Biomolecular Filters for Improved Separation of Output Signals in Enzyme Logic Systems Applied to Biomedical Analysis

Biomolecular logic systems processing biochemical input signals and producing “digital” outputs in the form of YES/NO were developed for analysis of physiological conditions characteristic of liver injury, soft tissue injury and abdominal trauma. Injury biomarkers were used as input signals for activating the logic systems. Their normal physiological concentrations were defined as logic-0 level, while their pathologically elevated concentrations were defined as logic-1 values. Since the input concentrations applied as logic 0 and 1 values were not sufficiently different, the output signals being at low and high values (0, 1 outputs) were separated with a short gap making their discrimination difficult. Coupled enzymatic reactions functioning as a biomolecular signal processing system with a built-in filter property were developed. The filter process involves a partial back-conversion of the optical-output-signal-yielding product, but only at its low concentrations, thus allowing the proper discrimination between 0 and 1 output values.

💡 Research Summary

The paper addresses a critical limitation in enzyme‑based biomolecular logic systems used for biomedical diagnostics: when normal (logic‑0) and pathological (logic‑1) concentrations of input biomarkers differ only modestly, the resulting output signals (typically optical read‑outs) overlap, making binary YES/NO decisions unreliable. The authors previously constructed AND‑gate platforms for detecting liver injury (ALT and LDH), soft‑tissue injury, and abdominal trauma, where the output is the decrease in NADH absorbance at 340 nm. Because the “0” inputs still contain low but non‑zero enzyme activity, NADH is gradually consumed even for non‑target input combinations, and after several minutes the absorbance changes for 0‑ and 1‑outputs become indistinguishable.

To solve this, the authors introduce a “biomolecular filter” that temporarily reconverts the product NAD⁺ back to NADH, thereby suppressing the output signal until the filter is exhausted. The filter consists of glucose‑6‑phosphate dehydrogenase (G6PDH) and its substrate glucose‑6‑phosphate (Glc6P). In the presence of Glc6P, any NAD⁺ generated by the primary cascade is reduced to NADH, keeping the absorbance essentially constant. Once Glc6P is depleted, NAD⁺ accumulates and the absorbance drops sharply, but only when both primary inputs (ALT and LDH) are at their elevated, logic‑1 levels. By adjusting Glc6P concentration (4 mM) and G6PDH activity (10 U mL⁻¹), the authors achieve a pronounced separation between 0‑ and 1‑outputs that persists for at least 600 seconds and up to three hours.

Experimental data show that without the filter, the absorbance decrease for the 1‑0 input combination eventually matches that of the 1‑1 combination, compromising the AND logic. With the filter, the 0‑output signals remain near zero while the 1‑output reaches the maximal normalized value (ΔA/ΔA_max ≈ 1). Receiver operating characteristic (ROC) analysis quantifies this improvement: the non‑filtered system yields an area under the curve (AUC) of 0.92 (95 % CI 0.79–1.00), whereas the filtered system achieves a perfect AUC of 1.00 (95 % CI 1.00–1.00), corresponding to 100 % sensitivity and specificity under the experimental variance.

The filter concept is also applied to the STI and ABT logic platforms, with comparable enhancements in signal discrimination. The authors discuss the trade‑off that the filter can reduce overall signal amplitude, potentially increasing relative noise, but this is acceptable for applications where the signal is measured after saturation (e.g., triggering drug‑release membranes). They note that the filter strategy is broadly applicable to any NAD⁺‑dependent dehydrogenase cascade, provided a suitable substrate (like Glc6P) is available that does not interfere with endogenous metabolites.

In conclusion, integrating a reversible enzymatic filter into biomolecular logic circuits dramatically improves the fidelity of binary output discrimination when input biomarker levels are only modestly separated. This advance paves the way for more reliable point‑of‑care diagnostics, smart therapeutic actuators, and complex information‑processing networks in physiological environments. Future work will involve testing with real patient samples, scaling to multi‑input networks, and exploring alternative filter chemistries for other metabolic pathways.