Week 2 HW

🧬 BioColdTrace — Homework #2

Part 1: Define your expression target

What exactly should the cells make?

In Homework #1, I proposed BioColdTrace, a biological Time–Temperature Indicator designed to transform cumulative thermal exposure into a persistent visible signal.

My original prototype was intentionally simple:

🦠 Yeast + nutrient source + colorimetric reporter

For this homework, I want to move from that general concept to a genetically defined output.

Instead of leaving the colorimetric response open, I would engineer Saccharomyces cerevisiae to produce a visible chromoprotein whose accumulation could eventually act as the readable output of the indicator.

1. 🎨 What molecule am I expressing?

The expression target I would use is amilCP, a blue-purple chromoprotein originally identified in the coral Acropora millepora.

Chromoproteins are particularly interesting for BioColdTrace because they produce a visible color without requiring fluorescence excitation or specialized instrumentation.

amilCP has an absorption maximum around 588 nm, giving it a strong blue-purple appearance. [1,2]

This makes it attractive for a system whose final output should ideally be readable simply by looking at it.

Why amilCP?

For BioColdTrace, I want the output to eventually satisfy several requirements:

• 👁️ Visible without specialized equipment
• 🧬 Genetically encoded
• 📈 Able to accumulate over time
• 📦 Compatible with a simple contained indicator
• 🔬 Still measurable quantitatively during development

amilCP fits these criteria particularly well because the reporter itself is colored.

This means the first engineered architecture could conceptually look like:

🌡️ Thermal exposure → 🦠 biological response → 🧬 amilCP expression → 🎨 accumulated visible color

Importantly, amilCP would not itself be the temperature sensor.

It would be the output or recording molecule.

The mechanism determining when amilCP is expressed would be designed separately through the promoter and regulatory architecture.

This separation is useful because it allows me to independently optimize:

temperature sensing → gene regulation → visible output

The accumulation of a colored protein is also conceptually compatible with the idea of biological memory proposed in Homework #1.

If warmer conditions trigger reporter production, returning the system to refrigeration should slow further biological activity without immediately eliminating the protein that has already accumulated.

2. 🧬 What is the sequence?

The protein I would encode is amilCP from Acropora millepora.

Database information

Amino acid sequence

MSVIAKQMTYKVYMSGTVNGHYFEVEGDGKGKPYEGEQTVKLTVTKGGPLPFAWDILSPQCQYGSIPFTKYPEDIPDYVKQSFPEGYTWERIMNFEDGAVCTVSNDSSIQGNCFIYHVKFSGLNFPPNGPVMQKKTQGWEPNTERLFARDGMLLGNNFMALKLEGGGHYLCEFKTTYKAKKPVKMPGYHYVDRKLDVTNHNKDYTSVEQCEISIARKPVVA

For an actual DNA construct, I would use this protein sequence as the reference and generate a coding sequence codon-optimized for Saccharomyces cerevisiae.

Therefore:

Native coral sequence → protein reference
Codon-optimized DNA sequence → yeast expression construct

The amino acid sequence would remain unchanged while the DNA sequence could be adapted to improve expression in the yeast chassis.

This distinction is important because the biological target is amilCP, while the exact nucleotide sequence used in the final construct can be engineered according to the codon preferences and design constraints of S. cerevisiae.

3. 🦠 What host or chassis?

I would keep Saccharomyces cerevisiae as the biological chassis proposed in Homework #1.

I originally selected yeast because:

• it is inexpensive and easy to culture;
• it grows on relatively simple media;
• it is one of the best-characterized organisms in biotechnology;
• extensive genetic and synthetic-biology tools are available;
• its metabolism and growth are strongly affected by temperature;
• it has already been explored in biosensing applications.

For the engineered version of BioColdTrace, I would narrow the chassis further and initially use:

Saccharomyces cerevisiae CEN.PK113-7D

CEN.PK113-7D is a haploid laboratory strain widely used in metabolic engineering, systems biology, and industrial biotechnology. Its genome has also been sequenced and characterized, making it a useful reference strain for quantitative engineering studies. [3]

Using a well-characterized strain is useful because I ultimately want BioColdTrace to establish a quantitative relationship between:

🌡️ temperature × ⏱️ exposure time × 🦠 cellular response × 🎨 reporter accumulation

A defined genetic background should make those relationships easier to study than starting with an undefined commercial yeast strain.

