Structure-based engineering of papillomavirus major capsid L1: controlling particle assembly
© Bishop et al. 2007
Received: 12 September 2006
Accepted: 08 January 2007
Published: 08 January 2007
Skip to main content
© Bishop et al. 2007
Received: 12 September 2006
Accepted: 08 January 2007
Published: 08 January 2007
The outer shell of the papillomavirus particle is comprised of 72 pentamers of the major capsid L1 protein arranged on a T = 7 icosahedral lattice. The recombinant L1 can form T = 7 virus-like particles in vitro. The crystal structure of a T = 7 papilloma virion has not yet been determined; however, the crystal structure of a T = 1 particle containing 12 pentamers is known. The T = 1 structure reveals that helix-helix interactions, through three helices–h2, h3, and h4–near the C-terminus of L1, mediate the inter-pentameric bonding that is responsible for T = 1 assembly. Based on the T = 1 crystal structure, we have generated a set of internal deletions to test the role of the three C-terminal helices in T = 7 assembly. We have demonstrated that the h2, h3, and h4 near the C-terminal end of L1 are important for the L1 structure and particle assembly. In particular, we found that h2 and h3 are essential for L1 folding and pentamer formation, whereas h4 is indispensable for the assembly of not only T1, but also of the T7 virus-like particle.
Papillomaviruses are non-enveloped DNA tumor viruses. Infection by papillomaviruses is associated with tumorigenesis in experimental animals and in humans (reviewed in ). The virus particle is comprised of 72 pentamers of the major capsid protein (L1) on the outer surface, arranged on a T = 7 icosahedral lattice (reviewed in [3, 5]). A minor capsid, L2, is located internal to the L1 shell. The viral genomic DNA is packaged within the L1/L2 capsid as a minichromosome (reviewed in ).
The recombinant L1 capsid alone can assemble into the virus-like particle structure in vivo and in vitro [4, 6, 7, 10–12, 15]. Thus, the L1 protein contains all the information needed for the particle assembly. The in vitro assembly of L1 occurs spontaneously under high salt or low pH conditions [1, 2]. The sizes of particle assembly can be regulated by N-terminal truncations. For example, HPV16 L1 lacking the first 10 residues assemble into T = 1 particle, but L1 deletions missing nine or fewer residues from the N-terminus assemble into T = 7 particles [1, 2].
By fitting the L1 atomic structure of from the T = 1 particle into the cryo-EM reconstruction of BPV , an atomic model for T = 7 assembly was proposed . In the fitted T = 7 model, the laterally projected C-terminal fold of L1 seen in the T = 1 structure is rearranged to extend into the neighboring pentamers. This C-terminal arm exchange between the neighboring pentamers is similar to the "invading arm" model of the polyomavirus VP1 T = 7 virion structure [8, 13]. However, the roles of the helix-helix interactions observed in the T = 1 structure are not defined in the T = 7 assembly model. In order to understand the functional roles of h2, h3, and h4 in T = 7 assembly, we have made structure-guided internal deletions of these helices on the L1 of two different HPV types. We have demonstrated that h2 and h3 are essential for L1 folding and the production of the pentameric L1 and that h4 is indispensable for the assembly of T = 7 particles.
Because the L1 expression level and solubility could not be easily detected by directly analyzing the cell lysates by SDS-PAGE, we performed partial purification of the GST-fusion proteins from the cell lysates using affinity chromatography. The GST-L1 protein from the supernatant fraction of the lysates was passed through a glutathione resin column. After washing the column with 10× bed volumes of buffer L (25 mM Tris-Cl, pH 8.0, 0.25 M NaCl, 1 mM EDTA, 2 mM DTT), the resin was incubated with thrombin to cleave L1 from the GST fusion.
The proteins bound to the resins before and after thrombin treatment, as well as those in the eluate after thrombin treatment, were analyzed by SDS-PAGE (Fig. 2C). D1 and D2 constructs produced large quantities of soluble GST-L1 fusion proteins (lanes 2 and 5, respectively). D3 produced some GST-D3-L1 fusion (lane 8 in Fig. 2C), but many degradation bands were observed in the fraction before thrombin treatment. These are presumably L1 degradation products that are still fused with GST at the N-terminus of L1. There was almost no detectable full-length GST-D4-L1 fusion (lane 11 in Fig. 2B), indicating that this protein was even more severely degraded than the GST-D3-L1 protein (compare lanes 8 and 11).
After thrombin digestion of the GST-L1 fusions on the glutathione resin, soluble D1- and D2-L1 were eluted from the column (lanes 3 and 6, respectively). For D3-L1, however, no intact protein was present in the elution after thrombin cleavage; D3-L1 may be degraded by thrombin due to an unfolded or misfolded conformation.
The L1 proteins eluted from the glutathione column were analyzed using size-exclusion chromatography on Superdex-200 (Fig. 2D). A single major peak with the elution position expected for a pentamer was present for both D1- and D2-L1. These results demonstrated that h4 deletion mutants D1-L1 and D2-L1 were soluble and stable and that they fold and oligomerize into pentamers. However, when the deletion extended to h3 (in D3) or to h2 (in D4), the proteins were degraded into multiple fragments, before and during thrombin treatment.
The T = 1 particle structure shows that h2, h3, and h4 near the C-terminal end of L1 are the solely responsible for making inter-pentameric contacts required for the particle assembly . There are the obvious differences between the T1 structure and the proposed assembly model of T7 , and the roles of h2, h3, and h4 in T7 assembly are not defined. We have selectively deleted each of the three helices from papillomavirus L1. The biochemical and structural characterization of the deletion mutant proteins demonstrated the important roles of these C-terminal helices in the folding and assembly of L1 capsid.
