Nucleic-acid-triggered NADase activation of a short prokaryotic Argonaute

Gao, Xiaopan; Shang, Kun; Zhu, Kaixiang; Wang, Linyue; Mu, Zhixia; Fu, Xingke; Yu, Xia; Qin, Bo; Zhu, Hongtao; Ding, Wei; Cui, Sheng

doi:10.1038/s41586-023-06665-6

Article
Published: 02 October 2023

Nucleic-acid-triggered NADase activation of a short prokaryotic Argonaute

Xiaopan Gao ORCID: orcid.org/0000-0002-1435-6636^1,2^na1,
Kun Shang^3,4^na1,
Kaixiang Zhu^1,2^na1,
Linyue Wang^1,2^na1,
Zhixia Mu^1,2^na1,
Xingke Fu³,
Xia Yu⁵,
Bo Qin^1,2,
Hongtao Zhu³,
Wei Ding ORCID: orcid.org/0000-0003-4368-8215³ &
…
Sheng Cui ORCID: orcid.org/0000-0001-6329-3582^1,2

Nature volume 625, pages 822–831 (2024)Cite this article

8173 Accesses
3 Citations
52 Altmetric
Metrics details

Subjects

Abstract

Argonaute (Ago) proteins mediate RNA- or DNA-guided inhibition of nucleic acids^1,2. Although the mechanisms used by eukaryotic Ago proteins and long prokaryotic Ago proteins (pAgos) are known, that used by short pAgos remains elusive. Here we determined the cryo-electron microscopy structures of a short pAgo and the associated TIR-APAZ proteins (SPARTA) from Crenotalea thermophila (Crt): a free-state Crt-SPARTA; a guide RNA–target DNA-loaded Crt-SPARTA; two Crt-SPARTA dimers with distinct TIR organization; and a Crt-SPARTA tetramer. These structures reveal that Crt-SPARTA is composed of a bilobal-fold Ago lobe that connects with a TIR lobe. Whereas the Crt-Ago contains a MID and a PIWI domain, Crt-TIR-APAZ has a TIR domain, an N-like domain, a linker domain and a trigger domain. The bound RNA–DNA duplex adopts a B-form conformation that is recognized by base-specific contacts. Nucleic acid binding causes conformational changes because the trigger domain acts as a ‘roadblock’ that prevents the guide RNA 5′ ends and the target DNA 3′ ends from reaching their canonical pockets; this disorders the MID domain and promotes Crt-SPARTA dimerization. Two RNA–DNA-loaded Crt-SPARTA dimers form a tetramer through their TIR domains. Four Crt-TIR domains assemble into two parallel head-to-tail-organized TIR dimers, indicating an NADase-active conformation, which is supported by our mutagenesis study. Our results reveal the structural basis of short-pAgo-mediated defence against invading nucleic acids, and provide insights for optimizing the detection of SPARTA-based programmable DNA sequences.

You have full access to this article via your institution.

Download PDF

Oligomerization-mediated activation of a short prokaryotic Argonaute

Article 26 July 2023

Structural basis for sequence-specific recognition of guide and target strands by the Archaeoglobus fulgidus Argonaute protein

Article Open access 14 April 2023

Programmable RNA targeting by bacterial Argonaute nucleases with unconventional guide binding and cleavage specificity

Article Open access 08 August 2022

Main

Ago proteins are molecular systems present in archaea, bacteria and eukaryotes. Ago systems have key roles in RNA- and DNA-guided inhibition of target nucleic acids^1,2. Eukaryotic Ago proteins (eAgos) use short RNAs as guides to target complementary RNA sequences, thereby mediating RNA interference. Unlike eAgos, which act exclusively on RNAs, many pAgos mediate DNA-guided target DNA interference^1,3,4,5, whereas other pAgos use guide RNAs to target RNAs and/or DNAs^6,7,8,9,10. pAgos therefore exhibit more diverse functions than eAgos, including targeting bacteriophages and plasmids through DNA interference^1,3, unlinking replicating chromosomes^11,12 and enhancing homologous recombination^13,14.

A search of prokaryotic genomes in the latest version of the NCBI Reference Sequence database revealed that pAgos are widespread in archaea (around 25% of all sequenced genomes) and bacteria (around 17% of genomes)¹⁵, and previous studies^16,17 have come up with estimates of 32% and 9%, respectively. pAgos are more diverse than eAgos and can be divided into long-A, long-B and short pAgo clades¹⁵. Structures of free-state long pAgos and their guide–target-strand-bound forms have been reported^6,7,18,19,20. Long pAgos possess a bilobal scaffold comprising an N (N-terminal) domain, a PAZ (PIWI–Argonaute–Zwille) domain, a MID (middle) domain and a PIWI (P-element-induced wimpy testis) domain. The N-PAZ and MID-PIWI of long pAgos form a pocket that accommodates the guide and target nucleic acids^{19,21,22,23,24}. The MID and PAZ domains bind to the 5′ and 3′ ends of the guide strand, and the 5′ end of the template strand is recognized by the PIWI domain. Most long-pAgo PIWI domains also contain an RNase H module and the catalytic tetrad responsible for cleavage^{20,21,25,26,27}, although some long-B pAgos have mutations in their catalytic tetrad. Notably, these pAgo genes cluster with genes that encode putative nucleases or DNA-binding proteins¹⁵. The N domain of pAgos is highly variable. The N domain unwinds the target–guide duplex and blocks duplex formation during target recognition^26,28.

Short pAgos contain only a MID domain and a catalytically inactive PIWI domain; the N and PAZ domains are missing^15,16,17. Nevertheless, short-pAgo genes are often associated with genes that encode proteins analogous to the PAZ domain in the same operon, termed APAZ¹⁶, which is more similar to N domains than to PAZ domains^18,29. Furthermore, many APAZ-containing proteins are fused to an additional domain, such as the SIR2 (silent information regulator 2), TIR (Toll–interleukin-1 receptor) or DUF4365 domains^15,16. The SIR2 and TIR domains possess NAD(P)ase activity, and the DUF4365 domain is similar to the RecB-like domain of Mrr-subfamily PD-(D/E)XK nucleases; these features compensate for the lack of catalytic activity of the short pAgo proteins themselves^15,16.

Several studies have revealed that the short pAgo TIR-APAZ-(SPARTA) and the short SIR2-associated pAgo systems recognize invading nucleic acids through guide-mediated target binding, and ultimately activate the TIR or SIR2 domains. Activation of SIR2 or TIR unleashes their NAD(P)ase activity and causes cell death through NAD(P)⁺ depletion, thereby preventing the propagation of the invader’s genetic elements^30,31. These studies suggest that short pAgos act as prokaryotic immune systems that induce cell death after the RNA-guided recognition of invader nucleic acids; however, the molecular mechanism that underlies the activity of short-pAgo systems remains unknown.

In this study, we determine five cryo-electron microscopy (cryo-EM) structures of the Crt-SPARTA system, and identify molecular determinants of RNA-guided binding to target DNA, thus coupling Crt-SPARTA supramolecular assembly and TIR activation.

The free-state Crt-SPARTA structure

We assembled free-state Crt-SPARTA by co-expressing its two components in Escherichia coli. We determined the cryo-EM structure of the recombinant Crt-SPARTA heterodimer to a resolution of 3.27 Å using single-particle analysis (Extended Data Fig. 1). Free-state Crt-SPARTA is a hollow heterodimer comprising a TIR-APAZ protein and a short pAgo (Fig. 1a,b). On the basis of the features identified here, we name them Crt-TIR-pseudoPAZ (Crt-TIR-pPAZ) and Crt-Ago. Crt-TIR-pPAZ comprises a TIR domain that connects to a linker domain containing L0, L1, L2 and an N-like domain, as in typical Agos, but lacking the PAZ domain; hence we use ‘pseudo’ to mark this feature.

The Crt-SPARTA heterodimer contains a short Ago and a TIR-pPAZ protein, intertwined through numerous protein–protein interactions (Fig. 1b,c). We searched the Crt-SPARTA structure against entries in the Protein Data Bank (PDB) using the Dali server and identified Thermus thermophilus Ago (TtAgo, PDB: 4N76) as the top hit, with a z-score of 29.2 and a root-mean-square deviation (RMSD) of 3.0 Å, despite it sharing only 17% sequence identity with Crt-SPARTA. The top ten Dali hits are listed in Supplementary Table 1.

The Crt-SPARTA structure exhibits a bilobal architecture similar to long pAgos and eAgos; however, it lacks the PAZ domain but contains a TIR domain (Extended Data Fig. 2a,b). Whereas the TIR domain and the pPAZ domain are located in Crt-TIR-pPAZ, the MID domain and the PIWI domain are in Crt-Ago. The TIR domain is connected to the N-like domain through the linker domain. The extreme C-terminal region of Crt-TIR-pPAZ folds into a helical domain, a previously unobserved feature in Ago architectures; we call this the ‘trigger’ domain. Crt-Ago contains MID and PIWI domains with folds similar to the MID and PIWI of long pAgos and eAgos (Extended Data Fig. 2a,b). Although Crt-Ago and Crt-TIR-pPAZ are separate polypeptide chains, the N terminus of Crt-Ago anchors in Crt-TIR-pPAZ through a hydrogen bond with Crt-TIR-pPAZ β19. The deep cavity formed between Crt-TIR-pPAZ and Crt-Ago is the nucleic-acid-binding site (Fig. 1d). All of the Crt-SPARTA domains are involved in the formation of the cavity except TIR. A molecular surface electrostatic potential plot shows that the interior of this cavity is positively charged, which suggests that it is the nucleic-acid-binding pocket—akin to those of long pAgos and eAgos.

Crt-SPARTA is composed of an Ago-like lobe that connects with a TIR domain, although the conserved Ago domains are spread across the two proteins held together through extensive interactions. The Ago-like lobe of Crt-SPARTA might negatively regulate the activity of the TIR domain to avoid autotoxicity.

