echemi logo
Product
  • Product
  • Supplier
  • Inquiry
    Home > Biochemistry News > Biotechnology News > Science: Artificial intelligence scientists develop more powerful gene vectors

    Science: Artificial intelligence scientists develop more powerful gene vectors

    • Last Update: 2020-06-07
    • Source: Internet
    • Author: User
    Search more information of high quality chemicals, good prices and reliable suppliers, visit www.echemi.com
    Before we can understand this heavyweight paper, published in Science by renowned molecular biologist Professor George MChurch, let's see what adeno-related viruses (AAVs) areAAV is a single-stranded DNA virus that scientists discovered in the preparation of adenoviruses in 1965 and is therefore named adenovirusAdenovirus will infect a variety of vertebrates, including humans, will induce human upper respiratory tract infection, and AAV is very special, the current scientific consensus is that it will not lead to any human disease, but also the current human discovery of a class of the simplest structure of single-stranded DNA defect virus, precisely because of its DNA structural defects, without the participation of auxiliary viruses (typical such as adenovirus, herpes simplex virus, etc.), AAV can not cause viral infectionThis characteristic makes AAV the most ideal gene carrier for scientistsunmodified natural-based wild AAVs consist of a protein crust (capside) and a 4.7kb single-stranded DNA genomeThe protein crust consists of three sub-bases, VP1, VP2, and VP3At both ends of the AAV genome are two "T" type end reverse repeat sequences (inverted terminal repeat, ITR)These two ITRs are the starting point for virus DNA replication and the signal that triggers the virus packagingThe rep gene in the AAV genome encodes four proteins associated with virus replication, namely Rep78, Rep68, Rep52, and Rep40scientists have acquired a good vector for gene therapy, recombinant adeno-related viruses (rAAVs), by modifying AAVsThe protein crust carried by rAAV is almost identical to that of a wild AAV, but the portion of the virus protein encoded in the genome of the crust is completely removed and replaced by therapeutic genetically modified (transgene)Now, the remaining parts of the AAV genome are mainly cap genes that encode shell proteins, as well as ITRs, which serve to guide genome replication and viral vector assemblyThe advantage of completely removing the portion of the encoded viral protein is that, on the one hand, it maximizes the capacity of the recombinant AAV to carry GM, and on the other hand, reduces the immunogenicity and cytotoxicity produced when the virus is delivered to the bodyBut the current rAAV still doesn't meet the needs of gene therapy, and we urgently need a more powerful AAV, and this time Professor George MChurch, a renowned molecular biologist at Harvard University, led the research team in conjunction with Dyno Therapeutics, a biotech company that uses artificial intelligence in gene therapy, to successfully acquire a number of high-performance AAVs by combining computer technologyprevious research strategies focused on optimizing the cap gene, making it more efficient to deliver, but also have the activity of infestation, research strategy is limited to random mutations, the effect is not ideal As a result, the team performed a single mutation of 735 amino acid sites in the adeno-related virus type 2 (AAV2) crust, resulting in a single mutation library containing about 200,000 variants To study their function, the researchers transferred the mutants to mice to see how rich they were in different organs in mice For example, some mutants are deliberately enriched in the liver, while others are in the blood This phenomenon is also known as the "homecoming" phenomenon At the same time, through clever experimental design, they identified the corresponding changes in the clothing and shell, and the corresponding relationship with the mutation site, and established a computer model but in the actual design, single mutation may not meet the needs of gene therapy, to adopt multi-mutation site design, but also aAV2 vitality To this end, they used computer models to predict some combinations of multiple point mutations, and compared them with random mutation combinations, and finally found that computer-designed mutants, many of which have high AVV2 vitality, but also maintain their "homecoming" potential Surprisingly, the team also found a new auxiliary protein hidden in the shell-coded DNA sequence, which binds to the target cell membrane this paper builds the most comprehensive AAV protein case library to date Professor Crunch said: "Using the data generated by this library, we were also able to design more shell mutants than previous natural or artificial mutations What's more, AI designs produce effective shells that are far more efficient than AAVs produced by random mutagenic methods "
