Nanotechnology of the Virus Capsid
(A rendition of a scanning electron micrograph of nanofabricated silicon subunits and an assembled silicon capsid)
A Fellowship proposal aimed towards
(1) modifying virus capsids for useful purposes,
and (2) the rational design and production of
self-arranging nanoparticles from first principles.
By Ranjan V. Mannige
2011
SUMMARY: Spherical capsids are nanometerscale virus armors formed from a large number
of self assembling protein subunits. Currently,
scientists are attempting to utilize capsids as
vaccines,
molecular scaffolds, and other
medicine-related applications. I hope to expand
on such traditional areas of nano-biotechnology
via rational protein interface modification, while
applying our recent understandings of theoretical
capsid design (Fig. 1) to the construction of
completely artificial capsids by solid state
techniques.
Gleaning insights from spherical capsids—nanometer-scale protective virus shells formed from a number of
protein subunits—may allow for unique solutions to problems in biomedical and materials sciences. Possible
examples of such applications include capsids designed to (1) attack cancerous cells, (2) cleanse greasestricken arteries, and (3) pose as scaffolds for hydrogen gas producing nano-factories. All of these
applications are viably unrealized but approachable, especially with a thorough understanding of the capsid.
Our studies showed that the structure and function of a large number of natural capsids are underlined by
simple mathematical rules, enabling the compilation of canonical capsid design principles. It is my hope that,
with our new theoretical understandings, many new and exciting applications will soon be realized.
My first interest lies in manipulating natural capsid subunits in two specific ways: (1) by protein
modification, which will involve the modification of existing subunit interaction sites to form non-native
assemblies, and (2) by assembly interference, which will
involve the design and production of peptide fragments that
impede the formation of infectious capsid forms by strategic
interface modification (supporting section 1.2)—a starting
point for design of anti-viral agents (not just vaccines). In
addition to exploring the predictive value of our geometric
rules, such techniques will allow us to expand the repertoire
of possible protein arrays to be used in the material
sciences,
and
produce
rational
vaccine
development
pipelines.
My second interest lies in exploiting the geometric
nature of virus capsids by designing artificial ones from
materials like silicon and porcelain. The need for biomimitic
nanostructures is especially important as replacements of
natural capsids, since proteins are linear folded polymers
that, although efficient at what they do, are notoriously difficult to commandeer by protein design or
modification techniques; certainly, the modification of subunit-subunit interactions discussed above are
possible and useful, but their utility is limited, and drastic changes to protein functionality has been shown
to be highly unsuccessful. I propose an iterative cycle beginning with theoretical capsid design and ending
with the mass produced nanoassemblies displaying novel function (supporting section 1.3).
These goals cannot be achieved alone but through exciting collaborations with angstrom and
nanometer-oriented scientists (especially Professor Whitesides and colleagues at Harvard), which I believe,
will allow for the confluence of theory, biology and nanofabrication technologies into solutions crucial to the
rapid design and production of therapeutics and assemblies that may better the human situation.
These aims will be pursued in concert with further theoretical studies on biological design criteria in
collaboration with my current colleagues and mentor, which, I hope, will further enable fluid translations of
theories and ideas into discoveries and solutions.
- Ranjan V. Mannige
Supporting material for the proposal titled “Nanotechnology of the Capsid” by Ranjan V. Mannige.
Introduction: The Emergent Future.
``Emergent features'', a phrase used to describe how the components of a system do not easily predict the
system's behavior, is an idea that has touched the face of diverse fields such as macro economics, chaos
theory, condensed matter physics, cell biology, and nanotechnology. Often, the term is used to explain the
inability to understand and predict systems by just looking at the system's components. From theory
(Supporting Section 2) we show that the emergent features arising from capsid subunit interactions are not
only predictable from first principles but these understandings may possibly be harnessed for medical and
industrial uses discussed in the main text and Supporting Section 1. Feynman's dream is not far away, and
the coming decade will dictate its direction. I am excited to hold an oar.
Supporting Section 1: Methods.
We discuss two ways in which molecular biological and inorganic materials may be utilized in creating nonnatural assemblies that may be further used as scaffolds for functional derivatization.
