ARTICLES PUBLISHED ONLINE: 16 MARCH 2015 | DOI: 10.1038/NNANO.2015.27
Light-emitting self-assembled peptide nucleic acids exhibit both stacking interactions and Watson–Crick base pairing Or Berger1, Lihi Adler-Abramovich1, Michal Levy-Sakin2, Assaf Grunwald2, Yael Liebes-Peer2, Mor Bachar1, Ludmila Buzhansky1, Estelle Mossou3,4, V. Trevor Forsyth3,4, Tal Schwartz2, Yuval Ebenstein2, Felix Frolow1,5†, Linda J. W. Shimon6, Fernando Patolsky2,7 and Ehud Gazit1,7* The two main branches of bionanotechnology involve the self-assembly of either peptides or DNA. Peptide scaffolds offer chemical versatility, architectural flexibility and structural complexity, but they lack the precise base pairing and molecular recognition available with nucleic acid assemblies. Here, inspired by the ability of aromatic dipeptides to form ordered nanostructures with unique physical properties, we explore the assembly of peptide nucleic acids (PNAs), which are short DNA mimics that have an amide backbone. All 16 combinations of the very short di-PNA building blocks were synthesized and assayed for their ability to self-associate. Only three guanine-containing di-PNAs—CG, GC and GG—could form ordered assemblies, as observed by electron microscopy, and these di-PNAs efficiently assembled into discrete architectures within a few minutes. The X-ray crystal structure of the GC di-PNA showed the occurrence of both stacking interactions and Watson–Crick base pairing. The assemblies were also found to exhibit optical properties including voltage-dependent electroluminescence and wide-range excitation-dependent fluorescence in the visible region.
N
ature has produced a basic set of building-block molecules through billions of years of molecular selection and evolution. The 20 coded amino acids and four primary nucleotides are the most fundamental elements in living systems. Protein and nucleic acid biomolecules allow the formation of an enormously diverse range of supramolecular assemblies, exhibiting a vast array of physical properties. Inspired by natural ordered assemblies, many studies have been directed towards the design of peptide and protein building blocks that self-assemble into preferred architectures1–7. The shortest, most simple peptide building block shown to self-assemble into ordered architectures is diphenylalanine, the core recognition motif of β-amyloid polypeptide8. This aromatic dipeptide self-assembles into discrete well-ordered nanotubular structures of notable persistence length in aqueous solution9. Various studies have shown the remarkable physical characteristics of these nanotubes, including metal-like rigidity and high thermal and chemical stability, as well as optical, semiconductive and piezoelectric properties10–12. In contrast to peptide self-assembly, structural DNA nanotechnology is solely derived from the specificity of the hydrogenbonding interactions between complementary Watson–Crick base pairs. The use of DNA as a structural building block instead of merely a genetic material was first recognized in the early 1980s, in a theoretical work by Seeman13. Based on the complementary nature of nucleic acids, it is possible to predict and design DNA structures of nanoscale order. This pioneering conceptual work materialized into a vivid field of research with tremendous growth. Now, with the aid of computer design and the DNA
origami method, the fabrication of any two- or even three-dimensional nanostructure shape has become considerably simpler14–16. Peptide and DNA building blocks offer two distinct approaches for the generation of supramolecular architectures by self-assembly. Peptide-driven materials are characterized by robustness, synthetic versatility, architectural flexibility and structural complexity, whereas DNA nanostructures are based on specific molecular recognition and base-pairing complementarity. Convergence of the peptide and nucleic acid assembly strategies could be very useful for the design of novel self-organized materials. Peptide nucleic acids are artificially synthesized polymers that were first described by Nielsen and co-workers in 1991 (refs 17, 18). The polymer is an oligonucleotide analogue in which the phosphodiester backbone is replaced by repeating N-(2-aminoethyl) glycine units linked by amide bonds. Thus, PNA can be regarded as either a DNA mimic with a neutral amide backbone, or as a peptide mimic with nucleobases as side chains. Another advantage of PNAs is their resistance to degradation by proteases or nucleases. A slightly different approach for the production of a peptide–nucleic acid hybrid is based on the conjugation of functionalized purine and pyrimidine bases to a peptide backbone19. The neutral peptide-like backbone replacing the negatively charged phosphodiester groups, which may limit self-association due to the repulsion of similar charges, adds structural elasticity and chemical adaptability. Another group of unnatural peptide–nucleobases hybrids (denoted nucleopeptides) have already been shown to self-assemble into nanofibres to generate supramolecular hydrogels20. Despite the great potential to converge the two distinct fields of peptide self-assembly and structural DNA
1
Department of Molecular Microbiology and Biotechnology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv 69978, Israel. 2 School of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Ramat Aviv 69978, Israel. 3 Partnership for Structural Biology, Institut Laue Langevin, 71 Avenue des Martyrs, Grenoble Cedex 9 38042, France. 4 Faculty of Natural Sciences, Keele University, Staffordshire ST5 5BG, UK. 5 Daniella Rich Institute for Structural Biology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv 69978, Israel. 6 Department of Chemical Research Support, Weizmann Institute of Science, Rehovot 76100, Israel. 7 Department of Materials Science and Engineering, Iby and Aladar Fleischman Faculty of Engineering, Tel Aviv University, Ramat Aviv 69978, Israel. †Deceased. *e-mail: [email protected] NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
1
ARTICLES
NATURE NANOTECHNOLOGY
a
b HN
N O
H2N N O N
H2N
T
A
AA
AC
AG
AT
C
CA
CC
CG
CT
G
GA
GC
GG
GT
T
TA
TC
TG
TT
H2N O
N
N
N N
O
H2N
H2N N O
O
N
N H
N
12.0
N
N
HN H2N O
NH2
N
H2N
N
N
N
O
f
11.9
O
HN
N
O NH2
O
NH2
O
e
h pH = 11.8
d
O
N
O N H
G
HN
N
N
C
c
O
NH2
A
DOI: 10.1038/NNANO.2015.27
N O
O N
N H
O NH2
g
12.1
12.2
12.3
12.4
Figure 1 | Self-assembly of guanine-containing di-PNAs into well-ordered architectures. a, The 16 different di-PNA building blocks were synthesized and assayed for their ability to undergo self-assembly into distinct structural entities. Molecular assembly was evident with three di-PNAs under alkaline conditions (highlighted in black). More assemblies were observed for three other di-PNAs on drying the sample (highlighted in grey). b–d, Chemical structures of the three assembly-forming di-PNAs: CG (b), GC (c) and GG (d). e–g, SEM micrographs of the structures formed by CG (e), GC (f) and GG (g). Scale bars, 10 µm. h, Light microscopy images of assemblies formed by GC di-PNA on dissolving with rising pH levels of disodium hydrogen phosphate buffer. Original magnification, ×40.
nanotechnology, no work has been performed on PNAs as building blocks to assemble into distinct structural entities by themselves. As described in a recent review21, PNA has so far only been used as a template or as a conjugate to other molecules that undergo the selfassembly process, in order to gain the hybridization properties of nucleic acids. A notable example is the work of Stupp and coworkers22, in which a PNA strand and a peptide sequence that promotes nanofibre formation were coupled to form a PNA–peptide amphiphile conjugate. The designed fibre-shaped nanostructures show binding of oligonucleotides with high affinity and specificity. Here, we report for the first time the efficient and rapid formation of supramolecular architectures based solely on PNA molecule self-assembly.
di-PNA synthesis, assembly and structure Inspired by the ability of simple aromatic dipeptides to form unique assemblies, we decided to examine the ability of short di-PNA 2
building blocks to form ordered supramolecular assemblies. All of the 16 different di-PNA combinations were synthesized using solid-phase peptide synthesis (AA, AC, AG, AT, CA, CC, CG, CT, GA, GC, GG, GT, TA, TC, TG, TT; italics are used to denote PNA) (Fig. 1a). The synthesis of the di-PNA was performed according to conventional peptide synthesis practice using standard methods and commercially available protected building blocks. The synthetic effort, yield and number of steps are comparable to those for the production of other aromatic dipeptides that are readily synthesized at an industrial scale, such as L-aspartyl-L-phenylalanine methyl ester. Favourable conditions for self-organization were determined by screening the di-PNAs in a variety of solvents including organic solvents (methanol, ethanol, dimethyl sulphoxide, hexafluoro-2propanol, and so on) and diverse buffer solutions with a range of pH values and concentrations. Under alkaline conditions, microscopic observation of molecular assembly was evident with only
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY
ARTICLES
DOI: 10.1038/NNANO.2015.27
a
b
c
d 3.57
2.87
3.48
3.47
3.48
3.57
2.93 3.47 2.85
e
f
Figure 2 | Crystal structure of GC di-PNA. The GC di-PNA building block was crystallized and its structure determined by synchrotron radiation at a resolution of 0.95 Å. a, Molecular structure of a single GC di-PNA molecule from single-crystal structure determination. The cytosine and guanine nucleobases form an intramolecular stacking interaction. b, Each molecule forms hydrogen bonds with a neighbouring unit between the cytosine and guanine residues. c, The hydrogen bond length between symmetry-related molecules is measured to be 2.85–2.93 Å, the same as in typical Watson–Crick base pairs. d, The bases are 3.5 Å apart, the same as in a DNA double-helix structure. e, The di-PNAs are packed in a continuous tilted stack through the crystal. f, When lining the stacks in the z direction it is evident that this form of packing results in rectangular-shaped pores comprising over 50% of the crystal volume.