The Crt-TIR-pPAZ structure

The Crt-TIR domain (residues 1–164) shares a typical fold with other TIR domains that are widely distributed across eukaryotes and prokaryotes^{32,33,34,35,36,37,38}. The core region of Crt-TIR contains a parallel five-stranded β-sheet (βA–βE) surrounded by five α-helices (αA–αE; Extended Data Fig. 3a,b and Supplementary Fig. 2). We identified a NADase catalytic cavity in Crt-TIR composed of αB, αC, βB and the loop connecting βB and αB, which contains conserved catalytic residues (Fig. 1e). An extra helix (αF) at the C terminus of the Crt-TIR core connects the downstream pPAZ domain. Helices αF, αA and αE of Crt-TIR form an interfacial area (around 819 Å²) that binds to the MID domain of Crt-Ago (Fig. 1f). The TIR–MID interaction is centred by hydrophobic interactions between residues W20, W124 of the TIR and a hydrophobic patch (P63, Y65, M74 and P76) of the MID. Five more intermolecular polar contacts surrounding the hydrophobic centre strengthen the TIR–MID interaction (Fig. 1f).

The Crt-TIR is connected to the N-like domain through a long loop (residues 164–183) that extends across the bottom of Crt-Ago, so that the TIR domain and the N-like domain clamp on both sides of Crt-Ago (Fig. 1b,c). The Crt-N-like domain (spanning only 59 amino acids, residues 184–242) is a miniaturized N domain compared with that of long Ago (for example, TtAgo) and eAgo (for example, human AGO2; Extended Data Fig. 2c). Nevertheless, the Crt-N-like domain still maintains a recognizable N-fold: four-stranded antiparallel β-sheets interact with five short α-helices on one side, and the other side of the β-sheets interact with the linker domain (Extended Data Fig. 3c,d and Supplementary Fig. 3).

The Crt-linker domain is composed of two discrete regions: the loop between the TIR domain and the N-like domain (denoted L0; β6–β7, residues 165–183); and the regions between the N-like domain and the trigger domain (denoted L1 and L2; α12–α16 and β12–β19, residues 243–415), which fuse into an intact domain (Extended Data Fig. 3c,d and Supplementary Fig. 3). L1 (residues 243–351) and L2 (residues 352–385) are topologically equivalent to the L1 and L2 regions found in long-pAgo and eAgo proteins (Extended Data Fig. 2d), but the PAZ domain between them is completely omitted in Crt-SPARTA. The L1 domain contains five-stranded antiparallel β-sheets surrounded by three α-helices (Extended Data Fig. 3c,d and Supplementary Fig. 3). Whereas the N-terminal long helix (α13) extends along the β-sheets on one side, two C-terminal helices (α15–α16) in the L2 domain orient perpendicular to the β-sheets on the opposite side. The C-terminal loop of the linker domain (β18–β19, residues 387–415) forms antiparallel β-sheets with the L0 loop at the bottom of the linker domain (Extended Data Fig. 3c,d and Supplementary Fig. 3).

The extreme C-terminal region of Crt-TIR-pPAZ (α17–α18, residues 416–450) is acidic region, with almost half of the residues (15 out of 34) being Asp or Glu (Extended Data Fig. 3c,d and Supplementary Fig. 3). The trigger domain folds into two α-helices (α17–α18) and adopts an L shape. This unique module joins the linker domain and Crt-Ago together, and mediates numerous polar and hydrophobic contacts with the adjacent domains. On one side, the trigger domain forms multiple polar contacts with helix α15 of the linker domain (Fig. 1g), so that three helices (α15, α17 and α18) constitute a triangular ‘lasso’. The lasso unit has a key role in recognizing target DNA, as described in detail below. On the other side, the trigger domain interacts with both the MID and the PIWI domain of Crt-Ago. As shown in Fig. 1g, whereas the upper part of α17 and the α17–α18 loop interact with the MID domain, the lower part of α17 interacts with the ‘PIWI box’ (defined in the section below). Superimposing the Crt-SPARTA structure with those of the long-pAgo TtAgo and the eAgo human AGO2 indicates that the trigger domain occupies the exit of the nucleic-acid-binding pocket, where the 3′ terminus of the target strand binds (Extended Data Fig. 4). It is possible that the trigger domain is involved in nucleic acid recognition.

The Crt-Ago structure

Crt-Ago comprises an MID domain (residues 1–269) and a PIWI domain (residues 270–507; Fig. 1a–c, Extended Data Fig. 3e,f and Supplementary Fig. 4). The Crt-MID domain contains a four-stranded parallel β-sheet (β1–β4) surrounded by nine α-helices (α1–α6 and α9–α11) packed against the β-sheet on both sides (Extended Data Fig. 3e,f and Supplementary Fig. 4). Unlike most known Ago structures, Crt-MID contains a unique insertion between α6 and α9 (residues 150–205; Extended Data Fig. 2e). This insertion folds into two orthogonally oriented helices (α7–α8), forming a protrusion on top of the MID domain (Extended Data Fig. 3e,f and Supplementary Fig. 4), we denote it ‘MID-protrusion’. We identified four conserved residues that might participate in the recognition of the 5′-phosphate (5′-P) of the guide strand, namely H207, K211, Q222 and K263, which lie within a putative phosphate-binding pocket formed by α9, α11 and β4 of the MID domain (Extended Data Fig. 3e,f and Supplementary Fig. 4). The C terminus of the PIWI domain also resides in proximity to the 5′-P-binding pocket, suggesting that it is involved in 5′-P recognition.

The Crt-PIWI domain (residues 270–507) comprises an RNase H module and a PIWI box (residues 390–449), similar to long pAgos and eAgos (Extended Data Fig. 2f). Although the RNase H module exhibits a typical RNase fold, comprising a four-stranded mixed β-sheet (β5–β9) sandwiched by five α-helices (α12–α16) on both sides, key catalytic residues are missing (Extended Data Fig. 3e,f and Supplementary Fig. 4). The conserved tetrad of negatively charged amino acid residues in the DEDX-containing motif is replaced by V284, Y328, R362 and N484, respectively (Extended Data Fig. 3e,f and Supplementary Fig. 4), suggesting that it lacks nuclease activity. The PIWI box contains a three-stranded antiparallel β-sheet (β10–β12) that is fused with the RNase H β-sheet, forming an intact β-plane. The β9–β10 loop and the β11–β12 loop of the PIWI box anchor deeply into Crt-TIR-pPAZ and interact with both the linker and the trigger domain (Extended Data Fig. 3e,f and Supplementary Fig. 4).

The Crt-Ago structure contains a MID domain that is required for 5′-P guide strand binding and a catalytically inactive PIWI domain, suggesting that it only participates in 5′-P guide-mediated target strand binding and not in nucleic acid cleavage.

The Crt-SPARTA–RNA–DNA structure

To gain structural insights into Crt-SPARTA guide-mediated target recognition, we assembled a quaternary complex containing a Crt-SPARTA heterodimer, a 5′-P 21-nucleotide (nt) guide RNA and a complementary 45-nt target DNA (Fig. 2a and Extended Data Fig. 5 and Supplementary Fig. 1). The guide and target strands form a partial duplex with a 3′ overhang (1-nt dT0′) and a 5′ overhang (23-nt dT22′-dA44′). Guide RNA was preincubated with Crt-SPARTA heterodimer before the addition of target DNA. The resulting nucleoprotein complex was purified by size-exclusion chromatography and yielded two oligomeric species: a low-order species corresponding to the size of one quaternary complex; and a higher-order species corresponding to dimeric or tetrameric Crt-SPARTA–RNA–DNA complexes (Extended Data Fig. 5b–d and Supplementary Fig. 1). We separated the two species for cryo-EM investigations.

**Fig. 2: Nucleic-acid-bound *Crt*-SPARTA.**

We determined the cryo-EM structure of the single quaternary complex to a resolution of 3.27 Å (Extended Data Fig. 1e–h). Guide RNA and target DNA form a hybrid duplex that occupies the nucleic-acid-binding cavity between Crt-Ago and Crt-TIR-pPAZ (Fig. 2b,c). The entire guide RNA (U1–U21) and the 3′ 24 nt of the target DNA (dT0′–dT23′) were identified in the EM density, but the 5′ portion of the target DNA (dC24′-dA44′) pointing away from the hybrid duplex was untraceable. The hybrid duplex adopts a double-helical conformation more similar to an ideal B-form duplex than to an ideal A-form duplex (Extended Data Fig. 6a). By contrast, many Ago-bound guide–target duplexes adopt A-form conformations, except the MpAgo-bound RNA–DNA hybrid duplex^{6,24,25,26,39} (in which Mp refers to Marinitoga piezophila). RNA–DNA duplexes naturally adopt an A-form conformation similar to double-stranded RNA^20,40. The A-form conformation is thought to benefit from reduced entropic costs associated with hybrid duplex formation²⁴. The rare B-form hybrid duplex observed in Crt-SPARTA hints at unusual Crt-SPARTA–nucleic acid interactions.

Indeed, guide and target strands are recognized by the Crt-SPARTA heterodimer in a mode that has not to our knowledge been reported previously (Fig. 2b and Extended Data Fig. 6d–g). Whereas the linker, N-like, trigger and PIWI domains are involved in binding the hybrid duplex, the MID and TIR domains have no contacts with the nucleic acids. At the hybrid duplex end pointing away from Crt-SPARTA, we observed clear EM density for seven more base pairs (G15–dC15′ to U21–dA21′) and a short tail (dT22′–dT23′), which protrude from the Crt-SPARTA heterodimer (Fig. 2b). At the hybrid duplex end pointing into Crt-SPARTA, the termini of guide and target strands cannot reach their canonical binding pockets in the MID and PIWI domains as seen in long-pAgo and eAgo systems, because the trigger domain blocks their access (Fig. 2c and Extended Data Fig. 6d).

Superimposing the TtAgo–DNA-DNA structure (PDB: 4N47) onto the Crt-SPARTA–RNA–DNA structure revealed that the 5′-P of the TtAgo guide strand reached the conserved 5′-P pocket in the MID domain (Fig. 2c, bottom), whereas the 5′-P of Crt-SPARTA guide RNA is 17.5 Å away from the 5′-P-binding pocket, approximately five base steps behind the TtAgo guide DNA (Extended Data Fig. 6b). The 3′ end of the TtAgo target strand reaches its PIWI pocket, whereas the 3′ end of Crt-SPARTA target strand is also five base steps behind (Fig. 2c, bottom and Extended Data Fig. 6c). This is because the trigger domain acts as a roadblock residing between the 3′ end of the target strand and Crt-Ago.

