Pharmacological Functionalization of Protein-Based Nanorobots as a Novel Tool for Drug Delivery in Cancer

Abstract

The delivery of hydrophobic therapeutic agents to tumors is a challenge in the treatment of cancers. Here, we review recent advances in coiled-coil protein origami and discuss a proposed programmable protein origami structure, switchable by a protein kinase A/phosphatase switch, as an example of functionalization for designing future protein nanorobots.

Introduction

Targeted drug delivery to the site of cancer cells is an important concept to increase pharmacokinetics/pharmacodynamics (PK/PD) properties and to reduce off-target toxicity of hydrophobic cancer therapeutics. Recent reports demonstrate that biological materials, such as proteins and DNA, have the capacity to be hijacked for targeted and programmable drug delivery.¹ DNA complementary base pairs provide a classic and simple tool to design complementary DNA modules that self-assemble to form ultrastructure origami nanorobots. Functionalization of these nanostructures is conducted by loading anticancer agents inside the structure and conjugation of an aptamer sequence outside the structure, which serves as a cancer-targeting domain and also as a molecular trigger for the mechanical opening of the DNA nanorobot to release the therapeutic agent. An alternate strategy is a pH-sensitive DNA origami designed to control the opening/closing of the structure by a pH switch. In the recent first functionalized nanorobot targeting cancer cells, a rectangular DNA origami sheet was built by assembling M13 bacteriophage genome DNA strands using several staple strands. This functional nanorobot was tested in vitro (cell cultures) and in vivo (mice and Bama miniature pigs), demonstrating they are able to deliver thrombin to specific targeting sites (tumor-associated blood vessels) and prevent tumor growth by induction of intravascular thrombosis and tumor necrosis.¹ DNA strands are hydrophilic and negatively charged because of the phosphate moieties, which make them useful tools for loading hydrophilic therapeutics, unless the surface is modified to be hydrophobic. DNA sequences which are composed of base pairs have complementary property. This chemical feature has been used to design modular units that self-assemble and form origami structures, whereas in the case of proteins and peptides, designing complementary units to build origami structures is not as feasible as DNA sequences and is challenging. Despite recent advances in coiled-coil protein origami (CCPO) structures that self-assemble in vitro and in vivo,²⁻⁴ functionalization of these structures to generate programmable nanorobots that act in cells or in vivo systems has not been described yet. While the building blocks of CCPOs, i.e., coiled-coil (CC) motifs, have been used as tools for peptide and hydrophobic drug delivery, the function of these motifs, especially in drug release, was not programmable.^5,6 Furthermore, a single CC structure does not provide a protecting shield for the efficient delivery of therapeutic agents. The self-assembled structure of multiple complementary CC motifs to form a CCPO is an excellent device for targeted drug delivery. We believe that recent advances in protein chemistry and biochemistry provide potential tools to design CCPO nanorobots that (a) sequester hydrophobic therapeutics and (b) act as a programmable robot under the control of enzyme-responsive switches, which provide pharmacological control of their assembly and disassembly as well as drug release. In this viewpoint, we discuss and highlight possible available molecular gadgets appropriate for this purpose and emphasize the importance of designing protein-based nanorobots as future tools for cancer therapeutics.

Coiled-Coil Protein Origami Structures

CCPO structures are complex polyhedron-shaped structures, resulting from the folding of a primary sequence consisting of several structural units; orthogonal modular coiled-coil connected with short disorder peptide linkers. For example, in order to build a tetrahedral shape CCPO, six complementary modular CCs (12 CCs) are required. Each module interacts with its complementary module to form intramolecular CC dimers as the edges of the polyhedron (in this example, tetrahedral) in the 3D structure.²⁻⁴ The order of modules and their position in primary sequence guide self-assembly of a sequence to form CCPO structures. In general, CCs are composed of two or more α-helix units that form a supercoiled bundle via intra-α-helix-to-α-helix interaction. Therefore, the structural units of CCPO are CC dimers of two α-helixes that wrap around each other. The helix–helix dimerization was first characterized by Crick as the knobs-into-holes (KIH) interaction, which still serves as a rational design method for antibody engineering to induce heterodimerization of the heavy (H) chains and generate bispecific IgG antibodies. KIH interaction is defined as hydrophobic interactions between asymmetric hydrophobic residues in which a knob (a hydrophobic residue with a large side chain, such as Met, Ile, and Tyr) from one helix inserts into a hole (a hydrophobic residue with a small side chain, such as Val or Ala) from another helix. Moreover, the CC structure is further stabilized by salt bridges—electrostatic interactions between charged side chains projecting from the wrapped helixes. Therefore, the KIH-based supercoiled structure requires repeats of hydrophobic (h) and polar (p) residues in an order to form helical structures via intra-hydrogen binding of a backbone in each individual helix as well as CC dimerization via interhydrophobic and interelectrostatic interactions.²⁻⁴ The most common helical motif in CCPO design are repeats of a heptamer peptide, known as heptad repeats. A cross section of this CC structure is represented as helical-wheel diagrams, which exhibit heptad repeats (represented as “abcdefg”). Helix dimerization requires a polarity order of “hpphppp” for the heptad repeat (abcdefg) in which “a” and “d” are hydrophobic (h) residues and b, c, e, f, and g are polar (p). Furthermore, “e” and “g” are charged residues that confer the interhelix salt bridges, whereas residues at “b”, “c”, and “f” positions do not form interhelix interactions but are usually α-helix facilitating and control the solubility of the structure (Figure 1a). These residues form the side surfaces of a CC and can form a trigger sequence consisting of unliked charged residues (negative and positive charged residues, such as Glu, Asp, Lys, and Arg) at i, i+3, and i, i+4 positions that facilitate intrasalt bridges (...b_i–4c_i–3--f_i--b_i+3c_i+4...), leading to an increase in the helical propensity as well as the stability of the CC structure.²⁻⁴

