Genetics Updates vol.61

November 3rd 2024

Check out latest research updates in the field

article image

TL;DR

  • Smart supramolecular assemblies
  • Naturally occurring DNA-protein hybrids
  • Breakthrough study on how new genes evolve
  • How cells ‘repress’ genomic remnants of ancient viruses
  • Double-edged sting: study identifies new pathway involved in aging

Overview

Genetic technology is defined as the term that includes a range of activities concerned with the understanding of gene expression, advantages of natural genetic variation, modifying genes and transferring genes to new hosts. Genes are found in all living organisms and are transferred from one generation to the next. Gene technology encompasses several techniques including marker-assisted breeding, RNAi, and genetic modification. Only some gene technologies produce genetically modified organisms.

Modern genetic technologies like genome editing would not be possible without all the previous generations of genetic technologies that have enabled scientists to discover what genes are, what they do, and how DNA can be modified to add, remove, or replace genes. You can find major genetic technology development milestones via the link.

Latest Research

Additive-assisted macroscopic self-assembly and control of the shape of assemblies based on host–guest interaction

by Akihito Hashidzume, Takahiro Itami, Masaki Nakahata, Yuri Kamon, Hiroyasu Yamaguchi, Akira Harada in Scientific Reports

If you’ve ever opened a box from a furniture store and wished the pieces inside could somehow spontaneously merge to form a table or chair, then a simple virus could have a thing to two to teach you.

Self-assembly of complex molecules is essential for a wide array of biological structures, including proteins, cell membranes, or even entire viruses. Supramolecular chemistry is a field of study that attempts to build large molecules out of a discrete number of smaller building blocks. By altering the strength of attraction between different polymers, complexes can be constructed on demand, leading to the development of ‘smart materials’ that respond to changes in their environment, such as the addition of a new chemical. However, many aspects of supramolecular chemistry remain poorly understood.

Now, researchers from Osaka University showed how additives can promote the self-assembly of spherical microparticles made of a super absorbent polymer, poly(sodium acrylate), as well as control the macroscopic shape of the resulting assemblies. Some of the polymer molecules were functionalized with the specific chemical, β-cyclodextrin (βCD), and others with adamantane (Ad) residues. However, the microparticles did not assemble until a critical threshold concentration of the additive 1-adamantanamine hydrochloride (AdNH3Cl) was introduced. The researchers took inspiration from biological proteins, which consist of long chains of smaller units called amino acids. Certain attractions or repulsion between amino acids, including hydrogen bonding, electrostatic interaction, or hydrophobic interactions, can control the folded shape of the resulting protein. Similar effects can occur among other large biomolecules, including DNA, polysaccharides, and lipids.

“In a certain sense, all living organisms are just collections of supramolecular polymers with sophisticated functions,” lead author of the study Akihito Hashidzume says.

The team analyzed the behavior of the macroscopic assembly of spherical microparticles, and found that the resulting shape, whether more spherical or elongated, could be controlled based on the AdNH3Cl concentration. This suggests that stimuli, such as heat and force, can be used to control the shape of the assemblies.

“The results in this study might help us understand the origin of various shapes of organisms,” senior author Akira Harada says. This research could also assist in future work on the control of macroscopic assemblies based on microscopic interactions, as well as novel active materials that change depending on their situation.

 

Biosynthesis of peptide–nucleobase hybrids in ribosomal peptides

by Zeng-Fei Pei, Natalia M. Vior, Lingyang Zhu, Andrew W. Truman, Satish K. Nair in Nature Chemical Biology

Thanks to a serendipitous discovery and a lot of painstaking work, scientists can now build biohybrid molecules that combine the homing powers of DNA with the broad functional repertoire of proteins — without having to synthesize them one by one, researchers report in a new study. Using a naturally occurring process, laboratories can harness the existing molecule-building capacities of bacteria to generate vast libraries of potentially therapeutic DNA-protein hybrid molecules.

