Three design approaches to engineer melanin-shielded yeast

We chose three approaches for melanin production in yeast to find an optimal strategy in terms of melanin production, the toxicity of its synthesis, and its protective effect (Fig. 1). In all three approaches, we constructed yeast strains for galactose-inducible expression of bacterial tyrosinase from Bacillus megaterium. As the simplest option, we expressed the enzyme (Tyr1 gene was codon-optimized for yeast expression) that is expected to localize to the cytoplasm (Fig. 1A). In the second approach, we aimed to concentrate the tyrosinase and melanin production into virus-like nanoparticles to minimize toxicity. To achieve that, we introduced gene encoding for viral VP1 protein, which self-assembles into nanocompartments. In addition to that, tyrosinase was fused to VP2C peptide that promotes its encapsulation into the nanocompartments by interacting with VP1 (Fig. 1B). In this way, VP1 and VP2C multimerize to form a particle so that tyrosinase is localized inside the particle. Thirdly, we expressed tyrosinase fused to a secretion signal and Aga2 protein, which is expected to direct accumulation of tyrosinase on the cell wall (Fig. 1C). In this case, melanin production takes place outside the cell, possibly decreasing the adverse effects of synthesis. Also, the fusion protein contains 4B4 peptide that binds melanin, which is necessary to create a protective shield on the cell surface.

Figure 1. Three different design approaches for melanin synthesis in engineered yeast.

Western blot confirms expression of the bacterial tyrosinase in yeast

Although protein domains are often modular, the assembly of five protein domains to form a novel protein always carries risks of poor protein expression, degradation, or proteolysis. So, it was important to confirm the expression of our engineered proteins. For this, we used western blot to detect tyrosinase with antibodies. In the Aga2-Tyr1 strategy, we added 3HA-epitope tag to the fusion protein to facilitate its detection (Fig. 2). To test the Aga2-Tyr1 fusion protein expression, we collected samples before (0 h) and 0.5 h, 1.5 h, and 24 h after galactose induction. We analyzed the samples by western blotting using antibodies against HA-epitope tag. The experiment showed that Aga2-Tyr1 is expressed at a comparable level to another protein used as a positive control, and that the tyrosinase expression is dependent on galactose (Fig. 2). This allowed us to proceed to further experiments.

Figure 2. Western blot confirms galactose-induced expression of Aga2-Tyr1-3HA. Samples were collected at the indicated time points from galactose addition. Western blot using anti-HA Epitope Tag antibodies is shown.

Pigmentation of yeast cells indicates synthesis of melanin

Having engineered these yeast strains, we set up experiments where we grew the strains in parallel cultures supplemented with either galactose (tyrosinase expression activated) or glucose (tyrosinase expression suppressed) (Fig. 3). We cultivated the cultures continuously for 72h, taking samples for different downstream experiments at different time points.

Figure 3. Experimental setup to test the melanin-producing yeast. The strains were grown in glucose or galactose-containing medium for 72h, taking samples for different measurements throughout the experiment. Samples were taken for culture optical density measurements, melanin quantification by fluorescence and UV resistance experiments.

Interestingly, at the 24h time point, we observed a significant color change in the cultures expressing Aga2-Tyr1 (cell-wall-targeted tyrosinase) but not in the other two tyrosinase-expressing strains or the background strain (DOM90) (Fig. 4A). At 72h, the color change also became apparent in the other two cultures with cytoplasmic tyrosinase (Tyr1) and viral-particle-directed tyrosinase (VP2C-Tyr1). At the same time, no color change occurred in the background strain or in cultures grown in glucose, where tyrosinase expression is suppressed (Fig. 4B). Previous studies have shown that production of melanin, which is a dark-colored pigment, also darkens the color of cells (Babaei et al., 2020; Gustavsson et al., 2016). These results show that our engineered yeast strains expressing cytoplasmic tyrosinase or cell-wall-directed tyrosinase are producing melanin.

