Our project leverages a sophisticated bioinformatics pipeline to design and evaluate synthetic peptides aimed at selectively binding the intrinsically disordered C-terminal region of progerin, a truncated form of lamin A implicated in Hutchinson-Gilford Progeria Syndrome (HGPS). Artificial Intelligence (AI)-based platforms like AlphaFold combined with generative models such as RFdiffusion and ProteinMPNN, have paved the way for new approaches in protein design, particularly in targeting intrinsically disordered regions (IDRs). In our work, we integrated these AI-based models with molecular docking platforms such as HADDOCK and ClusPro, which enable us to simulate and assess potential interactions between designed peptides and progerin.

The pipeline begins with structural modeling of progerin, followed by the design of peptide binders using RFdiffusion and ProteinMPNN, which generate sequences expected to fold into specific structural motifs. These peptides are then subjected to docking simulations, where their interaction with progerin is evaluated. The integration of advanced tools like AlphaFold3 allows us to model even disordered regions, which are notoriously difficult to study experimentally. The results of docking simulations are further refined through affinity estimation using PRODIGY, providing us with valuable insights into the binding strength of each candidate. This multi-step, AI-driven pipeline provides a powerful framework for designing selective peptide binders for progerin, a critical step in developing therapeutic strategies for HGPS.

Our computational pipeline enabled the exploration of a broad library of designed peptides to assess their potential binding to the flexible C-terminal tail of progerin. We identified candidates displaying consistent and promising binding behavior. Our workflow illustrates how rational design, sequence analysis, and energy-based evaluation can effectively guide the selection of peptide binders with high confidence for subsequent experimental validation.
More information about our pipeline design in our wiki Model page.

Our docking analysis revealed that our binders exhibited heterogeneous binding patterns. Some peptides designed with RFdiffusion on the full-length structure of progerin only weakly interacted with the C-terminal region or did so in clusters of limited size, suggesting that these interactions were not significant. In contrast, other binders consistently engaged the C-terminal tail and produced several clusters of considerable size, indicating reproducible and reliable interactions. This agreement with hotspot-driven docking adds confidence to the overall analysis.

For the binders modelled on individual C-terminal fragments, the picture is more complex. Since IDRs can adopt multiple conformations, the AlphaFold3 model of full-length progerin represents only one possible conformation. This means that the lack of interaction in docking simulations does not necessarily indicate a binder would fail, but rather that the interaction may require or induce alternative conformations of the protein to engage properly.

Based on these observations, we selected for affinity prediction only the binders that displayed interactions in proximity to the last 15 residues of their fragment, corresponding to the hotspot region defined during design with RFdiffusion. In addition, binders that also docked at the C-terminal tail of full-length progerin were considered as particularly promising candidates, as they suggested a stronger and more consistent affinity for that region. In total, we analyzed approximately 350 interactors; the sequences are provided in the supplemental materials.

Application of other techniques for binding design

Furthermore, for the binders designed against full-length progerin, the fifteen sequences retained from ProteinMPNN were subjected to multiple sequence alignment (MSA), revealing which positions were conserved and which varied across different designs.
Stability analyses were then performed using the Stabilize Protein protocol in ROSIE, which suggested specific substitutions to increase the thermodynamic stability of the candidates. From this process, a consensus sequence was derived and designated as LOGO, in which certain residues were substituted to enhance stability, while others were replaced according to the consensus from the alignment. LOGO represents our rational design approach, using a consensus sequence optimized based on previous observations (Figure 2).

Insights into LOGO: Rational Design and Interaction Analysis with progerin
Figure 2. Insights into LOGO, the interactor obtained through rational design. A: Schematic illustrating how the LOGO sequence was generated for rational design, by analyzing the most frequent amino acids and their properties. B and C: Analysis of the interaction between progerin and the LOGO interactor.

In addition to the main computational pipeline, we employed NeuroBind, which was not part of our main workflow, but was used to explore additional candidates outside the RFdiffusion–ProteinMPNN design process. NeuroBind generated 25 candidate binders in a single run, which were subsequently inspected in ChimeraX. Candidates that produced unrealistic conformations of progerin were discarded. The remaining models were retained for downstream analysis.

After refinement of the docking complex, the binding affinities of the complexes were estimated with PRODIGY at 36 °C. For the HADDOCK results, the first model from the top three clusters ranked by HADDOCK score was analyzed for each binder. For the ClusPro docking against full-length progerin, all clusters in which the interaction involved the C-terminal region were evaluated, while for the fragment-based docking, only the first two clusters by size were considered. For each case, the predicted Kd values were averaged by calculating a weighted mean, in which the weight was given by the cluster size. This procedure yielded one mean Kd per binder from HADDOCK and one from ClusPro.

Insights from the selected models

Table 1 summarizes key properties of the five interactors selected for laboratory validation. The conformations of these interactors vary to explore which structural features correlate with optimal binding to the progerin C-terminus. 62aa11_1 adopts an α-helical structure and stands out for its strong binding affinity to the 62-residue C-terminal fragment, with a Kd of 4.90 × 10⁻⁹ M. LOGO and n80_02 are both globular interactors. LOGO was designed rationally and demonstrates a satisfactory binding affinity to the progerin C-terminus (7.50 × 10⁻⁸ M), while n80_02 was generated through our computational pipeline.

Rank 15 and Rank 21 were produced using NeuroBind, which predicts very high binding affinities. We included these to assess the potential of this bioinformatic tool. Their structures are less well defined and more disordered than the other interactors, reflected in relatively low pTM scores of 0.20 for Rank 15 and 0.32 for Rank 21.

Interactor pTM HADDOCK Kd Progerin Sequence Structure
62aa11_1 0.76 9.32E-07 GLEEAQRRAEEARRQIALANRAGRDQEEAARLQRELEALEAEIEEAKTG 62aa11_1 structure
LOGO 0.44 7.50E-08 APGRGRCRGNPPVCCCPNCPRCDADCTQGGGSGCPACPCP LOGO structure
n80_02 0.33 6.92E-08 AAGCGRCIGNPPVCCCCNCPECGQDCTQCGGSGCPNCPCP n80_02 structure
Rank 15 0.2 1.08E-17 GLPLPELDLPEAMFRGKCAQANGAGASGTTHTAPPPEPREPLSGE Rank 15 structure
Rank 21 0.32 5.15E-15 ACAGSPRNCPAPCTGTDCPPCPGPAFEGDTEKPGPGEPPRGGAGG Rank 21 structure
Table 1. Key information about the interactors tested in the laboratory, including their pTM (prediction confidence), the predicted Kd from HADDOCK for the interaction with the full-length progerin C-terminus, as well as the peptide sequence and structural features. Other interactors were interesting for further analysis, but we only had the possibility to test 6 of them. The sequences of all the other interactors are present in the supplemental materials section.

Conclusion

We developed a computational pipeline for the de novo design of peptide binders selectively targeting the C-terminal tail of progerin, combining structural modeling, sequence design, docking, and PRODIGY-based affinity prediction. From an initial library of 96 candidates, nine binders were selected with consistent engagement of the C-terminal region and estimated dissociation constants in the nanomolar range or lower.

While limited by single AlphaFold3 conformations and rigid docking, the approach demonstrates the feasibility of designing selective binders for intrinsically disordered regions. Future experimental validation in yeast and human cells will test these predictions and guide iterative optimization, supporting the design of effective interactors for the ProgERASE therapeutic strategy.

These results must be interpreted in light of recent advances in the field. In particular, two studies recently published on Nature and Science by the Baker Lab have introduced pipelines for designing binders against intrinsically disordered regions, which have long been regarded as inaccessible to rational design [1], [2] . While our strategy shares some core elements with these pipelines, such as the integration of deep generative models and docking simulations, their approaches explicitly account for the conformational ensemble of the disordered region, either by co-folding the target and binder during diffusion or by guiding design with multiple conformations and secondary structure propensities. In contrast, our method relies on AlphaFold3 snapshots of individual conformations combined with rigid docking, which inevitably restricts the exploration of the full structural heterogeneity of the IDR. These differences underline both the strengths and the limitations of our method. While it demonstrates the feasibility of designing selective binders for disordered regions, it also highlights the need for broader conformational sampling. In particular, AlphaFold3 predictions could have been expanded by using multiple random seeds, thereby capturing a larger ensemble of conformations of the disordered C-terminal tail, and secondary structure predictions could have been integrated to better approximate its conformational landscape. Moreover, more extensive sampling during binder design would likely have increased the diversity of viable candidates. Such refinements would bring our pipeline closer to the state of the art in IDR-targeted binder design and represent clear directions for methodological improvement in future work.

Laboratory Experimentation

These computational predictions will require rigorous experimental validation. First, the interaction between the designed peptides and progerin has been assessed using a yeast two-hybrid assay to detect direct interactions in a cellular context (visit our yeast wiki page). In parallel, the NanoLuc® Binary Technology (NanoBiT) complementation assay will be applied in human MRC-5 fibroblasts to confirm binding under physiologically relevant conditions. Finally, quantitative affinity measurements will be carried out by Microscale Thermophoresis (MST) and Spectral Shift on a Monolith X instrument (NanoTemper Technologies), enabling the precise determination of dissociation constants for the peptide–progerin complexes (visit our Mammalian Cell Validation wiki page. Together, these experiments will provide critical validation of our computational predictions and allow us to refine the design pipeline through an iterative cycle of in silico modeling and in vitro testing.

The entire phase 2 was carried out in parallel with, and subsequently to, the experimental testing of the interactors, due to the time constraints of the iGEM competition. The results presented from this point onward did not have a direct impact on the project, but they will become relevant if the study is continued.

Step 1 - Without RING domain

After predicting the interactors using AI-driven tools and trying to target specifically progerin, further analyses involve the measurement of the binding affinity with lamin A to inspect if our peptides can discriminate between the two proteins. To do that, we continued our analyses focusing on the first 9 interactors that showed the strongest binding affinity with progerin. First, we predicted the interactors using Alphafold3 a second time, in order to be sure to continue the analyses with the best structures possible as we wanted to refine and test effectively the binding capability of both lamin A and progerin. After obtaining the AlphaFold structures, we ran the relax cycle and we checked the structure integrity using MolProbity as we previously described. For each interactor, we maintained the structure that showed the best features and compromise between a low energy level and a good MolProbity score. The same was performed for lamin A.

Interactor Energy MolProbity score
Pre-Relax Post-AMBER Post-Rosie Post-AMBER 2.0 Post-AMBER Post-Rosie Post-AMBER 2.0
lamin A 33.167,06 -86.321,45 -65.282,53 -88.179,65 1,90 0,90 1,84
LOGO -716,32 -3.606,94 -3.001,67 -3.914,43 1,93 0,86 2,70
n80_02 12.057,20 -2.528,40 -1.272,80 -2.425,82 2,02 1,68 1,56
Rank 15 316,67 -2.755,90 -1.375,60 -2.868,68 1,46 1,88 1,99
Rank 21 7.667,93 -1.766,30 -329,59 -1.912,69 2,02 1,88 1,19
62aa11_1 -6.279,97 -9.645,16 -7.290,12 -9.855,37 1,94 0,53 2,81
52aa2_1 -2.907,52 -6.645,47 -5.453,73 -6.911,33 3,01 1,01 2,35
42-2L03 2.855,60 -1.668,46 -844,73 -1.712,35 1,73 0,50 2,22
n51_02 -618,09 -2.918,23 -1.947,73 -3.036,73 2,43 1,65 1,54
Rank 7 1.030,02 101,65 1.247,85 -43.38 1,46 1,19 1,83
Table 2. Relaxation results for the predicted most promising interactors. The table shows the energy levels and MolProbity scores of the structures after each relaxation step.

In the docking step of our pipeline, ClusPro results for lamin A docking showed no binding between the interactors and the region of interest in the C-terminus. HADDOCK results were useful for assessing the binding capability between the interactor and lamin A. After defining the interacting region, we focused on clusters with low HADDOCK scores and large cluster sizes, as these indicate significant and reliable interactions. We focused on clusters with a HADDOCK score lower than -35, and we measured the weighted average Kd, with the cluster size as weighting factor.

