Functional insights into Solo-Cas4 in Methanosarcina mazei Gö1

Abstract

Solo-Cas4 homologs are Cas4-family proteins found outside of canonical CRISPR–Cas operons. Here, we present the biochemical characterization of Solo-Cas4 from Methanosarcina mazei Gö1. We found significantly upregulated solo-cas4 transcript levels during stationary phase, while remaining constant under oxygen exposure, temperature shifts, high salt conditions or virus challenge. Heterologously expressed as a SUMO-fusion, the purified tag-free protein displays an absorption peak at 420 nm, indicative of a [4Fe–4S]-cluster​. Size-exclusion-chromatography revealed that Solo-Cas4 forms a higher oligomeric complex, with an apparent molecular mass of 318 kDa. In vitro nuclease activity assays demonstrated that Solo-Cas4 cleaves metal-dependent linear dsDNA, with highest cleavage activity in the presence of Mn²⁺, followed by Mg²⁺, while Ca²⁺ and Cu²⁺ result in negligible cleavage. Isoleucine169 was identified to be crucial for catalysis, mutating it to alanine completely abolished nuclease activity​. Mutating any of the four conserved cysteines—proposed to coordinate the [4Fe–4S]-cluster did not affect nuclease activity; however, it abolishes metal cluster binding. Supercoiled circular dsDNA was preferentially nicked by Solo-Cas4 in the presence of Mg²⁺, whereas Mn²⁺ also led to linearization followed by complete degradation. Besides, ssDNA was cleaved by Solo-Cas4 but with lower activity. In agreement, Microscale thermophoresis analysis revealed strong dsDNA binding with highest affinity to supercoiled circular DNA, and weak ssDNA binding. Overall, these findings indicate that M. mazei Solo-Cas4 is a high oligomeric Cas4-family nuclease that preferentially targets supercoiled dsDNA and is upregulated during stationary growth.

Characterization of Solo-Cas4 from Methanosarcina mazei Gö1. The protein assembles into a high-molecular-weight oligomeric complex containing a Fe–S cluster and exhibits metal-dependent nuclease activity. Solo-Cas4 preferentially cleaves supercoiled double-stranded DNA, with expression upregulated during stationary phase.

Introduction

Prokaryotes hold CRISPR (Clustered Regularly Interspaced Short Palindromic Repeat) arrays and associated (Cas) proteins as an adaptive immune system against foreign genetic elements (Yang et al. 2024, Huang et al. 2025). In the process of CRISPR-Cas immunity, short fragments of invader DNA (protospacers) are captured and integrated into the host genome within the CRISPR array, providing a genetic memory of past infections​ (adaptation phase) (Mojica et al. 2005, Bolotin et al. 2005, Pourcel et al. 2005). Upon subsequent invasion, respective transcribed and matured CRISPR RNAs guide Cas nuclease complexes to recognize and destroy the matching invader nucleic acids (interference phase) (recently reviewed in Hille and Charpentier 2016).

Cas4 genes occur in approximately 90% of archaeal and 20% of bacterial genomes (Mojica et al. 2009, Hudaiberdiev et al. 2017). As part of CRISPR-Cas system subtypes I, II, and V, Cas4 proteins frequently cooperate with Cas1-Cas2 to process prespacers for integration, ensuring the exclusion of the protospacer adjacent motif (PAM) from the CRISPR array and ensure spacers are integrated in correct orientation and length (Shiimori et al. 2018, Zhang et al. 2019, Koonin and Makarova 2019). Moreover, Cas4 has been shown to interact directly with the Cas1-Cas2 complex (Lee et al. 2019, Hu et al. 2021), trimming prespacers before integration, as demonstrated in Alkalihalobacillus halodurans (previously Bacillus halodurans) (Lee et al. 2018, Dhingra et al. 2022), S. solfataricus (Rollie et al. 2018) and Synechocystis sp. 6803 (Kieper et al. 2018). Biochemically, Cas4 proteins belong to the RecB-family of nucleases and typically possess a conserved array of cysteine residues for binding an iron–sulfur (Fe-S) cluster (Zhang et al. 2012). Besides, Cas4 proteins exhibit low sequence similarity and are subdivided into DUF83 and DUF911 classes. DUF83-type enzymes degrade ssDNA in a 5′ to 3′ direction, whereas DUF911 members can process substrates bidirectionally (Kaushik et al. 2018).

In addition to CRISPR-encoded Cas4 (often termed CAS-Cas4), so-called solo-cas4 genes are located outside the CRISPR locus and are found in about 12% of archaeal and 4% of bacterial genomes (Hudaiberdiev et al. 2017), broadening the functional scope of Cas4 beyond canonical CRISPR-mediated defence. Solo-Cas4 were long considered to be a less functional or even an “orphan” version of Cas4 and the physiological function is largely uncharacterized. Phylogenomic analysis have revealed that solo-cas4 genes are often present in organisms that lack any other CRISPR-Cas components (Hudaiberdiev et al. 2017, Koonin and Makarova 2019). The high frequency of Solo-Cas4 raises the possibility that they might have a more general function in genomic defence or DNA repair independent of the canonical CRISPR pathway. Moreover, solo-cas4 homologs have been detected on mobile genetic elements such as plasmids and viruses (Hooton and Connerton 2014). The widespread occurrence of Solo-Cas4 in viral genomes suggests these proteins might confer an advantage to the mobile element, possibly by interfering with host defence systems. Hooton and Connerton (2014) hypothesized that a phage-encoded Cas4 could distort spacer acquisition, potentially leading to the incorporation of host DNA fragments into the CRISPR array and resulting in autoimmunity. Such a strategy could benefit the phage by incapacitating the host’s CRISPR immunity. The evolutionary dynamics of Solo-Cas4 appear to be distinct from other Cas proteins: the cas1-cas2-cas4 adaptation module is often horizontally transferred as a unit in CRISPR-containing organisms​, whereas solo-cas4 genes tend to be inherited vertically and are under normal selection pressures similar to housekeeping genes (Hudaiberdiev et al. 2017)​. This lack of strong diversifying pressure in Solo-Cas4 further suggests that they might not be engaged in arms races with viruses, but rather serve more cellular housekeeping roles.

