Model
Metabolic Model
Our project aims to optimize the production of a secondary metabolite from a cryptic cluster found in a Streptomyces strain. Initially, our target was chloramphenicol, an antibiotic of great importance in medicine, produced by Streptomyces venezuelae (FERNÁNDEZ-MARTÍNEZ; BORSETTO; GOMEZ-ESCRIBANO; BIBB; AL-BASSAM; CHANDRA; BIBB, 2014). However, due to difficulties in obtaining this microbial strain in time for the iGEM competition, we decided to shift our focus to cosmomycin.
Cosmomycin is an anthracycline with antibiotic properties, but it stands out for its potential as an antitumor agent in carcinosarcomas, in addition to having reduced side effects compared to other anthracyclines. It is produced by Streptomyces olindensis, a strain isolated from Brazilian soil in the 1960s (GARRIDO, L. M. et al. 2006).
The work titled "METABOLIC FLUX ANALYSIS FOR OPTIMIZING THE SYNTHESIS OF COSMOMYCIN ANTIBIOTIC BY STREPTOMYCES OLINDENSIS ICB20" (LIMA LOBATO, A. K. C., 2010) has provided a valuable contribution to the understanding of cosmomycin synthesis over the years. The study was based on a metabolic flux model that considered 245 reactions, encompassing the major pathways of primary metabolism and the cosmomycin production pathway in the strain of Streptomyces olindensis.
Metabolic fluxes allowed for correlating the degree of involvement among different metabolic pathways, focusing on intracellular reactions quantified from more easily observable extracellular fluxes. This model was constructed based on stoichiometric reactions independent of the biochemical network and included the necessary variables to solve all equations.
The author of the work mentioned that information about metabolic pathways and reactions was obtained from literature data, biochemistry books, and databases such as KEGG, related to the Streptomycetes genus. Additionally, the choice of data related to secondary metabolism was grounded in the work of Professor Leandro Garrido from ICB-USP (GARRIDO, L. M. et al. 2006), who dedicated himself to sequencing the genes involved in cosmomycin synthesis.
This study highlights the importance of integrating data from different sources and applying metabolic flux models for a better understanding and optimization of the synthesis of complex compounds like cosmomycin.
Figure 1: Steps to promote metabolic flux analysis. Taken from (LIMA LOBATO, A. K. C., 2010).
To solve this system of equations, an optimization model utilizing the mathematical tool of linear programming was employed. To achieve this, optimization was guided by defining an objective function and imposing constraints on the system, such as incorporating experimental parameters, enforcing positive rates for metabolic products, ensuring fluxes greater than or equal to zero for irreversible reactions, and maintaining a pseudo-steady state for intracellular metabolites.
As a result of this optimization model, the distribution of metabolic fluxes was obtained. These calculated metabolic fluxes were then transferred to the constructed flux map for improved visualization and utilization in metabolic flux analysis (AFM).
Figure 2: Proposed metabolic map to depict the distribution of fluxes in Streptomyces olindensis ICB20. Taken from (LIMA LOBATO, A. K. C., 2010).
One of the most significant results of the study is a graph where we can compare the production rates of cosmomycin obtained by the metabolic flux model with those obtained experimentally.
Figure 3: Specific product formation rates, both computational and experimental, as a function of specific feed rates in continuous cultures of S. olindensis ICB20 (LIMA LOBATO, A. K. C., 2010).
In black, we have the experimental curve, in red, we have the computational curve without considering nutritional constraints, and the green curve is also a computational curve, taking into account nutritional constraints, making it closer to the experimental result.
The results of this study provide us with insights into the most sensitive metabolic pathways for cosmomycin production. In our project, one of our constructions was based on the information obtained in this work. The synthesis of fatty acids competes with the cosmomycin pathway for the precursor Acetyl-CoA. Since fatty acid production is an essential activity for cell survival and cannot be inhibited, we proposed the incorporation of malonate from the culture medium to provide a greater amount of Acetyl-CoA for cosmomycin synthesis without compromising the cell's demand for fatty acids.
To estimate the increase in cosmomycin production using malonate from the medium as a synthesis precursor, we would only need to add a malonate input to the existing model.
References:
FERNÁNDEZ-MARTÍNEZ, Lorena T.; BORSETTO, Chiara; GOMEZ-ESCRIBANO, Juan Pablo; BIBB, Maureen J.; AL-BASSAM, Mahmoud M.; CHANDRA, Govind; BIBB, Mervyn J.. New Insights into Chloramphenicol Biosynthesis in Streptomyces venezuelae ATCC 10712. Antimicrobial Agents And Chemotherapy, [S.L.], v. 58, n. 12, p. 7441-7450, dez. 2014. American Society for Microbiology. http://dx.doi.org/10.1128/aac.04272-14.
GARRIDO, L. M. et al. Insights in the glycosylation steps during biosynthesis of the antitumor anthracycline cosmomycin: characterization of two glycosyltransferase genes. Applied Microbiology Biotechnology, v. 73, p. 122-131, 2006.
LOBATO, Ana Katerine de Carvalho Lima. Metabolic Flux Analysis for the Optimization of Cosmomicin Antibiotic Synthesis by Streptomyces olindensis ICB20. 2010. 209 p. Thesis (Doctorate in Research and Development of Regional Technologies) - Federal University of Rio Grande do Norte, Natal, 2010.
Protein Model
Regarding the proposed structure of SARP (Streptomyces Antibiotic Regulatory Protein), we used the genetic sequence found in NCBI (National Center for Biotechnology Information) databases to predict the protein structure, as it has not yet been resolved from of techniques in the field of crystallography, such as crystallization.
The prediction of the 3D structure of the protein SARP was done using AlphaFold2 on ColabFold, where the input was the aminoacid sequence of Steptomyces olindensis’s SARP. It works by taking the input and searching for homologous sequences on UniRef100 and a database of environmental sequences via MMseqs2, creating alignments that will be used to predict the structure based on residue index manipulation, next, the model was validated in Expasy, with the SWISS-MODEL program.
We can use the SARP structure to assess the interaction between it and potential ligands such as the SARP-box of DNA, the regulatory region of the cluster, and the RNA polymerase that will initiate the transcription of the cluster's genes (TANAKA; TAKANO; OHNISHI; HORINOUCHI, 2007).
We used the HDOCK program to measure the interaction between SARP and a sigma factor domain of the RNA polymerase.
To determine the interaction site of SARP with the DNA in the specific region, we opened the structure predicted by AlphaFold in the Pymol program and chose the surface representation to locate the presence of any cavity that could be a potential interaction site.
With the validated model, we used the protein with the DNA sequence, which the SARP interaction, to define how this interaction occurs through a molecular dynamics simulation, and the region was defined based on the helix turn helix sequences, where is known proteins interact with portions of DNA.
References:
TANAKA, Akiko; TAKANO, Yuji; OHNISHI, Yasuo; HORINOUCHI, Sueharu. AfsR Recruits RNA Polymerase to the afsS Promoter: a model for transcriptional activation by sarps. Journal Of Molecular Biology, [S.L.], v. 369, n. 2, p. 322-333, jun. 2007. Elsevier BV. http://dx.doi.org/10.1016/j.jmb.2007.02.096.