What if we could record everything that happens inside a cell, so we can just look at it later?
Well, GEMINI does exactly that! Computationally designed proteins, molecular recorders, and plenty of experiments. This paper has it all! Is it enough to unveil the mysteries of cells’ dynamic life?
Don’t keep this newsletter a secret: Forward it to a friend today!
Was this email forwarded to you? Subscribe here!
Cells Build 3D Memory
Researchers created GEMINI, an intracellular recording device that can record cellular events in a “treering”-like pattern. Image credits: Nature.
Cells are super dynamic.
They’re constantly changing their state, responding to internal or external cues, expressing the right molecules, moving to the best environment, or dealing with stress… The life of a cell is full of choices!
And it gets even more complicated, because cells talk to each other.
Cell communication is crucial for tissue development, organ function, and everything in between! Cell signalling must be perfectly coordinated to prevent disease and developmental problems.
Of course, scientists asked: Can we map all this complicated activity?
Imagine following the changes in a cell population, in a whole tissue, over time… Scientists dream of this! It would help us explain development, disease progression, and drug response.
We have incredible tools, but they all have limitations:
Live imaging → real-time, but limited to a few signals
RNA sequencing → powerful, but destructive, and you only get a snapshot
What you want is a tool that can reconstruct the history of individual cells, in intact tissues, with spatial and temporal precision.
Recording Events with DNA and Proteins
A solution? Intracellular recorders.
These are molecular systems that sit inside cells and physically record what’s happening. The most common ones use nucleic acids to store events, but they’re often slow and require tissue destruction for readout → you lose spatial info.
An alternative? Proteins.
Protein assemblies are easy to design and image. The idea is simple: the assembly grows, and when something happens, a fluorescent subunit is added. Later, you image the structure and read out where and when something happened!
The problem? Most are linear → hard to read if they’re tilted or out of plane. Plus, they can become big and mechanically annoying for the cell. Can we fix these problems and keep the imaging power?
Enter GEMINI
Enter today’s paper!
The authors introduced GEMINI (granularly expanding memory for intracellular narrative integration). With this cool name, GEMINI is a protein-assembly-based intracellular memory device that sidesteps the limitations of linear recorders.
How?
GEMINI records a cell’s history as a tree-ring-like fluorescent pattern inside living cells! It goes 1D → 3D: instead of growing into a linear assembly, its subunits create a 3D lattice. The result is beautiful intracellular octahedra.
This solves two problems at once:
Insensitive to tilt: Because the structure grows the same in 3 dimensions, GEMINI can be imaged from any angle without losing info
No mechanical disturbances: The final particles are smaller than 1D assemblies, so they disturb the cells less → no impact on cell behaviour
And it’s not just pretty. The system achieves hour-level resolution, fast recording times (down to 15 minutes!), and it even works in xenografts and the mouse brain! Crazy stuff.
What is GEMINI?
GEMINI has 3 parts:
Blank subunits for predictable assembly
Timestamp subunits that mark when the assembly has grown to a certain stage → a fluorescent dye binds to them, marking the time they were added
Reporter subunits that encode the biological event being recorded → with a fluorescent marker, they are produced only when a specific event happens
The authors used computational design + experimental screening to develop proteins that assemble into intracellular 3D structures. They tested over 30 designs in vitro (even more in silico!), before selecting the one with the traits they needed:
started nucleation early
all at the same time
formed one assembly per cell
Perfect for clean recording!
Now that they had a working scaffold, the team built the expression system. The recorder is induced with doxycycline, and timestamping is based on the HaloTag system. When turned on, GEMINI nucleates in ~4 hours, and the growth plateaus at ~10 hours.
The workflow looks like this:
You induce GEMINI expression in cells
You wait, and when a GEMINI nucleus is formed, you add the first timestamp, creating a fluorescent layer and marking the time (for example, hour 1)
Then you record what you want, creating an event layer with a different fluorescence
You can add more timestamps or events, creating different colored layers at different times
Once you finish with your experiments, you fix the cells and image them with fluorescence microscopy.
