Prenyl groups, the five-carbon isoprene-derived units installed by prenyltransferases across all domains of life, are among the most common chemical decorations found on natural products. They modulate membrane affinity, target binding, and the photochemical and oxidative stability of their host molecules. Conventionally, the further diversification of prenyl side chains has been understood to proceed through oxidation at the allylic positions, cyclization, or hydration chemistry that preserves or extends the preexisting double bond. The newly described enzymes break with that logic. Rather than modifying the prenyl unit from its periphery, they remove one of its methyl substituents in a demethylation reaction that reorganizes the bonding framework of the side chain itself, converting what was a simple isoprenoid branch into an allene or an alkyne.

The investigative team, whose analysis appears in the journal’s September 2026 issue, combined comparative genomics, in vivo gene activation, purified-enzyme biochemistry, isotopic labeling, and computational modeling to trace the reaction from gene to product. The biosynthetic gene clusters in question were flagged because they encode proteins annotated as radical S-adenosylmethionine enzymes alongside partners resembling methyltransferases and oxidative tailoring catalysts. When the clusters were expressed in heterologous bacterial hosts, the researchers observed the accumulation of metabolites bearing allene and terminal alkyne functionalities, structures that could only be explained if the prenyl substituents had lost a methyl group and undergone a formal dehydrogenation in the process.

At the heart of the discovery is the finding that the demethylating enzyme acts on the methyl group at the branched position of the prenyl chain. Isotopic feeding experiments, in which cultures were supplied with methyl-labeled precursors, demonstrated that the carbon of that methyl group is released from the product scaffold, while the hydrogen atoms retained on the adjacent carbons undergo stereospecific removal. The net result is a desaturation achieved by C–C bond cleavage followed by reorganization of the pi system, a sequence that organic chemists would typically accomplish with strong bases or transition-metal catalysts under strictly anhydrous conditions. The enzyme accomplishes the same transformation in water, at ambient temperature and neutral pH, using nothing more exotic than the cofactors already standard in secondary metabolism.

Mechanistically, the evidence supports a pathway in which an initial methylation or hydroxymethylation prime of the prenyl unit sets up an elimination-competent intermediate. Computational docking and quantum mechanical calculations suggest that the enzyme positions the methyl substituent adjacent to a catalytic base, enabling proton abstraction that drives fragmentation of the C–C bond to the methyl carbon. The resulting conjugated intermediate collapses either to an allene, when the geometry of the active site permits cumulation of the two double bonds, or to an alkyne, when successive dehydrogenation steps flatten the terminus into a linear acetylenic arrangement. The divergence between the two outcomes appears to be governed by subtle differences in the active-site architecture among enzyme variants, a conclusion the authors support with site-directed mutagenesis in which single residue swaps shifted product profiles from allene-bearing to alkyne-bearing metabolites.

The biological context of these transformations is as intriguing as the chemistry itself. Allene and alkyne motifs are rare but consequential in pharmacologically relevant molecules. The cytotoxic natural product families that contain them often owe their potency to the electrophilicity and strained geometry of the unsaturation, which can engage biological nucleophiles or undergo controlled activation to generate cytotoxic species. By revealing a biosynthetic route to these groups that runs through demethylation, the study supplies a missing link in the catalog of enzymatic reactions available to natural product assembly lines and suggests that many undiscovered metabolites bearing these motifs may be encoded in silent or poorly annotated gene clusters across microbial genomes.

From a biocatalysis standpoint, the implications are immediate. Synthetic routes to allenic and acetylenic compounds frequently demand multiple steps, protecting groups, and careful control of regioselectivity and stereochemistry. An enzyme that installs these functionalities regioselectively from a prenylated precursor offers a shortcut that synthetic chemists can borrow. The demonstrated tolerance of the enzymes for varied prenylated substrates raises the prospect of chemoenzymatic pipelines in which readily assembled prenylated intermediates are converted into allene- or alkyne-containing analogs for medicinal chemistry screening. Because the reactions proceed under mild aqueous conditions, they are compatible with sensitive molecular frameworks that would not survive conventional synthetic desaturation chemistry.

