluate scaffold ability in terms of three-dimensional cell culture. MCF-7 cells were seeded onto scaffolds and cultivated for 72 h. Then, attached cells were trypsinized and counted. No cells were counted on the 90◦ scaffolds. Under an optical microscope, no cells were observed on the filament, but rather attached at the bottom of the microplate well (Figure 5a), which is in agreement with cell counting. Scaffolds of 45◦ showed a subtle cell adhesion of 3.52 ± 1.16% when compared with the 2D control. We subsequently tested 60◦ scaffolds, which showed an increased cell adhesion of 26.50 ± 10.98%. In both cases, cells were previously observed at the well bottom and attached to the scaffold filaments, with the last ones are indicated by white arrows (Figure 5b,c, respectively).


**Table 2.** Scaffold designs with different deposition angles: 90◦, 45◦, and 60◦.

**Figure 5.** Optical microscope images of MCF-7 cells seeded on the scaffolds. In 90◦ scaffolds, cells were attached at the bottom of the well (**A**). In 45◦ and 60◦ scaffolds, cells were attached both on the scaffold and at the well (**B**,**C**, respectively). (**D**) MCF-7 cells in 2D culture. White arrows on the images indicate cells adhered to PCL filaments. Scale bars represent 100 μm.

Then, scaffolds were also evaluated through fibroblast cell cultures. Murine NIH/3T3 fibroblasts were seeded onto the three designs during 72 h and cell proliferation was assessed. In this case, fibroblasts adhered to all three scaffold models (Figure 6), with the highest cell proliferation value being found on the 90◦ design (56.30 ± 5.03% compared to the 2D control). The other two architectures exhibited slightly smaller values. For instance, 60◦ scaffolds presented a 49.52 ± 5.62% cell growth and 45◦ models, 39.11 ± 8.12%, compared to the monolayer culture.

**Figure 6.** Optical microscope images of NIH/3T3 cells seeded on the scaffolds. Cells were attached on scaffolds of 90◦ (**A**), 60◦ (**B**), and 45◦ (**C**). (**D**) NIH/3T3 cells in a 2D culture. White arrows on the images indicate cells adhered to PCL filaments. Scale bars represent 100 μm.
