Insights into a Viral Lytic Pathway from an Archaeal Virus-Host System

Abstract

Archaeal host cells infected by Sulfolobus turreted icosahedral virus (STIV) and Sulfolobus islandicus rod-shaped virus 2 (SIRV2) produce unusual pyramid-like structures on the cell surface prior to virus-induced cell lysis. This viral lysis process is distinct from known viral lysis processes associated with bacterial or eukaryal viruses. The STIV protein C92 and the SIRV2 protein 98 are the only viral proteins required for the formation of the pyramid lysis structures of STIV and SIRV2, respectively. Since SIRV2 and STIV have fundamentally different morphotypes and genome sequences, it is surprising that they share this lysis system. In this study, we have constructed a collection of C92/P98 chimeric proteins and tested their abilities, both in the context of virus replication and alone, to form pyramid lysis structures in S. solfataricus. The results of this study illustrate that these proteins are functionally homologous when expressed as individual chimeric proteins but not when expressed in the context of complete STIV infection.

INTRODUCTION

We have only a rudimentary understanding of archaeal virus-host interactions. Much of the limited knowledge that we do have comes from examining crenarchaeal viruses that infect members of the Sulfolobales species (1–9). Since its isolation from an acidic hot spring in Yellowstone National Park (YNP), Sulfolobus turreted icosahedral virus (STIV) has emerged as a model system for examining archaeal viruses due to its relative ease of propagation in the laboratory (4), the availability of multiple host and viral genome sequences, and the development of both host and viral genetic systems (9). The 72-nm icosahedral virion is built upon a T=31 lattice that contains an internal lipid membrane and packages a 17.6-kb circular double-stranded (dsDNA) genome (10–14). STIV has been the subject of transcriptomic, proteomic, genetic, and structural studies (2, 4, 7, 9–20).

A single virus-encoded gene product, C92, or its homologue in Sulfolobus islandicus rod-shaped virus 2 (SIRV2) (P98) has been determined to be the sole viral component responsible for the formation of the pyramid-like lysis structures that open to form holes that allow virion release (6, 7). This lysis system appears to be distinct from the holin/endolysin system described for many DNA bacteriophages (21). While many of the details of lysis structure formation remain to be elucidated, it is clear that pyramid formation begins by the dissolution or disassembly of the overlying cellular S layer, followed by the apparent extrusion of the underlying cell membrane in association with multiple copies of the viral C92 or P98 protein to form seven-sided pyramid-like structures that project ∼150 nm above the cell surface. The initially closed pyramids open along the junction between each face, like the petals of a flower. Expression of C92 in Sulfolobus results in the formation of pyramid-like structures that appear indistinguishable from pyramid-like structures produced by STIV infection, at least by transmission electron microscopy (TEM) analysis (7). The introduction of a premature stop codon into the C92 gene prevents virus replication, indicating that C92 is an essential viral gene (7).

This newly described viral lysis system is widespread in viruses infecting thermophilic acidophiles (20). By scanning several cellular and viral metagenomes constructed from YNP hot springs, we were able to identify four groups of viruses that likely contain this lysis system (20). Within YNP, it is likely there are three unique groups of viruses, including two groups of viruses that have yet to be identified (20).

Interestingly, another archaeal virus, SIRV2, appears to have a cellular lysis system similar to that of STIV (5, 6). SIRV2 is a 23-nm by 900-nm rod-shaped virion that packages a 35.8-kb dsDNA genome (22). A product of SIRV2 open reading frame 98 (ORF98) was identified as a constituent of the cell lysis system (5). The protein has a homologue in the genome of STIV, C92. Both of these viral proteins are small (C92, 276 bp and ∼10 kDa; P98, 294 bp and ∼10 kDa) alpha-helical rich proteins with a predicted N-terminal transmembrane domain of approximately 27 amino acids and were shown to form pyramidal structures when expressed in Sulfolobus species. The two proteins share 55.4% identity at the predicted amino acid level; however, the N-terminal domains are only 42% identical, and the C-terminal domains are 67% identical. As with STIV C92, when SIRV2 P98 was expressed in Sulfolobus cells, pyramid-like structures were formed (6).

For this study, we were interested in determining whether we could ascertain functional domains of these viral proteins by constructing chimeric C92/P98 proteins. The hypothesis that we are testing is that the STIV C92 and SIRV P98 proteins are functional homologs in terms of pyramidal structure formation both in vitro and in the context of viral replication. To test this hypothesis, we analyzed C92/P98 chimeric proteins both in the context of STIV infection and when expressed in Sulfolobus in the absence of STIV.