The rings in GEMINI tell you when a recording happens, with single-cell resolution, just like tree rings!
What Can GEMINI Measure?
After a deep study of the system (they even developed a mathematical model of GEMINI’s growth!), the team tested GEMINI’s capacities, in cells and even in vivo! Because the real question is: what can it do?
Cell Culture Recordings
The first recordings were done in cell cultures.
The biological proof-of-concept focused on NF-κB, a transcription factor linked to inflammation and activated by TNF. The team built a cell line in which NF-κB activates reporter expression and is recorded in GEMINI.
When TNF is added at different times, GEMINI records the expected timing! There is a ~2-hour delay because transcription and translation take time. For example, for TNF added at 3 hours, the reported time is 4.82 ± 1.35 h.
And the team could also record signal-off events by removing TNF at defined times!
In Vivo Recording
This part was incredible! The authors also tested what happens in vivo, with not one, but two models.
Xenograft: Grafted Tumor Models
The team tested GEMINI in a mouse xenograft model of inflammation. They implanted GEMINI-expressing cells under the skin, induced the recorder, and then LPS triggered inflammation. GEMINI recorded the inflammation signal, with a clear dose-response! This means that GEMINI can successfully quantify inflammatory intensity in tissues. And they could even distinguish the differences in response between highly or poorly vascularized regions!
Crossing the blood-brain barrier
Finally, the team delivered GEMINI using a viral system directly into mice's brains. GEMINI still works, nucleating efficiently, but more slowly. And, importantly, without harming the brain or the mice! Amazing biocompatibility.
And GEMINI is still functional. The authors recorded seizure-induced protein activation, with GEMINI showing bands at the right time, capturing activity-dependent gene expression in vivo!
Molecular Recorders: Cellular Memory Devices
An incredible paper!
The idea of physically recording what’s happening inside cells and later collecting it is fascinating. And it’s so useful! I’m sure there are so many cellular processes we know nothing about just because we can't measure them…
And GEMINI can for sure help! It’s fast, reliable, and extremely biocompatible! The authors found only minor changes in cell culture and in mouse brains, with no functional effects. It’s still not a perfect system:
Growth is not completely synchronized → differences between cells
Delays for the transcriptional events
You could get different responses in different cells!
But, it’s still an amazing paper; go and read it here!
If you made it this far, thank you! What do you think of molecular recorders? Do you think they have a place in biomedicine? Reply and let me know!
P.S: Know someone interested in SynBio? Share it with them!
What did you think of today's newsletter?
More Room:
Trapping Cells in DNA: Normally, DNA is trapped inside cells. What happens when the opposite is true? Single-cell encapsulation creates controlled microenvironments for individual cells, but existing materials often lack precision and versatility. This review highlights DNA nanomaterials as a powerful alternative, offering programmable structure, biocompatibility, and modular functionality. It summarizes how DNA-based systems enable precise cell-surface encapsulation, their applications in bioanalysis and cell therapy, and the key challenges for future development.
DNA Walks on Cells: DNA walkers are small molecular machines that “walk” between different targets. I find them funny, but they are also useful. Detecting cancer-related miRNAs is difficult due to their low abundance and similarity. In this study, the authors develop a DNA walker-based biosensor on gold nanoparticles that simultaneously detects two miRNAs. The system uses a triple-input AND logic gate, requiring both miRNAs and a tumor enzyme to activate signal amplification, ensuring high specificity and low false positives.
RNA Is Growing Up: RNA therapeutics have come a long way, from simple small RNAs to diverse molecules that regulate multiple cellular processes. This review presents a design framework combining chemistry, structural biology, and computation to improve RNA stability, reduce immunogenicity, enhance efficacy, and enable targeted delivery. It highlights successes like mRNA vaccines and outlines strategies for developing safer and more effective RNA-based medicines.