This unusual nucleic-acid-binding mode can largely be attributed to the unique lasso unit (α15, α17 and α18) in Crt-SPARTA. As its name indicates, the Lasso unit loops around the 3′ terminus of the target DNA and provides extensive contacts to nucleic acids (Fig. 2c,d). Helix α17 is positioned in front of the target strand 3′ end and provides three contacts (E438–T0′, N434–T0′ and E438–A1′) to the first and second bases. Helix α15 is inserted into the minor groove of the hybrid duplex and recognizes the G2–C2′ base pair through residues H358 and R362. Helix α18 is placed on top of the target strand and provides further contacts to the backbone phosphates of the duplex (Fig. 2d and Extended Data Fig. 6d).

Most nucleic acid contacts provided by the N-like, linker and PIWI domains are phosphate backbone interactions (Fig. 2d and Extended Data Fig. 6e–g), except the β12–β13 loop of Crt-TIR-pPAZ. The β12–β13 loop extends into a major groove of the hybrid duplex, where K289 recognizes the A7 and G8 bases of the guide RNA, and S288 recognizes the 2′-hydroxyl group of the G5 ribose (Fig. 2d and Extended Data Fig. 6e).

Collectively, the interactions between Crt-SPARTA and the hybrid duplex involve multiple base contacts in both guide and target strands, suggesting that the nucleic acid recognition is sequence-specific (Fig. 2d). Recognition of the ribose 2′-hydroxyl group on the guide strand underlies the preference of Crt-SPARTA for an RNA guide. Our structural findings support previous studies showing that SPARTA activation is sequence-specific, and that SPARTA is preferentially activated by 5′-P guide-RNA-mediated target DNA binding³¹. Crt-SPARTA recognizes more molecular features on the target DNA than the guide RNA. We found residues E438, N434 and H358 recognizing DNA bases and R362 and K289 recognizing RNA bases (Fig. 2d), consistent with previous results showing that guide RNA–target DNA combinations elicit the most robust NADase activity³¹. However, we did not observe interactions between the guide RNA 5′-P and the Crt-MID domain, even though the conserved 5′-P-binding pocket is available. It is possible that some intermediate steps during guide RNA binding were not captured by our cryo-EM investigations.

To validate our cryo-EM structures, we performed mutagenesis studies. First, we characterized the NADase activity of Crt-SPARTA. Whereas free-state Crt-SPARTA had undetectable NADase activity, the guide RNA–target DNA-loaded Crt-SPARTA monomer had minimal activity and the Crt-SPARTA tetramer exhibited the highest activity (Fig. 2e). We then measured the Michaelis–Menten parameters of Crt-SPARTA NADase activity (Fig. 2f), which gave a K_m of 651 μM and a catalytic rate constant k_cat of 1.34 s⁻¹. These kinetic parameters are similar to those of TIR-SAVED (SMODS-associated and fused to various effector domains) NADase³².

We then mutated a selection of Crt-SPARTA residues that are important for nucleic acid binding and measured the NADase activity of these mutants (Fig. 2g). Although all mutations ablated the NADase activity to various extents, mutations at the 5′-P pocket in the MID domain resulted in the greatest loss of NADase activity, suggesting that the 5′-P pocket has a key role in the nucleic- acid-activated NADase activity.

Next, we investigated the guide RNA binding capability of the Crt-SPARTA mutants containing mutations at the 5′-P pocket (Fig. 2h and Supplementary Fig. 1). Both single and combinational mutations had a limited effect on RNA binding. One possible explanation is that, because guide strand binding involves multiple Crt-SPARTA domains (MID, PIWI, N-like and so on), merely disrupting RNA binding at the 5′-P pocket probably caused a limited decrease in affinity, which is consistent with a previous study⁴¹. These results also imply that, despite the pivotal role of the 5′-P pocket in activating the NADase activity of Crt-SPARTA, its contribution to guide RNA binding affinity is of lesser importance.

We also investigated the target DNA binding of Crt-SPARTA mutants that lack the trigger domain. Partial deletion of the trigger domain (TIR-pPAZ 1–431) reduced the DNA binding affinity, whereas complete deletion of the trigger region (TIR-pPAZ 1–416) altered the migration of the Crt-SPARTA–RNA–DNA complex in native PAGE, indicative of the formation of higher-order nucleoprotein complexes (Fig. 2i and Supplementary Fig. 1). R362 is a key residue on the trigger domain that recognizes bases inside the minor groove of the hybrid duplex. As anticipated, the R362A mutation greatly reduced the target DNA binding affinity (Fig. 2i and Supplementary Fig. 1).

The conformational changes

We superimposed free-state and RNA–DNA-bound Crt-SPARTA structures (Fig. 3a) to reveal conformational changes associated with nucleic acid binding, which gave an RMSD of 1.94 Å. Overall, the entire nucleic-acid-binding pocket expands in the presence of RNA–DNA duplex, and structural rearrangements were observed in multiple regions, the most evident of which involves the MID domain (Fig. 3a–c). The MID domain expands at α6–α8, resulting in disorder of the MID-protrusion (α7–α8) region on top of the MID domain. Subsequently, the expanded MID pushes the PIWI domain outwards (Fig. 3b). This implies a cryptic involvement of the MID domain in nucleic acid binding, although we did not find direct contacts between nucleic acids and the MID domain. In addition, the lasso unit tilts outwards (Fig. 3c), the β12–β13 loop of the linker domain moves downwards and the β7–α12 loop and the β5–β6 loop of the PIWI domain shift upwards (Fig. 3d and Supplementary Video 1).

**Fig. 3: Nucleic-acid-induced conformational changes of *Crt*-SPARTA.**

Combining our structural investigations, we propose a four-step model for nucleic acid recognition by Crt-SPARTA (Extended Data Fig. 7a). The guide strand is first loaded into the Crt-SPARTA heterodimer and its 5′-P end is recognized by the conserved 5′-P-binding pocket in the MID domain. The guide-RNA-loaded Crt-SPARTA searches for a complementary target DNA and anneals with the guide RNA through base pairing, and the hybrid duplex propagates. The 3′ end of the target DNA is recognized by the lasso unit, and owing to the presence of the ‘roadblock’ trigger domain, the 3′ end of the target DNA cannot fully anneal with the guide RNA strands. Finally, the 5′-P end of the guide RNA detaches from its binding pocket in the MID domain so that it can fully base pair with the target DNA. We did not capture the structural intermediates for steps 2 and 3 in this study, probably because they are transient and unstable.

Higher-order Crt-SPARTA structures

We collected cryo-EM data for high-oligomeric species eluted from the size-exclusion column (Extended Data Fig. 5 and Supplementary Fig. 1). Three classes were observed after three-dimensional (3D) classification. Three-dimensional reconstructions uncovered two dimeric and one tetrameric Crt-SPARTA–RNA–DNA complexes to 3.49 Å, 3.50 Å and 3.41 Å resolution, respectively (Extended Data Fig. 8). As shown in Fig. 4a, Crt-SPARTA–RNA–DNA dimers are flat modules that comprise two molecules associating with each other through their MID domains. We further identified two dimer conformations during data analysis: the dimer Conf-1 contains two tail-to-tail-organized TIR domains (Fig. 4a, left), and the dimer Conf-2 contains two head-to-tail-organized TIR domains (Fig. 4a, right). In the dimer Conf-2, the TIR domain of one Crt-SPARTA flips upwards by around 180° so that it can interact with the other TIR of the adjacent molecule (Fig. 4a, right). A common feature of the two dimer types is the interaction at the MID–MID interface, involving 4 segments and 13 residues, most of which are polar contacts (Fig. 4b). The buried MID–MID interfacial areas in Conf-1 and Conf-2 are highly similar (1,581.6 Å² and 1,548.6 Å², respectively).

To validate the dimeric Crt-SPARTA structure, we introduced a selection of mutations at the dimerization interface. Mutations on four segments that mediate the MID–MID interface not only ablated the NADase activity of Crt-SPARTA to different levels (Fig. 4c), but also undermined Crt-SPARTA tetramerization (Fig. 4d), which supports the idea that the Crt-SPARTA NADase activity is dependent on the formation of tetramers. For example, Y37 has a key role in the MID–MID interaction by extending its aromatic side chain into a hydrophobic cavity in the adjacent molecule (Fig. 4b, right). As anticipated, Y37A not only lost the NADase activity, but also caused the complete disappearance of the tetramer in size-exclusion chromatograph: nearly all tetramer species shifted to monomers (Fig. 4d). K130 participates in a polar contact with the neighbouring molecule. Mutant K130A still retained a fraction of tetramer, which coincides with its residual NADase activity level: around half of the wild-type activity (Fig. 4c,d).

In the tetrameric structure, two Crt-SPARTA–RNA–DNA dimers are held together through their TIR domains. At the dimer–dimer interface, four TIR domains associate into a cluster (Fig. 4e), suggesting an NADase active TIR assembly. Of note, each Crt-SPARTA dimer of the tetramer contains one TIR domain in the ‘flipped’ conformation (TIR-2 or TIR-4), and one TIR in the original conformation (TIR-1 or TIR-3).

TIR association and NADase activation

Self-association of TIR domains is key to their activation and hence their biological function. We captured in our cryo-EM structure the association of four Crt-TIR domains induced by guide RNA–target DNA activators. As shown in Fig. 5a, the Crt-TIR assembly consists of two parallel head-to-tail-organized TIR dimers: the BB loop (βB–αB loop) of one TIR interacts with the EE (the βD and βE strands and the αE helix)⁴² interface of an adjacent TIR. Two head-to-tail TIR dimers associate with each other through their BCD interfaces (interface involving αB, αC and αD) to form a tetramer. It is worth noting that the Crt-TIR domain contains an extra αF helix bridging the TIR domain and the MID domain. The αF helix functions as a highly flexible linker that allows the MID domain to be oriented in either a cis conformation with TIR-1 or TIR-3, or a trans conformation with the ‘flipped’ TIR-2 or TIR-4 (Figs. 5a and 4e).

