Bioink: Model Validation

We devised a protocol to incorporate UTEX 2973 into the MGS-1 alginate gel, for which we can adopt in the future for validating the formation of bricks from calcium carbonate crystals. Here, we also introduce Design of Experiments, a modelling approach we can use to optimize our bioink composition for UTEX 2973 viability and calcium carbonate formation in the future.

Introduction to Design of Experiments

The objective of Design of Experiments (DoE) is to analyze and model the relationship between several input variables (i.e. factors) and output variables (i.e. response variables). Each factor has at least two levels, which are the specific values a factor can take (e.g. high, medium, low). This allows us to identify relevant factors that have the largest influence on the response variable by reducing the amount of experimental conditions and modelling the factor-response relationship of the system. Given that our bioink formulation is novel, this is a useful and cost-effective way to determine what factors (compositions) in the gel influences its structural integrity and viability of UTEX 2973, in order to optimize the formulation. Futhermore, successful formulation to produce bricks will require experimenting with various conditions, making DoE strategies, such as fractional factorials, highly desirable.

The four stages of DoE execution includes planning, screening, optimization, and verification. There are also many different types of DoE with unique advantages and disadvantages depending on the experimental goal. We proposed to mainly use screening designs as a way to determine optimal bioink formulations due to limited quantitative readouts.

DBTL 1 - Earth Sand

Earth Sand Bioink Optimization on JMP

Factors and levels for earth sand bioink formulation were decided through previous studies investigating alginate and methylcellulose-based bioprinting ([1]) and their rheological properties in ensuring proper printability, shape fidelity, and biocompatibility ([2]). Although these studies were applied towards tissue scaffolding and incorporating mammalian cells, we can use the level ranges of factors relevant to our experiments for screening purposes. This will allow us to determine main effect estimates for each factor using the JMP Statistical Discovery software, which are individual impacts of each factor on the response variable. Three factors, including weight % sodium alginate, weight % earth sand, and calcium chloride concentration, were selected to investigate at three levels each.

The JMP software allows us to generate an optimal experimental design for testing these three factors at three different levels with a fractional factorial screening, which reduces the amount of experimental conditions while optimizing the amount of statistically significant information of the system that can be obtained. For this experiment, a definitive screening design (DSD) will be used. In reducing the amount of runs, there is 17 experimental conditions as outputted in Figure 1.

Table 1. Factors and levels for the JMP design matrix

	-1	0	1
[sodium alginate - X1]	2wt%	3wt%	4wt%
[earth sand - X2]	10wt%	30wt%	50wt%
[CaCl2 - X3]	50mM	100mM	200mM

We can then input experimental responses for UTEX 2973 and determine the main effect estimates in the system, allowing us to first determine important variable ranges to maintain and carry these through for further response-surface modelling experiments.

Incorporating Microbes into Biomaterials

We adapted Reinhardt et al.’s methodology for preparing biomaterial inks to our Earth Sand and MGS-1 alginate gels ([3]). First, we investigated the most efficient method for sterilizing the bioink components. Once mixing Earth Sand, 3 wt% alginate solution, and CMC together, we autoclaved this homogeneous solution at 120 degrees celsius for 20 minutes. As documented on 2025.07.22 Sterilizing Earth Sand Bioink Composition (1), the gel was filled to half the volume of the container and this overfilled after the autoclaving process. Thus, we reduced the volume to 1/5 of the container and the Earth Sand alginate gel was successfully autoclaved in 2025.07.25 Sterilizing Earth Sand Bioink Composition (2). In subsequent experiments, we also autoclaved each component separately, where 3 wt% alginate solution was autoclaved on liquid cycle and Earth Sand/MGS-1 with CMC was autoclaved on dry cycle. Previous studies have shown that autoclaving alginate led it to become non-printable as structures collapsed, where it is characterized by water loss and disruption to hydrogel structure ([4], [5]). However, in 2025.07.25 Sterilizing Earth Sand Bioink Composition (2), we showed that the autoclaved 3 wt% alginate solution still successfully crosslinked to the same degree as non-autoclaved alginate solution.

In [3] ‘s methodology, mineralization medium was prepared by supplementing additional calcium chloride and sodium bicarbonate to the growth medium in order to induce calcium carbonate formation. We implemented this by increasing the concentrations of calcium chloride, sodium carbonate and/or sodium bicarbonate to 100 mM in the BG-11 growth medium. As documented on 2025.07.25 Sterilizing Earth Sand Bioink Composition (2) and 2025.07.22 Sterilizing Earth Sand Bioink Composition (1), all versions of mineralization media led to precipitate already spontaneously forming in the media. Thus, when incorporating engineered UTEX 2973, that should have the ability to biocement without carbonate supplementation, we will cure the gels in pure BG-11 growth medium.

