3.1 Reconfigurable Edge-modified DNA Tetrahedra for Sensing

The functionalization of the edge domains of DNA tetrahedra and their structural distortion upon interactions with auxiliary agents provides versatile means to use the DNA tetrahedra as sensing scaffolds. This is exemplified in Figure 6 with the multiplex analysis of miRNAs, aptamer-ligand complexes, or enzymes by the tetrahedra scaffold.⁵⁴ Figure 6A depicts a three-hairpins, H₁-, H₂-, and H₃-functionalized analysis of three different miRNAs, e.g., miRNA-21 by hairpin H₁, miRNA-221 by H₂, and miRNA-155 by H₃. The respective edges hybridized with the three hairpins frames were modified at the stem domains of H₁ with the FAM fluorophore/BHQ1 quencher, stem domain of H₂ with the ROX fluorophore/BHQ2 quencher, and stem domain of H₃ with the Cy5 fluorophore/BHQ2 quencher. Under these conditions, the fluorophore units existed in a quenched state. The loop regions of H₁, H₂, and H₃ included the recognition sequence for miRNA-21, miRNA-221 and miRNA-155, respectively. Subjecting the tetrahedra module to any of the miRNAs resulted in the selective opening of the respective hairpin units, and the miRNA concentration guided switching-on of the respective fluorophores, stimulated by the separation of the fluorophore-quencher pair, associated with the respective miRNA-responsive hairpins. Subjecting the tetrahedra sensing module to two different miRNAs or to three miRNAs enabled the multiplexed analysis of the respective miRNAs (Figure 6B). The miRNAs-guided optical sensing was further applied to selectively image miRNAs associated with specific cancer cells (Figure 6C). Subjecting MCF-7 breast cancer cells containing the miRNA-21 and miRNA-155 to the miRNA-responsive tetrahedra switches on the intracellular green fluorescence of FAM and the red fluorescence of Cy5, associated with the concomitant opening of hairpin H₁ and H₃. In turn, treatment of HepG2 liver cancer cells that include overexpressed miRNA-21 resulted in the selective green fluorescence of FAM, originating from the selective opening of H₁. Control experiments applying the tetrahedra sensing module on MCF-10 A epithelial breast cells, lacking any overexpressed miRNA, did not lead to any fluorescent signal, consisted with the caged configuration of the hairpin units in the sensing module.

In addition, a tetrahedra associated with hairpin on modified at their three edges with pre-engineered hairpins were applied for the optical detection of DNA cleavage enzymes⁵⁴ (Figure 6D). The tetrahedra edges included the FAM fluorophore/BHQ1 quencher pair, the ROX fluorophore/BHQ2 quencher pair, and the Cy5 fluorophore/BHQ2 quencher pair at the termini edge positions proximate to the hairpin stem domains. The stem regions of H₄, H₅, and H₆ included sequence-specific duplexes that are cleaved by the nicking enzyme Nt.BbvCI, and the endonuclease EcoRI, and HindIII, respectively. Treatment of the reaction module with any of the cleavage enzymes resulted in the selective, concentration-guided cleavage of the respective hairpins, leading to the switched-on fluorescence of FAM, ROX and Cy5, respectively. By subjecting the tetrahedra reaction module to any of the two enzymes or the three enzymes, the multiplexed analysis of the biocatalysts was demonstrated. The concept of hairpin-modified edges of tetrahedra was further realized with the multiplexed analysis of aptamer-ligand complexes.⁵⁵

The flexibility of designing the DNA tetrahedra structures was further used to develop an ATP sensing module⁵⁶ (Figure 7A). One of the edges of the tetrahedra was modified with the donor-acceptor pair (rhodamine green and dabcyl) that were spatially extended by the strand L hybridizing with the edge constituents. The strand L included, however, the ATP aptamer recognition sequence. In the presence of ATP, the loop structure of the ATP-aptamer complex forced the donor-acceptor pair into a spatially close configuration that resulted in effective ATP concentration-guided quenching of rhodamine green (Figure 7B).

3.2 Reconfigurable Corner-modified Tetrahedra for Sensing

A versatile approach to apply fluorescent tetrahedra functionalized nanostructures for in vitro sensing and intracellular optical imaging has involved the modification of corners of the tetrahedra with fluorescent labels. According to this approach, tethers associated with the corner fluorophores act as sensing elements that upon the recognition of the target analyte perturb the fluorescent probe associated with the tetrahedra thereby, transducing the sensing events. Alternatively, the edge domains of the tetrahedra were engineered to include the recognition sensing domains. The sensing events perturbed the edge structures, thereby distorting the fluorescent features of the corner fluorophores, thus allowing the optical transduction of the sensing processes.

Figure 8A depicts the multiplexed analysis of three different mRNAs translating c-myc, TK1 and GalNac−T proteins using a tetrahedra module that includes at three corners, tethers consisting of FAM/BHQ1, Cy3/BHQ2, and Cy5/BHQ2 fluorophore/quencher pairs-modified duplexes, where the fluorophore-modified strands are complementary to the C-myc, TK1, and GalNac−T mRNA targets hybridized with quencher-functionalized strand.⁵⁷ Subjecting the tetrahedra module to any of the respective mRNAs resulted in the concentration-guided removal of the respective fluorescent labels, thus allowing the selective and multiplexed analysis of the respective mRNAs (Figure 8B). This method was then applied to image the C-myc, TK1 and GalNac−T mRNAs overexpressed in MCF-7 breast cancer cells. The same concept was applied to develop a fluorophore/quencher corner-modified miRNA-responsive tetrahedra module for multiplexed sensing and cellular imaging of miRNAs.

Tetrahedra modified with different fluorophores were used as mitochondria-targeting probes imaging pH and Ca²⁺ ion changes in neurons.⁵⁸ The tetrahedron scaffolds were modified with AF660 fluorescent imaging tracer, fluorescein isothiocyanate (FITC) as responsive fluorophore, fluorescent C-dots (CD) modified with the calcium ligand (CaL) enhancing the fluorescence, and (3-azidopropyl) triphenylphosphonium bromide, TPP, as mitochondria targeting agent (Figure 9A). Under acidic conditions and in the presence of Ca²⁺-ions the FITC (λ_em=520 nm) is quenched while the CD fluorescence at 600 nm is enhanced (Figure 9B, Panels I and II). Applying the tetrahedra probe to wild-type neurons allowed the AF660 visualization of TPP targeting mitochondria, and pH and Ca²⁺-stimulated changes in the mitochondria by the respective probes (Figure 9C).