Lamin A Cluster information Prodigy
Cluster Haddock score Cluster size Kd [M] Weighted average
LOGO 2 -41.8 ± 3.1 22 8.60E-08 1.28E-07
3 -41.5 ± 4.6 16 3.80E-08
1 -36.9 ± 3.3 29 2.10E-07
n80_02 2 -38.2 ± 2.7 21 9.40E-09 3.49E-08
3 -31.2 ± 2.1 17 4.70E-08
1 -29.2 ± 3.7 80 3.90E-08
Rank 15 3 -22.3 ± 6.9 15 1.90E-06 9.10E-07
10 -20.3 ± 6.3 5 1.00E-06
Rank 21 2 -25.8 ± 2.2 17 9.70E-09 2.18E-07
5 -42.4 ± 2.1 9 3.50E-07
4 -35.1 ± 7.4 11 1.10E-07
62aa11_1 4 -40.6 ± 3.4 15 2.40E-08 1.73E-07
1 -40.6 ± 5.2 60 2.10E-07
52aa2_1 3 -34.0 ± 3.9 14 1.80E-06 1.80E-06
5 -40.9 ± 1.6 13 1.60E-07
42-2L03 2 -39.9 ± 2.6 39 6.40E-07 4.42E-07
4 -38.7 ± 0.5 20 2.40E-07
n51_02 6 -35.8 ± 1.3 6 1.30E-07 1.30E-07
Rank 7 8 -23.7 ± 3.6 6 1.30E-06 1.30E-06
Table 3. HADDOCK docking results between the interactors and lamin A. The final Kd was calculated as a weighted average of the Kd values from individual clusters, using cluster size as the weighting factor. For models with insufficient significant clusters, the first reliable cluster was retained.

We repeated the same procedure with progerin to improve our Kd predictions using these refined progerin structures. Table 7 shows the HADDOCK results for progerin. We notice that the best Kd value is now 5.70 × 10⁻⁹M, corresponding to Rank 21, which likely retains part of the optimistic Kd predicted by NeuroBind.

Progerin Cluster information Prodigy
Cluster Haddock score Cluster size Kd [M] Weighted average
LOGO 3 -35.7 ± 4.0 10 1.30E-08 1.30E-08
7 -50.3 ± 5.7 8 3.60E-08
n80_02 3 -42.4 ± 3.7 11 1.60E-07 7.81E-08
5 -41.9 ± 3.6 10 1.90E-09
4 -40.8 ± 1.7 10 2.30E-07
2 -38.4 ± 4.2 11 2.00E-09
1 -38.0 ± 2.7 48 6.80E-08
Rank 15 7 -39.9 ± 9.6 8 1.30E-07 1.30E-07
Rank 21 1 -43.4 ± 1.8 28 5.70E-09 5.70E-09
62aa11_1 1 -32.2 ± 2.0 29 1.40E-06 1.40E-06
52aa2_1 1 -56.0 ± 0.8 32 9.50E-08 4.17E-07
3 -53.7 ± 7.0 53 5.00E-07
2 -51.2 ± 2.1 29 6.20E-07
n51_02 7 -46.2 ± 2.8 8 3.00E-08 6.30E-08
15 -43.8 ± 7.2 5 4.20E-08
Rank 7 3 -37.5 ± 2.8 10 1.00E-07 1.60E-07
42-2L03 2 -29.9 ± 2.3 17 1.60E-07 4.53E-07
9 -39.8 ± 2.4 8 1.00E-06
10 -36.5 ± 3.4 7 4.00E-08
Table 4. HADDOCK results for the interaction between progerin and the binding peptides. The final Kd was calculated as a weighted average of the Kd values from individual clusters, using cluster size as the weighting factor. For models with insufficient significant clusters, the first reliable cluster was retained.

To assess the selectivity of each designed interactor, we compared the dissociation constants (Kd) obtained from docking simulations with progerin and lamin A. To enable direct comparison, we calculated the ratio Kd(progerin)/Kd(lamin A), where values below 1 indicate stronger affinity for progerin, while ratios above 1 suggest preferential binding to lamin A. The results are summarized in Table 5.

This analysis revealed that several candidates displayed marked selectivity toward progerin, with ratios as low as 0.03–0.23, corresponding to up to an order of magnitude stronger binding compared to lamin A. These include the models LOGO, Ranks 15 and 21, 7, and 52aa_1, which represent the most promising interactors for specific recognition of the pathogenic isoform. Rank 21 is the most promising interactor. Conversely, a few candidates (n80-02, 62aa11_1, and 42-2L03) showed ratios exceeding 1, suggesting potential cross-reactivity with lamin A due to the high structural similarity between the two C-terminal regions. In particular, 62aa11_1 exhibited a significantly higher affinity for lamin A. One possible reason is that since the two binding sites of lamin A and progerin differ by only a few amino acids and also correspond to an intrinsically disordered region, the peptide designed to bind progerin efficiently may also have the potential to bind lamin A equally well or even better.

Overall, these results confirm that our computational pipeline can generate interactors with differential binding profiles and demonstrate that both de novo (NeuroBind) and rational design approaches are valid and complementary strategies.

Interactor Haddock Kd [M]
progerin - second analysis
Haddock Kd [M]
lamin A
Kd (progerin) / Kd (lamin A)
LOGO 1.30E-08 1.28E-07 0.10
n80_02 7.81E-08 3.49E-08 2.24
Rank 15 1.30E-07 9.10E-07 0.14
Rank 21 5.70E-09 2.18E-07 0.03
62aa11_1 1.40E-06 1.73E-07 8.10
52aa2_1 4.17E-07 1.80E-06 0.23
42-2L03 4.53E-07 4.42E-07 1.02
n51_02 6.30E-08 1.30E-07 0.48
Rank 7 1.60E-07 1.30E-06 0.12
Table 5. Comparison of the Kd values for the interaction between the designed interactors and progerin versus lamin A. The third column reports the Kd ratio to facilitate the comparison of binding affinities. Values below 1 indicate stronger binding to progerin than to lamin A, while values above 1 correspond to higher affinity for lamin A. These results refer to the interactors without the RING domain.

Step 2 - With the RING domain

To analyse the behaviour of the interactors when attached to the TRIM21-RING domain, it is first necessary to predict and relax the structures of the interactors-RING chimeric protein. The interactors were combined with the RING domain in the N-terminus tail using a flexible linker composed of glycine and serine (GGGGSGGGGSGGGGS). Table 10 presents the energy level of the structure after each relaxation step, run using Amber, RosettaRelax and Amber; along with the MolProbity score for structure validation. Additionally, the pTM score obtained by Alphafold is provided, which estimates the global confidence of the protein folding, serving as an indicator of the reliability of the global topology of the structure. In the table, we highlighted the structures that were retained based on their balance between MolProbity score and energy.

Interactor pTM Energy MolProbity score
Pre-Relax Post-AMBER Post-RosettaRelax Post-AMBER 2.0 Post-AMBER Post-RosettaRelax Post-AMBER 2.0
LOGO 0.41 -1.570,60 -5.551,27 -3.702,61 -6.333,65 2.55 1.46 2.39
n80_02 0.4 -588,60 -4.514,32 -2.275,82 -4.553,58 2.37 1.51 2.12
Rank 15 0.39 365,13 -3.555,26 -850,96 -3.647,22 2.30 2.02 2.38
Rank 21 0.39 302,12 -2.899,76 -549,98 -3.690,36 2.27 1.03 2.42
62aa11_1 0.47 -7.308,03 -11.329,11 -8.128,80 -11.533,71 2.12 1.40 2.17
52aa2_1 0.41 -5.565,58 -8.299,75 -5.352,63 -8.400,86 2.18 1.41 2.02
42-2L03 0.43 -447,58 -2.869,33 -853,65 -3.031,81 2.18 1.55 2.46
n51_02 0.41 -241,29 -4.606,89 -1.848,39 -4.534,06 2.32 1.51 2.18
Rank 7 0.41 1.741,06 -2.435,16 124,32 -2.813,69 2.15 1.45 2.01
Table 6. Relaxation results for the predicted interactors attached to the RING domain. The table shows the energy levels and MolProbity scores of the structures after each relaxation step.

This phase has brought to light important insights about our interactors. We discovered that, when attached to the RING, not all structures are capable of maintaining the correct folding necessary for protein binding. This implies that, for some of the interactors and especially for the globular ones, more damaged by the presence of RING, the linker length needs to be optimized to achieve a functional and properly folded protein. Increasing the linker length can improve the folding of the interactor, but it should not compromise the activity of the RING domain by placing it too far from the target progerin. (We also discuss it in our Engineering page, cycle 5). The most affected structure was LOGO, but we continued the downstream analysis to investigate whether it could retain some binding activity. The following images show comparisons between the properly folded structure and the one whose folding was disrupted by the presence of the RING domain.

Folding Disruption of LOGO Interactor: Effects of the RING Domain on Globular Conformations
Figure 3. folding disruption of interactor LOGO. Globular conformations are the most affected by the presence of the RING domain.
Folding Disruption of Interactor n80_02: RING Domain Impact with Overall Structural Maintenance
Figure 4. folding disruption of interactor n80_02. Globular conformations are the most affected by the presence of the RING domain. Nevertheless, a global maintenance of the interactor’s folding is observed.

We then calculated HADDOCK binding Kd values for lamin A and progerin to investigate whether our interactors could discriminate between the progerin target and lamin A. The most promising interactor appears to be Rank 15, with a Kd of 1.20 × 10⁻⁸ M (Table 7). For other interactors, such as n80_02, the presence of the RING domain enhanced their ability to bind progerin, decreasing the Kd from 7.81 × 10⁻⁸ M to 5.55 × 10⁻⁸ M (Table 7), likely due to conformational rearrangements (Figure 4). Nevertheless, a Kd of 5.55 × 10⁻⁸M can still be considered acceptable.

progerin Cluster information Prodigy
Cluster Haddock score Cluster size Kd [M] Weighted average
LOGO 6 -40.4 ± 1.3 14 6,20E-07 1,38E-07
1 -38.3 ± 2.8 47 3,90E-09
8 -38.0 ± 3.2 6 3,80E-07
5 -36.1 ± 3.6 15 1,40E-08
n80_02 1 -48.5 ± 1.8 98 7,10E-09 5,55E-08
2 -48.3 ± 2.3 19 3,30E-07
3 -40.8 ± 8.8 17 4,40E-08
4 -35.5 ± 1.0 6 1,00E-08
Rank 15 1 -41.0 ± 3.1 49 1,20E-08 1,20E-08
Rank 21 2 -31.7 ± 2.5 26 7,80E-08 7,80E-08
62aa11_1 1 -23.4 ± 8.4 39 1,70E-07 1,70E-07
52aa2_1 1 -45.9 ± 1.3 49 3,80E-07 3,80E-07
42-2L03 2 -52.9 ± 1.7 24 1,60E-07 1,19E-07
4 -45.3 ± 3.3 9 8,50E-09
n51_02 1 -48.3 ± 3.6 74 2,00E-08 1,89E-07
10 -41.7 ± 4.9 5 7,00E-07
2 -41.7 ± 2.3 34 4,80E-07
3 -41.0 ± 1.4 10 1,90E-07
Rank 7 1 -31.1 ± 2.2 116 2,20E-08 2,20E-08
Table 7. HADDOCK docking results for the analysis between the interactors attached to the RING domain and progerin.
Lamin A Cluster information Prodigy
Cluster Haddock score Cluster size Kd [M] Weighted average
LOGO 1 -31.7 ± 1.8 19 1,30E-07 1,30E-07
n80_02 2 -32.5 ± 2.8 50 4,90E-08 4,90E-08
Rank 15 2 -28.6 ± 13.4 19 4,50E-07 4,50E-07
Rank 21 11 -38.9 ± 9.0 4 1,80E-07 1,80E-07
62aa11_1 1 -43.1 ± 1.6 71 3,00E-09 3,00E-09
52aa2_1 1 -36.3 ± 2.6 73 7,30E-06 7,30E-06
42-2L03 5 -52.7 ± 4.1 10 1,70E-07 1,19E-07
1 -50.2 ± 2.1 42 1,60E-07
6 -48.0 ± 0.7 9 3,50E-08
11 -42.7 ± 2.9 4 7,70E-08
4 -41.6 ± 3.8 13 2,40E-07
n51_02 1 -48.0 ± 4.2 74 2,00E-08 1,97E-07
2 -41.7 ± 2.3 36 4,80E-07
9 -41.7 ± 4.9 6 7,00E-07
3 -41.1 ± 1.3 10 1,90E-07
Rank 7 3 -44.6 ± 5.6 25 2,20E-07 2,20E-07
Table 8. shows HADDOCK docking results for the interaction between the interactors with the RING domain and lamin A.