**Fig. 5: Association of *Crt-*TIR domains.**

TIR assemblies are categorized into two types: scaffold assemblies and the enzyme assemblies³⁶. Acinetobacter baumannii TIR (Ab-TIR) belongs to the scaffold assembly group; these proteins are analogous to the TLR adaptor proteins MyD88 and MAL (MyD88 adaptor-like protein)³⁵. Plant TNLs (TIR-domain-containing NLRs), ROQ1 (recognition of XopQ 1), RPP1 (recognition of Peronospora parasitica 1) and SARM1 (sterile alpha and TIR motif containing 1) have similar TIR assemblies and belong to the enzyme assembly group. Both types of TIR assembly feature large open-ended complexes comprising two TIR strands, in which all TIR domains are organized head to tail. Whereas the orientation of two TIR strands is antiparallel in enzyme assemblies, the orientation in scaffold assemblies is parallel³⁶.

The organization of the Crt-TIR tetramer shows similarities with Ab-TIR and MAL-TIR assemblies (Fig. 5b and Extended Data Fig. 9a), but is different from mammalian SARM1 and plant TNLs^34,38 (Extended Data Fig. 9b), indicating that the Crt-TIR assembly belongs to the scaffold assemblies. However, it is unlikely that the Crt-TIR tetramer can further expand into long TIR strands like other scaffold TIR assemblies, owing to the steric hindrance imposed by the Ago part of the Crt-SPARTA (Fig. 5b).

The head-to-tail Crt-TIR dimer alone is more similar to the SARM1-TIR dimer (RMSD of 2.5 Å). Superimposing the head-to-tail Crt-TIR dimer onto the SARM1-TIR:3AD (an NAD⁺ analogue) structure (PDB:7NAI) allowed us to identify a putative NADase active site between two Crt-TIR domains, which is analogous to the SARM1-TIR active site; we located several conserved residues of Crt-SPARTA that are important for NADase activity (Fig. 5c). The structural counterparts of the SARM1-TIR active-site residues W638, L598, F603, L607 and D594 were found to be W46, L39, F45, I49 and D35 in Crt-TIR, implying that they are involved in recognizing the NAD⁺. Crt-TIR R71 is positioned close to the diphosphate group of NAD⁺, but it lacks a structural counterpart in SARM1-TIR; instead, SARM1-TIR R569 lies in proximity to R71. Crt-TIR E77 adopts a similar conformation to the catalytic residue E642 of SARM1-TIR, suggesting that E77 has a similar role in catalysis³¹ (Fig. 5c, right).

To investigate whether the tetrameric Crt-TIR assembly observed in the cryo-EM structure represents an active NADase conformation, we mutated a selection of residues either at the TIR–TIR interfaces or inside the putative NADase active site (Fig. 5c–e). We then compared the NADase activity of these mutants with that of the wild-type enzyme (Fig. 5f). The TIR–TIR interface mutations had a severe effect on NADase activity; most of these mutants retained less than 10% of wild-type NADase activity, and they also lost the ability to undergo tetramerization (Fig. 5g). Although most mutations in the putative active site also caused a severe loss of NADase activity, many of them had a limited effect on tetramerization (Fig. 5g). For example, the NADase activity of the D35A mutant was less than 20% of that of wild-type Crt-SPARTA, but this mutant formed more tetramers than the wild type did (Fig. 5g). In addition, we found active-site mutations that resulted in a severe loss of both NADase activity and tetramerization capability; for example, G42A and D44A. G42 and D44 are located on the BB loop, so their mutations might disrupt the head-to-tail TIR organization and undermine the formation of the active site. Collectively, our mutagenesis studies indicate that the tetrameric TIR assembly and the NADase active site identified in our cryo-EM structure are functionally relevant.

The BE-interface-mediated head-to-tail TIR arrangement is thought to be important for its TIR activation^32,34,35,36. Specifically, the BE interface within the TIR assembly comprises the BB loop on one TIR domain and the DE loop on an adjacent TIR domain; thus, the association of the TIR domain brings together two parts of the active site, thereby activating TIR NADase activity. Nevertheless, investigations into TIR-containing proteins have raised questions regarding the function of the BB loop in NADase catalysis. For example, human MAL-TIR and MyD88-TIR assemblies also involve the BB loop and the BE interface, but they are both catalytically inactive^42,43. Although a TIR-SAVED mutant forms a dimer containing the BE interface and a complete active site, it still lacks the NADase activity (Extended Data Fig. 9c); nevertheless, when TIR-SAVED forms filaments, its NADase activity is recovered³². In addition, a bifunctional plant TIR protein (L7-TIR) forms a TIR assembly that resembles neither the scaffold nor the enzyme assemblies, which further expands the diversity of TIR functions⁴⁴. Comparing the Crt-TIR assembly with a bacterial TIR-STING filament from Sphingobacterium faecium (Sf) revealed similarities as well as differences³³. Structure superimposition shows that the head-to-tail-organized Crt-TIR dimer aligns well with TIR dimers in the SfSTING-TIR strand (Extended Data Fig. 9d). Nevertheless, whereas the Crt-TIR tetramer contains two parallel TIR dimers, the SfSTING-TIR filament contain antiparallel double-TIR strands. The conserved NADase active site of both Crt-TIR and SfSTING-TIR lies at the BE interface between neighbouring TIR domains (Extended Data Fig. 9d). It is worth noting that a complete NADase active site of SfSTING-TIR also involves the DD loop extending from the antiparallel TIR dimer partner (Extended Data Fig. 9d), a feature absent in the Crt-TIR tetramer.

Together, our findings show that the Crt-TIR tetramer has a TIR assembly analogous to the scaffold assemblies, in support of the activation of the NADase activity of Crt-SPARTA³¹. Future investigations of Crt-TIR complexed with NAD⁺ or its mimetics will be necessary to elucidate the Crt-TIR catalytic mechanism.

Conclusions

We determined five high-resolution cryo-EM structures of the Crt-SPARTA system: the free-state heterodimer (3.27 Å); the guide RNA–target DNA-loaded quaternary complex (3.27 Å); two distinct dimer conformations (3.50 Å and 3.49 Å, respectively); and one tetramer of the quaternary complexes (3.41 Å). Combining these sequential states along the transition from target recognition to full activation of the Crt-SPARTA system allows us to propose a working model (Extended Data Fig. 7b). In the presence of invading DNA (phage infection or foreign DNA transformation), the Crt-SPARTA acquires a guide RNA from the invader transcriptome that is complementary to a target DNA segment. Binding of guide RNA to the Crt-SPARTA system is followed by recognition of the target DNA through base pairing. The presence of the trigger domain in the Crt-SPARTA nucleic-acid-binding pocket prevents the propagation of the RNA–DNA duplex. To fully anneal with the target DNA, the guide RNA probably detaches from its binding site in the Crt-MID domain so that the guide RNA–target DNA hybrid duplex can be fully formed. Recognition of the hybrid duplex causes a conformational change in Crt-SPARTA and promotes the dimerization of the nucleic-acid-loaded Crt-SPARTA through the MID domains. During the dimerization process, one TIR domain flips around to interact with another TIR in the adjacent molecule, so that two TIR are organized head to tail and a complete NADase active site is formed at the BE interface. Finally, two Crt-SPARTA dimers associate through their TIR domains to yield a tetrameric Crt-TIR assembly, similar to the scaffold assemblies. The tetrameric Crt-TIR assembly possesses the NADase activity that depletes the NAD(P)⁺ pool in bacterial cells, ultimately causing cell death. The structure of the TIR-APAZ–Ago system has also been independently confirmed by other groups^{45,46,47,48,49,50}.

Methods

Plasmid construction

The whole Crt-SPARTA operon, including genes encoding TIR-APAZ and pAgo, was chemically synthesized by GenScript (Supplementary Table 2). The Crt-SPARTA operon was amplified by PCR and cloned into the pET21a vector to include a tobacco etch virus (TEV)-cleavable N-terminal 6×His–maltose-binding protein (MBP) tag for co-expression of TIR-APAZ and pAgo proteins. In addition, a 45-bp spacer comprising an extra ribosomal binding site between the two genes and mutants was constructed using site-directed mutagenesis (QuickChange) as previously described³¹ (Supplementary Table 3).

Protein expression and purification

To express the Crt-SPARTA binary complex, the constructed pET21a-6×His-MBP-TEV-Crt-SPARTA plasmid was transformed into expression strain E. coli BL21 Star (DE3) cells (TransGen). A single colony was picked and cultured in 4 × 10 ml LB medium overnight. The pre-culture (10 ml) was transferred to 4 × 1l LB medium and incubated at 37 °C with shaking at 180 rpm until the optical density at 600 nm reached 0.8–1.0. The temperature of the incubator was then lowered to 16 °C, protein expression was induced by the addition of isopropyl-β-d-thiogalactopyranoside (IPTG) to a final concentration of 0.25 mM, and the culturing was continued at 16 °C for 16–18 h. For purification of Crt-SPARTA protein, the cells were collected by centrifugation (3,470g, 10 min) at 4 °C. The collected cells were resuspended in lysis buffer (50 mM Tris-HCl pH 8.0, 500 mM NaCl, 10 mM imidazole and 1 mM phenylmethylsulfonyl fluoride (PMSF)) and lysed by sonication. The cell debris was clarified by centrifugation (47,850g, 50 min) at 4 °C, and the clarified lysate was passed through a 0.45-μm syringe filter and, loaded onto Ni-NTA resin (Qiagen) pre-equilibrated with the lysis buffer. The resin was washed with 10 column volumes (CVs) of lysis buffer, and the His-tagged proteins were eluted with elution buffer containing 20 mM Tris-HCl pH 8.0, 500 mM NaCl and 300 mM imidazole. Subsequently, the eluted sample was applied to an amylose resin (New England Biolabs) column, and the resin was washed with 10× CVs buffer containing 20 mM HEPES pH 7.4, 500 mM NaCl and 1 mM DTT. The bound protein was eluted with amylose elution buffer (20 mM HEPES pH 7.4, 500 mM NaCl, 1 mM DTT and 10 mM maltose). The N-terminal 6×His–MBP tag was removed by adding TEV protease and dialysing overnight against 20 mM HEPES pH 7.4, 250 mM NaCl,1 mM DTT and 2 mM EDTA. Dialysed proteins were diluted with buffer containing 20 mM HEPES pH 7.4 and 1 mM DTT to reduce the NaCl concentration. The diluted proteins were centrifuged to remove precipitate, and further purified using a HiTrap Heparin HP column (Cytiva) pre-equilibrated with buffer A (20 mM HEPES pH 7.4, 1 mM DTT and 75 mM KCl), and eluted with a linear gradient of 20 mM HEPES pH 7.4, 2 M KCl and 1 mM DTT. Finally, the elution fractions were pooled, concentrated and further purified by gel filtration chromatography using a Superdex 200 10/300 column (GE Healthcare) in buffer containing 20 mM HEPES pH 7.4, 1 mM DTT and 150 mM KCl.