However, I would still consider the chassis an experimental variable.

If CEN.PK113-7D shows insufficient activity at refrigeration or intermediate temperatures, other S. cerevisiae strains adapted to lower temperatures could later be screened.

This would remain consistent with my original BioColdTrace proposal, where chassis selection itself was identified as one of the important questions to explore.

4. 📍 Where does the product need to go?

For the first BioColdTrace construct, amilCP would remain intracellular, primarily in the cytoplasm.

Therefore, I would not include a secretion signal peptide or organelle-targeting sequence.

The first architecture would simply be:

🦠 S. cerevisiae CEN.PK113-7D
↓
🧬 intracellular expression of amilCP
↓
🎨 accumulation of blue-purple color

Keeping the protein intracellular simplifies the first prototype because BioColdTrace does not require the reporter to perform any extracellular function.

Instead, the yeast cells themselves could act as small pigment-producing units.

There may also be an advantage for a future contained indicator: the reporter would remain associated with the biological material instead of needing to diffuse into the surrounding environment.

At this stage, I would therefore avoid adding unnecessary secretion or targeting sequences and keep the first construct as simple as possible.

🧠 What does this mean for BioColdTrace?

Defining amilCP solves one important part of the system:

What should the cells make?

The answer would now be:

A visible blue-purple chromoprotein whose accumulation can act as the readable output of BioColdTrace.

But this immediately creates the next — and more interesting — engineering question:

When should the cells make it?

The system should not simply become purple because yeast has been alive for a certain amount of time.

Instead, amilCP production should ultimately depend on thermal history.

For example, BioColdTrace should ideally be able to distinguish between conditions such as:

❄️ 48 h at 4 °C

and

🌡️ 4 h at 25 °C

even though both involve the passage of time.

The challenge is therefore not simply to express a colored protein.

The challenge is to create a regulatory architecture where temperature and exposure time determine how much reporter accumulates.

In that sense, the system can be separated into three modules:

🌡️ 1. Input

Temperature history

↓

🧬 2. Processing

Temperature-sensitive biological regulation

↓

🎨 3. Output

amilCP accumulation

This modular architecture is particularly useful because each component could potentially be optimized independently.

If amilCP works well as an output but the temperature-response mechanism does not, I could modify the regulatory system without replacing the reporter.

Likewise, if a better visible reporter becomes available later, the output module could be exchanged without completely redesigning the sensing mechanism.

Ultimately, the objective would be to transform:

thermal profile → biological response → reporter accumulation → visible warning

This is the first step toward turning the original BioColdTrace concept from:

“yeast + nutrients + color”

into a defined synthetic biological system.

Part 2: Design your genetic parts

How do I turn amilCP from a protein sequence into an expression system that yeast can actually use?

In Part 1, I defined amilCP as the visible output of BioColdTrace and Saccharomyces cerevisiae as the chassis.

The next step is to place that protein inside a complete yeast expression architecture.

For the first construct, I would deliberately separate two problems:

1. Can yeast produce enough amilCP to generate a useful visible signal?
2. Can that expression later be controlled by thermal history?

Trying to solve both questions simultaneously would make it difficult to determine why a construct succeeds or fails.

For that reason, my first construct would be a constitutive reporter-validation construct.

🧬 Proposed expression architecture

pTDH3 → 5′ UTR / Kozak → amilCP → CYC1 terminator

carried on a:

pRSII41K low-copy yeast backbone

The complete conceptual architecture is therefore:

🧬 pTDH3
↓
🔤 yeast 5′ UTR / Kozak
↓
🎨 amilCP CDS
↓
🛑 CYC1 terminator + 3′ UTR / poly(A) processing region
↓
⭕ pRSII41K — CEN6/ARS4 + kanMX/G418

No secretion signal or organelle-targeting peptide would be included because amilCP is intended to remain intracellular.

1. ▶️ Transcription start — TDH3 promoter

Part information

I would use the S. cerevisiae TDH3 promoter, also commonly called the GPD promoter.

