Integrated transcriptomic and hormonomic insights into cadmium tolerance of a Euglena mutabilis fungal–algal–bacterial consortium

Graphical Abstract

In a fungal-algal-bacterial consortium, cadmium-exposed Euglena mutabilis remains viable by sequestering cadmium in chloroplasts, accumulating paramylon and reprogramming metal transporter, photosynthetic and hormone pathways to coordinate whole-community metal detoxification.

Abstract

Acidic, metal-contaminated environments harbour specialized microbial consortia adapted to extreme stress. We examined an environmental Euglena mutabilis culture naturally associated with Talaromyces and Acidiphilium acidophilum and exposed it to cadmium (Cd). Integrated transcriptomic, hormonal, structural and taxonomic analyses revealed a coordinated Cd-tolerance strategy. RNA sequencing showed differential regulation of metal transporters consistent with a shift from Cd uptake to intracellular sequestration. Transmission electron microscopy confirmed Cd compartmentalization within chloroplasts and increased paramylon granules. Cd exposure suppressed light-harvesting complex genes and formate/nitrite transporters while maintaining core photosynthetic function. Hormone profiling indicated strong repression of bioactive auxin and cytokinin (CK) free bases, alongside accumulation of CK nucleotides and downregulation of CK biosynthetic and activation genes. Metagenomics revealed Cd-driven enrichment of Talaromyces and Acidiphilium, implicating them in detoxification and stress support. Together, these responses highlight early Cd uptake followed by chloroplast-based detoxification, metabolic buffering via paramylon, hormonal downregulation of growth and community-mediated resilience.

Availability of data and materials

All raw sequencing data generated from Illumina sequencing and used to assess both transcriptome and metatranscriptome analyses are available from the NCBI Sequence Read Archive; accession number PRJNA1337907; with accession numbers listed in Table S1. Raw mass spectrometry hormonomic data generated in this study have been deposited in the MetaboLIGHTS repository (https://www.ebi.ac.uk/metabolights/MTBLS13259; accession number MTBLS13259. Metadata including instrument settings, sample annotations and analysis workflows are also available at this location. Software and packages used in this study are open source and listed, with versions and parameters, in Methods. No custom code was used in this study.

Introduction

Mining activities have profoundly altered landscapes worldwide, leaving behind long-lasting environmental legacies in the form of acid mine drainage (AMD), heavy metal contamination and chemically unstable tailings ponds [1, 2]. These engineered reservoirs often remain biologically degraded for decades, resisting natural recovery due to their acidic, metal-laden conditions [3–5]. Conventional strategies for remediating heavy metal-polluted environments, such as chemical neutralizing agents, passive treatment systems and biological sulphate-reducing systems, are expensive and technically demanding [6], prompting growing interest in biological approaches that leverage the adaptive capacities of extremophilic micro-organisms [7, 8].

Micro-organisms can metabolize, immobilize or transform toxic pollutants into less harmful forms [9, 10]. While monoculture systems offer reproducibility, they often underperform in complex environments due to limited metabolic flexibility [11, 12]. In contrast, co-culture systems enhance tolerance, functional breadth and stability by exploiting complementary pathways and interspecies interactions [13–15]. For instance, acidophilic algae co-cultivated with bacteria or fungi have demonstrated improved metal uptake and detoxification in AMD waters [16–18]. Such synergistic strategies position microbial consortia as a promising frontier for sustainable bioremediation of AMD-affected sites and tailing ponds.

Among extremophilic algae, Euglena mutabilis has emerged as a particularly compelling organism. This acidophilic, facultatively autotrophic protist is considered an indicator species of AMD and can be found in acidic aquatic environments such as peat bogs and volcanic lakes [19]. E. mutabilis demonstrates remarkable resilience to low pH and high metal loads and is often observed within complex microbial mats or biofilms where it coexists with acidophilic bacteria and fungi [20, 21]. These associations are thought to contribute to its survival and functional stability in extreme environments. For instance, a culture of E. mutabilis isolated from gold mine tailings in Ontario, Canada, was found to be maintained as part of a stable facultative fungal–algal–bacterial (FAB) consortium alongside Acidiphilium acidophilum and Talaromyces [20]. Disrupting these microbial partners using antibiotic and antifungal treatments resulted in a significant reduction in Cd tolerance, suggesting that interspecies interactions are essential for E. mutabilis’ enhanced stress response [20].

Despite its ecological importance, E. mutabilis remains relatively understudied at the molecular level compared to its close relative and model species, Euglena gracilis. E. gracilis has been widely explored for its metabolic plasticity, genome structure and potential in biotechnological applications [22–24]. However, recent studies demonstrate that E. mutabilis possesses substantially greater physiological resilience to Cd stress, maintaining cellular viability at concentrations that severely impair E. gracilis [25]. In Euglena, the uptake and intracellular compartmentalization of heavy metals is strongly influenced by metabolic state and environmental conditions. Under photoautotrophic growth, Cd tends to accumulate in chloroplasts, whereas in heterotrophic or dark-grown cells, Cd is more often sequestered in mitochondria [26, 27]. These redistribution patterns likely reflect dynamic detoxification strategies based on energy status. Additionally, Euglena stores carbon as paramylon, a β-1,3-glucan polymer that has been implicated in oxidative stress regulation and energy buffering [28, 29]. Paramylon accumulation may also contribute to metal stress tolerance, though its role in heavy metal detoxification remains underexplored [28, 29].

An emerging aspect of microalgal stress response involves the production and regulation of phytohormones. Although traditionally associated with plant development, signalling molecules such as auxins and cytokinins (CKs) have been detected in bacteria [30], fungi [31] and algae [32, 33], where they influence growth, metabolism and interspecies communication. In eukaryotic microalgae, hormone biosynthesis and signalling are increasingly recognized as evolutionarily conserved systems with potential roles in modulating metal stress responses [34]. However, few studies have integrated hormone profiling with gene expression data to explore these pathways – particularly in complex microbial consortia or extremophilic systems such as E. mutabilis.

To address this gap, we investigated the transcriptomic, chemical, structural and taxonomic responses of an E. mutabilis FAB culture exposed to Cd stress over a 21-day time course. RNA sequencing (RNA-seq) was used to identify differentially expressed genes (DEGs) at three timepoints, followed by functional annotation using Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Clusters of Orthologous Groups (COG) classifications. We complemented this molecular analysis with targeted quantification of extracellular phytohormones to assess their involvement in stress signalling. To explore potential shifts in community structure, we also profiled microbial taxa using Kraken2 and Bracken-based metagenomics. Finally, transmission electron microscopy (TEM) was used to assess Cd-induced ultrastructural changes in E. mutabilis cells, with a particular focus on organelle integrity and paramylon accumulation. Together, this integrative approach provides new insight into the transcriptional and physiological strategies underlying heavy metal tolerance in E. mutabilis and highlights the role of microbial partnerships and phytohormones in shaping stress resilience in mixed algal cultures.

Methods

Culture selection and growth of E. mutabilis

Field samples of E. mutabilis were sourced from the Canadian Phycological Culture Centre (CPCC) at the University of Waterloo, Canada. The strain selected (CPCC 657) comprises E. mutabilis, which maintains a seemingly symbiotic relationship with a fungus belonging to the genus Talaromyces and the bacterium A. acidophilum. Stock cultures were maintained in modified acid medium (MAM) [35], following CPCC-recommended adjustments, and the pH was set to 4.3 [36]. After autoclaving, the medium was allowed to cool before the addition of filter-sterilized F/2 vitamins. Cultures were grown autotrophically in 250 ml Pyrex Erlenmeyer flasks, sealed with foam stoppers wrapped in aluminium foil and aerated at 100 r.p.m. using a Thermo Fisher Scientific MaxQ 3000 shaker. Growth conditions included a 16:8 light-dark cycle with controlled temperature (260 µmol s⁻¹ m⁻¹; 24 ± 0.5 °C during light and 18 ± 0.5 °C during dark) using a Conviron PCG20 environmental chamber.

