|Home | About | Journals | Submit | Contact Us | Français|
The Kaposi’s sarcoma-associated herpesvirus ORF6 has a 41% sequence identity with Balf2 protein of Epstein-Barr virus and 23% with ICP8 protein of Herpes Simplex type I. Balf2 and ICP8 are multi-functional DNA binding proteins with roles central to viral DNA replication and recombination. In this study we cloned the KSHV ORF6 gene, expressed the full length ORF6 protein in insect cells and purified it to homogeneity. Gel filtration revealed the protein to be present in a broad spectrum of sizes ranging from monomers to high molecular weight oligomers. Transmission electron microscopy (TEM) using negative staining under conditions favoring monomers and small oligomers revealed fields of globular particles measuring 11 nm in diameter consistent with the size of a protein monomer. Incubation of ORF6 protein at room temperature for extended periods of time resulted in the bulk of the protein forming very long helical filaments. Measurements from negative staining revealed that the filaments were up to 2600 nm in length, with a width of 13.7 nm and a long gentle helical periodicity of 42.9 nm along the filament axis. Using rapid freezing and freeze-drying, it was possible to show that the filaments consist of two protein chains wrapped around each other. The possibility that these protein filaments generate a scaffold upon which viral DNA replication, recombination, and encapsidation occur in the infected cell nucleus is discussed.
The lifecycle of herpesviruses is comprised of a latent and a lytic phase. During the lytic phase, viral DNA replication takes place resulting in the production of multiple progeny virions. Kaposi’s sarcoma-associated herpesvirus (KSHV), also known as human herpesvirus 8 (HHV-8), is the most recently identified member of the family of human herpesviruses  and along with Epstein-Barr virus (EBV), is a member of the γ-herpesvirus family. Although KSHV was originally identified as the etiological agent of Kaposi’s sarcoma, it is also associated with two other lymphoproliferative diseases: primary effusion lymphoma and multicentric Castleman’s disease . Similar to EBV , KSHV encodes six replication proteins. These include a DNA polymerase (ORF9), a polymerase processivity factor (ORF59), a primase (ORF56), helicase (ORF44), a primase-associated factor (ORF40/41), and a DNA binding protein (ORF6). Functional analysis of the KSHV replication proteins revealed that they could fully substitute for their EBV counterparts and promote replication of the lytic origin of replication of EBV in Vero cells . ORF6 was identified from sequencing the KSHV genome  as having sequence similarity to EBV Balf2 protein and Herpes Simplex type 1 (HSV-1) ICP8 protein. Both Balf2 and ICP8 have been characterized as DNA binding proteins. ICP8 has been shown to be multifunctional, having roles in DNA replication [6, 7], recombination [8–11] and origin activation . Both ICP8 [8, 9, 11, 12] and Balf2  will form helical filaments when bound to single stranded DNA. In addition, ICP8 has been found to assemble into helical filaments in the absence of DNA in a magnesium dependent manner . Balf2 has not been examined to date for such filament formation.
Light microscope studies of cells undergoing lytic infection with HSV-1 [15, 16], EBV , and recently KSHV  have revealed the presence of virus-specific replication bodies in the infected cell nuclei. These dense bodies appear early in infection and are thought to represent DNA replication factories. These structures stain strongly for ICP8 [15, 16] and Balf2  and in addition to the viral replication proteins, they contain host replication and repair factors. Wang et al.  demonstrated by immunostaining that the presence of ORF6 co-localized with a surprising number of classic DNA repair factors including MSH2, DNA-PK, Ku86, Topo I and Topo II. ORF6 remained highly abundant in these bodies throughout the lytic cycle. The high abundance of the DNA binding proteins in these structures and the propensity of ICP8 to form filamentous structures in the absence of DNA has suggested to us that in additional to playing specific roles in the DNA replication reactions that these DNA binding proteins may provide a structural role as well. For this reason, determination of whether other members of this class in addition to ICP8 are able to self-assemble into protein filaments is central to this hypothesis. In the present study, we describe the cloning, expression and purification of full length ORF6 protein from insect cells. Physical and biochemical analysis has been used to characterize the protein and we report here its propensity to form long helical protein filaments in the absence of DNA.
