01 / The hypothesis
Could LK11
protect the cells?
We wanted to test whether LK11, a short peptide, could protect cells from damage associated with amyloid fibres—protein molecules assembled into long aggregates. Our starting hypothesis was that LK11 might interfere with fibre growth.
We used insulin-derived amyloid material and C2C12 mouse muscle cells as an experimental model. The question was whether adding LK11 changed cell mortality in this culture system.
Working hypothesis · not a demonstrated mechanism
aggregationGrowing
fibersCell
damage?
LK11 → proposed inhibition of fiber growth
Test the components separately before interpreting their combination.
02 / The measurement
From fluorescence
to a cell count.
We used two fluorescent stains to turn microscopy images into a comparison. Hoechst made cell nuclei visible in blue; propidium iodide (PI) marked cells with compromised membranes. We counted the cells with ImageJ, an image-analysis tool, and calculated the fraction positive for PI.
NPI-positive: counted membrane-compromised cells. Ntotal: all counted cells.
This readout measures membrane damage. Staining, image acquisition and separating clustered nuclei during counting affect the estimate; it does not tell us which molecular interaction caused that damage.
We compared insulin with and without aggregates, LK11 on its own, and their combination. The medium-and-buffer control established a baseline; testing each component separately let us distinguish an effect of LK11 from an effect of insulin.
The six conditions and what they isolate
Final concentrations for the initial comparison. µM means micromoles per litre.
- Control
- Medium and buffer; the comparison baseline.
- Soluble insulin
- 63 µM insulin without added aggregates.
- Soluble + aggregated insulin
- 63 µM soluble and 18 µM aggregated insulin.
- Mixed insulin + LK11
- The same insulin mixture with 25 µM LK11.
- LK11 alone
- 25 µM LK11 without insulin.
- Aggregated insulin
- 18 µM aggregated insulin without soluble insulin.
03 / The surprise
The strongest toxicity
came from LK11 alone.
The controls challenged our starting hypothesis. Cells exposed to LK11 alone showed the highest observed mortality, while the insulin-only conditions were much lower. Before asking whether LK11 protected cells, we needed to understand its own toxicity.
Adding insulin to LK11 was associated with lower mortality than LK11 alone. That raised a different question about their interaction.
04 / The investigation
Handling first.
Concentration next.
The unexpected result prompted two separate checks: whether the order of addition changed mortality, and how the response varied with LK11 concentration.
The same final concentration, a different exposure.
Adding LK11 before the medium produced much higher mortality than adding it afterwards. Both conditions had a final LK11 concentration of 25 µM.
Approximately 90% versus 16% mortality, estimated visually from the two bars.
Before dilution into the medium, cells may briefly encounter a higher local LK11 concentration. This could explain why preparation matters even when the final concentration is identical; that local exposure was not measured directly.
A separate dose-response experiment.
We then examined concentration separately. Across the tested 0–25 µM range, mortality increased at higher LK11 concentrations. The two measurement series and their mean show the dose-related trend.
The initial controls, addition-order check and concentration series are separate experiments. Their preparation differed, so their values cannot be pooled into a single dose-response curve.
How we summarised the concentration response
c: LK11 concentration in µM, over the tested 0–25 µM range. M(c): fitted mortality (%).
The curve summarizes these observations. It is neither a biological law nor a validated toxicity threshold, and is not extrapolated beyond the tested range.
05 / The revised hypothesis
Could insulin reduce
LK11’s toxic effect?
The initial comparison suggested another direction: could insulin change LK11’s effect? In a follow-up experiment with 50 µM LK11, we varied the added soluble insulin. Mortality was lower with added insulin than with LK11 alone.
The result supports investigating an interaction. It does not establish how the molecules interact.
Insulin might bind LK11 or reduce how much of it remains available to affect cells. These are possible explanations to test; the mortality measurements alone do not establish binding or complete neutralisation.
06 / What changed
A revised question.
An unresolved mechanism.
We began by looking for protection from amyloid-associated damage. Testing the components separately redirected the work towards LK11 toxicity, the effect of preparation and a possible interaction with insulin. The controls changed the question we needed to ask.
The next step would be to control exposure during preparation and measure the interaction directly. These cell-culture results leave the molecular mechanism unresolved and do not establish a treatment or inhibition of fibre growth.
Supporting evidence: localisation and insulin controls
We also examined where labelled LK11 appeared and how cells responded to soluble, aggregated and combined insulin. These comparisons help interpret the mortality response, but do not identify a molecular mechanism.
The team
Experimental Protocol Design · Colin Schotte, Julien Le Boulch, Maxime Raffort, Nathan Luneau.
The academic foundations