Individual Final Project

Abstract

Emotional regulation is not solely a psychological phenomenon — gut bacteria actively produce neurotransmitters including GABA that modulate anxiety and stress responses through the gut-brain axis. Yet this biological reality remains invisible in how we understand selfhood and emotional autonomy. Most interventions treat stress as a generic physiological state to be suppressed. Meng Po Tang begins from a different premise: that the moment of emotional collapse is itself a form of biological knowledge, and that this knowledge can be encoded, stored, and returned to the body as care.

Meng Po Tang is an engineered probiotic system. The user records their own heartbeat during a moment of emotional crisis. This heart rate data is processed through a stochastic hidden Markov model to extract rhythmic parameters, quantised into an 8-bit binary string, and mapped to a 12-nucleotide DNA sequence — TCCTAGTTGGA, flanked by AAAA spacers. This sequence is synthesised and edited into the plasmid of Escherichia coli Nissle 1917. The bacterium carries a molecular sensing circuit: when the host enters a high-stress state again and norepinephrine concentration rises in the gut, the sensing system activates. The heart rate sequence begins to transcribe into RNA, triggering a toehold switch to open, driving expression of the downstream GAD gene, producing GABA. The body feels calm. What regulates your emotion is your own heartbeat from that day.

This is not a general anxiolytic. The circuit does not respond to stress in the abstract — it responds to this stress, the one whose rhythm you have already survived once. Named after the mythological soup administered at the threshold of reincarnation, Meng Po Tang operates simultaneously as a synthetic biology design and a speculative design provocation. Its central question is not only technical but philosophical: if a memory can be encoded into a living system that inhabits your body, and that system uses the memory to quiet you — who is doing the forgetting? Specific aims include designing and ordering the full genetic circuit via Twist Bioscience, developing in vitro validation protocols for stress-responsive GABA production, and framing the consent and encoding ritual as a designed bioethical intervention. Methods include SHM-based heart rate encoding, NUPACK toehold switch design, Benchling circuit mapping, fluorescence reporter assay, and GABA ELISA.

Inspiration

The day we broke up, I recorded my heartbeat. I have always been afraid to look back on those losses, sorrows, betrayals and pains. But each time I truly do, it gives me a sense of release about things that are not eternal. The memory of that heartbreak in my heart is like a bowl of Meng Po soup - when I truly face it, that part has gradually faded away. I began to think: Could this memory be turned into a ritual of food? I encoded that day’s heartbeat into an ATGC sequence, implanted it into probiotics, and let it enter my intestines. The stomach has always been an organ of emotion - when people feel nervous and anxious, it is always the first to react. If the memory I swallowed could be transformed into a calming substance there, helping me release and relax, and ensuring I never repeat the same mistakes - that would be the Meng Po soup in the sense of synthetic biology.

Main Idea

Mengpo Soup is an engineered probiotic system. Users record their heartbeats at a moment of emotional breakdown. This heartbeat data is encoded into a DNA sequence and edited into the plasmid of E. coli Nissle 1917. This strain of bacteria carries a set of molecular sensing circuits: when the host re-enters a high-stress state, the concentration of norepinephrine in the gut increases, activating the sensing system. The heartbeat sequence begins to be transcribed into RNA, triggering the toehold switch to open, and the downstream GAD gene is expressed, producing GABA, making the body feel calm. What regulates your emotions is your own heartbeat on that day.

system logic

Part 1: Biometric Data to DNA Encoding

This image focuses on the technological transformation of emotion. It depicts the flow from raw heart rate monitoring and mental input, visualized as ECG waves on a screen, into an algorithm that converts this raw data into a digital DNA sequence. We see the resulting ‘synthetic data’ string of ATGC base pairs ready for biological application.

Part 2: Synthio Memory Regulator and Gut-Brain Modulation

This image captures the biological synthesis. It features the physical precision bio-vessel (the capsule from the original diagram) containing the metabolic core. This vessel interacts with a complex model of the gut microbiome and the intestinal wall. We visualize the release of the engineered probiotic (E. coli Nissle 1917), which carries the specific genetic plasmid derived from Part 1, illustrating the full process of metabolic synthesis and subsequent gut-brain communication.

The Process

Part 1: Biometric Data to DNA Encoding

On March 29, 2026, 100 heart rate readings were recorded via Apple Watch across a single day — a day of personal significance. The data spans 12:21 to 22:45, with a several-hour gap in the middle that coincides with the most emotionally intense period of the day.

Step 1 — Modeling the rhythm The heart rate data is fitted to a Simple Harmonic Motion equation: x(t) = mean + A · cos(ωt + φ) This treats the heartbeat as a biological oscillator, extracting three parameters that together describe the rhythm of that day: ∙ A = 24.46 BPM — the amplitude of emotional fluctuation ∙ ω = 0.2848 rad/hr — the frequency, corresponding to a full cycle of approximately 22 hours ∙ φ = −0.568 rad — the phase offset, shaped by the silence in the middle The gap in recording is not treated as missing data. It is preserved in the model as a structural trace — embedded in the phase parameter as a period of arrested motion.

Step 2 — Quantization Each parameter is normalized within a defined biological range and mapped to an integer between 0 and 255, using 8 bits of precision.

