From gene editing to protein structure resolution to quantum computing, here are seven technologies that could have a significant impact on the scientific community.
The telomere-to-telomere (T2T) collaboration group is sequencing all chromosomes | Adrian T. Sumner/SPL
1 Full genomics
In 2019, Karen Miga, a genomics researcher at the University of California, Santa Cruz, and Adam Phillippy of the National Human Genome Institute of the United States established a telomere-to-telomere (T2T) collaboration, when about 1/10 of the human genome was still unknown. Today, that number has dropped to zero. In a preprint paper published last May, the T2T collaboration published the first end-to-end sequence of the human genome, adding nearly 200 million base pairs to the heavily used human reference genome sequence GRCh38 and completing the final chapter of the Human Genome Project[1].
GrCh38, first announced in 2013, is an important research tool and a framework for drawing sequencing sequences, but there are still many "skylights" on this framework. This is largely because the widely used sequencing technique (developed by Illumina, Calif.) is too long to clearly map highly repetitive genomic sequences, including telomeres at the ends of chromosomes and centromeres that coordinate the division of newly replicated DNA in cell division.
Long-reading long sequencing technology has been shown to change previous sequencing rules. Jointly developed by Pacific Biosciences and Oxford Nanopore Technology (ONT) in the United Kingdom, the technology can read tens of thousands or even hundreds of thousands of bases at once and sort them, although it was not absolutely accurate at first.
However, when the T2T collaboration group first recombined separate X and 8 chromosomes in 2020[2,3], Pacific Biosciences' sequencing technology has allowed scientists at T2T to detect tiny mutations in long fragment repeats. These tiny "fingerprints" make long, repetitive chromosomal fragments easier to handle, and the rest of the genome can be quickly pinpointed. ONT's platform also found a number of DNA modifications that regulate gene expression, and the T2T collaboration was able to map these "epigenetic markers" genome-wide [4].
The genome, which was resolved by the T2T collaboration, comes from a cell line containing two identical sets of chromosomes. In the normal diploid human genome, each chromosome has two versions, and researchers are working on a "genotyping" strategy that can accurately assign each sequence to the corresponding chromosome copy. Miga said: "We've got some really good parting assemblies. ”
The diploid assembly was done in collaboration with the Human Pangenome Reference Consortium, a T2T collaboration that hopes to create a more representative genomic map based on hundreds of donations from around the world.
Erich Jarvis, a geneticist at Rockefeller University in New York and principal investigator of the collaboration, said: "Our goal is to understand the average 97 percent of human allele diversity. As chair of the Vertebrate Genomes Project, Jarvis also hopes to obtain the complete genetic sequences of every vertebrate species on Earth through the assembly capabilities of these complete genomes. "I believe that in the next 10 years, telomere-to-telomere genome assembly will be the norm," he said. ”
2 Protein structure resolution
Structure determines function, but it is difficult to identify structure. Over the past two years, advances in experimentation and computing have provided more hands-on tools that allow researchers to resolve protein structures at unprecedented speed and resolution.
The AlphaFold2 structure prediction algorithm, developed by DeepMind, a London-based Subsidiary of Google, relies on a "deep learning" strategy to infer the shape of a folded protein's amino acid sequences.[5] At the 2020 protein structure prediction contest CASP, computational biologists competed on the same stage to compete with their respective protein structure prediction algorithms, but in the end, AlphaFold2 was invincible, and its popularity and popularity soared.
Janet Thornton, senior scientist and former director of the European Institute of Bioinformatics, said: "AlphaFold2's predictions of certain structures can be said to be surprisingly good. "Since its public release last July, AlphaFold2 has been applied to proteomics studies to determine the structure of all proteins in humans [6] and 20 patterns of biological expression (see Nature 595, 635; 2021; Disrupting Life Sciences! AlphaFold predicts the structure of the complete human proteome) and is used to identify the structures of nearly 440,000 proteins in the Swiss-Prot database, greatly increasing the number of proteins with high-confidence modeling data. AlphaFold2's algorithm also confirms its ability to resolve multi-stranded protein complexes [7].