The study also contributes to a broader reassessment of what demethylation can mean in enzyme chemistry. Demethylases are conventionally viewed as deactivating catalysts: they remove methyl groups in the course of detoxification, epigenetic regulation, or catabolism, restoring a parent structure without altering the carbon skeleton beyond the excised methyl. The enzymes described here instead use demethylation as a constructive act, coupling the loss of a one-carbon unit to a deep reorganization of unsaturation elsewhere in the molecule. This reframing widens the mechanistic repertoire attributed to S-adenosylmethionine-dependent and methyl-oxidizing enzyme families and predicts that other examples of desaturative demethylation await discovery, particularly in gene clusters encoding enzymes of mixed annotation whose functions have not been experimentally interrogated.

The authors buttress their mechanistic proposals with a combination of substrate analog studies and structural modeling that delineates how the active site discriminates between the two possible unsaturated outcomes. Key residues lining the substrate channel appear to enforce the trajectory of the departing methyl group and the orientation of the nascent pi system, effectively templating the geometry of the product. When the researchers perturbed these residues, the enzyme’s output shifted in ways consistent with the computed energy landscapes, reinforcing the picture of an active site that does not merely accommodate a reaction but actively choreographs which of two chemically plausible unsaturations emerges. That level of product control, achieved without metal cofactors beyond those required for the initial radical or oxidative priming steps, underscores the sophistication with which enzyme pockets can steer reactive intermediates toward defined outcomes.

Looking forward, the work opens several avenues. Genome-mining campaigns can now be retargeted to search for homologs of the demethylating enzymes, prioritizing clusters whose neighborhoods encode prenyltransferases and tailoring oxidases suggestive of comparable chemistry. Protein engineering efforts can explore whether the allene-versus-alkyne decision can be rationally inverted, converting these enzymes into programmable instruments for installing either motif at will. And for natural products research more broadly, the finding is a reminder that the chemical vocabulary of metabolism remains incompletely inventoried: even a modification as familiar as the prenyl group, appended to countless molecules and studied for decades, conceals transformations that redefine what enzymes can be asked to do. In removing a methyl group, these catalysts add two new bonds of unsaturation to the biosynthetic lexicon, and they invite chemists to reconsider how many other routine modifications might harbor equally unexpected chemistry.

Subject of Research: Enzymatic prenyl demethylation that generates allene and alkyne functionalities in microbial natural product biosynthesis

Article Title: Biosynthetic allene and alkyne formation by enzymatic prenyl demethylation

Article References: Liu, M., Ohashi, M., Han, W., Zhou, Q., Houk, K. N., & Tang, Y. (2026). Biosynthetic allene and alkyne formation by enzymatic prenyl demethylation. Nature Chemical Biology. https://doi.org/10.1038/s41589-026-02323-w

Image Credits: AI Generated

DOI: 10.1038/s41589-026-02323-w

Keywords: biosynthesis, allene, alkyne, prenylation, demethylation, radical SAM enzymes, natural products, biocatalysis, enzyme mechanism, secondary metabolism, isotopic labeling, genome mining

The investigation began with an integrated bioinformatics screen of colorectal cancer datasets, from which SUCLG2 emerged as a candidate gene whose expression tracks with disease severity. Multi-omics validation in clinical specimens confirmed the pattern: SUCLG2 levels were significantly downregulated in colorectal cancer tissue compared with healthy tissue, and the degree of loss correlated tightly with advanced TNM staging and poor patient prognosis. Patients whose tumors expressed the least SUCLG2 fared worst, positioning the enzyme not only as a mechanistic player but also as a potential prognostic biomarker that could help clinicians stratify risk at the time of diagnosis.

To probe function, the team manipulated SUCLG2 in colorectal cancer cell lines and in animal models. Functional assays showed that SUCLG2 restrains the proliferation of colorectal cancer cells in culture and suppresses xenograft tumor growth in vivo. When the enzyme was depleted, cells grew more aggressively; when it was restored, growth slowed appreciably. These results established SUCLG2 as a bona fide tumor suppressor rather than a metabolic bystander, raising the central question of how a tricarboxylic acid cycle enzyme exerts such direct control over cancer cell behavior.

The answer, uncovered through transcriptomic, metabolomic, and epigenetic analyses, lies in an unexpected biochemical pathway involving S-adenosylmethionine, or SAM, the universal methyl donor of the cell. SUCLG2 activity reduces intracellular SAM levels. Because SAM is the substrate required by DNA methyltransferases, or DNMTs, to attach methyl groups to DNA, lower SAM availability dampens DNMT activity genome-wide. The consequence for colorectal cancer cells is profound: hypermethylation of tumor suppressor gene promoters is alleviated, and genes silenced by this epigenetic brake can be switched back on.

