Supramolecular multiplexes from collagen mimetic peptide-PNA(GGG)₃ conjugates and C-rich DNA: pH-induced reversible switching from triplex–duplex to triplex-i-motif

Abstract

Peptides are well known for forming nanoparticles, while DNA duplexes, triplexes and tetraplexes create rigid nanostructures. Accordingly, the covalent conjugation of peptides to DNA/RNA produces hybrid self-assembling features and may lead to interesting nano-assemblies distinct from those of their individual components. Herein, we report the preparation of a collagen mimetic peptide incorporating lysine in its backbone, with alkylamino side chains radially conjugated with G-rich PNA [collagen-(PNA-GGG)₃]. In the presence of complementary C-rich DNA (dCCCTTTCCC) at neutral pH, the collagen mimetic triplexes were interconnected by PNA-GGG : DNA-CCC duplexes, leading to the formation of larger assemblies of nanostructures. Upon decreasing the pH to 4.5, the dissociation of the triplex–duplex assembly released the protonated C-rich DNA, which immediately folded into an i-motif. With an increase in the pH to 7.2 (neutral), the i-motif unfolded into linear DNA, which reformed the PNA-GGG : DNA-CCC duplex interconnecting the collagen triplexes. The pH-induced switching of the assembly and disassembly was reversible over a few cycles. The hybrid collagen-(PNAGGG)₃ : DNA-C₃T₃C₃ triplex–duplex and the individual components of the assembly including the i-motif were characterized by UV and CD melting, fluorescence, TEM and gel electrophoresis. The pH-induced reversible switching was established by the changes in the CD and fluorescence properties. Peptide-DNA conjugates have wide applications in both biology and materials science, ranging from therapeutics and drug delivery to diagnostics and molecular switches. Thus, the prototype ensemble of the triplex peptide-PNA conjugate and its duplex with DNA described herein has potential for elaboration into rationally designed systems by varying the PNA/DNA sequences to trap functional ligands/drugs for release in pH-controlled environments.

---

Introduction

The biopolymers nucleic acids and proteins are very well known to organize into assembled motifs induced by their unique chemical structures, which are responsible for their many biological functions. Although the assembly of individual biomolecules is reasonably well understood, the morphological and functional characteristics of their intermolecular hybrid self-assembly are less explored. In biological systems, proteins and nucleic acids co-assemble through both covalent and non-covalent interactions into distinct complexes, enabling their unique biological functions. Covalent conjugates between nucleic acids and proteins have been found to carry out several intricate functions, particularly in viruses.¹ Some examples of naturally occurring non-covalent supramolecular assemblies of proteins and nucleic acids include the infectious tobacco mosaic virus, in which the coat proteins self-assemble on the RNA template² the telomerase protein, which co-assembles with the RNA sequence (telomerase RNA) stabilizes the chromosomal ends;³ transcription factors involved in transcribing DNA into RNA; ribosomes that bind and read mRNA to decode genetic information into proteins; and chromatin, in which large DNA is condensed into a highly compact form to fit into confined spaces in cells.⁴

The development of solid-phase peptide and oligonucleotide synthesis with compatible coupling chemistry on the same resin has made it possible to synthesize peptide-DNA covalent conjugates.^5–7 These peptide-DNA conjugates can form filamentous-like superstructures that can be controlled by tuning the charge density on the superstructure.⁸ Genetically encoded staple proteins can be used to control the folding of double-stranded DNA into well-defined nanostructures. This helps to advance our understanding of the genomic architecture and modulation of the functions of genomic DNA.⁹ The intrinsic folding property of DNA into ordered secondary structures has been harnessed to bring protein bundles in close proximity for studying the functional role of helix-helix interactions generated by the proteins.^10–12 The triple helices of cationic-charged collagen mimetic peptides (CMPs) have been electrostatically linked to the DNA tiles generated by origami nanostructures.¹³ In these examples, the protein-DNA conjugates have functions distinct from that of their individual molecules.^8,10–13 Owing to the attributes of rational control and design, peptide-DNA nanostructures are finding emerging applications ranging from nanotechnology and biology to the nanomedicine field.^14–17

Short collagen peptides are comprised of repeating [X-Y-Gly] triad motifs, with proline (Pro) and 4R-hydroxyproline (Hyp) at the X and Y positions, respectively. Each strand of [X-Y-Gly]_n forms a left-handed polyproline PPII helix (extended 3.1 Å rise per residue with all trans peptide bonds) compared to the right-handed PPI helix (compact 1.9 Å rise with all cis amide bonds). The three left-handed PPII strands intertwine on a common axis to form a right-handed triple-helix super structure akin to that of natural collagen protein.¹⁸ In collagen mimetic peptides (CMP), one of the triad prolines is replaced with another natural or modified amino acid (4-exo substituted prolines: 4F, 4-NH₂, and 4-N₃)¹⁹ and this change may retain or modulate their innate ability to form higher-order triplex structures. The collagen mimetic peptides (CMP) have found applications in tissue engineering, cell encapsulation, drug delivery, bone regeneration and many other applications.^20–24 Recently, the Hyp at the Y-site of collagen peptide was replaced with lysine (Lys) and its Cα-propyl sidechain amino function was used to functionalize the peptide with a bipyridyl moiety.²⁵ Metal ions can bridge two bipyridyl moieties on adjacent chains and the resultant π–π stacking of bipyridyls in these interconnected chains favor the formation of stable triplexes.²⁵ This idea prompted us to consider the conjugation of Lys-CMP²⁵ at the lys-Cα-sidechain with peptide nucleic acid (PNA).²⁶ The natural stacking of bases in PNA chains can induce further secondary structures in CMP through self-assembly and form PNA : DNA duplexes with complementary DNA, in both standard²⁶ and modified PNAs.²⁷ Recently, PNAs containing amino acid side chain at Cα/Cγ termed bilingual twin coded molecules with both protein and nucleic acid languages were shown to have interesting assembling properties.²⁸ Additionally, peptide-PNA conjugates have been employed in antiviral, antibacterial, and drug-delivery applications.²⁹