At the same time, the evolution of cryo-EM has also enabled researchers to use experimental methods to deal with the most intractable proteins and their complexes. Cryo-EM uses an electron beam to scan rapidly frozen molecules, generate images of proteins from multiple angles, and then reassemble these images into a 3D structure through calculations. In 2020, after the hardware and software of cryo-EM were upgraded, the two teams achieved a structural resolution of less than 1.5 angstroms, determining the position of individual atoms [8,9].
Bridget Carragher, co-director of the Simmons Center for Electron Microscopy at the Center for Structural Biology in New York, said: "Before that, although we could say 'atomic resolution' at every turn, it could only be considered close to the atomic level, and this is the real atomic level." Carragher said that although both teams used deferitin, a well-studied pattern protein, their study showed that for other, more difficult targets, reaching near-atomic resolution is also feasible.
Images from cryo-EM can help resolve complex structures | Paul Emsley/MRC Molecular Biology Laboratory
Many experimentalists who initially believed in AlphaFold2 now see it as an effective complement to experimental methods such as cryo-EM. AlphaFold2's computational model can help with data analysis and reconstruction, while cryo-EM can discover structures that are currently impossible to predict with computers. For example, Carragher's team is using "time-resolved" cryo-EM to capture rapid conformational changes as proteins interact with other molecules. "We can freeze the change and see what's really happening in 100 milliseconds," she said.
Another related technique is called cryo-electronic tomography (cryo-ET), a method that captures naturally occurring protein behavior in thin sections of frozen cells. However, these complex images are very difficult to interpret. Carragher believes that advances in computing power in machine learning will be indispensable. She asked, "How else can we solve these nearly insoluble problems?"
3 Quantum simulation
Atoms are definitely atomic in size. But under the right conditions, the atomic energy is in a highly excited state, with a diameter of 1 micron or greater. By performing controlled excitation on a carefully arranged array of hundreds of atoms, physicists have demonstrated that they can solve challenging physics problems that push the limits of traditional computers.
Quantum computers process data in qubits. Qubits are coupled together by quantum mechanical phenomena called entanglements, which in turn affect each other over a certain distance. These qubits can significantly increase computing power relative to the same number of bits in classical computers.
Multiple teams have successfully utilized individual ions as qubits, but the charges they carry make it difficult to assemble at high densities. Physicists such as Antoine Browaeys of France's National Center for Scientific Research (CNRS) and Mikhail Lukin of Harvard University are working on another approach. Their team used optical forceps to precisely immobilize the uncharged atoms in tightly arranged 2D and 3D arrays, and then excited the particles into large-diameter Rydberg atoms with lasers, entangled them with nearby atoms [10,11]. Jaewook Ahn, a physicist at the Korea Advanced Institute of Science and Technology, explains, "The Reedberg atomic system is independently controllable, and their interactions can be turned on and off. "This, in turn, gives it programmability.
This approach shined in just a few years, with technological advances improving the stability and performance of the Reedberg atomic array, and the number of qubits rapidly expanding from dozens to hundreds. While the early applications of the technology focused on issues that have already been raised, such as predictions of material properties, its uses are very broad. Browaeys said: "So far, any theoretical model proposed by theorists has its own way of implementing it. ”
Pioneers in the field have already formed companies that are currently developing the Rydberg Atomic Array System for laboratory use, and Browaeys expects the quantum simulator to be commercially available within a year or two. The work also paves the way for a wide range of applications in quantum computers, including in the fields of economics, logistics, and cryptography. Researchers have yet to determine the innovative technology's place in computing, but Ahn likens it to the Wright brothers' early explorations in aviation. "The first plane had no transportation advantage to speak of, but it changed the world," he said. ”
4 Precise genome regulation
Despite its amazing genome-editing capabilities, CRISPR-Cas9 technology is better suited for inactivating genes than for gene repair. This is because the Cas9 enzyme targets the genome sequence, although it is still accurate, but the cell repairs the subsequent double-strand cleavage is not accurate. CRISPR-Cas9 fixes are usually mediated by a process called non-homologous end-joining, which is often mixed with problems with small fragment insertions or missing pieces.