Pre-growth and Cd exposure with E. mutabilis

E. mutabilis cultures were exposed to CdCl₂ following the protocol outlined by Kennedy et al. [25], with modifications. In brief, 21 stock cultures were grown in MAM in 250 ml Erlenmeyer flasks, each containing 120 ml of MAM and inoculated at a density of 20,000 cells ml⁻¹. Cultures were maintained under the same environmental conditions as the original stock cultures. Cell densities were monitored on days 7 and 14 post-inoculation using a haemocytometer (Hausser Scientific) at 10× magnification. By day 14, cell concentrations exceeded 200,000 cells ml⁻¹ in all flasks. At this point, each culture was divided into two conditions: a control (MAM only) and a treatment (MAM + 100 µM CdCl₂), resulting in a total of 42 flasks. The concentration of Cd, corresponding to ~18.3 mg l⁻¹ CdCl₂ or 11.2 mg l⁻¹ Cd, was selected to represent a high-Cd stress condition relevant to severely metal-contaminated environments [37]. Each flask contained 60 ml of MAM, with or without Cd, and a concentration of 200,000 E. mutabilis cells ml⁻¹. These were maintained under the same growth conditions as the pre-treatment phase. On days 7, 14 and 21 post-treatment, seven pairs of control and treated cultures were sampled to assess growth and viability, extract RNA, quantify hormone concentrations in the supernatant and measure intracellular Cd accumulation. Timepoints for analysis were chosen based on previous work by Kennedy et al. [25], designed to expose cells to Cd over extended periods in order to assess adaptability.

Euglena cell growth and viability after Cd exposure

At 7, 14 and 21 days post-inoculation, 100 µl aliquots were taken from each flask being assessed that day and diluted with 100 µl of a 0.4% Trypan Blue solution (Bioshop Cat. No. TRY477.100) in PBS (Bioshop, Cat. No. PBS404.100). Cell counts were determined using a haemocytometer (Hausser Scientific) at 10× magnification to assess the viability of E. mutabilis visually by counting the number of cells that had taken up the Trypan Blue stain.

Intracellular Cd accumulation

Intracellular Cd accumulation was assessed in treatment flasks 7, 14 and 21 days after inoculation. Sample preparation followed the protocol outlined by Metcalfe et al. [38], with modifications. A 1.0 ml aliquot was taken from each culture flask and transferred into 1.5 ml microcentrifuge tubes. After centrifuging at 3,824 relative centrifugal force (RCF) for 10 min, the supernatant was carefully removed using a pipette. The remaining cell pellets were washed three times by resuspension in 2.0 ml of sterile deionized water (dH₂O), followed by centrifugation and supernatant removal after each wash. Following the final rinse, 300 µl of 70% (v/v) nitric acid (HNO₃; Fisher Scientific Cat. No. A509P500) was added to each sample. Tubes were then incubated on a heat block at 95 °C for 20 min to facilitate digestion. The digested samples were diluted with 10.2 ml of sterile dH₂O to achieve a final acid concentration of 2% (v/v) HNO₃. Each sample was subsequently passed through a 0.45 µm polytetrafluoroethylene syringe filter (VWR Cat. No. 76479-004) into a sterile 15 ml Falcon tube. Seven media blanks of the treatment condition were processed in parallel.

Elemental quantification was performed using an Agilent 8900 triple quadrupole inductively coupled plasma mass spectrometer at the Water Quality Centre (Trent University, Peterborough, Canada). Samples were introduced via a MicroMist nebulizer (nominal uptake: 400 µl min⁻¹) attached to a Scott-type double-pass spray chamber to optimize aerosol generation. The inductively coupled plasma mass spectrometer (ICP-MS) was operated under the General-Purpose Plasma Preset with a radio frequency power setting of 1,550 W and a carrier gas flow rate of 1.05 l min⁻¹. The sample introduction depth was fixed at 10 mm, and an X-type extraction lens was employed to enhance ion focusing. Standard nickel cones (sampler and skimmer) facilitated ion transmission to the plasma. Cadmium (Cd) was detected in single quadrupole mode using helium (He) as the collision gas. Calibration standards were made by serial dilution from 1000 ng ml⁻¹ single-element stock solutions in 2% HNO₃. Rhodium was used as an internal standard and delivered inline continuously during measurement.

Transmission electron microscopy

A 1.0 ml aliquot of cultures grown in either MAM or MAM + 100 µM CdCl₂ after 21 days was transferred into 1.5 ml microcentrifuge tubes. Following centrifugation, the supernatant was carefully removed using a pipette, and the resulting pellet was resuspended in 500 µl of 2.5% glutaraldehyde (Sigma-Aldrich, Cat. No. G5882-100ML) prepared in 0.1 M phosphate buffer (pH 7.2). Samples were fixed at room temperature for 24 h. All subsequent preparation and embedding steps were carried out by the Cellular and Molecular Electron Microscopy Core Facility at The Hospital for Sick Children (Toronto, Canada). Fixed samples were cut into 2 mm cubes and rinsed in 0.1 M sodium cacodylate buffer containing 0.2 M sucrose (pH 7.3) for 10 min. Post-fixation was performed in 1% osmium tetroxide in the same buffer for 1.5 h, followed by a second rinse in the sucrose-buffer solution. Samples were dehydrated through a graded ethanol series (70%, 90% and three changes of 100% ethanol) and then washed twice with propylene oxide. Infiltration was carried out using a 1:1 mixture of propylene oxide and Spurr resin for 1 h, followed by two incubations in pure Spurr resin (1 h and overnight). Samples were then embedded in moulds and polymerized overnight at 65 °C. Ultrathin sections (90 nm) were cut using a diamond knife on an ultramicrotome, mounted on copper grids and stained with uranyl acetate and lead citrate prior to imaging with a Hitachi HT7800 transmission electron microscope.

All solutions were prepared using ultrapure water (18.2 MΩ·cm) and trace-metal grade HNO₃ to minimize contamination.

Hormone profiling

Extraction and solid-phase purification of hormones

Supernatants from culture samples (12 ml) were harvested from each flask at 7, 14 and 21 days following inoculation and filtered through a 0.2 µm polyethersulfone syringe filter (Fisher Scientific, Cat. No. 13-1001-06) into 50 ml Falcon tubes. Media controls underwent the same processing. All processed samples were flash-frozen in liquid nitrogen, lyophilized using a LabConco FreeZone freeze dryer (Kansas City, MO, USA) and stored at −80 °C until extraction.

A sequential extraction method, adapted from previously described protocols [39, 40], was used to isolate distinct hormone classes from a single sample. This approach allowed for the recovery of 36 CKs and acidic hormones, including abscisic acid, indole-3-acetic acid (IAA), salicylic acid, jasmonic acid and five key gibberellins (GA₁, GA₄, GA₇, GA₉ and GA₂₀) (Table S1, Method S1, available in the online Supplementary Material).