4.1 Physical Features of DNA Tetrahedra on Eectrode Surfaces

In general, the functionalization of the corners of tetrahedra scaffold with thiol functionalities provides a versatile means to assemble tetrahedra on metal, e.g., Au surfaces. By the functionalization of the DNA tetrahedra corners with one-, two- or three- thiol functionalities, it was demonstrated that the three thiol-functionalized tetrahedra form a stable dense monolayer on electrode surfaces, allowing the fourth vacant corner to be modified with a free flexible tether for subsequent anchoring of auxiliary agents, such as proteins or nucleic acids.

Understanding the charge transport properties of DNA tetrahedra scaffolds associated with conductive supports is a key feature to develop DNA tetrahedra-based electrochemical sensors.⁶⁰ This issue was addressed by the integration of Methylene Blue, MB, redox-active intercalator units into duplex sites of DNA tetrahedra scaffolds assembled through three thiol functionalities on an Au surface, where the MB⁺ units were positioned in spatially close or sterically remote in respect to the electrode surface (Figure 10A, panel I). The voltammetric responses of tetrahedra states were identical (Figure 10A, panel II) implying that the redox-active intercalator units experienced similar charge transport features within the DNA carrier scaffold. The cleavage of one of tetrahedra duplex edges (Figure 10B, panel I), significantly perturbed the voltammetric response of the MB intercalator (Figure 10B, panel II), indicating that an intact duplex configuration of tetrahedra framework is essential to retain the charge transport to the intercalated constituent. The flexible tethering of the redox label to tetrahedra scaffold is, however, significantly affected by the spatial position of the redox label within tetrahedra structure (Figure 10C). The voltammetric responses of ferrocene (a non-intercalating redox label) tethered to DNA tetrahedra at bottom, middle or top positions, panel I, are strongly affected by the distance separating the redox label from the electrode, panel II. The spatially close position redox label revealed a high electronical response, whereas the ferrocene units in a remote position demonstrated inefficient electron transfer communication with the electrode. Furthermore, the sizes of tetrahedra scaffold have an immense effect on the coverage of the electrodes by tetrahedra structures and on the control over the surface properties toward interaction with auxiliary strands, and the subsequent application of tetrahedra-modified surface for sensing.^49b As the size of tetrahedra scaffolds increases, the detection limit decreases consistent with the enhanced hybridization efficiencies of the larger tetrahedra scaffolds.

4.2 Electrochemical Sensing with DNA-tetrahedra Scaffolds

The possibility of modifying the corners of DNA tetrahedra with functional sequence-specific tethers was applied to develop diverse sensing platforms. Figure 11A exemplifies the electrochemical sensing of a target gene T by the DNA tetrahedra scaffold.⁶¹ A multi-thiol functionalized DNA tetrahedra structure was assembled on an Au electrode, and the corner of tetrahedra was modified with the tether T′ that was complementary to a part of the target gene, T. In the presence of the target gene T, the hybrid T′/T duplex was formed. The subsequent functionalization of the sensing interface with the biotinylated strand (L) that hybridized with the toehold single-strand of T, followed by the association of the avidin-HRP conjugate to the interface, resulted in the bioelectrocatalytic interface for the voltammetric detection of the gene, T. In the presence of 3, 3′, 5, 5′-Tetramethylbenzidine, TMB, as redox label, the HRP-mediated electrocatalyzed reaction of H₂O₂ proceeded. The resulting volumetric response related to the concentration of the target gene and enabled the analysis of T the gene with a detection limit corresponding to 1 pM (Figure 11B). Moreover, the DNA-tetrahedra-assisted electrochemical sensing of a target nucleic acid was further developed by demonstrating that integration of the nucleic acid recognition probe in-between the corners of two adjacent tetrahedra frameworks associated with the electrode, provides an efficient means to enhance the sensing efficacy.⁶² This enhancement was attributed to the stretching of the coiled analyte strand by the recognition probe positioned in between the two tetrahedra.

In addition, the conjugation of nucleic acid tethers to the DNA tetrahedra scaffold was further extended to include aptamer tethers and to develop electrochemical aptasensor devices, e.g., anti-exosome aptamer.⁶³ Alternatively, the structural reconfiguration of redox-labeled aptamer constituents associated with tetrahedra, as a result of formation of the aptamer-ligand complexes on tetrahedra, provided means to develop electrochemical aptasensors. For example, Figure 11C depicts the functionalization of two corners of tetrahedra with subunits of the ATP aptamer, where one of subunits is modified with methylene blue, MB, as redox label.⁶⁴ In the absence of ATP, the redox-label is spatially apart from the electrode surface, leading to inefficient electron transfer communication with the electrode and low voltammetric response. In the presence of ATP, the resulting reconfiguration of the ATP-aptamer subunits complex led to spatial proximity of the redox label to the electrode and to effective electron transfer communication and high voltammetric response. The ATP guided electrochemical responses of the electrode were controlled by the concentrations of ATP and allowed the analysis of ATP with a detection limit corresponding to 5 nM.