Methanosarcina mazei strain Gö1 provides an exciting archaeal model system to investigate protein functions, since it is genetically tractable (Ehlers et al. 2005, Hüttermann and Schmitz 2024). It is a mesophilic methanogenic archaeon containing two CRISPR-Cas systems (subtype I-B and III-B) encoded on its genome (Nickel et al. 2013). In addition to CRISPR locus encoded cas genes, a solo-cas4 gene is encoded unrelated to both CRISPR-Cas systems. Here, we present a comprehensive characterization of the Solo-Cas4 from M. mazei. Besides transcription analysis of solo-cas4 gene in various growth phases and under stress conditions, the protein was overexpressed in M. mazei to identify regulatory mechanisms and evaluate its impact on growth. In addition, biochemical characterization of the purified Solo-Cas4 and mutant derivatives were performed concerning the presence of an Fe-S cluster and nuclease activity using different DNA substrates.

Materials and methods

Growth, RNA preparation and qRT-PCR analysis

M. mazei was cultivated in sealed bottles of an anaerobic complex media with a gaseous phase consisting of N₂ and CO₂ (vol/vol, 80/20) (Ehlers et al. 2002, Deppenmeier et al. 2002). The media was supplemented with 150 mM methanol as carbon source. Cells were cultivated at 37°C and the optical density at 600 nm (OD₆₀₀) was measured at indicated time points.

Four biological replicates of 25 mL M. mazei cell cultures were prepared for each condition. and grown at 37°C and harvested mid exponential phase (OD₆₀₀ 0.6) unless stated otherwise. For standard conditions cultures were grown in minimal media with 150 mM methanol as carbon. For temperature stress, the cultures were incubated at the indicated temperature (30° or 42°C). For oxygen stress cultures were grown under standard conditions to an OD₆₀₀ of 0.25 and subsequent 20 mL filtered air was added and cells were incubated for 30 min upon harvesting. For high salt stress, cultures were grown in media containing 0,5 M NaCl. For nitrogen limiting conditions, cells were grown in nitrogen free media to OD₆₀₀ of 0.25. For H₂/CO₂ cultures, H₂/CO₂ was used as sole carbon source. For stationary phase stress, cultures were grown to stationary phase (OD₆₀₀ 1). For virus stress cultures were grown under standard conditions to early exponential phase with 6.3 × 10⁷ cells/mL (OD₆₀₀ 0.25) before infecting with 1.35 ± 0.45 × 10⁹ viruses/mL. The virus titer was estimated using the method described by Gehlert et al. (Gehlert et al. 2023). Cultures were harvested before infection and 10 min, 30 min and 60 min after infection.

For quantitative reverse transcription (qRT)-PCR analysis complete cultures were incubated on ice for 5 min, harvested at 4 000 x g at 4°C for 30 min and cell pellets resuspended in ROTI®Zol (Carl Roth, Karlsruhe, Germany). Total RNA was isolated by chloroform extraction and was followed by DNase I treatment and phenol-chloroform precipitation (see Veit et al. 2005). qRT-PCR was performed using the QuantiTect SYBR Green RT-PCR Kit (Qiagen, Cat. No. 204 243) according to manufacturer’s instructions and as described before (Veit et al. 2005). 40 ng/µL total RNA from M. mazei cultures, primers (listed in Table 1) and the ViiA 7 Real-Time PCR System from Applied Biosystems (Thermo Fisher Scientific, Darmstadt, Germany) were used. Ct values were normalized in respect to the corresponding Ct values obtained from the same RNA for three housekeeping genes (MM_1215, MM_1621, and MM_2181) (Veit et al. 2005, Veit et al. 2006). The fold change was calculated using the formula fold change equals 2 ^−∆∆Ct, as described before (Talaat et al. 2004).

Table 1.

qRT PCR Primer set in direction 5´ to 3´.

qRT_PCR_ RS04560_for	GTGATGCCCACAGATTCACAGA
qRT_PCR_ RS04560_rev	TTTACATGCATAAGTGAGAGCAGCTT
qRT_1215_for	TCAAGAGCGAGGGCATGAATG
qRT_1215_rev	GCACTACCGAGAACAATAGCC
qRT_1621_for	TAGGAGGTTTTCTCGGAAGCG
qRT_1621_rev	AAGCGTATCTCCATCAAGCCC
qRT_2181_for	GCCTCCATGAGAAGAATGCTC
qRT_2181_rev	CTTCAAGGTCTCCAACTCCTG

Construction of plasmids

Cloning of MM_RS04560 under the control of the constitutive promotor pmcrB chrom. M. mazei DNA was used as a template for amplifying MM_RS04560 with the primer pair MM_RS04560_NdeI_for/MM_RS04560_BamHI_rev (Table 2). The addition of the restriction sites NdeI and BamHI of the PCR product were introduced by the primers and the plasmid pRS2150 were used for ligation, resulting in plasmid pRS2197. The plasmid pRS2150 is a derivate of pRS1807 (Hüttermann and Schmitz 2024), modified to include a Twin-Strep-Tag, resulting in a 35.6 kDa fusion protein. The transformation of pRS2197 DNA into M. mazei 3A was conducted through liposome-mediated transformation as described by Ehlers et al. (Ehlers et al. 2005) and modified by Gehlert et al. (Gehlert et al. 2023). Western blot analysis with a specific anti-Solo-Cas4 antibody generated by David´s Biotechnology (Regensburg, Germany) was used for expression verification (Fig. S1). For cloning solo-cas4 in fusion to SUMO the respective ORF (MM_RS04560, codon optimized) was cloned into pET SUMO with a T7, lacO Promotor. First a codon optimized Solo-Cas4 was synthesized and cloned into pEX-A128 (Eurofins Genomics, Ebersberg, Germany). The plasmid was named pRS1883 and transformed into E. coli DH5α. MM_RS04560 was cloned into pET SUMO using pRS1883 as a template, primer pair MM_RS04560_opti_for/MM_RS04560_opti_rev (Table 2), and the Champion pET SUMO expression system (Thermo Fisher Scientific, Waltham, USA) according to the manufacturer’s instructions, yielding plasmid pRS1884 and strain E. coli BL21 Δiscr K5014. Plasmid DNA was introduced into E. coli DH5α or BL21 Δiscr using the method described by Inoue et al. (Inoue et al. 1990).

