Lysine Chemistry Shapes Protein Readouts

Evidence

Lysine is both a positively charged amino acid and a site for modifications including ubiquitination and methylation. Two recent studies approached that chemistry from different directions. One experimentally blocked primary amino groups on purified ribonucleases and measured the consequences across several protein properties. The other mapped ubiquitinated peptides in human ovarian granulosa cells and followed one candidate protein into a cell model. They do not test the same disease or intervention, but together they show why “modified protein” is not a single biological state.

In the engineering study, researchers used human ribonuclease 1 and QBI-139, an engineered cytotoxic ribonuclease variant. Each protein contained eight lysines plus an accessible N terminus. Reductive methylation at 4°C and pH 7.5 converted the primary amino groups to dimethylamino groups in less than six hours. Mass spectrometry showed the expected complete labeling, a 252.49-dalton increase across nine sites.

The modification had sharply different effects depending on the property measured. Differential scanning fluorimetry found melting-temperature shifts of no more than 1.5°C. Binding between QBI-139 and ribonuclease inhibitor weakened modestly: the measured dissociation constant rose from 0.64 ± 0.08 nanomolar to 1.8 ± 0.3 nanomolar after dimethylation. Catalysis was far more sensitive. Ribonuclease 1 retained 0.14% of its catalytic efficiency, and QBI-139 retained 0.035%. That loss was consistent with lysines participating directly in substrate binding and catalysis.

Cell experiments reinforced the distinction between entry and enzyme function. Dimethylated and unmodified QBI-139 conjugates showed no significant difference in uptake at 24 hours in HEK 293AAV cells, and both produced little signal decay over 72 hours. Yet dimethylated QBI-139 was much less cytotoxic in two leukemia cell lines, consistent with its reduced ribonucleolytic activity. These results support reductive methylation as a useful probe or engineering method only when the relevant function is tested rather than assumed.

The proteomics study examined luteinized granulosa cells from 60 people undergoing their first IVF or ICSI cycle: 30 in a younger-age group with normal ovarian reserve and 30 in an advanced-maternal-age group described as having ovarian aging. Nine samples from each group were selected for ubiquitinated-proteomic sequencing. The analysis identified 174 ubiquitinated peptides meeting the reported thresholds of an absolute log2 fold change above 1 and q below 0.05; 80 were higher and 94 lower in the advanced-age group.

Pathway analysis associated the differentially ubiquitinated proteins with metabolism, ubiquitin-mediated protein degradation, and the proteasome. Ubiquitination was reported as higher among proteins linked to the tricarboxylic-acid cycle and lower among proteins linked to glycolysis and the pentose-phosphate pathway. Alpha-enolase, or ENO1, emerged from the network analysis. Immunoprecipitation and Western blot experiments showed lower ENO1 ubiquitination but higher ENO1 protein abundance in the advanced-age granulosa cells, while ENO1 messenger RNA did not significantly change.

The investigators then used hydrogen-peroxide-treated KGN granulosa-like tumor cells as an oxidative-stress model. ENO1 inhibitors reduced the reported senescence signal, lowered p16 expression, and improved measured glycolytic-enzyme activity and metabolite levels. This cell-model result adds functional evidence around ENO1, but it does not establish that the human group differences were caused by ENO1 ubiquitination.

Analysis — Modification Context Changes Meaning

The cross-study inference is that lysine modification should be interpreted at the level of the property actually measured. This is analysis, not a conclusion directly tested across both systems. In the ribonuclease experiments, global dimethylation barely changed thermal stability and only modestly weakened one protein-protein interaction, yet it reduced catalytic efficiency by orders of magnitude. In the granulosa-cell study, lower ubiquitination of ENO1 occurred alongside higher ENO1 protein abundance without a significant messenger-RNA change, which is compatible with post-transcriptional regulation but does not identify the responsible degradation mechanism. The studies therefore converge on a measurement principle rather than a shared pathway: a modification-site readout, total protein abundance, binding affinity, stability, and enzyme activity answer different questions. An emerging experimental strategy would measure several of those layers together and perturb specific lysine sites rather than infer function from a ubiquitinated-peptide list alone. Whether that approach explains ovarian-aging biology remains unproven.

Limitations

The reductive-methylation study tested two closely related ribonucleases, global modification of all primary amino groups, one inhibitor interaction, and a limited set of human cell lines. Its results cannot establish that other proteins, buried lysines, interfaces, or enzymes will respond similarly. The cellular persistence assay also showed little decay for either protein, limiting its ability to reveal a degradation difference.

The granulosa-cell source was available to this briefing at abstract depth, restricting assessment of participant matching, treatment exposures, site localization, model adjustment, and the full proteomics workflow. Only nine samples per group entered ubiquitinated-proteomic sequencing. The human comparison was observational, and KGN cells are a tumor-derived model exposed to hydrogen peroxide rather than primary granulosa cells aging in vivo. ENO1 inhibition does not prove that changing a particular ENO1 ubiquitination site caused the human protein-abundance pattern. Independent cohorts, site-specific perturbations, direct turnover measurements, and primary-cell experiments would be needed to test that mechanism.