three di-PNAs: CG, GC and GG (for structures see Fig. 1b–d). Examination of the solutions using light and electron microscopy revealed well-organized architectures, including long rods (tens of micrometres long) for CG and GC (Fig. 1e,f ) and spheroids with a diameter of 2–3 μm for GG (Fig. 1g). Other di-PNAs did not form any type of ordered structure, or did so only upon drying of the solution (AG, GA and GT; Supplementary Fig. 1a–f ). Co-assembly of all the possible complementary di-PNA combinations at a 50:50 molar ratio was also examined. Well-defined structures were observed solely for the combination of CC and GG. When mixed together, instead of the spherical assemblies formed by GG alone, elongated structures similar to those generated by CG or GC were evident (Supplementary Fig. 1g). Based on these findings, we speculate that it is necessary to have a minimum of six hydrogen bonds between the bases for the stabilization of complementary di-PNAs and for further organization into supramolecular entities. Interestingly, all PNAs that were able to form structures contained
the guanine nucleobase. As it is known that the secondary structure of G-containing PNA oligomers may be altered under alkaline conditions due to deprotonation of the guanine bases23,24, the assembly of the GC di-PNA was further examined under increasing pH levels (Fig. 1h). Furthermore, alkaline conditions have been found to be essential for self-assembly, presumably due to the introduction of charges, so high salt conditions (which may shield these charges) were examined. Accordingly, GC di-PNA was dissolved in assembly buffer with 1 M sodium chloride. Under this condition, only small nucleation sites were observed. Typical structures initiated the assembly process following a tenfold dilution of the salt (Supplementary Fig. 1h). In this context it should also be noted that guanine is a key component in the assembly of various natural nucleic-acid structures. Nucleic-acid sequences that are rich in guanine are capable of forming G-quadruplexes, the main structural motif of the telomeric DNA. Moreover, guanosine analogues are able to self-associate into
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
3
ARTICLES
NATURE NANOTECHNOLOGY
a
00:00:00
00:00:30
00:01:00
00:01:30
00:02:00
b 45
Length (µm)
40
y = 0.2439x + 21.587 R2 = 0.9722
35 30 25 20 15 0
20
40
60
80
Time (s)
Figure 3 | The di-PNAs efficiently assemble into discrete architectures within a few minutes. a, Five snapshots taken from Supplementary Movie 1 demonstrate the assembly kinetics of the GC di-PNA architectures. A thin glass capillary was filled with a freshly dissolved GC di-PNA solution and sealed immediately at both ends with wax. The snapshots represented here were captured every 30 s. Each panel shows the full width of the capillary (200 µm). b, The average measured elongation rate is 0.25 µm s−1. Graphical representation of the elongation rate of a single structure.
dimers, ribbons and macrocycles that can further stack into supramolecular assemblies25. As a control we studied the ability of guanine-containing DNA dinucleotides to form ordered assemblies. No conditions under which ordered dinucleotides structures could be formed were found. To gain better insight into the assembly process and stabilizing interactions, we attempted to form ordered single crystals of the assemblies and acquire X-ray diffraction data. To this end, the di-PNAs were screened for crystallization conditions. Crystals of GC di-PNA were found to grow in a bicine-based crystallization buffer. Because bicine buffer also enables assembly of the structures, we strongly believe the crystal structure reflects the solution selfassembled architecture. The crystal structure of GC was determined at a resolution of 0.95 Å with data collected at the European Synchrotron Radiation Facility (ESRF). The determined structure revealed very unique packing of the PNA crystals. The cytosine and guanine in each molecule form stacking interactions (Fig. 2a). Each molecule then forms hydrogen bonds with a neighbouring unit between the cytosine and guanine residues (Fig. 2b). All of the bases are engaged in Watson–Crick hydrogen bonds, as observed in canonical DNA–DNA and PNA–DNA duplexes26. The hydrogen bond length between 4
DOI: 10.1038/NNANO.2015.27
symmetry-related molecules was measured to be 2.85–2.93 Å (Fig. 2c), which is similar to that of Watson–Crick base pairing. The bases are found to be ∼3.5 Å apart (Fig. 2d), which is characteristic of DNA double-helical structures27, but they do not exhibit any tilt or roll. The hydrogen-bonding molecules are related to each other via the two-fold dyad that passes between the stacked bases. Packing of the molecule in a centrosymmetric space group is possible due to the non-chiral nature of the polyglycine backbone. The bases are packed in a continuous tilted stack throughout the crystal (Fig. 2e). This form of packing results in rectangularshaped pores that comprise over 50% of the crystal volume (Fig. 2f ). The crystal structure reflects the dual identity of PNA. The di-PNAs form stacking interactions with each other, in the same way as aromatic peptides, while at the same time forming the Watson–Crick base pairs typical of DNA structures. This unique duality distinguishes these molecules from simple DNA dinucleotides that possess only Watson–Crick base pairing and do not form any self-assembled structures. It is also possible that the negatively charged phosphate backbone of simple DNA oligomers may limit the assembly properties. The obtained crystal structure enabled us to estimate the rate at which the di-PNA monomers self-assemble in solution to form the ordered structures. In Supplementary Movie 1 we capture the assembly process of the GC di-PNA in real time. Briefly, a thin glass capillary was filled with a fresh solution of the PNA building blocks and sealed immediately from both sides with wax to prevent evaporation and concentration changes of the solution. The capillary was monitored using light microscopy, and images were captured at a rate of one frame per second. Small nucleation seeds could be observed within a few seconds, and continual growth in one axis direction was sustained for a few minutes. Figure 3a depicts one capillary at five consecutive time points. To allow better visualization of the elongation of the structure as a function of time, one structure was tracked in a time frame in which the assembled structure is clearly seen and is not overlapped by other architectures (Supplementary Fig. 2). Distinct elongating structures from five different capillaries, each filled with freshly dissolved solution, were examined for their growth rate, and an average of 0.25 µm s−1 was measured. All the structures exhibited an excellent linear fit between the dimension of the assemblies in the long Z-axis and time (R 2 > 0.97). The elongation of a representative single structure as a function of time is given in Fig. 3b. For this specific structure, which is 1.77 µm wide, the elongation rate of 0.24 µm s−1 translates into a volume increase of ∼0.6 µm3 s−1. Because the crystal unit cell has a volume of 11,676 Å3 (Supplementary Table 1) and contains eight molecules, we can estimate that 4.11 × 108 di-PNA building-block molecules organize into the ordered structures each second. In comparison to other rapid elongating systems of a natural origin, such as the microtubule that elongates by an average rate of 0.66 µm per minute28, the assembly kinetics of the di-PNA is over 20 times faster. When the same experiment was carried out in an open environment, instead of a sealed capillary, the elongation rate was about ten times faster due to evaporation of the solution, leading to higher local concentrations of PNA. The rapid assembly of the di-PNA building blocks suggests they could be used in motor systems by converting the free energy emitted by the self-assembly process into mechanical motion29. Another very interesting property of this self-organization, important for technological applications, is its efficient, high-yield and uniform process, as only ordered homogeneous assemblies of the GC di-PNA could be observed under the examined conditions.
Optical characterization of PNA assemblies When trying to examine whether the PNA structures could bind DNA intercalators due to their Watson–Crick base pairing, we
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY
ARTICLES
DOI: 10.1038/NNANO.2015.27
a
Intensity (a.u.)
330 nm 340 nm 350 nm 360 nm 370 nm 380 nm 390 nm 400 nm 410 nm 420 nm 430 nm
Emission wavelength (nm)
c 500
b
y = 0.7073x + 187.51 R2 = 0.9943
480
460
440
420
400 350
400
450
500
550
320
600
360
400
440
Excitation wavelength (nm)
Wavelength (nm)
e
d
490 nm 470 nm
0.6 Δtmax (ns)
Normalized counts (a.u.)