RNA-seq analysis and bioinformatics

RNA was extracted from control and Cd-treated E. mutabilis cultures collected at 7, 14 and 21 days, with seven biological replicates per condition. RNA integrity was evaluated through gel electrophoresis ahead of RNA-seq (Fig. S1). Poly(A)-enriched, strand-specific libraries were sequenced using paired-end (PE) Illumina sequencing, and reads were quality filtered before de novo transcriptomic assembly. Transcript abundance was estimated by mapping reads to the assembled transcriptome, and differential gene expression (DGE) was assessed using both DESeq2 and edgeR, with shared significant genes retained for downstream analysis. Functional annotation and enrichment analyses were performed using blastx-based assignments, COG classifications and GO analysis. RNA-seq reads were also used for taxonomic profiling of the mixed culture using Kraken2 and Bracken (Method S2).

Statistical analysis

Differences in viable E. mutabilis cell counts and hormone content at each timepoint were evaluated after outlier removal and using an F-test of equality of variances, followed by Student’s t-test (P < 0.05) with appropriate consideration of variance structure.

A Mann–Whitney U-test was used to identify microbial taxa that were differentially abundant between control and Cd-treated cultures. Count data were derived from Bracken-refined Kraken2 classifications and aggregated into a sample-wise species-level count matrix. To correct for multiple testing, false discovery rate (FDR) adjustment was applied using the Benjamini–Hochberg procedure, and taxa with Padj < 0.05 were considered statistically significant. Log₂ fold changes were calculated to quantify the direction and magnitude of change in relative abundance between conditions.

Results

E. mutabilis cell viability after CdCl₂ exposure

Cd exposure significantly reduced E. mutabilis cell proliferation at all sampling times (Fig. 1). In control cultures (0 µM Cd), cell counts increased steadily from ~3.0×10⁵ cells ml⁻¹ on day 7 to 5.3×10⁵ cells ml⁻¹ by day 21 – representing a 72% increase. In contrast, cultures exposed to 100 µM CdCl₂ consistently exhibited lower cell densities at all timepoints. By day 7, Cd-treated cultures showed a significant reduction in cell counts compared to controls (P < 0.01). This trend persisted through day 14 (P<0.001) and day 21, where the difference became more pronounced (P < 0.001), despite continued growth in the treated group – representing a more modest 58% increase.

Fig. 1. Assessment of Cd tolerance of E. mutabilis. Growth of E. mutabilis over 21 days in 100 µM CdCl₂ revealed significant decreases in cell counts compared to E. mutabilis cells grown in MAM only. Error bars represent sd between biological replicates (n=7), and asterisks represent significant differences between 0 and 100 µM Cd on each day (*P<0.05; **P<0.01; ***P<0.001) as determined by an F-test for variance and Student’s t-test.

Despite reduced proliferation, cell viability remained high in all treatments. Trypan Blue staining revealed minimal E. mutabilis cell death across timepoints, with Cd-treated cultures exhibiting only a slight increase in the proportion of non-viable cells over 21 days ([number of dead cells/total cell count] x 100%; Table 1).

Table 1. Comparison of non-viable cells in control and Cd-treated cells after 21 days of growth (n = 7).

Dead cell frequency was assessed by staining with Trypan Blue and counting on a haemocytometer at 10× magnification. The frequency of dead cells was determined using the following formula: (# dead cells/total cell count) * 100%=frequency of dead cells (%).

	Day 7		Day 14		Day 21
	Average total cell counts (cell ml−1)	Frequency of dead cells (%)	Average total cell counts (cell ml−1)	Frequency of dead cells (%)	Average total cell counts (cell ml−1)	Frequency of dead cells (%)
0 µM Cd	302,678	0	465,104	0	538,988	0
100 µM Cd	237,648	0.69	318,228	2.25	379,910	1.75

Intracellular Cd accumulation

Following exposure to 100 µM Cd, intracellular Cd concentrations in E. mutabilis increased steadily over the 21-day period (Fig. 2). On day 7, the average intracellular Cd content was 161 ng/10⁵ cells, which rose to 204 ng on day 14 and 322 ng by day 21. When scaled to the total number of cells per 60 ml culture, this corresponded to ~22,800, 37,900 and 72,000 ng of total intracellular Cd on days 7, 14 and 21, respectively. Relative to the initial 674,000 ng dose, these values represent ~3%, 6% and 11% uptake, indicating progressive accumulation by E. mutabilis. However, direct ICP-MS measurements of the treatment culture media (MAM + 100 µM Cd with no cells) showed only 500,000 ng remaining after 7 days, 525,000 ng after 14 days and 491,000 ng by day 21, which increases the estimated sequestration efficiency to 5%, 7% and 15%, respectively. This indicates that a fraction of Cd was lost through abiotic processes. Cd ions are prone to adsorption to glassware surfaces [41] and forming insoluble precipitates within media components such as phosphate and sulphate [42], which would reduce the dissolved fraction available for detection and may explain the discrepancy between the amount of Cd introduced and the lower concentrations quantified.

Fig. 2. Intracellular Cd accumulation increases progressively over the 21 day exposure to Cd. Each light green dot represents an individual biological replicate (n=7), while the dark green point represents the average ±sd (black error bars). The dotted line illustrates the overall trend of increasing average Cd accumulation over time (R²=0.93; y=11.5x + 67.8).

TEM revealed distinct ultrastructural differences between E. mutabilis cells grown in control conditions and those exposed to Cd (Fig. 3). In control cells (MAM) grown for 21 days, organelles including chloroplasts, Golgi bodies, mitochondria, the nucleus and vacuoles appeared intact, with paramylon granules and a clearly defined pellicle. In contrast, Cd-treated cells displayed notable structural alterations, including the accumulation of electron-dense deposits within chloroplasts which appear limited in control cells. Additionally, Cd-treated cells exhibited increased vacuolization and a higher abundance of paramylon bodies. The appearance of these electron-dense inclusions correlates with the observed increase in intracellular Cd accumulation over time; however, these deposits were not quantified because the TEM analysis was designed as a qualitative ultrastructural assessment rather than a systematic particle-counting assay. Future work should consider quantifying the electron-dense deposits to reinforce the correlation between intracellular metal uptake data and visual assessments.

Fig. 3. TEM of E. mutabilis cells points to an elevated presence of paramylon and Cd accumulation after 21 days of growth. Representative cross-sections of E. mutabilis grown in (a) MAM and (b) MAM + 100 µM CdCl₂ display chloroplasts (c), Golgi bodies (g), mitochondria (m), the nucleus (n), vacuoles (v), the pellicle (black arrows), paramylon granules (white arrows) and electron-dense bodies within the chloroplast (red arrows). While not measured, paramylon granules appeared more numerous in Cd-treated cells. The scale bar on each image is 2 µm.

Hormone profiles

Identification and quantification of hormones in E. mutabilis culture supernatants revealed clear differences between control and Cd-treated conditions over the 21-day experimental period (Table 2). Of the hormones screened for, seven were detected, six of which were CKs (including one methylthiolated CK) and the other was IAA. No hormones were detected in culture media blanks.

Table 2. Concentrations of hormones detected in the culture supernatant of E. mutabilis under control and Cd-treated conditions over 21 days.

Hormone levels are reported as mean±standard error (SE) in picomolar per 100,000 cells (pmol/100,000 cells) or femtomolar per 100,000 cells (fmol/100,000 cells). CKs are grouped into free bases, ribosides and nucleotides, with total concentrations provided for each class where applicable. A dagger (†) indicates significant difference in hormone concentration between control and Cd-treated cultures at each timepoint using an F-test to assess variance followed by Student’s t-test (n=7).