David Liu, a chemical biologist at Harvard University, points out that most genetic diseases require genetic correction rather than gene destruction. Liu and his team have developed two promising approaches to this. Both methods take advantage of CRISPR's ability to precisely target while limiting Cas9's ability to cut DNA at that site.
The first method, called base editing, binds catalyzed damaged Cas9 to an enzyme that converts one nucleotide into another—for example, from cytosine to thymine, or from adenine to guanine (see Nature https://doi.org/hc2t; 2016; CRISPR technology is on the rise! It is now possible to edit individual bases of DNA). However, only specific base-base conversions can be achieved using this method.
The second approach is prime editing, the team's latest research, linking Cas9 to reverse transcriptase and using a modified guide RNA that integrates the desired edits into the genome sequence (see Nature 574, 464–465; 2019; Gene editing research is next: more accurate CRISPR tools are available). After a multi-stage biochemical process, these components copy the guide RNA into the DNA that eventually replaces the sequence of the genome of interest.
The point is that both base editing and guided editing cut only one strand of DNA, which is a safer and less destructive process for cells.
Base editing, first reported in 2016, is moving toward clinical applications: Beam Therapeutics, founded by Liu in Cambridge, received approval from the U.S. Food and Drug Administration (FDA) in November to test the technology for the first time in humans to repair the gene that causes sickle cell disease.
Although the emergence of guided editors is still not long, it has been upgrading and its development prospects are also very clear. Hyongbum Henry Kim, a genome editing expert at Seoul's Yonsei University School of Medicine, and his team demonstrated that using guided editing to correct retinal gene mutations in mice can achieve a 16 percent effectiveness rate [12]. "If we use the more advanced version of the latest report, there will be a greater increase in efficiency," he said. Liu's team also found that guided editing can help insert DNA sequences of gene size into the genome, promising to become a safer, more strictly controllable gene therapy [13]. Liu Ruqian said: "Although the effective rate of this method is not high, even a small number of repairs can be of great benefit. In some cases, if you can replace a gene with a 10% or even 1% effective rate, the disease can be saved. ”
5 Targeted gene therapy
Nucleic acid-based drugs can have a significant impact in the clinic, but there are still many limitations to the tissues they can apply. Most drugs are either administered topically or need to be transplanted back after cells are extracted from the patient for in vitro processing. But there is one exception – the liver. The liver can filter blood and has proven to be a reliable target for selective drug delivery. In this case, intravenous or even subcutaneous injections can be used.
Daniel Anderson, a chemical engineer at the Massachusetts Institute of Technology, said: "When you think about it seriously, it's hard enough to deliver drugs to any organization. Our bodies are inherently good at using existing genetic information and don't like to accept outsiders. "However, researchers are developing strategies that can direct drugs into specific organ systems without affecting other non-targeted tissues, and these efforts are making steady progress."
Adeno-associated viruses are the vector of choice for many gene therapies, and animal studies have shown that careful selection of appropriate viruses, coupled with tissue-specific gene promoters, enables efficient drug delivery confined to specific organs [14]. However, viruses are sometimes difficult to produce on a large scale and can mobilize immune responses, disrupt efficacy or produce adverse reactions.
Lipid nanoparticles are a non-viral vector, and multiple studies published over the past few years have demonstrated the potential to regulate their specificity. For example, the selective organ targeting (SORT) method developed by Daniel Siegwart et al., a biochemist at the University of Texas Southwestern Medical Center, can help rapidly synthesize and screen lipid nanoparticles to find nanoparticles that can effectively target tissue (such as lungs or spleen) cells [15].