Figure 1.

Schematic representation of the CC structure and interaction pattern before and after phosphorylation. CC heptad repeats are also shown by helical-wheel diagrams: (a) General helical-wheel diagram of heptad repeat “abcdefg” (b,c) depending on the location of PKA recognition site; phosphorylation has different effects on helix–helix dimerization (b) that for a motif of “...R ARRGSAR VR...” sequence which represents the CC sequence of “g abcfefg ab”; phosphorylation of Ser (pSer) at position “e” of first heptad repeat stabilizes interhelical salt bridge and, therefore, induces CC homodimerization (c) for a motif of “...K IAALRRK SA...” sequence which represents the CC sequence of “g abcfefg ab”; phosphorylation of Ser (pSer) at position “a” of second heptad repeat destabilizes hydrophobic interaction that direct its heterodimerization with “...E IAALEQE SA...” and, therefore, induces CC disassembly. Helical wheels were generated using DrawCoil 1.0 (http://www.grigoryanlab.org/drawcoil).

The CC structure is found in natural proteins, such as basic leucine zipper domain proteins (bZIP).^7,8 Recent advances in de novo design of these structures provide novel tools in synthetic biology to develop smart and functional CC structures with different applications. Gradišar et al. for the first time in 2013 designed a concatenated orthogonal polypeptide sequence composed of six pairs of complementary interacting CC-forming peptides connected with short peptide linkers that self-assembled and formed a tetrahedral CCPO structure (TET12) constructed by six CC structures as edges.² In a recently published paper, Ljubetic et al. improved the solubility limitation of TET12 by engineering surface residue (b, c, and f) substitution of negatively charged residues to generate soluble versions of TET12, i.e., TET12S and TET12SN. These soluble CCPOs undergo reversible folding under physiological conditions. The self-assembly of TET12S was confirmed in both HEK293 cells and mice liver.⁴ These results are promising and show the feasibility of the de novo design of CC-based nanostructures that undergo proper folding in vitro and in vivo and have potential applications for future drug delivery. However, they still lack a critical feature which is controlling their assembly/disassembly to become a fully functionalized nanorobot that is programmable and under control by a physiological switch.

Structural Effects of Phosphorylation on α-Helical Structures

There are natural modifications that cause conformational changes of proteins and peptides in cells, including post-translational modifications (PTMs) of specific amino acids, which change the conformation of peptides and proteins in cells⁹ and can be recruited as promising programming switches for assembly and disassembly of typical coiled-coil origami structures by pharmacologically controlling catalyzing enzymes. In this regard, to design a functionalized structure, the PTM must occur on the α-helical modules to disrupt the α-helical structure, facilitate α-helical formation in the individual helices, or to control dimerization of helices in CC structures. Most known PTMs target random coil or β-turn structures; however, phosphorylation signatures can be incorporated into α-helical structures and CCs to enable phosphorylation by different kinases that phosphorylate serine or threonine (Ser/Thr) residues on peptides and proteins.^7,8,10 The effect of this modification on the α-helical conformation is diverse and depends on the sequence of the peptide and site of phosphorylation, i.e., within the α-helix or N-terminal or C-terminal of the α-helix. Andrew et al.(11) investigated the effect of phosphoserine position (N-terminal, C-terminal, and central position) on the stability of an alanine-based α-helical peptide (Ac-AAAAAQRAAAARAGY-NH₂), assessing both the helix content and the stability upon pH titration. The result revealed that phosphoserine at or near N-terminus (N-cap or substituted within the three residues of the α-helix) stabilizes the α-helix structure more than phosphoserine at the central position of the peptide, whereas phosphoserine at the C-terminus destabilizes the structure.¹¹ This report exhibits the feasibility of using phosphorylation/dephosphorylation as a switch to stabilize or destabilize α-helix structures, which are building blocks of CCPOs.

