Periodic Reporting for period 1 - ISOS (ISOS-Implantable Ecosystems of Genetically Modified Bacteria for the Personalized Treatment of Patients with Chronic Diseases)
Période du rapport: 2023-12-01 au 2024-11-30
ISOS aims to develop the first biomedical product for in-situ fabrication and sustained delivery of therapeutics using probiotic genetically engineered bacteria (GEB) within a biomaterial-based bioreactor for chronic disease treatment.
Encapsulated GEBs produce drugs "on demand," responding to pathological signals like inflammation. The ecosystem ensures bacterial survival only within the biomaterial, meeting biosafety standards. Personalized designs use in-silico tools and synthetic biology.
As a Proof-of-Concept, ISOS will create an implantable GEB-based bioreactor for wet age-related macular degeneration (wAMD), replacing repeated Anti-VEGF injections with a single, responsive treatment, setting the foundation for next-gen personalized therapeutics.
Encapsulated GEBs produce drugs "on demand," responding to pathological signals like inflammation. The ecosystem ensures bacterial survival only within the biomaterial, meeting biosafety standards. Personalized designs use in-silico tools and synthetic biology.
As a Proof-of-Concept, ISOS will create an implantable GEB-based bioreactor for wet age-related macular degeneration (wAMD), replacing repeated Anti-VEGF injections with a single, responsive treatment, setting the foundation for next-gen personalized therapeutics.
We have engineered synthetic bacteria with genetically designed circuits (GEB). GEBs show intelligent behavior as a consequence of the analogy found between artificial neural networks (ANN) and Boolean logic (Figure A), which regulates the expression of the genetic circuits. First, we have studied different neural networks, i.e. elementary McCulloch-Pitts networks and perceptron, and performed in silico experiments with the gro simulator. Second, using the Cello platform, we obtained the DNA sequence, the logic circuit, and the genetic circuit that emulates the artificial neural network. Third, we edited the plasmid that implements the ANN. The use of this methodology was applied to the treatment of some chronic diseases.
We have developed silk-based hydrogels with different compositions, and analyzed their physical properties. We demonstrated their ability to confine and interact with genetically engineered bacteria at macro- and microscopic scales. Bulk analysis of silk fibroin hydrogels, with and without bacteria, revealed controlled bacterial growth and motility. These findings highlight the potential of silk hydrogels as versatile biomaterials for bacterial encapsulation and controlled microenvironments.
We designed peptides and miniproteins for inhibition of VEGF using a structure-based computational design workflow. The Mini-Z peptide and Z-1-2 protein were used as starting templates and their sequences were redesigned, aiming at improving their VEGF binding affinity and energetic stability. An innovative design pipeline consisting of the artificial intelligence-based tools ProteinMPNN and AlphaFold was established, yielding hundreds of high quality protein designs according to AlphaFold’s confidence metrics. The most promising peptides and proteins were finally selected for wetlab testing. Mini-Z designs were synthesized by solid-phase peptide synthesis and miniproteins were produced by recombinant expression in bacteria.
SPR measurements were established for the determination of binding affinities towards immobilized VEGF. A panel of six different anti-VEGF peptides were tested, establishing the Mini-Z peptide as the best binder with a KD of 13µM. Differences in binding affinity were observed between Mini-Z and its variants with variant M13 being slightly better with a KD of 6.2 µM. In addition, a cell-based bioassay to measure the inhibition of VEGF to the VEGF receptor was established for the functional characterization of anti-VEGF peptides. Amongst the tested peptides Mini-Z showed the strongest inhibition with an IC50 of 16.5 µM.
Engineered recombinant bacteria for survival in silk fibroin hydrogel and anti-VEGF production. Four anti-VEGF peptides were expressed, purified, and confirmed via SDS-PAGE and proteomics. A SLiDE E. coli strain requiring benzothiazole for growth was developed using lambda red recombination. A metabolic cross-feeding system was designed and implemented, optimizing D-alanine production by overexpressing racemase genes and evaluating transporters. A second E. coli strain, auxotrophic for D-alanine, is being engineered with racemase knockouts. A suitable ocular-adapted strain was identified, and a gene expression toolkit is under evaluation for stable gene expression.
We have developed silk-based hydrogels with different compositions, and analyzed their physical properties. We demonstrated their ability to confine and interact with genetically engineered bacteria at macro- and microscopic scales. Bulk analysis of silk fibroin hydrogels, with and without bacteria, revealed controlled bacterial growth and motility. These findings highlight the potential of silk hydrogels as versatile biomaterials for bacterial encapsulation and controlled microenvironments.
We designed peptides and miniproteins for inhibition of VEGF using a structure-based computational design workflow. The Mini-Z peptide and Z-1-2 protein were used as starting templates and their sequences were redesigned, aiming at improving their VEGF binding affinity and energetic stability. An innovative design pipeline consisting of the artificial intelligence-based tools ProteinMPNN and AlphaFold was established, yielding hundreds of high quality protein designs according to AlphaFold’s confidence metrics. The most promising peptides and proteins were finally selected for wetlab testing. Mini-Z designs were synthesized by solid-phase peptide synthesis and miniproteins were produced by recombinant expression in bacteria.