Table 9 compares the Kd values for the interaction between the RING-fused interactors and progerin or lamin A. The analysis revealed that, overall, the presence of the RING domain did not compromise the binding ability of most interactors. Several candidates maintained or even improved their selectivity toward progerin, as indicated by ratios below 1.

In particular, Rank 15 and 52aa2_1 showed the strongest preference for progerin, with Kd ratios of 0.03 and 0.05, respectively, confirming that RING fusion can preserve and even enhance target discrimination. Rank 21 also retained moderate selectivity (ratio = 0.43). In contrast, the LOGO model exhibited a reduction in affinity, with a ratio approaching 1, likely due to conformational disruption observed in Figure 10. It would be interesting to test whether modifying the linker length or the position of the RING domain could restore its optimal configuration.

Notably, 42-2L03 maintained the same Kd for both progerin and lamin A (ratio = 1.00), confirming its neutral binding behavior. n80_02 showed a slight improvement compared to the isolated form, but still displayed a ratio above 1, indicating a stronger affinity for lamin A. 62aa11_1 exhibited a pronounced shift in specificity (ratio = 56.67), suggesting that RING fusion induced unfavorable conformational rearrangements that impaired binding to progerin and favored lamin A instead. Both 62aa11_1 and n80_02 demonstrated low affinity for progerin and higher affinity for lamin A, and would likely be excluded from future optimization rounds.

Taken together, these data indicate that the inclusion of the RING domain generally preserves the designed binders’ selectivity for progerin, while also highlighting how structural compatibility between the effector domain and the interactor scaffold is crucial for maintaining target specificity.

Interactors HADDOCK Kd [M] Progerin HADDOCK Kd [M] Lamin] Kd(progerin) / Kd(lamin a)
LOGO 1,38E-07 1,30E-07 1.07
n80_02 5,55E-08 4,90E-08 1.13
Rank 15 1,20E-08 4,50E-07 0.03
Rank 21 7,80E-08 1,80E-07 0.43
62aa11_1 1,70E-07 3,00E-09 56.67
52aa2_1 3,80E-07 7,30E-06 0.05
42-2L03 1,19E-07 1,19E-07 1.00
n51_02 1,89E-07 1,97E-07 0.96
Rank 7 2,20E-08 2,20E-07 0.10
Table 9. Comparison of the Kd values for the interaction between the interactors and progerin versus lamin A. The third column reports the Kd ratio to facilitate the comparison of binding affinities. Values below 1 indicate stronger binding to progerin than to lamin A, while values above 1 correspond to higher affinity for lamin A. These results refer to the interactors without the RING domain.

Conclusion

In conclusion, as part of our project, we have developed a sophisticated bioinformatics pipeline for the design of peptide interactors specifically targeting an intrinsically disordered region—specifically, the C-terminal tail of progerin. This is a crucial aspect of our approach to tackling Hutchinson-Gilford Progeria Syndrome (HGPS). Our pipeline has undergone multiple iterations of trial and error, incorporating pages of engineering adjustments to refine the results, and we have validated the structures obtained using a variety of bioinformatics tools.

After generating these interactors, we proceeded to analyze their binding affinities. From the hundreds of initial interactors, we selected the top 9 candidates that demonstrated the most promising binding characteristics. These were then subjected to more in-depth analysis, both in laboratory experiments and further computational studies using software tools. As part of our investigation, we included both versions of the interactors: with and without the RING domain from the TRIM21 ubiquitin ligase, which is central to our project’s goal of targeted progerin degradation.

Our findings revealed that, from the many initial candidates, the 9 interactors we analyzed showed varying degrees of specificity. Some interactors performed better against progerin, while others were more effective against lamin A, which underscores the complexity of targeting intrinsically disordered regions with precision. All the interactors developed during this process are now available in the Registry, and their respective sequences are provided in the supplemental materials section for further reference. This resource will be crucial for future studies and experimental validation as we move forward with optimizing these interactors for therapeutic use in the ProgERASE approach.

In this section we’ll provide our experimental design, highlighting the rationale behind each methodological choice concerning our work on S. cerevisiae.

After this introductory framework, the experimental results are presented in four parts:

  • Progeria phenotyping on Saccharomyces cerevisiae;
  • Yeast two-hybrid (Y2H): the limitation we found and the change we decided to apply;
  • Translating of the known interactor BUBR1 into the Y2H;
  • Usage of the system to translate the modelled interaction of our bioinformatics predicted peptides

This structure allows the reader to follow a clear progression: from the strategy underlying our work, to the stepwise presentation, the analysis of experimental outcomes and the methodology choices we decided to apply along the way.

Yeast, including our chassis Saccharomyces cerevisiae, has already been used in the past to study the effects of aberrant protein accumulation, such as in Alzheimer’s disease research. Despite this, our literature research revealed no peer-reviewed papers concerning the expression and modeling of progeria in yeast. That is why we decided to implement this model.

To do so, we expressed progerin sequence BBa_25NDL8N0 using pYES2 (BBa_K555009) in S. cerevisiae CEN.PK. In this way, we were able to create a chassis capable of expressing progerin in a galactose-dependent manner by cloning our progerin sequence under the control of the GAL1 promoter.
All the yeast strains used in our experiments (CEN.PK and Y190) are auxotrophic for certain amino acids and nucleotides: histidine, leucine, tryptophan, uracil, and adenine.
This allows us to select correctly transformed yeast without the need for antibiotics; however, it also requires the preparation of a specific medium with the appropriate compounds to support the growth of the desired yeast.

pYES2 Plasmid (BBa_K555009): Galactose-Inducible progerin Expression in <i>S. cerevisiae</i>
Figure 2.pYES2 plasmid (BBa_K555009). This plasmid was cloned into S. cerevisiae CEN.PK in order to express progerin in a galactose-dependent manner under the GAL1 promoter. It carries the pathway for uracil synthesis (URA3), which complements the uracil auxotrophy of the laboratory yeast strain. When cloned in bacteria it replicates via the ColE1 origin and the transformed colonies can be selected by making use of the ampicillin resistance. Made with Benchling.

We employed this chassis in order to study the progerin phenotype through various assays: spot test and growth curves.

The spot test allows us to compare the growth of the same cultures on different mediums. Up to 8 cultures at the same time can be tested, and each culture is plated in four dilutions (1, 10-1, 10-2 , 10-3) in order to better understand the difference in growth between the different cultures. Each dilution is made from a culture standardized to OD 1; for this reason, the dilution 1 is equivalent to 1 OD concentration of cells.
Plating was performed by depositing a drop of culture onto the array created on the plate. After this, plates are incubated at 28°C for three days; with photos taken every 24 hours to allow results comparison.

The spot test offers the best results in terms of progerin phenotype; in the picture, it is possible to see the growth difference between the same cultures on SD Glucose -URA plate and SD Galactose -URA plate. These results confirm that the expression of progerin under galactose-inducing conditions exerts a toxic effect on yeast cells. Meanwhile, the presence of progerin had no impact under glucose repression, where the GAL1 promoter is inactive. Induction of progerin expression in a galactose-rich medium led to clear growth inhibition.

Spot Test on <i>S. cerevisiae</i> CEN.PK: Galactose-Inducible Progerin Expression Restricts Growth on SD Glucose -URA Medium
Figure 3. Spot test results on S. cerevisiae CEN.PK. progerin expression (galactose-inducible) is responsible for limiting yeast growth on SD Glucose -URA medium.

Find out more about the spot test assay here.
As shown in the picture, progerin expression (induced by the presence of galactose in the sample) is responsible for the growth limitation observed in S. cerevisiae CEN.PK.

Growth curve:

Despite performing the growth curve for pYES2+progerin multiple times, (BBa_25NDL8N0+BBa_K555009) we were unable to obtain consistent results comparable to those from the spot tests.

After many tries, the curve in the picture is the one that represents most closely what we had expected. All experimental measurements were normalized to the first data point of each series to correct for initial differences between samples and allow direct comparison of relative changes over time. This approach ensures that observed trends reflect the actual biological response rather than variability in the starting values.

Effect of progerin Expression on <i>S. cerevisiae</i> CEN.PK Growth: GAL1-Induced Cytotoxicity in Galactose Media
Figure 4.Effect of progerin expression on S. cerevisiae CEN.PK growth. pYES_∅ GLU and pYES2_progerin GLU cultures were growing in SD Glucose 2% -URA, while pYES_∅ GAL and pYES2_progerin GAL cultures were growing in SD Galactose 2% -URA. Galactose is the inductor of the GAL1 promoter in the pYES2 plasmid; for this reason, pYES2_progerin GAL is the only culture where progerin is expressed, and, as it is possible to see, the only one with a strong limitation in growth. The limited growth of pYES_∅ GAL is probably due to the usage of a less preferred sugar.

S. cerevisiae provides a convenient system to study protein interaction using the Yeast Two-Hybrid (Y2H) system. This approach allows us to test the interaction between two proteins via the use of two plasmids: one coding for a “prey” protein and another coding for a “bait” protein.

Yeast Two-Hybrid (Y2H) Mechanism: Protein–Protein Interaction Detection via Reporter Activation
Figure 3. Yeast two hybrid (Y2H) molecular mechanisms. The two plasmids coding for the prey and bait domain are cloned in frame with the two proteins to be assessed for interaction and co-transformed into the same S. cerevisiae cell. If the two proteins of interest interact, the reporter pathway is activated. Made with Biorender.com

The plasmids we used are commercially available, pGADT7 and pGBKT7:

  • pGAD serves as the prey plasmid, expressing the activation domain (AD)
  • pGBK serves as the bait plasmid, expressing the DNA-binding domain (BD)

Proteins of interest can be cloned into these plasmids in frame with the AD or BD domains, enabling the system to test their interaction.

When both plasmids are co-transformed into the same S. cerevisiae Y190 cells, interaction between the proteins of interest brings the BD and AD domains closer together; this way, the protein complex activates the HIS3 reporter gene, allowing yeast growth on a histidine-deficient medium. Because the HIS3 pathway is leaky, 3AT (a competitive inhibitor of the histidine pathway) is added to strengthen the selection when testing interactions.

First attempt: cloning of the whole progerin sequence

We initially attempted to clone the entire progerin sequence (an initial version non complied with RFC10 standard), into both our yeast two-hybrid plasmids, pGADT7 and pGBKT7, in order to determine whether we would be able to observe the expected self-interaction of progerin molecules via Y2H.

Unfortunately, we found that cloning the whole progerin sequence into either pGAD or pGBK plasmids was not feasible, as it rendered S. cerevisiae unable to grow, most likely due to the cytotoxic effect related to progerin.

This conclusion was supported by:

  • Abnormal colony morphology on selective plates (SD Glu -Trp or -Leu, when trying single transformation or -Trp -Leu, when trying co-transformation);
  • Absence of detectable protein expression, as confirmed by Western blot (WB).

S. cerevisiae Y190 Plates Transformed with pGBK_Progerin: Growth Impairment and Repression of Cytotoxic Progerin Expression
Figure 6. Plates of S. cerevisiae Y190 transformed with pGBK_progerin after 72h. Progerin expressed by the constitutive promoter ADH1 impairs colony growth. Western blot analysis confirmed the absence of progerin expression. The colonies we see likely repressed protein expression as a response to progerin cytotoxicity.

This is likely due to a key difference between pGAD/pGBK and pYES2 plasmid: in pGAD and pGBK progerin synthesis is constitutively expressed by ADH1 promoter, whereas in pYES2 it is controlled by inducible promoter GAL1.

Second attempt: cloning of progerin C-term

Since we were unable to obtain any growth by cloning the whole progerin sequence, we chose to repeat the cloning using only the C-terminal of the protein. Specifically, we decided to design two C-term proteins:

  • From amino acid 430 to 614 (BBa_25HM35ST); this sequence includes the Ig-like part of the protein;
  • From amino acids 545 to 614 (BBa_25IF4C4Z); this sequence does not include the Ig-like part of the protein

AlphaFold3-Predicted 3D Structure of Progerin: Visualization of the Ig-Like Domain and Farnesylated C-Terminal Tail
Figure 8. Progerin predicted 3D structure. AlphaFold3-predicted model of progerin visualized through Chimera X. The Ig-like domain is followed by an intrinsically disordered C-terminal tail which contains the CaaX motif recognized by farnesyltransferase (FTase). From Chimera X.