The ORF6 gene was cloned from BCBL-1 viral DNA . The gene was amplified using the following primers: Forward – 5’-GGC TAT GGA TCC GAT GGC GCA AAG GGA CCA CA-3’ and Reverse – 5’-TCA TCG ATA AGC TTC TAC AAA TCC AGG TCA GAG AG-3’. The PCR product was digested with BamHI and HindIII restriction enzymes, whose sites were incorporated into the primers flanking the ORF6 sequence. The gene was inserted into pFstBac HTa (Invitrogen) plasmid using the same restriction enzymes. The recombinant ORF6 has a 6xHis tag at the N-terminus. The Bacmid (baculovirus genome containing the ORF6 gene) was generated by transforming DH10Bac E.coli cells with the ORF6 construct according to the manufacturer’s procedures (Invitrogen, California). The incorporation of the ORF6 gene into the viral genome was verified by PCR.
Sf21 cells (9×105) grown in a 6-well plate in Sf900 II SFM media (Gibco BRL) were transfected with ~ 5µg of bacmid DNA using Cellfectin (Invitrogen, California). After 72 hours, the virus was collected and stored at 4 °C. For expression, 250 ml of Sf21 cells grown in suspension in Sf900 II SFM were inoculated with the virus at an MOI=10. The cells were incubated for 48 hours at 27 °C. The cells were then pelleted at 1500 × g for 10 minutes and washed once with ice-cold 1X PBS (Gibco BRL). After pelleting again, the cells were frozen and kept at −80 °C until purification. For purification, the batch purification was used following the Qiagen Nickel resin protocol with minor modifications. After thawing the cell pellet at 37 °C, the cells were resuspended in cold lysis buffer (20 mM Tris pH 7.4, 300 mM NaCl, 0.1 % NP-40, 8 mM β-mercaptoetanol and EDTA-free protease cocktail tablets (Roche)) and incubated on ice for 20 minutes. The cells were then dounce homogenized with a B (tight) pestle. The cell extract was clarified by spinning at 100,000 × g for 30 min at 4 °C. Qiagen Nickel resin (1 ml) was equilibrated with the lysis buffer and applied to the clarified cell free extract. The protein was allowed to bind to the column by rocking in the cold room for 1 hour. The resin was then washed several times with lysis buffer containing 5% glycerol and 25 mM Imidazole (pH 8.0). ORF6 protein was eluted with elution buffer (20 mM Tris pH7.4, 200 mM NaCl, 250 mM Imidazole) and dialyzed overnight in dialysis buffer (20 mM Hepes pH 7.4, 200 mM NaCl, 0.1 mM EDTA, 20% glycerol) and stored at −80 °C. The purity of the protein was determined by SDS-PAGE and Coomassie staining.
The physical state of ORF6 was examined by size exclusion chromatography using a Superdex 200 10/30 column (GE) attached to an AKTA FPLC (Pharmacia) device. Purified ORF6 protein (100 µg) was injected into the column, which was equilibrated with dialysis buffer containing 5% glycerol. The elution profile of ORF6 was examined by following the OD280 and SDS-PAGE and staining with Commassie blue.
Negative staining: aliquots (5 µl) of protein (~10–20 µg/ml) were absorbed to copper grids covered by thin glow-charged carbon foils and stained for 1 to 5 minutes with 2% (w/v) uranyl acetate in water. Samples were examined in a Philips CM12 TEM at 80 kV and images captured on Gatan 2kx2k fast scan and First Light high sensitivity CCD cameras.