Most of the reported peptide-DNA conjugated systems are comprised of either stable DNA duplexes or higher ordered structures such as triplexes and G-quadruplexes.^10–14 The C-rich DNA in an acidic environment (<pH 5.0) forms protonated C:C⁺ base pairs, which fold into a self-assembled tetrameric secondary structure called i-motif.³⁰ When the pH increases from acidic (4.5) to the physiological range (7.0), the folded i-motif unfolds into the open-stranded form, which can hybridize with the G-rich DNA in solution to form a DNA duplex.³¹ This reversible pH-switchable property of DNA secondary structures from folded (i-motif) to unfolded form (duplex) has not been exploited to date in tuning supramolecular assemblies of peptide-DNA conjugates. We surmised that the collagen mimetic peptide Lys(CMP) conjugated with PNA-(GGG)₃ [collagen-(PNA-GGG)₃, Fig. 1A] would form a collagen-like triplex at pH 7.2 (Fig. 1B). This triplex would have nine PNA-(GGG) chains (3 on each single strand of collagen peptide) projecting in a radial direction from the surface of the CMP triplex. In the presence of C-rich DNA (dC₃T₃C₃), the CCC sequences at both ends form a PNA : DNA hybrid with nine PNA-(GGG) sidechains on the collagen triplex, interconnecting two or more triplexes, which leads to the supramolecular assembly of the collagen triplex-PNA : DNA duplex (Fig. 1C). At acidic pH of <5.0, the interlinking PNA : DNA duplex would dissociate, liberating the C-rich DNA, which at acidic pH folds into a stable i-motif structure (Fig. 1B). Upon reverting to neutral pH 7.2, the i-motif DNA (C₃T₃C₃) again unfolds into a linear form, rehybridizing with peptide triplex-(PNA-GGG)₃ to regenerate the supramolecular assembly of collagen-PNA : DNA duplexes. The reversible association-dissociation process could be repeated over many cycles, making pH a switch to interconvert molecular and supramolecular assemblies.

Fig. 1 Schematic representation of collagen-(PNA-GGG)₃ triplex and pH-dependent reversible switching of collagen-(PNA-GGG)₃ : DNA self-assembly. (A) Collagen-(PNA-GGG)₃ peptide (red strands) displaying G-rich PNA (blue). (B) Collagen-(PNA-GGG)₃ triplex, triplex–duplex conjugate with C-rich DNA (green) at pH 7.2 and pH-dependent reversible switching from triplex–duplex (pH 7.2) to triplex/C⁺:C i-motif (pH 4.5). (C) Enlarged structure of collagen-(PNA-GGG)₃ triplex interconnected by DNA duplex at pH 7.2 and i-motif at pH 4.5. Duplexes may be formed in both parallel and antiparallel orientations.

Results and discussion

Synthesis of collagen-PNA-conjugated peptides

A convenient feature for the synthesis of peptide-PNA conjugates is the fact the solid phase (resin) and chemistry of peptide synthesis is compatible with that of PNA synthesis. The synthesis of a fully protected N-terminus-capped 27-mer collagen mimetic peptide (CMP-1) having lysine residues at the 8, 14 and 20 positions was performed by manual solid-phase peptide synthesis (SPPS) on rink amide resin (Scheme 1). The standard Fmoc protocol was employed with the sequential addition of the commercially available monomers Fmoc-Pro-OH (1), Fmoc-4-Hyp(O^tBu)-OH (2) or Fmoc-Lys(Mtt)-OH (3), and Fmoc-Gly-OH (4) as desired for the target peptide sequence. After the synthetic assembly of the peptide sequence, the N-terminus was capped as an acetyl derivative with acetic anhydride : pyridine (1 : 3) to obtain the resin-bound collagen peptide CMP-1. The lysine side chain protecting group 4-methyltrityl (Mtt) was selectively deprotected using 1.8% TFA in DCM³² to yield resin-bound CMP-2 (Scheme 1), which retained the O⁴–Bu^t protection in the Hyp residues. The free ε-amino groups on the lysine side chains of the resin-bound CMP-2 were used for the parallel stepwise synthesis of three PNA(GGG) chains, with one each at the 8, 14 and 20 sites by sequential coupling with the commercially available PNA-G monomer 5 [Fmoc-PNA(Bhoc)G-OH] (Scheme 1). This led to radial functionalization of CMP-2 with PNA(GGG) sidechains at the three lysine positions (8, 14 and 20). Subsequent removal of the Fmoc protecting group from the PNA N-terminus and cleavage of the collagen peptide-(PNA-GGG)₃ conjugate from the resin using TFA : H₂O : TIPS (95% : 2.5% : 2.5%) yielded the crude collagen-(PNA-GGG)₃ peptide-3 (CMP-3) (Scheme 1). Before the end-capping of the fully protected CMP-1, as shown in Scheme 1, NHFmoc-phenylalanine was coupled at the N-terminus to aid the determination of the concentration of the collagen peptides. The control collagen-lysine peptide CMP-4 (Scheme 2) was obtained from CMP-1 by first removing Fmoc, and then N-acylation, as stated above, to end-cap the peptide at the N-terminus, followed by cleavage from the solid support. The collagen-(PNA-GGG)₃ conjugate peptide (CMP-3) and the lysine collagen peptide (CMP-4) were purified by HPLC and characterized by MALDI-TOF (Fig. S1, ESI†) (Table 1).

Scheme 1 Synthesis and chemical structures of collagen-PNA(GGG)₃ peptide-3 (CMP-3) and intermediate resin-bound collagen mimetic peptides: CMP-1 and CMP-2.