510 nm
0.8 y = 0.0071x − 2.8495 R2 = 0.9697
0.4
450 nm 0.2 430 nm 410 nm 98
99
100
0.0 400
Time (ns)
+
+ −
−+
−+
+
−+ −+
−
+ − + − −+
+ −
−
−+
−
−+
−+
+
hν∞
−+
λR
−+
+
−+
−
−+ −+
hνI
−+
hν0
−+
−+ −+
λC
−+
−+
−+
−+
Solvent relaxed (R) state
−+
+
Intermediate (I) state
−+
−
520
−+
−+
−
+
480
−+
Franck−Condon (FC) state
f
440
Emission wavelength (nm)
−+
97
−+
96
−+ −+
95
Figure 4 | The PNA assemblies exhibit a red edge excitation shift with a broad range of emission wavelengths in the visible region. a, One bright-field and five fluorescence images of the same microscopic field of GC di-PNA structures. Each fluorescence image was taken with different excitation and emission filters: (excitation/emission) 387 nm/440 nm; 485 nm/525 nm; 537 nm/578 nm; 560 nm/607 nm; 650 nm/684 nm (from left to right). Pseudo-colours represent corresponding emission colour. Original magnification, ×100. b, Emission spectra of GC di-PNA assemblies at excitation wavelengths of 330, 340, 350, 360, 370, 380, 390, 400, 410, 420 and 430 nm. The emission peak shifts to the red with higher excitation wavelengths. c, Graphical representation of the relation between the excitation and emission wavelengths. The slope is measured to be ∼0.7, which suggests a dynamic Stokes shift. d, Time-resolved emission traces for different emission wavelengths following excitation at 390 nm. The emission onset is delayed with respect to the excitation peak as emission wavelength shifts to the red. e, Time delay of fluorescence as a function of wavelength. Δt is calculated to the fastest emission onset collected at 410 nm. The monotonic increase in Δt is a signature of REES. f, Continuous model of solvent relaxation. The I state refers to one of the intermediate states between the initial excited state and the final solvent relaxed state, in which the solvent molecules are partially relaxed. ν0 , νI and ν∞ represent the frequencies corresponding to the initially excited (Franck–Condon), intermediate and completely relaxed states, respectively, while λC and λR denote the wavelength maxima associated with these states. NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
5
NATURE NANOTECHNOLOGY
6
DOI: 10.1038/NNANO.2015.27
b
c
5
Isd (μA)
a
0
0V
4 Isd (μA)
made a surprising observation, as the control sample with no added dye had a fluorescence signal similar to that of the stained samples. Even more surprising was the fact that fluorescence emission was evident for a wide range of excitation wavelengths (Fig. 4a). Clear emission peaks were observed between 420 nm and 490 nm with the excitation wavelength ranging from 330 nm to 430 nm. Fair fluorescent images could be obtained even for higher wavelengths, but with lower intensity. As presented in Fig. 4a, the upper limit of the emission was observed at 684 nm when the excitation wavelength was 650 nm. To quantitatively analyse this phenomenon, fluorescence emission spectra were determined at different excitation wavelengths (Fig. 4b). It is clearly noticeable from the studies that the emission peak shifts in the direction of the change in excitation. To determine whether the shift is consistent for various excitation wavelengths, we plotted consecutive emission maxima as a function of excitation wavelengths (Fig. 4c). We observed a linear correlation between the excitation and emission peak wavelengths. Yet, the 0.7073 slope of the very strong fit (R 2 = 0.9943) indicates dynamic Stokes shift behaviour as the distance between the excitation and emission peaks gradually decreases with longer wavelengths. Such an observed change in fluorescence emission spectra in response to a shift in the excitation wavelength toward the red edge of the absorption band is termed a red edge excitation shift (REES). The phenomenon was originally described in rigid and highly viscous environments such as low-temperature glasses, graphene oxide layers or highly condensed polymeric states30. REES is assumed to be the result of the strong reduction in the dynamic environment of the excited fluorophores in organized molecular settings. A model for this spectroscopic behaviour assumes that the molecular lattice confinements slow the rates of matrix relaxation and reorientation around the excited state of the fluorophore relative to the fluorescence lifetime31. In biological and other organic molecules, such constraints could be imposed by exceptionally ordered hydration shells or rigid membranes. Time-dependent fluorescence measurements were used to determine whether this model could explain the origin of the excitationdependent emission behaviour observed with the PNA assemblies. The fluorescence decay was measured at several emission wavelengths following excitation at 390 nm (Fig. 4d). The results clearly indicate a process of time-dependent fluorescence onset, with longer wavelengths appearing with larger delays after the excitation pulse (Fig. 4e) (for example, the emission at 510 nm appearing 700 ps after the emission at 410 nm). The optical phenomenon observed here appears to be similar to the very strong REES effect shown in rigid graphene oxide31. The model used for the bulk covalent carbon system may be applicable to the non-covalent PNA supramolecular system in that the dipole of the environment molecules (either solvent or PNA backbone) aligns gradually with the excited PNA to minimize the interaction energy. Thus, various intermediate states are formed between the initial excited state and the final relaxed one (Fig. 4f ), resulting in the time-dependent fluorescence spectra collected and the observation of the shift in the fluorescence emission maximum. Such optical behaviour was not observed in parallel peptide assemblies that lack the distinct stabilization with both stacking and base pairing, as observed in the PNA crystal structure (Fig. 2). This indeed most likely represents a unique state of high polarizability in a motionally restricted environment induced by the condensed lattice packing. Intrigued by the optical properties distinguished for the characterized assemblies, we sought to study their capability to serve as an organic light-emitting material in optoelectronics. A simple fieldeffect transistor (FET) device composed of a silicon chip with printed gold source and drain electrodes was used (Fig. 5a). The FET architecture enables the investigation of elementary optoelectronic properties in organic materials and is emerging as a highly useful configuration for applications such as optical communication
2 −2 V 0 −2 −1
−4 V 0
1
2
3
−5
4
−5
d
0
5
Vsd (V)
Vsd (V)
Voltage (V)
ARTICLES
5
0
0
10
20
30
40
50
−5 Time (s)
Figure 5 | PNA-based light-emitting FET. a, Progressive magnifications of the chip, in which PNA assembly channels bridge the source (vertical electrode) with its surrounding drain electrodes. The channel width is 3 µm. b, Output electrical characteristics for negative drain–source voltages. The PNA structure responds to gate voltage in the same manner as a FET, with lower current rates as the voltage increases. The gate voltages are indicated near the curves. c, The PNA assemblies exhibited the nonlinear current– voltage characteristic of a diode. d, The voltage applied on the device was alternated repeatedly from 5 V to −5 V in time steps of 10 s. As the voltage reaches an absolute value of 5 V, the device emits bright luminescence. Images placed over the voltage–time graph are snapshots taken from Supplementary Movie 2, zooming in on the PNA channel.
systems, advanced display technology, electrically pumped organic lasers and solid-state lighting32. The PNA structures were deposited in the gap between the source and drain electrodes on top of the chip. Using this simple platform, the electrical properties of the assemblies were measured. The measured resistance was found to be between 0.1 and 1 MΩ, and the conductance between 0.9 and 1.8 μS. The PNA structures responded to gate voltage in the same manner as a FET, with lower current rates as the voltage increased (Fig. 5b), and exhibited the nonlinear current–voltage characteristic of a diode (Fig. 5c). When applying different voltages to the device, electroluminescence was observed at both 5 V and −5 V. Figure 5d illustrates an experiment in which the voltage was alternated repeatedly between 5 V and −5 V in time steps of 10 s. Supplementary Movie 2 displays one minute of the experiment. As seen in both Fig. 5d and the movie, the device emits bright light every 10 s, as the voltage reaches an absolute value of 5 V. The luminescence fades as the voltage is changed. The light produced by the device flickers dozens of times with no apparent decay. The newly characterized assemblies should have potential for optical biosensing and light emission-based applications including organic light-emitting diodes (OLEDs) and imaging labels with tunable emission via optical or electrical modulation. The blue luminescence produced by the previously reported diphenylalanine nanostructures has raised considerable scientific and industrial interest11. The novel PNA structures offer similar simplicity,
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY
DOI: 10.1038/NNANO.2015.27
prompt and efficient assembly, as well as the availability of industry standard deposition methods such as physical vapour deposition, together with extended and tunable spectral properties and over an order of magnitude higher molar absorptivity relative to the diphenylalanine nanostructures.
Conclusions We have demonstrated the ability of PNA building blocks to selfassemble into ordered architectures. Very simple guanine-containing di-PNA formed well-organized entities coordinated by both stacking interaction as well as Watson–Crick base pairing. The ultrastructures show the combination of intramolecular organization with ordered supramolecular arrangement. The structures were assembled into discrete and uniform entities, exhibiting very fast elongation kinetics. The optical properties of the newly characterized assemblies are especially intriguing. The excitation-dependent emission and dynamic Stokes shift are unique for such organic supramolecular systems. Furthermore, the wide spectral span may be useful for the fabrication of organic optical devices. This new and simple molecular system of notable physical properties could serve as an excellent model for molecular self-assembly as well as for the study of supramolecular polymers. The PNA architectures could also serve in a variety of technological applications in the fields of material science and bionanotechnology.
Methods Materials. Fmoc-protected PNA monomers were purchased from Polyorg, and Fmoc-protected PAL-PEG-PS resin was purchased from Life Technologies. All solvents (peptide grade) used in the synthesis process were purchased from Bio-Lab. All crystallization solutions and equipment were purchased from Hampton Research. PNA synthesis. di-PNAs were synthesized using standard solid-phase protocols24. The crude product was then purified by reversed-phase high-performance liquid chromatography using a C8 column. The product was verified by electrospray ionization time-of-flight mass spectrometry. Scanning electron microscopy. Lyophilized PNA powder was dissolved in 0.1M bicine buffer pH 9.0 to a concentration of 50 mg ml−1. The solution was then diluted with ddH2O to a final concentration of 10 mg ml−1. A 10 µl aliquot of the structures solution was dried at room temperature on a microscope glass coverslip and coated with chromium. Scanning electron microscopy images were taken using a JEOL JSM 6700F FE-SEM operating at 10 kV. Crystallization and X-ray diffraction analysis. The di-PNAs were screened for crystallization conditions using the hanging-drop vapour-diffusion method on siliconized glass coverslips in Linbro plates, using 146 pre-formulated crystallization solutions. All crystallization experiments were performed at 293 K in a temperaturecontrolled room. After 5days, colourless needle-like crystals appeared for GC in 0.1 M bicine pH 9.0, 2% vol/vol 1,4-dioxane, 10% wt/vol PEG 20,000. Before mounting, crystals were soaked for 1 min in a cryo-protecting solution (comprising 16% ethylene glycol, 18% sucrose, 16% glycerol, 4% glucose, mixed in a 1:1 ratio with the crystallization reservoir solution). Crystals were mounted on loops and flash-frozen in liquid nitrogen for transportation to the synchrotron. The data were measured at ESRF beamline ID29 using a Pilatus 6M-F detector and a wavelength of 0.80 Å. A full sphere of 360° of data were collected as 1° frames with a resolution of 0.95 Å (Supplementary Fig. 3). The data were autoprocessed using EDNA33. Two data sets, collected from different locations on the same crystal, were merged in XPREP. The structure was solved by direct methods in SHELXS. The refinements in SHELXL-97 were weighted full-matrix least-squares against |F2| using all data. In the final stages of refinement, SQUEEZE34 was used due to the large voids and remaining disordered solvent molecules. Atoms were refined independently and non-solvent atoms were refined anisotropically with the exception of hydrogen atoms, which were placed in calculated positions and refined in a riding mode. Crystal data collection and refinement parameters are given in Supplementary Table 1 and the complete data can be found in the Supplementary Information (Cif file). Spectroscopy measurements. Emission spectra were taken on a Horiba Jobin Yvon Fluorolog-3 spectrofluorometer at various excitation wavelengths as described in Fig. 4b. Emission was recorded between 350 and 600 nm at 25 °C. Emission and excitation slits were set at 2.5 nm. Measurements were performed in a 1 cm rectangular quartz cuvette containing 5 mg ml−1 of GC di-PNA structures in buffer solution. All spectra were normalized so that the emission maxima and minima were identical.