Table 2.

Primer used for cloning set in direction 5´ to 3´.

MM_RS04560_NdeI_for	TTTCATATGATGCCCACAGATTCAC
MM_RS04560_BamHI_rev	TTTGGATCCAATATGAAAAACTTGGAC
MM_RS04560_opti_for	ATGATGCCGACCGATAGCC
MM_RS04560_opti_rev	TTAAAAAAATTTGCTCATCAGGCTGC

Protein expression and purification

E. coli Bl21 (DE3) Δiscr cultures were grown in LB (Luria-Bertani) medium at 37°C under continuous shaking. At turbidity at 600 nm OD₆₀₀ = 0.4, the culture was cooled down and protein expression was induced by adding 1 µM IPTG (isopropyl-β-thiogalactoside) (final concentration) and the cultures were incubated overnight at 18°C. Cells were harvested by centrifugation at 4 000 x g at 4°C for 20 min, suspended in 6 mL of HEPES buffer A (50 mM HEPES, 1 mM DTT, 5% glycerol, pH 7.5 (Chemicals from Carl Roth, GmbH + Co. KG, Karlsruhe, Germany)) and lysed by passing through a French pressure cell two times with 80 N (mm²)⁻¹. Following, the extract was centrifuged for 30 min at 13,000 g and 4°C to get rid of cell debris and the remaining unlysed cells. His₆SUMO-tagged Solo-Cas4 was purified by affinity chromatography on Ni-NTA agarose (Qiagen, Hilden, Germany) gravity flow columns with 1 mL of bed volume 3×5 mL buffer containing 20 mM imidazole and eluted with 100 mM imidazole (SERVA, Heidelberg, Deutschland).

SUMO cleavage

SUMO-protease (Thermo Fisher Scientific, Waltham, Massachusetts) was used according to the manufacturer’s protocol to cleave the His₆-SUMO-Solo-Cas4 for 30 min at 4°C and obtain untagged Solo-Cas4 by passing through the Ni-NTA-column after the cleavage. Elution fractions of protein purifications were analyzed on 12% SDS-PAGE gels and the protein concentrations were determined by Bradford (Bio-Rad Laboratories, Hercules, California) or Qubit protein assay (Thermo Fisher Scientific, Waltham, Massachusetts).

Size exclusion chromatography (SEC) was conducted with 0.3 mg of untagged Solo-Cas4, purified from heterologous expression in E. coli using 50 mM HEPES buffer (1 mM DTT, 5% Glycerol, pH 7.5) and the analytical SEC column ENrich SEC 650 (BioRad, Hercules, USA). The protein was eluted at a flow rate of 1 mL min⁻¹ with 50 mM HEPES buffer (1 mM DTT, 5% Glycerol, pH 7.5). Elutions were collected in 1 mL fractions. In order to calibrate the chromatograph, a protein mix (BioRad size-exclusion standard; #151–1901, BioRad, Hercules, USA) was used as a standard (see calibration inset in Fig. 4A).

Figure 4.

Oligomeric state of M. mazei Solo-Cas4 analyzed by SEC (A) and native PAGE (B). (A) Elution profile of purified Solo-Cas4 on EnRich650 column (BioRad, Hercules, USA) with absorbance at 280 nm vs elution volume. The inset shows the calibration using a mix of standard proteins (Size-exclusion standard, BioRad, Hercules, USA). The main peak of Solo-Cas4 corresponds to an estimated native mass of approximately 318 kDa, indicating a decameric assembly. (B) 12% SDS PAGE of elution fraction 9-11 mL. (C) Coomassie-stained native PAGE of purified Solo-Cas4 showing a high-molecular-weight band consistent with a higher oligomer. Marker SERVA Marker Mix for Blue/Clear Native PAGE shown in kDa.

Site directed mutagenesis

For constructing mutant proteins, the primer pairs listed in Table 3 were constructed with an additional 5´ Phosphate (Eurofins Genomics, Ebersberg, Germany) and the template pRS1884 (solo-cas4 wildtype) was used for exchanging amino acids.

Table 3.

Site directed mutagenesis primer set in direction 5´ to 3´.

RS04560_C22A_for	CCCGCGCCGCGTGTA
RS04560_C22A_rev	GCTTAATATACAGCAGCAGATCGCT
RS04560_E102A_for	CGAAGGCGATGCGCGC
RS04560_E102A_rev	GCCAGAAAGCCTTTTTCGGTG
RS04560_N150A_for	CCTGAAAGGCAGCCCGA
RS04560_N150A_rev	GCCAGGCGTTTGCTAAAAATAAT
RS04560_S156A_for	CAAAATTGTGTGCTTTGAAGGCGC
RS04560_S156A_rev	GCCGGGCTGCCTTTCAG
RS04560_I169A_for	GCGCCATTCGCCCGGG
RS04560_I169A_rev	TCGGCACGTTCGCGCCTT
RS04560_Y210A_for	CGTGAGCTTTGGCATTATTCG
RS04560_Y210A_rev	GCTTCCACAAAGCCAAACG