Roy van der Meel, a biomedical engineer at Eindhoven University of Technology in the Netherlands, said: "This is one of the earliest papers that shows that if these lipid nanoparticles are systematically screened and their composition is changed, their biological distribution can be disturbed. Anderson says many teams are also investigating how protein components, such as cell-specific antibodies, can be used to assist in this targeting process.
Anderson is particularly excited about the preclinical advances made by companies like Beam Therapeutics and Intellia on targeting blood and immune cell precursors in the bone marrow, both of which use specially designed lipid nanoparticles. If these tissues are successfully targeted, he said, patients can get rid of the pain of current in vitro gene therapy, including killing existing bone marrow cells with chemotherapy before transplantation. Anderson said: "Putting these tasks in the body to complete may completely change the concept of treatment." ”
6 Spatial multiomics
In 2016, a team led by Joakim Lundeberg, a researcher at the Royal Swedish Institute of Technology, devised a strategy to address these problems. The team prepared slides with barcoded oligonucleotide — RNA or DNA short strands — that could capture messenger RNA from intact tissue slices so that each transcript could correspond to a specific location in the sample according to its barcode. Lundeberg said: "No one believed before that we could actually perform a full transcriptome analysis from a single tissue slice, but it turned out to be very simple. ”
Since then, the field of spatial transcriptomics has ushered in a major explosion. A variety of commercial systems are available, including the Visium Spatial Gene Expression platform developed by 10x Genomics using Lundeberg's technology. Many academic research teams are also working on new ways to map gene expression with better depth and spatial resolution.
The CRISPR-Ca9 gene-editing complex uses a guide RNA (red) to cut DNA (blue). Source: Mulekuul/SPL
Now, the researchers are superimposing omics data on their spatial atlases. For example, Rong Fan, a biomedical engineer at Yale University, has developed a platform called DBiT-seq16 that uses a microfluidic system that can simultaneously generate barcodes for thousands of mRNA transcripts and hundreds of proteins labeled with labeled oligonucleotide antibodies.
This gives a more accurate assessment of how cell gene expression affects protein production and activity than using transcriptome data alone, and Fan's team has been using it to study processes such as immune cell activation. "We're seeing early signs of how skin immune cells are responding to the Moderna covid-19 vaccine," he said. "Some commercial systems can also take spatial data from multiple proteins while acquiring transcriptome data, including the Visium platform here and Nanotring's GeoMx system.
At the same time, Lundeberg's team also improved their spatial transcriptomics approach to capture DNA sequence data simultaneously. This allowed his team to begin mapping the spatiotemporal events behind tumorigenesis. "We can track the spatial variations of these genes and see how they evolve additional genetic variants that ultimately lead to tumors," he said. ”
Fan's team has demonstrated how to spatially localize chromatin modifications in tissue samples, which can reveal cellular gene regulation that influences processes such as development, differentiation, and intercellular communication [17]. Fan believes this approach can be combined with spatial analysis of RNA and even proteins. "Our data shows preliminarily that this is possible," he said. ”
7 CRISPR-based diagnostics
The CRISPR-Cas system's ability to precisely cut specific nucleic acid sequences stems from its role as a bacterial "immune system" to fight viral infections. This connection inspired the first researchers to use the technique to think about its applicability for viral diagnosis. Pardis Sabeti, a geneticist at the MIT-Harvard Broad Institute, says, "It's the right thing to do with their innate function, after all they've evolved for billions of years." ”
But not all Cas enzymes are exactly the same. Cas9 is the enzyme of choice for CRISPR genome manipulation, but much of the crispr-based diagnosis has used Cas13, a family of targeted RNA molecules that was first discovered in 2016 by Feng Zhang, a molecular biologist at the Broad Institute, and his team.