MATERIALS AND METHODS

Sulfolobus strains and growth conditions.

Sulfolobus solfataricus strain P2-2-12 (P2³) was grown in DSMZ medium 182 at pH 3 to revive from glycerol or pH 2.5 for virus production (http://www.dsmz.de/microorganisms/medium/pdf/DSMZ_Medium182.pdf) as previously described (4, 7). Cultures were maintained in long-necked flasks at 80°C in a shaking incubator. S. solfataricus strain PH1-16 (kindly provided by C. Schleper [23]) was grown in Brock's medium at pH 3.5 at 80°C (24).

Construction of STIV C92 deletion mutant.

A complete deletion of C92 from the STIV genome was created using inverse PCR. Briefly, a previously constructed “C92 subclone” (7) was used as the template in an inverse PCR. The C92 gene was removed from the subclone using del-C92-AgeI-F and del-C92-AgeI-R (Table 1) and Phusion High-Fidelity DNA polymerase (New England BioLabs [NEB], Ipswich, MA) using the manufacturer's recommendations. The primers were designed to introduce a unique AgeI restriction endonuclease site in place of the C92 gene. The resulting linear PCR product was purified with the QIAquick PCR purification kit (Qiagen, Germantown, MD), digested with AgeI (NEB), heated to inactivate the enzyme, and purified a second time by excision of the appropriate band from an agarose gel. The construct was ligated back to a circular form using T4 DNA ligase (NEB). After confirmation by enzyme digestion and DNA sequencing, the C92 deletion fragment was introduced into the full-length STIV-TOPO clone. Briefly, the fragment containing the AgeI site in place of the C92 gene was PCR amplified from the subclone using the Topo-seq-F and Topo-seq-R primers (Table 1) and Phusion DNA polymerase (NEB), following the manufacturer's recommendations. The amplicon was verified using gel electrophoresis and purified using QIAquick PCR purification (Qiagen). The purified PCR product was digested with PmlI and SacII (NEB), followed by heat inactivation and another round of purification. The PmlI/SacII-digested PCR product was ligated to PmlI/SacII-digested STIV-TOPO and transformed into BP5α Super Competent Escherichia coli cells (BioPioneer, San Diego, CA). The resulting colonies were screened using colony PCR with the C92-amp-F and C92-amp-R primers (Table 1) and Phusion DNA polymerase, following the manufacturer's recommendations. The PCR products were digested with AgeI to verify the presence of the restriction endonuclease site and further verified by DNA sequencing. The resulting STIV (ΔC92) DNA was prepared for transfection into S. solfataricus P2³ by digestion with BamHI (NEB) to remove the pCRII-TOPO-TA vector, followed by ligation with T4 DNA ligase (NEB) to generate a full-length STIV genome containing an AgeI site in place of the C92 gene. This construct was transformed into S. solfataricus P2³ cells as described previously (7, 9).

Table 1.