Nuclease and binding assays

The reaction mixture contained 20 mM HEPES (pH 7.5), 1 mM DTT, 2 mM of divalent metal (various, as indicated) and 2 mM ATP as indicated. For dsDNA assays the substrate was produced via PCR using ATTO647N-labeled forward primer pCRII_topo_label_for ([ATTO647N] CAAAGAGAAAGCAGGTAGCTTGC), unlabeled reverse primer pCRII_topo_rev (GCCCAATAGCAGCCAGTCCCTTCCCGCTTC) and pCRII TOPO (Thermo Fisher Scientific, Waltham, USA) as template, yielding a 500 bp ATTO647N-labeled PCR fragment. 5 nM ATTO647N-labeled dsDNA PCR fragment was incubated with 15 µM tagless Solo-Cas4 (or increasing concentrations of 1, 3, 5, 15 µM). The reaction products were separated on a 5% native polyacrylamide gel and analyzed using ChemDoc (BioRad, Hercules, USA). To confirm the absence of reaction products for the mutant protein, an additional reaction was subjected to Proteinase K digestion (final concentration 1 mg/ml, 30 min at 60 °C) prior to gel loading in case of experiments for Fig. 6C. For ssDNA assays 50 nM ATTO647N-labeled ssDNA (either 5´or 3´as indicated, Table 4) (Eurofins, Ebersberg, Germany) was incubated for 30 min at 37°C with purified tagless Solo-Cas4 and exclusively 2 mM MnCl₂. Prior to incubation ssDNA substrates were heated to 95°C and cooled down fast to prevent secondary structures. The reaction products were separated and analyzed by 20% denaturing PAGE containing 7 M Urea and analyzed using ChemDoc (BioRad, Hercules, USA). For circular dsDNA assays, 12 nM plasmid DNA (3.9 kb, pCRII TOPO (Thermo Fisher Scientific, Waltham, USA)) was incubated with 6.5 µM tag-less Solo-Cas4 and 5 mM of MnCl₂ or MgCl₂ as indicated. The reaction products were separated using 1% agarose gel, stained using ethidium bromide and imaged (Decon Science Tec GmbH, Hohengandern, Germany).

Figure 6.

Metal-dependent nuclease activity and substrate specificity of Solo-Cas4. (A) Cleavage of [ATTO647N]-labeled 500 bp linear dsDNA (37 °C, 15 min) by 15 µM purified Solo-Cas4 in the presence of 5 mM various divalent metal ions, with or without addition of 2 mM ATP as indicated. A control reaction with 15 µM of the mutant protein (I169A) and a no-protein control (lane 1) are included. (B) Cleavage assay using [ATTO647N]-labeled 500 bp linear dsDNA and increasing concentrations of Solo-Cas4 (1–15 µM) in the presence of 5 mM Mn²⁺. Mutant I169A (15 µM) and a no-protein control (lane c) served as controls. (C) Cleavage of of [ATTO647N]-labeled 500 bp linear dsDNA (37 °C, 15 min) by 15 µM purified Solo-Cas4 in the presence of 5 mM Mn2+ as indicated. Mutant I169A (15 µM) and a no-protein control (lane c) served as controls. The reaction products were subjected to Proteinase K treatment (1 mg/ml, 30 min at 60 °C) and analyzed on a 5 % native polyacrylamide gel. (D) Analysis of nuclease activity of various Solo-Cas4 mutant proteins (15 µM each) compared to wild-type protein and a no-protein control (lane c). All reactions were performed with 5 mM Mn²⁺, 37 °C for 15 min. Reaction products in panels A–C were analyzed via 5% native PAGE. DNA size markers (in bp) were obtained from Jena Bioscience (Thuringia, Germany). (E) Cleavage of 3′- or 5′-[ATTO647N]-labeled ssDNA oligonucleotides (41 nt) by 20 µM Solo-Cas4 in the presence of 5 mM Mn2+ at 37 °C for 30 min. Reactions were analyzed on a 20% denaturing polyacrylamide gel (7 M urea). Cye3 and [ATTO647N]-labeled oligonucleotides of different length (Eurofins, Ebersberg, Germany) were used as a marker and are shown of which the length is indicated in nucleotides (nt). (F) Nuclease and nickase activity of 15 µM Solo-Cas4 on circular 3.9 kb dsDNA in the presence or absence of 5 mM divalent metal ions (Mg²⁺ or Mn²⁺). Samples were incubated at 37 °C for the indicated times (0, 5, 15, 60 min) and analyzed by 1% agarose gel electrophoresis. Plasmid forms (supercoiled, nicked, linear) are indicated. DNA size marker (kb) from Thermo Fisher Scientific (Waltham, USA).

Table 4.

Fluorescently labeled substrates.

Oligonucleotide	Sequence (5´-3´)	Figure	Source
PAM1A5	[ATTO647N] CCAGAACTAGCTTATAGTTTAGAAGAAAACAAACAAATAAT	Fig. 6E, ssDNA 5´	(Lemak et al. 2013)
PAM1C5_3	CCCGAACTAGCTTATAGTTTAGAAGAAAACAAACAAATAAT[ATTO647N]	Fig. 6E, ssDNA 3´	(Lemak et al. 2013)

Microscale thermophoresis (MST) assays

SUMO cleaved Solo-Cas4 was labeled with the RED-NHS, 2nd generation, 650 nm fluorescent dye using the Monolith NT RED-NHS lysine labeling kit according to manufacturer’s protocol (NanoTemper, Munich, Germany). RED-labeled untagged Solo-Cas4 at 50 nM and 500 bp dsDNA at 16 different concentrations ranging from 0.0253 nM to 0.83 µM or 25 nM of Solo-Cas4 -RED and 41 nt ssDNA in 16 dilutions ranging from 1.53 nM to 50 µM (all concentrations based on monomeric molecular mass) were used. The interactions were measured in standard capillaries (NanoTemper), 100% excitation power, and medium or high MST power (IR laser intensity). For each substrate two biological replicates were analyzed using MO.Affinity Analysis Software (NanoTemper).

Iron–sulfur cluster modeling

The iron-sulfur cluster in Solo-Cas4 was modeled based on a related protein that contains a 4Fe-4S cluster. The AlphaFold3 model of Solo-Cas4 (AF-A0A0E3PT89) was overlaid with the structure of a protein (Protein Sso0001, PDB ID: 4IC1), which contains a 4Fe-4S cluster, in order to identify potential ligand positions. Only the cysteine residues, predicted to coordinate the metal ions in the cluster, were aligned between the two structures. This approach was used to model the placement of the 4Fe-4S cluster in Solo-Cas4 based on the position of the cysteines in the related protein. To focus on the cluster in Solo-Cas4, the structure of the other protein was made invisible in the visualization. Although AlphaFold3 does not explicitly predict the coordination of metal ions by cysteines, this method provides a reasonable approximation based on the structural similarities in the position of the cysteine residues. The cluster placement was visualized and annotated using ChimeraX.