Supernatant (pmol/100,000 cells±SE)		0 µM control			100 µM Cd
	Compound	Day 7	Day 14	Day 21	Day 7	Day 14	Day 21
	IAA	1022.1±132.7	665.2±86.4	n.d.	n.d.	n.d.	n.d.
Supernatant (fmol/100,000 cells±SE)		0 µM control			100 µM Cd
	Compound	Day 7	Day 14	Day 21	Day 7	Day 14	Day 21
Free bases	cZ	32.8±8.5	146.6±7.8	761.0±41.5	507.9±59.6†	n.d.	n.d.
	iP	11.0±1.5	18.1±0.8	10.2±0.6	13.9±3.2†	n.d.	n.d.
Total		43.7±10.0	164.7±8.7	771.3±42.0	521.8±62.8	0	0
Ribosides	cZR	8.8±1.8	25.1±3.5	29.2±5.0	20.3±2.6	n.d.	n.d.
Nucleotides	cZNT	125.7±6.0	308.0±5.8	290.0±18.0	301.7±11.4†	325.9±8.0†	235.6±4.0†
	iPNT	n.d.	n.d.	n.d.	n.d.	495.1±27.9†	991.9±78.3†
Total		125.7±6.0	308.0±5.8	290.0±18.0	301.7±11.4	821.0±35.9	1,227.4±82.3
Methylthiols	MeSZR	1689.1±50.5	1099.3±32.9	1164.2±31.0	641.5±6.1†	778.5±13.0†	487.3±18.1†

CK free bases showed distinct patterns between control and Cd-treated conditions. In control cultures, cZ rose steadily from 33 fmol/100,000 Euglena cells on day 7 to 147 on day 14, reaching 761 fmol/100,000 cells by day 21. In contrast, cZ was only detectable in Cd-treated cultures at day 7, in significantly lower concentrations than the control, with levels falling below detection thereafter. A similar trend was observed for iP, which peaked in control conditions on day 14 (18.1 ± 0.8) but was significantly lower in Cd-treated samples on day 7 and undetectable at subsequent timepoints. The total concentration of CK free bases in control samples increased nearly 18-fold over 21 days of growth, while Cd-exposed cultures exhibited a sharp decline in free base detection after day 7, resulting in no detectable free bases at 21 days.

Riboside forms of CKs, such as cZR, remained consistently detectable in control cultures. Like cZ, it followed a similar increasing trend over 21 days, peaking at 30 fmol/100,000 Euglena cells (Table 2). In Cd-treated cultures, however, cZR was observed only on day 7. Among nucleotide forms, cZNT remained detectable throughout all timepoints in both control and treatment conditions. In control cultures, cZNT peaked on day 14 (308.0 ± 5.8 fmol/100,000 Euglena cells). Cd-treated samples also showed their highest cZNT concentration on day 14 (325.9 ± 8.0 fmol/100,000 Euglena cells), which was significantly higher than the control, although by day 21, the cZNT concentration was significantly higher in the control. By contrast, iPNT was not detected in control cultures but was present in Cd-treated samples on both day 14 (495.1 ± 27.9 fmol/100,000 Euglena cells) and day 21 (991.9 ± 78.3 fmol/100,000 Euglena cells). Overall, total nucleotide concentrations were consistently higher in Cd-treated cultures compared to controls at all timepoints.

IAA was exclusively detected in control cultures and declined over time, with concentrations decreasing from 1022.1 ± 132.7 pmol/100,000 cells on day 7 to 665.2 ± 86.4 pmol/100,000 Euglena cells by day 14. IAA was not detected in any Cd-treated cultures at any timepoint.

RNA-seq and de novo transcriptome assembly

RNA-seq was performed on control and Cd-treated E. mutabilis cultures after 7, 14 and 21 days of exposure, with seven biological replicates per condition (Table S2). Sequencing generated an average of 65.1 million PE reads per sample, of which 64.7 million PE reads remained after quality filtering (Table S3). A de novo Trinity transcriptome assembly was generated from normalized reads and used as a reference for downstream analysis; more than 90% of reads mapped back to the assembly (Table S4). After filtering low-abundance transcripts and removing one outlier control sample with its paired Cd-treated sample, 78,302 genes were retained for differential expression analysis (Fig. S2).

Comparative transcriptome analysis of E. mutabilis grown in the absence and presence of Cd

To distinguish Cd-responsive genes from genes that changed over time, DGE analyses were organized into two comparison types: (i) treatment comparisons, in which Cd-treated cultures were compared with control cultures within each sampling day, and (ii) temporal comparisons, in which days 7, 14 and 21 were compared within each treatment group.

DGE analysis was conducted using both DESeq2 and edgeR, with DEGs retained for downstream interpretation after adjusted P-value and fold-change filtering (Table 3, Figs S3 and S4). For treatment comparisons, Cd-treated cultures were compared with control cultures separately at days 7, 14 and 21. These comparisons revealed time-dependent transcriptional responses, with 480 unique DEGs detected on day 7, 352 on day 14 and 751 on day 21, while 427 DEGs were shared across all three timepoints (Fig. 4a, Table S5). Cd-treated cultures showed 676 upregulated and 416 downregulated genes on day 7, 680 upregulated and 336 downregulated genes on day 14 and 1,003 upregulated and 469 downregulated genes on day 21 (Fig. 4b–d). Separately, temporal comparisons were conducted within each treatment group to identify genes that changed between days 7, 14 and 21 under control conditions or during prolonged Cd exposure (Table 3).

Table 3. Summary of DEGs identified by DESeq2 and edgeR across treatment and temporal comparisons.

Treatment comparisons compare control and Cd-treated cultures within each sampling day, while temporal comparisons compare days 7, 14 and 21 within each treatment group. The number of DEGs with an adjusted P-value (Padj) <0.05 and those with both Padj <0.05 and |log₂ fold change|>1 are reported for each method. The final two columns show the overlap in DEGs identified by both DESeq2 and edgeR using the same thresholds.

	DESeq2		edgeR		Overlap of DESeq2 and edgeR
Comparison	Padj <0.05	Padj <0.05; |log2FC|>1	Padj <0.05	Padj <0.05; |log2FC|>1	Padj <0.05	Padj <0.05; |log2FC|>1
Control vs Cd day 7	7,503	1,980	3,578	1,880	2,700	1,092
Control vs Cd day 14	8,202	2,378	4,170	1,745	2,893	1,016
Control vs Cd day 21	12,981	2,908	4,087	2,286	3,192	1,472
Control day 7 vs 14	13,832	5,304	66	65	23	23
Control day 7 vs 21	10,877	846	122	121	42	42
Control day 14 vs 21	16,457	3,758	101	101	31	31
Cd day 7 vs 14	10,540	3,239	143	142	33	32
Cd day 7 vs 21	15,481	1,016	32	30	6	6
Cd day 14 vs 21	12,191	1,690	14,056	11,597	5,212	882

Fig. 4. DGE between control and Cd-treated E. mutabilis at days 7, 14 and 21. A Euler diagram (a) displays the number of unique and shared DEGs between control and Cd-treated cultures after 7 (red), 14 (yellow) and 21 (blue) days growing in the presence of Cd. Volcano plots comparing the log₂ fold change (x-axis) and -log₁₀ adjusted p-value (y-axis) from Cd-treated versus control cultures at days 7 (b), 14 (c), and 21 (d) illustrate the distribution of DEGs that were significantly (Padj <0.05; |log₂ fold change|>1) upregulated (red) and downregulated (blue).