5.1 Dynamic Reconfiguration of DNA Tetrahedra Structure

The dynamic reversible and switchable oligomerization of DNA tetrahedra structures was accomplished by the triggered formation and dissociation of stimuli responsive bridging units⁶⁷ (Figure 12). In one system, the pH-stimulated dimerization of DNA tetrahedra structures and the reversible separation of the dimer structures were demonstrated by the application of T−A ⋅ T triplex nucleic acid bridging units (Figure 12A). Two tetrahedra T₁-functionzalized with a tether p and T₂-modified with a duplex tether q/o yielded the triplex (T−A ⋅ T)-bridged T₁−T₂ dimer structure at pH=7.0. At pH=10.0, the T−A ⋅ T bridges were dissociated leading to the separated tetrahedra structures. By labeling the tether p associated of T₁ with Cy5 and the duplex tether q/o with Cy3, the resulting intramolecular FRET signal in the triplex-stabilized tetrahedra dimer provided a readout signal for the dimerization process (Figure 12B). By the pH-stimulated formation and dissociation of the triplex-bridged T₁−T₂ dimer, the resulting FRET signal provided a readout signal for the cyclic dimerization/separation of the DNA tetrahedra dimer, inset. Similarly, by tethering to the DNA tetrahedra structures T₃ and T₄ with G-quadruplex subunits (x and y), the K⁺-ion-stimulated dimerization of tetrahedra T₃−T₄ by interlinking the K⁺-ion-stabilized G-quadruplex was demonstrated (Figure 12C). In the presence of 18-crown-6-ether, the G-quadruplexes were separated to yield the individual tetrahedra. The formation of the K⁺-ion stabilized G-quadruplex bridging units was followed by the incorporation of hemin into the G-quadruplex bridges yielding the hemin/G-quadruplex DNAzyme bridging units that stimulated the H₂O₂-catalyzed oxidation of Amplex Red to the fluorescent Resorufin that provided a readout signal for the dimerization of tetrahedra structures (Figure 12D). The cyclic K⁺-ion induced formation of the T₃−T₄ dimer, and the reverse crown-ether guided separation of the dimer was then probed by the ON/OFF switching of the DNAzyme-mediated H₂O₂ oxidation of Amplex Red to Resorufin, inset. The switchable dimerization/separation of tetrahedra T₁−T₂ and T₃−T₄ using pH or K⁺-ion/crown ether triggers was then applied to dictate the selective dimerization of T₁−T₂ or T₃−T₄ in the presence of the respective inputs (pH or K⁺-ions), resulting in the dictated FRET signal or generated by the hemin/G-quadruplex catalyzed oxidation of Amplex Red by H₂O₂. Alternatively, upon addition of the auxiliary tetrahedra T₅ that were modified with the tethers×and duplex q/o, to the mixture of T₁ and T₄, the oligomerization of tetrahedra into the trimer structure T₄−T₅−T₁ proceeded (Figure 12E). The formation of trimer was confirmed by the parallel transduction of the FRET signal and Resorufin fluorescence output signal and concomitant gel electrophoretic separation experiments. The triggered dynamic oligomerization of DNA tetrahedra structures was further applied to selectively sense miRNAs and to image cancer cells.⁶⁷

The dynamic dimerization of DNA tetrahedra structures was further examined in cell-like (protocell) environments by probing biocatalytic functions of the dimer structures and their interactions with biocatalysts (enzymes) in the confined media of the protocells.⁶⁸ Hydrogel microcapsules loaded with tetrahedra T₁ and T₂ (functionalized at their corner with G-quadruplex subunits) in the absence, or eventually the presence of an enzyme, e.g., glucose oxidase, were prepared according to Figure 13A. CaCO₃ microparticles impregnated with the DNA tetrahedra and enzyme were modified with poly(allylamine hydrochloride), PAH, and subsequently with a promoter nucleic acid strand (15). The particles were subjected to an aqueous solution that included the carboxymethyl cellulose polymer chains P₁ and P₂ modified with hairpins H₁ and H₂/(16). The hairpins H₁ and H₂ were engineered to allow the promoter-induced hybridization chain reaction (HCR) that resulted in the coating of the CaCO₃ core particles with a hydrogel film crosslinked by the opened duplexes H₁/H₂. The subsequent etching of the CaCO₃ core with ethylenediaminetetraacetic acid (EDTA) yielded the microcapsule containments loaded with the DNA tetrahedra (and eventually the enzyme). Subjecting the DNA microcapsules loaded with the tetrahedra T₁ and T₂ to K⁺-ions and hemin resulted in the dynamic K⁺-ion-stimulated stabilization of the hemin/G-quadruplex DNAzyme in the protocell assembly that catalyzed the oxidation of Amplex Red to the fluorescent Resorufin (Figure 13B). Subjecting the microcapsules to 18-crown-6-ether, CE, resulted in the separation of the hemin/G-quadruplex-bridged tetrahedra dimer and to the switched-off DNAzyme activity. By the cyclic permeation of K⁺-ions/CE into the microcapsules, the DNAzyme activity was reversibly switched between “ON” and “OFF” states (Figure 13C). In the presence of co-loaded glucose oxidase, GOx, in the protocell, and treatment of the system with glucose (Figure 13D) the biocatalytic cascade consisting of the aerobic oxidation of glucose to gluconic acid and H₂O₂, and the subsequent hemin/G-quadruplex-bridged dimer DNAzyme catalyzed oxidation of Amplex-Red to the fluorescent Resorufin proceeded in the protocell. The biocatalytic cascade operating in the protocell was found to be 5-fold enhanced as compared to the biocatalytic cascade operation by an equal content of biocatalysts in a diffusional random configuration (Figure 13E). The enhanced biocatalytic cascade in the confined environment of the protocell was attributed to the effective channeling of the GOx-generated H₂O₂, formed at high local concentration in the protocell, into the adjacent DNAzyme catalytic units.

5.2 DNA Tetrahedra Dynamically Switchable Nucleic Acid Machineries

DNA tetrahedra provide functional frameworks for the dynamic switchable operation of nucleic acid machineries.⁶⁹ This is exemplified in Figure 14 with the assembly of a DNA tetrahedron that guides the dynamic switchable operation of a transcription machinery. The tetrahedra include a caged inactive partial transcription template composed of the domains G/P domains and modified with the Cy3/Quncher labels, Figure 14A. In the presence of the “key” strand G′/P′/H, the tetrahedron framework is partially unlocked to yield the separated Cy3-labelled separated edge, acting as a promoter-activated transcription template, that in the presence of added T7RNA polymerase/NTPs yields in the presence of 3,5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI) the fluorescent DFHBI spinach aptamer complex. Subjecting the transcription machinery to the counter-locking strand G/P/H′ displaces the lock strand, resulting in the separation of the transcription template and the reconfiguration of the caged transcription-inactive DNA tetrahedra. The parent DNA tetrahedron and unlocking DNA tetrahedron could be imaged by AFM, Figure 14B, Panels I and II and by the switched fluorescence of the Cy3-labeled uncaged tetrahedra frameworks, Panel III. The “ON”/“OFF” switchable operation of the transcription machinery is displayed in Figure 14C. By mixing two functional DNA tetrahedra engineered to be unlocked selectively by two different “unlocking keys” yielding the DFHBI spinach aptamer- or/and the Malachite Green (MG) aptamer-generation transcription machinery, Figure 14D, the gated switchable operation of the two transcription machineries guided by the two tetrahedra was demonstrated, Figure 14E.