Scheme 2 Synthesis and chemical structure of control collagen mimetic peptide CMP-4.

Table 1 HPLC retention time and MALDI-TOF data of collagen-(PNA-GGG)₃ (CMP-3) and CMP-4

Peptide	Ret. time	Calcd. mass	Obs. mass
Collagen-PNA-(GGG)3 CMP-3	12.7	C212H290N94O61 [M^+] = 5128.2480	5128.4866
		C212H290N94O61Na [M + Na]^+ = 5151.2378	5151.6318
CMP-4	14.1	C122H182N32O35 [M^+] = 2655.3445	2655.9037
		C122H182N32O35Na [M + Na]^+ = 2655.3445	2678.8124

---

Conformational study of collagen-(PNA-GGG)₃ peptide CMP-3

The formation of the left-handed polyproline II helix by the collagen mimetic peptides was evidenced by the characteristic CD signals, with a weak positive band at around 220–227 nm, followed by a strong negative band in the range of 205–210 nm.^33,34 The CMP-3 peptide as a positive peptide exhibited two CD bands, one with a positive maximum at 225 nm and another with a negative maximum at around 201 nm in the concentration range of 1 μM to 30 μM (Fig. 2A). This indicates that the PNA-conjugated peptide CMP-3 adopted a similar polyproline PPII structure. It has been demonstrated that value of R_p/n (the ratio of the intensity of the positive band at 225 nm to the negative CD minimum at 208 nm) in the range of >0.1 is a good indication of collagen triple-helix formation.³⁴ In the case of CMP-3, the R_p/n value enhanced with an increase in concentration, with greater association of the peptide strands and saturation at a R_p/n value of 0.1 at a concentration of peptide of 15 μM (Fig. 2B), which is its critical triple-helix concentration (CTC).³⁴ Since beyond this concentration, the peptide exists in triplex form, all further experiments were conducted at a peptide concentration of 20 μM. The sequence of negative lower wavelength band and the positive higher wavelength CD bands corresponded to a right-handed triple helix.^33,34

Fig. 2 (A) Concentration-dependent CD spectra of collagen-(PNA-GGG)₃ peptide CMP-3 in the range of 1–30 μM. (B) R_p/nvs. concentration plot extracted from data in A. (C) Comparative CD spectra of collagen-(PNA-GGG)₃ peptide (CMP-3) and collagen mimetic peptide CMP-4 at 20 μM. Buffer 10 mM Tris, 10 mM NaCl at pH 7.2.

The triplex-forming ability of the radially linked PNA-GGG chains to the collagen mimetic backbone at Cα-lysine in the conjugate (CMP-3) was examined by CD studies. The nine PNA chains (3 on each strand) arranged radially outwards of the collagen chain enhanced the PPII conformation (higher R_p/n) for the CMP-3 peptide compared to the unconjugated lysine collagen peptide CMP-4, which showed a negligible PPII conformation (lower R_p/n) (Fig. 2C). The lower propensity for triplex formation by lysine collagen peptide CMP-4 is a consequence of the replacement of the neutral Hyp with cationic lysine.³⁵ The nine positive charges in close proximity on the triplex surface induced electrostatic charge repulsion, lowering the propensity for triplex formation. In contrast, in collagen-(PNA-GGG)₃ peptide CMP-3, the G-rich PNA-conjugated to the lysine side chain enhanced its triple-helical stability, given that the nine N-terminal positive charges on PNA chain are located away from the helix surface. Similar to the π–π stacks in bipyridyl-collagen peptides,²⁵ the base stacking of PNA nucleobases with possible additional interchain H-bonding may enhance the stability of the triple helix. The CD data also indicated the absence of the formation of a strong PNA–PNA duplex helical structure.

CD thermal melting (T_m) study

Upon heating, the triplex from the collagen mimetic peptides showed a transition from triple-helix to random-coil conformation.³⁶ The effects of modifications on the backbone³⁷ and C4-substituents (F, NH₂, and N₃) at C4 on the proline ring at the Y position¹⁹ and radial functionalization by bipyridyl²⁵ or sugar moieties³⁸ on the stability of the collagen triple-helix have been well studied by monitoring the change in ellipticity at 225 nm in CD spectra as a function of temperature.

The temperature-dependent CD spectra of the collagen-(PNA-GGG)₃ conjugate CMP-3 and collagen peptide CMP-4 in aqueous buffer (pH 7.2) in the temperature range of 5 °C–90 °C are shown in Fig. 3. Beyond 35 °C, a continuous decrease in ellipticity at 225 nm was seen for the CMP-3 conjugate with the appearance of an isodichroic point near 212 nm (Fig. 3A). This suggests two-state equilibrium in the melting of the triple-helix to a random coil.^19b The plot of the ellipticity change at 225 nm versus temperature showed a smooth negative sigmoidal transition for the CMP-3 conjugate expected for co-operative melting of the triplex to random coil (Fig. 3C). The thermal melting (T_m) value obtained from the sharp minimum of the derivative plot (Fig. 3D) for the collagen-(PNA-GGG)₃ conjugate CMP-3 was 45 °C. In comparison, the lysine collagen peptide CMP-4 lacked any CD bands at around 220 nm and no changes were observed in its CD spectra (Fig. 3B) upon heating, indicating the lack of formation of a peptide triplex (Fig. 3D). The weak transition seen at around 65 °C may arise from the self-melting of the liberated single-strand peptide-PNA conjugate. The placement of three lysines in place of 3× Hyp in the collagen peptide destabilized the derived triplex more than that of the single lysine-incorporated peptide.³⁵ The CD pattern of the lys-collagen peptide conjugated with PNA-(GGG)₃ (CMP-3) indicated a right-handed triplex melting at 45 °C, with no formation of a strong PNA–PNA duplex helical structure.

