Bowdoin Science Journal

Jared Lynch

Unveiling the True Potential of Telomeres in Cellular Health

by

Telomeres have long been known as repetitive DNA elements protecting the ends of chromosomes, but recently, reports of their transcription and translation have revealed an emerging pathological role upon abnormal expression.

When studying the nuclear genome of eukaryotes, one of the most basic characteristics is the organization of DNA into chromosomes that condense and duplicate during mitosis and meiosis. Within each chromosome, a critical protective feature is a string of repeated sequences at each end known as telomeres. These repeat sequences exist as (TTAGGG)n in mammals, which serve to prevent chromosome fusing and shortening. In fact, the shortening of telomeres throughout one’s lifetime is linked closely to aging, reflecting one of the many important health implications of telomeres [1].

More recently, these implications have grown with the discovery that telomeres are both transcribed and translated into repeat sequences (Figure 1, top) [2]. Transcription creates what is known as telomeric repeat-containing RNA (TERRA), and while most of these molecules never leave the nucleus, some escape to the cytoplasm [3]. Initially, TERRA was believed to have no substantial effect on cell functioning due to the absence of AUG start codons. However, its unique 3D structure is able to bypass this roadblock and initiate AUG-independent translation [4]. The result is one of two repeated protein sequences, depending on which reading frame translation begins at: valine-arginine (VR) and glycine-leucine (GL) [4]. In appreciable quantities, these are expected to impact cell functioning, but in what way is largely unknown, particularly in humans.

Al-Turki and Griffith sought to fill this knowledge gap by assessing the properties of VR repeats and GL repeats in vitro and in human cells [5]. To start, the researchers hypothesized that VR would bind nucleic acids, which are negative charged, due to its positively-charged arginines. By contrast, GL was predicted to form large aggregates known as amyloids due to its hydrophobic properties. In both cases, these molecules could severely impact cell functioning, so determining the exact behavior of VR and GL was an essential start.

For their first experiment, Al-Turki and Griffith synthesized VR and GL, then added them to a salt buffer to mimic how they would self-interact in the cytoplasm. Visualization revealed that VR was unable to aggregate due to the repulsion of repeated arginines. However, GL demonstrated mild aggregation with the potential to form amyloids and induce inflammatory responses in extreme cases (Figure 1). 

Next, the binding of VR to nucleic acids was tested by adding the peptide to a solution of RNA or DNA. In support of their hypothesis, the authors found that VR bound to both with a high affinity. They then took this a step further and added DNA plasmids containing a replication fork (i.e., the structure that forms during DNA replication) to test the interaction with VR. Intriguingly, VR bound with high preference to the replication fork itself which altered its geometry, possessing the ability to disrupt the replication process (Figure 1). 

Finally, an antibody for VR was developed which allowed the authors to visualize its presence in human cells. VR levels were found to positively correlate with the abundance of TERRA, and in osteosarcoma (“U2OS”) cells specifically, these were both at unusually high levels. Moreover, VR was often found in the nucleus as discrete clumps or “foci,” contrary to the in vitro experiment, due to stabilization by the negatively-charged DNA. The authors then attempted to alter VR levels by either knocking down TERRA or promoting its transcription. Intriguingly, VR aggregates increased in size for both cases, which may result in a cytotoxic effect that has yet to be explored.

Telomeres were initially viewed as protective elements with no function beyond the DNA level. Yet, the discovery of TERRA and its translated peptide sequences have completely flipped this narrative. Al-Turki and Griffith expand on this by showing that VR and GL peptides generated by translation could have significant physiological effects on cells including altered DNA replication, inflammation, and cytotoxicity. In particular, VR can not only bind DNA but also aggregate in the nucleus. These aggregates increased in size when altering TERRA levels, indicating that dysfunctional telomeres capable of promoting TERRA could have negative consequences for cell functioning.

In spite of these discoveries, future work remains in order to understand the full picture of VR and GL. For one, GL was unable to be assessed in live cells as the authors were unable to create an antibody for it. Additionally, these peptides are likely to have other functions beyond aggregating and killing cells. Understanding these abilities is crucial for not only characterizing its pathological potential but also for developing treatments down the line. Nevertheless, the present study creates a strong foundation for exploring this underappreciated chromosomal feature.

References

1. Ren F et al. (2009). Estimation of human age according to telomere shortening in peripheral blood leukocytes of Tibetan. Am J Forensic Med Pathol, 30(3):252-5. https://doi.org/10.1097/PAF.0b013e318187df8e.

2. Azzalin CM et al. (2007). Telomeric repeat containing RNA and RNA surveillance factors at mammalian chromosome ends. Science, 318(5851):798-801. https://doi.org/10.1126/science.1147182. 

3. Schoeftner S, Blasco MA (2008). Developmentally regulated transcription of mammalian telomeres by DNA-dependent RNA polymerase II. Nat Cell Biol, 10(2):228-36. https://doi.org/10.1038/ncb1685.

4. Zu T et al. (2010). Non-ATG-initiated translation directed by microsatellite expansions. Proc Natl Acad Sci U S A, 108(1):260-5. https://doi.org/10.1073/pnas.1013343108. 

5. Al-Turki TM, Griffith JD (2023). Mammalian telomeric RNA (TERRA) can be translated to produce valine-arginine and glycine-leucine dipeptide repeat proteins. Proc Natl Acad SciU  S A, 120(9): e2221529120. https://doi.org/10.1073/pnas.2221529120.

CRISPR-izing the Mitochondrial Genome Is Here

by

CRISPR-Cas9 has become a standard method for editing the nuclear genome, but researchers have long debated whether the mitochondrial genome could be edited too. At last, we have an answer: yes.

