All reagents were analytical grade. Rhodotorula mucilaginosa ATCC 66034 was kept at room temperature in Petri dishes containing YPD agar (10 g yeast extract, (MCD Lab, Estado de México, Mexico) 20 g peptone (MCD Lab, Estado de México, Mexico), 20 g glucose (Sigma Chem Co, St. Louis Mo, USA) and 20 g agar (Difco, Detroit Mi, USA). Cells were used within three weeks. For experiments, a loophole was inoculated into 10 mL of YPD (1% yeast extract, 2% peptone, 2% dextrose) or YPLac (1% yeast extract, 2% peptone, 2% lactate, pH 6.0. For YPLac, titration of pH to 6.0 with NaOH was needed to neutralize added 85% lactic acid (Meyer, CDMX, Mexico) and grown overnight. Then, an aliquot was added to 100 mL of the corresponding medium to an O.D. = 0.05 (540 nm). Note that the final concentration for each carbon source was 2%, i.e., 0.11 M dextrose or 0.23 M lactic. Flasks were incubated in a Gyratory Shaker (G10 model, New Brunswick Scientific, New Jersey, USA) at 250 rpm and 30°C for 24 hours. All experiments were performed in triplicate. Carotenoid production was inhibited by adding different concentrations of diphenylamine (DPA) (Sigma-Aldrich, Darmstadt, Germany) as described by Moliné et al. (2010). To discard any vehicle effects, we adjusted DPA concentrations in stock solutions (e.g., a 3.75 mM DPA solution was used to add 4 μL/mL and attain 15 μM DPA). We always added 4 µl EtOH/mL alone in controls to discard any effects on growth or oxygen consumption (See below).

Cells were seeded introducing a loophole from a Petri dish culture into 50 mL of either YPD (where dextrose is a fermentable carbon source) or YPLac (where lactate is a non-fermentable carbon source) (Castañeda-Tamez et al., 2024). After 24 h, cells were added to 100 mL of the corresponding medium, adjusting concentration to O.D. = 0.05 and were cultivated at 30°C. We used 250 mL Erlenmeyer flasks modified in our glass shop by attaching a Klett-test tube to the wall (Pinocchios) and cell growth was evaluated every three hours in a Klett-Summerson Model 800 colorimeter (Green filter) (Klett Manufacturing Co., New York, USA). To discard any effects of DPA on growth, samples containing 15 and 40 µM DPA were also tested ( Supplementary Figure S1 ). In addition to absorbance data, biomass wet weight was measured. Although, dry weight is probably more accurate, wet weight measurement is very straightforward and is routinely used to produce an estimate of cell mass (Uribe et al., 1985; Godbey, 2022). Cells were harvested and washed with distilled water three times at 6000 x g for 5 min at 4°C and then, samples were centrifuged at 12,000 x g for 5 min and the supernatant was discarded (Mussagy et al., 2021a). Subsequently, pellets were weighed using a Highland^® Portable Precision Balance-HCB 602H (ADAM, Oxford, USA).

To test oxidative metabolism, the rate of oxygen consumption was measured in cells harvested at 24 h (Log phase) grown in either YPD or YPLac (Purvis and Gegogeine, 2003). Respiration buffer was 10 mM 4-morpholineethanesulfonic acid (MES) pH 6.0. Cells were added to a final concentration of 12.5 mg (ww)/mL (Dejean et al., 2000). Additions were: one minute after initiating a given trace 40 µM DPA and after another minute 32 µM carbonyl cyanide 3-chlorophenylhydrazone (CCCP) (Barrientos, 2002) (see Supplementary Figure S2 ). Measurements were made using a Clark-type electrode coupled to an oximeter (StrathKelvin instruments model 782, North Lanarkshire, Scotland) equipped with a 1 mL water-jacketed chamber. Temperature was kept at 30°C with a water bath (PolyScience 7, Warrington Pa, USA). Oxygen uptake was measured as a function of time from the tangent to the initial part of the progress curve and expressed as nanoatom-grams of oxygen per minute per milligram of cells (wet weight) (natgO min^-1. mg cells (ww)^-1) (Bari et al., 2010; Nicholls and Ferguson, 2013).

