SYSTEMS COMPRISING NON-IMMUNOGENIC AND NUCLEASE RESISTANT NUCLEIC ACID ORIGAMI DEVICES FOR MOLECULAR COMPUTATION

FIELD OF THE INVENTION

[0001] The present invention relates in general to systems of molecular origami devices and their uses for molecular computing, quorum sensing and programmable drug delivery.

BACKGROUND OF THE INVENTION

[0002] Inspired by communication mechanisms occurring in biological systems, molecular communication (MC) is a novel interdisciplinary paradigm in which the research areas of biotechnology, communication technology and nanotechnology converge [Hiyama, S. et al., 2005]. The rapid advances in these fields have brought about the miniaturization of mobile machines and robots down to nanometer dimensions. At this scale, a bio-inspired nanorobot (nanobot in short) is the most basic functional unit, consisting of nanoscale components, that is capable of performing specific tasks such as computing, data storing, sensing or actuation [Llatser, I. et al., 2012]. Such tasks can be executed through the capacity of receiving inputs and generating outputs, which in a molecular communication context requires transceiver capabilities, as nanobots receive information by reacting to specific molecules and broadcast information by releasing other molecules, according to predefined parameters.

[0003] To date, different MC systems have been proposed depending on the way message molecules propagate from transmitters to receivers [Pierobon, M. et al., 2010]. These systems have been categorized into three classifications: walkway-based, diffusion based and flow-based MC. For each category, several efforts have been undertaken to design systems and laboratory condition feasibility has been investigated [Hiyama, S. et al., 2010]. For walkway-based MC systems, a walkway-motor- interaction transport model has been proposed [Hiyama, S. et al., 2009]. In such systems, signal molecules are propagated over protein filaments (i.e. microtubules) via molecular motors [Hiyama, S. et al., 2007; Hiyama, S. et al., 2008(a); Hiyama, S. et al., 2008(b); Enomoto, A. et al., 2006]. Diffusion-based MC is achieved by encapsulating information molecules into vesicles that are emitted into a medium [Moritani, Y. et al., 2006; Moritani, Y. et al., 2007] where they subsequently propagate via diffusion or unpredicted turbulence of the medium. This approach also includes systems utilizing diffusion through gap- junctions between cells [Nakano, T. et al., 2005; Walsh, F. et al., 2010] and long range systems that are envisioned to make use of platforms such as hormones, pheromones, pollen or spores [Parcerisa Gine, L. et al., 2009]. There have also been accounts of a new group of longer range, actively propelled systems utilizing flagellated bacteria [Gregori, M. et al., 2011; Gregorim M. et al, 2010] and catalytic nanomotor [Gregori, M. et al., 2010] systems in which information is encoded in DNA plasmids and transmitted via bacteria or synthetic nano motors. In flow based MC systems, signal molecules are released into a fluid medium where they are guided to their destination via currents or drifts. Such systems offer some of the most biologically realistic scenarios; i.e. as nanomachines can be deployed in flows which introduce a drift for the motion of signal molecules, such as in hormonal communication through the blood stream. Interestingly, even though vesicle and long range hormones, pheromones, pollen or spores systems might be applicable in flow guided MC systems, no research on physical systems employing such propagation systems has been published.

[0004] The fundamental challenge is to devise a biological nanobot platform that can combine the extraordinary characteristics of nano-transceivers with application-orientated functionality. Ultimately such a system will provide the blue print for a scalable single network system that can interconnect large numbers of nanobots to perform complex tasks.

[0005] General efforts in MC research have mainly focused on bio-inspired propagation systems and theoretical qualification thereof, whilst research on actual transceivers has been fairly limited. To date, two transceiver engineering approaches have been proposed; the construction of simplified artificial cell-like structures made of biological materials or adaptation of existing biological cells [Suda, T. et al., 2008]. In comparison, the DNA-origami platform, disclosed in WO 2012/061719 and Douglas et al. (2012), herein incorporated by reference as if fully described herein, sanctions a novel non-cell like approach to construct autonomous, logic- guided nanobots, that can be programmed to transport molecular payloads to targets.

SUMMARY OF INVENTION

[0006] The present application is based on the copending PCT application titled "Non- Immunogenic and Nuclease Resistant Nucleic Acid Origami Devices And Compositions Thereof" of the same applicant, filed on April 13, 2013, and claims priority from US Provisional Patent Application No. 61/813,412 filed April 18, 2013, of which some of the present inventors are co-inventors, herein incorporated by reference as if fully described herein.

[0007] In one aspect, the present invention relates to a system acting as a logic gate, said system comprising at least one effector device, at least one regulator device, at least one input and at least one output, wherein each one of said at least one effector device and at least one regulator device each is a nucleic acid origami device comprising a scaffold strand and a plurality of staple strands, and independently has the structure A, B or C, wherein in the structure A: (i) one of the staple strands comprises either (a) an aptamer domain capable of binding to a binding partner comprising one of said at least one input; (b) an oligonucleotide capable of binding a DNA binding protein, said DNA binding protein comprising another of said at least one input; or (c) an oligoneucleotide attached to a nano-antenna capable of receiving an electromagnetic field comprising still another of said at least one input, or one of the staple strands comprises an aptamer domain of (a) and another of the staple strands comprises an oligonucleotide of (c); (ii) another of the staple strands comprises a latch domain hybridized or bound to said aptamer domain of (a) or oligonucleotide of (b) or (c), the latch domain sequence being selected such that the aptamer domain of (a) is capable of binding to the binding partner such that the binding partner displaces the latch domain, or the latch domain is capable of hybridizing with an external oligonucleotide selected such that it displaces the aptamer domain; said latch domain is linked to a binding partner that is selected such that it has a first configuration under a first condition and a different second configuration under a second condition, and the aptamer of (a) or the oligonucleotide of (b) is capable of binding to the binding partner having the first configuration but incapable of binding to the binding partner having the second configuration such that the latch domain is displaced from the aptamer of (a) or the oligonucleotide of (b) when the binding partner transitions from the first to the second configuration; or the nano-antenna of (c), upon receipt of said electromagnetic field, undergoes inductive coupling and subsequent heating thereby displacing the latch domain from the oligonucleotide of (c); and (iii) the aptamer domain of (a) or the oligonucleotide of (b) or (c), and the latch domain, when hybridized or bound to one another, hold the device in a closed configuration; and the device transitions to an open configuration when said aptamer domain or oligonucleotide, and the latch domain, are not hybridized or bound to one another,

in the structure B: (i) one of the staple strands comprises a first aptamer domain capable of binding to a first binding partner comprising one of said at least one input; (ii) another of the staple strands comprises a second aptamer domain capable of binding to a second binding partner comprising another one of said at least one input; (iii) still another of the staple strands comprises a first latch domain hybridized to the first aptamer domain, the first latch domain sequence being selected such that the first aptamer domain is capable of binding to the first binding partner such that the first binding partner displaces the first latch domain, or the first latch domain is capable of hybridizing with an external oligonucleotide selected such that it displaces the aptamer domain; (iv) yet another of the staple strands comprises a second latch domain hybridized to the second aptamer domain, the second latch domain sequence being selected such that the second aptamer domain is capable of binding to the second binding partner such that the second binding partner displaces the second latch domain or the second latch domain is capable of hybridizing with an external oligonucleotide selected such that it displaces the aptamer domain; and (v) said nucleic acid origami device is in a closed configuration when the first aptamer domain is hybridized to the first latch domain and/or the second aptamer domain is hybridized to the second latch domain; and the device transitions to an open configuration when the first aptamer domain is not hybridized to the first latch domain and the second aptamer domain is not hybridized to the second latch domain,

in the structure C: (i) two of the staple strands each comprises a latch domain linked to an oligonucleotide capable of hybridizing with an external oligonucleotide, said external oligonucleotides are positioned on another one of said nucleic acid origami devices; and; (ii) said nucleic acid origami device is in an open configuration when each one of the oligonucleotides capable of hybridizing with an external oligonucleotide is not hybridized to said external oligonucleotide; and the device transitions to a closed configuration when both of said oligonucleotides capable of hybridizing with an external oligonucleotide are hybridized to said external oligonucleotides, wherein each one of said at least one effector nucleic acid origami device and at least one regulator nucleic acid origami device is either alkylated, acylated or hydroxylated, or interacts with a compound capable of non-covalently binding to the major- or minor-groove of a double stranded nucleic acid, and therefore is resistant to nucleases, and/or each one of said at least one effector nucleic acid origami device and at least one regulator nucleic acid origami device lacks TLR9 recognition elements or the TLR9 recognition elements of said nucleic acid origami device are masked or modified and therefore said nucleic acid origami device is non-immunogenic.

[0008] In certain embodiments, the system of the present invention comprises one effector device having one or more handle domains each bound to a payload, at least one regulator device, at least one input and one output, wherein said output has a first output state when said effector device is in the closed configuration and a second output state when said effector device is in the open configuration. More particularly, the first output state corresponds to a logical off state and the second output state corresponds to a logical on state. Specific logical gates based on such a system are (i) a logical NAND gate, comprising two inputs and being in a logical on state if both inputs are present; (ii) a logical NOT gate, comprising one input and being in a logical on state if the input is absent; (iii) a logical implicit OR gate, comprising two inputs and being in a logical on state if at least one of the two inputs is present; and (iv) a logical exclusive OR (XOR) gate, comprising two inputs and being in a logical on state if only one of the two inputs is present, i.e., if the first of said two inputs is present and the second of said two inputs is absent, or the second of said two inputs is present and the first of said two inputs is absent.

[0009] In certain embodiments, the system of the present invention comprises two effector devices each having one or more handle domains each bound to a payload, at least one regulator device, at least one input and two outputs, wherein the first of said two outputs has a first output state when the first of said two effector devices is in the closed configuration and a second output state when the first of said two effector devices is in the open configuration; and the second of said two outputs has a first output state when the second of said two effector devices is in the closed configuration and a second output state when the second of said two effector devices is in the open configuration. In a system having this configuration, each one of said first output states corresponds to a logical off state and each one of said second output states corresponds to a logical on state. Specific logical gates based on such a system are (i) a logical controlled NOT gate, comprising two inputs and two outputs, wherein the first output acts as a XOR gate and is in a logical on state if only one of the two inputs is present, and the second output is sensitive to the presence of the first of said two inputs only and it is thus in a logical on state if either only the first of said two inputs or both inputs are present.

[0010] According to the present invention, whenever the system, i.e., the logic gate, of the invention is in a logical on state, irrespectively to the specific configuration of that system, the effector device comprised in the system is in an open configuration and each one of the payloads is positioned on an outer surface of said effector device.

[0011] In another aspect, the present invention relates to a system exhibiting quorum sensing comprising a plurality of effector nucleic acid origami devices each having the structure A, B or C as defined above in connection with the system acting as a logic gate or, alternatively, a plurality of a member each selected from a liposome, a particle or an artificial cell; at least one input; and at least one output, wherein said plurality of effector nucleic acid origami devices or members aggregates at a predetermined concentration, and wherein: each one of the effector nucleic acid origami devices comprises said handle domain bound to a first oligonucleotide capable of hybridizing with said aptamer domain of another of said effector nucleic acid origami devices and positioned on the outer surface of the device when the device is in the closed configuration; or each one of the members comprises on an outer surface thereof (a) a staple strand comprising a handle domain bound to a first oligonucleotide; and (b) a further staple strand comprising a second oligonucleotide selected such that it is capable of hybridizing to a first oligonucleotide of another of said plurality of members, each one of said effector nucleic acid origami devices or members having on open and a closed configuration and is initially in the closed configuration.

[0012] In a further aspect, the present invention provides a pharmaceutical composition comprising a system acting as a logic gate or a system exhibiting quorum sensing as defined above, and a pharmaceutically acceptable carrier. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figs. 1A-B visualizes nanorobots by agarose-gel (left panels) and transmission electron microscopy (TEM) micrographs (right panels) of nanorobots obtained at different folding durations. Both samples of A and B were folded at 20 nM scaffold concentration, 200 nM staples concentration, IX Tris/Acetic acid/EDTA (TAE) buffer, 10 mM MgCl ₂ . After folding, excess staples were removed using micon Ultra-0.5 mL 100K centrifugal filters (Millipore). (A) Folding duration is 80-60 °C at 2 min/°C and 60-10 °C at 150 min/°C. (B) Folding duration is 80-60 °C at 5 min/°C, 60-10 °C at 75 min/°C.

[0014] Fig. 2 depicts a 3D representation of a nanorobot in its inactive state (left panel in front view and middle panel in side view) and in its active open state (right panel).