ARTICLES
Time-resolved spectra were captured using time-correlated single photon counting with a pulsed LED light source at 390 nm. The emission slit was set at 10 nm. A stirred solution of 20 mg ml−1 of GC di-PNA structures in buffer solution was used for the measurements. Fluorescence imaging. A fresh solution of 5 mg ml−1 of GC di-PNA structures was prepared and a volume of 10 µl was deposited on a glass slide and covered with a coverslip. Images were acquired using five different excitation/emission filters as described in Fig. 4a. PNA-based electrical device array fabrication. The silicon wafers were cleaned by washing with acetone, isopropyl alcohol, rinsing thoroughly with deionized water, blowing with dry nitrogen followed by oxygen plasma treatment (100 W, 50 s.c.c.m. O2 for 200 s; Axic HP-8). Source and drain electrodes were defined by mask exposure of a multilayer resist structure consisting of 500 nm LOR-5A copolymer and 500 nm S-1805 photoresist (MicroChem). After exposure, development and gold metallization of the gate, drain and source electrodes pattern (VST), the PNA assemblies, at a controlled density, were deposited between the source and drain electrodes by the dropcasting approach. After the deposition step, the electrodes were briefly washed with deionized water and dried gently with nitrogen. The entire chip was then passivated with a 10–20 nm thin insulating dielectric layer of Si3N4 , deposited by inductively coupled plasma-enhanced chemical vapour deposition (Axic). Electrical characterization. I–V measurements were taken for each device using a probe station (Janis Research) at room temperature. The drain current (Ids) response to the applied Vds , varying between −5 V and +5 V at a rate of 100 mV s−1, was recorded at constant gate bias (taken from +6 V to −6 V).
Received 7 May 2014; accepted 2 February 2015; published online 16 March 2015
References 1. Ghadiri, M. R., Granja, J. R., Milligan, R. A., McRee, D. E. & Khazanovich, N. Self-assembling organic nanotubes based on a cyclic peptide architecture. Nature 366, 324–327 (1993). 2. Hartgerink, J. D., Beniash, E. & Stupp, S. I. Self-assembly and mineralization of peptide–amphiphile nanofibers. Science 294, 1684–1688 (2001). 3. Zhang, S. Fabrication of novel biomaterials through molecular self-assembly. Nature Biotechnol. 21, 1171–1178 (2003). 4. Banwell, E. F. et al. Rational design and application of responsive alpha-helical peptide hydrogels. Nature Mater. 8, 596–600 (2009). 5. Morris, K. L. et al. Chemically programmed self-sorting of gelator networks. Nature Commun. 4, 1480 (2013). 6. Hirst, A. R. et al. Biocatalytic induction of supramolecular order. Nature Chem. 2, 1089–1094 (2010). 7. De la Rica, R. & Matsui, H. Applications of peptide and protein-based materials in bionanotechnology. Chem. Soc. Rev. 39, 3499–3509 (2010). 8. Reches, M. & Gazit, E. Casting metal nanowires within discrete selfassembled peptide nanotubes. Science 300, 625–627 (2003). 9. Yan, X., Zhu, P. & Li, J. Self-assembly and application of diphenylalanine-based nanostructures. Chem. Soc. Rev. 39, 1877–1890 (2010). 10. Adler-Abramovich, L. et al. Self-assembled arrays of peptide nanotubes by vapour deposition. Nature Nanotech. 4, 849–854 (2009). 11. Amdursky, N. et al. Blue luminescence based on quantum confinement at peptide nanotubes. Nano Lett. 9, 3111–3115 (2009). 12. Adler-Abramovich, L. & Gazit, E. The physical properties of supramolecular peptide assemblies: from building block association to technological applications. Chem. Soc. Rev. 43, 6881–6893 (2014). 13. Seeman, N. C. Nucleic acid junctions and lattices. J. Theor. Biol. 99, 237–247 (1982). 14. Rothemund, P. W. K. Folding DNA to create nanoscale shapes and patterns. Nature 440, 297–302 (2006). 15. Ke, Y., Ong, L. L., Shih, W. M. & Yin, P. Three-dimensional structures selfassembled from DNA bricks. Science 338, 1177–1183 (2012). 16. Dietz, H., Douglas, S. M. & Shih, W. M. Folding DNA into twisted and curved nanoscale shapes. Science 325, 725–730 (2009). 17. Egholm, M., Buchardt, O. & Nielsen, P. E. Peptide nucleic acids (PNA). Oligonucleotide analogues with an achiral peptide backbone. J. Am. Chem. Soc. 1, 1895–1897 (1992). 18. Nielsen, P. E., Egholm, M., Berg, R. H. & Buchardt, O. Sequence-selective recognition of DNA by strand displacement with a thymine-substituted polyamide. Science 254, 1497–1500 (1991). 19. Ura, Y., Beierle, J. M., Leman, L. J., Orgel, L. E. & Ghadiri, M. R. Self-assembling sequence-adaptive peptide nucleic acids. Science 325, 73–77 (2009). 20. Li, X. et al. Supramolecular nanofibers and hydrogels of nucleopeptides. Angew. Chem. Int. Ed. 50, 9365–9369 (2011). 21. Bonifazi, D., Carloni, L. E., Corvaglia, V. & Delforge, A. Peptide nucleic acids in materials science. Artif. DNA PNA XNA 3, 112–122 (2012).
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
7
ARTICLES
NATURE NANOTECHNOLOGY
22. Guler, M. O., Pokorski, J. K., Appella, D. H. & Stupp, S. I. Enhanced oligonucleotide binding to self-assembled nanofibers. Bioconjug. Chem. 16, 501–503 (2005). 23. Böhler, C., Nielsen, P. E. & Orgel, L. E. Template switching between PNA and RNA oligonucleotides. Nature 376, 578–581 (1995). 24. Uhlmann, E., Peyman, A., Breipohl, G. & Will, D. W. PNA: synthetic polyamide nucleic acids with unusual binding properties. Angew. Chem. Int. Ed. 37, 2796–2823 (1998). 25. Davis, J. T. & Spada, G. P. Supramolecular architectures generated by self-assembly of guanosine derivatives. Chem. Soc. Rev. 36, 296–313 (2007). 26. Menchise, V. et al. Insights into peptide nucleic acid (PNA) structural features: the crystal structure of a D-lysine-based chiral PNA–DNA duplex. Proc. Natl Acad. Sci. USA 100, 12021–12026 (2003). 27. Watson, J. D. & Crick, F. H. C. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 171, 737–738 (1953). 28. Trushko, A., Schäffer, E. & Howard, J. The growth speed of microtubules with XMAP215-coated beads coupled to their ends is increased by tensile force. Proc. Natl Acad. Sci. USA 110, 14670–14675 (2013). 29. Ikezoe, Y., Washino, G., Uemura, T., Kitagawa, S. & Matsui, H. Autonomous motors of a metal–organic framework powered by reorganization of selfassembled peptides at interfaces. Nature Mater. 11, 1081–1085 (2012). 30. Demchenko, A. P. The red-edge effect: 30 years of exploration. Luminesence 17, 19–42 (2002). 31. Cushing, S. K., Li, M., Huang, F. & Wu, N. Origin of strong excitation wavelength dependent fluoresence of graphene oxide. ACS Nano 8, 1002–1013 (2014). 32. Muccini, M. A bright future for organic field-effect transistors. Nature Mater. 5, 605–613 (2006).
8
DOI: 10.1038/NNANO.2015.27
33. Incardona, M. F. et al. EDNA: a framework for plugin-based applications applied to X-ray experiment online data analysis. J. Synchrotron Radiat. 16, 872–879 (2009). 34. Spek, A. L. Structure validation in chemical crystallography. Acta Crystallogr. D 65, 148–155 (2009).
Acknowledgements This work was supported in part by grants from the Israeli National Nanotechnology Initiative and Helmsley Charitable Trust for a focal technology area on Nanomedicine for Personalized Theranostics. The authors thank members of the Gazit Laboratory for helpful discussions, Y. Salitra for help with PNA synthesis and O. Yaniv for advice on crystallization experiments. The authors acknowledge the ESRF for synchrotron beam time and the staff scientists of the ID29 beamline for their assistance. O. Berger is supported by a fellowship from the Argentinean Friends of Tel Aviv University Association.
Author contributions O.B., L.A-A., L.B., Y.E. and E.G. designed the study. O.B., M.L-S., Y.L-P. and M.B. performed the experiments. O.B., A.G., T.S. and Y.E. analysed the data. F.F., L.J.W.S., E.M. and T.F. performed and analysed the X-ray diffraction experiments. F.P. performed OFET experiments. O.B., L.A-A. and E.G. prepared the manuscript.
Additional information Supplementary information is available in the online version of the paper. Reprints and permissions information is available online at www.nature.com/reprints. Correspondence and requests for materials should be addressed to E.G.
Competing financial interests
The authors declare no competing financial interests.