We based our choice on a previous study [1] where they were able to prove progerin interaction with another protein: BUBR1, particularly its N-term.

Given the cytotoxic effect of progerin on cells described above, we decided to proceed with cloning progerin only into the pGBK plasmid. Since pGBK encodes the binding domain (BD), limiting progerin cloning to this plasmid should help prevent its cytotoxic effect on the cell.

Constitutive Expression of Progerin C-Terminal Sections with Gal4 DNA-Binding Domain: Selection via Tryptophan Auxotrophy
Figure 9. Progerin C-term (430 and 545) cloned in pGBK. The selected progerin C-term sections are constitutively expressed along with the Gal4 DNA-binding domain. Since this plasmid encodes the tryptophan synthesis pathway, correctly transformed cells can be selected by not adding tryptophan in the medium. Made with Benchling.

This time, we were able to obtain growth on our plates and we confirmed protein expression through Western Blot (only for aa 430 to 416 fragment).

S. cerevisiae Y190 Plates Transformed with pGBK_Progerin C-Term (430): Stable Growth and Confirmed Protein Expression
Figure 10. Plates of S. cerevisiae Y190 transformed with pGBK_progerin C-term (430). Growth is consistent and we were able to confirm protein expression through WB. This means that C-terminus of progerin does not appear to be cytotoxic for the cell when expressed in pGBK plasmid.

We used the colonies obtained from transformation to continue our studies. The cells were co-transformed with the other plasmid and tested for interaction on Y2H selective medium (SD Glu -Leu -Trp -His +3AT 30 mM).

Let's start with something we already know: BUBR1 N-term

Before testing our bioinformatically predicted interactors, we wanted to validate the Y2H assay using a known progerin interactor, but we were not able to find any progerin known interactor tested through Y2H.
For this reason, we decided to test whether a known interactor would work in our system. To do so, we chose the N-terminal region of BUBR1.

The interaction between progerin (progerin C-term, as stated above) and BUBR1 N-term had been previously reported in mammalian cells via co-immunoprecipitation and pull-down assay [1], but we had no information on whether this interaction could be reproduced in yeast.

We tested the interaction between progerin C-term (430 and 545) and BUBR1 N-term by cloning:

  • Progerin C-term (430 and 545) into pGBK, for the reasons stated above;
  • BUBR1 N-term in pGAD.

Our experimental results indicated that the previously reported interaction between progerin C-term and BUBR1 N-term (seen in previous studies) does not occur in the Y2H system. In fact, we were not able to obtain a positive result in our selective plates for Y2H (SD Glucose -Leu -Trp -His +3AT 30 mM).

Spot Test in Y2H Selective Medium: No Interaction Detected Between Progerin C-Term (aa 430–614) and BUBR1 N-Term
Figure 11. Spot test results in Y2H selective medium. We can conclude that no specific interaction occurs between progerin C-term (aa 430-614) and BUBR1 N-term, as no specific growth was observed where co-transformed culture spots were made.
Position Plate 1 Plate 2
1 CTR (+) CTR (+)
2 CTR (-) CTR (-)
3 pGBK_∅ + pGAD_BUBR1 #1 pGBK_∅ + pGAD_BUBR1 #1
4 pGBK_∅ + pGAD_BUBR1 #2 pGBK_progerin 430 + pGAD_∅ #1
5 pGBK_progerin 430 + pGAD_∅ #1 pGBK_progerin 430 + pGAD_BUBR1 #1
6 pGBK_progerin 430 + pGAD_∅ #2 pGBK_progerin 430 + pGAD_BUBR1 #3
7 pGBK_progerin 430 + pGAD_BUBR1 #1 pGBK_progerin 430 + pGAD_BUBR1 #4
8 pGBK_progerin 430 + pGAD_BUBR1 #2 pGBK_progerin 430 + pGAD_BUBR1 #5

We concluded that these negative results are likely due to the following factors:

  1. We used a different chassis and a different method to test protein interactions;
  2. We were unable to test the full-length protein, as it is toxic for the cell

Now it’s our turn: testing of the bioinformatically predicted interactors

Despite the negative results obtained when testing progerin C-term and BUBR1 N-term, we proceeded to test our predicted interactors.
We tested five colonies (#1,#2,#3,#4,#5) for each of the following interactors:

  • BBa_25R6U8N1: LOGO - bioinformatically predicted progerin interactor 1;
  • BBa_25RLZ3X4: n80_02 - bioinformatically predicted progerin interactor 2;
  • BBa_25G9QB6S: Rank_15 - bioinformatically predicted progerin interactor 3

The interactors were tested both against progerin 430 and progerin 545 by co-transformation of the plasmids pGBK_progerin430/545 + pGAD+interactor in S. cerevisiae Y190. To test whether the interaction occurred, we carried out a spot test assay on Y2H selective medium SD GLU -Leu -Trp -His +3AT 30 mM.
When interaction occurs, yeast cells are able to grow in this medium; if the interaction does not occur, no growth should be observed.

Unfortunately, the lack of growth observed in the spot test suggests that interactions with progerin C-term (both 430 and 545 fragments) do not take place in the Y2H system for the first three interactors.

Spot Test Assay: Assessment of Progerin_430 Interaction with Interactors 1–3 in Yeast Two-Hybrid System
Figure 12. (A) Spot test assay to evaluate the interaction between progerin_430 and Interactor 1 (colonies #1–#5), (B) Interactor 2 (colonies #1–#5) and (C) Interactor 3 (colonies #1–#5) SD_GLU represents our positive control, because this medium selects transformed yeasts. On the other hand, SD_GLU +3AT 30mM medium should select those yeast cells where interaction between progerin and interactors occurs. We can conclude that we are unable to determine progerin interaction with our interactors, at least in the yeast two-hybrid system.
No. PLATE A PLATE B PLATE C
1 pGBK_progerin 430 + pGAD_∅ pGBK_progerin 430 + pGAD_∅ pGBK_progerin 430 + pGAD_∅
2 pGBK_progerin 430 + pool pGAD_interactor 1 #1 pGBK_progerin 430 + pool pGAD_interactor 2 #1 pGBK_∅ + pool pGAD_interactor 3
3 pGBK_progerin 430 + pool pGAD_interactor 1 #2&3 pGBK_progerin 430 + pool pGAD_interactor 2 #2&3 pGBK_progerin 430 + pool pGAD_interactor 3 #1
4 pGBK_progerin 430 + pool pGAD_interactor 1 #4&5 pGBK_progerin 430 + pool pGAD_interactor 2 #4&5 pGBK_progerin 430 + pool pGAD_interactor 3 #2&3
5 pGBK_∅ + pool pGAD_interactor 1 pGBK_∅ + pool pGAD_interactor 2 pGBK_progerin 430 + pool pGAD_interactor 3 #4&5
Spot Test Assay: Assessment of Progerin_545 Interaction with Interactors 1–3 in Yeast Two-Hybrid System
Figure 13. (D) Spot test assay to evaluate the interaction between progerin_545 and Interactor 1 (colonies #1–#5), (E) Interactor 2 (colonies #1–#5) and (F) Interactor 3 (colonies #1–#5) SD_GLU represents our positive control, because this medium selects transformed yeasts. On the other hand, SD_GLU +3AT 30mM medium should select those yeast cells where interaction between progerin and interactors occurs. We can conclude that we are unable to determine progerin interaction with our interactors, at least in the yeast two-hybrid system.
No. PLATE D PLATE E PLATE F
1 pGBK_progerin 545 + pGAD_∅ pGBK_progerin 545 + pGAD_∅ pGBK_progerin 545 + pGAD_∅
2 pGBK_progerin 545 + pool pGAD_interactor 1 #1 pGBK_∅ + pool pGAD_interactor 2 pGBK_∅ + pool pGAD_interactor 3
3 pGBK_progerin 545 + pool pGAD_interactor 1 #2&3 pGBK_progerin 545 + pool pGAD_interactor 2 #1 pGBK_progerin 545 + pool pGAD_interactor 3 #1
4 pGBK_progerin 545 + pool pGAD_interactor 1 #4&5 pGBK_progerin 545 + pool pGAD_interactor 2 #2&3 pGBK_progerin 545 + pool pGAD_interactor 3 #2&3
5 pGBK_∅ + pool pGAD_interactor 1 pGBK_progerin 545 + pool pGAD_interactor 2 #4&5 pGBK_progerin 545 + pool pGAD_interactor 3 #4&5

Despite the negative results the remaining two predicted interactors (Interactor 4 and Interactor 5) have yet to be tested, and interactions with the wild type protein must also be verified to fully assess the reliability of our predictions.

To address this, five colonies (#1,#2,#3,#4,#5) of the following two additional interactors were analyzed:

  • BBa_25UZI9YM (Rank_21) – bioinformatically predicted interactor 4;
  • BBa_256U2NU6 (62aa11_1) – bioinformatically predicted interactor 5

Following successful single transformations in S. cerevisiae Y190, co-transformations were carried out. Both interactors will be assayed against progerin 545.
Unfortunately we still did not obtain positive results, confirming that our five predicted interactors cannot be used to determinate progerin interaction in yeast.

Spot Test Assay: Assessment of Progerin_545 Interaction with Interactors 4 and 5 in Yeast Two-Hybrid System
Figure 14. (G) Spot test assay to evaluate the interaction between progerin_545 and Interactor 4 (colonies #1–#5) and (H) Interactor 5 (colonies #1–#5) SD_GLU represents our positive control, because this medium selects transformed yeasts. On the other hand, SD_GLU +3AT 30mM medium should select those yeast cells where interaction between progerin and interactors occurs.
No. PLATE G PLATE H
1 pGBK_∅ + pGAD_interactor 4 pGBK_∅ + pGAD_interactor 5
2 pGBK_progerin 545 + pGAD_∅ pGBK_progerin 545 + pGAD_∅
3 pGBK_progerin 545 + pGAD_interactor 4 #1&2 pGBK_progerin 545 + pGAD_interactor 4 #1&2
4 pGBK_progerin 545 + pGAD_interactor 4 #3&4&5 pGBK_progerin 545 + pGAD_interactor 4 #3&4&5

To further investigate the specificity of these interactors, interactors 1, 4 and 5 were tested against lamin A, the wild type protein. For this purpose, lamin A was co-transformed into Y190 with different pGAD plasmids:

  • pGBK_LaminA + pGAD_∅
  • pGBK_LaminA + pGAD_interactor1;
  • pGBK_LaminA + pGAD_interactor4
  • pGBK_LaminA + pGAD_interactor5
The results demonstrated that lamin A is not affected by any of our tested interactors, ruling out off-target effects.

Spot Test Assay: Validation of Interactors 1, 4, and 5 for Absence of Off-Target Binding to Lamin A
Figure 15. Spot test assay to validate that interactor 1 (I), interactor 4 (L) and interactor 5 (M) do not show off-target binding to lamin A SD_GLU represents our positive control, because this medium selects transformed yeasts. On the other hand, SD_GLU +3AT 30mM medium should select those yeast cells where interaction between lamin A and interactors may occur.
No. PLATE I PLATE L PLATE M
1 pGBK_∅ + pGAD_interactor 1 pGBK_∅ + pGAD_interactor 4 pGBK_∅ + pGAD_interactor 5
2 pGBK_laminA + pGAD_∅ pGBK_laminA + pGAD_∅ pGBK_laminA + pGAD_∅
3 pGBK_laminA + pGAD_interactor 1 #1&2 pGBK_laminA + pGAD_interactor 4 #1&2 pGBK_laminA + pGAD_interactor 5 #1&2
4 pGBK_laminA + pGAD_interactor 1 #3&4&5 pGBK_laminA + pGAD_interactor 4 #3&4&5 pGBK_laminA + pGAD_interactor 5 #3&4&5

Conclusions

Although we expected to observe an interaction between the five interactors and the tested portions of progerin, no interaction was detected, indicating that (at least in the yeast two-hybrid system) these interactions do not occur.

On the other hand, for lamin A the results were consistent with our expectations, confirming that the interactors do not display off-target effects toward lamin A.