Preparation of oligonucleotides

The guide RNA (gRNA) and target DNA (tDNA) were chemically synthesized and purchased from GenScript. The sequences are listed in Supplementary Table 4.

In vitro assembly of Crt-SPARTA–gRNA–tDNA quaternary complexes

For reconstitution of Crt-SPARTA–gRNA–tDNA quaternary complexes, purified Crt-SPARTA complexes were pooled, diluted to 1 mg ml⁻¹ and mixed with gRNA for 30 min at 37 °C to allow the formation of a ternary Crt-SPARTA–gRNA complex. Next, tDNA was added to the mixture in a Crt-SPARTA:gRNA:tDNA molar ratio of 1:1.2:1.2, and the mixture was incubated for 1 h at 37 °C, yielding an assembled Crt-SPARTA–gRNA–tDNA quaternary complex. The assembled complex was purified by size-exclusion chromatography using a Superdex 200 10/300 column equilibrated with buffer containing 20 mM HEPES pH 7.4, 2 mM MgCl₂,1 mM DTT and 150 mM KCl. The first size-exclusion chromatography peak corresponds to tetrameric Crt-SPARTA–gRNA–tDNA complexes (Crt-SPARTA bound to gRNA and tDNA with 4:4 stoichiometry) and the second peak corresponds to the monomeric Crt-SPARTA–gRNA–tDNA complex (Crt-SPARTA bound to gRNA and tDNA with 1:1 stoichiometry). Both the first peak and the second peak fractions were separately collected and concentrated for subsequent cryo-EM analysis.

Cryo-EM sample preparation and data collection

For free-state Crt-SPARTA, the Crt-SPARTA sample was freshly prepared and diluted to 0.3–0.6 mg ml⁻¹. A 3.5-μl sample of purified Crt-SPARTA was applied to glow-discharged Quantifoil R1.2/1.3 Cu 300-mesh holey carbon grids. The grids were blotted for 7 s with a blot force of 4 at 100% humidity and 4 °C .

For Crt-SPARTA–gRNA–tDNA quaternary and tetrameric Crt-SPARTA–RNA–DNA complexes, the corresponding samples (3 μl) were applied to freshly glow-discharged Quantifoil R1.2/1.3 Au 300-mesh grids. After blotting for 2–4 s with a blot force of 4 under 100% humidity conditions at 4 °C, all grids were plunge-frozen into liquid ethane using a Vitrobot Mark IV instrument (Thermo Fisher Scientific). Images were collected at the Center for Biological Imaging, Core Facilities for Protein Science, Institute of Biophysics, Chinese Academy of Sciences; Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences; and Beijing Branch of Songshan Lake Laboratory for Materials Science. Specifically, cryo-EM data were collected on the FEI Titan Krios electron microscope (Thermo Fisher Scientific) at an acceleration voltage of 300 kV equipped with a Gatan K3 direct electron detector. Each movie stack was dose-fractionated to 32 frames with a total electron dose of 66 e⁻ Å⁻² and a total exposure time of 6 s, resulting a physical pixel size of 0.82 Å or 1.04 Å per pixel. Data were collected using EPU software and/or SerialEM software⁵¹ by using a beam-image shift data-collection strategy and a defocus value ranging from −1.8 to −2.5 μm (ref. ⁵²).

Cryo-EM data processing

Data processing was performed using cryoSPARC-3.3.2 (ref. ⁵³). All dose-fractionated movies were translationally aligned using patch motion correction⁵⁴, and the contrast transfer function (CTF) parameters were estimated using patch-based CTF estimation in cryoSPARC (ref. ⁵⁵).

Processing of free-state Crt-SPARTA

For free-state Crt-SPARTA, a total of 4,729,814 particles were automatically picked from 7,330 micrographs and subjected to several rounds of two-dimensional (2D) classification, rebalancing of 2D classes and 3D classification. The best 1,158,576 particles were then selected and subjected to homogeneous refinement, non-uniform refinement, local refinement and DeepEMhancer auto-sharpening, yielding a map with an overall resolution of 3.27 Å based on a gold-standard Fourier shell correlation (FSC) criterion of 0.143 (ref. ⁵⁶).

Processing of Crt-SPARTA–gRNA–tDNA quaternary complexes

The data for the Crt-SPARTA–gRNA–tDNA quaternary complexes were processed similarly. In brief, 7,080 micrographs were collected, and 4,231,254 particles were initially picked using blob picker and template picker, followed by several rounds of 2D classification, rebalancing of 2D classes and 3D classification to curate particle sets. The best 3D classification with gRNA–tDNA density containing 1,040,721 particles was refined using homogeneous refinement, non-uniform refinement, local refinement and DeepEMhancer auto-sharpening, yielding a cryo-EM map at an overall resolution of 3.27 Å based on an FSC criterion of 0.143 (ref. ⁵⁶). The local resolution was estimated in cryoSPARC.

Processing of dimeric and tetrameric Crt-SPARTA–gRNA–tDNA

Regarding data processing for higher-order oligomeric structures, a total of 6,411,451particles were first auto-picked and extracted from the 20,299 micrographs. After several rounds of 2D classification and rebalancing of 2D classes, bad particles and those with monomer features were removed. A total of 2,420,916 good particles were subsequently selected for 3D classification (n = 3) without imposing any symmetry. Class-1 (614,651) and class-2 (686,832) particles were used for further 3D homogeneous refinement, non-uniform refinement, local refinement and DeepEMhancer auto-sharpening, yielding 3.49-Å and 3.50-Å maps for the two dimers of the Crt-SPARTA–gRNA–tDNA assembly based on an FSC criterion of 0.143. Class-3 particles (1,119,433) were used to produce the tetrameric Crt-SPARTA–gRNA–tDNA assembly. All particles with this conformation were used for final non-uniform refinement, local refinement and DeepEMhancer auto-sharpening, resulting in density maps with a resolution of 3.41 Å (ref. ⁵⁶).

Model building and refinement

The AlphaFold2-predicted model of Crt-SPARTA was used as the initial model⁵⁷, and the model was rigid-body-fitted into the Crt-SPARTA heterodimer density map using UCSF Chimera⁵⁸ and adjusted manually using Coot⁵⁹. The gRNA–tDNA duplex was modelled as ideal B-form within Coot⁵⁹. Because tDNA is longer than gRNA, and because the 5′ end of gRNA is preferentially bound by Ago of Crt-SPARTA, by evaluating the density map length of gRNA and tDNA, we can clearly distinguish between gRNA and tDNA density. The models of dimeric and tetrameric Crt-SPARTA–gRNA–tDNA complexes were produced by docking the corresponding numbers of Crt-SPARTA–gRNA–tDNA quaternary complexes into the density maps using UCSF Chimera⁵⁸ and adjusting manually in Coot⁵⁹. The models were refined using real-space refinement in PHENIX⁶⁰.The overall model validations were performed using the comprehensive validation tool (cryo-EM) in PHENIX⁶⁰. The statistics of the cryo-EM data collection, 3D reconstruction and model refinement are summarized in Extended Data Table 1, and the representative EM densities for the different functional states are illustrated in Supplementary Fig. 5. Structural figures were prepared using PyMOL (Molecular Graphics System) or Chimera X (ref. ⁶¹).

Urea –PAGE analysis

The freshly prepared Crt-SPARTA–gRNA–tDNA quaternary complexes and tetrameric Crt-SPARTA–gRNA–tDNA complexes were incubated with loading buffer (Thermo Fisher Scientific) and resolved by 15% denaturing urea–PAGE with 1× TBE buffer for 70 min. After electrophoresis, the gels were strained with SYBR GOLD (Thermo Fisher Scientific) and visualized with the ChemiDoc Imaging System (Bio-Rad).

NADase assay

The NADase activity of Crt-SPARTA was measured using an NAD fluorescent analogue ɛNAD⁺ as the substrate. After cleavage, the product ɛADP-ribose serves as a fluorophore for fluorescence detection and quantification (excitation wavelength of 310 nm and emission wavelength of 410 nm). Each reaction (100 μl) contains reaction buffer, 10 mM MES pH 7.5, 75 mM KCl, 2 mM MgCl₂, and 0.2 μM Crt-SPARTA, and the indicated amount of ɛNAD⁺ (Sigma), gRNA and tDNA. The reaction was initialized by adding nucleic acids (activators). gRNA (slightly excess over protein concentration) was first added into the reaction mixture. After 15-min incubation at 37 °C, an excess amount of tDNA was added to initialize the reaction. The reaction mixtures were loaded into Greiner 96-well half-area plates and fluorescence was measured continuously for 60 min with 2-min intervals on a luminescence multimode microplate reader, SpectraMax M5. All reactions were conducted at 55 °C. A calibration line was plotted using the data measured for 60 min of reaction with 17, 33, 50, 67 and 83 μM of ɛNAD⁺ as the initial concentration.

To compare the NADase activity of various Crt-SPARTA mutants, the concentrations of ɛNAD⁺ and Crt-SPARTA were kept at 500 μM and 0.2 μM, respectively, for all reactions. Fluorescence was measured continuously for 60 min with 2-min intervals.

All experiments were repeated three times and results are presented as mean values with s.d.