Because a complete annotated genome is unavailable for E. mutabilis, functional annotation was performed using blastx searches against available protein databases, followed by COG and GO term analyses. Most DEGs lacked functional assignment (Table S6), highlighting the limited annotation resources available for this organism and the mixed-culture transcriptome. Among annotated DEGs, the predominant Cd-responsive categories were inorganic ion transport and metabolism, signal transduction, posttranslational modification/protein turnover/chaperones and energy production and conversion. Formate/nitrite transporter genes were generally downregulated, whereas several zinc transporter, cation efflux and metal-homeostasis genes were upregulated. Photosynthesis-related responses were mixed: light-harvesting complex (LHC) genes were downregulated, while core photosynthetic components, including ATP synthase and photosystem II (PSII)-related genes, were upregulated (Table S6).

Temporal comparisons within Cd-treated cultures revealed additional transcriptional remodelling, particularly between days 14 and 21, when 931 COG classifications corresponding to 882 DEGs were identified (Table S7). These DEGs were mainly associated with signal transduction, posttranslational modification/protein turnover/chaperones, RNA processing and modification, lipid transport and metabolism and inorganic ion transport and metabolism. Upregulated genes included those associated with kinases and cyclases, RNA recognition and processing, ubiquitin-mediated protein regulation, lipid metabolism, zinc transport, cation efflux and ABC phosphate transport, while formate/nitrite transporters were predominantly downregulated (Table S7).

DEGs were further assessed for GO enrichment using the PANTHER classification system. For treatment comparisons, GO enrichment was assessed using DEGs from Cd-treated cultures compared with control cultures within each sampling day. Across these Cd-versus-control comparisons, 369 significant annotated genes (FDR<0.05) were assigned GO terms and species information, including 105 on day 7, 115 on day 14 and 149 on day 21. These genes represented 39 unique GO terms related primarily to photosynthesis, chloroplast and thylakoid components, membrane transport, translation and ATP synthase activity (Table 4, Fig. S5, Table S8).

Table 4. GO terms enriched in treatment comparisons between control and Cd-treated cultures.

Thirty-nine statistically enriched (FDR<0.05) GO terms were identified from the list of DEGs using the PANTHER classification system when E. mutabilis cultures grown in MAM were compared to cultures grown in MAM + 100 µM Cd. Databases for GO term identification include Homo sapiens, Arabidopsis thaliana, Chlamydomonas reinhardtii and Synechocystis sp. in relation to COGs identified using blastx.

Ontology	GO term	Gene count	Average FDR	COG classification
Biological process	Photosynthesis, light harvesting (GO:0009765)	178	7.09×10−9	C, P, unclassified
	Photosynthesis (GO:0015979)	22	6.13×10−6	S, unclassified
	Translation (GO:0006412)	19	8.36×10−3	J, unclassified
	Cyclic nucleotide biosynthetic process (GO:0009190)	8	3.68×10−3	P, T, unclassified
	Photosynthesis, light reaction (GO:0019684)	6	4.58×10−7	C
	Nucleotide metabolic process (GO:0009117)	4	2.12×10−2	F
	Response to light stimulus (GO:0009416)	3	1.91×10−8	BDLTU
	Translational elongation (GO:0006414)	1	3.22×10−5	Unclassified
	Proton transmembrane transport (GO:1902600)	1	4.74×10−2	Unclassified
Cellular component	Cytoplasm (GO:0005737)	276	6.98×10−4	E, F, J, N, O, Q, S, T, UY, V, Z, unclassified
	Chloroplast (GO:0009507)	228	1.47×10−2	C, E, J, P, S, unclassified
	Membrane (GO:0016020)	218	2.45×10−3	BDLTU, C, DKT, E, J, P, Q, S, V, Z, unclassified
	Photosystem I (GO:0009522)	95	4.04×10−9	C, P, S, unclassified
	Chloroplast thylakoid membrane (GO:0009535)	50	2.43×10−6	C, S, unclassified
	Cytosol (GO:0005829)	48	3.86×10−2	B, E, F, G, H, O, T, Z, unclassified
	Plasma membrane (GO:0005886)	41	1.28×10−4	P, PQ, Q, T, U, V, Z, unclassified
	Extracellular region (GO:0005576)	14	7.39×10−3	E, O, Q, Z, unclassified
	Plastid (GO:0009536)	10	4.00×10−3	E, G, unclassified
	Ribosome (GO:0005840)	9	5.94×10−3	J, unclassified
	Photosystem (GO:0009521)	6	6.10×10−3	C
	Thylakoid (GO:0009579)	6	5.60×10−3	S
	Small ribosomal subunit (GO:0015935)	5	1.93×10−2	J, unclassified
	PSII (GO:0009523)	5	4.35×10−9	C, unclassified
	Extracellular exosome (GO:0070062)	5	1.34×10−4	E, unclassified
	Protein-containing complex (GO:0032991)	4	8.74×10−3	Unclassified
	Plastoglobule (GO:0010287)	3	1.23×10−2	S
	Thylakoid membrane (GO:0042651)	3	4.43×10−3	Unclassified
	Extracellular space (GO:0005615)	2	3.91×10−3	Unclassified
	Mitochondrial nucleoid (GO:0042645)	1	3.75×10−2	Unclassified
	Cytosolic ribosome (GO:0022626)	1	3.70×10−2	Unclassified
	Chloroplast envelope (GO:0009941)	1	1.37×10−2	E
Molecular function	Transmembrane transporter activity (GO:0022857)	46	4.38×10−2	E, P, unclassified
	ABC-type transporter activity (GO:0140359)	21	3.39×10−2	Q, V, unclassified
	Translation elongation factor activity (GO:0003746)	15	7.79×10−7	J, unclassified
	Proton-transporting ATP synthase activity, rotational mechanism (GO:0046933)	9	6.32×10−3	C, unclassified
	Chlorophyll binding (GO:0016168)	6	2.50×10−6	C
	Hydrolase activity (GO:0016787)	4	1.41×10−2	T, unclassified
	Formate transmembrane transporter activity (GO:0015499)	2	5.85×10−4	P
	Nucleotide binding (GO:0000166)	1	1.45×10−3	Unclassified

Temporal GO comparisons were assessed separately within Cd-treated cultures to identify functional changes during prolonged Cd exposure. In these within-Cd temporal comparisons, 124 significant annotated genes were detected only for the day 14 versus day 21 comparison, representing 24 GO terms consistent with continued cellular remodelling during prolonged Cd exposure (Table 5, Fig. S6, Table S9). No significant GO-annotated genes were detected for the day 7 versus 14 or day 7 versus day 21 comparisons within Cd-treated cultures.

Table 5. GO terms enriched in the temporal comparison between day 14 and day 21 Cd-treated cultures.

Twenty-four statistically enriched (FDR<0.05) GO terms were identified from the list of DEGs using the PANTHER classification system when E. mutabilis cultures grown in 100 µM Cd for 14 days were compared to cultures in 100 µM Cd for 21 days. Databases for GO term identification include Homo sapiens, Arabidopsis thaliana, Chlamydomonas reinhardtii and Synechocystis sp. in relation to COGs identified using blastx.