Fig. 3 Variable temperature CD spectra of (A) CMP-3 and (B) CMP-4 and temperature-dependent ellipticity changes at 225 nm for (C) CMP-3 and CMP-4. (D) First derivative plot of data in (C). Peptide concentration 20 μM. Buffer conditions 10 mM Tris, 10 mM NaCl at pH 7.2.

Effect of salt (Na⁺, K⁺) on collagen-(PNA-GGG)₃ peptide CMP-3

PNA is a true mimic of DNA given that it can hybridize with complementary DNA/RNA through Watson–Crick H-bonding to form a PNA : DNA duplex^26,27 and via Hoogsteen H-bonding to form triplex and tetraplex structures.^39,40 PNAs containing G-rich sequences follow the trait of DNA in forming a tetrameric G-quadruplex in the presence of salt (Na⁺/K⁺), as evidenced by the CD spectra having a positive band at a wavelength of 290 nm.³⁹ The absence of any band at higher wavelengths beyond 240 nm in the CD spectra of collagen-PNA conjugate CMP-3 (Fig. 3A), even in the presence of Na⁺ (100 mM) and K⁺ (200 mM) salt suggests absence of G-quadruplex from the PNA-(GGG)₃ side chains (Fig. S2, ESI†). The salt (Na⁺/K⁺) also had no effect on the stability of the peptide triplex from collagen-(PNA-GGG)₃ peptide CMP-3, which remained unaltered with a T_m of 45 °C (Fig. S2, ESI†).

Interaction of self-assembled collagen-(PNA-GGG)₃ peptide CMP-3 with C-rich DNA

The triplex derived from collagen-(PNA-GGG)₃ conjugate CMP-3 is endowed with nine PNA-GGG appendages, which radially project out of the triplex bundle (Fig. 1B). The PNA-(GGG)₃ sidechains were hybridized with C-rich complementary DNA 1 (5′-CCCTTTCCC-3′) to yield PNA : DNA duplexes, which interconnect the triplex bundles (Fig. 1B). At neutral pH 7.2, pure DNA 1 showed CD spectra with a positive band at around 278 nm crossing over at 254 nm to a broad negative CD band at around 250–210 nm (Fig. 4A, blue), which is typical of random-coil DNA.⁴¹ When collagen-(PNA-GGG)₃ peptide CMP-3 was mixed with DNA 1 in a 1 : 1 ratio, the positive CD band at 278 nm was blue-shifted by 6 nm to 272 nm and exhibited a negative CD band at 236 nm (Fig. 4A, red). This CD pattern is characteristic of PNA : DNA duplexes, which showed a moderate intensity positive CD band at a higher wavelength (260–270 nm), followed by a mild negative band at a lower wavelength (240 nm).²⁷ The positive band at 225 nm in the CD spectra corresponds to the collagen triplex component of CMP-3 : DNA 1 duplex, similar to that seen for peptide CMP-3 alone (Fig. 4A, black). When the relative stoichiometric ratios of CMP-3 and DNA 1 were changed to 1 : 0.33 and 1 : 0.66, only the intensities of the bands changed due to the difference in the concentration of DNA 1 (Fig. S3, ESI†) without any wavelength shifts. This indicates the progressive formation of PNA : DNA duplexes with an increase in the concentration of DNA 1. Upon the addition of non-complementary DNA T₉ (5′-TTT TTT TTT-3′) to collagen-(PNA-GGG)₃ CMP-3, no changes in the CD spectra of the peptide CMP-3 was observed (Fig. S4, ESI†), suggesting non-formation of the CMP-3 : DNA 2 duplex. This control further supports the sequence specificity in establishing the CMP-3 : DNA 1 duplex. The retention of the CD band at 225 nm in the CMP-3 collagen-(PNA-GGG)₃ : DNA 1 duplex confirmed the simultaneous presence of both the triplex from the peptide component CMP-3 and the duplex from the conjugated (PNA-GGG)₃ with DNA 1. Thus, collagen-(PNA-GGG)₃ CMP-3 : DNA 1 is a supramolecular multiplex of collagen triplexes interconnected with different possible PNA : DNA duplex combinations, resulting in polyplexes. To examine the formation of any possible CMP-PNA₂ : DNA 1 or CMP-PNA : (DNA 1)₂ triplexes at pH 4.5, CD spectra were recorded for the control samples using single-stranded CMP-3 at concentrations (5 μM) much lower than the critical triple-helix concentration (15 μM) in the presence of DNA 1. The CD spectra were dominated by bands at around 286 nm, typical of the i-motif, rather than PNA : DNA triplexes (CD bands at 275–280 nm). Even with PNA(GGG)₃ : DNA 1 at pH 4.5, DNA 1 occurs mostly in the i-motif form rather than as a PNA : DNA triplex or duplex (ESI, S3a–b†). It should be pointed out that the DNA 1 sequence with two continuous stretches of 4xC's is more favoured to form the i-motif than a triplex at acidic pH 4.5.

Fig. 4 (A) CD spectra of CMP-3 triplex and CMP-3 : DNA 1 triplex–duplex at pH 7.2. (B) CD spectra of CMP-3 triplex and CMP-3 : DNA 1 triplex-i-motif at pH 4.5.