Protein Kinase-Inducible Switches for Coiled-Coil Structures

To be able to use phosphorylation/dephosphorylation as a switch in a CCPO nanorobot, a reversible enzymatic reaction is required to install or remove phosphate to specific sites within the modular CC units. Possible mechanisms to install this phosphorylation switch to control CCPO nanorobot assembly and disassembly include substitution of phosphorylation sites in the heptad repeats to affect the helix–helix interaction and dimerization of CCs. Including the PKA recognition site “RRXS” within the heptad sequence has been tested as a switch to control CC dimerization. Depending on the position of the phosphorylation site in the heptad repeat, this modification controls assembly/disassembly of CCPOs through stabilizing/destabilizing of (a) the interhelical salt bridges or (b) the hydrophobic interaction that directs dimerization of helixes.^7,8,10 Specifically, in a motif of “...R ARRGSAR VR...”, which represents the CC sequence of “g abcfefg ab”; phosphorylation of Ser (pSer) at position “e” of first heptad repeat stabilizes an interhelical salt bridge and, therefore, induces CC homodimerization (Figure 1B).^7,8 While for a motif with the “...K IAALRRK SA...” sequence, which represents the CC sequence of “g abcfefg ab”; phosphorylation of Ser (pSer) at position “a” of second heptad repeat destabilizes the hydrophobic interaction that directs its heterodimerization with “...E IAALEQE SA...” and, therefore, induces CC disassembly (Figure 1C).¹⁰ These reactions are reversible, with the phosphatase able to reverse the stability of CC by removing the phosphate from the targeted Ser.

Conclusion and Discussion

Drug delivery is an important concept in targeted drug development, especially in the case of nonsmall molecule therapies that have large molecular weight and are not soluble in body fluids due to their nonpolar structures. The development of biomolecular superstructures that surround these therapeutic agents as a cage in order to protect them from degradation pathways and sort them to the target organs is an emerging concept in drug development.¹ Functionalization of these nanostructures to build a nanorobot requires controlling the open/close conformation of these structures to be able to induce upload/release of the drugs. In the case of protein-based origami CCPO structures, controlling the open/close conformation has not been reported and characterized. Here, we suggest the use of a phosphorylation-inducible switch as a key to control the release of drugs at the site of the targeted organ. These phosphorylation-inducible switches have been tested to control assembly/disassembly of CC conformations, which are subunits of CCPOs. These known and characterized CC motifs^7,8,10 can be used to design complex CCPO nanorobots under control by pharmacological inhibition or activation of PKA or phosphatases to switch on/off the nanorobot to load or release therapeutic agents (Figure 2). This novel approach will improve the drug delivery modalities and provide a basis for designing new peptide-based nanorobots that can be pharmacologically controlled by the administration of FDA-approved drugs that inhibit or activate PKA or phosphatase activities. Interestingly, PKA is a biomarker and a target in cancer therapy,¹² thus anti-PKA antibodies or aptamers can be conjugated to the CCPOs to guide the nanorobots to tumor areas. In this case, administration of PKA inhibitor/activator leads to the release of therapy in the desired cancerous organs which express high levels of PKA.

Figure 2.

Schematic representation of open/close conformation of a CCPO origami structure possessing PKA recognition sites within the pairs CC helixes as a switch to control its assembly/disassembly. This novel pharmacologically functionalized nanorobot has the potential for future drug delivery to control the rate of therapeutic release over time (pharmacokinetics).

Glossary

Abbreviations

bZIP

basic leucine zipper domain proteins

CC

coiled-coil

CCPO

coiled-coil protein origami

KIH

knobs-into-holes

PK/PD

pharmacokinetics/pharmacodynamics

PKA

protein kinase A

Author Present Address

^§ I.T. contributed to this paper at the Cellular Energetics Program of the Kavli Institute for Theoretical Physics; his current affiliation is C2i Genomics, 180 Varick St, sixth Fl New York, NY 10014

Author Contributions

The manuscript was drafted, revised, and edited by O.T. and I.T. O.T. prepared all the graphical items.

The authors declare no competing financial interest.