Under stress, R. mucilaginosa increases carotenoid production. To evaluate this, carotenoids were extracted from 24 h cells using a microwave method as described by Mohamadi et al. (2013) with slight modifications. Briefly, cells were washed with distilled water three times at 6000 x g for 5 min at 4°C. Then, samples were centrifuged at 12000 xg for 5 min and the supernatant was discarded while the pellet (1 g (ww) mL^-1) was spread on the surface of a glass Petri dish. Each dish was treated in a microwave oven with a concave reflection system for 30 sec at 700 watts (Daewoo, Seoul, Korea). A fine pink powder was obtained and dissolved in DMSO to 50 mg dry weight per mL. The sample was sonicated for 30 min (Sonics Vibra Cell, Newtown, CT, USA) at 20 kHz, 50% amplitude with pulses of 30 sec alternated with 30 sec resting periods on ice. These were incubated under agitation for 1 h at room temperature and then cyclohexane, 5 mL/0.1 g dry weight biomass was added and further incubated for 60 min at room temp. Extraction was performed twice. At the end, the sample was centrifuged at 12000 x g for 10 min and the remaining organic phase was evaporated under a mild airflow (3 L/min) in a dark chamber until a dry powder was obtained. Each sample was solubilized in 0.2 mL 96% ethanol (Jaeschke et al., 2017) and absorbance spectra, from 400 to 600 nm (POLARstar Omega luminometer, BGM LABTECH, Allmendgrün, Germany) were taken. Carotenoid identities were annotated as in Varmira et al. (2016). To avoid interference with the torularhodin peak at 480 nm, torulene was identified by its characteristic shoulder at 530 nm instead of 490 nm, Carotene concentration was determined as in Sharma and Ghoshal (2020).

TLC was used to estimate of carotenoid composition on YPD- and YPLac-cell extracts. The stationary phase was silica gel in commercial plates (TLC silica gel 60 F254, 6x9 cm (Merck, Darmstadt, Germany). The mobile phase was ether:hexane:acetone (90:30:10, v/v/v; Meyer, CDMX, Mexico). Samples were run for 20 min at room temperature (Kanno et al., 2021). Images were taken in visible light. The distance (Rf) between the baseline and each spot was estimated, and spots were tentatively identified comparing with the literature (Zeb and Murkovic, 2010; Cheng and Yang, 2016). In an effort to further explore carotenoid identity, each band from TLC was excised and eluted in the same solvent and its absorbance spectrum was read at 400 to 600 nm in a POLARstar Omega luminometer. Detected pigments were annotated as in Moliné et al. (2012).

R. mucilaginosa survival was evaluated in a dilution spot assay of cells grown in either YPD or YPLac at 1.0 O.D. First, cells were grown in 100 mL, at 250 rpm at 30°C in the presence of 0, 15 or 40 µM DPA (Sigma, USA) (Jamieson, 1992). At 15 μM DPA carotenoid production was inhibited by 80%, while as 40 mM DPA it was inhibited by 100%. After 24 hours, 0, 15 or 40 mM menadione (MD), a free-radical producing agent was added (Sigma-Aldrich, Darmstadt, Germany) and the mixture was further incubated under agitation for 2 more hours in an orbital shaker at 250 rpm (G10, New Brunswick Sci, NJ, USA). Then, samples were collected and concentration adjusted to O.D. = 1.0. These cells were used to conduct a spot assay using a 96 well plate with 200 μL in each well. Then, performing 1/10 serial dilutions in the same medium where they grew (either YPD or YPLac) (dots, from left to right in each panel). All samples were incubated at 30°C for three days. Petri dishes were distributed as follows: Cells grown in either YPD (upper panels) or YPLac (lower panels) were divided into three groups: No additions, Medium supplemented with 15 µM DPA, and Medium supplemented with 40 µM DPA. In each panel, rows were as follows: row 1, no additions; row 2, DMSO alone; row 3, 15 mM MD and row 4, 40 mM MD.

