Neuromorphic Circuits and Biomaterials

I was not able to attend this lab in person because I was traveling for an interview. I informed Ronan ahead of time. A makeup session was not offered because this lab depended on time-sensitive biological materials, including mammalian cells that had already been grown or plated for the scheduled experiment.

Even though I missed the hands-on part, I reviewed the lab materials and thought through what I would have done if I had been there.

Lab Goal

The lab was about building neuromorphic genetic circuits in mammalian cells. The word “neuromorphic” here does not mean that the cells are literally neurons. It means that the circuit is designed to process information in a more dynamic way than a simple on/off reporter.

Instead of just putting a fluorescent protein after a promoter, the circuit uses regulatory parts that interact with each other before producing an output. That makes the behavior depend on combinations of inputs, expression levels, and timing.

What I Would Have Tried

If I had attended the lab, I would have kept my circuit fairly simple. Since transfection itself can be noisy, I think it would be better to design something that is easy to debug rather than something too clever.

I would have tried a two-input circuit where Csy4 and CasE act as the main regulatory inputs, and mNeonGreen is the final output. I would also include separate fluorescent markers for the two input groups so that I could tell whether the cells actually received and expressed those inputs.

The main reason for including mKO2 and eBFP2 is that they would make the result easier to interpret. If the final green output is missing, I would want to know whether the circuit logic failed or whether one of the input groups just did not transfect well.

Proposed Circuit Design

The circuit I would propose is a two-input regulatory circuit. The rough idea is that the final green output depends on how the Csy4 and CasE-related parts interact.

I would not expect this to behave like a perfect electronic logic gate. In cells, expression is messy and continuous. Some cells might receive more DNA than others, and even cells with the same DNA can express it at different levels. So I would treat this as a biological signal-processing experiment, not as a clean digital circuit.

The three most useful readouts would be:

• mKO2 for the Csy4 input group
• eBFP2 for the CasE input group
• mNeonGreen for the final output

That way, the final output could be compared against the input markers.

Spreadsheet Plan

The lab used a spreadsheet to specify the circuit. Each row listed the circuit name, transfection group, part name, DNA concentration, and amount of DNA wanted. The concentration was fixed at 50 ng/uL, so the important design choices were which parts to include and how much DNA to use.

I would fill it out like this:

Total DNA: 650 ng.

I chose 650 ng because the protocol limit was 650 ng total DNA. I gave the input components and marker components 100 ng each, then used the remaining DNA for the output-related components. This feels like a reasonable first-pass design because it keeps the inputs visible while still giving enough DNA to the output module.

OT-2 Workflow

The spreadsheet would then act as the instruction layer for the OT-2 robot. The robot would use the part names and DNA amounts to prepare the transfection mixtures.

This part of the lab is interesting because it turns the spreadsheet into something executable. The design is not just notes for a human. It becomes a recipe for the robot to pipette the correct DNA parts into the correct mixtures.

Using the OT-2 also makes sense because these experiments can involve many small-volume transfers. Doing that by hand would be easy to mess up, especially if different groups are testing different circuit designs.

Transfection into HEK293 Cells

After the DNA mixtures were prepared, they would be transfected into HEK293 cells. Transfection introduces DNA into mammalian cells, and then the cells express the circuit components from that DNA.

I think the key point is that the cells are the system actually running the circuit. The DNA is the design, but the result depends on the cell state, transfection efficiency, expression level, and time after transfection.

HEK293 cells are commonly used for mammalian expression experiments, so they are a practical choice for this kind of lab.

What I Would Look For

If I had been able to run the experiment, I would compare the three fluorescence channels rather than only looking at the final green output.

The most useful comparison would be across different input combinations:

This would help separate real circuit behavior from boring experimental failure. For example, if mNeonGreen is low but mKO2 and eBFP2 are also low, then the problem might just be poor transfection. But if both input markers are high and mNeonGreen changes, then the circuit interaction is more meaningful.