[0015] Figs. 3A-C depict the basic design enabling collisions between DNA strands extending from E (effector), P (positive regulator), and N (negative regulator) robots. (A) Interaction between E and P robots; (B) Interaction between E and N robots; (C) Structural basis for differential keying of E and F (a second effector) robots.

[0016] Figs. 4A-F show gates mechanism for collision-based computing. (A) The nanorobot gate consists of two "arms", each comprised of a sensing strand hybridized to a complementary strand. Opening of the gate can be achieved by two modes: by a molecular cue (e.g. a protein, nucleic acid fragment, small molecule etc.) binding to the sensing strand (right), or by a DNA key presented by an adjacent nanorobot, hybridizing to the complementary strand and displacing the sensing strand by toehold-mediated strand migration. The DNA key can be presented by an adjacent nanorobot only when the latter is in state 1; nanorobots in state 0 cannot present any payload; (B) AND emulated by E architecture. The solitary E nanorobot turns to state 1 when it encounters both molecular cues kl (in this case - platelet-derived growth factor (PDGF)) and k2 (in this case - vascular endothelial growth factor (VEGF)), which are received by its sensing arms (kl* and k2*, respectively). Binding of only one of the cues is not sufficient for complete nanorobot opening. Therefore, the output generated by E is AND; (C) OR emulated by EP1P2 architecture. Here PI and P2 are added to E (here only PI is shown for clarity). PI and P2 require only one of the molecular cues kl or k2, respectively, to turn to state 1. Once in state 1, either PI or P2 can collide with E and present it with the DNA key necessary for opening of the reciprocal gate. For instance, if only kl is present, PI would turn to state 1 (by sensing kl through kl*) and present to E the DNA key necessary to open gate k2*. Therefore, collisions between E and PI (or P2) lead to a complex EP1 or EP2 in which E is at state 1, emulating an OR output; (D) XOR emulated by EP1P2N architecture. Here N is added to EP1P2 (here only PI is shown for clarity). Similar to E, N requires both cues kl and k2 to open, but unlike E, the DNA keys presented by PI and P2 do not open N (since the complementary strands of N are different than those of E). The payload presented by N is DNA strands which clasp both ends of

E, preventing it from opening (or promoting closure of E when it is already open). Therefore, when either kl or k2 are present, E is at state 1 (owing to PI and P2), but when both kl and k2 are present, it is pushed to state 0 by N. This emulates XOR output; (E) NAND emulated by E ₀ penN architecture. Here, E _ope n (nanorobots which lack sensing arms and are therefore constantly in state 1) are mixed with N. Only when both kl and k2 are present, N turns to state 1 and induce closure of E _ope n- This emulates a NAND output; and (F) NAND-constructed inverter by E _open Nx architecture. Here, N has only one cue required for turning to state 1. When the cue kl is present, N closes E, emulating an inverter based on the NAND gate shown above.

[0017] Figs. 5A-F depict Fluorescence Assisted Cell Sorting (FACS) analysis of fluorescently labeled nanorobots tested in vitro, in which each panel shows the response of the nanorobots to the presence or absence of either one of the two cues: (i) no cue, (ii) cue 1 present, cue 2 absent, (iii) cue 1 absent, cue 2 present, (iv) both cues present. (A) E architecture providing an AND gate; (B) EPl architecture providing an OR gate; (C) EP1P2N architecture providing a XOR gate; and (D, E) EopenN architectures providing a NAND gate (D) and a NOT gate (E).

[0018] Figs. 6A-C depict (FACS) analysis of fluorescently labeled nanorobots tested in vivo, in which each panel shows the response of the nanorobots to the presence or absence of either one of the two cues: (i) no cue, (ii) cue 1 present, cue 2 absent, (iii) cue 1 absent, cue 2 present, (iv) both cues present. (A) E architecture providing an AND gate; (B) EP1P2N architecture providing a XOR gate; and (C) EopenN architecture providing a NAND gate.

[0019] Fig. 7 depicts (FACS) analysis of fluorescently labeled nanorobots responding to two different cues on insect hemocytes in vivo in E architecture providing an AND gate; EPl architecture providing an OR gate; EP1P2N architecture providing a XOR gate; and EopenN architecture providing a NAND gate.

[0020] Figs. 8A-B depict NOT (A) and NAND (B) gates utilized in a multi-type "Asimov system" that shuts down robots doing harm to normal cells.

[0021] Fig. 9 shows a scheme demonstrating a quorum sensing (QS) signal generator consisting of three components: a selected key, a cleavable tether (substrate); and a DNA tail (Ulc) for loading onto a nanorobot loading site (the angled arms coming out from the square representing the nanorobot).

[0022] Fig. 10 shows a graph from a dynamic light scattering measurement showing the average diameter of the individual nanorobots as a function of their concentration (density).

[0023] Fig. 11 shows a FACS experiment done as follows: robots tagged fluorescently and loaded with biotin were mixed, in various densities (=concentrations) with avidin-coated microspheres. The results demonstrate that there are two separate affinities: the affinity of biotin to avidin determines fluorescence induced by unlocked robots (squares), while the quorum sensing mechanism determines fluorescence in the gated robots (triangles).

[0024] Fig. 12 depicts hemocyte fluorescence resulting from attached fluorescent nanorobots as a function of the amount of nanorobots injected into the insect Blaberus discoidalis. The nanorobots were equipped with a signal generator comprising the tether sequence KIEARC (SEQ ID NO: 1), a substrate of insect prophenol oxidase-activating protease (PAP) and fluorescently labeled anti-insect hemocytes antibody.

[0025] Fig. 13 depicts a scheme showing an example of a nanorobot capable of reversible self-aggregation and activation in a nanorobot concentration-dependent manner.

[0026] Fig. 14 depicts a scheme showing an example of a nanorobot capable of reversible self-aggregation and activation in a nanorobot concentration-dependent manner that is also dependent on the concentration of a ligand.

[0027] Fig. 15 shows a graph depicting dependency of cell cycle arrest or apoptosis in cancer cells on the concentration of robots loaded with a drug as compared with the free drug. The results demonstrate that the robot-induced effect is different than the drug-induced one and appears at different concentrations. In fact quorum sensing here improved the therapeutic window of the drug because it enabled a higher dose to accumulate without being active until the defined dose was reached.

DETAILED DESCRIPTION OF THE INVENTION

[0028] Computers capable of controlling molecules in a living organism could lead to a paradigm shift in how therapeutics are designed and used, had a proper interface been designed. The challenge of designing a computer-molecule interface is avoided when the computer itself is built from molecules, with computations driven by molecular forces [Adamatzky, A. et al., 2002]. DNA is a natural substrate for such a task, and DNA computing has been successfully demonstrated in a diverse set of purely mathematical-computational settings [Adleman, L. M. et al., 1994; Braich, R. S. et al., 2002; Yeh, C. W. et al., 2006; Chen, J. H. et al., 1991; Seeman, N. C, 1982; Winfree, E. et al., 1998; Rothemund, P. W., 2006; Adar, R. et al., 2004; Chang, W. L. et al., 2004; Qian, L. et al., 2011; Yin, Z. et al., 2002]. As a biomolecule, DNA also provides a natural conduit to affect genes, proteins and whole cells, although literature in this field is scarce. Benenson, Shapiro and colleagues have described elegant nucleic acid based computers, which generate logical outputs upon recognizing defined gene expression patterns [Xie, Z. et al., 2011; Rinaudo, K. et al., 2007; Benenson, Y. et al., 2004; Benenson, Y. et al., 2001]. Recently, we have used DNA origami [Rothemund, P. W., 2006; Dietz, H. et al., 2009; Douglas, S. M. et al., 2009] to fabricate an autonomous, logic guided nanorobot capable of delivering therapeutic payloads to the surface of target cells (WO 2012/061719).

[0029] The current work aims at setting interactions or collisions between these bio- responsive nanorobots to produce computations of higher complexity for controlling molecules. Collision-based computing is driven by the interactions between discrete mobile objects such as billiard balls, particles or solitons in nonlinear media [Adamatzky, A., 2002]. The DNA nanorobots disclosed herein provide a fit platform for universal collision-based computing if collisions can be designed to logically determine the state of the nanorobot carrying the effector payload.

[0030] The state of a single nanorobot can be either closed (0) or open (1) controlled by the nanorobot gate [Douglas, S. M. et ah, 2012], equivalent to an AND gate. Each of the two "arms" in the gate (one for each input) consists of a sensing strand (e.g. an aptamer to a specific cue) and a complementary strand; when these hybridize, they hold the nanorobot at state 0. To enable state-determining collisions between an "effector" nanorobot and a "regulator" nanorobot, the gate system was re-designed as follows. The gate typically opens only in response to the chosen cue, which displaces the complementary strand from the sensing stand. However, in the current design, the gate opens also in response to an external DNA "key", which hybridizes with the complementary strand, displacing the sensing strand by toehold-mediated migration [Zhang, D. Y. et ah, 2011; Zhang, D. Y. et ah, 2009] (Figure 4A). This external, cue-independent key to the effector nanorobot gate, is loaded as cargo into a regulator nanorobot, such that when the latter is at state 1, a collision between the two will subsequently alter the former's state from 0 to l.This assigns a "positive regulator" (P) phenotype to the nanorobot holding the external key. To design a "negative regulator" (N) phenotype as well, the regulator nanorobot is loaded with DNA strands that clasp two juxtaposed sides of the effector (E) nanorobot, thereby forcing it to close or preventing it from becoming open. Once E, P and N nanorobots have been designed and fabricated, various architectures can be achieved by mixing the proper types at defined stoichiometries in the presence or absence of their cognate cues.

[0031] Consider type E nanorobots carrying a payload and controlled by a gate, which opens if and only if both proteins X and Y are present. An architecture consisting only of E nanorobots (termed E architecture) would emulate a logical AND gate as previously described [Douglas, S. M. et ah, 2012] (Figure 4B). Now consider two additional nanorobot types, PI (opens in response to X and carries the key to the Y arm of E) and P2 (opens in response to Y and carries the key to the X arm of E). Adding PI and P2 to E (EP1P2 architecture) ensures the latter opens in response to X alone (by PI), Y alone (by P2), or XY (by itself), emulating a logical OR gate (Figure 4C). [0032] We now add type N nanorobots to the system described thus far (EP1P2N architecture), N being activated if and only if both X and Y are present. Here, in contrast to

EP1P2, the presence of both X and Y activates N, which simultaneously negates or prevents the activation of E. These outputs are identical to the ones generated by the logical XOR gate

(Figure ID). It is important to remember that in architectures EP1P2 and EP1P2N, E at its final state is always in complex with PI, P2 or N.

[0033] Architectures based on more than one effector nanorobot (like the EFP1P2N mentioned above) enable to relay output bits to additional therapeutic molecules, forming the basis for more complex gates, reversible logic and binary decoders. To demonstrate such architecture, a controlled NOT (CNOT) gate was designed, consisting of E, PI, P2 and N as described above, and in addition a second effector nanorobot, F, not negated by N, and a positive regulator nanorobot P3 that responds only to X and keys only F. Thus, one input bit is XORed by E and the second output bit is mapped unchanged by F.

[0034] Of particular importance is the emulation of functionally complete gates, such as NAND and NOR, which can be cascaded to build any other gate. NAND architecture can be based on two nanorobot types: Eopen, which is similar to E but lacks gate strands, hence it is constitutively open regardless of the inputs X and Y; and N, which opens in response to XY and negates Eopen as a result. This architecture therefore produces 0 in response to XY, and 1 otherwise.

[0035] We have recently proposed an actuating DNA device, to which the drug is tethered, logically switching the drug between two states - unavailable ("off) and available ("on"), without actually releasing it [Douglas, S. M. et ah, 2012]. By constructing various types of DNA logic gates based on aptamer recognition, strand displacement etc., such devices could be programmed to respond to diverse stimuli in many defined ways. Still, even the design described thus far does not take advantage of the more elaborate modes of control provided by multi-agent systems exhibiting collective behaviors.

[0036] The present invention is based, inter alia, on the finding disclosed in Example 2, that the utilization of a great number of different nucleic acid origami devices, some acting as effector devices because they carry a payload, such as fluorescent molecule, biotin or avidin, and some acting as regulator devices because in their active state they are capable of interacting with the effector devices by way of their oligonucleotide payload and either turning them on or off, in specific predetermined architectures, can act as any important logical gate. For example, it has been found in accordance with the present invention that certain combinations of effector and regulator nucleic acid origami devices exhibit logical AND, OR, XOR, NAND, NOT and CNOT behavior. These different gates can be combined to form any logical circuit required in order to present one or more payloads, such as a drug or combination of different drugs, under specific circumstances that may be defined in terms of the presence of specific ligands or enzymes, or certain environmental determinants such as pH. Since the nucleic acid origami devices can circle between an open and a closed configuration, and since the payload may be bound to the device via a linker and presented when the device is in the open configuration without being released, not only the delivery of a drug may be controlled but also the withdrawal of the drug.