5.3 Dynamic Reconfiguration of DNA Tetrahedra Networks

The development of artificial dynamic networks mimicking biological networks attracts recent research interest as a part of the rapidly developing area of System Chemistry.⁷⁰ Dynamic biological networks reveal characteristic features including adaptive,⁷¹ hierarchically adaptive,⁷² switchable,⁷³ and amplified reactivity,⁷⁴ signal promotion⁷⁵ and information transfer capacities.⁷⁶ A simple dynamic network aiming to emulating some of the features of biological network is depicted in Figure 15. The network includes a mixture of four dynamically equilibrated constituents AA′, BB′, AB′ and BA′ composed of the components A, A′, B, B′. Subjecting the constitutional dynamic network (CDN) X to trigger T₁ that stabilizes constituent AA′, provides a promoting signal, guiding the dynamic reconfiguration of CDN X to CDN Y, where constituent AA′ is upregulated due to its stabilization, and on the expense of AB′ and BA′ that are simultaneously downregulated. The resulting separation of AB′ and BA′ yields, however, the concomitant upregulation of BB′, a constituent that does not share components with the stabilized constituent AA′. That is, stabilization of the constituent AA′ results in the adaptive dynamic downregulation of the antagonist constituents AB′ and BA′ and the concomitant upregulation of the agonist constituent BB′. Triggering CDN Y with the counter trigger T₁′, that depletes the stabilized trigger T₁, reconfigures CDN Y to the parent CDN X. Integrating CDN X to trigger T₂ that stabilizes constituent AB′ results in the triggered upregulation of AB′, the adaptive downregulation of the antagonist constituents AA′ and BB′, and the concomitant upregulation of the agonist constituent BA′. The information encoded in the bare sequence of nucleic acid structures provides versatile means to engineer structurally interchangeable equilibrated constituents that include embedded structural switchable motives that are triggered by auxiliary stimuli.^18c The structural reconfiguration of the constituents provides, then, versatile means to control the stabilities of constituents, and guide dynamic transitions of DNA networks.

Indeed, substantial progress in the design and dynamic operation of DNA-based CDNs were reported in recent years.^20d Dynamic networks revealing adaptive,⁷⁷ hierarchically adaptive,⁷⁸ feedback-driven⁷⁹ and cascaded and switchable⁸⁰ reconfigurable properties were demonstrated. Different auxiliary triggers, such as fuel/anti-fuel strands,⁷⁷ the formation and displacement of triplex structures,⁸¹ the K⁺-ion-induced formation of G-quadruplex and their separation by crown ether⁷⁷ or the use of photoisomerizable intercalators⁸² were applied to dynamically reconfigure the CDNs. Different application of dynamically controlled DNA network assemblies were reported, including the operation of switchable CDN-guided biocatalytic cascades⁸³ or the engineering of network guided formation of hydrogels revealing regulated stiffness for controlled drug release.⁶⁶

The synthetic means to construct DNA tetrahedra of controlled sizes (3–8 nm) close to the dimension of protein scaffolds, and the feasibility to modify tetrahedra with nucleic acid tethers allowing the dynamic switchable dimerization and separation of DNA tetrahedra structures, provides a versatile artificial means to mimic native protein-protein interactions by these nanostructures. Not surprisingly, the DNA tetrahedra frameworks were used as building blocks to construct constitution dynamic networks. The stimuli-triggered reconfiguration of these CDNs provided, then, means to model dynamic control over protein-protein interactions by the DNA tetrahedra nanostructures.

Figure 16A introduces a dynamic network, CDN P, consisting of four equilibrated different-sized tetrahedra dimers AA′, BB′, AB′, and BA′. Each of the dimers is bridged by a different Mg²⁺-ion-dependent DNAzyme consisting of appropriately engineered subunits.⁸⁴ The DNAzymes associated with the constituents differ in their substrate binding sites, and the catalytic cleavage rates of the respective fluorophore/quencher-modified substrate (Figure 16A, panel I), provide readout signals quantifying the contents of the equilibrated constituents. That is, by designing appropriate calibration curves relating the rates of the cleavage of the respective substrates to different concentration of the DNAzymes, and monitoring the catalytic rate of the DNAzymes associated with constituents in the CDNs, the contents of the constituents are quantified. Subjecting CDN P to the fuel trigger T₁ that hybridizes with AA′ leads to the reconfiguration of CDN P into CDN Q, where AA−T₁ is upregulated, AB′ and BA′ are downregulated and concomitantly the agonist tetrahedra-dimer BB′ is upregulated. Treatment of CDN Q with the counter strand T₁′ regenerates CDN P. Similarly, subjecting CDN P to the trigger T₂ results in the stabilization of BA′−T₂, resulting in the dynamic reconfiguration of CDN P into CDN R, where BA′−T₂ is upregulated, AA′ and BB′ are downregulated, and BA′ acting as agonist constituent to AB′−T₂, is also upregulated. Subjecting CDN R to the counter fuel T₂′ regenerates CDN P. The reversible triggered fluorescence changes transduced by the DNAzyme reporter units associated with constituents upon the T₁/T₁′-triggered dynamic transitions of P→Q→P and the respective fluorescence changes of constituents are summarized in Figure 16B. Similarly, the fluorescence changes of DNAzyme reporter units associated with the constituents upon the dynamic T₂/T₂′-triggered transitions of P→R→P are presented in Figure 16C. The results demonstrate the dynamic control over the concentrations of the dimer-tetrahedra constituent by means of auxiliary triggers as a model system for guided dynamic control of protein-protein interactions.