Ethidium bromide (EB) displacement assay

The formation of the collagen-(PNA-GGG)₃ : DNA 1 duplex was substantiated by the ethidium bromide (EB) displacement assay.⁴² EB intercalates into the DNA : DNA duplex, leading to fluorescence from the bound EB. In contrast, EB does not intercalate with the PNA : DNA duplex. DNA 1 (CCCTTTCCC) was complexed with its perfect complementary DNA 2 (GGGAAAGGG) to generate a DNA 1 : DNA 2 duplex, which was treated with EB. It intercalated into the duplex, leading to a fluorescence emission at 600 nm (Fig. 5, black). When the EB-DNA duplex was mixed with collagen-(PNA-GGG)₃ peptide CMP-3 in relative equivalents of 0.33, 0.66 and 1.0, a decrease in EB fluorescence emission at 600 nm was seen with the complete loss of fluorescence at a 1 : 1 ratio (Fig. 5, red, blue and magenta, respectively). The yield of collagen-PNA : DNA hybridization was calculated from the loss of EB fluorescence intensity and it was found to be 8.2% (0.33 eq.), 75% (0.66 eq.) and 94% (1 eq.) for stochiometric additions of collagen-(PNA-GGG)₃ peptide (CMP-3). Due to the higher affinity of PNA to DNA, it is known to invade the DNA : DNA duplex, displacing the iso-sequential DNA strand to form a PNA : DNA duplex, and cationic PNA is even more efficient in this invasion.⁴² The PNA-(GGG)₃ component of CMP-3 displaced the DNA 2 strand from the DNA 1 : DNA 2 duplex to form the CMP-3 : DNA 1 duplex with higher stability, releasing the intercalated EB and leading to the loss of its fluorescence. This supports the formation of the CMP 3-PNA : DNA 1 duplex assembly.

Fig. 5 EB displacement fluorescence assay at λ_ex 310 nm and λ_em 600 nm. Buffer conditions: 10 mM Tris, 10 mM NaCl at pH 7.2.

CD thermal melting study of CMP-3 peptide triplex and CMP-3 : DNA 1 (triplex–duplex)

As shown in the preceding sections, the collagen-(PNA-GGG)₃ peptide (CMP-3) triplex melts at 45 °C, while its complex with DNA 1 (CMP-3 : DNA 1) showed slight destabilization of the peptide triple helix with a T_m of 40 °C, (monitored at 225 nm, pH 7.2). The T_m of the peptide triplex of CMP-3 : DNA 1 with the two components mixed at different relative ratios of 1 : 0.33, 1 : 0.66 and 1 : 1 showed progressive destabilization with values of 40 °C, 30 °C and 25 °C, respectively (Table 2 and Fig. 6). This suggests that the formation of the PNA : DNA interconnecting duplex lowers the stability of the triplex component CMP-3.

Fig. 6 CD melting curve for collagen-(PNA-GGG)₃ : DNA 1 complexes in 1 : 0.3, 1 : 0.6 and 1 : 1 ratios. (A) Normalized ellipticity curve at 260 nm and (B) first derivative plots. Buffer conditions: 10 mM Tris, 10 mM NaCl at pH 7.2.

Table 2 Triplex and duplex UV-T_m (°C) in CMP-3 and CMP-3 : DNA 1 complex at different pHs^a

		CMP-3		CMP-3 : DNA 1
	nm	225 (t)		225 (t)	260 (d)	280 (i)
Ratio	pH^a	7.2	4.5	7.2		4.5
a t = triplex, d = duplex, i = i-motif; buffer contains 10 mM NaCl.
1 : 0		45	40	45	—	—
1 : 0.33		—	—	40	30	30–35
1 : 0.66		—	—	35	30	35
1 : 1		—	—	35	30–35	25
0 : 1		—	—	—	—	30

---

The melting of the PNA : DNA duplexes is generally analyzed by monitoring their UV absorbance at 260 nm, which shows a positive sigmoidal transition with the maximum in the derivative curve representing the T_m for the PNA : DNA duplex.^26,27,44 The temperature-dependent CD spectra of the collagen-(PNA-GGG)₃ : DNA 1 complex monitored at 260 nm (pH 7.2) showed an inverse sigmoidal transition in the ellipticity vs. temperature plot (Fig. 6) arising from a decrease in ellipticity with an increase in temperature.

Peptide triplex and DNA i-motif formation at pH 4.5

The conformational change in C-rich DNA from random coil (pH 7.2) to ordered i-motif secondary structure (pH < 5.5) could be detected in the CD spectra, showing the i-motif characteristic positive band around 290 nm.^41,43 The cytosine-rich single-strand DNA 1 (5′-CCCTTTCCC-3′) can first fold into a hairpin at acidic pH through C:C⁺ base pairing, and then to a bimolecular i-motif structure. At acidic pH 4.5, DNA 1 displayed a strong positive CD band at 289 nm, accompanied by a lower intensity negative band at 264 nm with cross-over at 272 nm (Fig. 4B, blue), showing a T_m of 30 °C (Table 2 and Fig. S6, ESI†). Compared to its CD profile at neutral pH 7.2 (Fig. 4A, blue), the positive band is red-shifted by 11 nm with higher intensity (Fig. 4B, red), indicating the i-motif form for DNA 1 at pH 4.5. The addition of a stoichiometric equivalent of collagen-(PNA-GGG)₃ peptide CMP-3 to DNA 1 at pH 4.5 neither affected the i-motif structure, as seen by the unchanged CD signal at 289 nm (Fig. 4B, red), nor the peptide triplex structure of collagen-(PNA-GGG)₃ conjugate CMP-3, as observed by the band at 225 nm (Fig. 4B, black). The lower intensity CD bands suggest a lower propensity for triplex formation by CMP-3 at acidic pH.