Results

solo-cas4 expression is induced in stationary growth phase but not in response to stress conditions applied during exponential growth

To determine Solo-Cas4 expression in M. mazei, we quantified solo-cas4 transcript levels during growth and under stress conditions using quantitative reverse transcriptase PCR (qRT-PCR). Cultures were grown under anaerobic conditions to mid-exponential and stationary growth phase, as well as subjected to specific stresses in mid exponential growth (oxygen exposure, temperature shifts (30 and 42°C), increased salinity (0.5 M NaCl), and H₂/CO₂ as sole carbon source as well as under nitrogen limitation (data not shown). solo-cas4 transcript levels were determined normalized to housekeeping genes. We observed a significant transcriptionally upregulation of solo-cas4 when cells entered stationary phase. Approximately nine-fold higher solo-cas4 transcript levels were detected compared to exponentially growing cells (Fig. 1A). In contrast, solo-cas4 transcription levels remained unchanged under various stress conditions applied (Fig. 1A) including nitrogen limitation. To evaluate the expression level of solo-cas4 during virus infection, cultures were infected with the Methanosarcina Spherical Virus (MetSV) (Weidenbach et al. 2017) and the solo-cas4 transcription levels compared and normalized to cultures before infection. This showed that challenging M. mazei with MetSV did not change the solo-cas4 transcript level at all (Fig. 1B). These results indicate that solo-cas4 is specifically induced during stationary phase or under respective conditions, rather than being involved in general, nutritional or viral stress-response, where specific stress related gene clusters are expressed (Wong et al. 2017).

Figure 1.

qRT-PCR analysis of solo-cas4 expression. Relative fold-regulation of solo-cas4 transcript levels were determined in (A) exponentially grown cells under standard growth conditions vs. various stress conditions (oxygen, 30°C, 42°C, stationary phase and high NaCl (0.5 M NaCl)), and (B) 10 min, 30 min and 60 min after virus infection vs. before challenge. Transcript levels are normalized to an internal reference and expressed relative to the exponential phase value using the formula fold-change equals 2^−∆∆Ct as described before (Talaat et al. 2004). Error bars indicate standard deviation from three biological replicates. An unpaired two-sided t-test shows a significant difference (P < 0.05) between stationary phase and exponential phase with a fold-change of 9.

In vivo overexpression of Solo-Cas4 results in reduced growth

To gain a deeper understanding of the function of Solo-Cas4 in vivo, a genetic approach was employed by generating a M. mazei strain that overexpresses the protein from a plasmid under a constitutive promoter (pRS2197). The overexpression of Solo-Cas4 in this mutant strain was verified via western blot analysis showing expression in exponential and stationary phase (Supplemental Fig. S1B, lanes 4–6). In contrast, chromosomal expressed Solo-Cas4 was not detectable by western blot in the late exponential or stationary phase indicating very small amounts of protein under native conditions (Supplemental Fig. S1B, lanes 1–3). When grown under standard conditions with methanol as sole energy and carbon source, overexpression of Solo-Cas4 not only resulted in slower growth (Fig. 2, orange) but also the strain failed to reach the same optical turbidity as the control strain carrying an empty vector (M. mazei/pRS1595, green) (Fig. 2). These observations suggest that additional high levels of Solo-Cas4 adversely affect cellular fitness or replication, potentially disrupting normal growth dynamics.

Figure 2.

Growth analysis of M. mazei Solo-Cas4 overproducing strain in comparison to M. mazei wt. Growth of the M. mazei Solo-Cas4 overproducing strain (orange), and the empty vector control strain (green), grown under standard conditions on 150 mM methanol. In each case, a 50 mL anaerobic medium was inoculated with 1.6 × 10⁸ cells (T₀). The standard deviation of three biological replicates are shown.

Solo-Cas4 binds an [4Fe–4S] cluster and has no RecB motif

For heterologous expression and purification, Solo-Cas4 from M. mazei was fused to a N-terminal His₆SUMO-tag and expressed in E. coli BL21 (DE3, ∆iscr). After purification under aerobic conditions by Ni-NTA affinity chromatography, the SUMO-tag was cleaved and separated yielding a highly pure, untagged protein fraction, as demonstrated by SDS PAGE, suitable for subsequent biochemical analyses (Fig. 3A, Supplemental Fig. S2). Purified Solo-Cas4 protein fractions exhibited a brown color, suggesting the presence of an iron-sulfur cluster (Fig. 3B, inset). This observation was further supported by the UV–visible absorption spectrum of purified Solo-Cas4, which revealed a distinct peak at approximately 420 nm (Fig. 3B). This spectral signature is characteristic for an [4Fe–4S] iron–sulfur cluster, which typically peaks around 410–420 nm in proteins (Pieck et al. 2006, Lemak et al. 2013). Consequently, we conclude that Solo-Cas4 binds a [4Fe–4S] cluster, which is likely coordinated by the four identified cysteine residues, which are conserved in Cas4 proteins (Figs. 3C, 5A). The respective structure predicted by AlphaFold3 (Abramson et al. 2024) is depicted in Figs. 3C and 5B, including a close-up of the predicted [4Fe-4S] binding site and its coordinating cysteines in Fig. 3C. Sequence comparison of Solo-Cas4 with other Cas4 homologs revealed a conserved iron staple motif and the absence of the RecB motif (Fig. 5A, Supplemental Fig. S3). Although the primary sequence shows limited similarity to other Cas4 protein, structural comparison revealed a remarkable resemblance between Sso1392 from S. solfataricus and M. mazei Solo-Cas4 (Fig. 5B), suggesting an active site 22 Å apart from the 4Fe-4S cluster with different amino acids.

Figure 3.

Identification of an Fe–4S in purified Solo-Cas4. (A) 12% SDS PAGE of 112 µg purified Solo-Cas4. Marker shown in kDa (LMW marker (BioRad, Hercules, USA). (B) UV–visible absorption spectrum of aerobe purified Solo-Cas4 observed a characteristic peak for an [4Fe-4S] cluster at ∼420 nm. Inset: photograph of the purified Solo-Cas4 fraction, displaying a brown color typical for Fe–S proteins (right), compared to buffer (left). (C) AlphaFold3-based structure of Solo-Cas4 with an overlay of [4Fe–4S] coordinating cysteines of Sso0001 (PDB ID: 4IC1) (green) and Solo-Cas4 (yellow).