Ontology	GO term	Gene count	Average FDR	COG classification
Biological process	Cyclic nucleotide biosynthetic process (GO:0009190)	9	3.68×10−3	C, T
	Photosynthesis, light harvesting (GO:0009765)	3	7.09×10−9	S, unknown
	Translation (GO:0006412)	2	8.36×10−3	J, unknown
	Response to light stimulus (GO:0009416)	1	1.91×10−8	LT
Cellular component	Cytoplasm (GO:0005737)	23	6.98×10−4	C, H, I, J, L, LT, O, Q, S, T, unknown
	Membrane (GO:0016020)	18	2.45×10−3	DO, EG, I, J, P, Q, S, T, unknown
	Plasma membrane (GO:0005886)	16	1.28×10−4	C, O, P, PQ, S, T, U, unknown
	Cytosol (GO:0005829)	12	3.86×10−2	DO, F, G, H, J, O, S, Z, unknown
	Extracellular region (GO:0005576)	8	7.39×10−3	E, G, I, O, Q, S, unknown
	Chloroplast (GO:0009507)	6	1.47×10−2	I, O, S, unknown
	Photosystem I (GO:0009522)	3	4.04×10−9	S, unknown
	Ribosome (GO:0005840)	2	5.94×10−3	J, unknown
	PSII (GO:0009523)	2	4.35×10−9	S, unknown
	Chloroplast thylakoid membrane (GO:0009535)	2	2.43×10−6	S, unknown
	Plastoglobule (GO:0010287)	2	1.23×10−2	I, S
	Extracellular exosome (GO:0070062)	2	1.34×10−4	C, G
	Extracellular space (GO:0005615)	1	3.91×10−3	O
	Chloroplast envelope (GO:0009941)	1	1.37×10−2	S
	Protein-containing complex (GO:0032991)	1	8.74×10−3	S
Molecular function	Nucleotide binding (GO:0000166)	3	1.45×10−3	H, LT, unknown
	Hydrolase activity (GO:0016787)	3	1.41×10−2	F, I, M
	Chlorophyll binding (GO:0016168)	2	2.50×10−6	S, unknown
	Catalytic activity (GO:0003824)	1	3.97×10−3	I
	Transmembrane transporter activity (GO:0022857)	1	4.38×10−2	P

Integrating analysis of transcripts and hormones

Annotated genes associated with phytohormone biosynthesis were identified using keyword searches of SPROT blastx results, transcript predicted functions and KEGG identifiers (Table S10). Hormone-related DEGs were then evaluated using the same comparison structure as the broader DEG analysis. For treatment comparisons, Cd-treated cultures were compared with control cultures within each sampling day. This search identified 357 unique genes consistently present across days 7, 14 and 21. Of these, 34, 46 and 68 genes were deemed statistically significant at days 7, 14 and 21, respectively, while only 6, 5 and 4 genes exceeded the differential expression threshold at the corresponding timepoints (Padj<0.05; |log₂ fold change|>1; Table S11).

Across the 21-day exposure period, treatment comparisons showed modest but significant changes in genes associated with auxin and CK pathways. Several tryptophan-dependent auxin biosynthesis genes showed differential expression, including annotations related to tryptophan synthase, aminotransferase activity and flavin monooxygenase-like enzymes. However, these transcript patterns did not correspond with extracellular IAA profiles: IAA declined over time in control culture conditions and was undetectable in Cd-treated cultures (Table 2). This suggests that auxin-related transcriptional changes were not reflected in detectable extracellular IAA accumulation. This disconnect may reflect regulation beyond transcript abundance, including post-transcriptional control, altered enzyme activity, intracellular retention or degradation.

CK-related gene expression aligned more clearly with metabolite profiles. Genes annotated as tRNA isopentenyltransferase (IPT) (K00591), CK riboside 5′-monophosphate phosphoribohydrolase (LOG; K00588), homocysteine S-methyltransferases and O-methyltransferases were differentially expressed, particularly at days 14 and 21 (Table S12). Although fold changes were generally modest, these patterns were consistent with altered CK metabolism under Cd exposure. In parallel, CK nucleotide levels, including cZNT and iPNT, increased in Cd-treated cultures, while bioactive CKs and methylthiolated CKs remained low or undetectable by day 21 (Table 2).

Temporal comparisons were assessed separately within Cd-treated cultures to evaluate hormone-related transcriptional remodelling during prolonged Cd exposure (Table S12). In these within-Cd temporal comparisons, 99 hormone-related genes were significant between days 7 and 14, 193 between days 7 and 21 and 115 between days 14 and 21. Most changes were moderate in magnitude, but genes associated with tryptophan metabolism, putative TAA1/YUC-like auxin pathways, methyltransferase activity, LOG and IPT were among those identified. Together, these results suggest persistent, low-magnitude modulation of auxin- and CK-related pathways during prolonged Cd exposure.

Metagenomic analysis

Kraken2 classified an average of 15 and 14% of reads at the genus level in control and Cd-treated samples, respectively. At the species level, the average classification was 14% in both control samples and Cd-treated samples. Bracken refinement assigned most classified reads to E. gracilis, likely reflecting the limited availability of reference sequences for E. mutabilis (99.2%; Fig. 5).

Fig. 5. Relative abundance of the top ten most abundant species identified in metatranscriptomic samples across timepoints and treatment conditions using Kraken2 classification with Bracken refinement. (a) Full-scale relative abundance showing that E. gracilis overwhelmingly dominated all samples across control and Cd-treated cultures at days 7, 14 and 21. (b) Expanded view (0–1.5% scale) highlighting low-abundance taxa. The analysis was performed at the species level, and values represent the estimated fraction of reads assigned to each species normalized by Bracken. Bars represent the average relative abundance across replicates (n=7) for each condition.

Differential abundance analysis identified gene sequences whose closest matches in GenBank differed between conditions, without confirming that the corresponding organisms were present (Table S13, Fig. S7). Under Cd exposure, sequences most similar to Talaromyces amestolkiae and Exophiala oligosperma were strongly enriched (log₂ fold change>8; FDR-adjusted P-values < 0.001). In contrast, a separate set of bacterial-origin genes, most similar to Corynebacterium simulans, was more abundant in control samples (log₂ fold change=–2.32; Padj=0.01).

Discussion

This study highlights the stress resilience of E. mutabilis and its associated microbial partners under Cd exposure, revealing a coordinated response that spans molecular, physiological and community levels. Rather than simply tolerating toxicity, the FAB consortium actively modulates gene expression, hormone biosynthesis and ultrastructural organization to manage heavy metal stress, underscoring the adaptive capacity of co-cultured extremophiles in severe environments.

Cd tolerance and accumulation of E. mutabilis

Cd is a non-essential and toxic metal for microalgae that can disrupt photosynthetic processes and respiration [43–46]. In this study, E. mutabilis showed strong resilience to 100 µM Cd, maintaining high cell densities and over 97% viability after 21 days. This tolerance exceeded that reported for several freshwater microalgae, including Chlorella sp. Ankistrodesmus sp. and Scenedesmus, which show reduced growth or viability at comparable or lower Cd concentrations [47]. It also contrasts the related euglenoid E. gracilis, which is more sensitive to Cd under autotrophic conditions [25, 26, 48, 49].

This tolerance was accompanied by sustained intracellular Cd accumulation. By day 21, ~15% of total Cd added to the culture was internalized, exceeding values reported for Chlorella vulgaris [50] and Scenedesmus obliquus [51], where most Cd remains surface-associated after short exposures. Unlike diatoms such as Thalassiosira weissflogii, in which intracellular Cd accumulation can plateau within hours [52], E. mutabilis continued accumulating Cd over 3 weeks. This sustained accumulation, together with high viability, indicates that E. mutabilis can tolerate internal Cd loads that are inhibitory to many other algae. Intracellular sequestration is a major detoxification mechanism in algae because it prevents free Cd²⁺ from disrupting photosynthesis and metabolism [53].