[0037] Thus, in one aspect, the present invention relates to a system acting as a logic gate, said system comprising at least one effector device, at least one regulator device, at least one input and at least one output, wherein each one of said at least one effector device and at least one regulator device each is a nucleic acid origami device comprising a scaffold strand and a plurality of staple strands, and independently has the structure A, B or C, wherein

in the structure A (this structure has one key: either an aptamer, a nano-antenna or a "ligand-sensing molecule"-type key): (i) one of the staple strands comprises either (a) an aptamer domain capable of binding to a binding partner comprising one of said at least one input; (b) an oligonucleotide capable of binding a DNA binding protein, said DNA binding protein comprising another of said at least one input; or (c) an oligoneucleotide attached to a nano- antenna capable of receiving an electromagnetic field comprising still another of said at least one input, or one of the staple strands comprises an aptamer domain of (a) and another of the staple strands comprises an oligonucleotide of (c); (ii) another of the staple strands comprises a latch domain hybridized or bound to said aptamer domain of (a) or oligonucleotide of (b) or (c), the latch domain sequence being selected such that the aptamer domain of (a) is capable of binding to the binding partner such that the binding partner displaces the latch domain, or the latch domain is capable of hybridizing with an external oligonucleotide selected such that it displaces the aptamer domain; said latch domain is linked to a binding partner that is selected such that it has a first configuration under a first condition (comprising one of said at least one inputs) and a different second configuration under a different second condition (comprising another of said at least one inputs), and the aptamer of (a) or the oligonucleotide of (b) is capable of binding to the binding partner having the first configuration but incapable of binding to the binding partner having the second configuration such that the latch domain is displaced from the aptamer of (a) or the oligonucleotide of (b) when the binding partner transitions from the first to the second configuration; or the nano-antenna of (c), upon receipt of said electromagnetic field, undergoes inductive coupling and subsequent heating thereby displacing the latch domain from the oligonucleotide of (c); and (iii) the aptamer domain of (a) or the oligonucleotide of (b) or (c), and the latch domain, when hybridized or bound to one another, hold the device in a closed configuration; and the device transitions to an open configuration when said aptamer domain or oligonucleotide, and the latch domain, are not hybridized or bound to one another,

in the structure B (this structure has two different aptamer keys): (i) one of the staple strands comprises a first aptamer domain capable of binding to a first binding partner comprising one of said at least one input; (ii) another of the staple strands comprises a second aptamer domain capable of binding to a second binding partner comprising another one of said at least one input; (iii) still another of the staple strands comprises a first latch domain hybridized to the first aptamer domain, the first latch domain sequence being selected such that the first aptamer domain is capable of binding to the first binding partner such that the first binding partner displaces the first latch domain, or the first latch domain is capable of hybridizing with an external oligonucleotide selected such that it displaces the aptamer domain; (iv) yet another of the staple strands comprises a second latch domain hybridized to the second aptamer domain, the second latch domain sequence being selected such that the second aptamer domain is capable of binding to the second binding partner such that the second binding partner displaces the second latch domain, or the second latch domain is capable of hybridizing with an external oligonucleotide selected such that it displaces the aptamer domain; and (v) said nucleic acid origami device is in a closed configuration when the first aptamer domain is hybridized to the first latch domain and/or the second aptamer domain is hybridized to the second latch domain; and the device transitions to an open configuration when the first aptamer domain is not hybridized to the first latch domain and the second aptamer domain is not hybridized to the second latch domain,

in the structure C (this structure is initially open and is closed by external oligonucleotides on another nanobot): (i) two of the staple strands each comprises a latch domain linked to an oligonucleotide capable of hybridizing with an external oligonucleotide, said external oligonucleotides are positioned on another one of said nucleic acid origami devices; and (ii) said nucleic acid origami device is in an open configuration when each one of the oligonucleotides capable of hybridizing with an external oligonucleotide is not hybridized to said external oligonucleotide; and the device transitions to a closed configuration when both of said oligonucleotides capable of hybridizing with an external oligonucleotide are hybridized to said external oligonucleotides, wherein

each one of said at least one effector nucleic acid origami device and at least one regulator nucleic acid origami device is either alkylated, acylated or hydroxylated, or interacts with a compound capable of non-covalently binding to the major- or minor-groove of a double stranded nucleic acid, and therefore is resistant to nucleases, and/or each one of said at least one effector nucleic acid origami device and at least one regulator nucleic acid origami device lacks TLR9 recognition elements or the TLR9 recognition elements of said nucleic acid origami device are masked or modified and therefore said nucleic acid origami device is non-immunogenic.

[0038] The terms "nucleic acid origami device", "nanorobot", "nanobot", and simply "device", used herein interchangeably in connection with the effector devices and regulator devices composing the logic gate of the present invention refer to a nucleic acid, e.g., a DNA, origami device as defined above, in any one of the configurations described below.

[0039] The term "scaffold strand" as used herein refers to a long nucleic acid, e.g., DNA, strand, for example about 7250 bases long, that may be folded into a particular shape using a plurality of rationally designed "staple" DNA strands. However, there is no principal limit to the length of scaffold strand; it all depends on the size of device you want to build, so the scaffold strand could have any length between 15 and 10 13 bases. The sequences of the staple strands are designed such that they hybridize to particular portions of the scaffold strands and, in doing so, force the scaffold strands into a particular shape. Methods useful in the making of DNA origami structures can be found, e.g., in Rothemund, P.W., Nature 440:297-302 (2006); Douglas et al, Nature 459:414-418 (2009); Dietz et al, Science 325:725-730 (2009); and US Patent Publication Nos. 2007/0117109, 2008/0287668, 20100069621 and 2010/0216978, each of which is incorporated by reference in its entirety. Staple design can be facilitated using, e.g., CADnano software, available at http://www.cadnano.org.

[0040] The term "latch domain" as used herein refers to a nucleic acid domain capable of hybridizing to an opposing aptamer domain or other nucleic acid (e.g., attached to a nano- antenna) and thereby holds the nucleic acid origami device in a closed configuration.

[0041] The term "aptamer domain" as used herein refers to a nucleic acid molecule selected such that it is capable of specifically binding a target molecule, much like an antibody is capable of specifically binding an antigen. Aptamers can be designed to target essentially any antigen of interest using methods known in the art. For example, methods of designing aptamers specific for a target of interest can be found in US Patent Nos. 5,582,981, 5,756,291, 5,840,867 and 7,745,607, and in Tang et al, Anal. Chem. 79:4900-4907 (2007), each of which are incorporated by reference in their entirety.

[0042] A non-limiting example of a "oligonucleotide capable of binding a DNA binding protein" is an isolated response element originally found within a gene promoter. A specific example is the glucose responsive regulatory element.

[0043] The terms "hybridized" or "hybridizing" as used herein refer to the binding of two strands of nucleic acid molecules to each other, and are meant to include that the two strands are capable of hybridizing to each other, or that the two strands are complementary to each other, wherein the complementarity of the two strands may vary in order to calibrate the affinity between the two strands. For example, each one of the two strands may have a sequence of base pairs that is 100, 99, 98, 96, 94, 92, 90, 85, 80, 75, 70, 65 or 60% complementary to the sequence of the other strand.

[0044] The aptamer domain is capable of binding to a binding partner such that the binding partner displaces the latch domain. In the case the binding partner is a ligand such as a target antigen that binds to the aptamer by inducing a structural change in the aptamer domain enabling it to bind to the ligand, the ligand binds to the aptamer domain at a higher affinity than the latch domain causing the aptamer to lose its linear configuration that enables it to hybridize with the latch domain, and therefore releases the latch domain. In case the binding partner is a nucleic acid molecule, the nucleic acid molecule may have a higher degree of complementarity with the aptamer domain than has the latch domain and therefore displaces the latch domain. Conceptually, the nucleic acid molecule could be complementary to the latch domain and bind to it, thus displacing the aptamer domain. In both cases, the displacement would lead to the transitions of the device from a closed to an open configuration. In all embodiments wherein it is defined that the binding partner binds to the aptamer domain such that the binding partner displaces the latch domain, it should be understood that, in the case of the binding partner being an oligonucleotide, the binding partner may alternatively bind to the latch domain such that oligonucleotide displaces the latch domain.

[0045] The term "external oligonucleotide" as used herein refers to an oligonucleotide that is not comprised within a particular device being defined, but is found e.g. on a neighboring identical or different device.

[0046] The term "nuclease" as used herein refers to an enzyme capable of cleaving the phosphodiester bonds between the nucleotide subunits of nucleic acids. Nucleases are usually further divided into endo nucleases and exonucleases, although some of the enzymes may fall in both categories. Well known nucleases are deoxyribonuclease and ribonuclease.

[0047] The term "resistant to nucleases" as used herein refers to nucleic acid molecules that have been modified so that they are more stable in the presence of nucleases than unmodified molecules.

[0048] The term "TLR9 recognition element" as used herein refers to elements on a nucleic acid molecule, e.g., a nucleic acid molecule, that are recognized by Toll Like Receptor (TLR) 9 present in the cytoplasm of antigen presenting cells. TLR9 is expressed by numerous cells of the immune system such as dendritic cells, B lymphocytes, monocytes and natural killer (NK) cells. TLR9 is expressed intracellularly, within the endosomal compartments and functions to alert the immune system of viral and bacterial infections by binding to DNA rich in CpG motifs. TLR9 signals leads to activation of the cells initiating pro-inflammatory reactions that result in the production of cytokines such as type-I interferon and IL-12.

[0049] The terms "CpG motif" and "CpG island" are used interchangeably herein and refer to short single- stranded synthetic nucleic acid molecules that contain a cytidine triphosphate deoxynucleotide ("C)" followed by a guanidine triphosphate deoxynucleotide ("G"). The "p" refers to the phosphodiester link between consecutive nucleotides, although some DNA nucleic acids have a modified phosphorothioate (PS) backbone instead.

[0050] The term "masked nucleic acid origami device" refers to a device that is linked to a molecule capable of covering domains otherwise accessible to the environment, such as CpG islands and thereby making these domains inaccessible to the environment.

[0051] The term "non-immunogenic" as used herein refers to a molecule that does not induce a response against it by the immune system in a mammal, or induces a weaker response than would have been induced by the same molecule that differs only in that it has TLR9 recognition elements that have not been masked or modified.

[0052] Examples of non-covalent binding are binding involving ionic bonds, electrostatic interactions, hydrophobic interactions, hydrogen bonds or van der Waals forces.

[0053] In one embodiment, in at least one of the nucleic acid origami devices composing the logic gate of the present invention and having the structure A: (i) one of the staple strands comprises a first aptamer domain and another of the staple strands comprises a second aptamer domain wherein both aptamer domains are capable of binding to identical binding partners; (ii) still another of the staple strands comprises a first latch domain hybridized to the first aptamer domain and selected such that the first aptamer domain is capable of binding to the binding partner such that the binding partner displaces the first latch domain; (iii) yet another of the staple strands comprises a second latch domain hybridized to the second aptamer domain and selected such that the second aptamer domain is capable of binding to the binding partner such that the binding partner displaces the second latch domain; and (iv) said nucleic acid origami device is in a closed configuration when the first aptamer domain is hybridized to the first latch domain and/or the second aptamer domain is hybridized to the second latch domain; and the device transitions to an open configuration when the first aptamer domain is not hybridized to the first latch domain and the second aptamer domain is not hybridized to the second latch domain.

[0054] In other embodiments, each one of the nucleic acid origami devices composing the logic gate of the present invention is non-immunogenic, resistant to nucleases, or both non- immunogenic and resistant to nucleases. [0055] In certain embodiments, the nucleic acid of each one of the nucleic acid origami devices composing the logic gate of the present invention is DNA.

[0056] In certain embodiments, the TLR9 recognition elements are CpG islands, and the nucleic acid origami devices may be methylated, preferably at CpG dinucleotides.

[0057] The term "methylated" refers to a nucleic acid molecule to which a methyl group has been added, specifically to the cytosine or adenine nucleotide.

[0058] In certain embodiments, the nucleic acid origami devices composing the logic gate of the present invention are each methylated at the carbon atom in position 5 of cytosine residues; at the amino group linked to the carbon atom in position 4 of cytosine residues; or at the amino group linked to the carbon atom in position 6 of adenine residues, for example, the nucleic acid origami device may be methylated at the carbon atom in position 5 of cytosine residues in CpG dinucleotides.