Moreover, diverse intracellular transformations, such as gene transcription⁸⁵ and protein translation,⁸⁶ cell motility,⁸⁷ and proliferation⁸⁸ reveal spatial dissipative mechanisms demonstrating transient^76c or oscillatory^76b processes in clustered networks, leading to temporal structures and catalytic reactions. Substantial recent efforts are directed to develop artificial networks and circuits emulating such natural processes.^{70a, 70b, 89}

The size analogy between DNA tetrahedra and proteins was then employed to develop transient tetrahedra dimerization circuits as models for dynamic, transient, dimerization of proteins in nature⁹⁰ (Figure 16D). A mute reaction module composed of two tetrahedra T₁ and T₂, modified with Cy3 and Cy5 fluorophore-functionalized tethers, a duplex L₁/I₁ and a nicking enzyme (Nt. BbvCI) was employed to mimic dissipative protein dimerization by tetrahedra structures. Subjecting the reaction module to the triggering strand L₁′ displaced the duplex L₁/I₁ yielding the duplex L₁/L₁′ and the released I₁ interlinked tetrahedra T₁/T₂ into a dimer structure. The duplex L₁/L₁′ was engineered to be cleaved by the nicking enzyme, leading to the release of L₁′ that displaced I₁ from tetrahedra dimer, resulting in the separation of dimer and the recovery of inactive reaction module. The transient formation and depletion of tetrahedra structure was followed by the FRET signal changes associated with the Cy3/Cy5 fluorophores linked to the temporal tetrahedra dimer structure (Figure 16E). The transient operation of the circuit was controlled by the concentrations of the activating strand L₁′ (Panel I) and the nicking enzyme (Panel II).

6.1 Nanoparticle-Modified DNA Tetrahedra Structures

Using single-strand functionalized Au NPs (5 nm), the assembly of a DNA-bridged pyramidal tetrahedra structure carrying at each corner the Au NPs was demonstrated.⁹⁶ By applying DNA conjugated different-sized Au NPs (5, 10, 15, 20 nm) the assembly of asymmetric Au NPs-modified DNA tetrahedra was accomplished. The asymmetric structures were visualized by TEM (Figure 17A and 17B). Although, this structure reveals chiroplasmonic properties, the size homogeneity of the particle limited the intense circular dichroism (CD) response of the structures. Nonetheless, by engineering the length of the DNA edges bridging two-sized AuNPs, a symmetric, non-chiral structure of a pair of 10 nm sized AuNPs and a pair of 20 nm sized AuNPs⁹⁷ (Figure 17C) structure I, was synthesized. By the appropriate shortening of two mirror-image edges of tetrahedra structures, two asymmetric, chiral mirror-image structures of the four AuNPs were constructed, structures II_a and II_b. As expected, the structure I lacked any circular CD response, yet the asymmetric structures II_a and II_b revealed opposite CD spectra where the CD intensities were controlled by the distances shortening the respective edges, consistent with the distortion degree of the respective tetrahedra (Figure 17C). In a related study, the strand dictated formation of DNA tetrahedra modified at four edges with different types of particles leads to four edges-dictated mirror-images of the asymmetric chiral tetrahedra structures⁹⁸ (Figure 17D). The chiral structures of the mirror-image NPs-modified tetrahedra was evidenced by the CD bands corresponding to the plasmon absorbance of the AuNPs at λ=530 nm and the AgNPs at λ=410 nm (Figure 17E).

6.2 DNA Tetrahedra as Nanocages

The size-programmability of the DNA tetrahedra nanostructure⁴⁹ and the ability to functionalize intra-edge domains with nucleic acid tether⁵² provides versatile means to load the tetrahedra with payloads, e.g., nanoparticles⁹⁹ or enzymes,¹⁰⁰ and eventually introduce into the DNA tetrahedra instructive information to guide sequential chemical transformations within tetrahedra frameworks.¹⁰¹

Figure 18A introduces a versatile strategy to encapsulate nanoparticles, e.g., AuNPs into a self-assembled tetrahedra structure.⁹⁹ A tripod subunit, X, consisting of three arms, where each arm consisting of two duplexes was engineered as functional subunit to assemble the DNA tetrahedra structures. Each of the duplexes associated with the arms was further crosslinked with a nucleic acid strand P that includes domains k and t, and each of duplexes was further functionalized with a tether f. The tether k exhibits self-complementarity features, while the tether k is complementary the nucleic acids k′ coupling the auxiliary AuNPs. The tripod subunits self-assemble into tetrahedra structure, resulting in the intra tetrahedra functionalization of the edges with the tethers k. The permeation of the k′-modified AuNPs through the faces of the nanostructures led to the intra-tetrahedra loading of the structure with AuNPs that hybridized with the tethers k. By applying a counter strand k′′, the AuNPs payload was displaced, resulting in a process that regenerated the vacant tetrahedra structure. The AFM images and cryo-TEM images supported the expected AuNPs-tetrahedron encapsulated structures (Figure 18B, panels I and II). By the further addition of subunits (W) consisting of cross-bridged duplexes extended by tethers g′ and h′ and modified with single strand tethers l, to the tripod subunits (Y) modified with tethers g and h, the inter-bridging of the tripod units into an extended DNA tetrahedra structure was demonstrated¹⁰² (Figure 18C). Subjecting the extended DNA tetrahedra structure to AuNP_in capped with two functional strands l′ and m′, and then to AuNP_out capped with two functional strands m, the gold nanoparticles interacted with the extended tetrahedra assembled into a supramolecular structure consisting of five AuNPs. One AuNP_in was positioned in the center of tetrahedra through the hybridization of strand l/l′ and the other four AuNPs_out were positioned on the other tethers (m′) of AuNP_in through hybridization of the duplexes m/m′. The assembly of a 3D structure of the AuNPs was realized and confirmed by TEM reconstructed cryo-TEM imaging (Figure 18D, panels I and II). Furthermore, the tripod subunits crosslinking concept to assemble DNA tetrahedra cages was further developed to assemble multi-layered tetrahedra nanocages¹⁰³ (Figure 18E). Towards this goal, three types of tripods X, Y, and Z were designed and two types of extending cross-bridging duplexes W1 and W2 were used for the sequential self-assembly of the multi-layered tetrahedra caged assembly. In the first step, the tripod subunits X were allowed to assemble into the small-sized inner tetrahedra structure. Subjecting the small-sized tetrahedra structure to tripod Y and the extending duplex units W1 resulted in the X-tetrahedra-guided coating of the second layer tetrahedra by the crosslinking of the tripod Y, in the presence of the extending duplexes W1. The medium sized tetrahedra was crosslinked to the inner small tetrahedra by edged strand functionality associated with the two-sized tetrahedra. Similarly, treatment of the two-layered caged tetrahedra structure with the tripod Z and the extending duplex bridging units W2 resulted in the two-layer tetrahedra-guided assembly of the third layer tetrahedra coating where tripod Z were linked by the extender units W2 and concomitantly bridged by duplex formed between tether functionalities associated with the edge of medium sized tetrahedra and the extender units W2. The three-layer caged structure of tetrahedra was confirmed by reconstructed Cryo-TEM images (Figure 18F) and supported by electrophoretic and AFM measurements.