TDH3 encodes glyceraldehyde-3-phosphate dehydrogenase, and its promoter is widely used for strong constitutive gene expression in S. cerevisiae.

For this first BioColdTrace construct, strong constitutive expression is useful because I first need to answer a fundamental question:

Can accumulation of amilCP generate enough visible color in yeast to function as an output?

The promoter sequence can be imported from an existing yeast expression vector or from the S. cerevisiae TDH3 upstream region.

YEASTRACT reports the −1000 to −1 upstream sequence of TDH3:

CTATTTTCGAGGACCTTGTCACCTTGAGCCCAAGAGAGCCAAGATTTAAATTTTCCTATG
ACTTGATGCAAATTCCCAAAGCTAATAACATGCAAGACACGTACGGTCAAGAAGACATAT
TTGACCTCTTAACAGGTTCAGACGCGACTGCCTCATCAGTAAGACCCGTTGAAAAGAACT
TACCTGAAAAAAACGAATATATACTAGCGTTGAATGTTAGCGTCAACAACAAGAAGTTTA
ATGACGCGGAGGCCAAGGCAAAAAGATTCCTTGATTACGTAAGGGAGTTAGAATCATTTT
GAATAAAAAACACGCTTTTTCAGTTCGAGTTTATCATTATCAATACTGCCATTTCAAAGA
ATACGTAAATAATTAATAGTAGTGATTTTCCTAACTTTATTTAGTCAAAAAATTAGCCTT
TTAATTCTGCTGTAACCCGTACATGCCCAAAATAGGGGGCGGGTTACACAGAATATATAA
CATCGTAGGTGTCTGGGTGAACAGTTTATTCCTGGCATCCACTAAATATAATGGAGCCCG
CTTTTTAAGCTGGCATCCAGAAAAAAAAAGAATCCCAGCACCAAAATATTGTTTTCTTCA
CCAACCATCAGTTCATAGGTCCATTCTCTTAGCGCAACTACAGAGAACAGGGGCACAAAC
AGGCAAAAAACGGGCACAACCTCAATGGAGTGATGCAACCTGCCTGGAGTAAATGATGAC
ACAAGGCAATTGACCCACGCATGTATCTATCTCATTTTCTTACACCTTCTATTACCTTCT
GCTCTCTCTGATTTGGAAAAAGCTGAAAAAAAAGGTTGAAACCAGTTCCCTGAAATTATT
CCCCTACTTGACTAATAAGTATATAAAGACGGTAGGTATTGATTGTAATTCTGTAAATCT
ATTTCTTAAACTTCTTAAATTCTACTTTTATAGTTAGTCTTTTTTTTAGTTTTAAAACAC
CAAGAACTTAGTTTCGAATAAACACACATAAACAAACAAA

Source: YEASTRACT+, TDH3/YGR192C promoter sequence.

Why this part?

I chose pTDH3 because it provides strong constitutive expression in S. cerevisiae, allowing the first construct to test whether amilCP can accumulate to a visually detectable level before introducing temperature-responsive regulation.

Important design decision

pTDH3 is not intended to be the final temperature-sensing element of BioColdTrace.

Instead:

pTDH3 construct = reporter calibration

Later:

temperature-responsive regulation = BioColdTrace sensing layer

This gives me a baseline against which future temperature-responsive constructs can be compared.

2. 🔤 Translation start — yeast 5′ UTR + Kozak context

Part information

Unlike bacteria, yeast does not use a bacterial ribosome binding site or Shine–Dalgarno sequence.

Instead, translation initiation depends strongly on the sequence surrounding the ATG start codon.

A well-established Kozak-like sequence for S. cerevisiae is:

AAAAAAATGTCT

Because the ATG belongs to the coding sequence, I would define the separate 5′ UTR / Kozak part in Kernel as:

AAAAAA

and design the amilCP CDS to begin with:

ATGTCT...

When assembled, the junction becomes:

AAAAAAATGTCT...
||||||--- CDS ---
 5'UTR

Why this part?

I chose an A-rich yeast Kozak context to provide an efficient and host-appropriate translation initiation environment immediately upstream of the amilCP start codon.

3. 🎨 Coding sequence — amilCP

Part information

The coding sequence would encode the same amilCP protein defined in Part 1.