Figure 5.

Sequence and structural comparison. (A) Schematic multiple sequence alignment with conserved cysteines (red), RecB motif (blue) and length of the protein (left). (B) Structural comparison of AlphaFold3-predicted Solo-Cas4 (green) and Sso1392 (pink). Zoom-in highlights the RecB motif in Sso1392 (pink) and the corresponding amino acids in Solo-Cas4 (green), distance of 22.5 Å between 4Fe-4S cluster (yellow) and manganese (purple) measured with PyMol.

Solo-Cas4 shows a high oligomeric conformation

We further investigated the quaternary structure of the purified Solo-Cas4 protein in solution by size-exclusion chromatography (SEC) using a Superdex EnRich 650 column. The protein eluted in a single peak at a volume corresponding to an apparent molecular mass of approximately 318 kDa (Fig. 4A, B). Given that the molecular mass calculated from the amino acid sequence of the Solo-Cas4 monomer is approximately 30.7 kDa, the SEC result indicates that Solo-Cas4 forms a large oligomer in solution, most likely a decamer. Decameric structures of Cas4 proteins have been observed before (Lemak et al. 2013). To further validate the oligomerization, we performed native PAGE, the main band between 146–450 kDa corresponds to the main oligomeric form of Solo-Cas4. A faint band migrating near the 720 kDa marker likely represents higher-order oligomers or protein aggregates formed under native PAGE conditions (Fig. 4C).

Solo-Cas4 possesses a metal dependent nuclease activity

We assessed the nuclease activity of Solo-Cas4 on various DNA substrates to determine its nuclease properties and substrate specificity. Using purified and tag-less Solo-Cas4, we performed in vitro cleavage assays using three types of DNA: (i) a 500-bp linear dsDNA fragment (Fig. 6A-D), (ii) a 41-nucleotide ssDNA oligonucleotide (Fig. 6E), and (iii) supercoiled circular plasmid DNA (∼3.9 kb) (Fig. 6F). The ssDNA as well as the dsDNA fragment were fluorescently labelled (fluorophore ATTO647N, Eurofins, Ebersberg, Germany). For the dsDNA fragment, reactions were carried out in the presence of 5 mM divalent cations (MgCl₂, MnCl₂, CaCl₂, or CuCl₂) at 37°C for 15 min and fluorescently labelled products were analyzed by native PAGE (5%). ssDNA was incubated with 5 mM MnCl₂ and separated using denaturing PAGE (20%, 7 M Urea). Fluorescently labelled reactions products were analyzed using image reader ChemDoc (BioRad, Hercules, USA). Circular plasmid DNA was incubated with 5 mM of either MgCl₂ or MnCl₂ and analyzed on 1% agarose gel. The results show that Solo-Cas4 exhibits robust metal ion dependent nuclease activity on linear dsDNA (Fig. 6A-C) and supercoiled DNA (Fig. 6F), and less activity on ssDNA as a subset of unprocessed substrate remains even after 30 min of incubation (Fig. 6E). With dsDNA as substrate Solo-Cas4 showed maximal activity in the presence of MnCl₂, followed by MgCl₂ (Fig. 6A, lanes 4 and 7). CaCl₂ could only substitute for MnCl₂ to some extent, yielding detectable dsDNA cleavage, though to much lower amounts indicating significant lower activity (Fig. 6A, lane 6). In contrast, the presence of CuCl₂ did not support nuclease activity at all, but rather inhibited it. Apparently Solo-Cas4 binding to the dsDNA was even affected in the presence of CuCl₂, since shifting of the dsDNA due to binding was not obtained (Fig. 6A, lane 5). Incubating in the absence of divalent metal ion completely abolished the nuclease activity and only resulted in shifting the dsDNA (Fig. 6A, lane 3), confirming that a divalent metal ion, with a strict octahedral coordination geometry as found in Mg²⁺ and Mn²⁺ is crucially required for catalysis. ATP, which is required for other RecB-domain containing proteins, in contrast had no effect on nuclease activity (Fig. 6A, lanes 8–11). Under optimal conditions, with MnCl₂, Solo-Cas4 completely cut the 500 bp dsDNA substrate within 15 min at 37°C starting at concentration higher than 1 µM, as demonstrated by the disappearance of the 500 bp fragment and the appearance of very short fragments (Fig. 6B). These results strongly indicate that Solo-Cas4 is a metalloenzyme, likely using Mn²⁺ as a cofactor for its nuclease active site to cut dsDNA into short fragments.

Several mutant proteins with single amino acid exchanges were generated by site-specific mutagenesis selected based on structural similarity to Sso1392 from S. solfataricus. The respective mutant proteins were purified following the purification protocol of the wildtype protein, yielding tag-less mutant derivatives of apparent homogeneity (Supplemental Fig. S4A). Nuclease assays with labeled linear dsDNA revealed that substituting Isoleucine169 to Alanine (I169A) resulted in a complete loss of nuclease activity but had no effect on binding activity (Fig. 6B, C, Supplemental Fig. S4B). In contrast, all other mutants generated (S156A, C22A, E102A, N150A and Y210A) are still active (Fig. 6D). The generation of an inactive mutant derivative (I169A), provided valuable insights into the active site, with the main chain carbonyl of Ile169 potentially binding the Mn²⁺ ion (Fig. 5B). Moreover, it also ruled out the possibility, that the activity determined for the purified wildtype protein is due to contaminating proteins. Thus, we can conclude with certainty, that the observed activity is due to Solo-Cas4.

We further analyzed the activity using a ssDNA of 41 nt. When 5′-labeled ssDNA was incubated with purified Solo-Cas4, the resulting labeled fragments are approximately 5 nucleotides in length or smaller. Using 3′-labeled ssDNA results in one labeled product of approximately 35 nucleotides besides not detectable smaller fragments suggesting that the cleavage occurs near the 5′ end (Fig. 6E). Overall compared to dsDNA, the nuclease activity appears to be much lower.

