Giant viruses (GVs) are a hotspot of unresolved controversies since their discovery, including the definition of “Virus” and their origin. While increasing knowledge of genome diversity has accumulated, GV functional genomics was largely neglected. Here, we describe an experimental framework to genetically modify nuclear GVs and their host Acanthamoeba castellanii using CRISPR/Cas9, shedding light on the evolution from small icosahedral viruses to amphora-shaped GVs. Ablation of the icosahedral major capsid protein in the phylogenetically-related mollivirus highlights a transition in virion shape and size. We additionally demonstrate the existence of a reduced core essential genome in pandoravirus, reminiscent of their proposed smaller ancestors. This proposed genetic expansion led to increased genome robustness, indicating selective pressures for adaptation to uncertain environments. Overall, we introduce new tools for manipulation of the unexplored genome of nuclear GVs and provide experimental evidence suggesting that viral gigantism has aroused as an emerging trait.
Giant viruses (GV) have genomes up to 2.8 megabases and form viral particles that can match in size some cellular organisms1,2. Their dsDNA genomes can encode more than 1000 genes, 70% of them corresponding to proteins unseen in any other organisms2,3. These genes are referred to as ORFans4. In addition, these atypical viruses encode proteins involved in translation and energy metabolism, hallmark functions of the cellular world5,6,7. These unprecedented features have led to great controversies regarding the origin of such viruses8. The origin of these viruses has been attributed either to reductive evolution branching from an ancestral cell8 or evolution towards complexity from smaller viruses9.
Particularly, pandoraviruses possess the current record in genome size of the viral world with P. salinus harboring a double-stranded DNA molecule of 2.77 Mbp1. The particles of these viruses are amphora-shaped and formed by a membranous compartment encapsulated by a tegument-like envelope made of several layers1,10,11, one of them composed of cellulose11. An apical pore allows delivery of the content of the particle by fusion of the internal membrane with the phagosome1. Upon delivery of the virion content, pandoravirus enters an eclipse phase of 2 to 4hs, followed by a dramatic host nucleus reorganization (including shape, content and loss of membrane integrity)1. Eight to ten hours post-infection (PI), host cells round up and new virions become visible in the infected cell1. Interestingly, the tegument of pandoravirus particles and the content of the virion seem to be synthesized/assembled simultaneously, this process being initiated at the apical pore1. Virions are released by cell lysis or exocytosis1,10.
Here, we developed a battery of genetic tools to study nuclear giant viruses and their host Acanthamoeba castellanii. In addition, we used this technology to tentatively trace back the evolution from a small icosahedral to the amphora-shaped pandoravirus GV12. We demonstrate the existence of an essential core region in pandoravirus, which corresponds with the accumulation of evolutionary conserved genes across their proposed smaller viral ancestors. Finally, we demonstrate the essentiality of the icosahedral Major Capsid Protein in mollivirus, a phylogenetically related virus and previously proposed evolutionary intermediate in virion shape9,13,14. Overall, our analysis demonstrates the potential of genetic screens to uncover conserved biological processes and host–pathogen interactions for the most genetically complex viruses discovered so far.
CRISPR/Cas9 expression allows efficient modification of host DNA
To achieve genetic manipulation of GVs, we built an Acanthamoeba castellani CRISPR/Cas9 system, using a plasmid encoding a polycistronic tRNA-gRNA15,16 and the SpCas9 gene fused to the green fluorescent protein (GFP)17 gene (Fig. 1a). To test the system, we conducted Cas9-mediated modification of an episomally encoded red fluorescent protein (mRFP) using the vector vc241 (Fig. S1a). When amoebas were transfected with Cas9 and specific guides targeting the mRFP gene (Fig. S1a), we observed a significant decrease in the mRFP positive population, which correlated with Cas9 expression (GFP positive) (Fig. 1b, c). The presence of double-negative amoebas is likely explained by the amitotic nuclear division of these organisms, leading to aneuploidy and the potential loss of genetic material18,19. Double-labeled amoebas were not observed when on-target guides were present (Fig. 1b–c). The locus targeted by the gRNA was amplified by PCR, cloned into a TA vector and single clones were isolated for sanger sequencing. Accordingly, genotyping confirmed gene locus modification in ~50% of the screened library (10 clones) (Fig. 1d and S1b). Importantly, the expression of Cas9 in presence or absence of off-target gRNA did not affect the doubling time of the amoeba (Fig S1c) or increased mortality rate (Fig S1d). To increase the efficiency of genome modification, we hypothesized that the introduction of a “mutagenic chain reaction” (Fig. S1e)20 would allow the propagation of recombinant genes. Concordantly, 100% reduction of mRFP expression was demonstrated when this strategy was utilized (Fig. 1b–c and S1f). Correct integration and homozygosis were validated by PCR (Fig. 1f). No recombinant cells were detected when off-target guides were transfected (Fig. 1b, c), indicating that homologous recombination (HR) is rather inefficient in absence of double strand breaks.
Importantly, the genome of A. castellanii is highly polyploid (aprox. 25n21, (Fig. S1e)), which could make genetic manipulation difficult. To assess the efficiency of Cas9 to modify the genome of A. castellanii, we designed gRNA targeting the cellulose synthase genes. A. castellanii encodes three possible genes for cellulose synthase (CS)22, two of them containing a full sequence of the CS monomer. Concordantly, we designed two gRNA targeting conserved regions in the two full-length CS (cs1 and cs2) and one in the shorter one (cs3) (Fig. 2a). Expression of Cas9 led to complete modification of the targeting sequencing as demonstrated by PCR using reverse primers annealing on the gRNA targeting sequences (Fig. 2a). These results indicate that each paralogue gene, present in approximately 25 copies21, was fully targeted by Cas9 and modified by NHEJ (around 125 modified sites). Moreover, the locus targeted by the gRNA was amplified by PCR, cloned into a TA vector and 10 single clones were isolated for Sanger sequencing. Genotyping of this screened library confirmed gene locus modification in all clones of cs1 (Fig. 2b). Importantly, due to the heterozygosis shown by the library screening and, since not every modification resulted in a frame shift for translation, full depletion of gene expression cannot be confirmed. While no changes in the doubling time of trophozoites were observed (Fig. S1f), a strong inhibition of encystment was induced upon knock-out/downregulation of CS, as demonstrated by SDS resistance (Fig. S1g) and calcofluor (CFW) staining of cysts (Fig. 2c, d). These data reproduced previous results obtained by RNA silencing23,24. This phenotype was rescued by the expression of a Cas9-targeted resistant cs1 gene fused with GFP (Fig. 2c, d and Fig S1h). Moreover, similar defects were observed at the ultrastructure level when analyzed by electron microscopy (Fig. S1j). These differences included the presence of opened pre-cysts with no cellular content, or cysts lacking a cyst endolayer, which is majorly composed of cellulose23,24.