When you hear the word “genome,” what comes to mind? The collection of DNA present in the nucleus? While technically correct, our old friend, the “powerhouse of the cell,” would like to object: it too has a genome! In humans, the mitochondrial genome is a circularized strand of DNA that is around 16,500 base pairs long, and within a single cell, it can have hundreds to thousands of copies [1]. At first glance, mitochondrial DNA (mtDNA) doesn’t seem that impressive, especially compared to the 3.4 billion base pairs of the nuclear genome. However, it has significant relevance to human health: mtDNA mutations are associated with numerous diseases from the cardiomyopathy of Barth syndrome to the neurodegeneration of Leigh syndrome [2]. Thus, efforts have been made to “fix” these defects, particularly through CRISPR-Cas9 gene editing.

The CRISPR-Cas9 system consists of an enzyme (Cas9) and a guide RNA (sgRNA) that can locate a specific region of the genome and induce a double-stranded break in the DNA. Using DNA repair mechanisms, this can be exploited to delete or insert a new sequence. For instance, homology-directed repair (HDR) uses a DNA template, either natural or supplied by researchers, to rebuild the broken strand with that template’s sequence. This has been especially powerful for editing nuclear genomes, but it has been debated whether this method could be used for mtDNA as well [3].

This is where Bi et al. enter the scene with their recent study [4]. Their goal was simple: to demonstrate that the mitochondrial genome can, in fact, be “CRISPR-ized.” The first step was to design a CRISPR-Cas9 system capable of entering the mitochondrion. To accomplish this, they took an existing system for nuclear editing and inserted “localization signals” that tell the cell where the protein needs to go. One of these was the mitochondrial targeting signal (MTS) derived from a nuclear-encoded gene, COX8A, that allows the protein to travel to the mitochondrion. Next, an sgRNA was designed that targets the ND4 gene of the mitochondrial genome. Combining these, Bi et al. created a system that could be tested in human cells. Indeed, with the use of fluorescent markers and microscopy, they demonstrated that both the Cas9 protein and sgRNA could enter the mitochondrion (Figure 1).

Next, the researchers needed to determine whether this system could edit mtDNA once inside the organelle. In addressing this, they added the CRISPR-Cas9 system alongside a template strand (called an ssODN) that contained an inserted sequence, GAATTC, that they could screen for (Figure 1). Once the cells were treated, they confirmed the insertion through two methods: PCR and direct sequencing. The PCR method used a primer that could only bind to the inserted sequence, so if the mtDNA was amplified, that indicated a successful “knockin.” The sequencing method, “PCR-free 3rd generation sequencing,” provided additional confirmation without the risk of false-positives from PCR artifacts (Figure 1).

Finally, as a proof of concept, Bi et al. altered the sgRNA and template strand to insert a pathogenic mtDNA mutation (changing ACCTTGC to GCAAGGT on the ND1 gene). This is associated with mitochondrial myopathy (a disease that causes muscular problems), and if successfully inserted, it could provide researchers with a new cell that models this mutant. Indeed, PCR and sequencing confirmed its knockin, demonstrating the power of editing mtDNA.

The authors of this paper directly confirmed that mtDNA could be edited via CRISPR-Cas9 for the first time. This is in contrast to previous studies which have shown mitochondrial localization of the Cas9-sgRNA complex, but with no clear indication of knockin because of PCR-based artifacts or indirect screening [5]. However, the present study is not without its own limitations, particularly in the method’s feasibility at a clinical level. First and foremost, an individual cell has many more copies of mtDNA than nuclear DNA, so editing every strand is more difficult. Indeed, the “knockin efficiency,” or proportion of DNA with the GAATTC insertion, was only 0.03-0.05% (compared to the 20-30% seen in some nuclear genome editing studies) [6]. Furthermore, Bi et al. reported an increase of reactive oxygen species (ROS) formation upon the addition of CRISPR-Cas9, a class of reactive molecules that can degrade DNA and induce cell death. 

Work remains to be done, whether it is through improving the efficacy of the CRISPR-CAS9 system or confirming the method’s safety amidst ROS production. Nonetheless, Bi et al. give us a powerful starting point for treating mitochondrial dysfunction in ways never previously thought possible. In the coming years, CRISPR-izing mtDNA may even become a standard technique, just as we saw with nuclear DNA in the last decade.

References:

  1. Falkenberg M, Larsson NG, Gustafsson CM. (2007). DNA replication and transcription in mammalian mitochondria. Annu Rev Biochem, 76: 679-99. https://doi.org/10.1146/annurev.biochem.76.060305.152028. 
  2. Stenton SL, Prokisch H. (2020). Genetics of mitochondrial diseases: Identifying mutations to help diagnosis. EBioMedicine, 56: 102784. https://doi.org/10.1016/j.ebiom.2020.102784.
  3. Gammage PA, Moraes CT, Minczuk M. (2018). Mitochondrial Genome Engineering: The Revolution May Not Be CRISPR-Ized. Trends Genet, 34(2): 101-110. https://doi.org/10.1016/j.tig.2017.11.001. 
  4. Bi R et al. (2022). Direct evidence of CRISPR-Cas9-mediated mitochondrial genome editing. Innovation (Camb), 3(6): 100329. https://doi.org/10.1016/j.xinn.2022.100329. 
  5. Antón Z et al. (2020). Mitochondrial import, health and mtDNA copy number variability seen when using type II and type V CRISPR effectors. J Cell Sci, 133(18): jcs248468. https://doi.org/10.1242/jcs.248468. 
  6. Zhang JP et al. (2017). Efficient precise knockin with a double cut HDR donor after CRISPR/Cas9-mediated double-stranded DNA cleavage. Genome Biol, 18(1): 35. https://doi.org/10.1186/s13059-017-1164-8.