Among the genes reactivated by this mechanism, one stood out. GADD45G, a growth arrest and DNA damage-inducible gene with well-documented anti-proliferative functions, regained transcriptional activity when SUCLG2 was present. The researchers showed that SUCLG2 demethylates the GADD45G promoter through this SAM-dependent epigenetic remodeling, relieving promoter hypermethylation and restoring GADD45G expression. Gene set enrichment analysis reinforced the picture, revealing that SUCLG2 overexpression activates p53 and apoptosis signaling while inhibiting cell cycle pathways, consistent with GADD45G acting as a conduit between the mitochondrial enzyme and the cell’s central tumor-suppressive machinery.

Causality was tested directly. When the researchers knocked down GADD45G in cells engineered to overexpress SUCLG2, the tumor-suppressive effects of the enzyme were largely abolished. This loss-of-function experiment established GADD45G as a critical downstream mediator of the SUCLG2/p53 signaling axis. In other words, SUCLG2 does not simply slow cancer cells through metabolic exhaustion; it reactivates a specific genetic program, via demethylation of GADD45G, that engages p53-mediated growth inhibition and programmed cell death.

Clinical validation strengthened the mechanistic model considerably. Across patient cohorts, the team confirmed a robust positive correlation between SUCLG2 and GADD45G expression, and, in keeping with the proposed epigenetic mechanism, a negative correlation between SUCLG2 levels and methylation of the GADD45G promoter. Tumors with abundant SUCLG2 tended to carry unmethylated, transcriptionally active GADD45G, whereas SUCLG2-poor tumors showed the silenced, hypermethylated state. These correlative findings in human tissue mirror the experimental results and suggest that the SAM-DNMT-GADD45G axis operates in actual disease, not merely in laboratory models.

The study carries notable therapeutic implications. Epigenetic silencing of tumor suppressor genes is a hallmark of colorectal cancer, and demethylating agents exist but act globally, with limited specificity and considerable toxicity. If SUCLG2 activity, or downstream nodes of its pathway, could be pharmacologically enhanced, it might offer a more targeted way to lift methylation repression specifically at tumor suppressor promoters. Alternatively, the SUCLG2-GADD45G-p53 axis could be exploited indirectly, for example by screening for compounds that mimic the enzyme’s effect on SAM metabolism or DNMT activity. The authors position SUCLG2 as both a promising prognostic biomarker and a candidate therapeutic target, though translating these findings into clinical interventions will require further preclinical development and validation in larger patient populations.

Beyond its immediate clinical relevance, the research adds to a growing appreciation that metabolic enzymes can double as epigenetic regulators. Because metabolites such as SAM, alpha-ketoglutarate, acetyl-CoA, and NAD+ serve as substrates and cofactors for chromatin-modifying enzymes, shifts in cellular metabolism can directly reshape the epigenetic landscape. The SUCLG2 story is a vivid example: a change in the activity of a TCA cycle enzyme propagates through the methyl donor economy of the cell to determine whether a key anti-cancer gene is audible or silenced. As colorectal cancer remains a major clinical challenge, uncovering such regulatory mechanisms may open entirely new avenues for early detection, risk stratification, and treatment.

Subject of Research: SUCLG2-mediated epigenetic activation of the GADD45G-p53 axis in colorectal cancer progression

Article Title: SUCLG2 demethylates GADD45G to activate the p53 pathway and inhibit malignant progression in colorectal cancer

Article References: Guo, T., Lai, S., Yang, K., Tong, J., Liao, G., Lu, L., Jiang, C., Liu, H., Wu, Z., Yang, M., & Guo, J. (2026). SUCLG2 demethylates GADD45G to activate the p53 pathway and inhibit malignant progression in colorectal cancer. Cancer Cell International. https://doi.org/10.1186/s12935-026-04454-5

Image Credits: AI Generated

DOI: 10.1186/s12935-026-04454-5

Keywords: colorectal cancer, SUCLG2, GADD45G, p53 pathway, DNA methylation, demethylation, S-adenosylmethionine, DNA methyltransferase, tumor suppressor, cancer metabolism, epigenetics, prognostic biomarker