ROS concentrations at different carotenoid concentrations were measured in cells grown in the presence of different DPA concentrations (0 to 40 µM) (Moore et al., 1989; Irazusta et al., 2013; Tang et al., 2019). In each sample, both, carotenoids and ROS were measured. The reaction buffer (0.25 M Na₃PO₄, pH 7.4) was complemented with 10 μM Amplex^® Red, hydrogen peroxide/peroxidase kit (Invitrogen, Waltham Ma, USA), 0.2 U horseradish peroxidase/mL and 0.2 U superoxide dismutase/mL (Zhou et al., 1997). Cells from each medium were harvested and washed with distilled water three times at 6000 xg for 5 minutes at 4°C and then these were aliquoted in 5 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES) pH 7.0 in a small Eppendorf tube (1.5 mL), mixed 50/50 v/v with 0.5 mm glass beads, vortexed for 3 min and solubilized with sodium deoxycholate. Protein concentration in homogenates was measured by biuret (Gornall et al., 1949). From each suspension, 100 µg protein/well was added to a POLARstar Omega luminometer (BGM LABTECH) and samples were incubated for 40 min and read against a H₂O₂ standard curve (0 to 200 nmol) made in 5 mM HEPES pH 7.0 (Guerrero-Castillo et al., 2011; Morales-García et al., 2021). Experiments were conducted in triplicate and data are reported as H₂O₂ nmol/µg protein ± SD. See 2.9 for statistical analysis. Carotenoids were measured as described above (2.4).

To evaluate if carotenoids preserved protective activity after exposure to stress, these pigments were recovered from either YPLac- or YPD-grown cells (see section 2.4) sealed under a N flow and stored in the dark at -20°C. Cells grown in either YPD or YPLac plus 15 µM DPA produced a small amount of endogenous carotenoids (20% as compared to the control). After 24 h, 40 mM menadione without or with 40 µg/g (dry weight) cell carotenoid extract (from either YPD or YLac-cultures) were added to the new cells further incubating for 2 hours at 30°C. After incubation, cells were used in a Colony Forming Unit (CFU) assay (Tran and Green, 2019; Suarez-Diez et al., 2020). Briefly, 50 µL of a 10^-5 cell dilution were added to YPD or YPLac agar plates and incubated for 3 days at 30°C. Then, CFUs were counted. Results are reported as percentage of CFUs against a control without added carotenoids and menadione (Bhuyan et al., 2023).

Statistical differences were evaluated using one-way ANOVA (Fisher, 1992). Significant differences between means were evaluated with Fischer’s multiple comparison test to p<0.05. Data analysis and graphics were constructed with GraphPad Prism for Windows, version 8.0.2 (263).

Growth curves for R. mucilaginosa ( Figure 1A ) were complemented by biomass measurements in cells cultured for 24 h ( Figure 1B ). YPD-grown cells reached 570 Klett units at 24 h (mid-Log phase) while YPLac-grown cells reached 405 Klett units. The stationary phase was reached in both cases around 40 h, reaching 810 Klett units for YPD-grown cells and 542 Klett units for YPlac-cells ( Figure 1A ). YPD-cell biomass at 24 h was 17.6 g (ww)/L while YPLac-cells weighed 8.8 g (ww)/L ( Figure 1B ). Thus, cells grown in YPD grew about 1.8 times as much as YPLac-grown cells. This is comparable to reports by others (Elsanhoty et al., 2017) and to results from other yeasts such as S. cerevisiae (De Barros et al., 2023). All experiments were conducted in cells grown until mid-Log phase (24 hours). As DPA was used in other experiments, its effect on growth was tested. Consistent with findings in the literature (Moore et al., 1989; Irazusta et al., 2013), DPA did not affect growth in R. mucilaginosa ( Supplementary Figure S1 ).

Oxygen consumption was measured in YPD- and in YPLac-cells both in basal conditions and at a maximal flow of electrons evoked by the OxPhos-uncoupler CCCP ( Figure 2 ; Supplementary Figure S2 ). DPA was also tested, and it did not have any effects ( Figure 2 ; Supplementary Figure S2 ). The basal rate of oxygen consumption for YPD-grown cells was 8 natgO min^-1. mg cells (ww)^-1 and 15 natgO min^-1. mg cells (ww)^-1 in the uncoupled state ( Figure 2 , black bars). In YPLac-grown cells the basal rate of oxygen consumption was 15 natgO min^-1. mg cells (ww)^-1 and when CCCP was added it increased to 23 natgO min^-1. mg cells (ww)^-1 ( Figure 2 ). Thus, both in YPD-grown cells ( Figure 2 , black bars) and YPLac-grown cells ( Figure 2 , gray bars), the uncoupled rate of respiration was higher than in the basal state, while DPA had no effects. In addition, in all cases the rate of oxygen consumption was higher in YPLac- than in YPD-grown cells ( Figure 2 ), indicating cells were well coupled. Recently, it was reported that in R. mucilaginosa the mitochondrial respiratory chain components vary in concentration, depending on whether the growth medium is YPLac or YPD (Castañeda-Tamez et al., 2024). The higher rate of oxygen consumption observed in YPLac-grown cells suggested that ROS increased.