“Two of the most common building blocks in biology are nucleic acids — used for making RNA and DNA — and amino acids, which make up proteins,” said University of Illinois Urbana-Champaign biochemistry professor Satish Nair, who led the study with postdoctoral researcher Zeng-Fei Pei. “We have these two sets of biological molecules that do very different things, and, for decades, chemists have been trying to integrate them into the same molecule. If you can make a complex protein and then put a nucleic acid on it that makes it go exactly where you want it to go because it will bind to specific regions of DNA or RNA, you can build a precision drug.”

Such drugs can be used to interrupt various disease-promoting processes in the cell, blocking the transcription of mutated genes, for example, or binding to pathogenic noncoding RNA molecules to stall their activity, Nair said.

The initial discovery was serendipitous, he said. He and his colleagues had been looking for proteins that bind to metals when they noticed that a team at the John Innes Centre in Norwich, England, had reported on a bacteria-generated molecule of interest that appeared to be a DNA-protein hybrid.

The Illinois team contacted the Innes Centre scientists, Natalia Vior and Andrew Truman, suggesting they re-examine the molecule to determine if it was in fact what it appeared to be. Once that initial discovery was confirmed, the American and English scientists collaborated on a more in-depth analysis to discover the molecular mechanisms that formed the hybrid.

Finding a naturally occurring DNA-protein hybrid and determining how a bacterium can be induced to make it would streamline what is now a slow, labor-intensive process, Nair said.

“Lots of high-powered laboratories all over the world have been using various synthetic chemical methods to make biohybrid molecules, and that’s great: They’re all proof-of-concept and it works,” he said. “The problem is that you can’t do it at large scale. You can’t make 100 million compounds because that would require you to do the chemical synthesis 100 million times.”

In a series of experiments, Nair and his colleagues found that two bacterial enzymes together convert certain peptides into DNA-protein hybrids. The first enzyme, YcaO, modifies an amino acid in the peptide to convert the peptide into a ring structure like the bases that allow DNA and RNA to pair with other DNA or RNA molecules. The second enzyme is a protease that cuts off one part of the newly modified molecule, converting it into a fully functional nucleobase-protein hybrid.

The team was able to make the conversion in a test tube by adding only three ingredients: the original peptide and the two enzymes. But they also demonstrated that the process could be carried out by the bacterium E. coli.

Understanding this process will allow laboratories to create hybrid molecules that can attach to any region of the genome or any RNA molecules in cells, Nair said. Using bacteria to streamline the pipeline will speed the process of discovery.

 

Diverse Origins of Near-Identical Antifreeze Proteins in Unrelated Fish Lineages Provide Insights Into Evolutionary Mechanisms of New Gene Birth and Protein Sequence Convergence

by Nathan Rives, Vinita Lamba, C H Christina Cheng, Xuan Zhuang in Molecular Biology and Evolution

Where do new genes come from? That’s the question a team of biological sciences researchers from the U of A set out to answer in a new study.

They did so by examining the evolution of antifreeze proteins in fish — an essential adaptation that allows fish to survive in freezing waters by preventing ice formation through the binding of their antifreeze proteins to ice crystals.

The team investigated these proteins in three unrelated fish lineages and uncovered surprising results. While the proteins in each lineage are functionally and structurally similar, they evolved independently from different genetic sources. This phenomenon, known as convergent evolution, represents a rare case of protein sequence convergence. It demonstrates how the same adaptive traits — and even nearly identical protein sequences — can be produced through entirely different evolutionary trajectories.

Genomic loci of AFPI and the neighboring genes in the three focal AFPI-bearing species from separate lineages. Arrows and triangles represent genes in their respective directions. The size of the arrows is not proportional to the actual length of the genes.

The study provides concrete examples of different evolutionary mechanisms that can lead to the birth of new genes. Findings suggest that new genes can form by repurposing fragments of ancestral genes while incorporating entirely new coding regions (the protein-coding parts of the DNA). This innovative concept bridges the gap between entirely new gene formation from noncoding regions and the more traditional model in which new functions can arise from duplicated genes. Co-authors included Nathan Rives, Vinita Lamba, C-H Christina Cheng and Xuan Zhuang. The co-first authors, Rives and Lamba, are Ph.D. students in the Zhuang Lab at the U of A, which is led by assistant professor of biological sciences Xuan Zhuang, who oversaw the study. Cheng is a professor in the School of Integrative Biology at the University of Illinois Urbana Champaign.