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
Light-emitting self-assembled peptide nucleic acids exhibit both stacking interactions and Watson–Crick base pairing Or Berger1, Lihi Adler-Abramovich1, Michal Levy-Sakin2, Assaf Grunwald2, Yael Liebes-Peer2, Mor Bachar1, Ludmila Buzhansky1, Estelle Mossou3,4, V. Trevor Forsyth3,4, Tal Schwartz2, Yuval Ebenstein2, Felix Frolow1,5†, Linda J. W. Shimon6, Fernando Patolsky2,7 and Ehud Gazit1,7* The two main branches of bionanotechnology involve the self-assembly of either peptides or DNA. Peptide scaffolds offer chemical versatility, architectural flexibility and structural complexity, but they lack the precise base pairing and molecular recognition available with nucleic acid assemblies. Here, inspired by the ability of aromatic dipeptides to form ordered nanostructures with unique physical properties, we explore the assembly of peptide nucleic acids (PNAs), which are short DNA mimics that have an amide backbone. All 16 combinations of the very short di-PNA building blocks were synthesized and assayed for their ability to self-associate. Only three guanine-containing di-PNAs—CG, GC and GG—could form ordered assemblies, as observed by electron microscopy, and these di-PNAs efficiently assembled into discrete architectures within a few minutes. The X-ray crystal structure of the GC di-PNA showed the occurrence of both stacking interactions and Watson–Crick base pairing. The assemblies were also found to exhibit optical properties including voltage-dependent electroluminescence and wide-range excitation-dependent fluorescence in the visible region.
N
ature has produced a basic set of building-block molecules through billions of years of molecular selection and evolution. The 20 coded amino acids and four primary nucleotides are the most fundamental elements in living systems. Protein and nucleic acid biomolecules allow the formation of an enormously diverse range of supramolecular assemblies, exhibiting a vast array of physical properties. Inspired by natural ordered assemblies, many studies have been directed towards the design of peptide and protein building blocks that self-assemble into preferred architectures1–7. The shortest, most simple peptide building block shown to self-assemble into ordered architectures is diphenylalanine, the core recognition motif of β-amyloid polypeptide8. This aromatic dipeptide self-assembles into discrete well-ordered nanotubular structures of notable persistence length in aqueous solution9. Various studies have shown the remarkable physical characteristics of these nanotubes, including metal-like rigidity and high thermal and chemical stability, as well as optical, semiconductive and piezoelectric properties10–12. In contrast to peptide self-assembly, structural DNA nanotechnology is solely derived from the specificity of the hydrogenbonding interactions between complementary Watson–Crick base pairs. The use of DNA as a structural building block instead of merely a genetic material was first recognized in the early 1980s, in a theoretical work by Seeman13. Based on the complementary nature of nucleic acids, it is possible to predict and design DNA structures of nanoscale order. This pioneering conceptual work materialized into a vivid field of research with tremendous growth. Now, with the aid of computer design and the DNA
origami method, the fabrication of any two- or even three-dimensional nanostructure shape has become considerably simpler14–16. Peptide and DNA building blocks offer two distinct approaches for the generation of supramolecular architectures by self-assembly. Peptide-driven materials are characterized by robustness, synthetic versatility, architectural flexibility and structural complexity, whereas DNA nanostructures are based on specific molecular recognition and base-pairing complementarity. Convergence of the peptide and nucleic acid assembly strategies could be very useful for the design of novel self-organized materials. Peptide nucleic acids are artificially synthesized polymers that were first described by Nielsen and co-workers in 1991 (refs 17, 18). The polymer is an oligonucleotide analogue in which the phosphodiester backbone is replaced by repeating N-(2-aminoethyl) glycine units linked by amide bonds. Thus, PNA can be regarded as either a DNA mimic with a neutral amide backbone, or as a peptide mimic with nucleobases as side chains. Another advantage of PNAs is their resistance to degradation by proteases or nucleases. A slightly different approach for the production of a peptide–nucleic acid hybrid is based on the conjugation of functionalized purine and pyrimidine bases to a peptide backbone19. The neutral peptide-like backbone replacing the negatively charged phosphodiester groups, which may limit self-association due to the repulsion of similar charges, adds structural elasticity and chemical adaptability. Another group of unnatural peptide–nucleobases hybrids (denoted nucleopeptides) have already been shown to self-assemble into nanofibres to generate supramolecular hydrogels20. Despite the great potential to converge the two distinct fields of peptide self-assembly and structural DNA
1
Department of Molecular Microbiology and Biotechnology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv 69978, Israel. 2 School of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Ramat Aviv 69978, Israel. 3 Partnership for Structural Biology, Institut Laue Langevin, 71 Avenue des Martyrs, Grenoble Cedex 9 38042, France. 4 Faculty of Natural Sciences, Keele University, Staffordshire ST5 5BG, UK. 5 Daniella Rich Institute for Structural Biology, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv 69978, Israel. 6 Department of Chemical Research Support, Weizmann Institute of Science, Rehovot 76100, Israel. 7 Department of Materials Science and Engineering, Iby and Aladar Fleischman Faculty of Engineering, Tel Aviv University, Ramat Aviv 69978, Israel. †Deceased. *e-mail: [email protected] NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
1
ARTICLES
NATURE NANOTECHNOLOGY
a
b HN
N O
H2N N O N
H2N
T
A
AA
AC
AG
AT
C
CA
CC
CG
CT
G
GA
GC
GG
GT
T
TA
TC
TG
TT
H2N O
N
N
N N
O
H2N
H2N N O
O
N
N H
N
12.0
N
N
HN H2N O
NH2
N
H2N
N
N
N
O
f
11.9
O
HN
N
O NH2
O
NH2
O
e
h pH = 11.8
d
O
N
O N H
G
HN
N
N
C
c
O
NH2
A
DOI: 10.1038/NNANO.2015.27
N O
O N
N H
O NH2
g
12.1
12.2
12.3
12.4
Figure 1 | Self-assembly of guanine-containing di-PNAs into well-ordered architectures. a, The 16 different di-PNA building blocks were synthesized and assayed for their ability to undergo self-assembly into distinct structural entities. Molecular assembly was evident with three di-PNAs under alkaline conditions (highlighted in black). More assemblies were observed for three other di-PNAs on drying the sample (highlighted in grey). b–d, Chemical structures of the three assembly-forming di-PNAs: CG (b), GC (c) and GG (d). e–g, SEM micrographs of the structures formed by CG (e), GC (f) and GG (g). Scale bars, 10 µm. h, Light microscopy images of assemblies formed by GC di-PNA on dissolving with rising pH levels of disodium hydrogen phosphate buffer. Original magnification, ×40.
nanotechnology, no work has been performed on PNAs as building blocks to assemble into distinct structural entities by themselves. As described in a recent review21, PNA has so far only been used as a template or as a conjugate to other molecules that undergo the selfassembly process, in order to gain the hybridization properties of nucleic acids. A notable example is the work of Stupp and coworkers22, in which a PNA strand and a peptide sequence that promotes nanofibre formation were coupled to form a PNA–peptide amphiphile conjugate. The designed fibre-shaped nanostructures show binding of oligonucleotides with high affinity and specificity. Here, we report for the first time the efficient and rapid formation of supramolecular architectures based solely on PNA molecule self-assembly.
di-PNA synthesis, assembly and structure Inspired by the ability of simple aromatic dipeptides to form unique assemblies, we decided to examine the ability of short di-PNA 2
building blocks to form ordered supramolecular assemblies. All of the 16 different di-PNA combinations were synthesized using solid-phase peptide synthesis (AA, AC, AG, AT, CA, CC, CG, CT, GA, GC, GG, GT, TA, TC, TG, TT; italics are used to denote PNA) (Fig. 1a). The synthesis of the di-PNA was performed according to conventional peptide synthesis practice using standard methods and commercially available protected building blocks. The synthetic effort, yield and number of steps are comparable to those for the production of other aromatic dipeptides that are readily synthesized at an industrial scale, such as L-aspartyl-L-phenylalanine methyl ester. Favourable conditions for self-organization were determined by screening the di-PNAs in a variety of solvents including organic solvents (methanol, ethanol, dimethyl sulphoxide, hexafluoro-2propanol, and so on) and diverse buffer solutions with a range of pH values and concentrations. Under alkaline conditions, microscopic observation of molecular assembly was evident with only
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY
ARTICLES
DOI: 10.1038/NNANO.2015.27
a
b
c
d 3.57
2.87
3.48
3.47
3.48
3.57
2.93 3.47 2.85
e
f
Figure 2 | Crystal structure of GC di-PNA. The GC di-PNA building block was crystallized and its structure determined by synchrotron radiation at a resolution of 0.95 Å. a, Molecular structure of a single GC di-PNA molecule from single-crystal structure determination. The cytosine and guanine nucleobases form an intramolecular stacking interaction. b, Each molecule forms hydrogen bonds with a neighbouring unit between the cytosine and guanine residues. c, The hydrogen bond length between symmetry-related molecules is measured to be 2.85–2.93 Å, the same as in typical Watson–Crick base pairs. d, The bases are 3.5 Å apart, the same as in a DNA double-helix structure. e, The di-PNAs are packed in a continuous tilted stack through the crystal. f, When lining the stacks in the z direction it is evident that this form of packing results in rectangular-shaped pores comprising over 50% of the crystal volume.