Sequences of primers used in this study

Primer	Sequence (5′ to 3′)a
delC92-AgeI-F	CGACCGGTAATATCTTCTTTTTTGGTATTTCTCTTTTATATACTCTTTCTCTAAC
delC92-AgeI-R	TTACCGGTTCTCCTCCTCTTCTTATTTTTAGAAGATATAATCAGATAGGCGA
Topo-seq-F	CGAGCTCGGATCCACTAGTAA
Topo-seq-R	GCGAATTGGGCCCTCTA
C92-amp-F	GACGGGTAGAAGAGGTAGAC
C92-amp-R	GGAGAAAGTGCTAATCTCTCATC
STIV-5f	GGACTGCAGGGACAAATATAC
STIV-mid-r	CCAGATGCAGGTACAGATGTT
SIRV2-P98-AgeI-F	GATAATATTTTTTGACCGGTAATGGCTATAAC
SIRV2-P98-AgeI-R	GCTAAATAATTTTAATTTTTAACCGGTATTATTTTTTTAG
C92-AgeI-F	CTTTTTTAGACTCTAAGACAAACCGGTATTTTTTGATGCTAGCAGAAATAGC
C92-BsrGI-R	TGATTTCTGATCTTCTCGTTGTACAGATGAGCTGCTTC
P98-BsrGI-F	GTTGGAGAAATTGTCCACTTGTACAATGAAAAACAAAATAATG
P98-AgeI-R	GACTAAAAAAATAATACCGGTTAAAAATTAAAA
P98-AgeI-F	GATAATATTTTTTGACCGGTAAATGGCTATAACATTATTAG
P98-BsrGI-R	CATTATTTTGTTTTTCATTGTACAAGTGGACAATTTC
C92-BsrGI-F	GCAGTTTTCATTATCGCCGAAGCAGCTCATCTGTACAACGAGAAGATC
C92-AgeI-R	GAAGATATTACCGGTTCGAGCCAAGAAAAA
Np98Cc92-NdeI-F	GGAGGAGAACCGGCATATGGCTATAACATTATTAG
Np98Cc92-FLAG-SalI-R	CCAAGAAAAAAAAGATAGTCGACTTTATCATCATCATCCTTATAATCCTTTTGAGCGGGC
Nc92Cp98-NdeI-F	CCGGTATTTTCATATGCTAGCAGAAATA
Nc92Cp98-FLAG-SalI-R	CCGGTTAAAAATTAAAAGTCGACTTTATCATCATCATCCTTATAATCTTTAGCTTGCGTATTTGG
P98-NdeI-F	GGAGGAGAACCGCATATGGCTATAAC
P98-FLAG-SalI-R	ATTACCGGTCGACAAATTAAAATTTATCATCATCATCCTTATAATCTTTAGCTTGCGTAT
A510-seq-F	GTAATAGCGGCATTGGCAATGG

^aEngineered endonuclease sites are in bold and the FLAG tag sequence is italicized in each primer sequence.

After transformation, samples were collected every 24 h and analyzed for viral DNA replication via quantitative PCR (qPCR). Standard qPCR protocols were followed, using the manufacturer's recommended protocols (SsoFast EvaGreen supermix; Bio-Rad, Hercules, CA) and 7.5 pmol STIV-5F and mid-r primers (Table 1) to amplify a portion of the STIV major capsid protein (MCP). If STIV DNA was detected in qPCR, a Western blot analysis using the anti-B345 polyclonal antibody raised against the STIV MCP was performed using standard procedures as previously described (7, 9). If viral DNA and the viral MCP were detected, Sulfolobus P2³ cultures were then infected with virus particles to determine whether the genetically modified virus was infectious.

Full replacement of STIV C92 with SIRV2 P98.

In order to replace the STIV C92 ORF with the SIRV2 P98 ORF, the complete P98 ORF was PCR amplified from SIRV2 total DNA using primers designed to engineer an AgeI site at both the 5′ and 3′ ends. The P98 ORF was amplified from SIRV2 viral DNA using 50 pmol (each) of SIRV2-P98-AgeI-F and SIRV2-P98-AgeI-R (Table 1) with Phusion DNA polymerase (NEB). The amplicon size was verified using agarose gel electrophoresis. The PCR product was purified and digested with AgeI (NEB). After heat treatment and another purification, the AgeI-digested P98 product was ligated to AgeI-digested STIV-TOPO (C92 deletion construct) with T4 DNA ligase (NEB) and transformed into E. coli. The transformants were screened using colony PCR with the C92-amp-F and C92-amp-R primers (Table 1) and Phusion DNA polymerase. The resulting PCR products were digested with AgeI to verify the presence of SIRV2 P98 within the STIV genome and further verified by DNA sequencing. Full-length STIV DNA containing the P98 ORF was prepared for transformation and was transformed into S. solfataricus P2³ as described above.

Construction of STIV C92 and SIRV2 P98 chimeric genes.

Sets of C92/P98 chimeras were constructed by fusing the N-terminal domain of one protein with the C-terminal domain of the other protein. A series of PCRs were performed to engineer an AgeI site at each end of the protein and a single BsrGI site after the predicted N-terminal transmembrane domain of each protein (Fig. 1). Briefly, for the N-terminal C92 plus C-terminal P98 construct, STIV C92 was amplified using C92-AgeI-F and C92-BsrGI-R (Table 1 and Fig. 1), and SIRV2 P98 was amplified using P98-BsrGI-F and P98-AgeI-R (Table 1 and Fig. 1). The amplicons were PCR purified and digested with BsrGI (NEB). The endonuclease reaction mixtures were purified, and the C92 and P98 amplicons were ligated to each other using T4 DNA ligase (NEB). The ligation reaction mixture was digested with AgeI (NEB) and ligated to AgeI-digested del-C92-TOPO “subclone” (Fig. 1). The ligation reaction mixture was transformed into E. coli, and colonies were screened by colony PCR using the Topo-seq-F and Topo-seq-R primers (Table 1). The PCR products were digested with BsrGI and further confirmed by DNA sequencing.