Electrophoretic mobility shift assay

The electrophoretic mobility shift assay was performed as previously described⁶². In brief, unlabelled gRNA and tDNA containing sulfo-cyanine 5.5 (Cy5.5) dye at the 5′ terminus (GenScript) were synthesized. For the gRNA binding assay, 2.5 µM gRNA was incubated with Crt-SPARTA (0–1.6 µM) in a binding buffer containing 20 mM HEPES pH 7.4, 100 mM KCl, 5% glycerol, 2 mM MgCl₂ and 0.5 mM TCEP for 30 min at 37 °C. For the tDNA binding assay, Crt-SPARTA was first incubated with gRNA to allow formation of the Crt-SPARTA–gRNA complex, and then 0.5 µM Cy5.5-labelled tDNA was added to the Crt-SPARTA–gRNA complex (0–0.8 µM) in binding buffer for 30 min at 37 °C. The mixtures were loaded onto a 6% native PAGE running in 1× TBE buffer (Solarbio) at 4 °C. gRNAs in the gel were stained using SYBR Gold Nucleic Acid Gel Stain (Thermo Fisher Scientific) and visualized using the Bio-Rad Gel Doc XR System. The Cy5.5-labelled tDNAs in the gel were visualized using an Odyssey infrared imaging system (Li-COR Biotechnology).

Size-exclusion chromatography

Purified Crt-SPARTA and its mutants were concentrated to 1 mg ml⁻¹ before being incubated with gRNA for 30 min at 37 °C to allow formation of the Crt-SPARTA–gRNA complex. Next, tDNA was added to the mixtures, which were then incubated for 1 h at 37 °C to allow formation of the quaternary Crt-SPARTA–gRNA–tDNA complex. The overall molar ratio of Crt-SPARTA:gRNA:tDNA was 1:1.2:1.2. The quaternary complex was loaded onto a Superdex 200 Increase 10/300 column (Cytiva) pre-equilibrated with a buffer containing 20 mM HEPES pH 7.4, 2 mM MgCl₂, 1 mM DTT and 100 mM KCl. The size-exclusion column was calibrated using Gel Filtration Standard (BIO-RAD). The eluates were further analysed by SDS–PAGE.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The atomic models of the Crt-SPARTA binary complex, Crt-SPARTA–gRNA–tDNA quaternary complex, dimer Conf-1 Crt-SPARTA–gRNA–tDNA complexes, dimer Conf-2 and tetrameric Crt-SPARTA–gRNA–tDNA complexes have been deposited into the Protein Data Bank with accession codes 8ISY, 8ISZ, 8IT0, 8K9G and 8IT1, and the corresponding 3D cryo-EM density maps have been deposited into the Electron Microscopy Data Bank under the accession codes EMD-35700, EMD-35701, EMD-35702, EMD-36986 and EMD-35703, respectively. Publicly available protein atomic models with the following PDB codes were used in the study: 4N76, 7UXU, 7NAI, 4N47, 2BGG, 4F3T, 1U04, 4W5O, 5UZB, 7NAK, 7QQK and 7UN8. Source data are provided with this paper.

References

Swarts, D. C. et al. DNA-guided DNA interference by a prokaryotic Argonaute. Nature 507, 258–261 (2014).
Article CAS PubMed PubMed Central ADS Google Scholar
Sashital, D. G. Prokaryotic Argonaute uses an all-in-one mechanism to provide host defense. Mol. Cell 65, 957–958 (2017).
Article CAS PubMed Google Scholar
Kuzmenko, A. et al. DNA targeting and interference by a bacterial Argonaute nuclease. Nature 587, 632–637 (2020).
Article CAS PubMed ADS Google Scholar
Swarts, D. C. et al. Argonaute of the archaeon Pyrococcus furiosus is a DNA-guided nuclease that targets cognate DNA. Nucleic Acids Res. 43, 5120–5129 (2015).
Article CAS PubMed PubMed Central Google Scholar
Zander, A. et al. Guide-independent DNA cleavage by archaeal Argonaute from Methanocaldococcus jannaschii. Nat. Microbiol. 2, 17034 (2017).
Article CAS PubMed Google Scholar
Doxzen, K. W. & Doudna, J. A. DNA recognition by an RNA-guided bacterial Argonaute. PLoS One 12, e0177097 (2017).
Article PubMed PubMed Central Google Scholar
Kaya, E. et al. A bacterial Argonaute with noncanonical guide RNA specificity. Proc. Natl Acad. Sci. USA 113, 4057–4062 (2016).
Article CAS PubMed PubMed Central ADS Google Scholar
Kropocheva, E., Kuzmenko, A., Aravin, A. A., Esyunina, D. & Kulbachinskiy, A. A programmable pAgo nuclease with universal guide and target specificity from the mesophilic bacterium Kurthia massiliensis. Nucleic Acids Res. 49, 4054–4065 (2021).
Article CAS PubMed PubMed Central Google Scholar
Lapinaite, A., Doudna, J. A. & Cate, J. H. D. Programmable RNA recognition using a CRISPR-associated Argonaute. Proc. Natl Acad. Sci. USA 115, 3368–3373 (2018).
Article CAS PubMed PubMed Central ADS Google Scholar
Olovnikov, I., Chan, K., Sachidanandam, R., Newman, D. K. & Aravin, A. A. Bacterial argonaute samples the transcriptome to identify foreign DNA. Mol. Cell 51, 594–605 (2013).
Article CAS PubMed Google Scholar
Jolly, S. M. et al. Thermus thermophilus Argonaute functions in the completion of DNA replication. Cell 182, 1545–1559 (2020).
Article CAS PubMed PubMed Central Google Scholar
Swarts, D. C. Prokaryotic Argonautes function beyond immunity by unlinking replicating chromosomes. Cell 182, 1381–1383 (2020).
Article CAS PubMed Google Scholar
Fu, L. et al. The prokaryotic Argonaute proteins enhance homology sequence-directed recombination in bacteria. Nucleic Acids Res. 47, 3568–3579 (2019).
Article CAS PubMed PubMed Central Google Scholar
Lee, K. Z. et al. NgAgo possesses guided DNA nicking activity. Nucleic Acids Res. 49, 9926–9937 (2021).
Article CAS PubMed PubMed Central Google Scholar
Ryazansky, S., Kulbachinskiy, A. & Aravin, A. A. The expanded universe of prokaryotic Argonaute proteins. mBio 9, e01935-18 (2018).
Article PubMed PubMed Central Google Scholar
Makarova, K. S., Wolf, Y. I., van der Oost, J. & Koonin, E. V. Prokaryotic homologs of Argonaute proteins are predicted to function as key components of a novel system of defense against mobile genetic elements. Biol. Direct 4, 29 (2009).
Article PubMed PubMed Central Google Scholar
Swarts, D. C. et al. The evolutionary journey of Argonaute proteins. Nat. Struct. Mol. Biol. 21, 743–753 (2014).
Article CAS PubMed PubMed Central Google Scholar
Willkomm, S. et al. Structural and mechanistic insights into an archaeal DNA-guided Argonaute protein. Nat. Microbiol. 2, 17035 (2017).
Article CAS PubMed Google Scholar
Song, J. J., Smith, S. K., Hannon, G. J. & Joshua-Tor, L. Crystal structure of Argonaute and its implications for RISC slicer activity. Science 305, 1434–1437 (2004).
Article CAS PubMed ADS Google Scholar
Miyoshi, T., Ito, K., Murakami, R. & Uchiumi, T. Structural basis for the recognition of guide RNA and target DNA heteroduplex by Argonaute. Nat. Commun. 7, 11846 (2016).
Article CAS PubMed PubMed Central ADS Google Scholar
Wang, Y., Sheng, G., Juranek, S., Tuschl, T. & Patel, D. J. Structure of the guide-strand-containing argonaute silencing complex. Nature 456, 209–213 (2008).
Article CAS PubMed PubMed Central ADS Google Scholar
Elkayam, E. et al. The structure of human argonaute-2 in complex with miR-20a. Cell 150, 100–110 (2012).
Article CAS PubMed PubMed Central Google Scholar
Nakanishi, K., Weinberg, D. E., Bartel, D. P. & Patel, D. J. Structure of yeast Argonaute with guide RNA. Nature 486, 368–374 (2012).
Article CAS PubMed PubMed Central ADS Google Scholar
Schirle, N. T., Sheu-Gruttadauria, J. & MacRae, I. J. Structural basis for microRNA targeting. Science 346, 608–613 (2014).
Article CAS PubMed PubMed Central ADS Google Scholar
Wang, Y. et al. Nucleation, propagation and cleavage of target RNAs in Ago silencing complexes. Nature 461, 754–761 (2009).
Article CAS PubMed PubMed Central ADS Google Scholar
Sheng, G. et al. Structure-based cleavage mechanism of Thermus thermophilus Argonaute DNA guide strand-mediated DNA target cleavage. Proc. Natl Acad. Sci. USA 111, 652–657 (2014).
Article CAS PubMed ADS Google Scholar
Liu, Y. et al. Accommodation of helical imperfections in Rhodobacter sphaeroides Argonaute ternary complexes with guide RNA and target DNA. Cell Rep. 24, 453–462 (2018).
Article PubMed PubMed Central Google Scholar
Kwak, P. B. & Tomari, Y. The N domain of Argonaute drives duplex unwinding during RISC assembly. Nat. Struct. Mol. Biol. 19, 145–151 (2012).
Article CAS PubMed Google Scholar
Willkomm, S., Makarova, K. S. & Grohmann, D. DNA silencing by prokaryotic Argonaute proteins adds a new layer of defense against invading nucleic acids. FEMS Microbiol. Rev. 42, 376–387 (2018).
Article CAS PubMed PubMed Central Google Scholar
Zaremba, M. et al. Short prokaryotic Argonautes provide defence against incoming mobile genetic elements through NAD(+) depletion. Nat. Microbiol. 7, 1857–1869 (2022).
Article CAS PubMed Google Scholar
Koopal, B. et al. Short prokaryotic Argonaute systems trigger cell death upon detection of invading DNA. Cell 185, 1471–1486 (2022).
Article CAS PubMed PubMed Central Google Scholar
Hogrel, G. et al. Cyclic nucleotide-induced helical structure activates a TIR immune effector. Nature 608, 808–812 (2022).
Article CAS PubMed ADS Google Scholar
Morehouse, B. R. et al. Cryo-EM structure of an active bacterial TIR-STING filament complex. Nature 608, 803–807 (2022).
Article CAS PubMed PubMed Central ADS Google Scholar
Shi, Y. et al. Structural basis of SARM1 activation, substrate recognition, and inhibition by small molecules. Mol. Cell 82, 1643–1659 (2022).
Article CAS PubMed PubMed Central Google Scholar
Manik, M. K. et al. Cyclic ADP ribose isomers: production, chemical structures, and immune signaling. Science 377, eadc8969 (2022).
Article CAS PubMed Google Scholar
Nimma, S. et al. Structural evolution of TIR-domain signalosomes. Front. Immunol. 12, 784484 (2021).
Article CAS PubMed PubMed Central Google Scholar
Morehouse, B. R. et al. STING cyclic dinucleotide sensing originated in bacteria. Nature 586, 429–433 (2020).
Article CAS PubMed PubMed Central ADS Google Scholar
Horsefield, S. et al. NAD⁺ cleavage activity by animal and plant TIR domains in cell death pathways. Science 365, 793–799 (2019).
Article CAS PubMed ADS Google Scholar
Wang, Y. et al. Structure of an argonaute silencing complex with a seed-containing guide DNA and target RNA duplex. Nature 456, 921–926 (2008).
Article CAS PubMed PubMed Central ADS Google Scholar
Egli, M., Usman, N., Zhang, S. G. & Rich, A. Crystal structure of an Okazaki fragment at 2-Å resolution. Proc. Natl Acad. Sci. USA 89, 534–538 (1992).
Article CAS PubMed PubMed Central ADS Google Scholar
Ma, J. B. et al. Structural basis for 5′-end-specific recognition of guide RNA by the A. fulgidus Piwi protein. Nature 434, 666–670 (2005).
Article CAS PubMed PubMed Central ADS Google Scholar
Ve, T. et al. Structural basis of TIR-domain-assembly formation in MAL- and MyD88-dependent TLR4 signaling. Nat. Struct. Mol. Biol. 24, 743–751 (2017).
Article CAS PubMed PubMed Central Google Scholar
Clabbers, M. T. B. et al. MyD88 TIR domain higher-order assembly interactions revealed by microcrystal electron diffraction and serial femtosecond crystallography. Nat. Commun. 12, 2578 (2021).
Article CAS PubMed PubMed Central ADS Google Scholar
Yu, D. et al. TIR domains of plant immune receptors are 2′,3′-cAMP/cGMP synthetases mediating cell death. Cell 185, 2370–2386 (2022).
Article CAS PubMed Google Scholar
Guo, M. et al. Cryo-EM structure of the ssDNA-activated SPARTA complex. Cell Res. 33, 731–734 (2023).
Article CAS PubMed Google Scholar
Zhang, J.-T., Wei, X.-Y., Cui, N., Tian, R. & Jia, N. Structural basis for ssDNA-activated NADase activity of the prokaryotic SPARTA immune system. Preprint at bioRxiv https://doi.org/10.1101/2023.07.14.549122 (2023).
Guo, L. et al. Structural basis for auto-inhibition and activation of a short prokaryotic Argonaute associated TIR-APAZ defense system. Preprint at bioRxiv https://doi.org/10.1101/2023.07.12.548734 (2023).
Shen, Z. et al. Oligomerization-mediated activation of a short prokaryotic Argonaute. Nature 621, 154–161 (2023).
Article CAS PubMed ADS Google Scholar
Wang, X. et al. Structural insights into mechanisms of Argonaute protein-associated NADase activation in bacterial immunity. Cell Res. 33, 699–711 (2023).
Article CAS PubMed PubMed Central ADS Google Scholar
Ni, D., Lu, X., Stahlberg, H. & Ekundayo, B. Activation mechanism of a short argonaute-TIR prokaryotic immune system. Sci. Adv. 9, eadh9002 (2023).
Article CAS PubMed PubMed Central Google Scholar
Mastronarde, D. N. Automated electron microscope tomography using robust prediction of specimen movements. J. Struct. Biol. 152, 36–51 (2005).
Article PubMed Google Scholar
Wu, C., Huang, X., Cheng, J., Zhu, D. & Zhang, X. High-quality, high-throughput cryo-electron microscopy data collection via beam tilt and astigmatism-free beam-image shift. J. Struct. Biol. 208, 107396 (2019).
Article CAS PubMed Google Scholar
Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods 14, 290–296 (2017).
Article CAS PubMed Google Scholar
Zheng, S. Q. et al. MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nat. Methods 14, 331–332 (2017).
Article CAS PubMed PubMed Central Google Scholar
Rohou, A. & Grigorieff, N. CTFFIND4: fast and accurate defocus estimation from electron micrographs. J. Struct. Biol. 192, 216–221 (2015).
Article PubMed PubMed Central Google Scholar
Sanchez-Garcia, R. et al. DeepEMhancer: a deep learning solution for cryo-EM volume post-processing. Commun. Biol. 4, 874 (2021).
Article PubMed PubMed Central Google Scholar
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Article CAS PubMed PubMed Central ADS Google Scholar
Pettersen, E. F. et al. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605–1612 (2004).
Article CAS PubMed Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. D 66, 486–501 (2010).
Article CAS PubMed PubMed Central ADS Google Scholar
Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr. D 66, 213–221 (2010).
Article CAS PubMed PubMed Central ADS Google Scholar
Pettersen, E. F. et al. UCSF ChimeraX: structure visualization for researchers, educators, and developers. Protein Sci. 30, 70–82 (2021).
Article CAS PubMed Google Scholar
Gao, X. et al. Structural basis for Sarbecovirus ORF6 mediated blockage of nucleocytoplasmic transport. Nat. Commun. 13, 4782 (2022).
Article CAS PubMed PubMed Central ADS Google Scholar