three di-PNAs: CG, GC and GG (for structures see Fig. 1b–d). Examination of the solutions using light and electron microscopy revealed well-organized architectures, including long rods (tens of micrometres long) for CG and GC (Fig. 1e,f ) and spheroids with a diameter of 2–3 μm for GG (Fig. 1g). Other di-PNAs did not form any type of ordered structure, or did so only upon drying of the solution (AG, GA and GT; Supplementary Fig. 1a–f ). Co-assembly of all the possible complementary di-PNA combinations at a 50:50 molar ratio was also examined. Well-defined structures were observed solely for the combination of CC and GG. When mixed together, instead of the spherical assemblies formed by GG alone, elongated structures similar to those generated by CG or GC were evident (Supplementary Fig. 1g). Based on these findings, we speculate that it is necessary to have a minimum of six hydrogen bonds between the bases for the stabilization of complementary di-PNAs and for further organization into supramolecular entities. Interestingly, all PNAs that were able to form structures contained
the guanine nucleobase. As it is known that the secondary structure of G-containing PNA oligomers may be altered under alkaline conditions due to deprotonation of the guanine bases23,24, the assembly of the GC di-PNA was further examined under increasing pH levels (Fig. 1h). Furthermore, alkaline conditions have been found to be essential for self-assembly, presumably due to the introduction of charges, so high salt conditions (which may shield these charges) were examined. Accordingly, GC di-PNA was dissolved in assembly buffer with 1 M sodium chloride. Under this condition, only small nucleation sites were observed. Typical structures initiated the assembly process following a tenfold dilution of the salt (Supplementary Fig. 1h). In this context it should also be noted that guanine is a key component in the assembly of various natural nucleic-acid structures. Nucleic-acid sequences that are rich in guanine are capable of forming G-quadruplexes, the main structural motif of the telomeric DNA. Moreover, guanosine analogues are able to self-associate into
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
3
ARTICLES
NATURE NANOTECHNOLOGY
a
00:00:00
00:00:30
00:01:00
00:01:30
00:02:00
b 45
Length (µm)
40
y = 0.2439x + 21.587 R2 = 0.9722
35 30 25 20 15 0
20
40
60
80
Time (s)
Figure 3 | The di-PNAs efficiently assemble into discrete architectures within a few minutes. a, Five snapshots taken from Supplementary Movie 1 demonstrate the assembly kinetics of the GC di-PNA architectures. A thin glass capillary was filled with a freshly dissolved GC di-PNA solution and sealed immediately at both ends with wax. The snapshots represented here were captured every 30 s. Each panel shows the full width of the capillary (200 µm). b, The average measured elongation rate is 0.25 µm s−1. Graphical representation of the elongation rate of a single structure.
dimers, ribbons and macrocycles that can further stack into supramolecular assemblies25. As a control we studied the ability of guanine-containing DNA dinucleotides to form ordered assemblies. No conditions under which ordered dinucleotides structures could be formed were found. To gain better insight into the assembly process and stabilizing interactions, we attempted to form ordered single crystals of the assemblies and acquire X-ray diffraction data. To this end, the di-PNAs were screened for crystallization conditions. Crystals of GC di-PNA were found to grow in a bicine-based crystallization buffer. Because bicine buffer also enables assembly of the structures, we strongly believe the crystal structure reflects the solution selfassembled architecture. The crystal structure of GC was determined at a resolution of 0.95 Å with data collected at the European Synchrotron Radiation Facility (ESRF). The determined structure revealed very unique packing of the PNA crystals. The cytosine and guanine in each molecule form stacking interactions (Fig. 2a). Each molecule then forms hydrogen bonds with a neighbouring unit between the cytosine and guanine residues (Fig. 2b). All of the bases are engaged in Watson–Crick hydrogen bonds, as observed in canonical DNA–DNA and PNA–DNA duplexes26. The hydrogen bond length between 4
DOI: 10.1038/NNANO.2015.27
symmetry-related molecules was measured to be 2.85–2.93 Å (Fig. 2c), which is similar to that of Watson–Crick base pairing. The bases are found to be ∼3.5 Å apart (Fig. 2d), which is characteristic of DNA double-helical structures27, but they do not exhibit any tilt or roll. The hydrogen-bonding molecules are related to each other via the two-fold dyad that passes between the stacked bases. Packing of the molecule in a centrosymmetric space group is possible due to the non-chiral nature of the polyglycine backbone. The bases are packed in a continuous tilted stack throughout the crystal (Fig. 2e). This form of packing results in rectangularshaped pores that comprise over 50% of the crystal volume (Fig. 2f ). The crystal structure reflects the dual identity of PNA. The di-PNAs form stacking interactions with each other, in the same way as aromatic peptides, while at the same time forming the Watson–Crick base pairs typical of DNA structures. This unique duality distinguishes these molecules from simple DNA dinucleotides that possess only Watson–Crick base pairing and do not form any self-assembled structures. It is also possible that the negatively charged phosphate backbone of simple DNA oligomers may limit the assembly properties. The obtained crystal structure enabled us to estimate the rate at which the di-PNA monomers self-assemble in solution to form the ordered structures. In Supplementary Movie 1 we capture the assembly process of the GC di-PNA in real time. Briefly, a thin glass capillary was filled with a fresh solution of the PNA building blocks and sealed immediately from both sides with wax to prevent evaporation and concentration changes of the solution. The capillary was monitored using light microscopy, and images were captured at a rate of one frame per second. Small nucleation seeds could be observed within a few seconds, and continual growth in one axis direction was sustained for a few minutes. Figure 3a depicts one capillary at five consecutive time points. To allow better visualization of the elongation of the structure as a function of time, one structure was tracked in a time frame in which the assembled structure is clearly seen and is not overlapped by other architectures (Supplementary Fig. 2). Distinct elongating structures from five different capillaries, each filled with freshly dissolved solution, were examined for their growth rate, and an average of 0.25 µm s−1 was measured. All the structures exhibited an excellent linear fit between the dimension of the assemblies in the long Z-axis and time (R 2 > 0.97). The elongation of a representative single structure as a function of time is given in Fig. 3b. For this specific structure, which is 1.77 µm wide, the elongation rate of 0.24 µm s−1 translates into a volume increase of ∼0.6 µm3 s−1. Because the crystal unit cell has a volume of 11,676 Å3 (Supplementary Table 1) and contains eight molecules, we can estimate that 4.11 × 108 di-PNA building-block molecules organize into the ordered structures each second. In comparison to other rapid elongating systems of a natural origin, such as the microtubule that elongates by an average rate of 0.66 µm per minute28, the assembly kinetics of the di-PNA is over 20 times faster. When the same experiment was carried out in an open environment, instead of a sealed capillary, the elongation rate was about ten times faster due to evaporation of the solution, leading to higher local concentrations of PNA. The rapid assembly of the di-PNA building blocks suggests they could be used in motor systems by converting the free energy emitted by the self-assembly process into mechanical motion29. Another very interesting property of this self-organization, important for technological applications, is its efficient, high-yield and uniform process, as only ordered homogeneous assemblies of the GC di-PNA could be observed under the examined conditions.
Optical characterization of PNA assemblies When trying to examine whether the PNA structures could bind DNA intercalators due to their Watson–Crick base pairing, we
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY
ARTICLES
DOI: 10.1038/NNANO.2015.27
a
Intensity (a.u.)
330 nm 340 nm 350 nm 360 nm 370 nm 380 nm 390 nm 400 nm 410 nm 420 nm 430 nm
Emission wavelength (nm)
c 500
b
y = 0.7073x + 187.51 R2 = 0.9943
480
460
440
420
400 350
400
450
500
550
320
600
360
400
440
Excitation wavelength (nm)
Wavelength (nm)
e
d
490 nm 470 nm
0.6 Δtmax (ns)
Normalized counts (a.u.)