CD melting (T_m) of collagen-(PNA-GGG)₃ peptide CMP-3 triplex and CMP-3 : DNA 1 complex at pH 4.5

The collagen-(PNA-GGG)₃ peptide CMP-3 at pH 4.5 showed a triplex melting T_m of 40 °C, as monitored at 225 nm, destabilized by 5 °C compared to the T_m of 45 °C at pH 7.2 (Table 2). The CD melting of collagen-(PNA-GGG)₃ CMP-3 : DNA 1 hybrid at different relative compositions (DNA 1: 0.33, 0.66 and 1.0) at pH 4.5 was followed at two different wavelengths of 225 nm for the triplex (collagen peptide) and 289 nm for the i-motif (DNA 1). The triplex component of collagen-(PNA-GGG)₃ peptide CMP-3 in the complex showed destabilization (T_m = 40 °C, 35 °C and 25 °C) with an increase in the concentration of DNA 1 with respect to CMP-3 alone (Table 2 and ESI, S7†). In comparison, the T_m of i-motif DNA 1 remained the same except at a 1 : 1 stoichiometry, where a slight destabilization of 5 °C was observed (Table 2 and ESI, S8†). This suggested the non-complexation of i-motif DNA 1 with collagen-(PNA-GGG) peptide CMP-3 at pH 4.5.

Collagen-(PNA-GGG)₃ CMP-3 : DNA 1 self-assembly by gel electrophoresis

The formation of the collagen-(PNA-GGG)₃ CMP-3 : DNA 1 hybrid was also examined by native polyacrylamide gel electrophoresis (PAGE). The collagen-(PNA-GGG)₃ peptide CMP-3 is positively charged, and hence not expected to move in PAGE, but upon complexation with polyanionic DNA 1, the complex acquires an overall negative charge, and hence expected to migrate on the gel. Fig. 7 shows the native PAGE (non-denaturing) of collagen-PNA peptide CMP-3, together with the added complementary DNA 1 in different stoichiometric ratios at pH 7.2. The gel was stained with EB to visualize the dsDNA duplex (DNA 1 : DNA 2) (Fig. 7A, lane 5) and no bands were visualized in lanes 1–4 containing PNA : DNA 1 duplex, given that EB does not intercalate into the PNA : DNA duplex. After complete de-staining of EB by several washings with water, the same gel was stained with Coomassie brilliant blue dye to visualize the peptides⁴⁵ (Fig. 7B). The positively charged collagen-(PNA-GGG)₃ peptide CMP-3 in lane 1 failed to enter the gel. In contrast, lanes 2–4 exhibited intense blue bands with slower mobility compared to dsDNA (EB stain) in the gel (Fig. 7A). The increase in the intensity of the bands with an increase in the concentration of DNA 1 (0.33, 0.66 and 1 equivalent) progressively focused to a nice spot (lane 4), indicating the migration of peptide CMP-3 due to its complexation with anionic DNA 1 to yield the triplex–duplex complex (Fig. 7B), which reduced the overall positive charge. The poorly resolved smears seen in lanes 2 to 4 arise from the heterogeneous sizes of the CMP-3 : DNA 1 complexes evolving prior to the fully assembled triplex–duplex structures. The stationary bands on top of the gel in the wells correspond to the larger assembled structures, which did not enter the gel due to their high molecular weight and charged uncomplexed CMP-3. Thus, the PAGE results further support the formation of an assembled network of collagen peptide triplexes interconnected by PNA : DNA duplexes.

Fig. 7 PAGE (A) staining with EB and (B) staining with Coomassie brilliant blue. Lane 1 : collagen-(PNA-GGG)₃ (CMP-3), lanes 2–4 : collagen-(PNA-GGG)₃ (CMP-3) : DNA 1 in 1 : 0.33, 1 : 0.6 and 1 : 1 ratios, respectively, lane 5: double-stranded DNA 1:DNA 2 (20 μM). Peptide concentration in lanes 1–4 is 100 μM with equivalent DNA 1 concentration in respective ratios. Buffer: 40 mM Tris-acetate (pH = 7.2) with EDTA (1 mM).

The native PAGE in Tris buffer at pH 4.5 of DNA 1 alone and collagen-(PNA-GGG)₃ peptide CMP-3 with increasing amounts of DNA 1 showed faint smeared bands in only in lane 2, having excess CMP 3 and none of the other lanes showed blue staining (Fig. S9, ESI†). This indicates that DNA 1 was not complexed with peptide CMP-3 at pH 4.5.

Morphology of collagen-(PNA-GGG)₃ : DNA 1 assembly

Transmission electron microscopy (TEM) and FE-SEM are useful techniques to characterize the morphology of collagen-assembled structures in relation to the variation in their secondary structures.⁴⁶ In the TEM image, the collagen-(PNA-GGG)₃ conjugated peptide CMP-3 alone showed well-separated spherical grain particles with a size of 20–40 nm (Fig. 8A). This type of isolated spherical nanostructures was observed earlier for [Pro-Hyp-Gly]₆ collagen triplex peptides,⁴⁷ while DNA 1 did not show any regular morphology due to its unstructured single-strand form (Fig. 8B). Dark aggregates for DNA 1 at both pH (Fig. 8B and E) were seen, which may arise from the interaction of the uranyl acetate and phosphate ions with the DNA backbone. When peptide CMP-3 was mixed with DNA 1 in a 1 : 1 ratio at pH 7.2, larger linear networked structures with a length of ≥2–3 μm and width of ≥1 μm were observed (Fig. 8C). The enlarged region of the TEM image (Fig. 8C) shows that the particle assemblies are held together by short nanorods (as indicated by the yellow arrows in the zoom of image C) in the form of helical connects (Fig. 8C – zoom). This arrangement of interconnected particle assemblies is repeated with an overall helical twist over the larger lengths in μm dimensions (more images are provided in ESI, Fig. S10†). This is similar to the reported collagen-PNA conjugates upon the addition of DNA, leading to distinct nanostructures as in bipyridyl collagen, which showed curved disks alone and hollow spheres upon the addition of metal.²⁵ The TEM images of the peptide collagen-(PNA-GGG)₃ CMP-3 : DNA 1 complex suggested that the triplex regions from each CMP-3 are interconnected by short nanorods, corresponding to the sidechain PNA(GGG)₃ : DNA 1 duplex, with the formation of multiplexes from several triplex–duplex connects.