SPR measurements were established for the determination of binding affinities towards immobilized VEGF. A panel of six different anti-VEGF peptides were tested, establishing the Mini-Z peptide as the best binder with a KD of 13µM. Differences in binding affinity were observed between Mini-Z and its variants with variant M13 being slightly better with a KD of 6.2 µM. In addition, a cell-based bioassay to measure the inhibition of VEGF to the VEGF receptor was established for the functional characterization of anti-VEGF peptides. Amongst the tested peptides Mini-Z showed the strongest inhibition with an IC50 of 16.5 µM.
Engineered recombinant bacteria for survival in silk fibroin hydrogel and anti-VEGF production. Four anti-VEGF peptides were expressed, purified, and confirmed via SDS-PAGE and proteomics. A SLiDE E. coli strain requiring benzothiazole for growth was developed using lambda red recombination. A metabolic cross-feeding system was designed and implemented, optimizing D-alanine production by overexpressing racemase genes and evaluating transporters. A second E. coli strain, auxotrophic for D-alanine, is being engineered with racemase knockouts. A suitable ocular-adapted strain was identified, and a gene expression toolkit is under evaluation for stable gene expression.
Our work is a significant step forward in the state of the art both in the field of synthetic biology and its application in disease treatment, and therefore in the main objective of the present project. The possibility of endowing bacteria with intelligent behaviors by transforming them in the laboratory with a plasmid whose gene expression and regulation is equivalent to the information processing of an artificial neural network represents a milestone (Figure A). In the future, these synthetically intelligent bacteria will make it possible to tackle the elimination of cancerous tumors, pathogenic bacteria and viruses with greater success than has been achieved to date in experiments and approaches based on classical synthetic biology techniques.
This work advances the state of the art by developing an injectable, adaptive, and smart silk fibroin hydrogel matrix with tunable mechanical properties, and functionalization for encapsulating genetically engineered bacteria. By operating on biomaterial’s physical variables, we achieved precise control over the physical characteristics, creating hydrogels with tailored physical characteristics. Functionalization with bioactive molecules improve the performance of the material.
An innovative computational protein design pipeline integrating the artificial intelligence tools ProteinMPNN and AlphaFold was established. Its application to VEGF demonstrated that it can deliver several hundreds of good-scoring designs in a short time, majorly increasing the success chances for developing new protein binders.
A novel peptide with improved affinity (KD = 6.3 μM) compared to mini-Z-1 (KD = 13 μM) was found in only one round of design and testing. The chances of finding even better binders are high as more and more peptides and miniproteins will be tested in the next months.
Significant progress was made in engineering bacterial strains for safe, controlled anti-VEGF peptide production, achieving Deliverable 4.1. The peptides (Mini-Z, Z-1-2, Z-3B, V107) were successfully overexpressed in E. coli BL21(DE3) and validated via SDS-PAGE, Tricine-SDS, and proteomics. A SLiDE mutant strain was generated and awaits validation. Engineering of a D-alanine auxotroph and overproduction strain progressed, and a suitable ocular strain for therapeutic delivery was identified. These advancements support potential in vivo wAMD treatment, but further validation, regulatory assessment, and bioprocess optimization are crucial for clinical translation and sustained therapeutic efficacy.
This work advances the state of the art by developing an injectable, adaptive, and smart silk fibroin hydrogel matrix with tunable mechanical properties, and functionalization for encapsulating genetically engineered bacteria. By operating on biomaterial’s physical variables, we achieved precise control over the physical characteristics, creating hydrogels with tailored physical characteristics. Functionalization with bioactive molecules improve the performance of the material.
An innovative computational protein design pipeline integrating the artificial intelligence tools ProteinMPNN and AlphaFold was established. Its application to VEGF demonstrated that it can deliver several hundreds of good-scoring designs in a short time, majorly increasing the success chances for developing new protein binders.
A novel peptide with improved affinity (KD = 6.3 μM) compared to mini-Z-1 (KD = 13 μM) was found in only one round of design and testing. The chances of finding even better binders are high as more and more peptides and miniproteins will be tested in the next months.
Significant progress was made in engineering bacterial strains for safe, controlled anti-VEGF peptide production, achieving Deliverable 4.1. The peptides (Mini-Z, Z-1-2, Z-3B, V107) were successfully overexpressed in E. coli BL21(DE3) and validated via SDS-PAGE, Tricine-SDS, and proteomics. A SLiDE mutant strain was generated and awaits validation. Engineering of a D-alanine auxotroph and overproduction strain progressed, and a suitable ocular strain for therapeutic delivery was identified. These advancements support potential in vivo wAMD treatment, but further validation, regulatory assessment, and bioprocess optimization are crucial for clinical translation and sustained therapeutic efficacy.