Obelisk (Biology)

Obelisk (biology) is a class of newly discovered, viroid-like, covalently closed circular single-stranded RNA (circRNA) genetic elements—averaging ~1,000 nucleotides in length—that colonize commensal bacteria of the human oral and gut microbiomes as well as global environmental ecosystems. First reported in January 2024 and published in Cell in October 2024 by Ivan N. Zheludev, Robert C. Edgar, Maria Jose Lopez-Galiano, Marcos de la Peña, Artem Babaian, Ami S. Bhatt, and Nobel laureate Andrew Z. Fire at Stanford University, Obelisks were uncovered by screening petabase-scale metatranscriptomic sequencing repositories using a reference-free circular RNA bioinformatic algorithm named VNom (Viroid Nominator).

The genetic elements were named Obelisks because minimum-free-energy secondary structure predictions indicate that their ~1,000-nt circular genomes fold via extensive intramolecular base-pairing into elongated, symmetrical, rod-like double-stranded hairpins spanning the full length of the molecule, whose aspect ratio resembles Ancient Egyptian stone monuments. Phylogenetically and structurally, Obelisks occupy a previously unrecognized evolutionary niche between non-coding plant viroids (246–434 nt) and capsid-encoding RNA viruses: although Obelisks lack a protein capsid and often carry self-cleaving Variant III hammerhead ribozymes like viroids, they dedicate 70% to 90% of their circular RNA genome to open reading frames encoding a novel superfamily of RNA-binding proteins termed Oblins.

Reference-Free Discovery via VNom and Symmetric Hammerhead Ribozymes

Traditional viral metagenomics relies on sequence similarity searches against conserved viral proteins such as RNA-dependent RNA polymerases (RdRps) or capsid subunits. Because viroid-like circular RNAs lack RdRp genes and share zero nucleotide sequence homology with known reference genomes, they remained invisible in standard microbiome assemblies ("biological dark matter"). Zheludev and Fire designed the VNom computational pipeline to scan raw RNA-seq reads from the NCBI Sequence Read Archive (SRA) specifically for permutationally repeating k-mer signatures produced when reverse transcriptase transcribes around a covalently closed circular RNA monomer or rolling-circle concatemer multiple times.

Applying VNom across 5.4 million public sequencing datasets yielded 29,959 Obelisk species (clustered at 90% nucleotide identity) across all seven continents, wastewater, soil, marine metatranscriptomes, and vertebrate microbiomes. Canonical Obelisks share four defining molecular traits:

1. Conserved ~1,000-nt Circular Genome Length: Whereas plant viroids (Pospiviroidae and Avsunviroidae) are constrained between 246 and 434 nt, canonical Obelisks exhibit a tight length distribution centered at ~1,000 nucleotides (ranging from ~950 to 1,160 nt) with balanced ~40%–50% GC content.
2. Global Rod-Like Duplex Conformation: Across both sense (+) and antisense (−) polarities, base-pairing between the two halves of the circular RNA collapses the 1,000-nt ring into a single unbranched (or minimally branched) ~450-base-pair quasi-double-stranded RNA rod interrupted by short symmetric internal loops, shielding the RNA from bacterial single-strand endonucleases.
3. Symmetric Variant III Hammerhead Ribozymes: A prominent sub-clade of Obelisks encodes self-cleaving Variant III hammerhead ribozyme (HHR) motifs on both the (+) and (−) polarity strands. As in family Avsunviroidae viroids and hepatitis delta virus (HDV) ribozymes, dual-polarity self-cleavage indicates that Obelisks replicate via a symmetric rolling-circle mechanism catalyzed by host RNA polymerases, in which multimeric linear (+) and (−) concatemers self-cleave at specific phosphodiester bonds into unit-length 1,000-nt monomers that are subsequently circularized.

Oblin-1 and Oblin-2: A Novel Protein Superfamily

Unlike non-coding plant viroids, nearly all Obelisks contain one or two conserved open reading frames (ORFs) occupying the majority of their coding strand, translated using standard bacterial Shine–Dalgarno ribosome-binding sites:

• Oblin-1 (~180 to 220 amino acids): Encoded by every canonical Obelisk, Oblin-1 represents a previously unknown protein superfamily that shares zero detectable primary sequence homology (via BLASTp or HMMER profiles) with any protein in GenBank, UniProt, or viral databases. Tertiary structure predictions (AlphaFold2 and ESMFold) indicate that Oblin-1 folds into a globular multi-helical and β-sheet domain featuring a conserved patch of positively charged arginine and lysine residues along one surface (Domain A) and a conserved 18-amino-acid signature motif (GYxDxD...) predicted to bind divalent cations (Mg²⁺) or polyanionic RNA duplexes to assist rolling-circle replication or ribonucleoprotein stabilization.
• Oblin-2 (~50 to 60 amino acids): Present as a second ORF in roughly 40% to 50% of human-associated Obelisks, Oblin-2 is predicted by AlphaFold2 to form an amphipathic α-helical leucine-zipper coiled-coil homodimer capable of nucleic acid binding or host transcription factor modulation.