DNA tetrahedra were further applied as a structural subunit to self-assemble crystalline Au nanoparticle composites and particularly the challenging diamond superlattice structure¹⁰⁴ (Figure 19). Using the M13 phage DNA, as scaffold, and a set of DNA staple strands, a 3D DNA origami bundle-based tetrahedra structure functionalized at its corners with protruding tethers, l, and its edges with protruding tethers, m, was self-assembled (Figure 19A). The reconstructed cryo-TEM image of the assembled tetrahedron subunits, Panel I, reveals an edge length of ca. 36 nm and an internal void volume revealing a cross-section of ca. 9 by 6 nm. Treatment of tetrahedra with nucleic acid m′-functionalized AuNPs resulted in the modification of the internal void containment of tetrahedra with the AuNPs. The cryo-TEM image demonstrated the AuNPs-loaded tetrahedron structures, panel II. Subjecting the unloaded tetrahedra to the l′-tethered AuNPs led to the self-assembly of an ordered face-centered cubic (FCC) crystalline lattice of the aggregated AuNPs. In turn, subjecting the mixture of the AuNPs-loaded tetrahedra to the m′-modified AuNPs resulted in the self-assembly of tetrahedra into a diamond superlattice crystalline composite of the AuNPs (Figure 19A). Tetrahedra-guided assembly of the FCC and diamond crystalline structures was confirmed by small-angle X-ray scattering (SAXS) experiments (Figure 19B). The 2D scattering pattern and the associated structure factor S(q) revealed sharp scattering peaks. The ratio of the positions of the peaks matched $$ $$ …, consistent with an FCC lattice. Similarly, the SAXS peak and associated structure factor S(q) for the diamond superlattice AuNPs aggregate revealed sharp scattering peaks position at ratios corresponding to $$ $$ …, in agreement with the cubic diamond lattice.

7.1 DNA-tetrahedra as Cell Targeting and Nano-framework Delivery Vehicles

DNA tetrahedra provide an effective 3D structure to interact with cell membrane¹⁰⁹ and to assist cell sorting and separation.¹¹⁰ This has been exemplified with the functionalization of the corners of a DNA tetrahedra with one, two or three anti-EpCAM receptor aptamer units¹⁰⁵ (Figure 20A). The binding affinity of the different tetrahedra scaffolds to the clustered EpCAM receptor sites associated with circulating tumor cells (CTC) was controlled by the number of aptamer units tethered to tetrahedra backbone (Figure 20B). As the number of aptamer units increased, the synergistic binding of tetrahedra to the clustered receptor was enhanced. The control over the binding affinities of tetrahedra to the receptor sites was, then, applied to select and sort the cancer cells, from a mixture of cells. The different EpCAM aptamer-modified tetrahedra were anchored to magnetic beads and these were used to sort fluorescent (Cy5) stained MCF-7 CTC (Figure 20C). As the number of aptamer units linked to tetrahedra increased the capturing efficiency of the MCF-7 cancer cells was enhanced (Figure 20D). The control over the binding affinities of aptamer-functionalized tetrahedra with cell membranes was further used to separate different cells within a microfluidic device¹¹⁰ (Figure 20E). The glass channels of the device underwent functionalization with two distinct DNA tetrahedra: T₁, modified with an aptamer KDED-3a targeting DLD-1 cells, and T₂, carrying two aptamers KCHA-10 targeting HCT-116 cells. Despite the similar binding affinities of aptamer KDED-3a and aptamer KCHA-10 towards their respective cells, the microfluidic channel successfully sorted these cell types. Within the channels, tetrahedral scaffolds captured DLD-1 cells labeled with a blue fluorophore and HCT-116 cells labeled with a red fluorophore. The binding affinity of the respective cells to the aptamer-modified tetrahedra was influenced by the number of specific aptamers present. For instance, T₁ with one aptamer KDED-3a exhibited weak binding to DLD-1 cells, while T₂ with two aptamers KCHA-10 displayed strong binding to HCT-116 cells. Consequently, the separation of the cell mixture was determined by shear forces generated by varying flow rates through the channel. High-flow rates facilitated the release of both DLD-1 and HCT116 cells, releasing a mixture of the two. Conversely, at low flow rates, only weak DLD-1/aptamer complexes were released, with subsequent high flow rates leading to the separation of HCT-116 cells. This mechanism enabled the effective separation of the two cell types, as illustrated in Figure 20F.

Beyond the adhesion of DNA tetrahedra to cell interfaces by means of aptamer-cell membrane interactions,¹¹¹ DNA tetrahedra exhibit superior cell permeation properties thus guiding superior permeation of tetrahedra-modified payload-functional carriers.^{47, 112} Indeed, improved cell permeation of DNA tetrahedra carriers as compared to DNA-duplex functional carriers or Lipofectin transfection means were demonstrated.¹¹³ These features were employed to assemble miRNA responsive, doxorubicin (DOX)-loaded N₃-UiO-66 DNA-tetrahedra functionalized metal–organic framework nanoparticles, NMOFs¹¹⁴ (Figure 21A). Tetrahedra-modified DOX-loaded NMOFs revealed a ca. four-fold enhanced permeation into MDA-MB-231 cancer cell, as compared to duplex-modified NMOFs, and a ten-fold enhanced permeation as compared to epithelial MCF-10 A breast cells. The enhanced permeation features of the DOX-loaded miRNA-responsive NMOFs were reflected by selective cytotoxicity towards MDA-MB-231 breast cancer cells (miRNA-21 responsive) and HepG2 cancer cells (miRNA-155 responsive). Moreover, Zn(II)-protoporphyrin IX-loaded DNA tetrahedra-functionalized UiO-66 NMOFs, where tetrahedra units are conjugated to the NMOFs by a duplex (17)/(17′) comprising the Vascular endothelial growth factor (VEGF) aptamer in strand (17′), were prepared¹¹⁵ (Figure 21B). In the presence of VEGF, overexpressed in cancer cells, the formation of the VEGF/aptamer complex uncaged the NMOFs, resulting in the release of the Zn(II)-protoporphyrin IX. Its incorporation in the VEGF/aptamer G-quadruplex units resulted in an effective photosensitizer for the light-induced formation of reactive oxygen species (ROS) (Figure 21B). By integrating the Zn(II)-protoporphyrin IX-loaded DNA tetrahedra into MDA-MB-231 cancer cells, effective photodynamic treatment by the ROS agents was demonstrated (Figure 21C). Selective VEGF-triggered photodynamic treatment of the MDA-MB-231 cancer cell, as compared to epithelial MCF-10 A breast cells was realized.