Fig. 8 TEM images: upper panel pH 7.2; (A) collagen-(PNA-GGG)₃ peptide (CMP-3) alone, (B) DNA 1 alone and (C) collagen-(PNA-GGG)₃ peptide (CMP-3) + DNA 1 self-assembly in a 1 : 1 ratio. Lower panel pH 4.5 (D) collagen-(PNA-GGG)₃ peptide (CMP-3) alone; (E) DNA 1 at pH 4.5 alone and (F) CMP-3 + DNA 1 self-assembly in a 1 : 1 ratio. Peptide concentration of 20 μM, DNA 1 concentration of 30 μM. Buffer conditions: 10 mM Tris, 10 mM NaCl at pH 7.2 or pH 4.5.

At pH 4.5, CMP-3 alone showed granular assemblies similar in size to that at pH 7.2 (Fig. 8D). Free DNA 1 (i-motif) did not show any assembled structures (Fig. 8E). The TEM images showed no predominant nanostructures for collagen-(PNA-GGG)₃ CMP-3 : DNA 1 upon mixing in a 1 : 1 ratio at pH 4.5 (Fig. 8F), in contrast to that at pH 7.2. The few assembled nanoparticles with a length of ≤100 nm seen at pH 4.5 are small and different in comparison with the nanostructures characterized at neutral pH 7.2. This indicates the absence of interconnects due to no formation of PNA : DNA duplex when DNA 1 is mixed with collagen-(PNA-GGG)₃ (CMP-3) at pH 4.5, and hence no macro polyplex assemblies.

pH-dependent nanostructure switching

To demonstrate the reversible assembly–disassembly of the macro assemblies, we examined the effects of a change in pH on the polyplex assemblies. The CMP-3 : DNA 1 hybrid at pH 7.2 exhibited large networked nanostructures of several micrometers in length and width (Fig. 9A). When the pH was reduced to 4.5, the assembled network structure disappeared. Upon reversing the pH to 7.2, macro assemblies varying in length and dimensions reappeared (Fig. 9C), similar to that in Fig. 9A. A similar situation was seen in the subsequent 2^nd cycle of pH changes to acidic 4.5 pH (Fig. 9D) and neutral 7.2 (Fig. 9E). To examine the role of the collagen triplex assembly in inducing the formation of the supramolecular network through the PNA : DNA hybrid, TEM images were recorded for CMP-3 at a concentration of 5 μM, where it existed as a single strand. It was found that no supramolecular network structures were present in the TEM images upon the addition of DNA 1 at both pH of 7.2 and 4.5 (ESI, S10a–c†).

Fig. 9 TEM images of pH-dependent nanostructure switching of CMP-3 + DNA 1: pH 7.2, cycles 1–3 (A, C and E) and pH 4.5, cycles 1 and 2 (B and D), respectively. Peptide concentration of 20 μM and DNA 1 concentration of 30 μM. Buffer conditions: 10 mM Tris, 10 mM NaCl at pH 7.2 or pH 4.5.

Self-assembly of collagen-(PNA-GGG)₃ peptide CMP-3 : DNA 1 in solution

The self-assembly of collagen-(PNA-GGG)₃ CMP-3 : DNA 1 triplex–duplex in solution was examined using the thioflavin-T (ThT) assay.⁴⁸ Free ThT is weakly fluorescent and when present in a self-assembled/aggregation environment, it showed an enhanced fluorescence emission at 485 nm. In the presence of free collagen-(PNA-GGG)₃ peptide CMP-3 and DNA 1, ThT exhibited low fluorescence (Fig. 10A). Upon mixing them, the fluorescence intensity became enhanced with an increase in the concentration of DNA 1. At 1 : 1 equivalent of collagen-(PNA-GGG)₃ peptide CMP-3 and DNA 1, a 14-fold increment in fluorescence intensity was observed. ThT can intercalate and recognize nucleic acid secondary structures such as duplexes, triplexes and tetraplexes.⁴⁹ Thus, to examine if the ThT fluorescence arises from its complexation with the PNA : DNA duplex component of collagen-PNA : DNA 1, the ThT interaction with a short PNA-(GGG) : DNA 1 duplex (without the peptide) was studied (ESI, S11†). However, this showed a negligible increment in the ThT fluorescence intensity (ESI, S12†). Thus, the huge ThT fluorescence observed with collagen-(PNA-GGG)₃ CMP-3 : DNA 1 (Fig. 10A) arises from the large aggregates of triplexes of CMP-3-PNA peptide interconnected by PNA(GGG) : DNA 1 duplexes, in addition to the small contribution from the ThT interaction with the PNA : DNA duplexes. This result provided additional evidence for the formation of aggregates from the self-assembled complex of CMP-3-(PNA-GGG)₃ : DNA 1 through triplex–duplex interconnects, as seen in the TEM images.

Fig. 10 Thioflavin (ThT) emission spectra for collagen-(PNA-GGG)₃ CMP-3, DNA 1 and peptide CMP-3 + DNA 1 complexes in different ratios at (A) pH 7.2 and (B) pH 4.5. λ_ex = 412 nm; λ_em = 450–700 nm. Peptide concentration of 20 μM. Buffer conditions: 10 mM Tris, 10 mM NaCl.