1.1 The Modification of Natural Capsid Subunits (Fig. 2A). The first method will focus on modifying
existing canonical capsid proteins of trapezoidal shape to form capsids of unnatural sizes. Such modification
will be immediately useful for two reasons: (a) to test our understanding of virus capsid assembly
determinants gathered from geometric considerations, and (b) to develop and expand the repertoire of tools
available to bio-nanotechnology and vaccine development.
We propose to rationally modify capsid size by engineering the protein to interact with its neighbors
with non-native dihedral angles (Fig. 2A). This method is centered around the theory that changing specific
subunit-subunit interaction angles will result in non-infectious but still immunogenic capsid sizes (see section
2.2). The ``protein engineering'' step illustrated in Fig. 2A will involve the modification of specific interaction
angles by computationally forcing the desired dimer conformation (angle) and then performing a
combinatorial computational search for protein sequences that display the lowest free energies in that
conformation. Such searches have been successfully performed on other protein systems ( Bolon et al., 2002;
Sterner et al., 2008; Jiang et al., 2008).
1.2 Creating peptide fragments that impede capsid formation (Fig. 2B). The second procedure will
involve producing peptide fragments obtained from the capsid subunit-subunit interfaces, that, when
modified with proline brackets (denoted as flanking `P's in Fig. 2B), compete with the native subunit-subunit
formation. This form of fragment-based therapeutics has been successfully used by Kini and colleagues
against viper venom (Kini, 1998). Aside from this method, theory obtained from geometric principles will be
used to determine the interfaces (``soft spots'') that, when targeted, will most efficiently disrupt native
capsid assembly [this technique will be based on the notion of hexamer complexity and dihedral angle
diversity, that was used in (Mannige and Brooks, 2009b) to explain
elusive evolutionary constraints on the capsid].
From a medical perspective, the two methods will allow us to
rationally
design
and
then
produce
vaccines
with
method
one
(attenuated capsid shells) and antivirals with method two for presently un-treatable viral infections posed by,
for instance, arenaviruses (e.g., lassavirus), flaviviruses (e.g., dengue virus), filoviruses (e.g., ebola virus),
etc.
1.3 Creating inorganic virus-like particles. The steps involved in producing completely artificial capsids
will require iterations of the following: (1) subunit design, (2) subunit fabrication and (3) testing.
1.3.1 Design. From our geometric understanding of the requirements of subunit-subunit interactions, we
can design a trapezoidal subunit with etched interfaces (Fig. 3), where geometrically complementary
interfaces will interact to form pentamers, hexamers, dimers and trimers – components of the final capsid.
Other features must also be imposed such as a mechanism to prevent van der Waals driven subunit-subunit
stacking by introduction of a cavity in the
subunit (Fig. 3, right), etc. Finally, the
angles with which the subunits interact
can easily be modeled into the subunit by
modification of interface angles whereby
closed assemblies of only specific sizes
will form.
1.3.2
Fabrication: After obtaining a
sufficient subunit design (Fig. 4B), we will
employ photolithography techniques to
create
a
master
polydimethylsiloxane
stamp
(PDMS)
(Fig.
4D)
driven
followed
by
soft-lithography
techniques pioneered by Whitesides and colleagues (Xia and
Whitesides, 1998) to produce subunits in a quick and cheap
manner (the turnaround time for subunit production may be
~24 hours from design, while the master can be used to
produce a large number of cheap PDMS stamps reducing the
cost to mass produce subunits).
The size scale of the subunits will be contingent upon
available
lithography/soft-lithography
technologies.
Two
specific methods for subunit production from PDMS stamps is delineated in Fig. 5. Especially important in this
step is the testing of a variety of subunit materials such as plastics, porcelains and silicates for properties
conducive to proper assembly.
1.3.3 Testing. We propose to use light scattering and small angle neutron scattering experiments to
characterize the assembly solution structure/ensemble, which will be further analyzed using scanning
electron microscopy and atomic force microscopy.
Once the development of such scaffolds is complete, the next step (and my final goal) is to help
develop nano-to-angstrom resolution techniques for subunit modification, allowing for functional motifs to be
specifically engineered into the capsid (both biological and artificial) with high orientational fidelity.
Section 2: The Theory of Capsid Design.