Cryoshadowing: aliquots of protein were diluted to 10 µg/ml in 10 mM ammonium bicarbonate (pH 7.5) and the sample passed through a G50 spin column equilibrated with the same buffer to remove residual dialysis buffer. An aliquot (3 µl) was adsorbed to 400 mesh nickel grids covered by glow-charged carbon foils for 1–3 minutes. An FEI (Hillsboro, Oregon) Vitrobot Mark IV computer controlled robotic freezing instrument was used to blot away excess buffer and quickly freeze the sample by plunging the grid into liquid ethane chilled in liquid nitrogen. The samples, maintained at liquid nitrogen temperature, on a magnetic table were transferred to a liquid nitrogen cooled rotating stage in an oil-free, scroll/turbomolecular pumped vacuum pumping system based on the Balzers 300 instrument. Freeze drying at −110 °C for 4 hrs followed. The grids were then shadowed by rotary tungsten coating at 1 × 10−7 torr without breaking the vacuum. Samples were examined in an FEI Tecnai 12 TEM at 40 kV and images captured using a Gatan 2kx2k fast scan CCD camera. Adobe Photoshop was used to adjust and invert images and to arrange the images into panels for publication.
ORF6 was identified previously by its sequence similarity to EBV Balf2 protein . Sequence comparison of ORF6 to ICP8 is shown in Fig. 1. ORF6 has a 23% sequence identity to ICP8 at the amino acid level. As expected the sequence similarity between ORF6 and Balf2 was higher, 41% at the amino acid level (not shown). To understand whether ORF6 is also biochemically similar to the other DNA binding proteins in this family, the gene was cloned from KSHV genomic DNA and expressed as a His-tagged protein in insect cells. The protein was purified by affinity chromatography. The cloning procedure and baculovirus protein production is described in the Materials and Methods. From 250 ml of infected Sf21 insect cells, we purified more than 3 mg of ORF6 with a concentration of 1 mg/ml. The purity of the protein was >95% as determined by SDS-PAGE and Coomassie staining (Fig. 2A).
The physical state of ORF6 was examined by gel filtration chromatography. Purified protein (100 µg) was loaded onto a Superdex 200 column and the elution profile followed by measuring the UV absorption at 280 nm and by SDS-PAGE with Coomassie blue staining (Fig. 2 B,C). Both methods revealed that the bulk of the protein appeared in the column void volume with lesser amounts eluting at volumes consistent with monomers to tetramers. This indicated that under these conditions, most of the protein is present either as high molecular weight oligomers, or that the protein had formed aggregates during the purification. The results were further analyzed by negative staining TEM as described below.
ORF6 is large enough (126 KDa) so that single protein monomers should be visible when the protein is imaged by TEM using negative staining or shadowcasting. To examine the physical state of ORF6, the protein preparation used for gel filtration was incubated overnight at 21°C in 20 mM Tris pH 7.4, 50 mM NaCL, and 1mM DTT. Aliquots (unfixed) were adsorbed to thin carbon foils supported by copper mesh grids and stained with 2% uranyl acetate and examined (Materials and Methods). This revealed fields containing a mixture of small particles and very long thick filaments. Fig. 3A shows a field consisting of the small protein particles which were relatively uniform in size suggestive of a single protein viewed in several different projections. Measurement of the dimensions of individual particles revealed an average diameter of 11 ±0.12 nm (n=196). For comparison, in the crystal structure of the 121 KDa C terminal truncated ICP8 monomer, the protein measures 10.9 nm by 14.6 by 16.8 nm . This suggests that the ORF6 particles visualized here represent protein monomers.
In the fields of ORF6 protein, many long stiff filaments were observed by negative staining and these could also be seen when the protein was fixed with glutaraldehyde and prepared for EM by rotary metal shadowcasting (not shown). The (unfixed) negative stained filaments had uniform diameter and a very regular helical twist along the filament axis (Fig. 3F). Frequently the filaments were observed to fuse back on themselves to form circles or lariat structures (Fig. 3F inset). This circularization may self-limit the length to which the individual filaments can grow. Measurement of the filament width (n=>70) revealed a value of 13.7 ±0.2 nm. The long axial twist had a repeat value of 42.9±0.4 nm (n=57), with the helical striation making an angle of ~24 degrees to the axis of the helix. The filaments were often very long with some measuring as much as 2600 nm in length. Following even longer incubations the protein was divided between small, apparently monomeric particles, and very long filaments with few very short filaments. This suggests that once a filament nucleates that it grows rapidly in length. The kinetics of filament dissociation into monomeric particles is, if this occurs, unclear.