Download references

Acknowledgements

We thank X. Huang, B. Zhu, X. Li, L. Chen, T. Niu and other staff members at the Center for Biological Imaging, Core Facilities for Protein Science at the Institute of Biophysics, Chinese Academy of Sciences for support with cryo-EM data collection; D. Sun and other staff members from Beijing National Laboratory for Condensed Matter Physics at the Institute of Physics, Chinese Academy of Sciences for cryo-EM data collection and technical assistance; and J. Wan for support with cryo-EM data processing. This work was supported by the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (2021-I2M-1-037 to S.C. and X.G.); the National Natural Science Foundation of China (81971985 and 82272308 to X.G.; 81572005 to S.C.); the National Key Research and Development Program of China (2019YFC0840602 to X.G.); and the National Natural Science Foundation of China/RGC Joint Research Scheme (82261160398, N_HKU767/22).

Author information

These authors contributed equally: Xiaopan Gao, Kun Shang, Kaixiang Zhu, Linyue Wang, Zhixia Mu

Authors and Affiliations

NHC Key Laboratory of Systems Biology of Pathogens, Institute of Pathogen Biology, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
Xiaopan Gao, Kaixiang Zhu, Linyue Wang, Zhixia Mu, Bo Qin & Sheng Cui
Key Laboratory of Pathogen Infection Prevention and Control, Peking Union Medical College, Ministry of Education, Beijing, China
Xiaopan Gao, Kaixiang Zhu, Linyue Wang, Zhixia Mu, Bo Qin & Sheng Cui
Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing, China
Kun Shang, Xingke Fu, Hongtao Zhu & Wei Ding
Medical School, Yan’an University, Yan’an, China
Kun Shang
National Clinical Laboratory on Tuberculosis, Beijing Key Laboratory for Drug-resistant Tuberculosis Research, Beijing Chest Hospital, Capital Medical University, Beijing Tuberculosis and Thoracic Tumor Institute, Beijing, China
Xia Yu

Authors

Xiaopan Gao
View author publications
You can also search for this author in PubMed Google Scholar
Kun Shang
View author publications
You can also search for this author in PubMed Google Scholar
Kaixiang Zhu
View author publications
You can also search for this author in PubMed Google Scholar
Linyue Wang
View author publications
You can also search for this author in PubMed Google Scholar
Zhixia Mu
View author publications
You can also search for this author in PubMed Google Scholar
Xingke Fu
View author publications
You can also search for this author in PubMed Google Scholar
Xia Yu
View author publications
You can also search for this author in PubMed Google Scholar
Bo Qin
View author publications
You can also search for this author in PubMed Google Scholar
Hongtao Zhu
View author publications
You can also search for this author in PubMed Google Scholar
Wei Ding
View author publications
You can also search for this author in PubMed Google Scholar
Sheng Cui
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

S.C. and X.G. designed the study. X.G., K.S., K.Z., L.W. and X.F. prepared grids. S.C., W.D., H.Z. and X.G. determined the cryo-EM structures and performed the model building. S.C., X.G., H.Z. and W.D. created figures. S.C., X.G., H.Z. and W.D. wrote and revised the paper. X.G., K.S., K.Z., L.W., Z.M., B.Q. and X.Y. performed protein expression and purification, mutagenesis studies and NADase assays. X.G., H.Z., W.D., L.W., K.Z., K.S. and S.C. analysed the data. All authors reviewed the results and approved the final version of the manuscript.

Corresponding authors

Correspondence to Hongtao Zhu, Wei Ding or Sheng Cui.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks the anonymous reviewers for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Single-particle cryo-EM analysis of the Crt-SPARTA heterodimer and the Crt-SPARTA–gRNA–tDNA quaternary complex.

a, Workflow of cryo-EM data processing for Crt-SPARTA heterodimer. b, Left, representative cryo-EM micrograph of Crt-SPARTA heterodimer from 7,330 movies was shown; Right, representative 2D averaged classification from the particles used for final reconstruction. c, Gold-standard Fourier shell correlation (FSC) curves of the final Crt-SPARTA heterodimer. The blue line indicates the 0.143 cut-off criterion, indicating a nominal resolution of 3.27 Å. d, Local resolution cryo-EM density map calculated using cryoSPARC, blue to red indicates high to low resolution. e, Single-particle cryo-EM image processing workflow of Crt-SPARTA–gRNA–tDNA quaternary complex. f, Left, representative micrograph of Crt-SPARTA–gRNA–tDNA quaternary complex from 7,080 movies was shown; Right, representative 2D averaged class images from the particles used for final reconstruction. g, Gold-standard Fourier shell correlation (FSC) curves. The map-to-map FSC curve was calculated between the two independently refined half-maps after masking (blue line), and the overall resolution was determined by the gold-standard FSC = 0.143 criterion. h, Local map resolution of the final structure, blue to red indicates high to low resolution.