510 nm
0.8 y = 0.0071x − 2.8495 R2 = 0.9697
0.4
450 nm 0.2 430 nm 410 nm 98
99
100
0.0 400
Time (ns)
+
+ −
−+
−+
+
−+ −+
−
+ − + − −+
+ −
−
−+
−
−+
−+
+
hν∞
−+
λR
−+
+
−+
−
−+ −+
hνI
−+
hν0
−+
−+ −+
λC
−+
−+
−+
−+
Solvent relaxed (R) state
−+
+
Intermediate (I) state
−+
−
520
−+
−+
−
+
480
−+
Franck−Condon (FC) state
f
440
Emission wavelength (nm)
−+
97
−+
96
−+ −+
95
Figure 4 | The PNA assemblies exhibit a red edge excitation shift with a broad range of emission wavelengths in the visible region. a, One bright-field and five fluorescence images of the same microscopic field of GC di-PNA structures. Each fluorescence image was taken with different excitation and emission filters: (excitation/emission) 387 nm/440 nm; 485 nm/525 nm; 537 nm/578 nm; 560 nm/607 nm; 650 nm/684 nm (from left to right). Pseudo-colours represent corresponding emission colour. Original magnification, ×100. b, Emission spectra of GC di-PNA assemblies at excitation wavelengths of 330, 340, 350, 360, 370, 380, 390, 400, 410, 420 and 430 nm. The emission peak shifts to the red with higher excitation wavelengths. c, Graphical representation of the relation between the excitation and emission wavelengths. The slope is measured to be ∼0.7, which suggests a dynamic Stokes shift. d, Time-resolved emission traces for different emission wavelengths following excitation at 390 nm. The emission onset is delayed with respect to the excitation peak as emission wavelength shifts to the red. e, Time delay of fluorescence as a function of wavelength. Δt is calculated to the fastest emission onset collected at 410 nm. The monotonic increase in Δt is a signature of REES. f, Continuous model of solvent relaxation. The I state refers to one of the intermediate states between the initial excited state and the final solvent relaxed state, in which the solvent molecules are partially relaxed. ν0 , νI and ν∞ represent the frequencies corresponding to the initially excited (Franck–Condon), intermediate and completely relaxed states, respectively, while λC and λR denote the wavelength maxima associated with these states. NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
5
NATURE NANOTECHNOLOGY
6
DOI: 10.1038/NNANO.2015.27
b
c
5
Isd (μA)
a
0
0V
4 Isd (μA)
made a surprising observation, as the control sample with no added dye had a fluorescence signal similar to that of the stained samples. Even more surprising was the fact that fluorescence emission was evident for a wide range of excitation wavelengths (Fig. 4a). Clear emission peaks were observed between 420 nm and 490 nm with the excitation wavelength ranging from 330 nm to 430 nm. Fair fluorescent images could be obtained even for higher wavelengths, but with lower intensity. As presented in Fig. 4a, the upper limit of the emission was observed at 684 nm when the excitation wavelength was 650 nm. To quantitatively analyse this phenomenon, fluorescence emission spectra were determined at different excitation wavelengths (Fig. 4b). It is clearly noticeable from the studies that the emission peak shifts in the direction of the change in excitation. To determine whether the shift is consistent for various excitation wavelengths, we plotted consecutive emission maxima as a function of excitation wavelengths (Fig. 4c). We observed a linear correlation between the excitation and emission peak wavelengths. Yet, the 0.7073 slope of the very strong fit (R 2 = 0.9943) indicates dynamic Stokes shift behaviour as the distance between the excitation and emission peaks gradually decreases with longer wavelengths. Such an observed change in fluorescence emission spectra in response to a shift in the excitation wavelength toward the red edge of the absorption band is termed a red edge excitation shift (REES). The phenomenon was originally described in rigid and highly viscous environments such as low-temperature glasses, graphene oxide layers or highly condensed polymeric states30. REES is assumed to be the result of the strong reduction in the dynamic environment of the excited fluorophores in organized molecular settings. A model for this spectroscopic behaviour assumes that the molecular lattice confinements slow the rates of matrix relaxation and reorientation around the excited state of the fluorophore relative to the fluorescence lifetime31. In biological and other organic molecules, such constraints could be imposed by exceptionally ordered hydration shells or rigid membranes. Time-dependent fluorescence measurements were used to determine whether this model could explain the origin of the excitationdependent emission behaviour observed with the PNA assemblies. The fluorescence decay was measured at several emission wavelengths following excitation at 390 nm (Fig. 4d). The results clearly indicate a process of time-dependent fluorescence onset, with longer wavelengths appearing with larger delays after the excitation pulse (Fig. 4e) (for example, the emission at 510 nm appearing 700 ps after the emission at 410 nm). The optical phenomenon observed here appears to be similar to the very strong REES effect shown in rigid graphene oxide31. The model used for the bulk covalent carbon system may be applicable to the non-covalent PNA supramolecular system in that the dipole of the environment molecules (either solvent or PNA backbone) aligns gradually with the excited PNA to minimize the interaction energy. Thus, various intermediate states are formed between the initial excited state and the final relaxed one (Fig. 4f ), resulting in the time-dependent fluorescence spectra collected and the observation of the shift in the fluorescence emission maximum. Such optical behaviour was not observed in parallel peptide assemblies that lack the distinct stabilization with both stacking and base pairing, as observed in the PNA crystal structure (Fig. 2). This indeed most likely represents a unique state of high polarizability in a motionally restricted environment induced by the condensed lattice packing. Intrigued by the optical properties distinguished for the characterized assemblies, we sought to study their capability to serve as an organic light-emitting material in optoelectronics. A simple fieldeffect transistor (FET) device composed of a silicon chip with printed gold source and drain electrodes was used (Fig. 5a). The FET architecture enables the investigation of elementary optoelectronic properties in organic materials and is emerging as a highly useful configuration for applications such as optical communication
2 −2 V 0 −2 −1
−4 V 0
1
2
3
−5
4
−5
d
0
5
Vsd (V)
Vsd (V)
Voltage (V)
ARTICLES
5
0
0
10
20
30
40
50
−5 Time (s)
Figure 5 | PNA-based light-emitting FET. a, Progressive magnifications of the chip, in which PNA assembly channels bridge the source (vertical electrode) with its surrounding drain electrodes. The channel width is 3 µm. b, Output electrical characteristics for negative drain–source voltages. The PNA structure responds to gate voltage in the same manner as a FET, with lower current rates as the voltage increases. The gate voltages are indicated near the curves. c, The PNA assemblies exhibited the nonlinear current– voltage characteristic of a diode. d, The voltage applied on the device was alternated repeatedly from 5 V to −5 V in time steps of 10 s. As the voltage reaches an absolute value of 5 V, the device emits bright luminescence. Images placed over the voltage–time graph are snapshots taken from Supplementary Movie 2, zooming in on the PNA channel.
systems, advanced display technology, electrically pumped organic lasers and solid-state lighting32. The PNA structures were deposited in the gap between the source and drain electrodes on top of the chip. Using this simple platform, the electrical properties of the assemblies were measured. The measured resistance was found to be between 0.1 and 1 MΩ, and the conductance between 0.9 and 1.8 μS. The PNA structures responded to gate voltage in the same manner as a FET, with lower current rates as the voltage increased (Fig. 5b), and exhibited the nonlinear current–voltage characteristic of a diode (Fig. 5c). When applying different voltages to the device, electroluminescence was observed at both 5 V and −5 V. Figure 5d illustrates an experiment in which the voltage was alternated repeatedly between 5 V and −5 V in time steps of 10 s. Supplementary Movie 2 displays one minute of the experiment. As seen in both Fig. 5d and the movie, the device emits bright light every 10 s, as the voltage reaches an absolute value of 5 V. The luminescence fades as the voltage is changed. The light produced by the device flickers dozens of times with no apparent decay. The newly characterized assemblies should have potential for optical biosensing and light emission-based applications including organic light-emitting diodes (OLEDs) and imaging labels with tunable emission via optical or electrical modulation. The blue luminescence produced by the previously reported diphenylalanine nanostructures has raised considerable scientific and industrial interest11. The novel PNA structures offer similar simplicity,
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
NATURE NANOTECHNOLOGY
DOI: 10.1038/NNANO.2015.27
prompt and efficient assembly, as well as the availability of industry standard deposition methods such as physical vapour deposition, together with extended and tunable spectral properties and over an order of magnitude higher molar absorptivity relative to the diphenylalanine nanostructures.
Conclusions We have demonstrated the ability of PNA building blocks to selfassemble into ordered architectures. Very simple guanine-containing di-PNA formed well-organized entities coordinated by both stacking interaction as well as Watson–Crick base pairing. The ultrastructures show the combination of intramolecular organization with ordered supramolecular arrangement. The structures were assembled into discrete and uniform entities, exhibiting very fast elongation kinetics. The optical properties of the newly characterized assemblies are especially intriguing. The excitation-dependent emission and dynamic Stokes shift are unique for such organic supramolecular systems. Furthermore, the wide spectral span may be useful for the fabrication of organic optical devices. This new and simple molecular system of notable physical properties could serve as an excellent model for molecular self-assembly as well as for the study of supramolecular polymers. The PNA architectures could also serve in a variety of technological applications in the fields of material science and bionanotechnology.
Methods Materials. Fmoc-protected PNA monomers were purchased from Polyorg, and Fmoc-protected PAL-PEG-PS resin was purchased from Life Technologies. All solvents (peptide grade) used in the synthesis process were purchased from Bio-Lab. All crystallization solutions and equipment were purchased from Hampton Research. PNA synthesis. di-PNAs were synthesized using standard solid-phase protocols24. The crude product was then purified by reversed-phase high-performance liquid chromatography using a C8 column. The product was verified by electrospray ionization time-of-flight mass spectrometry. Scanning electron microscopy. Lyophilized PNA powder was dissolved in 0.1M bicine buffer pH 9.0 to a concentration of 50 mg ml−1. The solution was then diluted with ddH2O to a final concentration of 10 mg ml−1. A 10 µl aliquot of the structures solution was dried at room temperature on a microscope glass coverslip and coated with chromium. Scanning electron microscopy images were taken using a JEOL JSM 6700F FE-SEM operating at 10 kV. Crystallization and X-ray diffraction analysis. The di-PNAs were screened for crystallization conditions using the hanging-drop vapour-diffusion method on siliconized glass coverslips in Linbro plates, using 146 pre-formulated crystallization solutions. All crystallization experiments were performed at 293 K in a temperaturecontrolled room. After 5days, colourless needle-like crystals appeared for GC in 0.1 M bicine pH 9.0, 2% vol/vol 1,4-dioxane, 10% wt/vol PEG 20,000. Before mounting, crystals were soaked for 1 min in a cryo-protecting solution (comprising 16% ethylene glycol, 18% sucrose, 16% glycerol, 4% glucose, mixed in a 1:1 ratio with the crystallization reservoir solution). Crystals were mounted on loops and flash-frozen in liquid nitrogen for transportation to the synchrotron. The data were measured at ESRF beamline ID29 using a Pilatus 6M-F detector and a wavelength of 0.80 Å. A full sphere of 360° of data were collected as 1° frames with a resolution of 0.95 Å (Supplementary Fig. 3). The data were autoprocessed using EDNA33. Two data sets, collected from different locations on the same crystal, were merged in XPREP. The structure was solved by direct methods in SHELXS. The refinements in SHELXL-97 were weighted full-matrix least-squares against |F2| using all data. In the final stages of refinement, SQUEEZE34 was used due to the large voids and remaining disordered solvent molecules. Atoms were refined independently and non-solvent atoms were refined anisotropically with the exception of hydrogen atoms, which were placed in calculated positions and refined in a riding mode. Crystal data collection and refinement parameters are given in Supplementary Table 1 and the complete data can be found in the Supplementary Information (Cif file). Spectroscopy measurements. Emission spectra were taken on a Horiba Jobin Yvon Fluorolog-3 spectrofluorometer at various excitation wavelengths as described in Fig. 4b. Emission was recorded between 350 and 600 nm at 25 °C. Emission and excitation slits were set at 2.5 nm. Measurements were performed in a 1 cm rectangular quartz cuvette containing 5 mg ml−1 of GC di-PNA structures in buffer solution. All spectra were normalized so that the emission maxima and minima were identical.