Future perspectives

Unfortunately, due to time constraints, we were unable to explore the following topics, despite the potential relevance of the results they could have produced:

  • Post-translational modifications (PTMs): We would like to further investigate how progerin is processed in yeast and whether it undergoes the same post-translational modifications as in human cells, with farnesylation being the most critical. At present, we cannot determine whether progerin produced in yeast is fully comparable to the human form. However, it is likely that many of the processing steps occur, given the similarities in protein synthesis pathways between yeast and humans. For example, S. cerevisiae expresses Ste24, homologous to the human protease ZMPSTE24[2], which mediates the final cleavage of prelamin A. This step is absent in progerin due to the mis-splicing event that generates it. Mass spectrometry analysis of purified protein would help clarify the extent of these modifications in yeast-expressed progerin;
  • Subcellular localization: Another key open question is whether progerin expressed in yeast correctly localizes to the nucleus. So far, we lack direct evidence of its localization. This could be addressed using immunofluorescence or immunohistochemistry with specific anti-progerin antibodies, which we were able to obtain thanks to our sponsor, Diatheva.

In terms of scientific impact, although our results represent just the beginning for modeling progeria (HGPS) in yeast, we believe that our work can serve as a valuable starting point for further studies. In fact, we anticipate that the phenotype observed in S. cerevisiae CEN.PK could be further standardized and subsequently used as a tool to study phenotype rescue in the context of chemical screening.
This is the reason why even though this tool represents foundational research, it has the potential to contribute to the development of new drugs aimed at mitigating the effects of progeria in patients.

  • [1] Zhang, N., Hu, Q., Sui, T., Fu, L., Zhang, X., Wang, Y., Zhu, X., Huang, B., Lu, J., Li, Z., & Zhang, Y. (2023). Unique progerin C-terminal peptide ameliorates Hutchinson-Gilford progeria syndrome phenotype by rescuing BUBR1. Nature aging, 3(2), 185–201. https://doi.org/10.1038/s43587-023-00361-w
  • [2] Spear, E. D., Alford, R. F., Babatz, T. D., Wood, K. M., Mossberg, O. W., Odinammadu, K., Shilagardi, K., Gray, J. J., & Michaelis, S. (2019). A humanized yeast system to analyze cleavage of prelamin A by ZMPSTE24. Methods (San Diego, Calif.), 157, 47–55. https://doi.org/10.1016/j.ymeth.2019.01.001

Structure of the Mammalian Results Section

In this experimental phase, we successfully expressed and confirmed the production of progerin in mammalian cells, establishing a working model for testing degradation and rescue strategies. Expression was verified through Western blot analysis of HA-progerin, confirming correct translation of the construct, and through fluorescence microscopy, which allowed direct visualization of the mEGFP-tagged proteins. In particular, fluorescence imaging confirmed the expression and localization of progerin-mEGFP, SpyTag-mEGFP-NLS-SpyTag, and RING-NLS-mEGFP, with all constructs showing the expected nuclear or perinuclear distribution patterns, albeit in a limited number of transfected cells due to the low transfection efficiency of MRC-5 fibroblasts.

Functional assays showed that progerin expression significantly reduced cell viability at four days post-transfection, as measured by the alamarBlue TM

We also tested the RING-SpyCatcher system for its ability to degrade both mEGFP and progerin, using flow cytometry and Western blot analysis, respectively. However, the low transfection efficiency in MRC-5 cells prevented reliable detection of double-transfected populations and clear evaluation of degradation effects. Despite these limitations, the experiments provided valuable insights for future optimization, including adjustments to transfection strategies, timing of assays, and choice of cell lines, which will guide the next phase of testing and validation of the ProgERASE degradation system.

Overcoming Low DNA Quantity:

The DNA fragments from A to M, ordered from TWIST, were delivered with blunt ends and in a limited amount of 1000 ng, which was insufficient for direct restriction. To overcome this, the fragments were first ligated into the pJET1.2/blunt plasmid to allow for amplification. The resulting plasmid was then digested with NEB restriction enzymes BamHI HF and XhoI and analyzed on a 1% agarose gel. Two distinct bands were observed: one corresponding to the plasmid backbone and the other to the insert, at the expected size. The insert band was excised and purified from the gel. Finally, the purified insert was ligated into the pcDNA3.1 plasmid, pre-digested with the same restriction enzymes, to achieve directional cloning compatible with the RFC 10 standard.

pcDNA3.1 plasmid map
Figure 1.The pcDNA3.1 vector (5,428 bp) contains a strong CMV promoter for expression in mammalian cells and a high-copy pUC origin for propagation in E. coli. Cloning is carried out via restriction enzymes. Selection markers include Ampicillin resistance for bacteria and Neomycin resistance for eukaryotic cells.
pJET1.2 plasmid map
Figure 2.pJET1.2 plasmid (2947 bp, Thermo Fisher Scientific). The vector features a positive selection system that enables efficient identification of recombinant colonies. Insertion of a DNA fragment disrupts a lethal restriction enzyme gene, allowing only cells carrying recombinant plasmids to survive, while self-ligated vectors lead to host cell death. The plasmid is supplied in a linearized blunt-end form, optimized for blunt-end cloning reactions.

Further analyis of our methods

pJET1.2 cloning

All constructs listed in the preview table were received in lyophilized form, rehydrated, and preliminarily ligated into pJET1.2. The ligation products were used to transform competent Escherichia coli DH5α, plated on LB agar with Ampicillin, and incubated overnight. The first constructs processed were A, B, and C. To assess pJET1.2 transformation efficiency, six colonies per plate were screened; once confirmed, three colonies were selected for each of the remaining constructs (D, E, F, I, K, M). Individual inocula were prepared, labeled, and grown overnight at 37°C with shaking (150 rpm). The following day, plasmid DNA was extracted with the PureLink™ HQ Mini Plasmid Purification Kit (Invitrogen) and quantified via NanoDrop.

Extracted plasmids were digested with BamHI HF and XhoI, and fragments analyzed by electrophoresis. Positive samples showed two bands (backbone and insert), with those matching the expected insert size considered correct.

pcDNA3.1 cloning

Restriction digestion was performed on all plasmid DNA from positive samples. Since NanoDrop measurements indicated low concentrations, the entire 42 µL of DNA was used instead of the 5 µg specified by the protocol. Digestion products were run on a 1% agarose gel with wide wells, clearly separating the plasmid backbone (higher molecular weight) from the insert (lower molecular weight). Insert bands were excised and purified using the QIAquick Gel Extraction Kit (QIAGEN).

The first three samples were quantified by NanoDrop to determine the volume required for ligation at a 3:1 insert-to-plasmid ratio. Purified inserts were ligated into the previously digested pcDNA3.1 plasmid, and the ligation reaction was incubated overnight.

Two control ligation reactions were prepared:

  • Plasmid with restriction enzymes: This control was included to confirm complete digestion of the plasmid DNA by both restriction enzymes. In cases of incomplete digestion, the plasmid may remain intact or be cleaved at a single site, thereby generating compatible ends that could be re-circularized by the ligase. Transformation with uncut or re-circularized plasmid DNA results in colony formation;
  • Plasmid only: This control was performed to exclude the occurrence of colonies arising from spontaneous plasmid re-circularization or external contamination, thus ensuring that colony growth in experimental reactions reflected specific ligation events

The following day, chemically competent E. coli DH5α cells were transformed with the ligation products and plated on LB agar with ampicillin. Two vials of competent cells were also transformed with the control ligation reactions. After overnight incubation, all plates showed growth except for the control plate transformed with linearized plasmid only. In contrast, ~180 colonies appeared on the plate from control “With T4 Ligase”, due to T4 ligase re-ligating a 69 bp fragment left between the restriction sites, allowing colony growth.

This unintended re-ligation may also have occurred in experimental ligations. To verify, six colonies from each plate were selected, grown overnight at 37 °C with shaking (150 rpm), and plasmid DNA was extracted using the PureLink TM HQ Mini Plasmid Purification Kit (Invitrogen). Restriction digestion with BamHI-HF and XhoI followed by agarose gel electrophoresis confirmed positive clones as those showing two bands: the plasmid backbone and the expected insert.

Sequencing

Sequencing was performed for positive samples. Extracted DNA was amplified using the Applied BiosystemsTM BigDyeTM Terminator v3.1 Cycle Sequencing Kit (Fisher Scientific). A total of 20 PCR reactions were set up, two per sample, using the CMV forward and BGH reverse primers.

PCR products were purified using OPTIMA DTRTM 96-Well Plates (EdgeBio) prior to automated Sanger sequencing. Obtained sequences were aligned with the expected sequences using Blast. Sequence identity ranged from 92% to 100% for all constructs.

All plasmids that showed the correct sequence were amplified through transformation of E. coli DH5α cells followed by plasmid purification.

Insert Name Forward Sequence Identity Reverse Sequence Identity
HA-progerin 98.35% 99.12%
SpyTag-progerin 100.00% 99.66%
SpyTag-laminA 99.12% 98.90%
mEGFP-progerin 98.18% 100.00%
mEGFP-NLS 92.47% 93.82%
mEGFP-NLS-SpyCatcher 97.60% 99.01%
RING-NLS-HA 100.00% 99.74%
RING-NLS-mEGFP 98.38% 98.57%
RING-NLS-SpyCatcher 99.35% 100.00%
Table 1. Identity % for both forward and reverse sequences obtained through pcDNA3.1-insert sequencing by PCR.

A deeper analysis of the alignment readings can be found in the following file: pCDNA3.1_sequencing_results.pdf

The Spy System

To assess the efficacy of our RING-bait strategy in human fibroblasts, we initially considered using HGPS-affected cells transfected with our RING-progerin interaction-fusion protein. However, to gain more reliable insights into the degradation dynamics of progerin, we opted for the Spy System, which facilitates protein-protein interactions between molecules fused with SpyTag and SpyCatcher. This two-part system allowed the expression of SpyTagged progerin in MRC-5 fibroblasts, enabling the formation of SpyTag-SpyCatcher specific progerin-RING complexes.

Structure of Spt System, both attached and detached
Figure 3. Spy System structure prediction. SpyTag (yellow) and SpyCatcher (blu) are depicted both detached (left) and attached (right). When they interact, they form a covalent isopeptide bond that stabilizes the complex.

The SpyTag/SpyCatcher system originates from the Streptococcus pyogenes FbaB protein, a fibronectin-binding protein and virulence factor. The CnaB2 domain of FbaB forms a stable isopeptide bond between lysine K31 and aspartic acid D117, catalyzed by glutamate E77. By splitting the domain into SpyTag (13 amino acids, containing Asp) and SpyCatcher (138 amino acids, containing Lys and Glu), a covalent bond is reconstituted when the two fragments are combined. This system is highly specific, demonstrating high affinity (≈0.2 µM) and low intrinsic reactivity unless the residues are aligned appropriately. The SpyTag can be incorporated at any position on a protein, including the N- or C-terminus or internally, provided the position is surface-exposed. We selected this system for its precision and one-to-one binding stoichiometry, which mirrors the architecture of our designed chimeric protein. [1]

A comparison between HGPS and healthy fibroblasts cell lines

Cell Lines Used

After defining the experimental strategy, the next step was to identify suitable cell models for testing the constructs. To this end, we established a collaboration with Prof.ssa Giovanna Lattanzi of the National Research Council (CNR) in Bologna, a leading expert in laminopathies. Through her laboratory, we obtained two human fibroblast lines, cc00366s and cc00500s.

The cc00366s line derives from a patient affected by Hutchinson–Gilford Progeria Syndrome (HGPS), while cc00500s corresponds to a healthy control. From their arrival in the laboratory, both lines were amplified to obtain a sufficient number of cells for subsequent experimental applications.

To assess morphological differences between healthy and progeria-affected cells, two T75 flasks were observed under bright-field microscopy: one containing fibroblasts derived from a healthy donor, and the other from a Hutchinson–Gilford Progeria Syndrome (HGPS) patient.

On the left side Control fibroblasts exhibit the typical elongated, spindle-shaped morphology and form dense, well-organized monolayers. On the right side, HGPS fibroblasts display irregular and elongated shapes, with thinner cytoplasmic extensions and a disrupted growth pattern
Figure 4.Bright-field microscopy comparison between healthy control fibroblasts (left) and HGPS patient-derived fibroblasts (right); image acquired with Zeiss Primo Vert inverted microscope with 20x magnification. Control fibroblasts exhibit the typical elongated, spindle-shaped morphology and form dense, well-organized monolayers. In contrast, HGPS fibroblasts display irregular and elongated shapes, with thinner cytoplasmic extensions and a disrupted growth pattern. The cells appear less aligned and often form sparse, unevenly distributed clusters.

