Virus Particles as Templates for Materials Synthesis

Research News

Virus Particles as Templates for Materials Synthesis** By Trevor Douglas* and Mark Young 1. Introduction The interface between biology, chemistry, and materials science has provided inspiration for novel approaches to the formation of materials. Increasingly, organized biomolecular architectures are being used as templates for the precise patterning of inorganic materials in a biomimetic approach to materials synthesis. For example bacterial S-layers,[1] lipid assemblies,[2] DNA,[3] and multicellular superstructures[4] have been used to direct the patterning and deposition of inorganic materials. In addition, self-assembled protein cage structures[5,6] provide spatially well defined host systems which can be used for the templating of guest materials having complementary size and shape.

2. Previous Investigations on Mineralized Protein Cages The protein cage derived from the iron storage protein ferritin has previously been very effectively used in the synthesis of nanophase materials.[5] In the biological mineralization reaction of ferritin, an enzymatic ferrous oxidase activity coupled with a protein cage structure having electrostatically dissimilar interior and exterior surfaces, work together to spatially control mineral deposition.[7,8] Thus, soluble ferrous ion enters the protein cage and is catalytically oxidized to the less soluble ferric ion. Ferric ions undergo hydrolysis to form a ferric oxyhydroxide mineral (ferrihydrite), constrained in size by the volume of the protein cage. Even in the absence of enzymatic activity, the cage structure can be used as a constrained reaction environment for mineralization.[9] This is due to the nature ± [*] Prof. T. Douglas Department of Chemistry Temple University Philadelphia, PA 19118 (USA) Prof. M. Young Department of Plant Sciences Montana State University Bozeman, MT 59107 (USA) [**] This work was funded in part by a grant from the National Science Foundation (CHE-9801685). The authors are grateful to Prof. Timothy S. Baker for permission to use the images in Figure 1. Adv. Mater. 1999, 11, No. 8

of the electrostatics of the protein±solution interface which influences mineral deposition. However, the use of ferritin as a protein cage is limited to 12 nm diameter protein spheres with a roughly 7 nm diameter cavity, regardless of the source (animal, plant, bacterial). Therefore, the size of the encapsulated mineral is limited to this size regime (although a smaller iron mineralizing ferritin-like protein has been reported with a cavity of only 4.5 nm diameter[10]).

3. Virus-Derived Protein Cages We have recognized the conceptual similarities between the ferritins and a large class of protein cage structures; viruses. Both systems can be thought of in the context of host±guest interactions. While ferritin functions to store and transport an inorganic polymer of iron oxyhydroxide, virus protein cages (capsids) serve to store and transport organic polymers of nucleic acid (DNA, RNA). It has been our objective to utilize the capsid proteins, which occur in a wide array of sizes and shapes (a few of which are shown in Fig. 1), for constrained materials synthesis. For example, equine herpes virus exists as a 100 nm diameter sphere,[11] cowpea chlorotic mottle virus is a 26 nm diameter sphere,[12] while the tobacco mosaic virus forms stable 300 nm rod-like structures.[13] These assemblies are unique in that they exist as well defined large individual molecules rather than an aggregate ensemble, such as lipid vesicles, and thus provide templates for the nano-engineering of materials having precise and controllable morphologies.

3.1. Spherical Viruses (Cowpea Chlorotic Mottle Virus, Norwalk Virus) We have used the well defined plant virus cowpea chlorotic mottle virus (CCMV)[6] as well as the animal virus Norwalk virus (NV)[14] as model systems for nanophase crystal growth. CCMV capsids are 26 nm in diameter and the protein shell defines an inner cavity approximately 20 nm in diameter as measured by electron microscopy. The dimensions of the capsid cavity define the upper limit for crystal growth of the entrapped mineral. CCMV is composed of

Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1999

0935-9648/99/0806-0679 $ 17.50+.50/0

679

Research News

Fig. 1. Image reconstructions of three examples of spherical (icosahedral) viruses ranging in size from the 100 nm diameter equine herpes virus (on the left), to the 50 nm diameter cauliflower mosaic virus (center), and to the 26 nm diameter CCMV (right).