The deletion mutants of HPV16 L1 lacking h4, h3, or h2 were expressed as GST-fusions in E. coli. The mutant proteins with partial or entire h4 deletion, D1-L1 and D2-L2, were similar to the full-length wild-type L1 in protein expression and behavior by size-exclusion chromatography. Like wild type, D1-L1 and D2-L2 were expressed at quite high levels, and a good fraction of each expressed proteins was soluble and could be purified as pentamers (Fig. 2C). This result demonstrated that h4 is not required for L1 folding and pentamerization.
In contrast, when h3 and h2 were deleted and fusion proteins (GST-D3-L1 and GST-D4-L1) were expressed, little intact GST-L1 fusion protein was detected (Fig. 2C). Instead, multiple smaller protein bands were present in the SDS-PAGE gels. Because only a GST-fusion can bind the glutathione resin of the column used for purification, the smaller protein bands must contain intact GST plus L1 fragments of various lengths due to C-terminal degradation. Thrombin treatment of these GST-fusions did not produce discreet L1 fragments, indicating that both D3-L1 and D4-L1 were unfolded and sensitive to protease degradation. Consistent with this severe degradation of these deletion mutants of L1 is that there are six potential thrombin cleavage sites within the HPV16 L1 sequence and many Arg and Lys residues can serve as cleavage sites for other proteases. Thus, we conclude that h2 and h3 are important for the folding and integrity of HPV L1 structure.
Interestingly, a small helix at the C-terminus, h5 (Fig. 1B), was also shown to be critical for L1 folding [1, 2]. A common feature of h2, h3, and h5 in the x-ray structure  is that all helices have a highly hydrophobic side that packs with the core structure of L1. Disrupting these hydrophobic packing interactions will obviously affect the folding of L1.
The deletion of h4 in another papillomavirus, HPV18, resulted in a mutant that behaved like the HPV16 D2-L1 mutant. This result confirmed that the L1 structure is conserved among different HPV types, which is not surprising considering the amino acid sequences of L1 proteins are highly conserved. We predict that the same D2 deletion of any other papillomavirus L1 should have the same properties.
When the D1-L1 and D2-L2 proteins of HPV16 and 18 were incubated under assembly conditions, pentamer structures, but no T = 7 particles, were visualized by EM examination (Fig. 3C–E). If the deletion mutants assemble into particles, these particles might be unstable to the acidic conditions used during EM sample preparation, which requires uranyl acetate treatment. To rule out this possibility, size-exclusion column chromatography was used to analyze the deletion constructs after assembly treatment (Fig. 4). The size-exclusion analysis showed that, in contrast to the full length L1, the D1-L1 and D2-L2 mutants lacking h4 failed to form assemblies larger than the pentamers, confirming the results from EM study.
In conclusion, we have demonstrated that the structural elements h2, h3, and h4 near the C-terminal end of L1 are important for the assembly of papillomaviruses into particles. The h2 and h3 regions are essential for L1 folding and pentamer formation, whereas the h4 region is indispensable for the assembly of not only T1, but also of the T7 virus-like particle.
The full-length L1 clones of HPV16 and HPV18 in pGEX-2T were used as parent clones for generating the deletion mutants of D1, D2, D3, and D4 as shown in Fig. 2A, 2B. The L1 proteins were cloned downstream of GST to express as GST-L1 fusions in this E. coli expression vector. The standard molecular cloning methods, including PCR amplification, enzyme digestion, DNA end ligation, and transformation, was used to obtain the deletion mutants of L1 as shown in Fig. 2A, 2B. All deletion clones were confirmed by sequencing the whole DNA insert to ensure the correct deletion and the wilt type sequences for this study.
The protein expression and purification of all L1 deletion mutants was carried out essentially as described preciously [1, 2]. Briefly, 0.2 mM IPTG was use to induce protein expression overnight at room temperature (RT). After cell lysis by sonication in buffer L (50 mM Tris-HCl, pH 8.0, 0.2 M NaCl, 1 mM DTT, 1 mM EDTA, 10 mM PMSF), urea (ultrapure grade) was slowly added to the lysate to a final concentration of 3.0 M. The mixture was incubated at RT for one hour with gentle shaking, and then dialyzed against three changes of buffer over an 18 hour period. After centrifugation at 25,000 × g for 75 min, the supernatant was passed through a glutathione affinity column to bind GST-L1 fusions. Usually, 10 ml glutathione resin was used for supernatant from 12 liter cell culture. The column was washed with 10× bed volumes of buffer L to wash away contaminating proteins. At this point, an additional washing step using 10× bed volumes of 2.3 M urea can clean up the small amount of the contaminated groEL. L1 was then cleaved from the GST fusion using thrombin (Sigma, T6634) in an approximate ratio of 100 μg GST-L1 to 1 NIH unit of enzyme. The digestion was carried out at 4°C overnight. L1 was further purified by Superdex-200 (60/16 column) size-exclusion chromatography. During the purification process, keep all the reagents at 4°C whenever possible.
Wild-type L1 can assemble into particles under known in vitro assembly conditions [1, 2]. The assembly reaction was performed by incubating purified L1 protein in assembly buffer (1 M NaCl, 40 mM sodium acetate, pH 5.4) at 25°C for 30 minutes, at a protein concentration of approximately 0.1 mg/mL. For EM analysis, samples were adsorbed to glow-discharged, carbon-coated copper grids (EM Science) and negatively stained with 2% uranyl acetate. Grids were examined in a JEOL 1200 transmission electron microscope at 80 kV and 60000× magnification. For size-exclusion chromatography analysis, samples were injected directly into the sample loop and onto a Superdex-200 column (16/60, Pharmacia) at a flow rate of 1 mL/minute in the assembly buffer. The protein peaks were detected by an UV monitor at a wavelength of 280 nM.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.