[0059] In certain embodiments, the nucleic acid origami devices composing the logic gate of the present invention are each modified at either cytosine or guanine residues in CpG dinucleotides. For example, the cytosine or guanine residue may be modified by covalently linking it via a linker to a macromolecule having a functional group, wherein said macromolecule is selected from a polymer such as poly(ethylene)glycol, polystyrene, poly(vinyl)chloride, pectin, polygalacturonic acid, polygalacturonic acid and poly(lactic-co- glycolic acid) (PLGA), a peptide, a lipid or a polysaccharide. The functional group may be, but is not limited to, an amino, mercapto, and carboxyl group.

[0060] In certain embodiments, the compound capable of non-covalently binding to the major- or minor-groove of a double stranded nucleic acid is selected from netropsin, distamycin, an oligoamide, a sugar-oligoamide conjugate or a bis-amidine. In a particular embodiment, the compound is netropsin, and the netropsin is further covalently linked through two of its terminal amino groups, optionally via a linker, to the double stranded nucleic acid, as shown, e.g., in Figs. 8 and 9 of copending US provisional application titled "Non-Immunogenic and Nuclease Resistant Nucleic Acid Origami Devices And Compositions Thereof", filed on April 18, 2013, of which some of the present inventors are co-inventors.

[0061] In certain embodiments, one or more further staple strands in at least one of the nucleic acid origami devices composing the logic gate of the present invention each comprises a handle domain bound to a payload, more specifically, to a payload moiety, optionally via a linker, wherein at least one of the payload(s) is one of said at least one output, and optionally, at least one of the payload(s) is one of said at least one input or emulates one of said at least one input. [0062] In one embodiment, the linker comprises an oligonucleotide having a sequence complementary to the sequence of the handle domain and optionally comprising a further domain comprising a recognition site for enzymatic cleavage, and the payload is bound to the handle domain through the hybridization of the oligonucleotide to the handle domain. This further domain may comprise a peptide linker comprising a protease recognition site for cleavage by a protease, such as a matrix metalloproteinase. Alternatively, the linker may comprise a protein capable of binding a small molecule such as, but not limited to, a cyclooxygenase protein capable of binding paracetamol, a sodium channel subunit capable of binding tetrodotoxin and an anti-digoixin antibody capable of binding digoxin.

[0063] The term "protease recognition site" refers to an amino acid sequence recognized by an endo peptidase, such as but not limited to: Trypsin - cuts after Arg or Lys, unless followed by Pro; Chymotrypsin - cuts after Phe, Trp, or Tyr, unless followed by Pro; Elastase - cuts after Ala, Gly, Ser, or Val, unless followed by Pro; Thermolysin - cuts before He, Met, Phe, Trp, Tyr, or Val, unless preceded by Pro; Pepsin - cuts before Leu, Phe, Trp or Tyr, unless preceded by Pro; and Endopeptidase V8 (also known as Glu-C) - cuts after Glu.

[0064] In certain embodiments, the payload each independently is a drug, such as insulin, an antibody or a fragment thereof, a cell surface receptor ligand or a biologically active fragment thereof, a small molecule, a nucleic acid, such as an oligonucleotide, a nuclease, an aptamer, a lipid, a glycan, a protein a glycoprotein, a glycolipid, a nanoparticle, a fluorophore, a radioactive compound, a nano-antenna, or a liposome. The payload being an oligonucleotide may function as an "external oligonucleotide" for another device.

[0065] The terms "payload" and "payload moiety" are used herein interchangeably and refer both to the free payload and to said payload when covalently linked directly to the handle domain or to the linker while maintaining its biological activity.

[0066] In certain embodiments, the nano-antenna of a nucleic acid origami device included within the logic gate of the present invention, e.g., a nucleic acid origami device having the structure A, or being used as a payload when linked to the handle domain, optionally via a linker, each independently comprises a metal quantum dot, a metal nanoparticle, or a metal nanocrystal, wherein said metal is preferably Au.

[0067] In other embodiments, the plurality of staple strands are selected such that at least one of the payloads is positioned on an inner surface of a nucleic acid origami device when the device is in the closed configuration; and the transition to the open configuration causes said payload to be positioned on an outer surface of the nucleic acid origami device.

[0068] The term "inner surface" with respect to the nucleic acid origami device of the present invention in any one of the configurations defined herein, refers to any surface area of the device that is sterically precluded from interacting with members in the immediate environment surrounding the nucleic acid origami device, such as the surface of a cell, while an "outer surface" is any surface area of the device that is not sterically precluded from interacting with members in the immediate environment surrounding the nucleic acid origami device, such as the surface of a cell.

[0069] In still other embodiments, the staple strands in at least one of the nucleic acid origami devices composing the logic gate of the present invention comprises a handle domain positioned on an outer surface of the device when the device is in the closed configuration, and in this case, the handle domain is bound to a payload preferably selected from an oligonucleotide or a liposome. In certain embodiments, the handle domain positioned on the outer surface of the device when the device is in the closed configuration becomes positioned on an inner surface of the device when the device is in the open configuration.

[0070] The shape of each one of the devices composing the logic gate of the present invention may be chosen according to the purpose of the device, and is easily obtained by defining the shape in a specialized computer program well known in the art of DNA origami, such as CADnano software, available at http://www.cadnano.org.

[0071] In certain embodiments, the plurality of staple strands composing at least one of the nucleic acid origami devices are selected such that the nucleic acid origami device is substantially barrel-shaped and in other embodiments the plurality of staple strands are selected such that the nucleic acid origami device has a substantially hexagonal tube shape. The plurality of staple strands composing at least one of the nucleic acid origami devices may be selected such that the nucleic acid origami device comprises an open end and they may be selected such that the nucleic acid origami device comprises two open ends.

[0072] In certain embodiments: (i) in at least one of the nucleic acid origami devices composing the logic gate of the present invention and having the structure A or B, the plurality of staple strands are selected such that the nucleic acid origami device comprises a first domain and a second domain, wherein the first domain comprises said aptamer domain(s) of (a) each capable of binding to a binding partner; or said oligoneucleotide(s) of (b), each capable of binding a DNA binding protein; or said oligoneucleotide(s) of (c), each attached to a nano- antenna; and the second domain comprises said latch domain(s), wherein a first end of the first domain is attached to a first end of the second domain by at least one single-stranded nucleic acid hinge and the second end of the first domain is attached to the second end of the second domain by the hybridization or binding of each one of said aptamer domains or said oligonucleotides to said latch domains, respectively; or (ii) in at least one of the nucleic acid origami devices composing the logic gate of the present invention and having the structure C, the plurality of staple strands are selected such that the nucleic acid origami device comprises a first domain and a second domain, wherein each one of said first and second domains comprises one of said latch domains linked to an oligonucleotide capable of hybridizing with an external oligonucleotide, wherein a first end of the first domain is attached to a first end of the second domain by at least one single-stranded nucleic acid hinge and the second end of the first domain is not attached to the second end of the second domain.

[0073] In particular such embodiments: (i) the plurality of staple strands in at least one of the nucleic acid origami devices having the structure A or B are selected such that the second end of the first domain becomes unattached to the second end of the second domain if each one of said aptamer domains is contacted by its respective binding partner and/or if each one of said nano-antennas receives an electromagnetic field and undergoes inductive coupling and subsequent heating; or (ii) the plurality of staple strands in at least one of the nucleic acid origami device having the structure C are selected such that the second end of the first domain becomes attached to the second end of the second domain if each one of said latch domains is hybridized to a different one of said external oligonucleotides.

[0074] In certain embodiments, the binding partner each independently is an antigen selected from a tumor associated antigen; a cell-membrane receptor; a secreted or membrane bound growth factor; a hormone; a cytokine; a ligand; a chemokine; a bacterial, a viral or parasitic antigen; a lipid; an oligonucleotide; a sugar, an enzyme or a DNA binding protein; or it may be a "damage cue" or "damage indicator", examples of which are an obligatory intracellular molecule such as ATP, ribosome fragments, rRNA, nuclear pore components, histones, etc. for discerning cell damage in a non-discriminatory way; or microRNA, or certain isoforms or analogs of the above mentioned cues, such as glycosylation or phosphorylation variants, differentially present in a specific type of cell, a normal cell or a tumor cell.

[0075] The system of the present invention may comprise a glucose-sensing nano-device that exhibits insulin on its surface at high glucose concentration. For this purpose, as explained above, the nano-devise may comprise a binding partner that is selected such that it has a first configuration under a first condition and a different second configuration under a different second condition, and the aptamer of (a) or the oligonucleotide of (b) is capable of binding to the binding partner having the first configuration but incapable of binding to the binding partner having the second configuration such that the latch domain is displaced from the aptamer of (a) or the oligonucleotide of (b) when the binding partner transitions from the first to the second configuration, thereby opening the device and exposing the insulin. In particular, this binding partner may be an enzyme, for example a glucokinase (see Example 6). [0076] Thus, in certain embodiments, the enzyme is a glucokinase and the aptamer domain of (a) is capable of binding to the glucokinase having the first configuration but is incapable of binding to the glucokinase having the second configuration; or the DNA binding protein is a glucose response factor and said oligonucleotide of (b) is a glucose responsive regulatory element capable of binding to the glucose response factor having the first configuration but incapable of binding to the glucose response factor having the second configuration. The glucokinase may be a mammalian glucokinase, such as but not limited to human, mouse or rat beta cell glucokinase.

[0077] In certain embodiments, the first condition (or input) is a glucose concentration in the range of 0-4.5 mM and the second condition (or input) is a glucose concentration above 4.5 mM, for example in the range of 5-10 mM.

[0078] The system of the present invention may be used to specifically kill target cells while sparing non-target cells, e.g. normal healthy cells. This is achieved by administering, by means of one or more effector nanorobots, a toxin that preferentially kills the target cells, and by stopping the action of the nanorobots when it senses, as a system also comprising one or more regulatory nanorobots, a distress or damage cue released from normal healthy cells that are also damaged in the process. The nanorobot make-up of the system depends on the logic gate required to control the effector nanorobots, for example logic gates disclosed herein or combinations thereof.

[0079] In certain embodiments, the system of the present invention comprises one effector device having one or more handle domains each bound to a payload, at least one regulator device, at least one input and one output, wherein said output has a first output state when said effector device is in the closed configuration and a second output state when said effector device is in the open configuration. In particular such embodiments, the first output state corresponds to a logical off state and the second output state corresponds to a logical on state.

[0080] In certain more particular such embodiment, the system of the present invention is a logical NAND gate, comprising two inputs and being in a logical on state if both inputs are present. In a specific embodiment, the system of the invention comprises one effector device and one regulator device, wherein (i) said effector device has the structure C and is initially in an open configuration; and (ii) said regulator device is a negative regulator device having the structure B and comprising two handle domains each bound to a nucleic acid molecule acting as a signal molecule capable of hybridizing to a different one of said latch domains of said effector device, said negative regulator device is initially in a closed configuration wherein said signal molecules are positioned on an inner surface of the negative regulator device and, when both inputs are present, transitions to an open configuration wherein said signal molecules are positioned on an outer surface of the negative regulator device,

wherein the gate is in a logical on state when each one of said two signal molecules hybridizes to a different one of said latch domains of said effector device thereby promoting closure of the effector device or preventing transition of the effector device to an open configuration.

[0081] In certain more particular such embodiment, the system of the present invention is a logical NOT gate, comprising one input and being in a logical on state if the input is absent. In a specific embodiment, the system of the invention comprises one effector device and one regulator device, wherein (i) said effector device has the structure C and is initially in an open configuration; and (ii) said regulator device is a negative regulator device having the structure A and comprising two handle domains each bound to a nucleic acid molecule acting as a signal molecule capable of hybridizing to a different one of said latch domains of said effector device, said negative regulator device is initially in a closed configuration wherein said signal molecules are positioned on an inner surface of the negative regulator device and, when the input is present, transitions to an open configuration wherein said signal molecules are positioned on an outer surface of the negative regulator device,

wherein the gate is in a logical on state when each one of said two signal molecules hybridizes to a different one of said latch domains of said effector device thereby promoting closure of the effector device or preventing transition of the effector device to an open configuration.