    " these high-throughput technologies combined with computer technology, for the future gene therapy laid a solid foundation Dr Eric Kelsic, current CEO of Dyno Therapeutics and co-author of the paper, said that past methods, such as man-made design or random mutations, had their own shortcomings, either due to the size of the mutation library or poor quality Machine-assisted design is a data-driven protein engineering method, and a simple mathematical model with sufficient data can successfully produce a viable synthetic shell With the help of the power of the computer, it is possible to fully combine the iteration and experience of the above-mentioned protein engineering to produce a large number of high-quality shell variants , the authors also found that the cap gene can also encode a new protein, the pop-associated accessy protein The authors speculate that the MAAP protein may be related to the high genome-shell coupling phenomenon previously found in the engineered AVV2 library MAAP is present in most AAV serotypes, and researchers believe it will play a role in the virus's natural life cycle "Studying the function of MAAP is an exciting area and helps people better understand AAV and design better AAV gene therapies," the researchers said The findings are encouraging, but they are only the first step Using this data and data from future experiments, we can build machine learning models to optimize the AAV carrier shell and solve the challenges of various gene therapy "This study is a landmark development and a good start So since 2015, the research team has focused on overcoming existing technical limitations by developing new machine-guided technologies, and today announced the development of a faster and more efficient tool-based AAv This study is a landmark, using new high-throughput measurement techniques to collect large amounts of data and teach them how to build a better multi-point mutant library, ultimately optimizing aAV delivery performance "This is just the beginning of a machine-led AAV shell engineering gene-change therapy, and the success of this study has shown us the limitless potential of the quest for more data and a larger capacity of machine learning models to be applied to gene therapy," the researchers said the researchers' idea is that, given the different mutants in the AAV2 single mutant library have been observed to be rich in different organs, and the corresponding AAV2 has corresponding mutation sites and shell structures, the researchers built a computer model to link the two To simplify the model, they selected those AAV2 mutants that were enriched in the liver and sequenced them one by one to study the role of these mutants in the body, the researchers infected AAV2 one by one in mice, where the mutants had different biological distribution characteristics, such as distribution in the kidneys, heart, liver, lungs, and so on The researchers then performed a master component analysis that linked the structural characteristics of different AAV2 mutants to their distribution characteristics within the organism; the results of the cluster analysis showed that some mutants were specifically removed from the liver, while they were enriched in the blood, heart and kidneys, and some were the opposite image source: Reference s3
    Considering that many previous random mutations produced AAV2 can not be effective gene transmission, the researchers also came up with the idea of "can create a computer method to improve AAV2 more effectively", to verify the idea they tried to use the computer mutation site design in view of the different mutants of aAV2 single mutant library are rich in different organs, and the corresponding AAV2 has corresponding mutation site and shell structure, the researchers built a computer model to link the two To simplify the model, they selected those AAV2 mutants enriched in the liver, sequenced them one by one, and sequenced the AAV2 mutants enriched in the liver region, whose mutation sites were limited to the 561-588-bit of amino acids encoded by the cap gene, so they identified this region as the target area of multiple point mutation selection scan amino acid sites in the target area of the cap gene candidate, the computer model will rate the amino acid sites according to the probability calculated by the model, the higher the score, the greater the probability The researchers then mutated the high-value dislocations together to build a multi-point mutant library at the same time, according to the principle of amino acid site effect and randomness, they selected some amino acid sites for mutation, as a control In this way, they designed 1,271 AVV2 mutants and 1,047 random mutants, which they then transferred to mice to detect their distribution The final results showed that about 25.6% of the computer-designed mutants were functional (i.e distributed in the liver), while nearly half (4,477) randomly produced mutants were ineffective (no distribution or weak distribution in the liver) this result shows that computer design is quite efficient References: 1 S Improved AAV Vector Capsid for Gene Therapy Engineered with a NewMachine-Guided Approach Retrieved Nov 29, 2019 from https:// https://science.sciencemag.org/content/366/6469/1139 https:// .
    This article is an English version of an article which is originally in the Chinese language on echemi.com and is provided for information purposes only. This website makes no representation or warranty of any kind, either expressed or implied, as to the accuracy, completeness ownership or reliability of the article or any translations thereof. If you have any concerns or complaints relating to the article, please send an email, providing a detailed description of the concern or complaint, to service@echemi.com. A staff member will contact you within 5 working days. Once verified, infringing content will be removed immediately.

    Contact Us

    The source of this page with content of products and services is from Internet, which doesn't represent ECHEMI's opinion. If you have any queries, please write to service@echemi.com. It will be replied within 5 days.

    Moreover, if you find any instances of plagiarism from the page, please send email to service@echemi.com with relevant evidence.