The group’s work also introduces a new model that advances understanding of the mechanisms behind new gene evolution: Duplication-Degeneration-Divergence. This model explains how new gene functions can arise from degenerated pseudogenes — formerly functional genes that lost their original role. This model also highlights how genes that appear to be nonfunctional or “junk” can evolve into something entirely new, a concept that holds significant implications for understanding adaptation under extreme environmental stress.

In the context of molecular evolution, this work represents a significant step forward in understanding how new genes are born and evolve, offering fresh perspectives on functional innovation — or gene recycling and adaptation.

 

Histone H3.3 lysine 9 and 27 control repressive chromatin at cryptic enhancers and bivalent promoters

by Matteo Trovato, Daria Bunina, Umut Yildiz, Nadine Fernandez-Novel Marx, Michael Uckelmann, Vita Levina, Yekaterina Perez, Ana Janeva, Benjamin A. Garcia, Chen Davidovich, Judith B. Zaugg, Kyung-Min Noh in Nature Communications

For any organism to survive and thrive, its cells must strictly control which genes are active when and where. New research from EMBL Heidelberg’s Noh Group and their collaborators from EMBL Australia sheds light on some of the key control sites that regulate this process, especially with respect to the activity of ancient viral sequences in the genome.

Our genomes are huge — a typical human cell contains DNA with over 6 billion units of information (measured in ‘base pairs’). However, this treasure trove of information poses a challenge when it comes to looking up the right information at the right time to perform a specific function. This is where epigenetic signatures come into play.

If you imagine the genome as a book, epigenetic marks are the highlights on its pages and the notes in its margins. Now, it is not always easy to know whether these marks are ‘instructive’ — i.e. do they tell the cell “Here, read this” or “Don’t read this”? Or are they simply marks left behind by a previous reader, indicating that that portion of the book was visited before?

It was this question that interested Kyung-Min Noh, Group Leader at EMBL Heidelberg, and her team. The researchers decided to focus on a molecule called H3.3, which belongs to a class of proteins called histones. Histones tightly bind to DNA in cells and help form its functional structure.

Transcriptional changes in H3.3K27A and K9A mESCs.

The H3.3 protein has a couple of spots on its tail (called K9 and K27) which are frequently chemically modified. It is hypothesised that these modifications are epigenetic marks that help the cell make decisions regarding gene expression. However, until now, it had never been experimentally proven that these are true control sites that instruct gene expression.

The researchers decided to experimentally mutate these sites, thus creating a version of H3.3 which could not be chemically modified at these spots. Considering our book analogy above, this created a protected page which could not be highlighted or marked, allowing the scientists to directly explore what the consequences of losing such marks would be.

Moreover, this system allowed the researchers to vary which page was protected, allowing them to draw comparisons between the loss of modifications at one or the other control site.

The scientists found that the mutation of these sites in mouse stem cells not only resulted in defects in cell differentiation, growth, and survival, it also caused spurious activation of genes across the genome. This included genes that should not be expressed in stem cells, such as immune system-specific genes.

This suggested that a normal function of these sites is to maintain those genes in an inactivated — or ‘repressed’ — state, allowing stem cells to remain stem cells. These effects were also different for the two control sites studied, showing that each of them plays a distinct role in gene regulation.

Upon further analysis, the researchers found that some of these regions, which are typically repressed but were activated upon mutating the histone sites, are ancient remnants of viruses that have integrated into our genomes.

“These regions are also called endogenous retroviruses (ERVs),” explained Matteo Trovato, former PhD student in the Noh group and the first author of the study, currently a postdoc at IFOM, Italy. “Throughout evolution, they have been co-opted by the host’s genome to exert regulatory functions. In immune cells, for example, 30% of the enhancers (a specific type of regulatory DNA element) are derived from ERVs.”

The researchers found that by modifying the K9 site in stem cells, many such ‘cryptic’ enhancers — regulatory DNA regions that are normally silenced — became active.

“Repression of these unique genomic regions is crucial for preserving the cell’s gene expression program balance,” said Noh. “Activation of the cryptic enhancers triggers a widespread rewiring of the gene regulatory network, ultimately impacting stem cell identity and functionality.”