[0082] In certain more particular such embodiment, the system of the present invention is a logical OR gate, comprising two inputs and being in a logical on state if at least one of the two inputs is present. In a specific embodiment, the system of the invention comprises one effector device and two regulator devices, wherein (i) said effector device has the structure B, is initially in a closed configuration and, when both inputs are present, transitions to an open configuration; (ii) the first of said two regulator devices is a first positive regulator device having the structure A and comprises one handle domain bound to a nucleic acid molecule acting as a signal molecule capable of emulating the second of said two inputs by hybridizing to the aptamer domain of said effector device that is capable of binding to the second of said two inputs, said first positive regulator device is initially in a closed configuration wherein said signal molecule is positioned on an inner surface of said first positive regulator device and, when the first of said two inputs is present, transitions to an open configuration wherein said signal molecule is positioned on an outer surface of the positive regulator device; and (iii) the second of said two regulator devices is a second positive regulator device having the structure A and comprises one handle domain bound to a nucleic acid molecule acting as a signal molecule capable of emulating the first of said two inputs by hybridizing to the aptamer domain of said effector device that is capable of binding to the first of said two inputs, said second positive regulator device is initially in a closed configuration wherein said signal molecule is positioned on an inner surface of said second positive regulator device and, when the second of said two inputs is present, transitions to an open configuration wherein said signal molecule is positioned on an outer surface of the positive regulator device,

wherein the gate is in a logical on state when the first of said two inputs binds to the aptamer domain of the effector device capable of binding to the first of said two inputs and the second of said two inputs binds to the aptamer domain of the effector device capable of binding to the second of said two inputs, such that said inputs displace the latch domains initially hybridized to said aptamer domains, respectively;

the first of said two inputs binds to the aptamer domain of said effector device capable of binding to the first of said two inputs and the signal molecule of the first positive regulator device hybridizes to the aptamer domain of said effector device capable of binding to the second of said two inputs, such that the first input and the signal molecule of the first positive regulator device displace the latch domains initially hybridized to said aptamer domains, respectively; or

the second of said two inputs binds to the aptamer domain of said effector device capable of binding to the second of said two inputs and the signal molecule of the second positive regulator device hybridizes to the aptamer domain of said effector device capable of binding to the first of said two inputs, such that the second input and the signal molecule of the second positive regulator device displace the latch domains initially hybridized to said aptamer domains, respectively. The reference to an 'input that is present' is meant to represent for example the presence of an antigen that is capable of opening a nano-device by binding to an aptamer or the presence of a ligand, such as glucose at a certain concentration, that is capable of binding to a concentration-sensitive binding partner, such as a glucokinase, which upon binding of glucose alters its configuration thereby opening the nano-device.

[0083] In certain more particular such embodiment, the system of the present invention is an exclusive OR (XOR) gate, comprising two inputs and being in a logical on state if only one of the two inputs is present, i.e., if the first of said two inputs is present and the second of said two inputs is absent, or the second of said two inputs is present and the first of said two inputs is absent. In a specific embodiment, the system of the invention comprises one effector device and three regulator devices, wherein (i) said effector device has the structure B, is initially in a closed configuration and, when both inputs are present, transitions to an open configuration; (ii) the first of said three regulator devices is a first positive regulator device having the structure A and comprises one handle domain bound to a nucleic acid molecule acting as a signal molecule capable of emulating the second of said two inputs by hybridizing to the aptamer domain of said effector device that is capable of binding to the second of said two inputs, said first positive regulator device is initially in a closed configuration wherein said signal molecule is positioned on an inner surface of said first positive regulator device and, when the first of said two inputs is present, transitions to an open configuration wherein said signal molecule is positioned on an outer surface of the positive regulator device; (iii) the second of said three regulator devices is a second positive regulator device having the structure A and comprises one handle domain bound to a nucleic acid molecule acting as a signal molecule capable of emulating the first of said two inputs by hybridizing to the aptamer domain of said effector device that is capable of binding to the first of said two inputs, said second positive regulator device is initially in a closed configuration wherein said signal molecule is positioned on an inner surface of said second positive regulator device and, when the second of said two inputs is present, transitions to an open configuration wherein said signal molecule is positioned on an outer surface of the positive regulator device; and (iv) the third of said three regulator devices is a negative regulator device having the structure B and comprising two handle domains each bound to a nucleic acid molecule acting as a signal molecule capable of hybridizing to a different one of said latch domains of said effector device, said negative regulator device is initially in a closed configuration wherein said signal molecules are positioned on an inner surface of the negative regulator device and, when both inputs are present, transitions to an open configuration wherein said signal molecules are positioned on an outer surface of the negative regulator device,

wherein the gate is in a logical on state when: the first of said two inputs binds to the aptamer domain of said effector device capable of binding to the first of said two inputs and the signal molecule of the first positive regulator device hybridizes to the aptamer domain of said effector device capable of binding to the second of said two inputs, such that the first input and the signal molecule of the first positive regulator device displace the latch domains initially hybridized to said aptamer domains, respectively; or

the second of said two inputs binds to the aptamer domain of said effector device capable of binding to the second of said two inputs and the signal molecule of the second positive regulator device hybridizes to the aptamer domain of said effector device capable of binding to the first of said two inputs, such that the second input and the signal molecule of the second positive regulator device displace the latch domains initially hybridized to said aptamer domains, respectively, and wherein both inputs are present, the first of said two inputs binds to the aptamer domain of the effector device capable of binding to the first of said two inputs and the second of said two inputs binds to the aptamer domain of the effector device capable of binding to the second of said two inputs, such that said inputs displace the latch domains initially hybridized to said aptamer domains, respectively, and each one of said two signal molecules of the negative regulator device hybridizes to a different one of said latch domains of said effector device thereby promoting closure of the effector device or preventing transition of the effector device to an open configuration.

[0084] In other embodiments, the system of the present invention comprises two effector devices each having one or more handle domains each bound to a payload, at least one regulator device, at least one input and two outputs, wherein the first of said two outputs has a first output state when the first of said two effector devices is in the closed configuration and a second output state when the first of said two effector devices is in the open configuration; and the second of said two outputs has a first output state when the second of said two effector devices is in the closed configuration and a second output state when the second of said two effector devices is in the open configuration. In particular such embodiments, each one of the first output states corresponds to a logical off state and each one of the second output states corresponds to a logical on state.

[0085] In certain more particular such embodiment, the system of the present invention is a logical controlled NOT gate, comprising two inputs and two outputs, wherein the first output is in a logical on state if only one of the two inputs is present and the second output is sensitive to the presence of the first of said two inputs only and it is thus in a logical on state if either only the first of said two inputs or both inputs are present.

[0086] In a specific embodiment, the system of the invention comprises two effector devices and four regulator devices, wherein (i) the first of said two effector devices has the structure B, is initially in a closed configuration and, when both inputs are present, transitions to an open configuration; (ii) the first of said three regulator devices is a first positive regulator device having the structure A and comprises one handle domain bound to a nucleic acid molecule acting as a signal molecule capable of emulating the second of said two inputs by hybridizing to the aptamer domain of said effector device that is capable of binding to the second of said two inputs, said first positive regulator device is initially in a closed configuration wherein said signal molecule is positioned on an inner surface of said first positive regulator device and, when the first of said two inputs is present, transitions to an open configuration wherein said signal molecule is positioned on an outer surface of the positive regulator device; (iii) the second of said three regulator devices is a second positive regulator device having the structure A and comprises one handle domain bound to a nucleic acid molecule acting as a signal molecule capable of emulating the first of said two inputs by hybridizing to the aptamer domain of said effector device that is capable of binding to the first of said two inputs, said second positive regulator device is initially in a closed configuration wherein said signal molecule is positioned on an inner surface of said second positive regulator device and, when the second of said two inputs is present, transitions to an open configuration wherein said signal molecule is positioned on an outer surface of the positive regulator device; (iv) the third of said three regulator devices is a negative regulator device having the structure B and comprising two handle domains each bound to a nucleic acid molecule acting as a signal molecule capable of hybridizing to a different one of said latch domains of said effector device, said negative regulator device is initially in a closed configuration wherein said signal molecules are positioned on an inner surface of the negative regulator device and, when both inputs are present, transitions to an open configuration wherein said signal molecules are positioned on an outer surface of the negative regulator device; and (v) the second of said two effector devices has the structure A, is initially in a closed configuration and, when the first of said two inputs is present, transitions to an open configuration,

wherein the first of said two outputs of the gate is in a logical on state when: the first of said two inputs binds to the aptamer domain of said first of said two effector devices capable of binding to the first of said two inputs and the signal molecule of the first positive regulator device hybridizes to the aptamer domain of said first of said two effector devices capable of binding to the second of said two inputs, such that the first input and the signal molecule of the first positive regulator device displace the latch domains initially hybridized to said aptamer domains, respectively; or

the second of said two inputs binds to the aptamer domain of said first of said two effector devices capable of binding to the second of said two inputs and the signal molecule of the second positive regulator device hybridizes to the aptamer domain of said first of said two effector devices capable of binding to the first of said two inputs, such that the second input and the signal molecule of the second positive regulator device displace the latch domains initially hybridized to said aptamer domains, respectively, and

wherein both inputs are present, the first of said two inputs binds to the aptamer domain of the first of said two effector devices capable of binding to the first of said two inputs and the second of said two inputs binds to the aptamer domain of the first of said two effector devices capable of binding to the second of said two inputs, such that said inputs displace the latch domains initially hybridized to said aptamer domains, respectively, and each one of said two signal molecules of the negative regulator device hybridizes to a different one of said latch domains of said first of said two effector devices thereby promoting closure of the effector device or preventing transition of the first of said two effector devices to an open configuration; and

wherein the second of said two outputs of the gate is in a logical on state if only the first of said two inputs is present, when the first of said two inputs binds to each one of the aptamer domains of the second of said two effector devices capable of binding to the first of said two inputs such that each signal molecule displaces a different one of said latch domains.

[0087] In certain embodiments of the various logic gates of the present invention, i.e., NAND, NOT, OR, XOR, and controlled NOT gates, when said effector device(s) is in an open configuration, each one of the payload is positioned on an outer surface of the effector device(s).

[0088] In this work we focused on artificial quorum sensing (QS) [Bassler, B. L. et al, 2006] as a bio-inspired collective behavior. QS is a phenomenon where microorganisms communicate and coordinate their behavior by the accumulation of signaling molecules. QS enables bacteria to regulate phenotypes such as biofilm formation [Davies, D. G. et al., 1998], bioluminescence [Engebrecht, J. et al., 1984], sporulation [Grossman, A. D. et al., 1995] and virulence factor secretion [Miller, M. B. et al., 2002; Hentzer, M. et al., 2003], by population density. Although group behaviors were not initially thought to exist in bacteria, they are now known to occur in many species and likely appeared early in evolution [Lerat, E. et al., 2004]. Population density-dependent group behaviors are also common in animals [Bowler, D. E. et al., 2005; Hauzy, C. et al, 2007; Kim, S. Y. et al, 2009; Karels, T. J. et al, 2000], and are mediated by neural-hormonal-behavioral circuits [Nephew, B. C. et al, 2003].

[0089] Collective behavior is a powerful paradigm in computing and robotics, and QS was demonstrated as a strategy to achieve this goal. Interestingly, QS has been proposed as a mechanism to coordinate behaviors of nanorobot swarms. Particularly, nanorobots exhibiting collective behaviors could have an enormous impact on medicine. However, this technology has not been implemented in reality as yet. Here we demonstrate a preliminary design towards this goal, which is based on the DNA nanorobot described above.

[0090] QS is based on a signal generated by each individual nanorobot, which builds up in the environment proportionally to the nanorobot population density. Subsequently, each nanorobot has to be able to detect this signal and respond to it by a concentration-dependent mechanism. The DNA nanorobot is activated in response to protein molecules termed "keys" through an aptamer-based gate [Douglas, S. M. et al, 2012] in a concentration-dependent manner, making these proteins good candidates for use as QS signals.

[0091] It has been found in accordance with the present invention that nucleic acid origami devices carrying quorum sensing molecules may aggregate in a concentration dependent manner both in vivo and in vitro as shown in Example 2. In this Example it is also shown that nanorobots can be made to aggregate by contacting one or more oligonucleotides presented on the outer surface of each individual nanorobot that is complementary to one or more other oligonucleotide also presented on the outer surface of each individual nanorobot. Furthermore, Example 3 below shows that the delivery of a drug may be controlled so that the drug is presented to the cells only when the nucleic acid origami devices carrying the drug have accumulated to a high concentration. In this way, the target cells do not experience low levels of the drug that may enable them to develop resistance, as is often the case when the drug is given systemically, but are challenged with a high level of the drug that is likely to kill all cells in a short time span.