ARTICLES
Time-resolved spectra were captured using time-correlated single photon counting with a pulsed LED light source at 390 nm. The emission slit was set at 10 nm. A stirred solution of 20 mg ml−1 of GC di-PNA structures in buffer solution was used for the measurements. Fluorescence imaging. A fresh solution of 5 mg ml−1 of GC di-PNA structures was prepared and a volume of 10 µl was deposited on a glass slide and covered with a coverslip. Images were acquired using five different excitation/emission filters as described in Fig. 4a. PNA-based electrical device array fabrication. The silicon wafers were cleaned by washing with acetone, isopropyl alcohol, rinsing thoroughly with deionized water, blowing with dry nitrogen followed by oxygen plasma treatment (100 W, 50 s.c.c.m. O2 for 200 s; Axic HP-8). Source and drain electrodes were defined by mask exposure of a multilayer resist structure consisting of 500 nm LOR-5A copolymer and 500 nm S-1805 photoresist (MicroChem). After exposure, development and gold metallization of the gate, drain and source electrodes pattern (VST), the PNA assemblies, at a controlled density, were deposited between the source and drain electrodes by the dropcasting approach. After the deposition step, the electrodes were briefly washed with deionized water and dried gently with nitrogen. The entire chip was then passivated with a 10–20 nm thin insulating dielectric layer of Si3N4 , deposited by inductively coupled plasma-enhanced chemical vapour deposition (Axic). Electrical characterization. I–V measurements were taken for each device using a probe station (Janis Research) at room temperature. The drain current (Ids) response to the applied Vds , varying between −5 V and +5 V at a rate of 100 mV s−1, was recorded at constant gate bias (taken from +6 V to −6 V).
Received 7 May 2014; accepted 2 February 2015; published online 16 March 2015
References 1. Ghadiri, M. R., Granja, J. R., Milligan, R. A., McRee, D. E. & Khazanovich, N. Self-assembling organic nanotubes based on a cyclic peptide architecture. Nature 366, 324–327 (1993). 2. Hartgerink, J. D., Beniash, E. & Stupp, S. I. Self-assembly and mineralization of peptide–amphiphile nanofibers. Science 294, 1684–1688 (2001). 3. Zhang, S. Fabrication of novel biomaterials through molecular self-assembly. Nature Biotechnol. 21, 1171–1178 (2003). 4. Banwell, E. F. et al. Rational design and application of responsive alpha-helical peptide hydrogels. Nature Mater. 8, 596–600 (2009). 5. Morris, K. L. et al. Chemically programmed self-sorting of gelator networks. Nature Commun. 4, 1480 (2013). 6. Hirst, A. R. et al. Biocatalytic induction of supramolecular order. Nature Chem. 2, 1089–1094 (2010). 7. De la Rica, R. & Matsui, H. Applications of peptide and protein-based materials in bionanotechnology. Chem. Soc. Rev. 39, 3499–3509 (2010). 8. Reches, M. & Gazit, E. Casting metal nanowires within discrete selfassembled peptide nanotubes. Science 300, 625–627 (2003). 9. Yan, X., Zhu, P. & Li, J. Self-assembly and application of diphenylalanine-based nanostructures. Chem. Soc. Rev. 39, 1877–1890 (2010). 10. Adler-Abramovich, L. et al. Self-assembled arrays of peptide nanotubes by vapour deposition. Nature Nanotech. 4, 849–854 (2009). 11. Amdursky, N. et al. Blue luminescence based on quantum confinement at peptide nanotubes. Nano Lett. 9, 3111–3115 (2009). 12. Adler-Abramovich, L. & Gazit, E. The physical properties of supramolecular peptide assemblies: from building block association to technological applications. Chem. Soc. Rev. 43, 6881–6893 (2014). 13. Seeman, N. C. Nucleic acid junctions and lattices. J. Theor. Biol. 99, 237–247 (1982). 14. Rothemund, P. W. K. Folding DNA to create nanoscale shapes and patterns. Nature 440, 297–302 (2006). 15. Ke, Y., Ong, L. L., Shih, W. M. & Yin, P. Three-dimensional structures selfassembled from DNA bricks. Science 338, 1177–1183 (2012). 16. Dietz, H., Douglas, S. M. & Shih, W. M. Folding DNA into twisted and curved nanoscale shapes. Science 325, 725–730 (2009). 17. Egholm, M., Buchardt, O. & Nielsen, P. E. Peptide nucleic acids (PNA). Oligonucleotide analogues with an achiral peptide backbone. J. Am. Chem. Soc. 1, 1895–1897 (1992). 18. Nielsen, P. E., Egholm, M., Berg, R. H. & Buchardt, O. Sequence-selective recognition of DNA by strand displacement with a thymine-substituted polyamide. Science 254, 1497–1500 (1991). 19. Ura, Y., Beierle, J. M., Leman, L. J., Orgel, L. E. & Ghadiri, M. R. Self-assembling sequence-adaptive peptide nucleic acids. Science 325, 73–77 (2009). 20. Li, X. et al. Supramolecular nanofibers and hydrogels of nucleopeptides. Angew. Chem. Int. Ed. 50, 9365–9369 (2011). 21. Bonifazi, D., Carloni, L. E., Corvaglia, V. & Delforge, A. Peptide nucleic acids in materials science. Artif. DNA PNA XNA 3, 112–122 (2012).
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
7
ARTICLES
NATURE NANOTECHNOLOGY
22. Guler, M. O., Pokorski, J. K., Appella, D. H. & Stupp, S. I. Enhanced oligonucleotide binding to self-assembled nanofibers. Bioconjug. Chem. 16, 501–503 (2005). 23. Böhler, C., Nielsen, P. E. & Orgel, L. E. Template switching between PNA and RNA oligonucleotides. Nature 376, 578–581 (1995). 24. Uhlmann, E., Peyman, A., Breipohl, G. & Will, D. W. PNA: synthetic polyamide nucleic acids with unusual binding properties. Angew. Chem. Int. Ed. 37, 2796–2823 (1998). 25. Davis, J. T. & Spada, G. P. Supramolecular architectures generated by self-assembly of guanosine derivatives. Chem. Soc. Rev. 36, 296–313 (2007). 26. Menchise, V. et al. Insights into peptide nucleic acid (PNA) structural features: the crystal structure of a D-lysine-based chiral PNA–DNA duplex. Proc. Natl Acad. Sci. USA 100, 12021–12026 (2003). 27. Watson, J. D. & Crick, F. H. C. Molecular structure of nucleic acids: a structure for deoxyribose nucleic acid. Nature 171, 737–738 (1953). 28. Trushko, A., Schäffer, E. & Howard, J. The growth speed of microtubules with XMAP215-coated beads coupled to their ends is increased by tensile force. Proc. Natl Acad. Sci. USA 110, 14670–14675 (2013). 29. Ikezoe, Y., Washino, G., Uemura, T., Kitagawa, S. & Matsui, H. Autonomous motors of a metal–organic framework powered by reorganization of selfassembled peptides at interfaces. Nature Mater. 11, 1081–1085 (2012). 30. Demchenko, A. P. The red-edge effect: 30 years of exploration. Luminesence 17, 19–42 (2002). 31. Cushing, S. K., Li, M., Huang, F. & Wu, N. Origin of strong excitation wavelength dependent fluoresence of graphene oxide. ACS Nano 8, 1002–1013 (2014). 32. Muccini, M. A bright future for organic field-effect transistors. Nature Mater. 5, 605–613 (2006).
8
DOI: 10.1038/NNANO.2015.27
33. Incardona, M. F. et al. EDNA: a framework for plugin-based applications applied to X-ray experiment online data analysis. J. Synchrotron Radiat. 16, 872–879 (2009). 34. Spek, A. L. Structure validation in chemical crystallography. Acta Crystallogr. D 65, 148–155 (2009).
Acknowledgements This work was supported in part by grants from the Israeli National Nanotechnology Initiative and Helmsley Charitable Trust for a focal technology area on Nanomedicine for Personalized Theranostics. The authors thank members of the Gazit Laboratory for helpful discussions, Y. Salitra for help with PNA synthesis and O. Yaniv for advice on crystallization experiments. The authors acknowledge the ESRF for synchrotron beam time and the staff scientists of the ID29 beamline for their assistance. O. Berger is supported by a fellowship from the Argentinean Friends of Tel Aviv University Association.
Author contributions O.B., L.A-A., L.B., Y.E. and E.G. designed the study. O.B., M.L-S., Y.L-P. and M.B. performed the experiments. O.B., A.G., T.S. and Y.E. analysed the data. F.F., L.J.W.S., E.M. and T.F. performed and analysed the X-ray diffraction experiments. F.P. performed OFET experiments. O.B., L.A-A. and E.G. prepared the manuscript.
Additional information Supplementary information is available in the online version of the paper. Reprints and permissions information is available online at www.nature.com/reprints. Correspondence and requests for materials should be addressed to E.G.
Competing financial interests
The authors declare no competing financial interests.
NATURE NANOTECHNOLOGY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturenanotechnology
© 2015 Macmillan Publishers Limited. All rights reserved