Finally, using supercoiled circular plasmid DNA as a substrate revealed distinct cofactor-dependent differences in cleavage patterns (Fig. 6F). In the presence of MnCl₂, Solo-Cas4 initially introduces a nick in the plasmid, subsequently linearizes it, and ultimately leads to its complete degradation within 60 min. In contrast, with MgCl₂, significant lower and preferentially nicking activity is observed while linearization only occurred after 60 min incubation (Fig. 6F). The absence of a divalent metal as well as ATP addition had no effect on activity (Supplemental Fig. S5). These findings highlight the importance of metal-ions impacting and modulating the Solo-Cas4 nuclease activity.

Binding of DNA substrates

The interaction between Solo-Cas4 and DNA substrates was further demonstrated and evaluated by microscale thermophoresis (MST) analysis using fluorescently labeled untagged Solo-Cas4 (RED- NHS) and (i) supercoiled plasmid DNA (∼3.9 kb) (Fig. 7A), (ii) the 500-bp linear dsDNA fragment (Fig. 7B), and (iii) the 41-nucleotide ssDNA oligonucleotide. The MST measurements revealed the strongest binding affinity between Solo-Cas4 and supercoiled circular plasmid DNA with an estimated dissociation constant of K_D = 1.6 ± 0.61 nM. dsDNA yielded a dissociation constant of 0.14 ± 0.6 µM, and therefore does not suggest binding activity against dsDNA substrate under tested conditions (Fig. 7 B). In contrast, MST measurements did not allow to determine a dissociation constant when using ssDNA as substrate, indicating very low binding affinity (Supplemental Fig. S6).

Figure 7.

Binding affinities of Solo-Cas4 to different DNA substrates using microscale thermophoresis (MST). 50 nM RED-labeled untagged Solo-Cas4 was incubated with (A) circular dsDNA (3.9 kb, pCRII (Thermo Fisher Scientific, Waltham, USA) at different concentrations ranging from 1.53 pM to 0.05 µM. Based on two biological replicates, the K_D was estimated to be 1.6 ± 0.61 nM. (B) dsDNA (500 bp) at different concentrations ranging from 25.3 pM to 0.83 µM. Based on two independent biological replicates, the K_D was estimated to be 0.14 ± 0.6 µM.

Discussion

Cas4 proteins exhibit high sequence variability, and their mechanisms of action and specific functions differ significantly across various CRISPR systems (Lemak et al. 2014). Cas4 homologs are widespread in archaea (∼90% of genomes encode Cas4) and even persist in CRISPR-free species, hinting at different cellular roles (Hudaiberdiev et al. 2017).

To date, four Cas4 proteins have been characterized biochemically all from archaeal systems: Sso0001 and Sso1391 from S. solfataricus, Pcal_0546 from P. calidifontis (PDB ID: 4R5Q) and TON_0321 from Thermococcus onnurineus. Here TON_0321 is the only protein that is encoded in connection with a CRIPSR-Cas system (CAS-Cas4) whereas the other three proteins are Solo-Cas4 enzymes. CAS-Cas4 proteins are known to play essential roles in CRISPR spacer acquisition and DNA end processing, where their activity is normally tightly coordinated with other Cas components to ensure precise DNA cleavage and repair upon PAM sequence recognition (Zhang et al. 2019). Interestingly, no sequence specificity could be obtained for all studied Cas4 proteins in vitro which indicates that a binding partner might be needed for sequence-dependent cleavage of DNA substrates. Cas4 family proteins share homology with the endonuclease domains present in the DNA repair complexes RecBCD, AddAB, and AdnAB (Amundsen and Smith 2019). They display little sequence similarity but share common core fold and a few critical active site residues (Knizewski et al. 2007). Cas4 proteins possess divalent metal ion-dependent nuclease activity which is coordinated by a RecB-like motif. Most of the characterized Cas4 proteins require Mn²⁺ and to some extent also Mg²⁺ for nuclease activity, most likely due to their octahedral coordination geometry. An exception is TON_0321 which shows activity even with Cu²⁺ and Ca²⁺ (Jain et al. 2024). M. mazei Solo-Cas4 cleaves linear ssDNA and dsDNA fragments, as well as circular DNA fragments, despite lacking a distinct RecB-like motif. In the presence of Mn²⁺, it is capable of completely degrading circular supercoiled dsDNA for which it shows also the highest affinity, implying that this substrate might be one major natural substrate in vivo. The Solo-Cas4 cofactor-dependent cleavage patterns observed on circular dsDNA, with Mn²⁺ promoting endonuclease activity, whereas with Mg²⁺ dsDNA is preferably nicked (Fig. 6F), are strikingly similar to the activity exhibited by E. coli DNase I. Under Mg²⁺ conditions, E. coli DNase I introduces random single-strand nicks, which convert supercoiled dsDNA into nicked circles. It can further utilize Mn²⁺ as a catalytic cofactor in place of Mg²⁺, and this substitution causes DNase I to linearize circular dsDNA (Campbell and Jackson 1980). Hence, intracellular metal availability seems to modulate Solo-Cas4 activity, highlighting its metal-triggered specificity.

At the molecular level, proper metal coordination is essential for nuclease activity. Cas4 enzymes generally utilize a mechanism based on divalent metal ions to initiate a nucleophilic attack on the phosphodiester bond (Kovall and Matthews 1997, Yang 2011, Zhang et al. 2011). Crystal structures have revealed that the main-chain carbonyl of a hydrophobic residue acts as one of the key metal ligands in the active site. This is evident in crystal structure of S. solfatarics Cas4 (Sso0001) that coordinates Mn²⁺ by His62, Asp99, Glu113, and the backbone carbonyl of Ile114, while P. calidifontis Cas4 (Pcal_0546) uses the carbonyl of Ile137 along with Asp123, Glu136, and His146 to chelate a metal ion (Lemak et al. 2013, Lemak et al. 2014). In the predicted structure of M. mazei Solo-Cas4, Ile169 might occupy a corresponding position and may contribute directly or indirectly to metal coordination (Fig. 5B). Although the backbone of Ile169 points in a different direction compared to Ile114 in Sso0001, deletion of this residue abolishes nuclease activity (Fig. 6 A-C; Suppl. Fig. S4 B), suggesting a critical functional role, either in catalysis or structural integrity.