[0092] Thus, in another aspect, the present invention relates to a system exhibiting quorum sensing comprising a plurality of effector nucleic acid origami devices each having the structure

A, B or C as defined above in connection with the system acting as a logic gate or, alternatively, a plurality of a member each selected from a liposome, a particle or an artificial cell; at least one input; and at least one output, wherein said plurality of effector nucleic acid origami devices or members aggregates at a predetermined concentration, and wherein: (i) each one of the effector nucleic acid origami devices comprises said handle domain bound to a first oligonucleotide capable of hybridizing with said aptamer domain of another of said effector nucleic acid origami devices and positioned on the outer surface of the device when the device is in the closed configuration; or (ii) each one of the members comprises on an outer surface thereof (a) a staple strand comprising a handle domain bound to a first oligonucleotide; and (b) a further staple strand comprising a second oligonucleotide selected such that it is capable of hybridizing to a first oligonucleotide of another of said plurality of members, each one of said effector nucleic acid origami devices or members having on open and a closed configuration and is initially in the closed configuration.

[0093] In certain embodiments, upon aggregation of said plurality of effector nucleic acid origami devices, said first oligonucleotide of each one of said effector nucleic acid origami devices binds to said aptamer domain of another of said effector nucleic acid origami devices such that the first oligonucleotide, acting as the first of said at least one input, displaces the latch domain of said another of said effector nucleic acid origami devices, thereby promoting transition of said another of said effector nucleic acid origami devices to an open configuration.

[0094] In certain embodiments, each one of said effector nucleic acid origami devices or each one of said members comprises yet a further staple strand comprising an aptamer capable of binding to a binding partner acting as another of said at least one input, in particular wherein said binding partner is a target antigen. In this way, the effector nucleic acid origami devices are likely to aggregate and be activated at the desired target. [0095] In other embodiments, each one of said effector nucleic acid origami devices or, in particular, each one of said members, further comprises a handle domain bound to a quorum sensing molecule, optionally via a linker, said quorum sensing molecule acting as a further one of said at least one input.

[0096] In particular embodiments, said linker comprises a third oligonucleotide having a sequence complementary to the sequence of the handle domain and optionally a further domain comprising a recognition site for enzymatic cleavage, and the quorum sensing molecule is bound to the handle domain through the hybridization of the third oligonucleotide to the handle domain. This further domain may comprise a peptide linker comprising a protease recognition site for cleavage by a protease, such as a matrix metalloproteinase.

[0097] In some embodiments, each one of said effector nucleic acid origami devices or members comprises still a further staple strand positioned on its outer surface, said further staple strand comprising a fourth oligonucleotide selected such that it is capable of hybridizing to said handle domain of another of said effector nucleic acid origami devices or members, such that the fourth oligonucleotide displaces the third oligonucleotide of said another of said effector nucleic acid origami devices or members, linked to the quorum sensing molecule, thereby releasing said quorum sensing molecule linked to said third oligonucleotide. In this way, the concentration of the released quorum sensing molecule represents the concentration of the effector nucleic acid origami devices or members present.

[0098] In certain embodiments, the released quorum sensing molecule is capable of causing transition of one of said effector nucleic acid origami devices or members from its closed configuration to its open configuration. In particular, the quorum sensing molecule is selected from a channel activating agent and a membrane permeable compound such as a peptide.

[0099] A non-liming example of how a peptide may function as a quorum sensing molecule is liposomes that can be triggered by an external signal such as a small peptide, that is capable of diffusing into the liposome and trigger a biosynthetic machinery. Moreover, we have demonstrated DNA arms decorated with a positively-charged peptide (we used GRKKRRQRRRPQ; SEQ ID O: 2) on the end, which can by themselves penetrate into the liposome and trigger its machinery. The process is as follows: 1) DNA arms on separate liposomes interact based on liposome density; 2) arrn interaction displaces a DNA strand containing a positive peptide as in above; 3) the displaced arm with peptide penetrates into the liposome and activates its internal machinery. Since this process initiates with strand displacement (step 2), which occurs only upon a certain density of liposomes (step 1), it represents quorum sensing. [00100] In certain embodiments the quorum sensing peptide is capable of penetrating membranes and in particular embodiments has the sequence KPLGMWSRC SEQ ID NO: 3) or GRKKRRQRRRPQ.

[00101] The concentration at which the effector nucleic acid origami devices or members of the quorum sensing system aggregate can be calibrated by adjusting the affinities between the interacting oligonucleotides or quorum sensing molecules attached to the devices and members.

[00102] In certain embodiments, said further nucleic acid molecule is capable of hybridizing to said handle domain of another of said plurality of at least one effector device or to said handle domain of another of said plurality of said member if the binding affinity of said handle domain to said further nucleic acid molecule is higher than the binding affinity of said handle domain to said nucleic acid molecule linked to the quorum sensing molecule.

[00103] In particular embodiments, the association constant of said binding affinity of said handle domain to said further nucleic acid molecule is selected by adjusting the degree of complementarity of said further nucleic acid molecule with said handle domain; and the association constant of said binding affinity of said handle domain to said nucleic acid molecule linked to the quorum sensing molecule is selected by adjusting the degree of complementarity of said nucleic acid molecule linked to the quorum sensing molecule with said handle domain.

[00104] In further embodiments, each one of said association constants of said binding affinity of said handle domain to said further nucleic acid molecule and to said nucleic acid molecule linked to the quorum sensing molecule is selected such that the binding affinity of said handle domain to said further nucleic acid molecule is higher than the binding affinity of said handle domain to said nucleic acid molecule linked to the quorum sensing molecule at a predetermined concentration of said further nucleic acid molecule.

[00105] In a further aspect, the present invention provides a pharmaceutical composition comprising a system acting as a logic gate or a system exhibiting quorum sensing in any one of the configurations defined above, and a pharmaceutically acceptable carrier.

[00106] Pharmaceutical compositions for use in accordance with the present invention may be formulated in conventional manner using one or more physiologically acceptable carriers or excipients. The carrier(s) must be "acceptable" in the sense of being compatible with the other ingredients of the composition and not deleterious to the recipient thereof.

[00107] The following exemplification of carriers, modes of administration, dosage forms, etc., are listed as known possibilities from which the carriers, modes of administration, dosage forms, etc., may be selected for use with the present invention. Those of ordinary skill in the art will understand, however, that any given formulation and mode of administration selected should first be tested to determine that it achieves the desired results. [00108] Methods of administration include, but are not limited to, parenteral, e.g., intravenous, intraperitoneal, intramuscular, subcutaneous, mucosal (e.g., oral, intranasal, buccal, vaginal, rectal, intraocular), intrathecal, topical and intradermal routes. Administration can be systemic or local. In certain embodiments, the pharmaceutical composition is adapted for intra- brain administration.

[00109] The term "carrier" refers to a diluent, adjuvant, excipient, or vehicle with which the active agent is administered. The carriers in the pharmaceutical composition may comprise a binder, such as microcrystalline cellulose, polyvinylpyrrolidone (polyvidone or povidone), gum tragacanth, gelatin, starch, lactose or lactose monohydrate; a disintegrating agent, such as alginic acid, maize starch and the like; a lubricant or surfactant, such as magnesium stearate, or sodium lauryl sulphate; and a glidant, such as colloidal silicon dioxide.

[00110] The compositions may be formulated for parenteral administration by injection, e.g., by bolus injection or continuous infusion. Formulations for injection may be presented in unit dosage form, e.g., in ampoules or in multidose containers, with an added preservative. The compositions may take such forms as suspensions, solutions or emulsions in oily or aqueous vehicles, and may contain formulatory agents such as suspending, stabilizing and/or dispersing agents. Alternatively, the active ingredient may be in powder form for constitution with a suitable vehicle, e.g., sterile pyrogen free water, before use.

[00111] For administration by inhalation, for example for nasal administration, the compositions for use according to the present invention are conveniently delivered in the form of an aerosol spray presentation from pressurized packs or a nebulizer, with the use of a suitable propellant, e.g., dichlorodifluoromethane, trichlorofluoromethane, dichlorotetrafluoroethane, carbon dioxide or other suitable gas. In the case of a pressurized aerosol the dosage unit may be determined by providing a valve to deliver a metered amount. Capsules and cartridges of, e.g., gelatin, for use in an inhaler or insufflator may be formulated containing a powder mix of the compound and a suitable powder base such as lactose or starch.

[00112] In certain embodiments the pharmaceutical composition is formulated for administration by any known method as described above.

[00113] The invention will now be illustrated by the following non-limiting examples.

EXAMPLES

Material and Methods

[00114] DNA scaffold. A 7249bp circular single-strand DNA molecule was used in the following examples (M13mpl8 DNA; New England Biolabs; NEB #N4040; SEQ ID NO: 4. [00115] Staples. Purchased from Integrated DNA Technologies SEQ ID NOs: 5-282. All sequences are in the 5' to 3' direction. The oligonucleotides were reconstituted in ultrapure,

DNase/RNase-free water to 100 μΜ concentration and stored at -20°C.

[00116] Robot preparation. Robots were initially produced by mixing M13mpl8 ssDNA as scaffold strand (final concentration of 20 nM) and staple strands (final concentrations of 200 nM of each strand). Buffer and salts of solution included 5mM Tris, ImM EDTA (pH 8.0 at 20 °C) and 10 mM MgCl ₂ . The mixture was subjected to a thermal- annealing ramp for folding. Initially the following program was used: 80 °C to 60 °C at 2 min/°C, 60 °C to 20 °C at 150 min/°C.

[00117] Purification of folded robots: After folding, excess staples were removed by centrifugal filtration using Amicon Ultra-0.5 mL 100K centrifugal filters (Millipore). Folding buffer was added to reach a total volume of 500 μί, after which samples were centrifuged at 12,000 g for lOmin. this was repeated three times. DNA concentration was measured by spectrophotometer (Thermo Sci. NanoDrop 2000c).

[00118] Gel purification of folded samples. Leading monomer bands were visualized on a UV table and excised from a 1.5%-2% agarose gel (running buffer is 0.5X TBE supplemented with 10 mM MgCl ₂ ), frozen at -20°C for 5 min, chopped to small pieces and centrifuged at 13,000 g for 3 min inside a Quantum Prep Freeze N' Squeeze DNA Gel Extraction spin column (Bio-Rad). Recovered solution was measured for DNA concentration by spectrophotometer (Thermo Sci. NanoDrop 2000c) and prepared for imaging by transmission electron microscopy (TEM) (Figs. 1A-B).

[00119] TEM negative-stain. Briefly, 5 μΐ _^ of 0.5 M NaOH were added to a pre-made frozen aliquot of 100 μΐ _^ 2% uranyl formate solution (Poly sciences, 24762) followed by rigorous vortexing for 3 minutes, after which solution was centrifuged at 18,000 g for 5 minutes and precipitate was removed. Robot samples at 1-5 nM concentration were loaded onto a TEM Grid (Science Services, EFCF400-Cu-50) immediately after glow-discharge treatment (Emitech K100X), followed by two consecutive washes with 0.1% uranyl formate solution. During the third wash the grid was incubated with uranyl formate solution for 30 seconds. Samples were visualized using a TEM microscope (JEM- 1400, JEOL) an hour to one week after negative staining. The robots were folded during 80-60 °C at 2 min/°C and 60- 10 °C at 150 min/°C or during 80-60 °C at 5 min/°C, 60-10 °C at 75 min/°C (Not shown). A graphic depiction of the nanorobots used in the examples is shown in Fig. 2.

[00120] Payload preparation. Antibodies were digested using a commercial kit (Thermo) with immobilized ficin in mouse IgG digestion buffer with 25 mM cysteine by shaking at a 37 °C water bath for 4 hours. Antibody Fab' fragments were purified by centrifugal filtration (Amicon, 10K MWCO Millipore) and evaluated by spectrophotometer (Thermo Sci. NanoDrop 2000c). Fab' fragments were buffer-exchanged into 0.05 M sodium borate buffer, pH 8.5, and incubated with DyLight Amine-Reactive Dye (Thermo) for 1 hour at room temperature on a rotary shaker. Excess dye was thoroughly removed using Amicon 10K MWCO (Millipore). Fab' fragments were incubated for 1 minute with 5'-amine-modified linker oligonucleotide (5AmMC6/GAACTGGAGTAGCAC (SEQ ID NO: 283 , Integrated DNA Technologies) at a molar ratio of 1 to 10, in a 0.1 M MES -buffered saline, pH 4.7 (Pierce #28390). EDC (Thermo, #22980) was added at a molar ratio of 5000 to 1 Fab' fragment and incubated at room temperature for 1 hour on a rotary shaker. Afterwards, Tris was added to a final concentration of 10 mM and solution was filtered via Amicon column 30K MWCO (Millipore).

[00121] Loading of robot. Oligonucleo tide-Fab' concentration was evaluated via absorption at 260 and 280 nm. Loading was performed for 2 hours on a rotary shaker at room temperature in folding buffer (10 mM MgCl ₂ in IX TAE) at a 2-fold molar excess of payloads to loading sites. Finally, loaded robots were cleaned by centrifugal filtration with a 100K MWCO Amicon column (Millipore) as described above

Example 1. Robot interaction design.