180 identical coat protein subunits which can be easily assembled in vitro into empty cage structures.[12] Each coat protein subunit presents at least nine basic residues (arginine and lysine) to the interior of the cavity. This creates a positively charged interior cavity surface, which provides an interface which can be utilized for mineral deposition. The outer surface of the capsid is not highly charged, thus the inner and outer surfaces of this molecular cage provide electrostatically dissimilar environments which have been exploited to spatially localize and control the mineralization reactions within the protein cage. Some virus capsids exhibit dynamic structural transitions, induced by defined chemical switches, which can be used to provide unique molecular gating mechanisms to control the containment and release of entrapped materials. CCMV undergoes a reversible pH dependent swelling which results in a 10 % increase in virus dimension.[12] Structural analysis has revealed that this transition is the result of an expansion at the pseudo 3-fold axis of the virus particle which causes the formation of 60 separate 2 nm diameter openings in the protein shell. Under swollen conditions (pH > 6.5) these openings allow free molecular exchange between the virus cavity and the bulk medium. In contrast, in the non-swollen form (pH < 6.5) there is no apparent exchange of large molecules between the cavity and the bulk medium. This transition (gating) allows the virus particle to selectively entrap and release materials from within the central cavity.

3.2. Inorganic Mineralization We selectively mineralized a range of polyoxometalate species (vanadate, molybdate, tungstate) within the CCMV capsid (Fig. 2A) and have also shown the same mineralization reaction (with tungstate) is possible in the NV.[14] The empty virions were incubated with the precursor ions (WO42±, VO3±, MoO42±) at approximately neutral pH. Under these conditions the virus exists in its open (swollen) form and allows all ions access to the cavity. The pH of the virus solution was then lowered to approximately pH 5.0. This 680

Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1999

Fig. 2. Transmission electron micrographs of A) assembled CCMV virions mineralized with cores of (NH4)10H2W12O42 showing electron dense mineral particles. B) Negatively stained samples of mineralized CCMV, showing the presence of the protein coat surrounding each mineral particle. C) Negatively stained sample of polymer encapsulated virions showing no stain intrusion to the interior of the particle. D) Fibers of TMV, mineralized on the outside with a thin coating of iron oxide.

induced two important complementary effects. i) The inorganic species underwent a pH dependent oligomerization to form large polyoxometalate species such as H2WO4210± which were readily crystallized as ammonium salts, ii) the viral capsid particle underwent a structural transition in which the pores in the protein shell closed, trapping crystallized mineral or mineral nuclei within the virus. Conditions were adjusted so that mineralization occurred selectively only within the viral capsid and no bulk mineralization was observed in solutions containing assembled viral capsids or virus-free controls. The mineralized virus particles were purified by centrifugation on sucrose gradients and subsequently examined by transmission electron microscopy (TEM). Electron dense mineral cores were observed commensurate in size (18 nm in diameter) and shape with the internal diameter of the virus particle (Fig. 2A). Negative stain of these samples demonstrated the presence of the intact capsid protein shell surrounding these cores (Fig. 2B). Crystallization of the polyoxometalates occurred in a spatially selective manner within the virus cavity. We believe that the crystallization is electrostatically induced at the basic interior surface of the protein. The negatively charged polyoxometalate ions (such as H2W12O4210±) aggregate at this interface facilitating crystal nucleation. Thus, we have suggested that the protein shell acts as a nucleation catalyst in the polyoxometalate mineralization, in addition to its role as a size and shape constrained reaction vessel. While the gating phenomenon exhibited by CCMV certainly facilitates the mineralization and entrapment of these species, it is not an 0935-9648/99/0806-0680 $ 17.50+.50/0

Adv. Mater. 1999, 11, No. 8

Research News

absolute requirement since the NV is not known to swell yet under similar conditions it forms cores virtually identical to those observed for CCMV.

20±30 nm in diameter. The question of whether the central cavity of the virus is also mineralized under these conditions is currently under investigation.

3.3. Organic Polymer Encapsulation

4. Conclusions and Outlook

To investigate the host-guest interaction and the influence of gating on the molecular entrapment, we undertook the pH dependent encapsulation of an anionic organic polymer (poly-anetholesulfonic acid) in CCMV. The polymer was incubated with the empty virion at high pH (7.5) followed by a lowering of the pH to well below the gating threshold (pH 4.5). This resulted in selective encapsulation of the polymer. The polymer loaded virion was isolated by gradient centrifugation and imaged by TEM (Fig. 2c) revealing the intact protein shell surrounding an electron transparent core (i.e., no stain intrusion into the virion interior). Stain intrusion is prevented by the presence of the organic polymer. In addition, the isolated material exhibited the characteristic UV-visible spectrum of the aromatic anionic polymer. Control reactions in which the empty virions were incubated at low pH (under non-swollen conditions) yielded no measurable uptake of the polymer.