The DNA-tetrahedra-stimulated enhanced cell permeation efficacy of drug carrier has been further applied to develop stimuli-responsive lentivirus-loaded DNA-based hydrogel microcapsules for immunogenetic therapy of liver cancer cells¹¹⁶ (Figure 22). DNA-based hydrogel microcapsules loaded with vector-modified lentivirus (green fluorescence protein, GFP or COVID-19 spike protein, SP, vectors) were prepared as outlined in Figure 22A. The microcapsules boundaries were functionalized with the anti-α-fetoprotein aptamer (α-fetoprotein, AFP, is overexpressed biomarker in liver cancer cells), and the DNA tetrahedra, as cell permeating agents. The immunogenetic therapeutic principles of the lentivirus-loaded microcapsules in schematically depicted in Figure 22B. The DNA tetrahedra-facilitated permeation of the microcapsules into liver cancer cells, as compared to normal cells, and the AFP-guided unlocking of the microcapsules in the liver cancer cells resulted in the selective release of the lentivirus in the cancer cells and the superior GFP (or SP) expression in the liver cancer cells. A selective factor of 50,000 for the GFP expression in the cancer cells over healthy liver cells, was observed (Figure 22C). These results originated from the cooperative DNA-tetrahedra-stimulated cancer cell permeation and AFP-induced uncaging of the microcapsules in the liver cancer cells. The effective and selective protein expression in the liver cancer cells was, then, utilized for the efficient immunogenetic therapeutic treatment of liver cancer cells, through the activation of the COVID-19 immune system (Figure 22D). The selectively expressed SP, by the SP-vector-loaded lentivirus permeated into the liver cancer cells, and accumulated in the cancer cell membranes. This led to the activation of the COVID-19 immune system of vaccinated or COVID-19 infected individuals, resulting in the targeted killing of the liver cancer cells. Indeed, the treatment of the liver cancer cells with human Peripheral Blood Mononuclear Cells, PBMCs (White Blood Cells, WBCs) of vaccinated or COVID-19-recovered individuals resulted in ca. 75 %–80 % cancer cell death (within exposing the cells to two doses of the WBCs), whereas no cytotoxic effect on the liver cancer cells was observed upon treatment with WBCs originating from non-vaccinated individuals (Figure 22E).

7.2 DNA-tetrahedra as Therapeutic Carriers

The DNA tetrahedra do not only facilitate the permeation of the nanostructure into cells but protect oligonucleotides against degradation.^{47, 112} There two features were employed to use DNA tetrahedra as nano-vehicles delivering therapeutic constituents, such as siRNAs into cells (Figure 23A). The edges of DNA tetrahedra were engineered to include duplexes comprising of a target siRNA (e.g., luciferase gene siRNAs) that are further modified with folate ligands targeting KB HeLa cells.¹¹⁷ Treatment of mice bearing xenograft KB tumors with the siRNA-functionalized tetrahedra resulted in over 50 % inhibition of luciferase activity in the tumors as compared to non-treated tumors or tumors treated with the siRNA duplex, in the absence of tetrahedra carrier (Figure 23B) demonstrating the effective delivery and protection of the siRNA by tetrahedra. Similarly, inhibition of the transcription/translation of the mRNAs, e.g., GFP¹¹⁸ or fisZ bacterial gene¹¹⁹ were demonstrated.

The superior cell permeation and stability features of DNA tetrahedra, as compared to other DNA scaffolds introduced useful DNA nanostructures for drug delivery. To reach this goal, the development of means to integrate the drugs as stable non-leaked configurations and the design effective release mechanisms, is essential. In addition, targeting of the drug-loaded tetrahedra to the specific cell environment is important.

One approach to load tetrahedra carrier utilized the double-strand edges of tetrahedra as scaffolds to intercalate the drug.¹²⁰ This is exemplified in Figure 24A with the assembly of DNA tetrahedra modified at their edges with the nucleic acid tethers that provided anchoring sites for EGFR-receptor nanobodies (fragments of the respective antibody) modified with the tether. The cis-Pt(II)-bisamine/bipyridine anti-cancer drug (56MESS) was intercalated into the duplex edges of tetrahedra duplexes acting as carriers. The supramolecular nanostructures consisting of the nanobody-conjugated, drug-loaded tetrahedra acted as functional targeted carriers that bind to the EGFR receptors associated with A431 skin cancer cells, resulting in tetrahedra facilitated endocytosis. The permeation of the carriers into the lysosome resulted in the enzymatic cleavage of the DNA tetrahedra and the release of the therapeutic drug load. Validation of the efficacy of the drug-loaded nanobody-modified tetrahedra as chemotherapeutic drug carrier toward cytotoxicity of a A431 cancer cell line, as compared to the cytotoxicity of a reference carrier composed of an analog drug-loaded duplex DNA functionalized with the nanobody, was demonstrated (Figure 24B). The cytotoxicity of the drug-loaded tetrahedra was two-fold higher as compared to the reference duplex carrier (90 % cell death as compared to 50 %), due to the enhanced permeation of tetrahedra drug carrier. The enhanced cytotoxicity of the drug-loaded tetrahedra carrier, as compared to the reference duplex-functionalized carrier, was further reflected in in vivo experiments (Figure 24C). In these experiments, mice bearing A431 tumors were subjected to treatment with 56MESS-loaded nanobody-functionalized tetrahedra, and their tumor growth was compared with reference groups. These references included mice treated with unloaded nanobody tetrahedra or 56MESS-loaded nanobody-functionalized duplex carriers. After 20 days, tumors in mice treated with drug-unloaded tetrahedra reached a volume of 1500 mm³, while those treated with drug-loaded duplex carriers reached 1000 mm³. Remarkably, mice treated with 56MESS-loaded nanobody-functionalized tetrahedra exhibited virtually no tumor growth (<20 mm³). This significant inhibition of tumor growth was ascribed to the effective permeation and localized cytotoxicity of the drug-loaded tetrahedra carrier. Related studies¹²¹ reported the intercalation of doxorubicin into the duplex domains of DNA tetrahedra carriers, and the use of tetrahedra as a vehicle to enhance the cytotoxic efficacies. Moreover, the DNA tetrahedra scaffold allowed, in addition to the intercalation of the doxorubicin drug, the further modification of tetrahedra corners with the anti-EGFR antibody and the functionalization of one of the corners with a fluorophore (Cy3), as a means to target and visualize the transport and localization of the carrier in cancer cells.

