Page 397 - IJB-10-2

International Journal of Bioprinting Automated bioink mixer improves bioprinting

The mixing process was further evaluated by varying Subsequently, the cell distribution and viability were
the number of exchanges between 25 and 200 times at examined. As depicted in Figure 6A and B, the cell
mixing speed of 10 mm/s. Again, cell distribution and distribution and cell viability were considered generally
viability were assessed following the mixing process satisfactory after mixing by the human operator. The cells
(Figure 5C and D). Predictably, the analysis indicated that were well-dispersed in the bioink, and the average cell
the homogeneity of the bioinks improved with increasing viability was relatively high. However, cell viability had
numbers of exchanges. Unexpectedly, however, the comparatively high variability in cells taken from either
metabolic activity was almost unaffected by the number adjacent sites or different positions within the cartridge
of exchanges, and even after 200 exchanges, only a slight (Figure 6B). This can be attributed to insufficient mixing
decrease in metabolic activity was observed. Furthermore, in the local regions and the longitudinal distribution of
the standard deviation between different experiments the cells by a poorly controllable mixing process. As a
decreased at higher numbers of exchanges. Staining of result, the average cell viability varied by more than 40% in
dead cells with ethidium homodimer-1 did not indicate some experiments.
a substantially higher number of dead cells after 200 In comparison, the machine had a substantially
exchanges compared to fewer exchanges (Figure S3 in superior mixing performance in terms of both the cell
Supplementary File). homogeneity and cell viability (Figure 6C and 6D).
Next, we tested various combinations of mixing speed The cells were distributed more evenly across different
and numbers of exchanges, as higher speeds tend to achieve positions within the cartridge. The average cell viability
homogenous bioinks at lower numbers of exchanges. As was found to be slightly higher than that of human
can be seen in Figure 5E and F, the combination of a low mixing. Importantly, the deviations of cell viability were
speed of 5 mm/s with a high number of exchanges (100 lower for different positions in the cartridge and between
times) failed to achieve homogeneity, consistent with the independent experiments. Similar results were also found
above-described results of the mixing speed test (Figure and confirmed in zigzag bioprinting of the bioinks (Figure
5A). The high-speed combinations of 30 mm/s and 70 6E and F). The favorable capability of the automated device
exchanges and of 50 mm/s with 50 exchanges resulted in was also validated using another cell line without GFP
homogeneous bioink mixtures with well-dispersed cells. expression (HepaRG) immediately after bioink mixing and
However, cell viability was substantially reduced, with bioprinting process (Figure S5 in Supplementary File).
the difference for 50 mm/s with 50 exchanges reaching The automated device thus produced bioinks with higher
statistical significance. In contrast, the combination of homogeneity, comparable cell viability, and superior
reproducibility between multiple experiments compared
a mixing speed of 10 mm/s and a mixing number of 80 to bioinks produced by a skilled human operator, who has
exchanges achieved comparatively good cell homogeneity, rich experience in bioink mixing in the long-term practice.
while maintaining high cell viability. To further improve To investigate the influence on cell functions, the albumin
the cell distribution in the bioink, the mixing speed secretion and cytochrome P450 3A4 (CYP3A4) activity
was slightly increased to 15 mm/s at a mixing number were evaluated using HepaRG cell line, both of which are
of 80 exchanges. Compared to the speed of 10 mm/s, important liver function markers. The results indicate that
homogeneity was improved further without a decrease in HepaRG cells maintained the secretion level of albumin for
cell viability (Figure S4 in Supplementary File). Thus, a both mixing groups from day 1 to day 7, while the values
speed of 15 mm/s and a mixing number of 80 exchanges for CYP3A4 activity increased substantially (Figure S6 in
gave the best overall performance, and the parameter Supplementary File). No significant difference was detected
combination was used for all subsequent experiments. The for the two mixing groups during the testing period.
mixing process of an alginate hydrogel (6%) and methylene Therefore, bioink mixing executed by automated machine
blue solution (0.001%) under these conditions is shown in did not negatively impact cell functions compared to that
Video S1 (Supplementary File). by experienced human operator.

3.4. Reproducibility of bioink mixing by machine In addition, to evaluate cell distribution and viability,
As described above, we found that bioinks produced by we also measured the rheological properties of the bioinks
different human operators had inconsistent properties. produced by the human operator or the automated mixing
We therefore assessed the bioink mixing by the automated device. The rheological properties are of utmost importance
device and compared its reproducibility to that of a for extrusion-based 3D bioprinting. The analysis revealed
relatively experienced human operator (No. 3 in the that all the bioinks after either human or machine mixing
experiments above). Both man and machine mixed the 6% exhibited stability under the testing frequency ranging
alginate hydrogel with the HEK293-GFP cell suspension. from 0.1 to 100 Hz (Figure 7A and B) and demonstrated

Volume 10 Issue 2 (2024) 389 doi: 10.36922/ijb.1974