Extended Data Fig. 2 Structure comparison of short pAgo, long pAgo, eAgo and Crt-SPARTA, and structural alignment of conserved Ago domains of Crt-SPARTA with TtAgo and human Ago2.

a, The domain architectures of short, long pAgos and/or eAgos. b, Structural comparison of Crt-SPARTA and the three representative Ago proteins: short AfAgo (PDB:2BGG), human Ago2(PDB:4F3T) and long PfAgo (PDB:1U04). The different domains are coloured according to the schematic in a. The additional trigger and TIR domain in Crt-SPARTA are coloured in blue and marine, respectively. c–f, Superimpositions of the individual domains of Crt-SPARTA with the TtAgo (yellow, PDB: 4N76) and human Ago2 (light blue, PDB: 4F3T). c, Comparison between N-like domains. d, Comparison between linker domains; the missing PAZ domain in Crt-SPARTA is labelled with a red dotted line circle. e, Comparison between MID domains; the unique insertion in MID domain of Crt-SPARTA is labelled with a red dotted line circle. f, Comparison between PIWI domains.

Extended Data Fig. 3 Ribbon models of Crt-TIR, Crt-pPAZ and Crt-Ago and secondary-structure diagrams.

a, Top, cryo-EM density map of Crt-SAPARTA heterodimer (white) with the TIR domain highlighted in light blue. Bottom, ribbon model of Crt-TIR using the same colour code; secondary structures are labelled. b, Secondary-structure diagram of Crt-TIR; TIR core domain is highlighted with a dashed box. c, Top, cryo-EM density map of Crt-SAPARTA heterodimer (white) with the N-like domain highlighted in cyan, linker domain highlighted in grey and trigger domain highlighted in dark blue. Bottom, ribbon model of Crt-pPAZ with the secondary structures labelled. d, Secondary-structure diagram of Crt-pPAZ; linker domain is composed of L0, L1 and L2, which are indicated by dashed regions. e, Top, cryo-EM density map of Crenotalea thermophila (Crt) SAPARTA heterodimer; Crt-Ago is highlighted by pink (MID) and magenta (PIWI), whereas the rest of the complex is coloured white. Bottom, ribbon model of Crt-Ago with the secondary structures labelled. f, Secondary-structure diagram of Crt-Ago.

Extended Data Fig. 4 Superimposition of Crt-SPARTA with TtAgo and human Ago2.

a, The structure of the Crt-SPARTA heterodimer is superimposed with the structure of TtAgo complexed with gDNA and tDNA (PDB: 4N47) b, The structure of the Crt-SPARTA heterodimer is superimposed with the structure of human Argonaute 2 bound to a gRNA and tRNA (human Ago2, PDB: 4W5O). Crt-SPARTA is coloured by chain. Crt-TIR-pPAZ is coloured light blue, Crt-Ago is coloured magenta; individual domains of Crt-SPARTA are labelled. TtAgo and human Ago2 are coloured grey and the guide DNA or RNA and target DNA or RNA are coloured orange and red, respectively. Crt-pPAZ lacks the PAZ domain but retains an N-like and a linker domain, topologically equivalent to the N and L1–L2 domains of TtAgo and human Ago2. Crt-pPAZ contains a unique C-terminal trigger domain unseen in previous Ago structures. The trigger domain seals the exit of the target DNA–RNA-binding tunnel.

Extended Data Fig. 5 Preparation of the Crt-SPARTA heterodimer and the quaternary protein–nucleic acid complex containing the target DNA and the guide RNA.

a, Schematic of Crt-SPARTA system protein domains, and the guide RNA and target DNA used in sample preparation. b, Purification of the Crt-SPARTA–gRNA–tDNA system. When Crt-SPARTA was incubated with guide RNA and target ssDNA in a 1:1.2:1.2 molar ratio and subjected to size-exclusion chromatography, Crt-SPARTA forms Crt-SPARTA–gRNA–tDNA monomer and Crt-SPARTA–gRNA–tDNA tetramer. The absorbance at A280 (protein) and absorbance at A260 (nucleic acid) are indicated by the blue line and the red line, respectively. c, SDS–PAGE analysis of the elution fractions corresponding to the gel filtration chromatography shown in b. The peak fractions containing the target complex were used for cryo-EM analysis. Data are representative of three independent experiments. d, Elution peaks of the co-eluted nucleic acids shown in b are detected by 15% urea–PAGE. Data are representative of three independent experiments.

Extended Data Fig. 6 Structural details of the interactions between Crt-SPARTA and nucleic acids.

a, The nucleic acid substrates of Crt-SPARTA form an unusual B-form duplex. The structure of Crt-SPATA–RNA–DNA hybrid duplex (coloured green and yellow) is superimposed to an ideal B-form duplex (coloured grey, left) and an ideal A-form duplex (coloured grey, right) using the software Matchmaker. b, Magnified view of the nucleic-acid-binding pocket of Crt-SPARTA–RNA–DNA complex with superimposed TtAgo–DNA–DNA complex (PDB: 4N47). Guide RNA (green) of Crt-SPARTA superimposed with the guide DNA of TtAgo (black) are shown with ribbon models. The trigger domain (blue) is close to the guide RNA but there are no direct contacts between them. The superimposed guide DNA of TtAgo extends near the 5’-phosphate-binding pocket in Crt-PIWI, whereas the 5′ terminus of Crt-guide RNA is around five base steps behind. c, Target DNA (yellow) of Crt-SPARTA with superimposed template DNA of TtAgo (black); TtAgo template DNA extends around five base steps ahead of Crt-SPARTA template DNA. The trigger domain recognizes the 3′ terminus of Crt-SPARTA template DNA, whereas the 3′ terminus of the superimposed TtAgo DNA template strand is around five base steps ahead. d–g, Magnified views of the Crt-SPARTA–nucleic acid interactions. Residues and nucleotides participating in the interaction are shown as stick models and labelled. d, Nucleic acid recognition by the lasso unit. e, Nucleic acid recognition by the Crt-TIR-pPAZ domain. f, Nucleic acid recognition by the Crt-PIWI RNase H domain. g, Nucleic acid recognition by the Crt-PIWI-Box domain.

Extended Data Fig. 7 Working models.

a, A proposed working model for the guide RNA–target DNA-induced conformational changes of Crt-SPARTA. The process is divided into four steps; steps 1 and 4 are supported by the cryo-EM structures determined in this study, and steps 2 and 3 (encircled by a dashed box) are postulated. b, A working model of the Crt-SPARTA system. The working model describes the process of Crt-SPARTA activation through guide RNA-mediated DNA target binding. Following phage infection or foreign DNA transformation, the autoinhibitory Crt-SPARTA loads guide RNA to recognize a segment in the invading DNAs. Guide RNA-mediated target DNA binding induces conformational changes in the Crt-SPARTA that facilitate dimerization Crt-SPARTA–RNA–DNA complexes. Next, two dimeric Crt-SPARTA–RNA–DNA complexes form tetrameric Crt-SPARTA–RNA–DNA complexes, in which the NAD(P)ase activity of the TIR assemblies is activated, resulting in NAD(P)⁺ depletion and cell death.

Extended Data Fig. 8 Cryo-EM single-particle analysis of dimeric and tetrameric Crt-SPARTA–RNA–DNA complexes.

a, Flow chart of single-particle cryo-EM image processing for two dimeric and tetrameric Crt-SPARTA–RNA–DNA complexes b, Representative cryo-EM micrographs of two dimeric and tetrameric Crt-SPARTA–gRNA–tDNA complexes from 20,299 movies was shown are shown at 290,000× magnifications; Right, representative reference-free 2D-class averages of two dimeric and tetrameric Crt-SPARTA–gRNA–tDNA complexes. c, The gold-standard Fourier shell correlation (FSC) curves of two dimeric and tetrameric Crt-SPARTA–gRNA–tDNA complexes for the reconstruction are indicated with resolutions at FSC = 0.143 criterion. d, Local map resolution of the two dimeric and tetrameric Crt-SPARTA–gRNA–tDNA complexes, blue to red indicates high to low resolution.

Extended Data Fig. 9 Comparison of the TIR domain observed in MAL-TIR and TIR-SAVED.

a, Structural superimposition of the Crt-TIR (magenta) with MAL-TIR (PDB:5UZB) (yellow), the BCD interface consisting of αB, αC and αD is labelled. b, The TIR domain tetramers observed in SARM1 (PDB: 7NAK), the BE interface and AE interface are labelled. c, The BE interface of a TIR-SAVED dimer composed of the BB loop and DE loop is labelled (PDB:7QQK). d, Structural comparison of the Crt-TIR assembly with bacterial STING-TIR assembly. A head-to-tail-organized Crt-TIR dimer (blue) is superimposed with SfSTING-TIR filaments (coloured by monomer, PDB: 7UN8). The BB loop of TIR domains, the NADase activate site and the DD loop extending from the opposite TIR filament are labelled. The direction of SfSTING-TIR filaments is indicated by arrows.

Extended Data Table 1 Cryo-EM data collection, refinement and validation statistics

Full size table

Supplementary information

Supplementary Information

This file contains Supplementary Figures 1–5, Supplementary Tables 1-4 and legend for Supplementary Video 1

Reporting Summary

Supplementary Video 1

Binding of guide RNA–target DNA hybrid duplex induce conformational changes in Crt-SPARTA

Source data

Source Data Fig. 2

Source Data Fig. 4

Source Data Fig. 5

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Cite this article

Gao, X., Shang, K., Zhu, K. et al. Nucleic-acid-triggered NADase activation of a short prokaryotic Argonaute. Nature 625, 822–831 (2024). https://doi.org/10.1038/s41586-023-06665-6

Download citation

Received: 05 July 2023
Accepted: 21 September 2023
Published: 02 October 2023
Issue Date: 25 January 2024
DOI: https://doi.org/10.1038/s41586-023-06665-6

This article is cited by

Insights into the modulation of bacterial NADase activity by phage proteins
- Hang Yin
- Xuzichao Li
- Heng Zhang
Nature Communications (2024)
Structural basis of antiphage immunity generated by a prokaryotic Argonaute-associated SPARSA system
- Xiangkai Zhen
- Xiaolong Xu
- Songying Ouyang
Nature Communications (2024)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.