VIPR reads DNA with a ‘match two, skip one’ code

Our usual picture of genetic recognition is zipper-like: one molecular letter pairs with the next, continuously down the line. VIPR shows that biology can build a precise address from something stranger—a repeating rhythm of match two, skip one.

That matters because CRISPR is only one solution to the problem of finding a particular stretch of DNA. VIPR, a system encoded by viruses that infect bacteria, demonstrates another molecular logic. It is not yet a practical biotechnology; it is evidence that programmable recognition can work even when the matching sequence contains deliberate holes.

Read two, skip one

VIPR combines a protein called Vipr with a guide RNA. The RNA contains repeating five-letter units: three relatively fixed bases, described as GGY, followed by two variable bases, NN. Those two variable bases carry the target information.

The resulting rhythm can be pictured like this:

DNA position:  1  2  3 | 4  5  6 | 7  8  9
VIPR checks:    ✓  ✓  – | ✓  ✓  – | ✓  ✓  –

This is a reading aid, not a guide-design recipe. The central discovery is that every third target-DNA nucleotide can remain unpaired while the surrounding pairs specify the address.

How does a molecule reliably preserve those gaps instead of closing them like an ordinary zipper? The peer-reviewed structural study, published in Science as “VIPR RNA-guided DNA recognition by noncontiguous geometric triplex formation” (DOI: 10.1126/science.aei3472), reports 21 cryo-electron microscopy reconstructions that support an answer. The publication and DOI are identified in the associated Electron Microscopy Data Bank record; an accessible manuscript supplies the detailed structural account.

A protein scaffold keeps the gaps open

The researchers observed multiple Vipr protein copies assembled along the guide RNA as a right-handed filament. Each protein copy grips one RNA repeat. Neighboring copies tuck away the repeat’s GGY bases while presenting its two variable bases for pairing with DNA.

When DNA engages, protein domains displace both the hidden guide bases and each skipped DNA base. The proteins therefore do more than hold the guide: they enforce the punctuation in its molecular address.

Mutations provide a functional check on that interpretation. In E. coli, replacing several amino acids involved in holding the RNA, accommodating a skipped DNA base or gripping the other DNA strand abolished or reduced the reported repression of a fluorescent marker. That result supports the importance of the contacts seen in the reconstructions, although it does not by itself establish how broadly VIPR could be redirected or used outside the tested system.

Not an ordinary three-strand helix

The fully engaged structure contains three nucleic-acid strands, but simply calling it a “triplex” hides what is unusual about it.

The guide RNA pairs intermittently with one DNA strand, producing a gapped RNA–DNA hybrid. The other DNA strand passes through the center of that arrangement. All three follow a shared helical pitch—the distance covered by one complete turn—even though the RNA travels farther around each turn than either DNA strand.

The authors call this a geometric R-loop triplex. Unlike conventional triplexes held together by direct contacts among all three strands, this arrangement is stabilized by the surrounding Vipr proteins. Each protein copy clamps the otherwise unpaired DNA strand while supporting the gapped guide–target hybrid.

Experiments with different substrates sharpened that interpretation. VIPR could form guide–target hybrids with either single-stranded DNA or RNA, but the third strand appeared in the reported reconstructions only when it was DNA. Samples unable to form the three-strand state also contained a larger proportion of incomplete protein filaments. The particle distributions are consistent with the geometric triplex stabilizing assembly, though they are not a movie of the process.

Twenty-one still pictures, not molecular footage

The researchers reconstructed complexes containing different numbers of protein subunits and progressively different DNA arrangements. They interpret these structures as snapshots along a pathway.

In their proposed sequence, Vipr first bends intact double-stranded DNA into the guide RNA’s helical register. The protein holds the guide and one DNA strand relatively steady while bases on the other DNA strand rotate away from their original partners and pair with the guide. The process advances locally along the complex—a “target-strand handoff” that converts the strained intermediate into the final geometric R-loop triplex.

That ordering is an inference from structural comparisons. Cryo-electron microscopy sorts many frozen particles into reconstructed states; it does not record one complex moving from beginning to end. The collection makes the proposed pathway physically coherent, but real-time or kinetic experiments would be needed to establish its timing directly.

The substrate choice matters too. EMD-77183 is a 3.13-ångström reconstruction of a 12-subunit complex made with pre-unwound double-stranded DNA. It supports the final engaged arrangement, but that map alone cannot show VIPR opening an intact duplex.

The intact-DNA pathway is instead supported by intermediate reconstructions. EMD-77735 is identified by its title and overview as a 3.4-ångström, five-subunit complex with duplex DNA. Its displayed sample-description field conflicts with that identification by repeating wording about a 12-subunit, pre-unwound sample, so the record should not carry more interpretive weight than its consistent fields allow.

Did VIPR come before CRISPR?

Vipr proteins share structural elements with components of Class 1 CRISPR systems. Both use protein projections to flip selected bases out of an RNA–DNA pairing, although VIPR applies that architecture to a different recognition rule.

The authors suggest that CRISPR machinery might have emerged through specialization of modules already present in a Vipr-like ancestor. That is an evolutionary hypothesis prompted by structural similarities—not a direct observation of ancient molecular history, nor proof that the studied VIPR system predates every CRISPR system.

The firmer conclusion is already surprising enough: a molecular search does not have to read an unbroken phrase. With a protein scaffold holding the spaces in exactly the right places, even a dotted line can function as an address.