[00122] Cue-driven collisions between robots were based on toehold-mediated strand displacement reactions. The basic design enabling collisions between DNA strands extending from E, P, and N robots was based on a 3-level interaction: the first level is the interaction of the cue itself (e.g. PDGF or VEGF in this study) with the sensing strands of the gates of all robots (E, P, and N). Alternatively, the E gates can also interact with either P or N as described in Figs. 3A C.

[00123] Interaction between E and P. Robots are depicted in Fig. 3A from a side view schematically as rectangles revolving around a vertical axis. The E gate complementary strand contains four regions (from 5' to 3'): 5 (toehold for collision with N), 4 (toehold for collision with P), 3* (partially complements with the aptamer), and 1 (anchors the gate to the robot chassis). The sensing strand contains two regions (from 5' to 3'): 2 (anchors the gate to the robot chassis) and 3 (the aptamer). The key loaded into P contains 3 regions (from 5' to 3'): 3 (portion of the sensing strand which only appear identical in this scheme), 4* (hybridizes with toehold 4), and 6 (loading sequence to robot).

[00124] Interaction between E and N. E robots are shown in Fig. 3B first from a front view with the virtual sides stated for proper orientation. The arms (clasps) extending from N include 4 regions (from 5' to 3'): 3 (portion of the sensing strand which only appear identical in this scheme), 4* (hybridizes with toehold 4), 5* (hybridizes with toehold 5), and 6. Multiple arms extending from N clasp the two gates of E. Since these are located on opposite sides of the robot

(up and down), the result is inability of E to open properly.

[00125] Structural basis for differential keying of E and F robots. E and F consist of different regions dictating collisions with P and N. PI and P2 robots key E by interacting with region 4 on the E gate (Fig. 3C). N robots close E by interacting with region 5. Since F robots contain neither region 4 nor 5, robots PI, P2 and N cannot interact with it. However, P3 keys F by interacting with region 7, which is also why it cannot key E.

[00126] It is important to note that the N gate complementary strands lack regions 5 and 4, so the N clasp arms cannot hybridize with them and inactivate N itself.

[00127] Based on this basic design, collision-mediating sequences were chosen and the resulting systems EP, EN and EPN were modeled and prototyped in visual DSD (vDSD; Lakin et al. (2011) Microsoft Research 27: 3211-3213). The sequences were then altered as necessary to achieve the desired performance and kinetics.

[00128] Simulated kinetics of the EP complex. E and P robots in varying stoichiometries - from 0.1 to 10 - were simulated in vDSD (not shown), and it was found that the fastest reaction occurs when P is at a molar excess of 5 over E, consistent with the ex-vivo prototyping.

Example 2: A nanorobotic collision-based computer controlling therapeutic molecules in a living animal

[00129] A scheme showing the nanorobot used in these studies is shown in Fig. 2 and Fig. 4A. The insect Blaberus discoidalis was chosen as model animal for various reasons, including simplicity, previous experience with closely related models, and availability of reliable custom made reagents [Bulmer, M. S. et ah, 2009]. Importantly, the hemolymph of B. discoidalis expresses negligible nuclease activity and contains large amounts of free DNA as reported for other Dictyoptera [Garbutt, J. S. et ah, 2012], and its salt/metal composition well supports DNA origami structures. Thus, the DNA origami nanorobots can survive for relatively long periods in the insect hemolymph. Resident DNA fragments are used here as endogenous biological cues. Stability in hemolymph was confirmed by quantitative PCR (not shown), and DNA cues were defined by sequencing (not shown).

[00130] The system was first examined on isolated insect hemocytes ex-vivo, with fluorescent anti-insect antibodies loaded as the effector payload, and synthetic DNA fragments taken from the sequencing data as cues. A typical nanorobot mixture contained 0.1 pmol of the effector nanorobot, and 0.1 pmol or more of each regulator nanorobot, as described below. E architecture exhibited AND gate behavior as previously observed with protein cues (Fig. 4B and Fig. 5A). For clarity, the truth table for the AND gate is shown below (k represents "key"): Truth table for AND gate

[00131] EP1P2 architecture exhibited partial to good activity at a 1 : 1 : 1 stoichiometry, and good activity at 1 :>2:>2, which correlated well with an OR gate behavior. Binding of the target cells by an EP1 or EP2 nanorobot complex instead of just by solitary E nanorobots was shown by loading PI and P2 with a reporter payload emitting at a wavelength different from that of the effector payload (Fig. 4C and Fig. 5B). For clarity, the truth table for the OR gate is shown below:

Truth table for OR gate

k ! k2 1*1 l»2 Iv

o 0 0 0

o 1 0 1

I Bill 1

[00132] The third architecture, EP1P2N, performed similar to EP1P2 (OR) at first. XOR behavior emerged only when N was at a molar excess of -10 over E (not shown).

Truth table for XOR gate

k l k2 PI 1*2 \ K

o 0 0

1 iiiiiiiiiiiiiiiiiiii 0

[00133] EP1P2N architecture integrates XOR (E output) with AND (N output), thus emulating a 1- bit adder in which the sum bit is relayed to the effector payload by E in the EP1 and EP2 complexes. Hence, it can be programmed to respond to a cue count rather than to the identity of particular cues. The carry bit cannot be relayed by N as both nanorobots in the EN complex (unlike in the EP complexes) directly face each other and sterically prevent each other from binding target cells; however it can be relayed by adding a second effector nanorobot (F), which is not keyed by PI and P2 and not negated by N. Thus, X or Y alone activate E, while XY activate only F (E is negated by N) (Fig. 4D and Fig. 5C). For clarity, the truth table for the XOR gate is shown above.

[00134] Of particular importance is the emulation of functionally complete gates, such as NAND and NOR, which can be cascaded to build any other gate. NAND architecture can be based on two nanorobot types: Eopen, which is similar to E but lacks gate strands, hence it is constitutively open regardless of the inputs X and Y; and N, which opens in response to XY and negates Eopen as a result. This architecture therefore produces 0 in response to XY, and 1 otherwise (Fig. 4E), and performed well at a 1:5 stoichiometry of EopenN (Fig. 5D). For clarity, the truth table for the NAND gate is shown below:

Truth table for NAND gate

[00135] A simple inverter was constructed from this NAND gate (resulting in a NOT gate), by programming N to respond to X only (by placing two identical X arms instead of X and Y arms on the nanorobot), termed NX. EopenNX architecture produces 1 if X is absent (by Eopen) and 0 if X is present (by NX negating Eopen) (Fig. 4F), and performed well on insect cells at a 1:5 stoichiometry (Fig. 5E). The truth table for the NX gate is shown below:

Truth table for NX gate

[00136] Architectures based on more than one effector nanorobot (like the EFP1P2N mentioned above) enable to relay output bits to additional therapeutic molecules, forming the basis for more complex gates, reversible logic and binary decoders. To demonstrate such architecture, a controlled NOT (CNOT) gate was designed, consisting of E, PI, P2 and N as described above, and in addition a second effector nanorobot, F, not negated by N, and a positive regulator nanorobot P3 that responds only to X and keys only F. Thus, one input bit is XORed by

E and the second output bit is mapped unchanged by F (Fig. 5F). Architectures E (AND; Fig. 6A and Fig. 7), EP1P2 (OR; Fig. 7), EP1P2N (XOR; Fig. 6B and Fig. 7), Eopen:N (NAND; Fig. 6C and Fig. 7), and EFP1P2P3N (CNOT; not shown), prepared at the stoichiometries defined ex-vivo, were examined in living insects and produced good outputs.

[00137] A central feature of the collision-based system presented here, as in other forms of unconventional computing, is reversible logic [Liu, D. et ah, 2011; Cervera, J. et ah, 2010]. This feature enables the original cues to be explicitly inferred from the final states of the nanorobots, consequently revealing important information about the physiological and biochemical condition of the animal. Hence, the system can be used as a diagnostic tool by extracting activated nanorobot complexes and reading their output.

[00138] Notably, while aptamer-based gates are very difficult to model in terms of opening kinetics and statistics, toehold-driven displacement gating can be reproducibly designed [Zhang, D. Y. et ah, 2011; Zhang, D. Y. et ah, 2009] and proved very efficient in this study; however, this obligates the use of nucleic acid cues, ostensibly limiting the application range of our design.

[00139] Our findings demonstrate the feasibility of stochastic collision-based computing with diffusing particles in a biological system. The nanorobots, with a -55,000 cubic nanometer volume and 5 megadalton mass, have a diffusion coefficient in the medium, which is adequate for efficient collisions. However, the findings also emphasize the importance of stoichiometry. Regulators were required at higher quantities than effectors to ensure that effector-regulator collisions result in a high incidence of proper complexes. At a 1: 1 ratio, complex formation is plausibly Poisson distributed around -50% effector-regulator pairs, and -25% of uncomplexed effectors and regulators. The architectures described here are capable of processing two bits at most. However, the outputs from two "processors" can be cascaded to a third one and increase the processing capacity from 4, 8, 16 bits and so on, with the capacity limited only by the possible number of unique gate-key systems that can be designed.

[00140] The logical NOT gate explained above can be utilized for a so called Asimov multitype system, in which an effector robot which potentially may harm normal cells is prevented from doing so according to one of the laws of robotics devised by Isaac Asimov which states that "[a] robot may not injure a human being or, through inaction, allow a human being to come to harm". A NOT gate used in an Asimov system may consist of two robot populations: (1) Eo (Eopen) robots - carrying the effector molecule (e.g. drug); and (2) N robots - which close Eo robots in response to damage (i.e. the presence of a damage indicator molecule).

[00141] Whenever damage indicators leak outside of off-target cells (e.g. a microRNA reflecting damage of normal cells, such as miR-16), N becomes active, and closes E. This interaction requires N to be in a molar excess of -10 over Eo (Fig. 8A). If one damage indicator is not enough to achieve damaged cell selectivity and discrimination, a multitype Asimov system can also be achieved by a NAND gate. By combining two or more indicators, the identity of the damaged cell can be inferred more precisely (Fig. 8B).

Example 3: DNA nanorobots exhibiting group behavior

[00142] A QS signal generator was constructed by chemically conjugating three components: a selected key, a cleavable tether; and a DNA tail (Ulc) for loading onto a nanorobot loading site as previously described [Douglas, S. M. et ah, 2012]. The tether is designed to undergo cleavage and generate a diffusing signal in response to certain stimuli. As tethers, we used the peptide tether sequence KPLGMWSRC (SEQ ID NO: 3), which is cleaved by matrix metalloproteinases (MMP) [Park, H. I. et ah, 2002]. As key molecule, human platelet-derived growth factor BB (PDGF-BB) was chosen first (Fig. 9). The signal generator Ulc-KPLGMWSRC-PDGF-BB was loaded into the nanorobots, which were then placed in MMP-2-containing buffer at various population densities (from 29 to 18,000 pM nanorobots. Dynamic light scattering demonstrated a density-dependent structural change of the nanorobot indicating switching between states in agreement with previously-described observations ¹ (Fig. 10). This observation was repeated using nanorobots loaded with GFP in parallel with the signal generator and mixed with microspheres coated with anti-GFP antibody. Upon addition of MMP-2, the generated signal induced a concentration-dependent activation of the nanorobots and exposure of GFP, subsequently guiding their attachment to the microspheres. Flow cytometric analysis revealed that the behavior of gated nanorobots correlates with the gate's affinity to the QS signal, while unlocked nanorobots showed slightly shifted kinetics correlating with the higher affinity of GFP to its antibody (Fig. 11).

[00143] This behavior was also examined in a living animal, using the insect Blaberus discoidalis as a model organism. Here, a different signal generator was constructed using the tether sequence KIEARC (SEQ ID NO: 1), a substrate of insect prophenol oxidase-activating protease (PAP) [Gupta, S. et ah, 2005]. Nanorobots loaded with this generator and with fluorescently labeled anti-insect hemocytes antibody were injected into the hemocoel of adult B. discoidalis in increasing quantities. Later, the hemolymph was extracted and hemocytes were analyzed by flow cytometry to reveal the state (open/closed) of the nanorobot population. Our analysis demonstrated a density dependent transition of the nanorobots to an active state in the living insect, relayed to controlling a cell-interacting molecule (Fig 12).

[00144] A potential drawback of the design showed thus far is that the signal generator operates immediately and only once, making it impossible to coordinate a group of nanorobots over a long period. To address this, a continuous signal generator was constructed by storing large amount of QS signal inside small liposomes, which were linked to the nanorobots. The nanorobots were capable of releasing QS signals to the environment over a period of time (not shown).