The reference protein sequence is:

MSVIAKQMTYKVYMSGTVNGHYFEVEGDGKGKPYEGEQTVKLTVTKGGPLPFAWDILSPQCQYGSIPFTKYPEDIPDYVKQSFPEGYTWERIMNFEDGAVCTVSNDSSIQGNCFIYHVKFSGLNFPPNGPVMQKKTQGWEPNTERLFARDGMLLGNNFMALKLEGGGHYLCEFKTTYKAKKPVKMPGYHYVDRKLDVTNHNKDYTSVEQCEISIARKPVVA

The natural amilCP sequence originates from the coral Acropora millepora and is available as GenBank AY646075.1 / AAU06854.

For my final Kernel part, however, I would create a derived DNA sequence codon-optimized for S. cerevisiae while preserving this amino acid sequence.

Therefore:

AAU06854 protein sequence
↓
codon optimization for S. cerevisiae
↓
amilCP_ySC CDS

The CDS must include:

• an ATG start codon;
• the complete amilCP open reading frame;
• a stop codon.

Sequence note

The exact nucleotide sequence recorded here should be the final codon-optimized DNA sequence generated/imported in Kernel, rather than a manually reverse-translated sequence.

Final Kernel sequence:

[PASTE FINAL amilCP_ySC CODON-OPTIMIZED SEQUENCE EXPORTED FROM KERNEL HERE]

This ensures that the sequence documented on the website is identical to the sequence actually used in the genetic design.

Why this part?

I chose amilCP because it produces a directly visible blue-purple chromoprotein, allowing reporter accumulation to be observed without fluorescence excitation or specialized instrumentation.

4. 📍 Targeting — none

Part information

No secretion signal, transmembrane domain, or organelle-targeting peptide would be added.

amilCP does not need to leave the cell to perform its role in BioColdTrace.

Instead:

yeast cell → intracellular amilCP accumulation → visible cellular color

Keeping the protein cytoplasmic also makes the initial construct simpler and avoids introducing secretion efficiency as another experimental variable.

Why no targeting sequence?

No targeting sequence is required because amilCP is intended to accumulate intracellularly as the visible reporter, and secretion would add unnecessary complexity to the first prototype.

5. 🛑 Transcription stop — CYC1 terminator

Part information

I would use the S. cerevisiae CYC1 terminator.

CYC1-derived 3′ sequences are widely used in yeast expression vectors and provide the signals required for proper transcript termination and 3′-end processing.

The sequence used is:

ACAGGCCCCTTTTCCTTTGTCGATATCATGTAATTAGTTATGTCACGCTTACATTCACGCCCTCCTCCCACATCCGCTCTAACCGAAAAGGAAGGAGTTAGACAACCTGAAGTCTAGGTCCCTATTTATTTTTTTTAATAGTTATGTTAGTATTAAGAACGTTATTTATATTTCAAATTTTTCTTTTTTTTCTGTACAAAC

Why this part?

I chose the CYC1 terminator because it is a well-established S. cerevisiae 3′-end element that provides reliable termination and transcript processing in yeast expression constructs.

6. ⭕ Vector context — pRSII41K

Instead of designing a yeast plasmid backbone from scratch, I would reuse an existing backbone.

Selected backbone

pRSII41K / pRSII41K-TDH3pr

Addgene plasmid #194522 / #194528

The particularly convenient version for this project is:

pRSII41K-TDH3pr — Addgene #194528

because it already contains:

TDH3 promoter → MCS → CYC1 3′ UTR

Part information

The backbone uses the CEN6/ARS4 system, producing a relatively low-copy plasmid in yeast.

For BioColdTrace, I prefer a low-copy backbone over a high-copy 2µ plasmid for the first quantitative experiments because I ultimately want to compare:

temperature → biological activity → reporter accumulation

and reducing plasmid copy-number effects should make expression easier to interpret.

The kanMX marker also provides G418 resistance for selection of transformed yeast.

Why this backbone?

I chose pRSII41K because it is a low-copy S. cerevisiae expression backbone with convenient G418 selection, and an existing version already provides the TDH3 promoter and CYC1 3′ UTR needed for this expression cassette.