R. mucilaginosa cells grown in YPLac medium were orange, while YPD-grown colonies were pale pink (See Figure 3 , “No addition” rows) suggesting that carotenoid production was higher in cells grown in YPLac medium. Thus, we decided to evaluate carotenoid concentrations by extracting them from either YPD- ( Figure 3A , black trace) or YPLac-grown cells ( Figure 3A , gray trace) and running absorbance spectra from 400 to 600 nm (Mussagy et al., 2021b). Carotenoids extracted from YPD-grown cells presented an initial absorbance nearing 0.5 units at 400 nm, and an absorbance increase reaching a peak at 490 nm. At higher wavelengths, absorbance decreased except for a shoulder at 520 nm, nearing zero at 600 nm. Under YPLac growth conditions, initial absorbance was close to 0.75 units and steadily increased until maxing out at 490 nm and decreasing at wavelengths higher that 520 nm. Notably, spectra exhibited peaks that were like those reported for Rhodotorula sp, i.e., β-carotene (λmax 450 nm), torularhodin (λmax 490 nm), and torulene (λmax 520 nm) (Park et al., 2007; Varmira et al., 2016; Udensi et al., 2022). To have a rough estimate of carotenoid concentrations in these samples, the extinction coefficient 0.16 cm^-1 M^-1 was used as in Sharma and Ghoshal, 2020 and Mussagy et al., 2021a ( Figure 3B ). In YPD samples, carotenoids were 90 µg/g cells dry weight, and in YPLac-grown samples 161 µg/g cells dry weight. The large increase in carotenoid synthesis observed in YPLac-grown cells suggested that these were under high oxidative stress.

To further characterize carotenoid production in R. mucilaginosa, a TLC assay was performed ( Figure 3C ). Extracts from either YDP- or YPLac-grown cells were included, along with a β-carotene standard (Std.). Both extracts revealed three colored bands, with bands a and b running very close to each other and to the large band in the standard (Rf= 0.92 to 0.85). A third band (c) with decreased migration was also detected in YPD- (Rf= 0.33) and in YPLac-growth extracts (Rf=0.14). All bands were scrapped from the silica plate, and their absorbance spectra were analyzed from each, YPD- ( Figure 3D , black traces) or YPLac ( Figure 3D , gray traces). Bands a and b were too close, so they were pooled together. The spectra revealed two peaks at 450 and 484 nm, suggesting the presence of a mixture of β-carotene (450 nm) and torulene (484 nm) ( Figures 3C, D , top panel) (Moliné et al., 2012). In YPLac-cells absorbance was higher than in YPD-cells. In addition, an absorbance shoulder at 520 nm was proportionally decreased only in YPLac, suggesting that carotenoid contents were different ( Figure 3D traces for bands a and b). The lower band c ran at slightly different Rfs depending on whether it came from YPD- or YPLac-grown cells. However, the deep red color and the curved shape of both bands suggested that it was the same carotenoid. This was tested running separate spectra for bands c from each, YPD- ( Figure 3D bottom spectrum, black trace) or YPLac ( Figure 3D bottom spectrum, gray trace). Spectra from bands c were almost superimposable. In addition, these exhibited a peak at 490 nm, suggested that both bands c were the same pigment, possibly torularhodin. These results, together with data from other authors, suggest that all three carotenoids usually found in R. mucilaginosa were present in extracts from both YPD- and YPLac-cells (Perrier et al., 1995; Park et al., 2007; Moliné et al., 2012; Cheng and Yang, 2016; Varmira et al., 2016; Kot et al., 2019; Tang et al., 2019; Lucas et al., 2020). Additionally, data suggest that carotenoid proportions vary with the carbon source as observed in the TLC results and spectra ( Figures 3C, D ). Still, TLC results are only suggestive, and further analyses using mass spectrometry are needed to unequivocally identify each band. It is puzzling that band c ran different distances in the YPLac or YPD-cell extracts. It is suggested that as torularhodin contains oxygen, it may be more susceptible to modification by ROS, changing slightly its structure and its affinity for the stationary phase, thus exhibiting a different Rf (Britton, 2008).

