<!DOCTYPE html> Electrode



Injuries–resulting from traffic collisions, drowning, poisoning, falls or burns - and violence - from assault, self-inflicted violence or acts of war–kill more than 5 million people worldwide annually and cause harm to millions more. They account for 9% of global mortality and are a threat to health in every country of the world.

In real emergency situations, there is probably no immediate access to hospitals, no professional analytical instruments and no medical specialists. In most cases, the lack of medical conditions can easily lead to delays in the treatment window and even death.

Under normal situations, simple biochemical detection requires patients to wait for hours, which is not friendly in modern fast-paced era at all, let alone high price of inspections. Moreover, some indicators cannot be detected with one single enzyme or inspection, while some inspections check unnecessary indicators.

Therefore, we used a well-designed multi-enzyme complex immobilized on electrodes to detect injuries-related biomarkers. The whole process only takes 30s! The multi-enzyme complex forms a specific logic that makes the detection process faster than ever before.

And, our products are called A detector I and A detector II.

Introduction of our project (in Q&A way) Why do we use Protein Logic Circuit? Why is it the best choice?

At the beginning, like many iGEMers, we were ready to work at the genetic (DNA) level. However, plenty of time needed for gene expression is completely inconsistent with the emergency situation of first aid. Protein (enzyme) is almost our best choice.

We spent a lot of thoughts on how to further shorten the testing time. After many rounds of brainstorming, we realized that instead of sticking to shorten the time required for a single test, it would be more efficient to research how to find more in one detection. This is where our design originated - logical gates.

A typical enzyme electrode sensor can tell us the concentration of a specific substance in a sample. This reminds us - if we use a multi-enzyme complex, can we detect multiple substances at the same time? Based on this, our thoughts continued to diverge outwards - If the multi-enzyme complex is well-designed and its catalytic reactions and sequencing of the enzymes meet a certain logic, they can even form logic gates! Give the simplest example: two enzymes in series (the product of the enzyme A is the substrate for the enzyme B) is an AND gate. If multiple pathological indicators of a disease meet a certain logic (This situation is very common), they can be the inputs of the protein logic gate. After a certain logic operation, we can instantlythe results of whether the subject is sick. This is probably the fastest test method! By contrast, traditional detection methods require different sensors to detect different substances, which requires precious time.

Our tools for building multi-enzyme complexes are three sets of robust, highly specific and orthogonal linkers - Tag/Catcher system.

How do we apply our ideas to real life?

We created two products, A Detector I and A Detector II. The first generation of products met all our expectations for protein logic gates. The second generation has been greatly improved in size so that it can even be put in your pocket. We went to the hospital for blood sample testing and finally implemented our project!

The work we have done

We constructed an artificial multi-enzyme complex to perform a logic gate at the protein level

We made the first generation of products, A detector I, and achieved satisfactory results!

We upgraded the former to A detector II and tested it in the hospital!

Main result Protein logic circuit

Based on protein structure prediction and sequence analysis, we DIY some enzymes ourselves. They are linked together by the Tag/Catcher system and form a multi-enzyme complex. Through the relationship between the substrates and the products, they form well-run logic gates.

Fig. 1 Protein structure modeling made by de novo way of Robetta.
Fig. 2 Proof of the specificity and orthogonality of the three tag-catcher systems.
Fig. 3 SDS-PAGE results indicated that the three enzymes were successfully expressed.
A detector I

By using the multi-enzyme complex immobilized on a electrode and an electrochemical workstation, we constructed enzyme electrodes, which we called A detector Ⅰ!This sensor has a unique way to immobilize the enzymes - matrix formed by the Curli protein.This method is not only stable, but also capable of retaining enzyme activity. The results of the single enzyme test and the detection of logic gates both met our expectation well.

Fig. 4 Quantitative Congo Red (CR) Binding Assays to verify the successful expression of Curli.
Fig. 5 The current-time curve shows that A detector 1 distinguishes different sets of inputs well.
A detector II

Although the results of the A detector Ⅰ are satisfactory, through many expert consultations and field research, We found that the product may encounter difficulties in practical uses. Therefore, we designed the second generation of products - A detector II! Compared to the huge size of the A detector I, the generation 2 is so small that it can even be carried in your pocket! We used A detector II to test the patients' blood samples in the hospital and found that it can successfully distinguish the blood samples of patients and normal people and give the results within 30s.

If the video format is not supported in Google Chrome, please open this page in Microsoft Edge or Apple Safari
Fig. 6(A)From samples we get amperometric I-t curves and amperometric I-t curve shows whether a person is injured (ill) or not. (B) Machine learning is used to determine brain disease (meningitis and cerebral infarction) and abdominal trauma. We used 72 data to test our products. There were 28 samples of brain diseases, 17 normal samples and 27 abdominal trauma. It can be seen that data larger than LineA (0.721 μA) is classified as a brain disease sample, data smaller than LineB (0.254 μA)is classified as abdominal disease sample, and data between the two lines are classified as normal condition. Our products have a good ability to distinguish sick from normal.
Future work

Design a software that can give different multi-enzyme complex design strategies according to different needs, so that our project can achieve higher applicability.

Improve user experience with A detector II, and make UI more user-friendly

Use a larger sample size to train A detector II making the results of machine learning more accurate