Strikingly, the tone of the color was clearly different between the cultures with cytoplasmic tyrosinase or particle-constrained tyrosinase and Aga2-Tyr1 (Fig. 4B). There are several types of melanin with different absorbance profiles. The color difference between the strains with cytoplasmic and cell surface melanin synthesis indicates that different types of melanin are produced. Also, the observation of an earlier color change in the cells with cell-wall-anchored tyrosinase could indicate that melanin synthesis is more efficient in an extracellular environment. The pathway from tyrosine to melanin includes oxidation steps, whose rate could depend on the redox characteristics of the environment (Cao et al., 2021). This is supported by our finding of enhanced melanin synthesis in the extracellular space, which is a more oxidizing environment than the cytoplasm (Cao et al., 2021; López-Mirabal & Winther, 2008).

Figure 4. Color change in cultures with tyrosinase expression indicates melanin production. (A, B) The cultures grown at 30 ^°C in the presence of 2% galactose (tyrosinase expression ON) or 2% glucose (tyrosinase expression OFF) were imaged at 24h (panel A) and 72h (panel B) time points. (C) The cultures from 72h after galactose induction were left to sediment to divide the cells and media.

To investigate whether the synthesized melanin is attached to the cells or secreted, we pelleted the cells. The cell pellet of galactose-induced cultures was dark brown in comparison to the non-induced cultures (Fig. 4C). Importantly, all three strategies resulted in the accumulation of melanin that is attached to the cells because we did not observe any coloring of the culture supernatant.

Fluorescence measurements confirm melanin production in the engineered yeast

To gain further insights into the differences in the produced melanin forms between the three approaches, we investigated the fluorescence properties of these cells that had been cultured in the presence (tyrosinase expression ON) or absence (tyrosinase expression OFF) of galactose for 48 hours. Due to the structural diversity of melanin types and the observation that melanin fluorescence is affected by oxidation (Fernandes et al., 2016; Kayatz et al., 2001; Perna et al., 2009), we measured the fluorescence of these cells using 470 nm and 509 nm or 567 nm and 610 nm wavelengths for excitation and emission, respectively. This experiment revealed an increase in fluorescence when excited at 470 nm for cells expressing tyrosinase of all three approaches (Fig. 5). Interestingly, the cells expressing Aga2-Tyr1 produced much higher fluorescence when excited at 567 nm compared to cells expressing intracellular tyrosinase (VP2C-Tyr1 and Tyr1, Fig. 5). This could relate to different types of melanin being synthesized intra- and extracellularly, or to different oxidation states of melanin in different environments (Kayatz et al., 2001). As the fluorescence is driven by tyrosinase expression and melanin is a fluorescent compound, this experiment further supports that melanin is being produced in these cells and that the properties of the produced compounds are different between the strains using different engineering approaches.

Figure 5. Cellular fluorescence measurements indicate accumulation of melanin in all three approaches. The cells were imaged at 48h after activation of tyrosinase expression. Exemplary cells are shown.

Next, we set out to extract melanin from these cells for further characterization. For this, we collected cells at 60 hours after activating tyrosinase expression. We used a previously developed method, where melanin is solubilized during a 1-hour incubation at 80 ^°C in the presence of NaOH and DMSO (Fernandes et al., 2016). In this protocol, melanin is oxidized, followed by measurements of fluorescence using 470 nm light for excitation and 550 nm for emission.

he experiment showed high levels of melanin in the sample from Aga2-Tyr1 strain after 60h of galactose induction (Fig. 6). While the melanin fluorescence levels from induced cytoplasmic Tyr1 and particle-constrained VP2C-Tyr1 were above the level of that measured from uninduced cultures or the background strain, the difference was quite minor (Fig. 6). Melanin synthesis pathway is complex and can lead to the formation of diverse products (Cao et al., 2021; Fernandes et al., 2016) and the results of the fluorescence analysis indicate that the compound produced in Aga2-Tyr1 is melanin. At the same time, there is more uncertainty in the identity of the compounds produced intracellularly in Tyr1 and VP2C-Tyr1 strains. The differences between the three design approaches observed in these fluorescence assays and in the color changes of the cultures support our initial strategy of addressing this problem in different ways.