These morphological alterations reflect the nuclear deformation and cytoskeletal instability characteristic of progeria-affected cells. Even under standard culture conditions, the phenotypic differences between the two fibroblast lines are evident, visually confirming the impact of progerin accumulation on cellular structure and organization.

Although both lines were successfully expanded, they exhibited slow proliferation rates, with extended doubling times that made them less suitable for repeated transfection assays or large-scale functional studies. To perform the first experimental phase, we therefore adopted an alternative fibroblast line already available in the laboratory: MRC-5 human lung fibroblasts.

The MRC-5 line consists of diploid, non-transformed fibroblasts derived from human fetal lung tissue. These adherent cells display the typical spindle-shaped morphology and form organized monolayers under standard culture conditions. Their doubling time ranges from 40 to 48 hours in serum-supplemented medium, and they can undergo approximately 42–48 population doublings before entering senescence.

Thanks to their moderate transfection efficiency, MRC-5 fibroblasts allow only a fraction of cells to express the introduced constructs. This condition prevents excessive accumulation of reactive metabolites and stress factors that would otherwise occur in highly transfected cultures, potentially affecting neighboring cells and leading to widespread toxicity. Maintaining a partially transfected population therefore enables a more physiological response and makes it easier to observe potential recovery effects or compensatory mechanisms within the culture.

MRC-5 fibroblasts are particularly suited for short-term assays, such as the evaluation of morphological alterations, oxidative stress, and viability changes associated with progerin expression, while their limited lifespan and sensitivity to nutrient depletion and culture stress make them less suitable for long-term experiments.

The choice to use fibroblast lines rather than immortalized cell models was guided by both biological and translational considerations. Since the HGPS-derived cells obtained from Prof.ssa Lattanzi are fibroblasts, maintaining the same cell type ensures experimental continuity and phenotypic comparability. Moreover, fibroblasts are among the most affected tissues in Progeria, where nuclear envelope defects and cytoskeletal disorganization are most evident, making them a biologically relevant and disease-representative model.

Bright-field images of MRC-5 fibroblasts at 90% confluence (left panel) and 40% confluence (right panel)
Figure 5.Bright-field images of MRC-5 fibroblasts acquired using LEICA: DMi1 inverted microscope. The left (10x magnification) panel shows cells at approximately 90% confluence, forming a dense and uniform monolayer with elongated spindle-like morphology. The right panel (20x magnification) shows the same cell line at around 50% confluence, where individual fibroblasts are more dispersed and clearly distinguishable, maintaining their characteristic elongated shape.

Overall, MRC-5 fibroblasts provided a reliable and physiologically relevant system for studying the cellular effects of progerin accumulation, validating construct expression, and laying the groundwork for future optimization in more easily transfectable cell models such as HEK293T.

Unexpected Culture Loss and Impact on Experimental Timeline

During the second half of August, a technical failure in the CO2 delivery system of the cell culture incubators led to a complete loss of all our fibroblast cultures. As a result, we were left without viable cells from mid-August until the second week of September.
During this period, the team focused exclusively on cloning procedures and plasmid amplification, which had already been initiated in the preceding weeks. Although this allowed us to advance the molecular part of the project, the incident significantly delayed the experimental phase.
Because of the time constraints imposed by the Wiki Freeze deadline, we were only able to perform each cellular assay once, without the opportunity to repeat or optimize the experiments. This limitation prevented us from obtaining more robust and statistically consolidated data, even though the results provided valuable qualitative insights for future repetitions.

Details on the experimental rationale and control design are provided in the following section (Experiments).

Fluorescence Microscopy

To visualize the expression and localization of the fluorescently tagged constructs, MRC-5 fibroblasts were analyzed by fluorescence microscopy on Day 3 (single transfections) and Day 4 (double transfections).
Fluorescence was acquired at an excitation wavelength of 488 nm and emission at 525 nm in the FITC channel. Images were taken from cultures transfected as part of other experiments, and not from newly transfected samples. All the constructs are cloned into the same pcDNA3.1 backbone.

The following constructs were evaluated:

  • SpyTag-mEGFP-NLS-SpyTag: visualization of mEGFP expression;
  • RING-NLS-mEGFP: visualization of RING expression through mEGFP fusion;
  • progerin-mEGFP: visualization of progerin expression through mEGFP fusion;
  • SpyTag-progerin + mEGFP-SpyCatcher: visualization of progerin localization through forced interaction (SpyTag/Catcher system) with mEGFP

Fluorescence microscopy allowed the visualization of mEGFP expression, acting as a first visual control.

SpyTag-mEGFP-NLS-SpyTag visualization

visualization of SpyTag-mEGFP-NLS-SpyTag in MRC-5 fibroblasts. Phase contrast (top-left), merged fluorescence (low-center), and GFP channel (top-right.
Figure 6.Fluorescence induced visualization of SpyTag-mEGFP-NLS-SpyTag in MRC-5 fibroblasts. Phase contrast (top-left), merged fluorescence (low-center), and GFP channel (top-right). A few fluorescent cells can be observed, confirming the expression of the construct. Low cell confluence can be observed, providing insight on low transfection rates. Image acquired utilizing Olympus APX100 with 20x magnification.

RING-NLS-mEGFP visualization

Fluorescence-based visualization of RING-NLS-mEGFP expression in MRC-5 fibroblasts. Phase contrast (top-left), merged fluorescence (low-center), and GFP channel top-right.
Figure 7. Fluorescence induced visualization of RING-NLS-mEGFP in MRC-5 fibroblasts. Phase contrast (top-left), merged fluorescence (low-center), and GFP channel (top-right). Cells expressing RING-NLS-mEGFP display a diffuse fluorescence, and for this reason subcellular localization was not possible. Only a limited number of fluorescent cells were detected, consistent with the low efficiency of transfection in this cell line. Image acquired utilizing Olympus APX100 with 20x magnification.

mEGFP-progerin visualization

Fluorescence-based visualization of RING-NLS-mEGFP expression in MRC-5 fibroblasts. Phase contrast (left) and merged fluorescence (right).
Figure 8.Fluorescence induced visualization of mEGFP-progerin in MRC-5 fibroblasts; phase contrast (left) and merged fluorescence (right. Cells expressing mEGFP-progerin show a faint but detectable GFP signal, confirming construct expression. As for the previous construct, only a limited number of fluorescent cells were observed, reflecting the low transfection efficiency of MRC-5 fibroblasts and the moderate expression levels achieved under these conditions. Image acquired utilizing Olympus APX100 at 20x magnification.

Progerin-Spytag + mEGFP-SpyCatcher visualization

Fluorescence induced visualization of progerin-SpyTag + mEGFP-SpyCatcher in MRC-5 fibroblasts; phase contrast (left) and merged fluorescence (right.
Figure 9. Fluorescence induced visualization of progerin-SpyTag + mEGFP-SpyCatcher in MRC-5 fibroblasts; phase contrast (left) and merged fluorescence (right). The image shows no detectable fluorescence providing no information on mEGFP–SpyCatcher expression. suggesting difficulties during the second transfection. Image acquired utilizing Olympus APX100 at 20x magnification.

Fluorescence microscopy confirmed that all tested constructs were correctly expressed in MRC-5 fibroblasts, though with very low transfection efficiency. This limitation was particularly evident in double transfection experiments, where the probability of cells simultaneously expressing both constructs was extremely low. As a result, fluorescence patterns and subsequent readouts, such as those for degradation or interaction assays, are likely affected by the low transfection efficiency, reducing the statistical robustness of the observations. Future experiments will focus on optimizing transfection conditions or employing alternative cell lines with higher transfectability to ensure more reliable expression and localization analyses.

Western Blot Analysis – Progerin Expression

To confirm the successful expression of progerin, we performed a Western blot analysis on lysates obtained from MRC-5 fibroblasts transfected with the HA-progerin construct. Cells were collected on Day 3, corresponding to 48 hours after transfection. All the constructs are cloned into the same pcDNA3.1 backbone.

The following conditions were analyzed:

  • Control (untransfected cells): used to assess background signal and confirm antibody specificity;
  • HA-progerin: to verify the expression of the HA-tagged progerin construct using an anti-HA antibody

Figure 10. The left panel shows the Western blot probed with an anti-HA antibody. The Thermo Scientific™ Spectra™ Multicolor Protein Ladder is visible on the left. A distinct HA–progerin–specific band is detected at approximately 74 kDa, while the untransfected control lane (to the right of the HA–progerin lane) shows no detectable signal. The right panel displays the Western blot for the housekeeping protein β-actin, developed using the Thermo Scientific™ PageRuler™ Plus Prestained Protein Ladder.

Western blot results confirmed the expression of HA-progerin, with a distinct band detected at approximately 74 kDa, corresponding to the expected molecular weight of the protein. No signal was observed in the untransfected control, confirming the specificity of the anti-HA antibody and the absence of non-specific background bands.

These results demonstrate that the HA-progerin construct was correctly expressed in MRC-5 fibroblasts and can serve as a reliable reference for subsequent analyses of progerin variants and degradation assays.

alamarBlueTM Cell Viability assay

To evaluate the effect of progerin expression and its targeted degradation through the RING-SpyCatcher system, cell viability was measured in MRC-5 fibroblasts four days and six days after transfection using the alamarBlueTM assay. The assay provided a quantitative readout of metabolic activity, reflecting the overall viability of cells transfected with different construct combinations. Each condition was analyzed in eight replicates to ensure statistical robustness, and fluorescence values were normalized as Relative Fluorescence Units (RFU).

Unlike the experimental samples, the control conditions were performed with a different number of replicates.The negative control was measured in duplicate, while the laminA-SpyTag condition was tested in six independent replicates. This difference in replication number should be taken into account when comparing standard errors between conditions, as lower replication increases variability and reduces statistical confidence in the mean values obtained. This difference derives from the limited amount of cells we had in culture. All the constructs are cloned into the same pcDNA3.1 backbone.

The following transfection conditions were tested:

  • Control: Untransfected cells, serving as a baseline for cell viability;
  • Progerin-SpyTag: To assess the impact of progerin expression tagged with SpyTag on cell viability;
  • SpyTag-mEGFP-SpyTag + RING-SpyCatcher: To evaluate whether the degradative activity of RING towards mEGFP affects cell viability;
  • Progerin-SpyTag + RING-SpyCatcher: To investigate whether Progerin degradation via the RING-SpyCatcher system restores cell viability to normal levels or improves it;
  • LaminA-SpyTag: A lamin A control construct to compare the effects of progerin-induced viability defects with a non-pathological lamin protein

Chart plotting the mean Relative Fluorescence Intensity against the different transfected wells.
Chart 1. The bar chart plots the mean Relative Fluorescence Intensity against different transfected wells. Mean RFI and Standard Error are noted on the bottom of the chart.

Technical note on the four-day assay.

During the four-day alamarBlueTM assay, the first row of the 96-well plate was affected by a handling issue that compromised fluorescence readings. Consequently, all experimental conditions except laminA-SpyTag were analyzed using seven wells instead of eight, while the negative control was measured in a single well. The laminA-SpyTag condition, which was plated starting from the third row, was unaffected by this problem and was consistently measured in six replicates.

The results at four days revealed a marked decrease in viability in cells expressing Progerin-SpyTag, confirming the cytotoxic effect of progerin accumulation. Mean fluorescence intensity was approximately 250 RFU for progerin-expressing cells, significantly lower than that observed in both the laminA-SpyTag (≈ 550 RFU) and Control (≈ 570 RFU) groups, which displayed comparable viability levels.

Co-expression of progerin-SpyTag and RING-SpyCatcher did not fully restore viability, with fluorescence values remaining close to those of progerin alone, indicating that progerin degradation via the RING-SpyCatcher system was only partially effective or not sufficient to counteract the toxic effects of progerin accumulation at this time point. However, fluorescence microscopy and FACS analysis revealed a low transfection efficiency in MRC-5 fibroblasts, suggesting that the second transfection (Day 2) may not have been fully effective. This limitation could explain the absence of a clear recovery in cell viability, as a significant proportion of cells might not have received both constructs required for efficient progerin degradation.