Fig 1.

Schematic overview of the construction of the C92/P98 chimeric genes. (A) Construction of N-terminal C92 + C-terminal P98 for replication within STIV. P1, c92 + AgeI-F; P2, c92 + BsrGI-R; P3, p98 + BsrGI-F; P4, p98 + AgeI-R. (B) Construction of N-terminal P98 plus C-terminal C92 for replication within STIV. P5, p98 + AgeI-F; P6, p98 + BsrGI-R; P7, c92 + BsrGI-F; P8, c92 + AgeI-R. See Table S1 in the supplemental material for primer sequences.

The N-terminal C92 plus C-terminal P98 construct was inserted into the STIV genome by amplifying the construct from the “subclone” using the Topo-seq-F and Topo-seq-R primers (Table 1). The resulting PCR product was purified and digested with PmlI and SacII (NEB). The PmlI/SacII-digested construct was ligated into PmlI/SacII-digested STIV-TOPO using T4 DNA ligase (NEB) and transformed into E. coli. The resulting transformants were verified by DNA sequencing with A510-seq-F (Table 1).

To construct the N-terminal P98 plus C-terminal C92 chimera, the above protocol was followed but using opposite ends of each ORF (Fig. 1). Briefly, for the N-terminal P98 plus C-terminal C92 construct, SIRV2 P98 was amplified using P98-AgeI-F and P98-BsrGI-R (Table 1 and Fig. 1), and STIV C92 was amplified using C92-AgeI-F and C92-BsrGI-R (Table 1 and Fig. 1). The amplicons were PCR purified and digested with BsrGI (NEB). The endonuclease reaction mixtures were purified, and the C92 and P98 amplicons were ligated to each other using T4 DNA ligase (NEB). The ligation reaction mixture was then digested with AgeI (NEB) and ligated to the AgeI-digested del-C92-TOPO “subclone” (Fig. 1). The ligation reaction mixture was transformed into E. coli, and the transformants were screened via colony PCR using the Topo-seq-F and Topo-seq-R primers (Table 1). The PCR products were then digested with BsrGI and further confirmed by sequencing.

The N-terminal P98 plus C-terminal C92 construct was inserted into the STIV-TOPO construct by amplifying the construct from the “subclone” using the Topo-seq-F and Topo-seq-R primers (Table 1). The resulting PCR product was purified and digested with PmlI and SacII (NEB). The PmlI/SacII-digested construct was ligated into PmlI/SacII-digested STIV-TOPO using T4 DNA ligase (NEB) and transformed into E. coli. The resulting transformants were verified by DNA sequencing with A510-seq-F (Table 1).

The STIV constructs containing both chimeric sequences were prepared for transfection into S. solfataricus P2³ as described above. After transformation, samples were collected every 24 h and analyzed for viral DNA replication via qPCR as described above. If STIV genomes were detected in qPCR, a Western blot analysis using the anti-B345 polyclonal antibody raised against the STIV MCP was performed.

Expression of C92 and P98 chimeras in Sulfolobus.

Each C92/P98 chimeric gene and full-length STIV C92 and SIRV2 P98 were all expressed in the uracil auxotroph S. solfataricus PH1-16 (23). A method similar to that described above was employed to clone the chimeric genes into a Sulfolobus expression vector, pSeSD1 (7). Using the above subclones as templates, each construct was engineered to have a 5′ NdeI and a 3′ SalI restriction endonuclease site, along with a FLAG tag sequence (Table 1 shows primer sequences). Amplification with each primer set was carried out using Phusion DNA polymerase, following the manufacturer's recommendations. After PCR amplification, the amplicon size was verified using gel electrophoresis, and the products were purified. Each PCR product was digested with NdeI and SalI, heat treated, and further gel purified. The PCR products were ligated into an NdeI/SalI-digested pSeSD1 plasmid using T4 DNA ligase (NEB). The ligation reaction mixtures were transformed into E. coli, and plasmid was purified from the resulting colonies. Plasmid constructs were DNA sequenced to verify each construct.

