A look at a Cas3 HD domain

In the news you might only hear about CRISPR-Cas9, but there are actually several types of CRISPR systems – and the most common are Type I systems, defined by the cas3 gene. In 2011, Sabin Mulepati, then a PhD student, and Scott Bailey were studying the Cas3 protein in Thermus thermophilus. At the time, the current research indicated that the Cascade complex was responsible for recognizing target DNA, while other work had identified Cas3 as a DNA nuclease, proposed to cleave the targets identified by Cascade.

Cas3 has two domains – an N-terminal histidine-aspartate (HD) domain with nuclease activity and a C-terminal helicase domain. In some genomes these are separately encoded subunits called Cas3” (the HD domain) and Cas3’ (the helicase domain).

At the time, reports about two other Cas3 proteins had reported different nucleic acid substrates and metal ion requirements. The S. thermophilus Cas3, which, as in T. thermophilus, is a single protein, was reported to cleave ssDNA and showed the most activity in the presence of Mn²⁺, and less activity with Mg²⁺ and Ca²⁺, but had no activity with Co²⁺, Ni²⁺, Zn²⁺. In contrast, the Sulfolobus solfataricus Cas3” subunit was reported to specifically cleave double-stranded oligonucleotides–DNA or RNA–and only in the presence of Mg²⁺, but not Mn²⁺, Ca²⁺, Ni²⁺, Zn²⁺.

Sabin and Scott wanted to know more. In a paper published in the Journal of Biological Chemistry in 2011, they reported on how they created Cas3^HDdom, a truncated version of the T. thermophilus protein that contained the HD domain region to answer several questions, including: What is the structure of the nuclease activity domain? What types of nucleic acids does it target? What types of metal ions does it need to function, and how does it bind them?

To answer these questions and better understand the molecular basis of this CRISPR system, Sabin used x-ray crystallography and biochemical analysis to characterize Cas3^HDdom and several versions of the protein with single point mutations in three groups of key residues – residues on the surface near the active site, predicted metal ion-binding residues in the active site, active site residues not predicted to be involved in metal ion binding.

What is the structure of the nuclease domain?

After purifying and crystalizing the Cas3^HDdom protein, Sabin generated a structure using x-ray crystallography. This structure, refined to a resolution of 1.8 Å, was the first structural view of an HD domain that has nuclease activity. Shortly after this paper was published, the Yakunin lab published a study characterizing the Cas3 HD domain of Methanocaldococcus jannaschii, including a crystal structure of the domain and its active site.

Sabin’s structure showed that Cas3^HDdom is globular, with the five conserved motifs forming with a concave surface. By comparing this structure with other HD domain super-family structures, they were able to identify the core fold of the HD domain, including a common core of five α-helices and their loops that house the five conserved motifs, highlighted in the ribbon structure below.

With the structure solved, Sabin and Scott moved on to biochemistry to answer the other questions.

What types of nucleic acids does Cas3HDdom target? What types of metal ions does Cas3HDdom need to function?

The activity assays tested different combinations of the protein, DNA substrates and other reaction components to see which resulted in DNA cleavage by running the products on an agarose gel and looking for either a single, clear band (uncut DNA) or smears (cleaved DNA).

The activity assays showed that, like the S. thermophilus Cas3, Cas3^HDdom only cleaved single-stranded, not double-stranded, DNA.

By testing different metal ions in the activity assays, Sabin showed that Cas3HDdom is active in the presence of Mn²⁺, Co²⁺, Ni²⁺, Cu²⁺ or Zn²⁺, but not Mg²⁺ or Ca²⁺.

How does the protein bind the metal ions?

To further examine the protein’s metal binding, Sabin and Scott examined the normal protein and the versions with point mutations using both the activity assay and a protein stability assay.

The protein stability assay monitors protein folding to give an apparent melting temperature (T_m) that was used as an indication of thermal stability. This works as a readout of metal binding because, in general, binding to ligands increases the stability of proteins. In fact, in the presence of the Ni²⁺ metal ions (which activates Cas3^HDdom), the apparent T_m increased by about 15 °C, while Mg²⁺ (which does not activate Cas3^HDdom), only increased it by about 6 °C, even at 200-fold higher concentrations.

By looking at the mutant proteins using the activity and metal ion binding assays, Sabin and Scott confirmed that the predicted residues bound the metal ions. Only the mutations in the predicted metal ion-binding residues prevented the decreased stability when Ni²⁺ was added (and the only metal ion binding mutation that showed a normal increase had a notably lower stability in the absence of metals), and all of those lost the ssDNA cleavage activity.

Combined with the structural data, this suggests that the metal ion-binding site binds two metal ions, and that binding is required for nuclease activity. It’s not known which are used by the bacterial protein in vivo, but based on the intracellular concentrations it’s likely Mn²⁺ or Ni²⁺.

Taken together, the experiments answered Sabin and Scott’s main questions – the Cas3 HD domain has a concave, globular structure, which binds two metal ions. These metal ions are required for the Cas3 nuclease activity, which is specific to single-stranded DNA.

The structural data and activity data also answered other questions about the domain and its active site, and pointed to differences in the HD domain activity and active site geometries between the Cas3 proteins with both a HD and helicase domain and the Cas3” subunits – to learn more about those conclusions, and the additional questions they raised, check out the paper.

The work was funded by the NIH’s National Institutes of General Medical Sciences and start-up funds from the Department of Biochemistry and Molecular Biology and the Johns Hopkins Malaria Research Institute.

Want to learn more about the work this paper builds on? Check out these papers:

Cascade’s role in targeting DNA

Jore MM, Lundgren M, van Duijn E, et al. Structural basis for CRISPR RNA-guided DNA recognition by Cascade. Nat Struct Mol Biol. 2011;18(5):529–536. doi:10.1038/nsmb.2019

Cas3 structure and activity

Sinkunas T, Gasiunas G, Fremaux C, Barrangou R, Horvath P, Siksnys V. Cas3 is a single-stranded DNA nuclease and ATP-dependent helicase in the CRISPR/Cas immune system. EMBO J. 2011;30(7):1335–1342. doi:10.1038/emboj.2011.41

Haft DH, Selengut J, Mongodin EF, Nelson KE. A guild of 45 CRISPR-associated (Cas) protein families and multiple CRISPR/Cas subtypes exist in prokaryotic genomes. PLoS Comput Biol. 2005;1(6):e60. doi:10.1371/journal.pcbi.0010060

Makarova KS, Haft DH, Barrangou R, et al. Evolution and classification of the CRISPR-Cas systems. Nat Rev Microbiol. 2011;9(6):467–477. doi:10.1038/nrmicro2577

Han D, Krauss G. Characterization of the endonuclease SSO2001 from Sulfolobus solfataricus P2. FEBS Lett. 2009;583(4):771–776. doi:10.1016/j.febslet.2009.01.024