Theory crucial to capsid design and useful to nanotechnology is described. All theoretical developments
discussed arise from a simple outcome: that capsids may be represented by monohedral tilings—polyhedra
whose faces are identical in shape—which is possible because most natural capsids are formed from
similarly shaped and often chemically identical protein subunits and display very few subunit-subunit
overlaps and capsid holes (Mannige and Brooks, 2008). With the help of monohedral tilings, we can surmise
the following from topology, geometry and numerical analysis:
2.1 Subunit Properties. Capsid subunits must have five edges (i.e.,
five interacting neighbors) represented commonly by the ``bisected
trapezoid'' (Fig.6; also see Fig 5 in Mannige and Brooks, 2008 and Fig
3 in Mannige and Brooks, 2009). Those edges must partake in three
distinct subunit-subunit interactions (dotted arrows in Fig. 6, left)
allowing for the formation of dimers, trimers, pentamers and hexamers that come together to form an
icosahedrally symmetric structure (the capsid). We will start with designing artificial capsids by mimicking
this trapezoidal subunit shape and bonding pattern. However, our work also showed that the bisected
trapezoid and its relatives are theoretically ``size invariant'' and may be assembled into capsids of all
permitted sizes (Mannige and Brooks, 2008), which behooves the need to explore the determinants of capsid
size specificity.
2.2 Hexamer Shape Defines Capsid Size. Capsids of various sizes are formed from groupings of 12
pentamers and a variable number of hexamers, that are formed from identical interactions (``w-z''
interactions in Fig. 2) that theoretically allow for an infinite range of capsid sizes (Horne and Wildy, 1961;
Caspar and Klug 1962). We find that, due to capsid monohedrality, capsids of different sizes may be formed
by imposing specific shapes onto the capsid hexamers dictated by intra-hexameric subunit-subunit
interactions [Fig. 7 shows a 180 subunit T=3 capsid and a 240 subunit T=4 capsid that have specific
hexamer shapes and dihedral angle profiles; see (Mannige and Brooks, 2009) for such size-specific geometric
traits]. Knowledge of this will allow for the design of subunits whose size-specificity is imparted by encoding
specific dihedral angle values at specific hexamer angles.
2.3
Additional
Capsid
Design
Criteria.
From
sections 2.1 and 2.2, capsids of any size may be
theoretically created.
However, from a geometric
perspective, we find that some capsids of specific size
(``h,k>1
capsids'')
display
complicated
design
``blueprints'', and are drastically underrepresented in
nature [see Fig. 3 in (Mannige and Brooks, 2009b)].
Knowledge of such design properties, aside from
explaining the geometric pressures on capsid evolution, will allow for more informed design of
artificial
capsids.
Other properties of geometric nature that may help in designing antivirals and elaborating on
assemby mechanics are the capsid's size-specific properties of rigidity (and associated events in some virus
life cycles) and the geometric requirement of crucial auxiliary proteins (Mannige and Brooks, 2009).
References
Bolon,D.N., Voigt,C.A. and Mayo,S.L. (2002) De novo design of biocatalysts. Curr Opin Chem Biol. 6:125-9.
Caspar,D.L. and Klug,A. (1962) Physical principles in the construction of regular viruses. Cold Spring Harb Symp Quant
Biol. 27:1-24.
Horne,R.W. and Wildy,P. (1961) Symmetry in virus architecture. Virology. 15:348-73
Jiang,L. et al. (2008) De novo computational design of retro-aldol enzymes. Science. 319:1387-91.
Kini,R.M. (1998) Proline brackets and identification of potential functional sites in proteins: toxins to therapeutics . Toxicon.
36(11):1659-70
Mannige,R.V. and Brooks,C.L. (2008) Tilable nature of virus capsids and the role of topological constraints in natural
capsid design. Phys Rev E, 77:051902.
Mannige,R.V. and Brooks,C.L. (2009) Geometric considerations in virus capsid size specificity, auxiliary requirements, and
buckling. PNAS USA, 106(21):8531-6.
Mannige,R.V. and Brooks,C.L. (2009b) Periodic table of virus capsids: implications for natural selection and design.
Submitted to PloS One.
Sterner,R., Merkl,R. and Raushel,F.M. (2008) Computational design of enzymes. Chem Biol. 15:421-3.
Xia,Y. and Whitesides,G.M. (1998) Soft Lithography. Angew. Chem. Int. Ed. 37:550-75