The protocol adapted from [3] for incorporating UTEX 2973 into the Earth Sand alginate gel can be found here 2025.07.29 Incorporating UTEX 2973 into Earth Sand Bioink. Prior to this experiment, we plated serial dilutions of UTEX 2973 and counted the number of colonies to convert OD measurements to UTEX 2973 cell concentration (cells/mL). Using this calibration equation, we seeded 70 million cells for 1 mL (assumed to be 1 g) of Earth Sand alginate gel ([3]). We mixed the bacteria suspension with the Earth Sand alginate gel in a microcentrifuge tube by aspirating up and down with the syringe. This proved to be challenging and ineffective in ensuring homogeneous mixing. Furthermore, incubating wild type UTEX 2973 in the Earth Sand alginate gel in a light incubator at 37C and in regular BG-11 media did not show any sign of calcium carbonate formation. Although the green colour of the cyanobacteria began to fade after five days of incubation, a live-dead assay will need to be developed to quantifiably determine the viability of UTEX 2973 in the alginate gels.

DBTL 2 - MGS-1

MGS-1 Bioink Optimization on JMP

We adapted the factors we used for the Earth Sand alginate gel fractional factorial design model, but following the issues we encountered with the MGS-1 alginate gel in Bioink Composition Testing, we wanted to explore weight % of CMC as a factor. 2025.08.08 Finalizing CMC Composition for MGS-1 Alginate Gel provided us a range of factor levels we can proceed with a similar, definitive screening design (DSD) for weight % MGS-1, weight % CMC, and weight % alginate in Table 2. Note that since we observed the sodium alginate was quite insoluble in water at 3 wt%, we set that as the maximum (highest level). Given that MGS-1 alginate gels were holding structure without submerging in calcium chloride solution in 2025.08.13 Testing Different Crosslinking Methods for MGS-1 Alginate Gels, we removed calcium chloride concentration as one of the factors for optimizing MGS-1 alginate gels. The DSD, identical to the one for earth sand bioink, was generated using JMP as shown in Figure 3.

Table 2. Factors and levels for the JMP design matrix in Figure 3.

	-1	0	1
[sodium alginate]	1wt%	2wt%	3wt%
[MGS-1]	20wt%	30wt%	40wt%
[CMC]	2wt%	3wt%	4wt%

Again, once the DSD is complete and main effect estimates are determined with experimentation, we can proceed to a response surface model (RSM). The RSM will allow us to optimize the system through multivariate analysis of factors and their responses.

Incorporating Commercial Carbonic Anhydrase in Alginate Gels

To demonstrate the MGS-1 alginate gel can self-mineralize into bricks, we incorporated commercial carbonic anhydrase (extracted from bovine erythrocytes) into the MGS-1 alginate gels. Using our hardware team’s in-house carbon dioxide chamber from dry ice, we bubble carbon dioxide into a solution of 1M calcium chloride that submerged scaffolds of MGS-1 alginate gels containing carbonic anhydrase. The experiment and protocol can be found here: 2025.09.21 Testing MGS-1 Alginate Gels With Carbonic Anhydrase.

After curing the scaffolds in calcium chloride and carbon dioxide for 24 hours, we qualitatively observed differences in appearance, touch and stiffness compared to gels that did not include carbonic anhydrase from previous experiments. For instance, the gels were stiffer and in the gels with no MGS-1, there is a significant increase in opaqueness. Under a light microscope, we can see freckles around the borders of the alginate gel, which may be calcium carbonate precipitate. This experiment however did not include any negative controls where scaffolds did not include carbonic anhydrase. Thus, the increase in stiffness may be caused by increased concentration of calcium chloride and many other variables.

More quantitative measures, such as compressibility tests, can be implemented to compare more definitely between samples. Furthermore, technologies like scanning electron microscopy would be needed to confirm that calcium carbonate precipitate has formed.

Future Directions

Since this is a novel bioink formulation, we designed many protocols and troubleshooted a variety of issues to design a working composition. To demonstrate its functionality as a bioink capable of housing and growing microbes, we need a live-dead assay to assess the viability of UTEX 2973 in the bioink. This is especially important for when we use screening experimetns and response-surface modelling, as we require quantitative readouts for our JMP model input. Our team has proposed using fluorescent dyes, such as DAPI and SYTOX Green, to stain the scaffolds containing UTEX 2973, in order to count the amount of dead cells under a fluorescent microscope. Once a live-dead assay is finalized, we can proceed with screening and fractional factorial designs for modelling bioink factor responses. This will be a monument step forward in optimizing our bioink compositions for maximizing UTEX 2973 viability, and its capabilities to produce carbonate through the MICP pathway. We did reach out to an iHP contact, listed below, that provided us some insights in developing Martian Regolith media and what is important to consider when growing microbes in Martian environments.

To further validate our bioink, we would need to explore characterization of the MICP process. Previous studies have used scanning electron microscopy to observe if calcite was precipitated ([6], [3]). Furthermore, we should do uniaxial compression testing to determine and compare the maximum compressive strengths across different compositions of MGS-1 ([3]). To quantitatively demonstrate the printability of the bioink itself, we would need to conduct rheological tests to measure the ink’s viscosity and resistance to strain ([3]). To evaluate the ability of the hydrogel to maintain its printed structure, we can proceed with classic fusion and collapse tests, which are commonly used to characterize bioinks ([7]). ‌

‌