Step 3 — Binary to nucleotide Each integer is converted to 8-bit binary. Every two bits are then translated into a nucleotide base:

00 → A 01 → T
10 → G 11 → C

The sequence The three parameters produce a 12-base sequence:

TCCTAGTTGGA

Flanked by neutral spacers for biological stability:

AAAA TCCTAGTTGGA AAAA

Part 2: Synthio Memory Regulator and Gut-Brain Modulation

Core design concept: Negative feedback biological loop

We divide the system into four logical levels: perception (Input), recognition (Logic), synthesis (Output). A. Perception Layer: Pressure Threshold Setting The system must be able to detect “abnormal” stress. It utilizes the QseC/QseB two-component system of Escherichia coli (which naturally responds to the host’s adrenaline and noradrenaline). Design Optimization: By analyzing P qseB Mutations in the promoter region were screened, and a response threshold was set (for example: activation only when the adrenaline concentration exceeds the normal resting level by three times). This prevents the system from accidentally triggering during normal activities.

B. Identification Layer: “Molecular Mask” of Heartbeat Sequence Your 20bp sequence is not only an identity identifier, but also plays the role of “unleashing inhibition”. Trigger RNA: When the promoter is activated, it transcribes AAAAUCUCUAGUUGGAAAAA. Toehold Switch: This is a hairpin-shaped RNA sequence that acts like a lock, securing the initiation site (RBS) of the GABA synthase in the stem. The hairpin will only open when your 20bp sequence appears and hybridizes perfectly with it.

C. Synaptic Layer: Metabolic Flow Control Substrate conversion: The selected E. coli strain needs to enhance its ability to absorb glutamate. Enzyme Engineering: Utilize highly effective GadB mutants to ensure that once the switch is activated, a sufficient concentration of GABA can be rapidly produced.

Background chose：E. coli Nissle 1917

Toehold Switch Step 1: Design complementary toehold regions 3’- TTTT AGGAUCAACCT TTTT -5’ 即：5’- TTTT TCCTAGTTGGA TTTT -3’（DNA） RNA形式：5’- UUUU UCCUAGUUGGA UUUU -3'

Group Final Project

L-Protein Engineering 

The idea is to transform L-Protein into a self-stabilizing, autonomously localized structure, eliminating the need for DNAJ assistance. This will improve the stability of L-Protein.

STEP 1

Gather Information about Lysis Protein /Sequence /DNAj Sequence / Conserved sites if any Known mutational effects from research

Lysis Protein Sequence (UniProtKB ID: https://www.uniprot.org/uniprotkb/P03609/entry)

METRFPQQSQQTPASTNRRRPFKHEDYPCRRQQRSSTLYVLIFLAIFLSKFTNQLLLSLLEAVIRTVTTLQQLLT

DnaJ sequence (UniProtKB ID: https://www.uniprot.org/uniprotkb/P03609/entry)

MAKQDYYEILGVSKTAEEREIRKAYKRLAMKYHPDRNQGDKEAEAKFKEIKEAYEVLTDSQKRAAYDQYGHAAFEQGGMGGGGFGGGADFSDIFGDVFGDIFGGGRGRQRAARGADLRYNMELTLEEAVRGVTKEIRIPTLEECDVCHGSGAKPGTQPQTCPTCHGSGQVQMRQGFFAVQQTCPHCQGRGTLIKDPCNKCHGHGRVERSKTLSVKIPAGVDTGDRIRLAGEGEAGEHGAPAGDLYVQVQVKQHPIFEREGNNLYCEVPINFAMAALGGEIEVPTLDGRVKLKVPGETQTGKLFRMRGKGVKSVRGGAQGDLLCRVVVETPVGLNERQKQLLQELQESFGGPTGEHNSPRSKSFFDGVKKFFDDLTR

Known Mutational Effect

LS dipeptide: Positions 44 and 45 (Leucine-Serine) in the MS2 L protein are crucial. Domain division: The L protein is divided into four domains. Domain 1 (N-terminus), while positively charged and important, is not essential for the cleavage function itself (primarily responsible for binding to the host chaperone protein DnaJ); while Domains 2 to 4 (C-terminal half), containing the LS motif, are the key components for performing the cleavage function.

Therefore, the design mainly focuses on the Domain 1 region and replaces the water-soluble amino acids in it to improve its hydrophobicity and make it stable to generate spontaneous folding.

STEP 2

Select an approach to make sequence variants

Plan 1: METRFPQQSQQTPASTNRRRPFKHEDYPCRRQQRSSTLLVLIFLAIFLSLFTNQLLLSLLEAVIRTVTTLQQLLT

Design Actions: According to the LLR score, the K mutation at position 50, changing to L, yields a score of 2.56. Replacing with L (leucine) significantly enhances the hydrophobic anchoring force at the Domain 2/4 junction. Changing the Y at position 39 to L, located in Domain 1, yields a score of 2.24. Y contains a polar hydroxyl group; replacing it with L makes the transmembrane helix purer and more stable.

Plan 1.5: METRFPQQSQQTPASTNRRRPFKHEDYPRRRQQRSSTLLVLIFLAIFLSLFTNQLLLSLLEAVIRTVTTLQQLLT

Design Actions: According to the LLR, the C mutation, changing to R, yields a score of 2.39.

Principle: Increasing positive charge enhances the protein’s autonomous attraction to the negatively charged cell membrane, thereby reducing dependence on DNAJ escort.

Plan 2: METRFPQQQQQTPASTNRRRPFKHEDYPRRRQQRSSTLLVLIFLAIFLSLFTNQLLLSLLEAVIRTVTTLQQLLT

Design Action: According to LLR, mutating S to Q results in a score as high as 2.39. Principle: Increasing the rigidity of Domain 1 allows it to fold into a helical state.