Moreover, an alternative approach has involved the covalent tethering of the anti-cancer drug to the DNA tetrahedra carriers and release of the drug by overexpressed chemical agents present in cancer cells. For example,¹²² the anti-cancer drug camptothecin was covalently linked to a nucleic acid strand being a part of the DNA tetrahedra carrier using a disulfide bridging bond. The drug-modified DNA tetrahedra was, then, released from the carrier by glutathione (GSH), overexpressed in cancer cells, through cleavage of the S−S bridging bonds. Effective inhibition of tumors growth by the drug-loaded carriers, as compared to treatment with the bare-drug composites, was demonstrated, an effect attributed to the enhanced drug permeation into the cells, assisted by the DNA carrier.

7.3 DNA Tetrahedra for Regenerative Medicine

Recent research activities demonstrated the use of DNA tetrahedra frameworks for regenerative medicine. Particularly, DNA tetrahedra were found to enhance cell proliferation¹²³ and differentiation¹²⁴ and revealed anti-inflammatory,¹²⁵ antioxidant¹²⁶ and immune regulation properties.¹²⁷ These features of the DNA tetrahedra scaffolds were employed to promote the regeneration of bones¹²⁸ and blood vessels,¹²⁹ to induce wound repair¹³⁰ and tissue regeneration,¹³¹ and to affect neurological disorders,¹³² joint-related inflammatory diseases,¹³³ and immune diseases.¹²⁷ These broad prospects of DNA tetrahedra nanostructures were addressed in detail in several review articles¹³⁴ and will be highlighted here with several examples.

Figure 25 introduces the use of the miRNA-2861-loaded DNA tetrahedra as a carrier for controlling bone repair by bone marrow mesenchymal stem cells (BMSCs).¹²⁸ The DNA tetrahedra carrier were functionalized, at the corners, with the nucleic acid tethers h. The miRNA-2861 that inhibits the expression of the histone deacetylase 5 (HDAC5) protein (that results in the upregulated expression of the runt-related transcription factor 2, Runx2, that ultimately promotes osteogenic differentiation and bone repair) was extended by the tether h′ and was hybridized with tetrahedra framework, yielding a functional carrier for transporting the miRNA into the stem cells (Figure 25A). The mechanism of the bone repair by the BMSCs, interacted with the bone injury, presented schematically in Figure 25B. The cellular RNase H degrades the ribonucleic base domain h′ of the h/h′ resulting in the release of the miRNA-2861 that inhibits the HDAC5 protein, resulting in the promotion of bone injury healing. Figure 25C, panel I, depicts in vitro immunofluorescence images of the HDAC5 protein developed in stem cells treated with the miRNA-2861 functionalized DNA tetrahedra, and respective control systems, and Panel II introduces the statistical analysis of the fluorescence intensities in the respective cell assays. Evidently, the development of the HDAC5 protein is inhibited by the miRNA-functionalized carrier. Similarly, Figure 25D depicts the western blot bands of the HDAC5 proteins formed in the stem cells treated with the miRNA-2861 modified tetrahedra, Panel I, and the quantitative analysis of the bands, Panel II. The results demonstrate significantly inhibited yields of the HDAC5 protein in the cells treated with the miRNA-2861-functionalized carrier, panel II. Figure 25E depicts the effect of the stem cells on the injured bones upon treatment with the miRNA-2861-functionalized tetrahedra frameworks and control systems. After a time-interval of 14 days, the bone-injury treated with BMSCs, interacted with the miRNA-2861 functionalized tetrahedra, is fully healed whereas the control systems did not show a significant bone repair. Furthermore, cells subjected to DNA tetrahedral structure demonstrated enhanced proliferation and migration and retardation of inflammatory processes.¹³⁵ These features were used to apply DNA tetrahedral framework to promote scarless healing of cutaneous wounds by intervening the AKT-signalling pathway.¹³⁶ Subjecting keratinocyte (HeCaT cells) or fibroblast (HSF cells) to the DNA tetrahedra frameworks demonstrated enhanced proliferation and migration capacities of the cells (reflected by faster bridging of a bare cell domain, as compared to a control system).

This section has outlined various nanomedical applications of DNA tetrahedra. These applications encompass the utilization of aptamer-modified tetrahedra for targeting and imaging cells, as well as for the controlled separation of cells. Additionally, DNA tetrahedra have been employed to facilitate the permeation of drug-loaded carriers into cells. Specifically, the section introduced the use of DNA tetrahedra to mediate the permeation of drug-loaded nanoparticles that release ROS-generating agents and vector-modified viral agents for the activation of immune-gene therapies targeting cancer cells. Furthermore, it discussed the use of drug-loaded DNA tetrahedra carriers, which are functionalized with specific cell-binding antibodies, for selective chemodynamic and photodynamic treatment. Additionally, the application of DNA tetrahedra carrying bioagents in regenerative medicine presents promising pathways for medical treatment. While the medical functions of DNA tetrahedra as carriers are limited to their small sizes and loading capacities, their major nanomedical advantages are reflected by their integration with carriers exhibiting higher loading capacities, such as porous particles or microcapsules.