Jennifer Doudna of the University of California, Berkeley, explains, "Cas13 uses its RNA guide to identify RNA targets through base pairing and activate ribonuclease activity, which can be utilized through the reporter RNA as a diagnostic tool." Doudna and Emmanuelle Charpentier, who currently works at the Max Planck Institute for Pathogen Science, were jointly awarded the 2020 Nobel Prize in Chemistry for developing CRISPR-Cas9's genome editing capabilities. That's because Cas13 not only cuts the RNA targeted by the guide RNA, it also performs "collateral cleavage" on all other nearby RNA molecules. Many Cas13-based diagnostics use a reporter RNA that tethers fluorescent markers to quenched molecules that inhibit fluorescence.
When Cas13 is activated after recognizing viral RNA, it cuts off the reporter gene and releases fluorescent markers from quenched groups, producing a detectable signal. Some viruses release strong signals that can be detected without amplification, greatly simplifying point-of-care diagnostics. For example, last January, Doudna and Melanie Ott of the Glese Institute of Virology in San Francisco demonstrated a rapid NASSPR-Cas13-based detection method that uses mobile phone cameras to perform non-amplified detection of the coronavirus (SARS-CoV-2).
RNA amplification can improve sensitivity to trace viral sequences, and Sabeti and her colleagues have developed a microfluidic system that uses genetic material amplified from just a few microliters of sample to screen for multiple pathogens at the same time [19]. "Currently, we have a way to detect 21 viruses at the same time, and each sample costs less than $10," she said. She also said they have developed a CRISPR-based tool that can detect more than 169 human viruses at the same time.
Doudna said other Cas enzymes are expected to continue to expand the diagnostic toolbox, including the Cas12 protein, which has similar properties to Cas13 but targets DNA rather than RNA. These enzymes can allow for a wider range of detection of pathogens and even rapid diagnosis of other noncommunicable diseases. "It would be very useful if you could increase the speed, especially considering that different cancer subtypes have now begun to be classified by specific types of mutations," Doudna said. ”
bibliography
[1] Nurk, S. et al. Preprint at bioRxiv https://doi.org/10.1101/2021.05.26.445798 (2021).
[2] Miga, K. H. et al. Nature585, 79–84 (2020).
[3] Logsdon, G. A. et al. Nature593, 101–107 (2021).
[4] Gershman, A. et al. Preprint at bioRxiv https://doi.org/10.1101/2021.05.26.443420 (2021).
[5] Jumper, J. et al. Nature596, 583–589 (2021).
[6] Tunyasuvunakool, K. et al. Nature596, 590–596 (2021).
[7] Evans, R. et al. Preprint at bioRxiv https://doi.org/10.1101/2021.10.04.463034 (2021).
[8] Nakane, T. et al. Nature587, 152–156 (2020).
[9] Yip, K. M., Fischer, N., Paknia, E., Chari, A. & Stark, H. Nature587, 157–161 (2020).
[10] Bernien, H. et al. Nature551, 579–584 (2017).
[11] Barredo, D., Lienhard, V., de Léséleuc, S., Lahaye, T. & Browaeys. A. Nature561, 79–82 (2018).
[12] Jang, H. et al. Nature Biomed. Eng. https://doi.org/10.1038/s41551-021-00788-9 (2021).
[13] Anzalone, A. V. et al. Preprint at bioRxiv https://doi.org/10.1101/2021.11.01.466790 (2021).
[14] Levy, J. M. et al. Nature Biomed. Eng. 4, 97–110 (2020).
[15] Cheng, Q. et al. Nature Nanotechnol.15, 313–320 (2020).
[16] Liu, Y. et al. Cell 183, 1665–1681 (2020).
[17] Deng, Y. et al. Preprint at bioRxiv https://doi.org/10.1101/2021.06.06.447244 (2021).
[18] Fozouni, P. et al. Cell184, 323–333 (2021).
[19] Ackerman, C.M. et al. Nature582, 277–282 (2020).
Originally written by Michael Eisenstein
The original article was published under the title Seven technologies to watch in 2022 on a technical feature of Nature on January 25, 2022
nature
Copyright Notice:
2022 Springer Nature Limited. All Rights Reserved
If you want to know more about industry and technology venture capital news,
Welcome to pay attention to the fruit shell hard technology