Thus, while these experiments demonstrate the potential for the use of CRISPR/Cas9 to modify the highly complex genome of A. castellanii, future efforts must be directed into optimizing vectors and protocols to allow controllable modifications and/or homozygosis in the cells.
CRISPR/Cas9 expression allows efficient modification of viral DNA
Most giant viruses encode an RNA polymerase (RNAP), which modeled the RNAPII of modern eukaryotes25. Due to its likely essentiality for viral replication, we targeted the rpb1 gene to assess the capacity of Cas9 to target viral genomes (Fig. S2a). Nuclear Cas9 expression in amoeba cells (Fig. 3a) showed efficient regulation of Pandoravirus neocaledonia3 and Mollivirus kamchatka26 replication (Fig. 3a, b), while the cytosolic version of the protein (Fig. 3a) was unable to target the genome of Pithovirus sibericum27 or Mimivirus reunion28 (Fig. 3a, b). The incapability of Cas9 to target the genome of strictly cytoplasmic viruses might reside in the shielding of the DNA by these viruses, excluding GFP-Cas9 from the viral factory as shown by fluorescence (Fig. S2b). A similar strategy was previously demonstrated for some bacteriophages29. Coherently with the likely essentiality of the rnap, only virions containing wild-type genomes (which escaped Cas9 targeting) were recovered in genomic libraries (Fig. S2c).
The efficient targeting of rpb1 of pandoravirus was further confirmed by utilizing multiple gRNA, which rendered similar phenotypic consequences (Fig. 3c). Moreover, to confirm the depletion of the target proteins upon knock-out, we built a plasmid that allows endogenous tagging of targeted genes (Fig. 3d). This plasmid contains a triple HA tag or triple FLAG tag and a nourseothricin N-acetyl transferase (NAT) resistance cassette, which confers resistance to nourseothricin30. The NAT selection cassette was expressed under the control of a pandoravirus promoter (Fig. 3d). NAT selection was used to enrich the recombinant viruses, which were later cloned to obtain homogeneous populations (Fig. 3e and Fig. S2d). We opted to tag the rpb2 subunit since a similar strategy was previously used in poxviruses without phenotypic consequences31. As a result, while pandoravirus rpb2-3HA was easily visualized by immunofluorescence in cells infected with recombinant viruses, targeting of the gene by CRISPR/Cas9 led to the disappearance of the protein indicating successful depletion (Fig. 3f, g).
A CRISPR/Cas9 genetic screen reveals a core essential genome in pandoravirus
To gather information on the repair mechanisms of double-strand breaks in nuclear GVs, we targeted genes with suspected different degrees of importance for virus replication. Fitness reduction of pandoravirus was highly dependent on the targeted gene locus (Fig. 4a) and ranged from 100-times reduction in particle numbers when genes at the 5′ of the genome were targeted, to no fitness cost when genes at the 3′ of the genome were targeted. Surprisingly, targeting of genes at the 3′ of the genome resulted in large deletions rather than discrete modification of the targeted locus (Fig. 4b–d and S3a), suggesting these viruses are not efficient in repairing double-strand breaks. Quantification of the ratio between PCR products using 5′ and 3′ regions of the genome as templates demonstrates that large deletions are rather frequent at the 3′ of the genome, but not detectable at the 5′ (Fig. 4b). Moreover, viruses harboring such 3′ large deletions were able to complete the full cycle of replication since cloning was achievable (Fig. 4c). Unexpectedly, pandoravirus was capable of losing at least over 600,000 bp (Δ1195458–1838505 base pairs) and more than 300 genes (Δpneo_612-974) (Supplementary Data 1), as shown by pulsed-field gel electrophoresis (Fig. 4d) and PCR amplification of specific loci (Fig. S3a). Interestingly, while the deletion of almost one-third of the genome led to a significant reduction of viral fitness (Fig. 4e), no changes in virion morphology were observed (Fig. 4f). These data strongly correlate with the strikingly low abundance of virion proteins encoded by genes located at the 3′ end of the genome (Fig. 4g). Moreover, viral DNA replication was significantly impaired in viruses lacking the genes present at the 3′ of the genome (Fig. 4h). Taken together, these data demonstrate that these genes are important during the intracellular stage of the infectious cycle, most likely for host–pathogen interaction.
Interestingly, fitness differences during the first generation of viral production upon targeting with CRISPR/Cas9 were associated with the gene location rather than the gene target itself (Fig. 5a). These data indicate that large deletions were not viable when generated at the 5’ end of the genome (core genome). The core genome includes the genes pneo_77 (dispensable in blue, Fig. 5a) to pneo_544 (essential in red, Fig. 5a). The identification of the precise limits of this area would require further characterization. Meanwhile, amplifications of the viruses produced under a constant selection with Cas9 targeting the core genome, led occasionally to the production of resistant viruses (Fig. S3b), which contain modification of the genome at the target site (Fig. 5b). Coherently, shorter deletions were observed in viral progenies when genes present at the core genome were targeted (Fig. 5c), demonstrating the capability of these viruses to repair by non-homologous end joining, even if not efficiently. Taken together, we were able to determine a core essential region of the genome associated with the 5’ end (with a high density of essential genes, labeled in red), and a non-essential one located at the 3’ of the genome (labeled in blue) (Fig. 5a). Finally, not all genes present at the core genome are essential since knock out of some of these genes was accomplished (Fig. 5a–c). This essential core region was even smaller in P. macleodensis, which lacks a recent genomic inversion present in P. neocaledonia (Fig. 5d), located at the intersection of the essential and non-essential zone of the genome (Fig. 5a)3. Concordantly, this inversion in P. neocaledonia has introduced a fragment of the non-essential region of the genome into a newly fragmented essential core region of its genome (Fig. 5d).