TEM observations were consistent with intracellular compartmentalization, revealing electron-dense deposits within chloroplasts of Cd-treated cells. These deposits are consistent with Cd sequestration in algae [54, 55], and chloroplast-associated Cd accumulation has also been observed in autotrophic E. gracilis [56]. However, unlike E. gracilis, which often shows substantial Cd-induced growth inhibition within days [26, 49], E. mutabilis maintained high viability while intracellular Cd increased over the full exposure period. TEM also suggested increased paramylon granules in Cd-treated cells, although this was not quantified. Similar paramylon accumulation has been reported in E. gracilis exposed to Cd [28], chromium (Cr) [57, 58] and osmotic stress [29, 59], supporting the interpretation that carbon storage may be part of the stress response of Euglena [60, 61].

Transcriptomic patterns were consistent with these physiological and ultrastructural responses. Cd exposure altered genes associated with ion transport and metal homeostasis, including zinc transporters, cation efflux proteins and formate/nitrite transporters. Similar transport families, including cation diffusion facilitators (CDFs), metal tolerance proteins, ABC transporters and P-type ATPases, contribute to detoxification in green microalgae such as Chlamydomonas reinhardtii and Auxenochlorella protothecoides [62–64]. In E. mutabilis, the mixed expression of transporter-related genes suggests that Cd tolerance involves coordinated regulation of ion transport rather than uniform activation of a single transporter class.

Cd exposure also affected photosynthesis-related genes. ATP synthase subunits and PSII components were upregulated, whereas LHC genes were downregulated. Similar repression of light-harvesting genes under metal stress has been reported in C. reinhardtii and may reduce photooxidative damage when photosystem integrity is compromised [65]. Enrichment of these terms related to catalytic activity and carbohydrate metabolism, together with the TEM observation of increased paramylon, further supports metabolic adjustment during Cd exposure (Fig. 3; GO) [59].

In contrast to these interpretable functional categories, a substantial proportion of DEGs across all timepoints corresponded to genes of unknown functions. While these genes cannot currently be assigned clear mechanistic roles, their prevalence underscores the limited functional annotation available for E. mutabilis and suggests that poorly characterized or lineage-specific responses may also contribute to Cd tolerance.

The high Cd tolerance of E. mutabilis may also be influenced by its microbial partners within the FAB consortium. Previous work showed that disrupting this consortium through antifungal or antibiotic treatments reduced Cd tolerance, suggesting that cohabiting microbes contribute to stress adaptation [20]. In the present study, the combined phenotype of high viability, sustained Cd accumulation and transcriptional remodelling is consistent with a consortium-supported response. Acidophilic fungi and bacteria can influence metal tolerance through metal binding, sequestration and changes to local chemical conditions [66–69], and fungal–bacterial biofilms have been shown to reduce bioavailable Cd, Pb and Zn more effectively than individual strains [70]. Although the transcriptomic data were dominated by E. mutabilis, these findings support the view that the FAB consortium provides environmental buffering and functional complementarity.

Hormones as regulators of Cd toxicity

Cytokinins

Transcriptome and hormone analyses indicate that Cd exposure strongly alters CK metabolism in the E. mutabilis FAB consortium. The near-complete suppression of CK free bases and ribosides – the forms directly involved in active signalling [71] – together with reduced expression of CK biosynthetic and activation genes such as IPT and LOG, suggests reduced CK signalling under Cd stress. This pattern is consistent with the observed decrease in growth and supports a shift away from growth-promoting metabolism during prolonged metal exposure.

In Euglena, CKs are primarily synthesized via the tRNA degradation pathway, in which IPT enzymes modify tRNAs that later release CK nucleotides upon degradation [72, 73]. LOG enzymes then convert CK nucleotides into active free base CKs [74], while S-adenosylmethionine (SAM)-dependent methylation contributes to the production of methylthiolated CK derivatives [75, 76]. Under Cd exposure, the accumulation of CK nucleotides alongside reduced CK free bases and ribosides suggests that CK metabolism was not completely halted, but that conversion to active signalling forms was limited. Changes in genes associated with SAM metabolism and methyltransferase activity further indicate altered methylthiolated CK metabolism under stress [75, 76].

Temporal comparisons within Cd-treated cultures showed modest upregulation of genes encoding LOG enzymes and SAM-dependent methyltransferases between days 14 and 21. Rather than indicating a simple, sustained shutdown of CK metabolism, this pattern suggests time-dependent regulation of the CK pathway during acclimation to Cd. Similar stress-associated regulation of LOG genes has been reported in higher plants, where altered LOG expression affects growth recovery and stress tolerance under abiotic stress [77, 78]. SAM metabolism is also linked to methylation reactions involved in stress responses and hormone metabolism [79]. This response may help limit growth-associated resource demands while maintaining the capacity for later hormonal reactivation during acclimation or recovery [80].

Auxins

Extracellular IAA was undetectable in Cd-exposed cultures, despite modest expression changes in genes annotated near the tryptophan-dependent IAA biosynthesis pathway. These included putative tryptophan synthase, aminotransferase-like and flavin monooxygenase-like genes [81]. However, because these annotations were broad and most fold changes were small, the transcriptomic evidence for direct activation of auxin biosynthesis remains inconclusive. The absence of extracellular IAA may therefore reflect reduced IAA production, intracellular retention, rapid metabolism or diversion of tryptophan into other stress-related pathways.

Reduced auxin accumulation has been associated with growth suppression under abiotic stress in other systems. In Pisum sativum, Cd and Cu exposure increased IAA oxidase activity, lowering auxin concentrations and reducing growth [82]. In S. obliquus, Pb exposure reduced intracellular IAA and cell proliferation, although IAA supplementation enhanced antioxidant defenses, indicating that auxin effects depend on context and concentration [83]. In E. mutabilis, the absence of detectable extracellular IAA is therefore consistent with reduced growth-associated signalling under Cd stress, but the available data do not distinguish between intracellular auxin metabolism and reduced auxin export.

The lack of extracellular IAA may also reflect altered microbial contribution to auxin pools. Auxin produced by co-cultured bacteria can promote growth in E. gracilis, suggesting that IAA may function as a cross-kingdom signalling molecule within microbial consortia [84]. In the present study, Cd exposure may have reduced extracellular auxin production by the algal host, microbial partners or both. Regardless of source, the loss of detectable extracellular IAA aligns with the broader stress phenotype observed in the E. mutabilis FAB culture, including reduced proliferation, increased paramylon storage, suppressed light-harvesting gene expression and maintained chloroplast function under Cd exposure.

Microbial contributions and community-level adaptation

Genes with mixed phylogenetic origins may reflect either poorly characterized E. mutabilis transcripts or contributions from other FAB members under Cd stress. Because metatranscriptomic taxonomic assignments can be biassed by incomplete databases, uneven genome representation and limited phylogenetic resolution, especially for fungi, we interpret these data conservatively at broad taxonomic levels rather than as species-level identifications [85–87]. This approach is consistent with previous sequence-based work identifying the prominent members of this FAB consortium and E. mutabilis, a Talaromyces sp. fungus and A. acidophilum [20].

Although some transcripts matched specific fungal or bacterial taxa, these signals are best interpreted as evidence of community-level functional potential. Metal-stress tolerance in microbial consortia can involve broad distributed processes such as biosorption, chelation, pH regulation and redox modification [7, 88]. Fungi associated with plants can buffer oxidative stress and alter nutrient mobility under heavy metal exposure [89], while acidophilic bacteria adjust pH and redox conditions in extreme environments [67]. Thus, the increased abundance of fungal- and bacterial-associated transcripts in Cd-treated cultures supports a consortium-level response, but independent confirmation would be required to assign these functions to specific taxa beyond the previously identified FAB members. Because euglenoids also contain genes with plant- and bacteria-like similarities [90, 91], functional interpretation is more robust than species-level assignment.

