Description

Methanol is a single-carbon, low-carbon feedstock that can be synthesized from industrial flue gases and renewable energy sources. It is abundant and inexpensive[1]. Compared with traditional sugar-based substrates derived from crops, methanol is not affected by seasonal variation, possesses a higher energy density, and can enhance the efficiency of biosynthetic conversion[2]. Moreover, the increasingly mature CO2 hydrogenation process for methanol production endows it with renewable characteristics. These features make methanol an ideal low-carbon alternative carbon source for microbial biosynthesis and industrial biomanufacturing

Currently, various naturally methylotrophic microorganisms have been explored as chassis strains, such as the Methylorubrum extorquens, the Bacillus methanolicus, and the methanol-utilizing yeast Pichia pastoris[2]. However, these microorganisms often suffer from limitations such as low methanol utilization efficiency, restricted product diversity, and underdeveloped genetic engineering tools[3]. To construct more efficient methanol-based chassis strains, the development of new microbial systems has become an emerging trend.

In this project, Aureobasidium melanogenum P16 is selected as the research subject. A. melanogenum, a widely distributed fungus, exhibits tolerance to multiple extreme environments and has a broad substrate spectrum[4]. It is capable of producing diverse metabolites, including glucose and polysaccharides. Furthermore, this strain is classified as biosafety level 1 (BSL-1) and has been recognized as a promising producer of polysaccharides. However, as an emerging chassis organism, A. melanogenum still lacks comprehensive genomic information and a fully elucidated methanol metabolic pathway. In addition, its genetic editing and expression systems remain underdeveloped[3], posing both challenges and opportunities for further investigation in this work.

In this study, multiple Aureobasidium strains were screened to identify a suitable methylotrophic chassis. The strain Aureobasidium melanogenum P16 was selected for its robust tolerance to extreme environmental conditions and its ability to utilize diverse carbon sources, consistent with the characteristics of polyextremotolerant Aureobasidium species [4].

To evaluate its methanol utilization capacity, P16 was compared with a typical methylotrophic yeast and a non-methylotrophic strain. Komagataella phaffii (formerly Pichia pastoris) can use methanol as its sole carbon source via a complete methanol utilization pathway [6]; in contrast, non-methylotrophic species such as Saccharomyces cerevisiae are incapable of methanol assimilation. Experimental results showed that P16 exhibited good growth in methanol-containing medium and demonstrated superior tolerance to high methanol concentrations compared to non-methylotrophic strains, reaching a level comparable to that of typical methylotrophs. These results confirmed P16 as a promising chassis for studies on methanol utilization.

To address the lack of efficient genetic manipulation tools in P16, we constructed and optimized a CRISPR-Cas9-Am system tailored for A. melanogenum. This system was adapted from an AMA1-based Cas9 expression vector previously developed for Aureobasidium pullulans, and incorporated gRNA expression elements compatible with P16.

To validate system functionality, the ADE2 gene involved in purine biosynthesis was selected as a knockout target. The resulting ade2Δ mutants exhibited the expected red colony phenotype on medium supplemented with adenine precursors, consistent with the accumulation of the purine intermediate AIR. This result confirmed that the CRISPR-Cas9-Am system enables efficient gene disruption in P16, providing a reliable platform for subsequent functional studies and metabolic engineering [7].

Based on genome annotation and literature review, key enzymes involved in methanol metabolism were identified and analyzed for their putative subcellular localization. Alcohol oxidase (AOX), catalyzing the first step of methanol oxidation, typically contains a C-terminal peroxisomal targeting signal type 1 (PTS1), directing it to the peroxisome. To verify this prediction, an AOX–GFP fusion protein was constructed and expressed in P16. Fluorescence microscopy revealed peroxisomal localization of AOX, consistent with the bioinformatic prediction.

However, the downstream enzymes dihydroxyacetone synthase (DAS) and dihydroxyacetone kinase (DAK) lacked canonical PTS1 (e.g., SKL) or PTS2 motifs at their C-termini, suggesting that their peroxisomal localization may rely on unconventional mechanisms. Previous studies have shown that translational readthrough of stop codons in fungal cells can generate C-terminally extended proteins harboring cryptic PTS1 signals, enabling dual localization[8]. We therefore hypothesized that DAS and DAK in P16 might undergo ribosomal readthrough to produce peroxisome-targeted variants.