Genomic organization of pandoravirus conserves traces of their putative evolution from smaller viral ancestors
The distribution of essential genes observed in pandoravirus would be parsimoniously explained by the expanding genome hypothesis9 and rather hardly with the reductive one8. Concordantly, reductive evolution in pathogenic bacteria or eukaryotes did not lead to the concentration of essential genes in particular regions of their genomes32,33. To test the genome expansion hypothesis in pandoravirus, we classified genes according to their conservation in their distant relatives within Nucleocytoviricota (Fig. 6a)9,25, and analyzed their position in the genome of P. neocaledonia (Fig. 6b). P. neocaledonia genes with orthologs in Phycodnaviridae and mollivirus are enriched at the 5’ of the genome, coinciding with the essential core of the genome (Fig. 6b). On the other hand, clade-specific genes are more uniformly distributed and strain-specific genes (ORFan genes) tend to accumulate at the 3’ end of the genome (Fig. 6b). Concordantly, these data support the hypothesis of a biased expansion of the pandoravirus genome from a Phycodnaviridae-like virus towards the 3’ end of the genome.
Importantly, gene persistence throughout evolution correlates with the fitness cost associated with the loss of function of that gene34. Thus, to further test the genome expansion theory of pandoravirus, we attempted gene KO for approximately 10% of the genes present in the essential region of the genome (Fig. 6c and S4). In agreement with the genome expansion theory, the density of essential genes increased at the positions where Phycodnaviridae and mollivirus conserved genes are also highly represented (Fig. 6c). The phenotype of non-essential genes was not associated with the percentage of infectious particles produced but rather the total number of particles produced per cell as shown by the infectious dose 50 (Fig S4c). Moreover, the assessment of the depth of conservation of genes across Nucleocytoviricota versus the phenotype associated with their deletion indicates that shared genes have a greater tendency to produce a more drastic effect upon deletion (Fig. 6d). This distribution of phenotypes followed the trend predicted by the genome expansion theory and correlates with the proposed evolutionary pathway from a phycodnavirus-like ancestor to pandoravirus (Fig. 6a). Moreover, genes largely conserved across the Nucleocytoviricota (despite multiple independent deletions rendered only one gene strictly conserved in all viruses35) that were used for the phylogenetic reconstruction of the phylum are essential for pandoravirus replication (Fig. 6d). These genes include: B DNA Polymerase (PolB), A32-like packaging ATPase (A32), virus late transcription factor 3 (VLTF3), large RNA polymerase subunits (RNAP1 and RNAP2), TFIIB transcriptional factor (TFIIB) and D5-like primase-helicase9,25,35. These data align with their use for phylogenetic reconstructions by indicating their likely ancestry for the phylum.
Finally, analysis of gene essentiality according to gene broad functional annotations indicates that essential genes accumulate particularly under “replication” and “transcription” functions, while “Translation and amino-acid metabolism” includes none (Fig. 6e). This observation further supports the viral origin of pandoravirus and indicates that translation-related genes might have been acquired recently to maximize the expression of viral genes36. Genes with unknown functions, including all the ORFan genes, englobe a large amount of non-essential genes. Thus, these genes are also likely to be more recent acquisitions of the pandoravirus.
Previous reports have shown an unexpected prevalence of multiple-copy genes in Pandoravirus genomes3. Moreover, the speed of acquisition of these multi-copy genes increased with its putative genome expansion (Fig. 6a). These data suggest that genome expansion was accompanied by a strong increase in genetic redundancy and thus, led to increased robustness of the biological system37. To test this hypothesis, we analyzed a family of SNF2-like genes present in Phycodnaviridae, mollivirus, and pandoravirus. SNF2 genes are helicase-related ATPases which in eukaryotic cells drive chromatin-remodeling38. While Phycodnaviridae and mollivirus encode only one SNF2-like gene, this gene is present in two copies in the genome of pandoravirus. Interestingly, phylogenetic analysis of these genes indicates that the duplication occurred in the ancestor of pandoravirus and mollivirus (Fig. S5). A followed-up deletion of one of these genes would explain the presence of a single gene in mollivirus (Fig. S5). Such evolutionary history can be explained by the proposed “genomic accordion” evolution proposed for GVs39,40. Importantly, both genes present in pandoravirus were shown dispensable individually (Fig. 7a) and viruses encoding a deletion of these genes were easily cloned (Fig. 7b, c). On the other hand, double KO led to synthetic lethality (Fig. 7b, c). Concordantly, these data demonstrate the presence of epistatic effects and genetic redundancy in GVs. Regardless, both genes are not strictly redundant since a fitness cost associated with single deletion was observed. To assess if genetic redundancy increased with genome expansion, we targeted the SNF2-like gene in mollivirus, which is present in a single copy. We were not able to knock-out mk_236, which strongly supports its essentiality (Fig. 7a–c). Thus, such redundancy displayed by pandoravirus would indicate that the flexibility for adaptation to uncertain environments might generate a strong selective pressure in these viruses. Concordantly, increased genetic robustness might be a key component that explains the selective advantage that led to viral giantism.