The use of virus-derived protein assemblies represents a new direction for the systematic use of a class of supramolecular assemblies for the precise patterning of materials. Control over the dimensions (size and shape) of inorganic materials has now been shown to result from templating interactions with a range of virion architectures. It is apparent that this approach to biomimetic materials chemistry is a general one with a great deal of flexibility both in terms of existing virion libraries and the possibility for site directed mutagenesis to engineer specific additional interactions and architectures.

3.4 Anisotropic StructuresÐTobacco Mosaic Virus Recently we reported the use of the rod-shaped tobacco mosaic virus (TMV) as a template for mineralization.[15] The TMV assembly comprises approximately 2130 subunits arranged as a helical rod around a single strand of RNA to produce a hollow tube 300 nm ´ 18 nm with a central cavity 4 nm in diameter.[13] The exterior protein assembly of TMV provides a highly polar surface, which has successfully been used to initiate mineralization of iron oxyhydroxides, CdS, PbS and silica. These materials form a thin coating over the protein and result in formation of mineral fibers, having diameters in the 20±30 nm range (Fig. 2D). In addition, there is evidence for ordered end-to-end assembly of individual TMV fibers with very high aspect ratios to form mineralized fibers, of iron oxide or silica, over 1 mm long and

± [1] W. Shenton, D. Pum, U. B. Sleytr, S. Mann, Nature 1997, 389, 585. [2] D. D. Archibald, S. Mann, Nature 1993, 364, 430. [3] E. Braun, Y. Eichen, U. Sivan, G. BenYoseph, Nature 1998, 391, 775. [4] S. A. Davis, S. L. Burkett, N. H. Mendelson, S. Mann, Nature 1997, 385, 420. [5] F. C. Meldrum, V. J. Wade, D. L. Nimmo, B. R. Heywood, S. Mann, Nature 1991, 349, 684. T. Douglas, in Biomimetic Approaches in Materials Science (Ed: S. Mann), VCH, New York 1996, pp. 91±115. [6] T. Douglas, M. J. Young, Nature 1998, 393, 152. [7] P. M. Harrison, P. Arosio, Biochim. Biophys. Acta 1996, 1275, 161. [8] T. Douglas, D. Ripoll, Protein Sci. 1998, 7, 1083. [9] V. J. Wade, S. Levi, P. Arosio, A. Treffry, P. M. Harrison, S. Mann, J. Mol. Biol. 1991, 221, 1443. [10] M. Bozzi, G. Mignogna, S. Stefanini, D. Barra, C. Longhi, P. Valenti, E. Chiancone, J. Biol. Chem. 1997, 272, 3259. [11] B. Roizman, in Fields Virology, 3rd ed. (Eds: B. N. Fields, D. M. Knipe, P. M. Howley, et al.) Lippincott-Raven, Philadelphia 1996, pp. 2221± 2230. [12] J. A. Speir, S. Munshi, G. Wang, T. S. Baker, J. E. Johnson, Structure 1995, 3, 63. [13] G. Stubbs, Seminars in Virology 1990, 1, 405. G. Stubbs, in Biological Macromolecules and Assemblies, Vol. I: The Viruses (Eds: A. McPherson, F. Jurnak), Wiley, New York 1984, pp. 149±202. [14] M. Young, T. Douglas, unpublished results. A. Kapikian, M. Estes, R. Chanock, in Fields Virology, 3rd ed. (Eds: B. N. Fields, D. M. Knipe, P. M. Howley, et al.) Lippincott-Raven Publishers, Philadelphia 1996, pp. 783±810. [15] W. Shenton, T. Douglas, M. Young, G. Stubbs, S. Mann, Adv. Mater. 1999, 11, 253.

_______________________

Adv. Mater. 1999, 11, No. 8

Ó WILEY-VCH Verlag GmbH, D-69469 Weinheim, 1999

0935-9648/99/0806-0681 $ 17.50+.50/0

681