The study was carried out in collaboration with Chen Davidovich’s group in EMBL Australia, Benjamin Garcia’s lab at Washington University, St. Louis, and Judith Zaugg’s team at EMBL Heidelberg.

“This is one of the first few studies conducted in a mammalian system showing that these histone residues play a causal role in gene regulation,” said Noh. “Understanding this process could have broader implications for developmental biology and disease research, particularly in cancer and neurological disorders, where gene regulation plays an essential role.”

 

A TBK1-independent primordial function of STING in lysosomal biogenesis

by Bo Lv, William A. Dion, Haoxiang Yang, Jinrui Xun, Do-Hyung Kim, Bokai Zhu, Jay Xiaojun Tan in Molecular Cell

A protein called STING, previously shown to control a pathway that contributes to antiviral signaling, also plays an important role in cellular stress clearance and cell survival, according to a new paper.

It was quite surprising that STING has a protective function for cells to reduce stress and damage in addition to its well-known role in inflammation,” said senior author Jay Xiaojun Tan, Ph.D., assistant professor at the University of Pittsburgh and UPMC Aging Institute and Pitt’s Department of Cell Biology.

“Our findings suggest that balance of STING’s two functions is important for the health of cells and could have implications for future development of therapeutics for age-related diseases,” added first author Dr. Bo Lv, Ph.D., a postdoctoral researcher in Tan’s lab.

In healthy human cells, DNA is packaged up inside the nucleus and mitochondria. When DNA leaks out into the fluid component of the cell known as the cytosol, it means that something is wrong.

“Cytosolic DNA is a danger signal associated with infections, cellular stress, cancer and other diseases,” explained Tan. “Cells have a warning system to detect DNA in the cytosol, which involves activation of STING, which in turn coordinates inflammation necessary to combat these threats.”

While short bursts of STING-mediated inflammation are crucial, in some people this pathway is chronically “on,” a state that has been linked with neurodegeneration and other diseases of aging, as well as normal aging.

To learn more about potential benefits of STING activation in response to diverse stresses, Tan and his team analyzed the full set of proteins within cells. They found that when STING was activated, two transcription factors called TFEB and TFE3 were shuttled to the nucleus of cells, where they activated genes that resulted in the production of more lysosomes.

“Lysosomes are organelles that are involved in autophagy, a cellular process that cleans up damaged material, almost like a housekeeping or recycling system,” said Tan. “In response to STING activation, cells used TFEB and TFE3 to produce more lysosomes and increase autophagy.”

Both lysosomes and autophagy are tightly linked with longevity and healthspan, the length of time that a person is healthy, suggesting that this protective function of STING is important for healthy aging.

STING-blocking therapies are currently being explored within the context of age-related diseases, but according to Tan, the new findings suggest that this strategy should be reconsidered because it would also block the autophagy/lysosome-promoting functions of STING. Instead, selectively targeting components of the inflammation pathway downstream of STING may be a better approach because it would preserve the protein’s beneficial functions.

Notably, TFEB and TFE3 are present across the animal kingdom, indicating that STING-induced autophagy-lysosome pathway is more evolutionarily ancient than its inflammation function, which is only found in vertebrates. The newly discovered function of STING may be an ancient way that cells maintain quality control, clear abnormal materials and manage cellular stress. Tan hypothesizes that mild cellular stress that activates STING may be important for maintaining lysosome quality and autophagy responses just like exercise improves our health by challenging our bodies.

“When we exercise regularly, we cause physical damage to our muscles, which triggers repair systems that over-repair and ultimately build muscle,” he said. “I want to understand whether challenging our cells with mild stress in general could boost stress response systems, including lysosome activity, and help delay age-related diseases and improve healthspan.”

Subscribe to Paradigm!

Medium. Twitter. Telegram. Telegram Chat. Reddit. LinkedIn.

Main Sources

Research articles

Nature Genetics

GEN: Genetic Engineering & Biotechnology News

National Institutes of Health

PLOS Genetics

Science

Science Direct

Tags

genetics
gt