Major capsid protein (MCP) is essential for virion formation in mollivirus
Viruses belonging to the Nucleocytoviricota tended to lose some core genes independently throughout evolution9,35. In particular, pandoravirus lacks a Nucleocytoviricota signature MCP protein9. MCP is a major component of the viral particles of icosahedral viruses and a key factor for the shaping of their capsids41. Interestingly, despite its spherical shape, mollivirus encodes an MCP-like protein14, and it has been speculated that this gene would be a trace of their icosahedral ancestry (Fig. 8a)12. AlphaFold predictions42 confirmed the jelly-roll folding unit of ms_334 which closely resembles the MCP of other icosahedral viruses like Paramecium bursaria chlorella virus 1 (PBCV-1)41 (Fig. 8b). To assess the role of this protein in the formation of mollivirus particles, we first expressed an ectopic copy of the gene in the amoeba. Fluorescent N- or C-terminally tagged MCP localized into the cytoplasm of non-infected amoebas but re-localized to specific sub-compartments upon infection by mollivirus (Fig. 8c). Moreover, incorporation of MCP to the viral particle was confirmed by fluorescence (Fig. 8d). To attempt genetic deletion of mollivirus MCP, we infected transgenic amoebas expressing Cas9 and gRNAs targeting the mcp gene in mollivirus. Targeting of mcp efficiently reduced the viral fitness of mollivirus (Fig. 8e). Trans-complementation, as we have previously described43, was not successful to restore viral growth (Fig. 8f). Expression quantities or timing might explain the failure of trans-complementation. Thus, we developed a system for cis-complementation of gene KO (Fig. 8f). To achieve cis-complementation, we built a plasmid that encodes for a second copy of the mcp gene containing silent mutations at the gRNA target sites; and a nourseothricin N-acetyl transferase (NAT) resistance cassette, which confers resistance to nourseothricin30. The NAT selection cassette was expressed under the control of a pandoravirus promoter, which can be used for expression of genes in mollivirus (Fig. 8f). Importantly, such recombinant viruses contain two copies of the mcp: one susceptible to cleavage by Cas9, and one resistant to its targeting (Fig. 8f and Fig. S5). Concordantly, double-strand breaks would still occur in the cis-complemented strain which might be repaired in order to produce viable progeny. Contrary to trans-complementation, cis-complementation efficiently restored viral fitness upon Cas9 cleavage of mcp (Fig. 8e). Thse data demonstrate that defects in the viral cycle associated with Cas9 cleavage are due to the knockout of the targeted gene rather than a general response to double-strand breaks. Moreover, while targeting of mcp did not impact DNA replication of the virus (Fig. 8g), biogenesis of virions was largely impaired (Fig. 8h). In contrast to untargeted or cis-complemented viruses, targeting of mcp on wild-type viruses led to the accumulation of membranous material in the cytoplasm and absence of mature neosynthesized virions (Fig. 8h). These data indicate a failure to initiate tegument biogenesis upon recruitment of an open cisterna44 and suggests a scaffolding function for mollivirus MCP. Overall, these data support a replacement of the MCP functions during the evolution of phycodnavirus-like icosahedral viruses to mollivirus-like.
The study of the origin of viruses and their evolutionary history remains rather challenging for a reliable phylogenetic reconstruction due to their fast evolution and quick exchange of genetic markers9,45. Regardless, such fast evolution in dsDNA viruses (with low mutation rates but high burst size46) might have been key to the evolution of all living cells47,48, and would have contributed to shaping modern genes25, genetic elements49, or even organelles like the nucleus50,51. Here, we introduced a battery of tools that allows genetic manipulation of host and nuclear GV and thus, the trackability of these gene exchanges. This includes the characterization of the neglected functional genomic of nuclear GV and their up to 70% ORFan gene-containing genomes. ORFan genes have contributed substantially as evolutionary innovations4 and are expected to arise as lineage-specific adaptations to the environment52,53. The role of such innovations and unpredictable functions in pandoravirus are now open to rigorous investigation.
Particularly, we demonstrate the existence of an essential core of the genome located at the 5´ end of pandoravirus genome (Fig. 5a), which coincides with the location of Phycodnaviridae and mollivirus-conserved ancestral genes (Fig. 6b). These data, together with the phylogeny recently accepted by the International Committee on Taxonomy of Viruses54, support the hypothesis that nuclear GVs likely originated from smaller icosahedral viruses36. Moreover, the presence of an icosahedral mcp gene in mollivirus, which is essential for virion formation despite its ovoid shape further supports this theory. If pandoravirus ancestors possessed mollivirus-like particles and which molecular changes allowed the dispensability of the scaffolding functions of MCP are yet to be studied.
Interestingly, we also show that while discrete mutations can easily be isolated when the 5’ of the genome is targeted by CRISPR/Cas9, large deletions are frequent when the 3’ of the genome is targeted. These differences might be due to several reasons. First, the repair mechanism might not defer across the genome but the high density of essential genes at the 5’ of the genome could select for the viruses with small deletions. Second, DNA compaction and transcription highly affect Cas9 targeting in mammalian cells55. Differences in density of late versus early genes at different positions of the genome might also indicate differences in DNA compaction or transcription activity along the viral infectious cycle. Third, it has been previously shown for herpesvirus that replication of the DNA influences the insertion and deletion lengths56. Moreover, replication of the DNA in vaccinia virus has been shown to initiate near the end of the genome57. While the location of the origin of replication of pandoravirus is unknown, if such origin would be located at the extremities of the genome, it might influence repair efficiencies similarly to what was described for herpesvirus. However, the exact mechanism governing these differences in different regions of the viral genome remains to be studied.
Overall, we anticipate that the development of reverse genetic approaches in the field of GVs will lead to a vast expansion of the knowledge about these outliers of the viral world and shed light on the intricated relationship with their cellular hosts.
Viral strains utilized in this work
The following viral strains have been used in this study: Pandoravirus neocaledonia3, Pandoravirus macleodensis3, Pandoravirus kuranda (this study, ON887157), Mollivirus kamchatka26, Pithovirus sibericum27 and Mimivirus reunion28.
Cloning of DNA constructs
CRISPR/Cas9 expression vector
Vectors for CRISPR/Cas9 expression in A. castellanii cells were generated by inserting a codon optimized SpCas9 (with or without a nuclear localization signal (NLS)) and a cassette for polycistronic gRNA expression into the pEF1-GFP-NEO vector43. The cassette for polycistronic gRNA expression was synthesized by GenScript and designed to contain an A. castellanii U6 promoter-tRNA-gRNA1-tracerRNA1-tRNA-gRNA2-tracerRNA2. This cassette was further modified by PCR and circularized by recombination (InFusion Takara) to replace gRNA1-tracerRNA1-gRNA2 by a NotI site. gRNAs were included in primers synthesized by eurofins genomics, and inserted into the NotI site by InFusion (Takara).
MCP-mRFP and mRFP-MCP expression vectors
PCR product of the mcp was generated using M. kamchatka genomic DNA produced by Wizard genomic DNA purification kit (PROMEGA) according to manufacturer’s specifications. Primers used are HB84/HB85 and HB86/HB87. Vectors for expression of mcp were generated using the pEF1-mRFP-NAT vector by InFusion (Takara) into the NdeI or XhoI site for N- or C-terminal RFP tagging, respectively.
MCP second copy
The plasmid for cis-complementation was generated by sequential cloning of the 3’ UTR of pneo_480, the promoter of pneo_650, a Nourseothricin N-acetyl transferase (NAT) selection cassette, the 3’ UTR of pneo_650, the promoter of mcp and the mcp coding sequence. Each cloning step was performed using the Phusion Taq polymerase (ThermoFisher) and InFusion (Takara). Silent mutations at the gRNA targeting site were also introduced using InFusion. Finally, 500 bp homology arms were introduced at the 5’ and 3’ end of the cassette in order to induce homologous recombination with the viral DNA. Prior to transfection, plasmids were digested with EcoRI and NotI.