🧩 Final part list

🧬 Final construct

The first BioColdTrace genetic construct can therefore be summarized as:

                    BioColdTrace v1
                    Reporter module

      ┌────────┐    ┌──────────┐    ┌───────────┐    ┌────────┐
      │ pTDH3  │ →  │ yKozak   │ →  │  amilCP   │ →  │ tCYC1  │
      │promoter│    │  5' UTR  │    │    CDS    │    │ 3' UTR │
      └────────┘    └──────────┘    └───────────┘    └────────┘
                                          │
                                          ▼
                                  🎨 blue-purple color

              on pRSII41K — CEN6/ARS4 — kanMX

or more simply:

🧬 pTDH3 → 🔤 yKozak → 🎨 amilCP → 🛑 tCYC1 → ⭕ pRSII41K

🌡️ Why is the promoter constitutive if BioColdTrace should sense temperature?

This is an intentional design decision.

The final BioColdTrace system should ultimately make amilCP accumulation dependent on thermal history, but I first need to characterize the reporter independently of that sensing mechanism.

The first construct therefore asks:

Can my selected yeast chassis produce enough amilCP to generate a measurable and visible output?

If the answer is yes, this construct provides a baseline response.

A later version can replace or regulate the constitutive transcriptional input with a temperature-responsive mechanism:

BioColdTrace v1

constitutive expression
        ↓
Does the reporter work?
        ↓
amilCP characterization
        ↓
baseline

followed by:

BioColdTrace v2

thermal exposure
        ↓
temperature-responsive regulation
        ↓
amilCP accumulation
        ↓
persistent visible signal

This modular approach lets me debug the system one layer at a time rather than trying to determine simultaneously whether a failure comes from the temperature sensor, the regulatory architecture, or the chromoprotein reporter.

Ultimately, the goal remains:

thermal history → biological computation → accumulated visible memory

💻 What I would create in Kernel

I would create or import the following records:

1. pTDH3

   • Type: Promoter
   • Source: S. cerevisiae TDH3 promoter

2. yKozak_A6

   • Type: 5′ UTR
   • Sequence: AAAAAA

3. amilCP_ySC

   • Type: CDS
   • Source protein: AAU06854 / Q66PV0
   • Designed: codon-optimized for S. cerevisiae

4. tCYC1

   • Type: Terminator
   • Source: S. cerevisiae CYC1 terminator

5. pRSII41K

   • Type: Backbone
   • Source: Addgene #194522 / #194528

I would then assemble them in the following order:

pTDH3 → yKozak_A6 → amilCP_ySC → tCYC1

inside the pRSII41K backbone.

📚 References

[1] Alieva, N. O., Konzen, K. A., Field, S. F., et al. (2008). Diversity and evolution of coral fluorescent proteins. PLoS ONE, 3(7), e2680.
https://doi.org/10.1371/journal.pone.0002680

[2] NCBI GenBank. Acropora millepora chromoprotein mRNA, complete cds. Accession AY646075.1.
https://www.ncbi.nlm.nih.gov/nuccore/AY646075

[3] FPbase. amilCP. GenBank AAU06854; UniProtKB Q66PV0.
https://www.fpbase.org/protein/amilcp/

[4] YEASTRACT+. TDH3 / YGR192C — Saccharomyces cerevisiae promoter sequence.
https://yeastract.com

[5] Dao Duc, K., Song, Y. S. (2018). The impact of ribosomal interference, codon usage, and exit tunnel interactions on translation elongation rate variation. PLoS Genetics. [Used here together with published analyses of the S. cerevisiae Kozak-like initiation context.]

[6] Osborne, B. I., & Guarente, L. (1989). Mutational analysis of a yeast transcriptional terminator. Proceedings of the National Academy of Sciences, 86(11), 4097–4101.
https://doi.org/10.1073/pnas.86.11.4097

[7] Smith, J. R., et al. (2024). Auxotrophy-Independent Plasmid Shuttle Vectors for Applications in Diverse Yeasts. Applied Microbiology, 4(1), 453–469.
https://doi.org/10.3390/applmicrobiol4010031

[8] Addgene. pRSII41K-TDH3pr. Plasmid #194528.
https://www.addgene.org/194528/