Figure 6. Fluorescence quantification of extracted melanin. Samples from 0h and 60h after activation of tyrosinase expression were collected and processed to extract melanin, which was then oxidized and measured by fluorescence. Plot shows the fluorescence measurements using 470 nm wavelength for excitation and 550 nm for emission.

UV resistance experiments show the shielding effect of cell-wall-localized melanin

After the detection of melanin in our yeast cells, we proceeded to test whether the melanin-containing cells had enhanced resistance to radiation. Due to safety considerations, we used UV radiation, which is easier to work with compared to ionizing radiation. We used an incubator with a UV light that can be precisely controlled to expose the culture to distinct doses of UV radiation. For the UV resistance experiments, we used cultures that were grown in media supplemented with galactose or glucose for 48 hours to accumulate sufficient amount of melanin (Fig. 3).

We irradiated the cultures with UV light for 3 and 10 minutes, made serial dilutions of the cultures, and plated these out to evaluate the survival rate of different strains. A 3-minute exposure caused around the 10-fold decrease in viable cell number compared to non-exposed cultures, as estimated by comparing the number of colonies in different dilutions in the 0 min UV exposure plate and the 3 min UV exposure plate of the uninduced cultures (Fig. 7). 10-minute exposure caused around a 100-fold decrease in viability compared to non-exposed cultures. Importantly, we observed a slightly increased survival of cells expressing Aga2-Tyr1 upon 3 and 10 minutes of UV exposure (Fig. 7). The experiments did not show any improved survival of cells expressing cytoplasmic tyrosinase or VP2C-Tyr1. The results of the UV resistance assay are in agreement with the melanin fluorescence measurements, which indicated that the product in Aga2-Tyr1 cells is indeed melanin. At the same time, the other two strains contain some other pigment or a different type of melanin. Interestingly, eumelanin, which is brown or black, has stronger radioprotective properties than pheomelanin, which is a red pigment (Brenner & Hearing, 2008; Pacelli et al., 2017). Based on the increased UV resistance and the brown color, we speculate that Aga2-Tyr1 strain produces eumelanin, while the cells expressing intracellular tyrosinase and turned red during cultivation, accumulate pheomelanin (Fig. 4, 7). The enhanced survival rate of Aga2-Tyr1-expressing cells is an important proof of concept showing that the engineered yeast cells can express melanin to protect themselves from radiation.

Figure 7. Cell viability assay shows increased survival of Aga2-Tyr1 expressing cells upon UV exposure. Yeast cultures grown for 48h in the presence of glucose (for negative control) galactose (to induce melanin synthesis) were exposed to varying times of UV light. Following the exposure, the cultures were serially diluted, the dilutions were plated and grown for 24h to evaluate cell survival rates.

Monitoring growth rate to investigate toxicity of melanin synthesis

One of our concerns while designing these strains was that the precursors of melanin synthesis can be cytotoxic (Madhusudhan et al., 2014; Sofian et al., 2014). To test this, we collected samples for cell number quantification at different time points while growing the engineered yeast cultures in the presence of galactose or glucose for activation or suppression of melanin synthesis, respectively. During a 60-hour experiment, most of the cultures reached the stationary phase with OD600 in the range of 5-9. The strains with intracellular melanin production grew very similarly to the background strain in galactose-containing media and also to the same strains in glucose-containing media, where melanin synthesis is switched off (Fig. 8). The strain expressing cell-wall-targeted Aga2-Tyr1, however, did have a slight growth defect in galactose-containing media (Fig. 8). Although this growth defect was minor, further optimization of the strain would be beneficial so that the melanin synthesis would not hinder the performance of the cell factories based on this approach for radiation protection.

Figure 8. Growth curve shows normal growth for the strains with intracellular melanin production.