[00145] The nanorobots can be made to aggregate by contacting one or more oligonucleotides presented on the outer surface of each individual nanorobot that is complementary to one or more other oligonucleotide also presented on the outer surface of each individual nanorobot (Fig. 13). The hybridization of each pair of oligonucleotides cause the nanorobots to aggregate. If one of the oligonucleotides is an aptamer initially hybridized to a latch domain, which together keep the nanorobot in a closed configuration, then the hybridization of the other nucleotide to the aptamer will cause the displacement of the latch domain and subsequent opening of the nanorobot. Thus, the interaction between the two oligonucleotides present on the outer surface of the nanorobots may cause both aggregation and opening of the nanorobots. In this way, a drug stowed away inside the nanorobots when in a closed configuration can be released only when the concentration of the nanorobots is high enough to allow aggregation and opening.

[00146] The other oligonucleotide mentioned above may be initially kept in an inactive form by binding to a complementary aptamer oligonucleotide. When a ligand binds to the aptamer, the other oligonucleotide is released and free to interact with complementary oligonucleotides on other nanorobots as described above (Fig 14). The size of an individual nanorobot, as assessed by measuring light scattering, is larger when it is in a closed configuration than when it is in an open configuration. The results (not shown; see figure 10 for a similar result) shows that, indeed, the diameter of the individual nanorobots decrease with the increase in nanorobot concentration (density).

Example 4: Controlled delivery of drugs by nanorobots exhibiting quorum sensing.

[00147] Robots loaded with a cancer drug were mixed in various concentrations with cancer cells. In a separate sample the free drug was mixed too. The results demonstrate that the robot-induced effect is different than the drug-induced one and appears at different concentrations. In fact, quorum sensing here improved the therapeutic window of the drug because it enabled a higher dose to accumulate without being active until the defined dose was reached (Fig 15). REFERENCES

Adamatzky, A., Collision-based computing, 2002, Springer.

Adamatzky, A., de Lacy Costello, B., Collision-free path planning in the Belousov- Zhabotinsky medium assisted by a cellular automaton. Die Naturwissenschaften, 2002, 89, 474- 478.

Adar, R., Benenson, Y., Linshiz, G., Rosner, A., Tishby, N., Shapiro, E., Stochastic computing with biomolecular automata. Proceedings of the National Academy of Sciences of the United States of America, 2004, 101, 9960-9965

Adleman, L. M., Molecular computation of solutions to combinatorial problems. Science, 1994, 266, 1021-1024

Bassler, B. L., Losick, R., Bacterially speaking. Cell, 2006, 125, 237-246

Benenson, Y., Gil, B., Ben-Dor, U., Adar, R., Shapiro, E., An autonomous molecular computer for logical control of gene expression. Nature, 2004, 429, 423-429

Benenson, Y., Paz-Elizur, T., Adar, R., Keinan, E., Livneh, Z., Shapiro, E., Programmable and autonomous computing machine made of biomolecules. Nature, 2001, 414, 430-434

Bowler, D. E., Benton, T. G., Causes and consequences of animal dispersal strategies: relating individual behaviour to spatial dynamics. Biological reviews of the Cambridge Philosophical Society, 2005, 80, 205-225

Bulmer, M. S., Bachelet, I., Raman, R., Rosengaus, R. B., Sasisekharan, R., Targeting an antimicrobial effector function in insect immunity as a pest control strategy. Proceedings of the National Academy of Sciences of the United States of America, 2009, 106, 12652-12657

Braich, R. S., Chelyapov, N., Johnson, C, Rothemund, P. W., Adleman, L., Solution of a 20-variable 3-SAT problem on a DNA computer. Science, 2002, 296, 499-502

Cervera, J., Mafe, S., Multivalued and reversible logic gates implemented with metallic nanoparticles and organic ligands. Chemphyschem : a European journal of chemical physics and physical chemistry, 2010, 11, 1654-1658

Chang, W. L., Ho, M. S., Guo, M., Molecular solutions for the subset-sum problem on DNA-based supercomputing. Bio Systems, 2004, 73, 117-130

Chen, J. H., Seeman, N. C, Synthesis from DNA of a molecule with the connectivity of a cube. Nature, 1991, 350, 631-633

Dietz, H., Douglas, S. M., Shih, W. M. Folding DNA into twisted and curved nanoscale shapes. Science, 2009, 325, 725-730

Douglas, S. M., Bachelet, I., Church, G. M. A logic-gated nanorobot for targeted transport of molecular payloads. Science, 2012, 335, 831-834 Douglas, S. M., Dietz, H., Liedl, T., Hogberg, B., Graf, F., Shih, W. M., Self-assembly of

DNA into nanoscale three-dimensional shapes. Nature, 2009, 459, 414-418

Engebrecht, J., Silverman, M., Identification of genes and gene products necessary for bacterial bioluminescence. Proceedings of the National Academy of Sciences of the United States of America, 1984, 81, 4154-4158

Enomoto, A., Moore, M., Nakano, T., A molecular communication system using a network of cytoskeletal filaments. Nanotechnology Conference and Trade Show, 2006, 725-728

Garbutt, J. S., Belles, X., Richards, E. H., Reynolds, S. E. Persistence of double stranded RNA in insect hemolymph as a potential determiner of RNA interference success: Evidence from Manduca sexta and Blattella germanica. Journal of insect physiology, 2012

Gupta, S., Wang, Y., Jiang, H., Manduca sexta prophenoloxidase (proPO) activation requires proPO-activating proteinase (PAP) and serine proteinase homologs (SPHs) simultaneously. Insect biochemistry and molecular biology, 2005, 35, 241-248

Gregori, M., Llatser, I., Cabellos-Aparicio, A., Alarcon, E., Physical channel characterization for medium-range nanonetworks using flagellated bacteria. Computer Networks, 2011, 55, 779-791

Gregori, M., Llatser, I., Cabellos-Aparicio, A., Alarcon, E., Physical channel characterization for medium-range nanonetworks using catalytic nanomotors. Nano Communication Networks, 2010, 1, 102-107

Gregorim M., Akyildizm I., A new nanonetwork architecture using flagellated bacteria and catalytic nanomotors. IEEE Journal on Selected Areas in Communications, 2010, 28, 612- 619

Grossman, A. D., Genetic networks controlling the initiation of sporulation and the development of genetic competence in Bacillus subtilis. Annual review of genetics, 1995, 29, 477-508

Hauzy, C, Hulot, F. D., Gins, A., Loreau, M., Intra- and interspecific density-dependent dispersal in an aquatic prey-predator system. The Journal of animal ecology, 2007, 76, 552-558

Hentzer, M., Wu, H., Andersen, J. B., Riedel, K., Rasmussen, T. B., Bagge, N., Kumar, N., Schembri, M. A., Song, Z., Kristoffersen, P., Manefield, M., Costerton, J. W., Molin, S., Eberl, L., Steinberg, P., Kjelleberg, S., H0iby, N., Givskov, M., Attenuation of Pseudomonas aeruginosa virulence by quorum sensing inhibitors. The EMBO journal, 2003, 22, 3803-3815

Hiyama, S., Inoue, T., Shima, T., Moritani , Y., Suda, T., Sutoh, K., Autonomous loading, transport, and unloading of specified cargoes by using DNA hybridization and biological motor-based motility. Small, 2008(a), 4, 410-415 Hiyama, S., Moritani, Y., Molecular communication: Harnessing biochemical materials to engineer biomimetic communication systems. Nano Communication Networks, 2010, 1, 20- 30

Hiyama, S., Moritani, Y., Suda, T., A biochemically-engineered molecular communication system. Nano-Net, 2009, 85-94

Hiyama, S., Moritani, Y., Suda, T., Molecular communication, Proceedings NSTI Nanotechnology Conference and Trade Show, 2005, 3, 391-394

Hiyama, S., Moritani, Y., Suda, T., Shima, T., Sutoh, K., An autonomous molecular transport system using DNAs and motor proteins in molecular communication, Bio-Inspired Models of Network, Information and Computing Systems, 2007, 135-138

Hiyama, S., Takeuchi, S., Gojo, R., Shima, T., Sutoh, K., Biomolecular motor-based cargo transporters with loading/unloading mechanisms on a micro-patterned DNA array. IEEE 21st International Conference on Micro Electro Mechanical Systems, 2008(b), 144-147

Karels, T. J., Boonstra, R., Concurrent density dependence and independence in populations of arctic ground squirrels. Nature, 2000, 408, 460-463

Kim, S. Y., Torres, R., Drummond, H., Simultaneous positive and negative density- dependent dispersal in a colonial bird species. Ecology, 2009, 90, 230-239

Lerat, E., Moran, N. A., The evolutionary history of quorum-sensing systems in bacteria. Molecular biology and evolution, 2004, 21, 903-913

Liu, D., Chen, W., Sun, K., Deng, K., Zhang, W., Wang, Z., Jiang, X., Resettable, multi- readout logic gates based on controllably reversible aggregation of gold nanoparticles. Angewandte Chemie International Edition, 2011, 50, 4103-4107

Llatser, I., Cabellos-Aparicio, A., Alarcon, E., Networking Challenges and Principles in Diffusion-based Molecular Communication, Wireless Communications, IEEE, 2012, 19, 36 - 41

Miller, M. B., Skorupski, K., Lenz, D. H., Taylor, R. K., Bassler, B. L., Parallel quorum sensing systems converge to regulate virulence in Vibrio cholerae. Cell, 2002, 110, 303-314

Moritani, Y., Hiyama, S., Nomura, S. M., Akiyoshi, K., Suda, T., A Communication interface using vesicles embedded with channel forming proteins in molecular communication. Bio-Inspired Models of Network, Information and Computing Systems, 2007, 147-149

Moritani, Y., Hiyama, S., Suda, T., Molecular communication among nanomachines using vesicles. Proceedings of 2006 NSTI Nanotechnology Conference, 2006, 705-708

Nakano, T., Suda, T., Moore, M., Molecular communication for nanomachines using intercellular calcium signaling. Proceedings of 2005 5th IEEE Conference on Nanotechnology, 2005 Nephew, B. C, Romero, L. M., Behavioral, physiological, and endocrine responses of starlings to acute increases in density. Hormones and behavior, 2003, 44, 222-232

Parcerisa Gine, L., Akyildiz, I. F., Molecular communication options for long range nanonetworks. Computer Networks, 2009, 53, 2753-2766

Park, H. I., Turk, B. E., Gerkema, F. E., Cantley, L. C, Sang, Q. X., Peptide substrate specificities and protein cleavage sites of human endometase/matrilysin-2/matrix metalloproteinase-26. The Journal of biological chemistry, 2002, 277, 35168-35175

Pierobon, M., Akyildiz, I., A physical end-to-end model for molecular communication in nanonetworks. IEEE Journal on Selected Areas in Communications, 2010, 28, 602-611

Qian, L., Winfree, E., Bruck, J., Neural network computation with DNA strand displacement cascades. Nature, 2011, 475, 368-372

Rinaudo, K., Bleris, L., Maddamsetti, R., Subramanian, S., Weiss, R., Benenson, Y., A universal RNAi-based logic evaluator that operates in mammalian cells. Nature biotechnology, 2007, 25, 795-801

Rothemund, P. W., Folding DNA to create nanoscale shapes and patterns. Nature, 2006, 440, 297-302

Seeman, N. C. Nucleic acid junctions and lattices. Journal of theoretical biology, 1982, 99, 237- 247

Suda, T., Nakano, T., Moore, M., Biologically inspired approaches to networks: The bio- networking architecture and the molecular communication. Bio-Inspired Computing and Communication. Lecture Notes in Computer Science, 2008, 5151, 241-254

Walsh, F., et al., Synthetic protocols for nano sensor transmitting platforms using enzyme and DNA based computing. Nano Communication Networks, 2010, 1, 50-62

Winfree, E., et al, Design and self-assembly of two-dimensional DNA crystals. Nature, 1998, 394, 539-544

Xie, Z., et al., Y. Multi-input RNAi based logic circuit for identification of specific cancer cells. Science, 2011, 333, 1307-1311

Yeh, C. W., Chu, C. P., Wu, K. R., Molecular solutions to the binary integer programming problem based on DNA computation. Bio Systems, 2006, 83, 56-66

Yin, Z., et al. Chinese Postman Problem based on DNA computing. Journal of chemical information and computer sciences, 2002, 42, 222-224

Zhang, D. Y., Seelig, G., Dynamic DNA nanotechnology using stranddisplacement reactions. Nature chemistry, 2011, 3, 103-113

Zhang, D. Y., Winfree, E., Control of DNA strand displacement kinetics using toehold exchange. Journal of the American Chemical Society , 2009, 131, 17303-17314