In contrast, the mEGFP + RING condition, used as a functional control for RING activity, showed no evident reduction in cell viability compared to the control group; however, this observation should be interpreted cautiously, as the low efficiency of the second transfection might also have affected this condition, potentially masking subtle effects of RING expression on cell metabolism.

Overall, these findings demonstrate that progerin expression significantly reduces cell viability at four days, whereas lamin A and mEGFP + RING maintain metabolic activity comparable to untransfected controls. The partial or absent recovery observed in the progerin + RING condition may be linked to the limited transfection efficiency rather than to a complete lack of effect of the RING-SpyCatcher system. These data are consistent with fluorescence microscopy observations showing limited co-expression of the two constructs, and future experiments with optimized transfection conditions will be required to confirm the functional activity of the degradative system.

Cell viability was evaluated again at six days post-transfection using the alamarBlueTM assay to assess the proliferation progression dependent from progerin expression and the RING-SpyCatcher system.

Chart plotting the mean RFI against the different transfected wells.
Chart 2. The bar chart plots the mean Relative Fluorescence Intensity against the different transfected wells. Mean RFI and Standard Error are noted on the bottom of the chart.

At this time point, all transfected conditions showed a strong decrease in metabolic activity compared to the negative control, indicating a general reduction in cell viability. Mean fluorescence values ranged between 52–84 RFU for all transfected samples, while the negative control displayed a markedly higher mean value (≈ 320 RFU).

Notably, cells were not provided with a medium change after transfection, which likely contributed to the overall decline in viability observed across all conditions. This lack of medium renewal, combined with prolonged culture stress, may have led to nutrient depletion and waste accumulation, resulting in a widespread loss of metabolic activity independent of transfection type.

These results are consistent with the Total ROS Production Assay, where increased oxidative stress was detected in parallel cultures at the same time point, confirming that cellular stress and decreased metabolic function contributed to the loss of viability at six days.

Overall, the six-day data indicate that while progerin expression exerts cytotoxic effects, the generalized cell death observed here is largely attributable to culture conditions rather than to specific construct effects, emphasizing the need for medium renewal in extended assays to accurately evaluate viability differences over time.

Total ROS Production Assay

To assess potential oxidative stress associated with progerin expression and its possible modulation by the RING-SpyCatcher system, we performed a Total ROS Production Assay in MRC-5 fibroblasts. The assay was conducted six days after seeding, at the same timepoint as the alamarBlueTM viability assay. Fluorescence intensity was measured as Relative Fluorescence Units (RFU), providing a quantitative indicator of intracellular ROS accumulation. All the constructs are cloned into the same pcDNA3.1 backbone.

The following transfection conditions were tested:

  • HA-progerin: To evaluate ROS production due to the presence of progerinl;
  • SpyTag-progerin: To determine the effect of progerin tagged with SpyTag on ROS levels;
  • mEGFP-progerin: A fusion construct of progerin with mEGFP, used to investigate whether the addition of a fluorescent tag alters ROS production;
  • RING-mEGFP: To assess the effect of the RING domain (fused to mEGFP) on ROS production, testing whether the RING domain influences oxidative stress;
  • RING-SpyCatcher: To analyze whether the combination of the RING domain with SpyCatcher modifies ROS production, potentially due to the structural impact of SpyCatcher;
  • mEGFP-NLS: Used as a control to evaluate whether the mEGFP tag alone contributes to ROS production;
  • mEGFP-SpyCatcher: A control to test the effect of the SpyCatcher tag in the context of mEGFP-tagged constructs;
  • laminA-SpyTag: A lamin A control construct to compare progerin-induced ROS with a non-pathological lamin protein;
  • Negative control: Untransfected cells used to establish baseline ROS production, representing normal ROS levels in cells not subjected to transfection or overexpression

Each condition was analyzed in multiple replicates, and mean RFU values with their respective standard deviations are reported below the chart.

chart plots the mean RFI against the different transfected wells.
Chart 3. Total Ros assay readout. The bar chart plots the mean Relative Fluorescence Intensity against the different transfected wells. Mean RFI and Standard Error are noted on the bottom of the chart.

As shown in Chart 3, all transfected conditions exhibited relatively similar fluorescence levels, with progerin-expressing constructs (HA-progerin, progerin-SpyTag, progerin-mEGFP) showing mean RFU values around 0.5, comparable to those of RING and SpyCatcher constructs. The negative control displayed a slightly higher fluorescence value (≈ 0.84 RFU).

At this stage of the experiment, cell viability was markedly reduced across all conditions, indicating that most cells were undergoing late-stage stress or death. Under these conditions, any potential differences in ROS production between constructs were likely masked by the generalized loss of viable cells. The residual fluorescence detected primarily reflects the number of surviving cells, rather than genuine differences in oxidative stress levels. The higher RFU observed in the control wells is therefore consistent with a greater number of viable cells, which naturally produce more ROS as a consequence of normal metabolic activity.

Overall, these results indicate that by six days post-seeding, cell death had leveled out differences in ROS production, resulting in comparable oxidative readouts across all transfected conditions. Future repetitions of this assay should be performed at four days post-transfection, when the alamarBlueTM viability assay revealed the greatest difference in cell survival between conditions. Performing the ROS test at that timepoint will allow a more accurate assessment of the relationship between progerin expression, oxidative stress, and cell viability.

RING-SpyCatcher Functionality

To evaluate whether RING-SpyCatcher could induce mEGFP degradation in the presence of SpyTag, we performed a flow cytometry (FACS) assay on MRC-5 fibroblasts transfected with different construct combinations. The analysis was conducted on Day 4, following two sequential transfections. Fluorescence was detected using an excitation wavelength of 488 nm and an emission wavelength of 525 nm, collected in the FITC channel. All the constructs are cloned into the same pcDNA3.1 backbone.

The following conditions were tested:

  • Negative Control: untransfected cells, used to determine baseline fluorescence and gating parameters;
  • SpyTag-mEGFP-SpyTag: single transfection to measure mEGFP expression levels in the absence of RING-SpyCatcher;
  • RING-SpyCatcher + SpyTag-GFP-SpyTag: co-transfection to test whether the RING-SpyCatcher system induces mEGFP degradation

Each condition was analyzed in duplicate, and data were acquired from gated populations excluding debris and artifacts.

FACS analysis of untransfected cells. Left: FSC-A vs SSC-A. Center: Fluorescence (B525-A) vs SSC-A plot of P1. Right: Histogram of fluorescence intensity
Figure 11. Negative control (untransfected MRC-5 fibroblasts). FACS analysis of untransfected cells. Left: FSC-A vs SSC-A plot showing the gated cell population (P1). Center: Fluorescence (B525-A) vs SSC-A plot of P1, indicating absence of mEGFP-positive cells (0.19%). Right: Histogram of fluorescence intensity showing only background autofluorescence
FACS analysis of MRC-5 fibroblasts expressing mEGFP tagged with SpyTag. Left: Gated population (P1) of single-transfected cells. Center: Fluorescent population (P2, 1.79%. Right: Histogram showing low fluorescence intensity.
Figure 12. SpyTag-mEGFP-SpyTag single transfection. FACS analysis of MRC-5 fibroblasts expressing mEGFP tagged with SpyTag. Left: Gated population (P1) of single-transfected cells. Center: Fluorescent population (P2, 1.79%) indicating limited mEGFP expression. Right: Histogram showing low fluorescence intensity (~10⁵ RFU), consistent with low transfection efficiency.
FACS analysis of MRC-5 fibroblasts co-transfected with RING-SpyCatcher and SpyTag-GFP-SpyTag. Left: Gated population (P1) of co-transfected cells. Center: Fluorescent population (P2, 2.10%. Right: Histogram showing similar fluorescence intensity
Figure 13. RING-SpyCatcher + SpyTag-mEGFP-SpyTag co-transfection. FACS analysis of MRC-5 fibroblasts co-transfected with RING-SpyCatcher and SpyTag-mEGFP-SpyTag. Left: Gated population (P1) of co-transfected cells. Center: Fluorescent population (P2, 2.10%) comparable to single mEGFP transfection, suggesting no evident degradation. Right: Histogram showing similar fluorescence intensity (~105 RFU) to mEGFP-only cells, indicating no measurable reduction in mEGFP signal.

Data visualization and gating strategy

Data visualization and gating strategy

Each set of plots is composed of three panels:

  • Left panel (FSC-A vs SSC-A): represents forward scattering (FSC) versus side scattering (SSC), used to evaluate cell size and internal complexity. A gating region (P1) was defined to select the main cell population and exclude debris, aggregates, or air bubbles;
  • Middle panel (B525-A vs SSC-A): displays fluorescence intensity versus side scatter, filtered according to the P1 gate. The fluorescent population (P2) is highlighted in the upper region of the plot, representing cells expressing mEGFP. The corresponding percentage of fluorescent cells within the total population is indicated;
  • Right panel (Fluorescence histogram): shows fluorescence intensity distribution (x-axis) against cell count (y-axis) for the gated population. The histogram allows direct comparison of fluorescence magnitude between samples

Results

The negative control (untransfected cells) exhibited only background autofluorescence, confirming the validity of the gating strategy. In the SpyTag-mEGFP-SpyTag condition, a small fluorescent population was detected (~1.8% of total), indicating a low transfection efficiency in MRC-5 fibroblasts. In the RING-SpyCatcher + SpyTag-mEGFP-SpyTag co-transfection, the fluorescent population (~2.1%) and overall fluorescence intensity were comparable to the single mEGFP transfection, with no measurable reduction in mEGFP-positive cells or fluorescence magnitude.

These results suggest that no significant mEGFP degradation was detectable under the present conditions. The very low transfection efficiency likely limited the number of double-transfected cells, preventing clear observation of RING-mediated degradation. Moreover, the fluorescence intensity of mEGFP-positive populations was in the range of 10⁵ RFI, whereas in efficiently transfected cell lines, signals typically reach 106–107RFI, producing a distinct separation between fluorescent and non-fluorescent populations—an effect not observed here.

Overall, these data indicate that the RING-SpyCatcher system could not be effectively evaluated in this experiment due to low transfection rates in MRC-5 cells. The experiment will be repeated using cell lines (HEK293T) with higher transfection efficiency to confirm the degradative function of RING under optimized conditions.

Western Blot Analysis – progerin Degradation

To evaluate whether the RING-SpyCatcher system could promote progerin degradation, we performed a Western blot on MRC-5 fibroblast lysates collected from cells transfected with various constructs. All samples were detected using the anti-progerin antibody (Diatheva Srl, ANT0046; Rabbit Anti-Human Cleaved-Farnesylated Prelamin A).
Cells were transfected on Day 1 and Day 2 (for double transfections) and harvested on Day 4 for protein extraction.

The following conditions were analyzed:

  • Control (untransfected cells): used to verify the absence of background signal;
  • Progerin-SpyTag: single transfection to confirm the expression of SpyTag-tagged Progerin;
  • Progerin-mEGFP: to evaluate the expression of the mEGFP-fused progerin variant;
  • RING-Catcher: used to confirm antibody specificity and ensure no cross-reactivity with RING-containing proteins;
  • mEGFP-NLS: used as a negative control for the anti-progerin antibody;
  • Progerin-SpyTag + mEGFP: baseline control for progerin expression in a double-transfection setup;
  • Progerin-SpyTag + RING-SpyCatcher: experimental condition to test potential Progerin degradation induced by RING-SpyCatcher co-expression.

Figure 14. Western blot readout (low) after incubation with primary antibody (Rabbit Anti-Human Cleaved-Farnesylated Prelamin A antibody; ANT0046; Diatheva Srl). No distinct bands could be detected under the concentrations and incubation time used. Longer exposure did not make any specific bands visible. Marker visualization is represented on top.

The resulting Western blot did not show any progerin-specific bands in any of the analyzed conditions. The Thermo ScientificTM SpectraTM Multicolor Protein Ladder bands were visible on the right side of the membrane, confirming proper electrophoretic migration and transfer. However, no distinct signal at the expected ~70 kDa band corresponding to progerin was detected in any sample.

This result rendered the assay inconclusive, as it was not possible to evaluate whether RING-SpyCatcher induced progerin degradation. The most plausible explanation is the suboptimal performance of the primary anti-progerin antibody, which had not been validated for Western blot detection.

The antibody was provided by Diatheva Srl within an active collaboration and had been previously validated only for immunofluorescence. As agreed with the company, the present results, whether positive or negative, will be shared to support further antibody validation.