To test this hypothesis, we designed experiments simulating ribosomal readthrough. Specifically, the C-terminal regions of DAS and DAK were fused with GFP, and the stop codons were mutated or extended to allow continued translation beyond the native termination site. The extended DAS–GFP and DAK–GFP fusion proteins exhibited strong peroxisomal localization, whereas the wild-type versions remained primarily cytosolic. This observation supports the model that translational stop-codon suppression allows the synthesis of C-terminally extended variants containing functional PTS1 sequences, thereby directing the enzymes to peroxisomes. These findings align with previous reports[9] and demonstrate that translational readthrough can modulate the subcellular localization of methanol-metabolizing enzymes, deepening our understanding of this regulatory mechanism.

To enhance methanol utilization efficiency in P16, we introduced the Das gene from the highly efficient methylotrophic yeast K. phaffii for heterologous overexpression. The Das enzyme catalyzes a key formaldehyde fixation step in the methanol assimilation pathway, converting three molecules of formaldehyde into dihydroxyacetone phosphate (DHAP). Previous studies have demonstrated that Das co-expression in K. phaffii significantly improves the conversion efficiency of methanol to target products [9].

We constructed a strong-promoter-driven Das expression cassette and integrated it into the P16 genome. The engineered strain exhibited improved growth rates and increased production of methanol assimilation products (e.g., CO₂ fixation intermediates) in methanol-containing medium, indicating successful enhancement of methanol assimilation through Das overexpression.

To further direct carbon flux toward glucose biosynthesis, we overexpressed the yihX (HAD4) gene from Escherichia coli, encoding α-D-glucose-1-phosphatase, which catalyzes the hydrolysis of glucose-1-phosphate to free glucose. Simultaneously, the pfk gene encoding phosphofructokinase was knocked out to block a key step in glycolysis, redirecting metabolic flux toward glucose synthesis and alternative pathways. The engineered strain showed significantly higher extracellular glucose levels than the wild type under identical conditions. Overexpression of yihX enhanced glucose formation, while pfk deletion reduced competitive glycolytic consumption, jointly improving glucose synthesis efficiency.

This integrated metabolic engineering strategy, built upon prior gene knockout and overexpression validation, effectively optimized the overall carbon metabolic network, achieving a systematic research workflow from chassis selection to functional verification and metabolic optimization.

We developed the SL-AttnESM model to predict the subcellular localization of proteins. This model integrates embeddings from sequence-based pre-trained models with multi-layer self-attention modules, enabling the capture of both global semantic features and critical sequence motifs. The model takes amino acid sequences as input and outputs localization probabilities through Transformer-based attention blocks and fully connected classification layers.

In this project, SL-AttnESM was used to predict the subcellular localization of key enzymes in P16 and guide the design of targeting signals. For instance, for enzymes required to localize to peroxisomes, the model was used to validate the effectiveness of canonical C-terminal PTS1 motifs.

Additionally, we constructed an auxiliary tool named LocAgent to assist in protein localization engineering. LocAgent integrates protein localization rules and a knowledge graph to recommend the addition or modification of targeting sequences. For example, to direct a protein to the peroxisome, it suggests appending a canonical PTS1 motif at the C-terminus; for secretory proteins, it proposes optimized signal peptide sequences. By combining machine learning with rule-based reasoning, LocAgent provides a rational framework for localization optimization.

Together, these computational tools enabled in silico prediction and design of protein targeting modifications prior to wet-lab experiments, substantially improving experimental efficiency and design accuracy.

To automate methanol substrate regulation, we developed a real-time methanol monitoring and automatic feeding system. The hardware component employs an electrochemical methanol sensor to continuously measure methanol concentrations in the fermenter. The signal is processed by a microcontroller and transmitted to a control algorithm module implementing PID or fuzzy logic control. The algorithm compares real-time concentration values with predefined setpoints, calculates the required feed amount, and activates a micro-pump or valve for automated methanol addition.