The plasmid for endogenous tagging was generated by sequential cloning of the 3xHA/FLAG tag, 3’ UTR of pneo_480, the promoter of pneo_650, a Nourseothricin N-acetyl transferase (NAT) selection cassette, the 3’ UTR of pneo_650. Each cloning step was performed using the Phusion Taq polymerase (ThermoFisher) and InFusion (Takara). Finally, 500 bp homology arms were introduced at the 5’ and 3’ end of the cassette in order to induce homologous recombination with the viral DNA. Prior to transfection, plasmids were digested with EcoRI and NotI.
cs1 (cellulose synthase) second copy
The cs1 gene (XP_004335167.1) was amplified by PCR from genomic DNA and cloned into a pGAPDH-GFP amoebal expression plasmid58. The plasmid was linearized by NdeI and the gene was inserted in the plasmid Vc243 used for recombination (InFusion Takara).
Cell culture and establishment of cell lines
Acanthamoeba castellanii (Douglas) Neff (American Type Culture Collection 30010TM) cells were cultured at 32 °C in 2% (wt/vol) proteose peptone, 0.1% yeast extract, 100 μM glucose, 4 mM MgSO4, 0.4 mM CaCl2, 50 μM Fe(NH4)2(SO4)2, 2.5 mM Na2HPO4, 2.5 mM KH2PO4, pH 6.5 (home-made PPYG) medium supplemented with antibiotics [ampicilline 100 μg/mL, and Kanamycin 25 μg/mL]. 100 μg/mL Geneticin G418 or Nourseothricin was added when necessary.
1.5×105 Acanthamoeba castellanii cells were transfected with 6 μg of each plasmid using Polyfect (QIAGEN) in phosphate saline buffer (PBS) according to the manufacturer’s instructions. Selection of transformed cells was initially performed at 30 μg/mL Geneticin G418 or Nourseothricin and increased up to 100 μg/mL within a couple of weeks.
Establishment of viral lines
Cas9 KO viral lines
Cell lines expressing Cas9 and gRNA were infected with a MOI (Multiplicity Of Infection) of 1. Infection was allowed to proceed for 24 h. This new generation of viruses was used to quantify the number of particles produced or cloned for isolation of viral strains.
Cas9 KO cell lines for cellulose synthase
1.5×105 Acanthamoeba castellanii cells were transfected with 6 μg of the plasmid expressing Cas9 and gRNAs against the cs1-3 using Polyfect (QIAGEN) in phosphate saline buffer (PBS). Selection of transformed cells was initially performed at 30 μg/mL Geneticin G418. Upon recovery of the cells, amoeba culture was cloned and genotyped.
Insertion of MCP second copy and endogenous tagging. 1.5 × 105 Acanthamoeba castellanii cells were transfected with 6 μg of linearized plasmid using Polyfect (QIAGEN) in phosphate saline buffer (PBS). One hour after transfection, PBS was replaced with PPYG and cells were infected with 1.5 × 107 mollivirus particles for 1 h with sequential washes to remove extracellular virions. 24 h after infection the new generation of viruses (P0) was collected and used to infect new cells. An aliquot of P0 viruses was utilized for genotyping in order to confirm integration of the selection cassette. New infection was allowed to proceed for 1 h, then washed to remove extracellular virions and nourseothricin was added to the media. Viral growth was allowed to proceed for 24 h. This procedure was repeated one more time before removing the nourseothricin selection to allow viruses to expand more rapidly. Once, viral infection was visible, selection procedure was repeated one more time. Viruses produced after this new round of selection were used for genotyping and cloning.
Cloning and genotyping. 150,000 A. castellanii cells were seeded on 6 wells plates with 2 mL of PPYG. After adhesion, viruses were added to the well at a multiplicity of infection (MOI) of 2. One hour post-infection, the well was washed 5 times with 1 mL of PPYG and cells were recovered by scraping. Amoebas were then diluted until obtaining a suspension of 1 amoeba/μL. One μL of such suspension was added in each well of a 96-well plate containing 1000 uninfected A. castellanii cells and 200 μL of PPYG. Wells were later monitored for cell death and 100 μL collected for genotyping. Genotyping was performed using Terra PCR Direct Polymerase Mix (Takara) following manufacturers specifications.
Virus Purification. The wells presenting a recombinant genotyping were recovered, centrifuged 5 min at 500 × g to remove the cellular debris, and used to infect ten 75 cm² tissue-culture flasks plated with fresh Acanthamoeba cells. After lysis completion, the cultures were recovered, centrifuged 5 min at 500 × g to remove the cellular debris, and the virus was pelleted by a 45 min centrifugation at 6800 × g prior purification. The viral pellet was then resuspended and washed twice in PBS and layered on a discontinuous CsCl gradient (1.2/1.3/1.4/1.5 g/cm3), and centrifuged at 100,000 × g overnight. An extended protocol is shown in ref. 59.
Transfected A. castellanii cells were grown on poly-l-lysine coated coverslips in a 12-well plate and fixed with PBS containing 3.7% formaldehyde for 20 min at room temperature. After one wash with PBS buffer, coverslips were mounted on a glass slide with 4 μl of VECTASHIELD mounting medium with DAPI and the fluorescence was observed using a Zeiss Axio Observer Z1 inverted microscope using a 63x objective lens associated with a 1.6x Optovar for DIC, mRFP or GFP fluorescence recording. Quantification of fluorescence was performed in at least 200 cells from no less than 10 different fields of view.
Amoeba encystation and quantification by SDS resistance
Acanthamoeba cysts were induced in encystation media (100 mM KCl, 80 mM MgSO4, 0.4 mM CaCl2, and 20 mM 2-amino-2-methyl-1,3-propanediol, pH 9.0) for 3 days, as previously reported23. Cysts were counted under a light microscope upon treatment with 0.5% SDS for 10 minutes and encystation ratios were calculated23.
Calcofluor white staining
Calcofluor white staining was performed as previously described23. Briefly, A castellanii cells were incubated in encystation media for 3 days and amoeba suspensions were incubated for 20 minutes at room temperature with 2.5% calcofluor white staining solution. After washing with PBS, samples were observed under a fluorescence microscope.