In this assay, the antibody was used at a 1:2000 dilution with a 1-hour incubation, conditions that may have been insufficient for effective detection. Future optimization will include longer incubation times and alternative antibody dilutions, with the goal of establishing a reliable protocol for progerin detection by Western blot and enabling robust evaluation of RING-SpyCatcher–mediated degradation in subsequent experiments.

NanoLuc Binary Technology (NanoBiT Assay)

Starting from the in-silico designed library of interactors, the NanoBiT complementation assay.
Figure 15. NanoBiT pipeline: image created with BioRender. Starting from the in-silico designed library of interactors, the NanoBiT complementation assay will provide quantitative information on the actual interaction the peptides have onto progerin in a physiological environment.

To extend the validation of our system beyond degradation-focused assays and to directly assess protein–protein interactions between progerin and its designed interactors, we implemented the NanoLuc Binary Technology (NanoBiT) system [2].
This method is based on the reversible reconstitution of NanoLuc luciferase through the association of its two fragments, Small BiT (SmBiT) and Large BiT (LgBiT), each fused to a protein of interest. When the two fusion partners interact, the SmBiT and LgBiT fragments are brought into close proximity, allowing the luciferase enzyme to refold into its active conformation. This conformational change restores the catalytic activity of NanoLuc, producing a strong luminescent signal that can be quantitatively measured in living cells.

Although interaction between progerin and its binders was previously verified through yeast two-hybrid screening and will be further evaluated by Microscale Thermophoresis (Monolith X), the NanoBiT assay provides a crucial complementary validation. Unlike those systems, which operate in in vitro or yeast contexts, NanoBiT enables the detection of protein–protein interactions within mammalian cells, reproducing the physiological environment in which the designed degradation system must ultimately function. This ensures that binding events observed in earlier assays are not artifacts of heterologous systems but are maintained in the cellular context relevant for therapeutic applications.

Figure 16. Schematic representation of the NanoBiT system showing Small BiT (yellow) and Large BiT (light green) in both a detached, inactive conformation (left) and a complexed, active conformation. When the fusion partners interact, the proximity of Small BiT and Large BiT induces a conformational rearrangement that restores NanoLuc catalytic activity, resulting in luminescence.

Although experimental validation could not yet be completed due to the limited number of MRC-5 fibroblasts available, the entire preparatory phase of the assay was successfully accomplished. This included DNA sequence optimization for NanoBiT cloning, vector assembly, and bacterial transformation. The constructs are currently awaiting sequence verification by Sanger sequencing using NanoBiT-specific primers, which will confirm successful cloning before proceeding with in-cell luminescence assays.

Through this experimental phase, we successfully established the complete workflow required to evaluate our degradation system in mammalian cells, from DNA amplification and cloning into the mammalian expression vector pcDNA3.1, to construct validation, transfection, and subsequent functional assays. The results confirmed the correct expression of the designed constructs, including HA-progerin, progerin-SpyTag, RING-SpyCatcher, progerin-mEGFP, and SpyTag-mEGFP-SpyTag.

Despite the low transfection efficiency observed in MRC-5 fibroblasts, fluorescence microscopy and Western blot analyses confirmed that the constructs were properly expressed. Cell viability tests provided preliminary but consistent indications that progerin accumulation negatively impacts cell metabolism and proliferation, while co-expression with the RING-SpyCatcher system may partially modulate these effects.

These results mark an important step toward the experimental validation of the ProgERASE strategy, in order to confirm the feasibility of using the RING-SpyCatcher platform as a programmable degradation system. Although several optimizations are still required, particularly regarding transfection efficiency, antibody validation, and assay sensitivity, the foundation for further development in mammalian models has been successfully established.

Experiment Improvements

Cell Line Optimization: HEK293T as a High-Transfection Model

To improve the consistency and interpretability of future experiments, we plan to complement the current fibroblast-based assays with HEK293T cells, which exhibit one of the highest transfection efficiencies among mammalian cell lines (often exceeding 70–90% with standard lipid-based reagents such as Lipofectamine 2000 or PEI). We chose to perform our experiments using MRC-5 fibroblasts, as their lower transfection efficiency results in progerin expression at sub-saturating levels. This condition helps preserve phenotypic variability and prevents the extensive cell loss observed under high-expression scenarios.

However, to confirm the functionality of the RING-SpyCatcher system, we plan to continue our studies using HEK293T cells. These cells are immortalized, robust, and fast-growing, providing a more homogeneous model for testing genetic constructs and fluorescent reporters. Their high metabolic activity and resilience to transfection-induced stress make them particularly suitable for optimizing plasmid ratios and analyzing the kinetics of the degradation system.
Using HEK293T cells will allow us to:

  • Optimize time windows for assays (24–72 h post-transfection) to capture degradation dynamics before secondary cytotoxic effects dominate;
  • Achieve clearer and more statistically reliable fluorescence and FACS readouts due to a higher number of successfully transfected cells;
  • Perform various Western blot assays on progerin-transfected cell’s lysates, optimizing Diatheva ANT0046 antibody concentrations and exposure times in cells with high transfection efficiency

Refinement of the mEGFP Degradation Assay

To enhance the evaluation of mEGFP degradation mediated by the RING–SpyCatcher system, future experiments will employ a bicistronic plasmid designed to co-express RING–SpyCatcher and RFP from a single transcriptional unit. In the initial setup, MRC-5 fibroblasts were co-transfected with SpyTag–mEGFP–SpyTag and RING–SpyCatcher constructs; however, low transfection efficiency hindered the reliable detection of double-positive cells. The inclusion of an RFP reporter will make it possible to identify and gate RFP and mEGFP-positive cells, those effectively expressing both mEGFP-SpyTag and RING–SpyCatcher, and quantitatively compare mEGFP intensity between RFP-positive and RFP-negative populations. This approach will:

  • Provide an internally normalized measurement of degradation efficiency;
  • Reduce noise caused by uneven transfection or expression variability;
  • Allow FACS-based sorting and downstream validation (e.g. Western blot or microscopy) of the RING–SpyCatcher expressing population

Optimization of the ROS Assay

To improve data reliability in oxidative stress assays, future replicates of the Total ROS assay will be performed either on Day 4 after seeding, or after a medium change at Day 3 to maintain cell viability.
Previous tests showed that by Day 6, a significant proportion of cells had detached or died, leading to inconsistent fluorescence values. Adjusting timing and refreshing the medium will help ensure that fluorescence reflects true intracellular ROS accumulation, rather than artifacts from population loss.
Normalization to cell viability (e.g., via alamarBlue, DNA content, or confluence measurements) will be integrated into future runs to provide a quantitative correction factor linking ROS signal intensity to living cell number.

Optimization of the alamarBlueTM Viability Assay

To refine the assessment of cellular viability and better capture the dynamic effects of progerin expression and RING–SpyCatcher co-transfection, we plan to optimize the alamarBlueTM Cell Viability assay protocol. In the initial experiments, viability was measured at Day 4 and Day 6 post-transfection. While the Day 4 assay revealed clear and significant differences in metabolic activity between the experimental conditions, by Day 6 most cells had already lost viability, preventing further quantitative interpretation.

Future experiments will therefore include daily viability monitoring over an extended time course to generate a detailed growth profile. Specifically, fluorescence will be first recorded at Day 0 (immediately after seeding) to establish a common baseline for all wells, and then measured daily for approximately ten days, with medium replacement performed as needed to maintain optimal culture conditions.

This approach will enable us to:

  • Monitor the progressive effects of progerin expression and RING–SpyCatcher co-expression on cell proliferation and metabolic recovery;
  • Identify the exact time window in which viability differences are most pronounced;
  • Distinguish between acute cytotoxicity and long-term adaptive responses;
  • Ensure that the assay reflects true metabolic variations rather than cumulative cell death due to culture stress

Tracking viability across multiple timepoints will thus provide a more accurate and continuous representation of cell health and growth dynamics in response to progerin expression and its targeted degradation.

Future Perspectives

NanoBiT Assay

After confirming successful cloning through vector sequencing, we will proceed with testing the NanoBiT control vectors (N203, N204) to verify:

  • The functionality of the assay reagents;
  • The compatibility of MRC-5 fibroblasts with the NanoBiT detection system
Following control validation, we will assess the interaction between LgBiT–laminA and SmBiT–progerin, as well as between SmBiT–progerin and LgBiT–progerin. Detection of a luminescent signal in these conditions will confirm that progerin is correctly folded and functionally expressed within the NanoBiT assay environment. Moreover, the observation of progerin–progerin interaction would provide further insight into its ability to self-associate, supporting the possibility of RING dimerization within the degradation system.
Once these preliminary validations are completed, we will perform assays between SmBiT–Progerin and LgBiT–Interactor constructs to evaluate specific binding, while SmBiT–lamin A will be used in parallel as a specificity control. This comparison is essential to ensure that the designed interactors selectively recognize progerin without binding to wild-type lamin A.

If the NanoBiT system performs as expected, we will extend the screening to a broader panel of interactors, enabling a systematic evaluation of binding strength and specificity directly within mammalian cells.

Toward Viral-Mediate Delivery

In the next stages of development, we aim to translate our approach into a disease-relevant context by testing the system in HGPS primary fibroblasts. This step will be undertaken only after completing the characterization of our degradation mechanism in laboratory-engineered fibroblasts, ensuring full validation under controlled conditions.

To enable targeted delivery in patient-derived cells, we plan to use viral delivery systems, such as Adeno-Associated Virus (AAV) or Lentiviral Vectors, to deliver our therapeutic construct.

AAVs represent a well-established platform for gene delivery, offering stable, long-term expression without genomic integration, and the possibility of tailoring tissue tropism and minimizing immunogenicity through serotype and capsid selection. On the other hand, Lentiviral Vectors are effective gene delivery tools that enable stable, long-term expression by integrating into the host genome. They can transduce both dividing and non-dividing cells, making them versatile for therapeutic and research applications. By modifying their envelope proteins, lentiviruses can be tailored for specific tissue targeting and minimized immune responses, making them ideal for gene therapy and functional gene studies.

This preclinical phase will bridge our current in vitro validation to future translational applications, providing a robust framework for assessing safety, efficiency, and specificity in a physiopathological model of HGPS.

AAV-Immune Considerations and Pediatric Window

Although AAV can elicit innate and adaptive immune responses to both capsid and transgene, studies indicate that anti-AAV antibody seroprevalence is significantly lower in children (often <15% in ages 3–12, increasing to >50% in adults). [3]
This suggests a therapeutic window in pediatric populations, including HGPS patients, in which viral transduction could be achieved with minimal immune interference. Nonetheless, standard NAb screening will remain essential prior to any application.
Given that HGPS cells exhibit a pro-inflammatory phenotype—characterized by elevated NF-κB activity and cytokine secretion—careful immunological monitoring (IL-6, TNF-α, interferon-stimulated genes) will be implemented in translational stages.

  • [1] Long Li, Jacob O. Fierer, Tom A. Rapoport, Mark Howarth, Structural Analysis and Optimization of the Covalent Association between SpyCatcher and a Peptide Tag, Journal of Molecular Biology, Volume 426, Issue 2 (2014), pp 309-317.https://doi.org/10.1016/j.jmb.2013.10.021
  • [2] Andrew S. Dixon, Marie K. Schwinn, Mary P. Hall, Kris Zimmerman, Paul Otto, Thomas H. Lubben, Braeden L. Butler, Brock F. Binkowski, Thomas Machleidt, Thomas A. Kirkland, Monika G. Wood, Christopher T. Eggers, Lance P. Encell, Keith V. Wood. NanoLuc Complementation Reporter Optimized for Accurate Measurement of Protein Interactions in Cells. ACS Chemical Biology, Volume 11, Issue 2 (2016), pp. 400–408. https://doi.org/10.1021/acschembio.5b00753
  • [3] Amit Chhabra, George Bashirians, Christos J. Petropoulos, Terri Wrin, Yuvika Paliwal, Peter V. Henstock, Suryanarayan Somanathan, Candida da Fonseca Pereira, Ian Winburn, John E.J. Rasko. Global seroprevalence of neutralizing antibodies against adeno-associated virus serotypes used for human gene therapies. Molecular Therapy - Methods & Clinical Development. Volume 32, Issue 3. (2024) https://doi.org/10.1016/j.omtm.2024.101273.