This setup achieves a closed-loop “sensing–computation–execution” control process. In pilot tests under simulated fermentation conditions, the device accurately maintained methanol concentrations within safe limits and automatically replenished the substrate according to consumption rates. This system improved fermentation stability and production efficiency and can be applied in both laboratory-scale and industrial bioreactors, supporting large-scale methanol-based bioprocessing.

We sought guidance from leading researchers and industry professionals. Interviews with Professors Guanglei Liu, Meng Wang, Changwei Bian, and Hongting Tang provided crucial insights into enzyme localization, gene-editing efficiency, and hardware modeling, as well as perspectives on the project’s economic feasibility and carbon fixation potential.

We also visited Qingdao Kangqiao Pharmaceutical Group and discussed commercialization strategies with Mr. Xiangwei Liu, gaining a deeper understanding of cost analysis and equipment development priorities, thereby aligning our research more closely with real-world applications.

We tailored science communication activities for different age groups. At Huancheng Town Experimental Primary School in Weishan County, we introduced children to synthetic biology through storytelling and drawing. Our summer camp with Qingdao No. 9 High School enabled students to explore marine biology via lectures and dissection experiments. On campus, we organized an iGEM 2026 recruitment workshop focusing on literature review and one-on-one mentorship, fostering interest in synthetic biology and ensuring team continuity.

We paid special attention to underrepresented groups. During volunteer teaching in Jining, we brought cutting-edge scientific concepts to children in resource-limited areas. At Qingdao Xingfu Home for the Elderly, we conducted interviews and surveys to understand climate change from the perspective of older adults, while engaging in meaningful intergenerational dialogue that fostered empathy and awareness.

We hosted three academic exchange meetings with teams from different universities and participated in the CCiC conference, which provided valuable opportunities for feedback and collaboration. These interactions helped us refine our project design and improve the clarity and coherence of our wiki presentation.

Through these comprehensive and inclusive Human Practices activities, we continuously integrated external expertise, reflected on societal impact, and optimized our research direction. Our ultimate goal is to create a synthetic biology project that is not only scientifically innovative but also educationally valuable, industrially relevant, and socially inclusive.

1. [1] Leiden iGEM Team. “Project Description.” iGEM 2023 Wiki, 2023, https://2023.igem.wiki/leiden/description. Accessed 6 Oct. 2025.
2. [2] Singh, H. B., et al. (2022). Developing methylotrophic microbial platforms for a methanol-based bioindustry. Frontiers in Bioengineering and Biotechnology, 10:1050740.
3. [3] Reiter, M. A., et al. (2024). A synthetic methylotrophic Escherichia coli as a chassis for bioproduction from methanol. Nature Catalysis, 7(5), 560–573.
4. [4] Xiao, D., et al. (2024). Advances in Aureobasidium research: Paving the path to industrial utilization. Microbial Biotechnology, 17(8): e14535.
5. [5] Liu, N.-N., et al. (2017). Simultaneous production of both high molecular weight pullulan and oligosaccharides by Aureobasidium melanogenum P16 isolated from a mangrove ecosystem. International Journal of Biological Macromolecules, 102, 1016–1024.
6. [6] Krainer, F. W., et al. (2012). Recombinant protein expression in Pichia pastoris strains with an engineered methanol utilization pathway. Microbial Cell Factories, 11:22.
7. [7] Masi, A., et al. (2024). Genomic deletions in Aureobasidium pullulans by an AMA1 plasmid for gRNA and CRISPR/Cas9 expression. Fungal Biology and Biotechnology, 11:6.
8. [8] Stiebler, A. C., et al. (2014). Ribosomal readthrough at a short UGA stop codon context triggers dual localization of metabolic enzymes in Fungi and animals. PLOS Genetics, 10(10): e1004685.
9. [9] Krainer, F. W., et al. (2012). Recombinant protein expression in Pichia pastoris strains with an engineered methanol utilization pathway. Microbial Cell Factories, 11:22.