Using EM as an assay for filamentation, different incubation conditions including temperature, addition of monovalent and divalent cations, incubation time, and the presence of reducing agents were examined. Incubation for 1–4 hours resulted in what appeared to be partially formed filaments that collapsed during uranyl acetate staining, as contrasted to the very well formed stable filaments observed upon incubation for 8 or more hours. While the addition of magnesium ions had been found necessary for ICP8 filamentation , magnesium was not required for ORF6 filamentation. However the presence of a reducing agent (DTT) was crucial in observing consistent filament formation.
ORF6 was cloned based on its homology to Balf2, a known single strand DNA binding protein. To ensure that the filaments observed here do not represent ORF6 binding to a small amount of contaminating DNA which we had not detected in the preparations, an aliquot of ORF6 was incubated with 2 units of DNase1 for 15 min on ice and then passed over a gel filtration column as in Fig. 2. No change in the elution profile was observed arguing that the high molecular weight complexes represent protein self-filaments and not DNA-protein complexes. Examination of the samples by negative staining shawed that the very long filaments remained, arguing that these filaments do not depend on ssDNA for assembly. This conclusion was bolstered by several further observations. First, treatment of a portion of the preparation with SDS and Proteinase K followed by analysis of the product for any residual DNA by optical density measurement showed that DNA could amount to less than 1% of the preparation by mass. Moreover, in work to be described in detail elsewhere, incubation of ORF6 with M13 ssDNA leads to the formation of very distinct DNA-protein filaments that are easily distinguished from the protein-only filaments by their thinner width and lack of a long gentle helical twist. An example of a field with both filament forms is shown in the supplemental figure (Fig. S1).
A further argument against the absence of DNA in the ORF6 filaments was the observation of protein filaments that had dissociated locally on the EM grid into scattered chains of individual protein monomers with no evidence of DNA between them (see Fig. 3F). Some of these images also suggested that the filaments might be composed of two protein threads rather than a single protein helix. This would parallel findings with ICP8 protein which forms a two-start helix consisting of two protein chains twisted around themselves . To further investigate this, the ORF6 filaments were prepared by an EM preparative method we termed cryoshadowing.
A gentle means of preparing macromolecules for EM visualization termed cryoshadowing, which is a variation of rapid freezing and freeze-drying, has been developed in this laboratory. These methods provide a means of avoiding chemical fixation, exposure of the sample to organic solvents, low pH solutions of heavy metal compounds (such as uranyl acetate) or air drying. In this method the sample is adsorbed to the thin carbon support and the excess liquid blotted away followed by immediate plunging into liquid ethane using an FEI Vitrobot freezing robot. The sample is then transferred to an ultrahigh vacuum system pumped with non-oil containing pumps where the sample is slowly freeze-dried followed by rotary shadowcasting with tungsten prior to breaking the vacuum (Materials and Methods).
The preservation of macromolecular structure by very rapid freezing, freeze-drying or freeze-etching, and finally metal shadowcasting rests on a long history of freeze-etch studies. This approach was extended by Heuser [23–25]. His method involves freezing macromolecules suspended in a slurry of mica chips by impact onto a metal block chilled in liquid nitrogen or helium followed by etching in a vacuum and rotary shadowcasting. In our approach we take advantage of the FEI Vitrobot robotic freezing instrument which freezes samples applied to an EM grid very rapidly under highly reproducible conditions including humidity and temperature. By freezing samples adsorbed to the EM grid, we avoid the additional steps of generating replicas and can shadowcast lightly with tungsten would be dissolved by the acids used to clean replicas. We have found the method to be highly reproducible from one grid to the next, and 10 different samples can be prepared in a single run. In work to be described elsewhere it has been possible to freeze cells grown directly on the EM grid using this approach with excellent results. Thus while the method may or may not offer higher resolution than previous methods, it represents an important step in providing a more reproducible and rapid method for preparing samples by this approach and one that can be easily combined with tagging methods.