It is important to note that the predicted distances between possible putative active site residues (Arg170; His186, Asp183, and Ile169) in Solo-Cas4 are somewhat larger (∼7.5–8.0 Å) than those observed in crystallized Cas4 proteins (e.g. ∼4.5–4.7 Å in Sso0001, Supplemental Fig. S7). However, as the Solo-Cas4 structure has not been experimentally determined, these distances must be interpreted with caution. Albeit AlphaFold predictions often match experimental maps remarkably closely, large differences on local and global scale between experimental maps and even very high-confidence predictions have been observed (Terwilliger et al. 2024). Side-chain flexibility and local rearrangements in vivo could permit the coordination of a divalent metal ion such as manganese or magnesium, both of which are required for the observed nuclease activity of Solo-Cas4. Furthermore, the predicted distance (∼22.5 Å) between the active site region and the iron–sulfur cluster in Solo-Cas4 closely matches the 20 Å separation found in the Sso0001 crystal structure (Lemak et al. 2014), supporting overall structural similarity. Nevertheless, the precise architecture of the active site remains unresolved. While Solo-Cas4 appears to adopt a RecB-like fold, it lacks the conserved RecB motif typically associated with catalytic activity in this protein family. Therefore, the exact location and composition of the active site remains speculative in the absence of crystallographic data.

We obtained strong evidence that Solo-Cas4 from M. mazei contains an [4Fe–4S] cluster. This Fe–S cluster, presumably coordinated by four conserved cysteine residues, is a hallmark of Cas4 proteins and related RecB-family nucleases (e.g. AddB and Dna2) (Zhou et al. 2015) and serves as an “iron-staple” that stabilizes the enzyme’s fold​ (Lemak et al. 2014, Zhou et al. 2015). However, in some Cas4 proteins, the cluster is closely tied to function. Mutating the coordinating cysteines in a S. solfataricus Cas-Cas4 abolishes its characteristic decameric assembly and weakens DNA binding and cleavage (Lemak et al. 2014). Other Cas4 homologs however, can retain activity without the cluster, implying that the [4Fe–4S] primarily reinforces structural integrity rather than directly catalyzing phosphodiester bond hydrolysis (Dixit et al. 2021). One further example is the archaeal Solo-Cas4 from P. calidifontis (Pcal_0546), where an [2Fe–2S] cluster was observed, the loss of which however did not abolish nuclease activity​ (Lemak et al. 2014). Thus, while Fe-S clusters are common in Cas4, their structural role and necessity for catalysis can vary. Consequently, our findings for M. mazei Solo-Cas4 are in line with previous reports that Cas4 family nucleases typically contain Fe–S clusters coordinated by conserved cysteine residues (Zhang et al. 2012). Moreover, the [4Fe–4S] cluster in M. mazei Solo-Cas4 appears not to serve as a catalytic cofactor, as mutating one of the cysteine residues produced a cluster free and thus colorless protein without any loss of activity. Besides a structural function, the [4Fe–4S] cluster may aid in DNA damage sensing, while the proper metal cofactor aligns catalytic residues for efficient hydrolysis. The homo-oligomeric state of Solo-Cas4 likely plays a crucial role in fine-tuning its catalytic efficiency and substrate interactions. S. solfataricus Cas-Cas4 assembles into a toroidal decamer with ten active sites facing a positively charged central channel (Lemak et al. 2014). This ring-shaped configuration likely facilitates cooperative DNA engagement and processive 5′-end trimming, analogous to ring nucleases like λ-exonuclease (Zhang et al. 2011). Hence, together, the [4Fe–4S] cluster, the cofactor (Mn²⁺ or Mg²⁺) and the multimeric assembly equip M. mazei Solo-Cas4 for its role in DNA processing.

The widespread occurrence of Solo-Cas4 proteins especially in organisms lacking CRISPR-Cas systems led to the suggestion that Cas4 may perform uncharacterized defense or repair functions​ (Hudaiberdiev et al. 2017). In Cyanobacteria Solo-Cas4 enzymes interact with Argonaut proteins where they interfere with plasmid DNA, arguing for a function as independent immune systems (Taton et al. 2023, Bobadilla Ugarte et al. 2025). Although M. mazei Solo-Cas4 also shows a preferential activity against plasmid DNA, M. mazei harbors no Argonaut proteins based on a DefenseFinder (Rousset et al. 2022) and PADLOC (Payne et al. 2022). Interestingly, expression of Solo-Cas4 was found to be moderately upregulated (∼1.6 log₂ fold change; Gehlert et al. 2022) at 180 minutes post-infection a time point corresponding to the late phase of virus replication cycle (assembly). Together with the observed growth retardation upon overexpressing Solo-Cas4 in M. mazei, this could superficially resemble an abortive infection-like response. However, since the host undergoes lysis at this stage, active transcription is likely limited, making a functional role in abortive infection unlikely. Notably, we also observe upregulation of Solo-Cas4 during the stationary phase (Fig. 1A), suggesting a broader role in stress or DNA metabolism rather than dedicated virus defense. Nevertheless, we cannot rule out the possibility that there are other, as yet unknown components involved which are working with Solo-Cas4 in a defense or repair related way in M. mazei.

Conclusion

Our study on the Solo-Cas4 protein from M. mazei provides novel insights into its biochemical and functional properties, clearly distinguishing it from its CRISPR-associated counterparts, highlighting the remarkable diversity in structure and functionality of Cas4 proteins. The Mn²⁺ dependent endonuclease activity with a preference for circular, supercoiled dsDNA may indicate a role in yet unknown defence mechanisms or DNA repair like it was shown for Cyanobacteria.

Author contributions

Conceptualization, R.A.S. and L.H.; methodology, R.A.S, L.R., L.H and S.S.; experimental investigation, L.R.; writing and editing, R.A.S., L.H. L.R. and S.S.; visualization, L.R. and R.A.S.; supervision, R.A.S.; funding acquisition, R.A.S. All authors have read and agreed to the published version of the manuscript.

Conflict of interest

None declared.

Funding

This research was funded by the Deutsche Forschungsgemeinschaft (DFG) within the priority program SPP2141 (SCHM1052/21–2 to R.A.S.).