Approximately 1 μg of purified plasmid DNA was transformed into S. solfataricus PH1-16 as previously described (7). Once a single colony containing the plasmid was isolated, protein production was induced with 0.4% arabinose or repressed with 0.4% galactose. Samples were collected prior to induction (or repression), 24 and 48 h postinduction or -repression, and screened using scanning electron microscopy (SEM) to look for the presence of pyramid structures, and protein expression was measured by Western blotting with anti-FLAG antibodies (Sigma).

Preparation of samples for SEM.

After transformation of PH1-16 cells with C92/P98 constructs, samples were collected for visualization under the SEM. Briefly, at 0, 24, and 48 h post-arabinose induction, 0.5 ml of culture was collected and mixed with an equal volume of electron microscopy (EM)-grade 8% glutaraldehyde (Ted Pella, Redding, CA) for at least 1 h. Silicon wafers were coated with several microliters of 1-mg/ml polylysine. After air-drying, these chips were washed by dipping in distilled water. The fixed cells were centrifuged for 5 min at 6,000 rpm, resuspended in high-performance-liquid-chromatography (HPLC)-grade distilled water, and then placed on the coated silicon chips and allowed to air dry. The silicon chips were dipped again in distilled water to wash. Cells were then coated with a thin film of iridium for 15 s at 20 mA in an Emitech sputter coater and viewed with a Supra 55VP field emission SEM (Zeiss).

RESULTS AND DISCUSSION

We determined the effect of C92/P98 chimeras and complete C92/P98 exchanges on STIV replication and pyramid lysis structure formation in the absence of viral replication. We had previously shown that introduction of a premature stop codon in the predicted N-terminal transmembrane domain of STIV C92 resulted in a nonreplicating virus (7). Therefore, it was not surprising that when the complete C92 gene was deleted from the STIV genome, we did not detect STIV genome replication within Sulfolobus cells. Furthermore, we were unable to detect STIV replication when the STIV C92 ORF was replaced with the SIRV2 P98 ORF. Likewise, we did not detect viral genome replication when the STIV C92 ORF was replaced with either of the C92/P98 chimeric ORFs. All transformed cultures were tested by qPCR for viral genome replication, and Western blotting for the MCP of STIV was performed to test for virus assembly. None of the STIV constructs containing C92/P98 chimeras resulted in virus replication within Sulfolobus P2³ cells. It was somewhat surprising that we did not detect STIV replication within Sulfolobus cells when the C92 gene was replaced with the P98 gene from SIRV2, considering that the SIRV2 P98 protein directed pyramid formation in S. solfataricus PH1-16 cells in the absence of other viral proteins (see below).

In contrast, both C92/P98 chimeras, as well as full-length P98 and C92, resulted in pyramids on the cell surface when expressed in S. solfataricus PH1-16 cells in the absence of other viral proteins. All three of the transformed constructs (P98 only, N-terminal P98 plus C-terminal C92, and N-terminal C92 plus N-terminal P98) resulted in pyramid structures on the surface of S. solfataricus PH1-16 cells when induced with arabinose (Fig. 2). This result is consistent with the expression of STIV C92 in S. solfataricus (7) and SIRV2 P98 in Sulfolobus acidocaldarius (6). Expression of each protein was verified with a Western immunoblot assay probed with antibodies against the FLAG tag present on the C-terminal end of each protein (Fig. 3). As was the case for wild-type C92 expression, there is no induction of protein prior to the addition of arabinose and when repressed with galactose (Fig. 3) (7). SEM images reveal similarly sized pyramids on the surfaces of cells regardless of the construct that is being expressed (Fig. 2).

Fig 2.

Expression of all protein constructs resulted in similarly sized pyramids on the cell surface of S. solfataricus PH1-16. SEM images of C92 (A), P98 (B), N-terminal C92 + C-terminal P98 (C), or (D) N-terminal P98 + C-terminal C92 (D) expressed in Sulfolobus are shown. The arrows point to pyramid structures. Scale bars, 200 nm.

Fig 3.

Western blot detection of C92, P98, Nc92 + Cp98, and Np98 + Cc92 expressed in S. solfataricus PH1-16. In each blot, the samples were collected 24 h post-repression with galactose (lane 1 of each panel) and post-induction with arabinose (lane 2 of each panel). Western blotting was performed with anti-FLAG antibody against the C-terminal FLAG tag present on each construct. M, size standard (kDa).