Indirect immunofluorescence analysis (IFA)
A. castellanii cells were grown on 96-well plates and infected with pandoravirus (MOI = 10). Cells were fixed at 6 hpi (post-infection) using 4% paraformaldehyde. Immunofluorescence analysis was performed as previously described60. Nuclei were stained with Hoescht, and coverslips were mounted in Fluoromount G (Southern Biotech).
Infective dose 50
ID50 were calculated as previously described61. In brief, particle numbers were calculated and adjusted in order to obtain a stock solution of viruses containing 1×107 particles/mL. Serial dilution of this stock was used to infect a 96 wells plate containing 1000 A. castellanii per well. Cytopathic effect was then observed one-week post-infection. The end point was later calculated and expressed as 50% infective dose (ID50).
Statistics and reproducibility
All data are presented as the mean ± s.d. of 3 independent biological replicates (n = 3), unless otherwise stated in the figure. All data analysis were carried out using Graphpad Prism. The null hypothesis (α = 0.05) was tested using unpaired two-tailed Student’s t tests.
Viral fitness determination
Viral burst was calculated by manual counting using a hemocytometer chamber. In the case of purified viral samples, optical density was utilized for viral quantification. The purity of the viral samples were analyzed by microscopy as previously described59.
Pulsed-field gel electrophoresis
Viral suspensions were prepared according to ref. 62. Drops of 45 μL of the viral suspension were embedded in 1% low melting agarose, and the plugs were incubated in lysis buffer (50 mM Tris-HCl pH 8.0, 50 mM EDTA, 1% (v/v) laurylsarcosine, and 1 mg/mL proteinase K) for 24 h at 50 °C with light shaking (500 rpm). The lysis buffer was renewed every 8 h and 1 mM DTT was added 30 min before the second buffer change. After lysis, the plugs were washed once in sterile water and three times in TE buffer (10 mM Tris HCl pH 8.0 and 1 mM EDTA), for 15 min at 50 °C. Electrophoresis was carried out in 0.5× TBE using a 1% agarose gel for 20 h 10 min at 6 V/cm, 120° included angle and 14 °C constant temperature in a CHEF-MAPPER system (Bio-Rad) with pulsed times ranging from 0.5 s to 3 min 10 s with a linear ramping factor.
Quantitative PCR analysis
Viral genomes or gDNA from infected amoebas were purified using Wizard genomic DNA purification kit (PROMEGA). To determine the amplification kinetic, the fluorescence of the EvaGreen dye incorporated into the PCR product was measured at the end of each cycle using SoFast EvaGreen Supermix 2× kit (Bio-Rad, France). A standard curve using gDNA of purified viruses was performed in parallel of each experiment. For each point, a technical triplicate was performed. Quantitative real-time PCR (qRT-PCR) analyses were performed on a CFX96 Real-Time System (Bio-Rad).
Extracellular virions or A. castellanii-infected cell cultures were fixed by adding an equal volume of PBS with 2% glutaraldehyde and 20 min incubation at room temperature. Cells were recovered and pelleted 20 min at 5000 × g. Virions were recovered and pelleted 45 min at 6800 × g. The pellet was resuspended in 1 mL PBS with 1% glutaraldehyde, incubated at least 1 h at 4 °C, and washed twice in PBS prior coating in agarose and embedding in Epon resin. Each pellet was mixed with 2% low melting agarose and centrifuged in microvettes tubes (Sarstedt), the agarose coated sample was recovered and cutted to obtain small flanges of approximately 1 mm3. These samples were then prepared using the osmium-thiocarbohydrazide-osmium method: 1 h fixation in 2% osmium tetroxide with 1.5% potassium ferrocyanide, 20 min in 1% thiocarbohydrazide, 30 min in 2% osmium tetroxide, overnight incubation in 1% uranyl acetate, 30 min in lead aspartate, dehydration in increasing ethanol concentrations (50, 70, 90 and 100% ethanol) and embedding in Epon-812. Ultrathin sections of 70 nm were observed using a FEI Tecnai G2 operating at 200 kV63.
Comparative genomics and phylogeny
The protein-coding genes annotated in the reassembled Pandoravirus neocaledonia genome were compared to the ones of the following viral genomes: Emiliania huxleyi virus 86 (NC_007346.1), Feldmannia species virus (NC_011183.1), Ectocarpus siliculosus virus 1 (NC_002687.1), Ostreococcus tauri virus 1 (NC_013288.1), Paramecium bursaria Chlorella virus 1 (NC_000852.5), Mollivirus sibericum (NC_027867.1), Mollivirus kamchatka (MN812837.1), Pandoravirus dulcis (NC_021858.1), Pandoravirus salinus (NC_022098.1), Pandoravirus quercus (NC_037667.1), Pandoravirus macleodensis (NC_037665.1) and Pandoravirus kuranda (ON887157, this study, https://www.ncbi.nlm.nih.gov/nuccore/ON887157). We used OrthoFinder (version 2.5.4) to compute the orthogroups (OG) of shared genes using the following parameters: “-M msa -S diamond_ultra_sens -y”. From a set of 438 OG selected by OrthoFinder (“Orthogroups_for_concatenated_alignment.txt” file) we kept the 435 OG without paralog, performed a multiple alignment of each OG using Clustal omega (version 1.2.4) and concatenated the multiple alignments using Catsequence. Next a partitioned phylogeny was computed using IQ-Tree (version 2.1.2) with the “-bb 1000 -m MFP + MERGE -rcluster 10” parameters. The root of the tree was calculated using the midpoint rooting method.
Density of shared genes along the Pandoravirus neocaledonia genome
Each Pandoravirus neocaledonia gene was given an “ancestry score” based on the OGs and the phylogenetic tree. This score indicates the node of the last common ancestor of all viruses having a homologue of the given gene. It ranges from 0 (genes only present in Pandoravirus neocaledonia) up to 5 (ancient, shared with at least one phycodnavirus). We next sliced the Pandoravirus neocaledonia genome in 100 bins and computed the fraction of genes of each category. Finally, these values were smoothed by computing the average values over a sliding window of 20 bins.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information files. Source data are provided with this paper. Plasmids generated in this study can be obtenaid from Addgene or upon request to the authors. Viral strains can be obtained upon request to the authors. Viral genomes: Pandoravirus kuranda (ON887157) Source data are provided with this paper.