Insights into Cd tolerance of E. mutabilis from transcriptome analysis

GO-term enrichment and differential expression analyses indicate that Cd exposure altered pathways associated with metal transport, ion homeostasis, photosynthesis and oxidative stress mitigation. Genes linked to metabolism, chloroplast thylakoid membrane and thylakoid functions were enriched, while genes associated with chloroplast, photosynthesis/light harvesting, transmembrane transporter activity and formate transmembrane transporter activity were consistently downregulated. ZIP (ZRT/IRT-like Protein) family transporters and cellular zinc homeostasis were significantly upregulated across timepoints. ZIP and IRT transporters mediate the uptake and intracellular transport of divalent metal ions, including Zn²⁺, Fe²⁺ and Cd²⁺, across membranes in plants and microalgae [92, 93], and ZIP transporters have been implicated in Cd accumulation within chloroplasts of E. gracilis [27, 44]. Their upregulation in E. mutabilis is consistent with the electron-dense deposits observed by TEM.

Temporal comparisons within Cd-treated cultures showed additional enrichment of membrane-associated functions between 14 and 21 days, including cation transmembrane transporters and CDF proteins. These transporters can move metals from the cytosol into organelles or out of the cell [94, 95], and both ZIP and cation transporters have been linked to Cd translocation and tolerance in E. gracilis and C. reinhardtii [96, 97]. Together, these patterns support a model in which Cd tolerance involved coordinated regulation of metal uptake, redistribution and compartmentalization rather than reliance on a single transporter class.

Unlike many microalgae, E. mutabilis did not exhibit increased expression of phytochelatin or glutathione synthesis genes, suggesting that other detoxification mechanisms may be important. Upregulation of a proton-phosphate symporter may indicate a role for phosphate transport in Cd management. Phosphate can reduce Cd bioavailability through Cd-phosphate complexation and can influence oxidative stress response [98, 99], although its effects can vary among systems [100–102]. In E. mutabilis, concurrent changes in phosphate transporters, ZIP family genes and CDF transporters suggest that phosphate-associated redistribution or complexation may contribute to Cd tolerance, but this mechanism requires direct biochemical confirmation.

Cd exposure also produced a clear photosynthetic stress response. Core photosynthetic genes including cytochrome b6, ATP synthase β-subunit and PSII reaction centre proteins were upregulated, while LHC genes were downregulated. Because Cd can disrupt photosynthetic proteins and metal cofactors, including Mg²⁺ or Fe²⁺, increased expression of core photosynthetic components likely reflects repair or maintenance of essential chloroplast function rather than enhanced photosynthetic efficiency [96, 103–108]. In contrast, repression of light-harvesting genes is consistent with photoprotective responses that limit excitation pressure and reduce reactive oxygen species (ROS) formation during photosynthetic stress [59, 109–111].

Downregulation of formate/nitrite transporters may further reduce redox stress. Limiting nitrite influx can reduce reactive nitrogen species formation and conserve reducing power for antioxidant defenses [110, 112]. Similarly, reduced formate transport may limit the protonated state of nicotinamide adenine dinucleotide (NADH) production from formate oxidation, lowering the risk of electron transport chain over-reduction and ROS generation [110, 113]. These patterns suggest that E. mutabilis maintains core energy-generating machinery while limiting input that could increase oxidative burden.

Together, the transcriptional, physiological, and ultrastructural data support a unified model of Cd tolerance in the E. mutabilis FAB consortium (Fig. 6). This model includes Cd uptake through metal transporters, intracellular redistribution and compartmentalization, reduced light harvesting and nitrogen/formate transport to limit oxidative stress, paramylon accumulation as a metabolic buffer and suppression of growth-associated hormone signalling. The enrichment of FAB-associated microbial signals under Cd further suggests that microbial partners may contribute to consortium stability through broad metal-tolerance functions such as biosorption, chelation and environmental modulation. Overall, Cd tolerance in this system appears to reflect coordinated algal and community-level responses that support long-term survival in metal-contaminated environments.

Fig. 6. Conceptual model of the multifaceted response of the E. mutabilis FAB consortium to Cd stress, integrating metal handling, metabolic regulation, antioxidant defence and hormonal modulation. Cd uptake is initially mediated by ZIP family transporters, followed by upregulation of CDFs and CDF transporters to support intracellular sequestration or export. Chloroplasts may serve as key detoxification compartments, limiting damage to other organelles. Cd exposure induces ROS (blue shape and up arrow), prompting downregulation of LHC and formate/nitrite transporter genes to reduce light harvesting and electron leakage. Despite suppression of light-harvesting activity, upregulation of ATP synthase and PSII core components suggests maintenance of basal photosynthetic function (depicted as dark pigmented chlorophyll). Paramylon accumulation (yellow) provides metabolic buffering under prolonged stress. Hormone responses include suppressed active CKs and auxin with the downregulation of IPT and LOG (grey star and down arrow) but increased CK nucleotide forms (purple star and up arrow), indicating a shift toward hormonal dormancy. Symbiotic partners such as Talaromyces and A. acidophilum may further enhance tolerance via metal chelation, pH stabilization and phytohormone signalling. Together, the model highlights a coordinated strategy for survival in metal-contaminated environments. The figure was made using BioRender.com.

Conclusions

This study reveals the unique molecular and physiological adaptations of a naturally occurring E. mutabilis FAB consortium to Cd stress, integrating transcriptomics, hormone profiling, structural imaging and community analysis. Notably, the combined response of E. mutabilis and its microbial partners involves a transition from early metal uptake to intracellular detoxification, strategic suppression of ROS-generating pathways, metabolic reorganization and hormone downregulation to maintain cellular stability.

The novelty of this work lies in the holistic demonstration that natural microbial consortia – rather than isolated monocultures or synthetic co-cultures – coordinate at multiple biological levels to achieve extreme stress resilience. By preserving photosynthetic function, activating detoxification pathways and dynamically modulating hormone metabolism, this system offers a blueprint for leveraging ecological interactions in engineered biotechnologies. The E. mutabilis FAB consortium represents a compelling model for designing biologically robust, adaptive remediation systems for acidic and metal-laden environments. These findings underscore the potential of co-cultured extremophiles in shaping the next generation of sustainable bioremediation strategies.

Abbreviations

AMD

acid mine drainage

Cd

cadmium

CDF

cation diffusion facilitator

CK

cytokinin

COG

Cluster of Orthologous Groups

CPCC

Canadian Phycological Culture Centre

cZ

cis-zeatin

cZNT

cis-zeatin nucleotide

cZR

cis-zeatin riboside

DEG

differentially expressed gene

DGE

differential gene expression

FAB

fungal–algal–bacterial

GO

Gene Ontology

IAA

indole-3-acetic acid (auxin)

IPT

isopentenyltransferase

KEGG

Kyoto Encyclopedia of Genes and Genomes

LHC

light-harvesting complex

LOG

LONELY GUY (cytokinin-activating enzyme)

MAM

modified acid medium

MeSZR

methylthiolated cis-zeatin riboside

NT

nucleotide

PE

paired-end (sequencing reads)

PSII

photosystem II

RNA-seq

RNA sequencing

ROS

reactive oxygen species

SAM

S-adenosylmethionine

TEM

transmission electron microscopy

ZIP

ZRT/IRT-like protein (metal transporter family)