Previously it had been reported that STIV C92-induced pyramids were observed in both the open and closed states in S. solfataricus PH1-16 (7); however, in contrast, SIRV2 P98-induced pyramids were never observed in the open state when P98 was expressed in S. acidocaldarius or E. coli (6). In this study, we observed both open and closed pyramids when P98 was expressed in S. solfataricus PH1-16 in the absence of other viral proteins (Fig. 4 and Table 2). However, we did not observe open pyramids in either of the chimeric C92/P98 proteins expressed in S. solfataricus PH1-16 (Table 2).

Fig 4.

Expression of both the C92 (A) and P98 (B) proteins in S. solfataricus PH1-16 cells resulted in open (red arrows) and closed (blue arrows) pyramid structures. Scale bar, 1 μm (A) or 200 nm (B).

Table 2.

Percentage of open pyramids in each of the expressed protein constructs within S. solfataricus PH1-16 cultures

Sample name	No. of open pyramids in PH1-16a	No. of closed pyramids in PH1-16	% of open pyramids
C92 alone	29	71	29
P98 alone	21	79	21
N-p98 + C-c92	None	100	0
N-c92 + N-p98	None	100	0

^aA total of 100 pyramids from arabinose-induced cultures were counted.

The results of this study confirm that both STIV and SIRV2, two distinct Sulfolobus viruses, share a lysis mechanism, which, however, is not fully interchangeable in the context of STIV replication. Because of sequence and suspected functional similarities between STIV C92 and SIRV2 P98, we wanted to determine if these proteins contained functional domains that could be exchanged between STIV and SIRV2. Surprisingly, none of the exchanges resulted in STIV replication, indicating that even the most conservative exchanges of the more conserved putative transmembrane domain was nonfunctional. The addition of the AgeI site resulted in the introduction of 3 bp upstream of the C92 start codon. This 3-bp introduction resulted in the amino acid sequence changing from a threonine and lysine to an asparagine and arginine. We eliminated the possibility that the introduced AgeI site upstream of the C92 ORF or an unintended base change distal to the C92 ORF was responsible for the nonreplicating phenotype by reintroducing the C92 ORF into STIV (ΔC92) using the AgeI site and tested for virus replication. We were able to detect virus replication, indicating that the changes in the sequence necessary for engineering the AgeI site were not essential for C92 or that unintended mutations were not responsible for the nonreplicating phenotype. A more likely explanation for why SIRV2 P98 or C92/P98 chimeras interfere with STIV replication is the presence of cis-acting elements within either P98 or C92 that interact with other viral or host factors that operate only within the context of their wild-type replication cycle. When placed out of context, they may have affected viral and or host elements in a manner that is currently unknown. However, the protein is functional, because pyramids were detected in the protein expression experiments.

We have previously speculated that the opening of the pyramid lysis structures is dependent on an unknown host (or viral) protein in a currently undescribed manner. In the absence of other viral proteins, we were able to detect both open and closed pyramids for C92 and P98 expressed in Sulfolobus cells (Fig. 3 and Table 2). This indicates that the opening of the lysis structures is independent of other viral proteins; however, it is plausible that host proteins are involved. The fact that pyramids produced in E. coli never open (6) supports the conclusion that a native host factor is required for the opening of the lysis structures. Additionally, neither of the expressed chimeric proteins resulted in open pyramid structures on Sulfolobus membranes. This could be indicative of the specificity of the interacting host proteins with each viral protein. It is likely that the protein interaction is specific to each of the individual viral proteins and that combining domains (even conserved domains) from two different viral proteins is enough to prevent the interaction required to open the pyramid structures.

The observation that the STIV C92/SIRV2 P98 genes are capable of forming pyramid-like lysis structures indistinguishable from those of either homologous protein alone indicates that they are functionally equivalent with respect to lysis structure formation. One plausible way of both STIV and SIRV2 obtaining this protein is through horizontal gene transfer (HGT). It is highly possible that at one time both viruses were infecting the same Sulfolobus host cell and the gene was transferred between the two viruses. Over evolutionary time, as these viruses have been separated from each other, the proteins have each evolved independently, with each maintaining its lysis function. In support of this hypothesis, several closely related lineages of C92 are evident in viral metagenomic data sets from YNP high-temperature environments (20). It will be worthwhile to determine if the pyramid-like lysis system is specific to only crenarchaeal viruses or is more broadly distributed with the Archaea and other domains of life.