Fractions from the Superdex 200 sizing column (Fig. 2) containing long filaments, as seen by negative staining, were prepared by cryoshadowing. Long protein filaments were observed similar in length and appearance to the filaments observed by negative staining and were present in abundance (Fig. 4 A, B). Frequently regions along the length of the filaments were observed in which two separate protein chains could be observed twisted around each other (Fig. 4 C, D).
In this study, we purified the KSHV DNA binding protein ORF6 and examined its physical state using TEM and gel filtration. The protein was expressed and purified from insect cells. Size exclusion chromatography demonstrated that it exists as a heterogeneous mixture of small oligomers ranging in size from monomers to what are likely to be tetramers and then very large species beyond the Superdex 200 resolution limit. EM analysis revealed both species, with the smallest particles having a size expected for monomers. The very large molecular weight material was present in the form of long protein filaments, which form in the absence of DNA. Using a gentle EM preparative method involving rapid freezing and freeze-drying, images of the filaments were obtained that support a model in which the filament consists of two protein chains wrapped about each other.
Formation of the long ORF6 filaments required prolonged incubations, however less ordered filaments were seen after a few hours of incubation. While HSV-1 ICP8 and KSHV ORF6 both form these long protein filaments, ICP8 filamentation requires magnesium while ORF6 filamentation does not and while reducing agents were essential for consistent ORF6 filament assembly they are not needed for ICP8.
Whether these filaments are also engaged in an active dissociation into monomers and small oligomers was unclear, however incubation at 4 °C was seen to push the distribution toward smaller forms arguing that the filaments will dissociate. EM inspection did not reveal a significant number of short protein filaments in the mixtures, which are seen in preparations of RecA protein [26, 27]. With RecA, these short polymers were presumed to nucleate the formation of much longer filaments [26, 27]. If the ORF6 filaments dissociate or grow from the filament ends, then the circular species in which two ends have fused would be blocked for further elongation or dissociation and one would expect to see a build-up of circles over time. Future studies using other physical methods will be needed to probe the nature and kinetics of filament formation.
The filament formed by ICP8 in the absence of DNA measures 18 nm in diameter with a 25 nm helical repeat . It was also determined to be a two-start left handed filament formed by two protein chains wrapping about each other . We measured the ORF6 filament to be 14 nm in diameter with a longer, 43 nm helical repeat. We have not yet assigned handedness for the ORF6 filament. Since ORF6 has higher sequence similarity to BALF2 than ICP8, it will be interesting in the future to determine if full length Balf2 protein is able to form filaments in the absence of DNA. In a previous study  we found that neither the C terminal truncated form of ICP8 nor the C terminal truncated Balf2 protein could form protein-only filaments, but both do bind single strand DNA.
ORF6 and ICP8 form long helical protein filaments in the absence of DNA, and for ICP8, the filaments are similar to those formed along single-stranded DNA. It has generally been assumed that in vivo, the DNA binding proteins polymerize from a pool of free monomers onto the DNA to generate the helical filaments that catalyze strand annealing and recombination. However in the case of KSHV, for example, the high local concentration of ORF6 in the nuclear replication bodies raises an alternative scenario. Here self-assembled ORF6 filaments are proposed to be present in the replication foci where they may provide a structural scaffold upon which other viral and host proteins involved in replication and recombination might bind. Rather than assembling filaments de novo along the viral DNA, DNA may meld into the pre-assembled filaments, an event which could greatly speed steps of strand annealing and recombination. This is a very different perspective on how these proteins would catalyze DNA transactions from the classic view in which the proteins assemble from a pool of free monomers. Further experiments will be required to probe this hypothesis.
This work was supported by grants from the NIH to J.D.G (GM31819, CA19014) and to B.D (CA096500).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.