Philippe, N. et al. Pandoraviruses: amoeba viruses with genomes up to 2.5 Mb reaching that of parasitic eukaryotes. Science 341, 281–286 (2013).
Queiroz, V. F. et al. Amoebae: hiding in plain sight: unappreciated hosts for the very large virsuses. Annu. Rev. Virol. 9, 79–98 (2022).
Legendre, M. et al. Diversity and evolution of the emerging Pandoraviridae family. Nat. Commun. 9, 2285 (2018).
Tautz, D. & Domazet-Loso, T. The evolutionary origin of orphan genes. Nat. Rev. Genet. 12, 692–702 (2011).
Abergel, C. & Claverie, J. M. Giant viruses. Curr. Biol. 30, R1108–R1110 (2020).
Brahim Belhaouari, D. et al. Metabolic arsenal of giant viruses: host hijack or self-use? Elife 11, e78674 (2022).
Moniruzzaman, M., Martinez-Gutierrez, C. A., Weinheimer, A. R. & Aylward, F. O. Dynamic genome evolution and complex virocell metabolism of globally-distributed giant viruses. Nat. Commun. 11, 1710 (2020).
Claverie, J. M. Viruses take center stage in cellular evolution. Genome Biol. 7, 110 (2006).
Aylward, F. O., Moniruzzaman, M., Ha, A. D. & Koonin, E. V. A phylogenomic framework for charting the diversity and evolution of giant viruses. PLoS Biol. 19, e3001430 (2021).
Pereira Andrade, A. et al. New isolates of Pandoraviruses: contribution to the study of replication cycle steps. J. Virol. 93, e01942–18 (2019).
Brahim Belhaouari, D. et al. Evidence of a cellulosic layer in pandoravirus massiliensis tegument and the mystery of the genetic support of its biosynthesis. Front. Microbiol. 10, 2932 (2019).
Krupovic, M., Yutin, N. & Koonin, E. Evolution of a major virion protein of the giant pandoraviruses from an inactivated bacterial glycoside hydrolase. Virus Evol. 6, veaa059 (2020).
Yutin, N. & Koonin, E. V. Pandoraviruses are highly derived phycodnaviruses. Biol. Direct 8, 25 (2013).
Legendre, M. et al. In-depth study of Mollivirus sibericum, a new 30,000-y-old giant virus infecting Acanthamoeba. Proc. Natl Acad. Sci. USA 112, E5327–E5335 (2015).
Sekine, R., Kawata, T. & Muramoto, T. CRISPR/Cas9 mediated targeting of multiple genes in Dictyostelium. Sci. Rep. 8, 8471 (2018).
Mefferd, A. L., Kornepati, A. V., Bogerd, H. P., Kennedy, E. M. & Cullen, B. R. Expression of CRISPR/Cas single guide RNAs using small tRNA promoters. RNA 21, 1683–1689 (2015).
Bisio, H., Krishnan, A., Marq, J. B. & Soldati-Favre, D. Toxoplasma gondii phosphatidylserine flippase complex ATP2B-CDC50.4 critically participates in microneme exocytosis. PLoS Pathog. 18, e1010438 (2022).
Cervantes, M. D., Coyne, R. S., Xi, X. & Yao, M. C. The condensin complex is essential for amitotic segregation of bulk chromosomes, but not nucleoli, in the ciliate Tetrahymena thermophila. Mol. Cell Biol. 26, 4690–4700 (2006).
Gicquaud, C. & Tremblay, N. Observations with Hoechst staining of amitosis in Acanthamoeba-Castellanii. J. Protozool. 38, 221–224 (1991).
Gantz, V. M. & Bier, E. Genome editing. The mutagenic chain reaction: a method for converting heterozygous to homozygous mutations. Science 348, 442–444 (2015).
Byers, T. J. Molecular biology of DNA in Acanthamoeba, Amoeba, Entamoeba, and Naegleria. Int. Rev. Cytol. 99, 311–341 (1986).
Anderson, I. J. et al. Gene discovery in the Acanthamoeba castellanii genome. Protist 156, 203–214 (2005).
Moon, E. K., Hong, Y., Chung, D. I., Goo, Y. K. & Kong, H. H. Down-regulation of cellulose synthase inhibits the formation of endocysts in Acanthamoeba. Korean J. Parasitol. 52, 131–135 (2014).
Aqeel, Y., Siddiqui, R. & Khan, N. A. Silencing of xylose isomerase and cellulose synthase by siRNA inhibits encystation in Acanthamoeba castellanii. Parasitol. Res. 112, 1221–1227 (2013).
Guglielmini, J., Woo, A. C., Krupovic, M., Forterre, P. & Gaia, M. Diversification of giant and large eukaryotic dsDNA viruses predated the origin of modern eukaryotes. Proc. Natl Acad. Sci. USA 116, 19585–19592 (2019).
Christo-Foroux, E. et al. Characterization of Mollivirus kamchatka, the First Modern Representative of the Proposed Molliviridae Family of Giant Viruses. J. Virol. 94, e01997–19 (2020).
Legendre, M. et al. Thirty-thousand-year-old distant relative of giant icosahedral DNA viruses with a pandoravirus morphology. Proc. Natl Acad. Sci. USA 111, 4274–4279 (2014).
Villalta, A. et al. The giant mimivirus 1.2 Mb genome is elegantly organized into a 30 nm diameter helical protein shield. Elife 11, e77607 (2022).
Mendoza, S. D. et al. A bacteriophage nucleus-like compartment shields DNA from CRISPR nucleases. Nature 577, 244–248 (2020).
Zahringer, U., Voigt, W. & Seltmann, G. Nourseothricin (streptothricin) inactivated by a plasmid pIE636 encoded acetyl transferase of Escherichia coli: location of the acetyl group. FEMS Microbiol Lett. 110, 331–334 (1993).
Grimm, C. et al. Structural basis of poxvirus transcription: vaccinia RNA polymerase complexes. Cell 179, 1537–1550.e1519 (2019).
Moule, M. G. et al. Genome-wide saturation mutagenesis of Burkholderia pseudomallei K96243 predicts essential genes and novel targets for antimicrobial development. mBio 5, e00926–00913 (2014).
Sidik, S. M. et al. A genome-wide CRISPR screen in toxoplasma identifies essential apicomplexan genes. Cell 166, 1423–1435.e1412 (2016).