Colonization of Streptococcus sanguinis and Persistence in Humans

Screening of human longitudinal metatranscriptomic cohorts revealed that Obelisks are widespread inhabitants of the human body, detected in ~7% of human gut (stool) metatranscriptomes (440 out of 5,452 individuals across five continents) and ~50% of human oral metatranscriptomes (17 out of 32 supragingival plaque and saliva datasets). Individual human subjects harbor anatomical site-specific Obelisk strains—with separate oral and gut populations—that persist stably inside the same host for over 300 days.

To identify the cellular host of an Obelisk experimentally, Zheludev and colleagues screened monocultures of human oral bacteria and discovered a specific 1,137-nt Obelisk (Obelisk-S.s, encoding both Oblin-1 and Oblin-2) replicating stably inside the model human dental-plaque commensal bacterium Streptococcus sanguinis strain SK36. In genomic sequencing controls, S. sanguinis SK36 contains no DNA sequence matching the Obelisk in its bacterial chromosome or plasmids, proving that Obelisk-S.s is an autonomous extrachromosomal RNA replicon rather than a spliced host transcript. Subsequent laboratory cures demonstrated that S. sanguinis strains cleared of Obelisk-S.s grow normally in rich broth, indicating that Obelisks behave as intracellular RNA episomes ("RNA plasmids") or subviral commensals. Together with ribozymes and viroids, Obelisks provide a living window into the RNA World and Ribonucleoprotein World, demonstrating how minimal circular RNA replicons co-evolved with small RNA-binding peptides (like Oblins) before acquiring viral capsids.

Why did scientists not discover Obelisks in the human microbiome until 2024?

Standard metagenomic pipelines identify viruses and microbes by matching DNA or RNA sequencing reads against reference databases of known genes—such as the universal RNA-dependent RNA polymerase (RdRp) gene found in all conventional RNA viruses. Obelisks have tiny (~1,000-nucleotide) circular RNA genomes that lack an RdRp gene and encode Oblin proteins that share zero sequence similarity with any previously known organism or virus. They were only discovered when Stanford researchers wrote a custom bioinformatic tool (VNom) that searches raw sequencing data for circular, self-complementary RNA rings regardless of sequence homology.

How do Obelisks differ from plant viroids and standard RNA viruses?

Plant viroids (such as potato spindle tuber viroid) are tiny (246–434 nt) naked circular RNAs that infect crops and do not code for any proteins at all. Standard RNA viruses (such as poliovirus or influenza) have larger genomes (>3,000 nt) that encode a protective protein capsid shell and their own RNA polymerase. Obelisks sit between the two: like viroids, they are naked, rod-shaped circular RNAs (~1,000 nt) that lack a protein capsid and often carry self-cleaving hammerhead ribozymes; unlike viroids, Obelisks infect bacteria and use most of their genome to encode one or two novel RNA-binding proteins (Oblin-1 and Oblin-2).

Do Obelisks cause disease in humans?

No link to human disease has been established. Although Obelisks are present in roughly 50% of human oral microbiomes and 7% of human gut microbiomes, they do not infect human eukaryotic cells directly; instead, they live inside commensal bacteria such as Streptococcus sanguinis (a normal inhabitant of healthy dental plaque). Laboratory experiments show that S. sanguinis cells carrying Obelisks divide and grow at the same rate as bacteria cured of their Obelisks, suggesting that Obelisks function like harmless intracellular RNA plasmids or bacterial symbionts.

1. Zheludev, I. N., Edgar, R. C., Lopez-Galiano, M. J., de la Peña, M., Babaian, A., Bhatt, A. S., & Fire, A. Z. (2024). "Viroid-like colonists of human microbiomes." Cell, 187(23), 6521–6536.e18. https://doi.org/10.1016/j.cell.2024.09.033
2. Lee, B. D., Neri, U., Roux, S., Wolf, Y. I., Camargo, A. P., Krupovic, M., & Koonin, E. V. (2023). "Mining metatranscriptomes reveals a world of viroid-like circular RNAs." Cell, 186(3), 646–661.e4. https://doi.org/10.1016/j.cell.2022.12.039
3. Diener, T. O. (1989). "Circular RNAs: Relics of precellular evolution?" Proceedings of the National Academy of Sciences, 86(23), 9370–9374. https://doi.org/10.1073/pnas.86.23.9370
4. Flores, R., Gago-Zachert, S., Serra, P., Sanjuán, R., & Elena, S. F. (2014). "Viroids: Survivors from the RNA world?" Annual Review of Microbiology, 68, 395–414. https://doi.org/10.1146/annurev-micro-091313-103416