YDP-cell survival was not affected by 15 μM DPA ( Figure 4A , central panel, row 1) and only mildly by 40 μM DPA ( Figure 4A , right panel, row 1). In addition, the FR-producing agent MD inhibited growth only slightly ( Figure 4A rows 3 and 4). In contrast, in YPLac-cells ( Figure 4B ), even in the controls, growth decreased slightly as dilution increased and it was more evident at each DPA concentration ( Figure 4B all panels, rows 1 and 2). At 40 μM DPA and 15 mM MD, YPLac-cell growth was absent at all dilutions ( Figure 4B , center panel, row 4) and at 40 μM DPA both MD concentrations fully inhibited growth ( Figure 4B , right panel, rows 3 and 4). In contrast to YPD-cells, YPLac-cells were highly susceptible to MD, suggesting that they were already under oxidative stress (Biryukova et al., 2009; Tauffenberger et al., 2019). These results suggest that the carotenoid increase in YPLac-cells was due to oxidative stress.

In spite of their higher carotenoid content ( Figure 3A ), YPLac-cells were more sensitive to DPA and MD that YPD-cells ( Figure 4 ). These results suggest that even control YPLac-cells were under oxidative stress (Castañeda-Tamez et al., 2024). To test this, we decided to unmask ROS production by inhibiting carotenoid synthesis. Thus, we measured both carotenoids and ROS in the presence of 1.5 to 40 μM DPA ( Figure 5 ). Carotenoid concentrations are reported as the percentage of absorbance at 490 nm observed in the control, without DPA, which in YPLac-cells was O.D. = 1.75, while in YPD cells was O.D. = 0.95 (See Figure 3 ) (Sharma and Ghoshal, 2020; Mussagy et al., 2021a). At each DPA concentration, measurements of ROS ( Figure 5 black squares) and total carotenoids ( Figure 5 , circles) showed that DPA led to a proportional decrease in carotenoids, both in YPD- ( Figure 5A ) and in YPLac-grown cells ( Figure 5B ). In contrast, ROS concentration variations were different for either YPD or YPLac-cells. In YPD cells, ROS remained below 1.0 nmol H₂O₂/µg protein except at 40 µM DPA, a slight increase was observed, to 1.8 nmol H₂O₂/µg protein ( Figure 5A , black squares). In contrast, in YPLac-cells, ROS were already at 1.4 nmol H₂O₂/µg protein even without DPA and then, different DPA concentrations led to increased ROS, reaching 6.3 nmol H₂O₂/µg protein at 40 µM DPA. Thus, it is proposed that in YPLac-cells carotenoids increased due to high ROS concentrations and DPA unmasked these high concentrations of ROS. In contrast, YPD-grown cells did not exhibit high carotenoid production because they did not produce as much ROS ( Figure 5 ).

After carotenoids interact with free radicals, they may become inactive or even pro-oxidant (Ribeiro et al., 2018; Lucas et al., 2020). To test whether carotenoid inactivation contributed to the increased ROS susceptibility observed in YPLac-cells, we quantified the effects of adding extracted carotenoids to new cells ( Figure 6 ). YPLac-cell carotenoid extracts ( Figure 6 , vertical striped bars) or YPD-cell carotenoid extracts ( Figure 6 , horizontal striped bars) were added to cells grown with MD plus DPA in either YPD ( Figure 6 , black bars) or YPLac ( Figure 6 , gray bars). Then, Colony Forming Units (CFUs) were measured. In controls without extracted carotenoids, YPD- ( Figure 6 , black plain bar) and YPLac-grown cells ( Figure 6 , gray plain bar) produced 2.3x10⁸ and 1.2x10⁸ CFUs, respectively (See Supplementary Figure S3 ). When carotenoid extracts from YPD-cells were added back to new cell cultures, a non-significative increase in CFU numbers both in YPD- and YPLac-grown cells was observed ( Figure 6 , vertical striped bars). In contrast, adding YPLac-carotenoid extract resulted in a decrease to about half the number of CFUs both in YPD- and YPLac-cells ( Figure 6 , horizontal-striped bars). These results suggest that in YPLac-cells aerobic metabolism induced high levels of ROS reacted with carotenoids, which were inactivated or even became pro-oxidizing species themselves (Ribeiro et al., 2018; Lucas et al., 2020).