Acevedo-Rocha, C. G., Fang, G., Schmidt, M., Ussery, D. W. & Danchin, A. From essential to persistent genes: a functional approach to constructing synthetic life. Trends Genet. 29, 273–279 (2013).
Monttinen, H. A. M., Bicep, C., Williams, T. A. & Hirt, R. P. The genomes of nucleocytoplasmic large DNA viruses: viral evolution writ large. Micro. Genom. 7, 000649 (2021).
Yutin, N., Wolf, Y. I. & Koonin, E. V. Origin of giant viruses from smaller DNA viruses not from a fourth domain of cellular life. Virology 466-467, 38–52 (2014).
Krakauer, D. C. & Plotkin, J. B. Redundancy, antiredundancy, and the robustness of genomes. Proc. Natl Acad. Sci. USA 99, 1405–1409 (2002).
Durr, H., Flaus, A., Owen-Hughes, T. & Hopfner, K. P. Snf2 family ATPases and DExx box helicases: differences and unifying concepts from high-resolution crystal structures. Nucleic Acids Res. 34, 4160–4167 (2006).
Elde, N. C. et al. Poxviruses deploy genomic accordions to adapt rapidly against host antiviral defenses. Cell 150, 831–841 (2012).
Filee, J. Genomic comparison of closely related Giant Viruses supports an accordion-like model of evolution. Front. Microbiol. 6, 593 (2015).
De Castro, C. et al. Structure of the chlorovirus PBCV-1 major capsid glycoprotein determined by combining crystallographic and carbohydrate molecular modeling approaches. Proc. Natl Acad. Sci. USA 115, E44–E52 (2018).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Liu, Y. et al. Virus-encoded histone doublets are essential and form nucleosome-like structures. Cell 184, 4237–4250.e4219 (2021).
Quemin, E. R. et al. Complex membrane remodeling during virion assembly of the 30,000-year-old Mollivirus Sibericum. J. Virol. 93, e00388–19 (2019).
Claverie, J. M. Fundamental difficulties prevent the reconstruction of the deep phylogeny of viruses. Viruses 12, 1130 (2020).
Lynch, M. Evolution of the mutation rate. Trends Genet. 26, 345–352 (2010).
Irwin, N. A. T., Pittis, A. A., Richards, T. A. & Keeling, P. J. Systematic evaluation of horizontal gene transfer between eukaryotes and viruses. Nat. Microbiol. 7, 327–336 (2022).
Forterre, P. & Prangishvili, D. The major role of viruses in cellular evolution: facts and hypotheses. Curr. Opin. Virol. 3, 558–565 (2013).
Fischer, M. G. & Suttle, C. A. A virophage at the origin of large DNA transposons. Science 332, 231–234 (2011).
Bell, P. J. The viral eukaryogenesis hypothesis: a key role for viruses in the emergence of eukaryotes from a prokaryotic world environment. Ann. N. Y Acad. Sci. 1178, 91–105 (2009).
Takemura, M. Poxviruses and the origin of the eukaryotic nucleus. J. Mol. Evol. 52, 419–425 (2001).
Khalturin, K., Hemmrich, G., Fraune, S., Augustin, R. & Bosch, T. C. More than just orphans: are taxonomically-restricted genes important in evolution? Trends Genet. 25, 404–413 (2009).
Domazet-Loso, T. & Tautz, D. An evolutionary analysis of orphan genes in Drosophila. Genome Res 13, 2213–2219 (2003).
International Committee on Taxonomy of Viruses Executive, C. The new scope of virus taxonomy: partitioning the virosphere into 15 hierarchical ranks. Nat. Microbiol. 5, 668–674 (2020).
Verkuijl, S. A. & Rots, M. G. The influence of eukaryotic chromatin state on CRISPR-Cas9 editing efficiencies. Curr. Opin. Biotechnol. 55, 68–73 (2019).
Oh, H. S. et al. Herpesviral lytic gene functions render the viral genome susceptible to novel editing by CRISPR/Cas9. Elife 8, e51662 (2019).
Senkevich, T. G. et al. Mapping vaccinia virus DNA replication origins at nucleotide level by deep sequencing. Proc. Natl Acad. Sci. USA 112, 10908–10913 (2015).
Bateman, E. Expression plasmids and production of EGFP in stably transfected Acanthamoeba. Protein Expr. Purif. 70, 95–100 (2010).
Bertaux, L., Lartigue, A. & Jeudy, S. Giant mimiviridae CsCl purification protocol. Bio Protoc. 10, e3827 (2020).
Burda, P. C., Bisio, H., Marq, J. B., Soldati-Favre, D. & Heussler, V. T. CRISPR/Cas9-based knockout of GNAQ reveals differences in host cell signaling necessary for egress of apicomplexan parasites. mSphere 5, e01001–e01020 (2020).
Boratto, P. et al. Acanthamoeba polyphaga mimivirus prevents amoebal encystment-mediating serine proteinase expression and circumvents cell encystment. J. Virol. 89, 2962–2965 (2015).
Blanca, L., Christo-Foroux, E., Rigou, S. & Legendre, M. Comparative analysis of the circular and highly asymmetrical Marseilleviridae genomes. Viruses 12, 1270 (2020).
Fabre, E. et al. Noumeavirus replication relies on a transient remote control of the host nucleus. Nat. Commun. 8, 15087 (2017).
Crotty, S. M. et al. GHOST: recovering historical signal from heterotachously evolved sequence alignments. Syst. Biol. 69, 249–264 (2020).
The authors thank Jean-Michel Claverie for providing the samples to isolate Pandoravirus kuranda and for critical discussion to improve the manuscript. We thank the PiCSL-FBI core facility (Nicolas Brouilly, Fabrice Richard, and Aïcha Aouane, IBDM, AMU-Marseille) and the Plateforme Transcriptome-IMM/CNRS (Yann Denis). This study was founded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No 832601; N.P., S.J., and C.A.). H.B. is the recipient of an EMBO Long-Term Fellowship (ALTF 979-2019).
The authors declare no competing interests.
Peer review information
Nature Communications thanks Jonatas Abrahao and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Bisio, H., Legendre, M., Giry, C. et al. Evolution of giant pandoravirus revealed by CRISPR/Cas9. Nat Commun 14, 428 (2023). https://doi.org/10.1038/s41467-023-36145-4
This article is cited by
Nature Protocols (2023)