The ThT fluorescence intensity of CMP-3-(PNA-GGG)₃ : DNA 1 at pH 4.5 was much lower compared to neutral pH 7.2 (with only 9-fold increment (Fig. 10B)). A considerable contribution to the ThT fluorescence arises from the i-motif structure of DNA 1, as seen in Fig. 10B.⁴⁹

pH-induced reversible switching of collagen-(PNA-GGG)₃ CMP-3 : DNA 1 complex between triplex–duplex and triplex/i-motif

The C-rich DNA 1 alone exhibited an i-motif tetraplex structure at pH 4.5, as shown by its CD spectra with a shift in the positive band at 275 nm band (pH 7.2) to 289 nm with higher intensity at pH 4.5 (Fig. 11A). Reversing the pH to 7.2, the i-motif of DNA 1 unfolded into a linear structure, with the disappearance of the i-motif CD band at 289 nm and restoration of the band at 275 nm. This conformational switch was reproducible over several cycles of pH switching, as seen in Fig. 11A. A similar experiment of pH cycling was performed with CMP-3-PNA(GGG)₃ : DNA 1, which at pH 7.2 showed duplex interconnecting the triplexes. At pH 4.5, the duplex dissociated into individual components (CMP-3 and DNA 1), with the liberated single-strand DNA 1 forming an i-motif tetraplex. This process was followed by CD (Fig. 11B), which at pH 4.5 showed the disappearance of the broad band at 275 nm and appearance of a new band at 290 nm, which is characteristic of the i-motif. The process of conformational change between pH 7.2 and 4.5 was reproducibly reversible over a few cycles, as shown by the CD profiles in Fig. 11B. The changes in the intensity of ellipticity at 289 nm (data extracted from Fig. 11A and B, respectively) at pH 7.2 and 4.5 in each cycle for CMP-3 : DNA 1 and DNA 1 are qualitatively similar in nature, but the magnitude slightly decreased with successive cycles (Fig. 11C), indicating the reversibility. With C-rich DNA 1 alone, the ellipticity decreased with successive cycles (Fig. 11C, black line), but in the case of the triplex–duplex (CMP-3 : DNA 1), a negligible decrease in ellipticity over successive cycles was seen (Fig. 11C, red line). This could arise from the difference in the kinetics of association/dissociation of the complexes formed in each cycle, arising from the different possible combinations of triplex–duplex interconnects, which may influence the overall triplex structure and stabilities.

Fig. 11 (A) Reversible CD spectra of DNA 1; (B) CD spectra of collagen-(PNA-GGG)₃ CMP-3 + DNA 1 between pH 7.2 and pH 4.5; and (C) ellipticity at 289 nm at pH 7.2 and 4.5. DNA 1 (black) & collagen-(PNA-GGG)₃ CMP-3 : DNA 1 (red) during cycles 1–4. (D) Maximum fluorescence intensity of ThT in the presence of CMP-3 : DNA 1 at 485 nm at pH 7.2 (red square) and pH 4.5 (blue circle) during cycling.

The switching between interconnected triplex–duplex to isolated triplex/i-motif on changing the pH was also followed by the fluorescence emission of the ThT complexes. Three cycles of pH switching between 7.2 and 4.5 of collagen-(PNA-GGG)₃ (CMP-3) : DNA 1 was examined through the changes in the ThT fluorescence intensity at 485 nm (Fig. 11D and Fig. S13, ESI†). The ThT fluorescence was higher in the case of the more aggregated CMP-3 : DNA 1 complex (pH 7.2) compared to either CMP-3 or DNA 1 (pH 4.5) and reversibly switching pH between 7.2 and 4.5 leading to the corresponding oscillation of the fluorescence intensity (Fig. 11D). This pattern agrees with that seen by the CD ellipticity at 289 nm (Fig. 11C). Given that the ThT assay recognizes only aggregation, the intensities remained relatively constant with successive cycles. The combined results from the CD and fluorescence data support the reversible switching of collagen-(PNA-GGG)₃ CMP-3 : DNA 1 from triplex–duplex to triplex/i-motif between pH 7.2 and 4.5.

Conclusions

This study illustrated the formation of polyplexes from the collagen mimetic peptide-PNA conjugate CMP-PNA-(GGG)₃ CMP-3 in the presence of complementary DNA 1. The PNA component of the conjugate formed a PNA : DNA duplex with C-rich DNA 1 at neutral pH 7.2, which interconnected the triplexes formed by the collagen mimetic peptide [CMP-PNA(GGG)₃]₃. The formation of this triplex–duplex hybrid was shown by thermal melting, CD, fluorescence, gel-shift experiments, ThT aggregation assay and TEM images. The installation of PNA chains on the collagen-type triplex enhanced the triplex-forming propensity of the collagen peptide without hampering the ability of the PNA sidechain to form a duplex with complementary DNA 1. This was proven through the CD signature bands for both the peptide triplex and PNA : DNA duplex and individual T_m monitored at two different wavelengths in the same set of experiments. Upon lowering the pH to acidic range (pH 4.5), the polyplexes dissociated, generating the component triplex CMP-3 and C-rich DNA 1. At pH 4.5, the liberated DNA 1 folded into an i-motif tetraplex, which was characterized by UV and CD spectroscopy. An increase in the pH to neutral 7.2 regenerated the triplex–duplex polyplexes and repeated switching of the pH from 7.2 to 4.5 and back to pH 7.2 over several cycles demonstrated the reversibility of this process. This was established by CD spectral changes, ThT fluorescence spectroscopy and switching of the nanostructure morphology by TEM.

The polyplex nanostructures made by the collagen-(PNA-GGG)₃ : DNA hybrid system described herein have potential for expansion through further functional designs to entrap small molecules or drugs of interest for their pH-controlled release at specific sites, for instance, in solid tumors, where the pH of the microenvironment is slightly acidic (e.g., release of doxorubicin in pegylated nanoparticles in Hela cell xenograft nude mice,⁵⁰ and by RGD nanoparticles in hepatoma,⁵¹ release of zorubicin hydrochloride using hollow silica nanocarriers,⁵² and many other specific examples⁵³). Given that the T_m of CMP-3 : DNA 1 conjugate is around 35–40 °C, the acid-sensitive nature of C-rich DNA present in the hybrid nanostructure would disassemble at physiological temperature only at the acidic tumor site, thereby releasing the embedded drug. The use of ionizable 4-aminoproline in collagen peptide,^19b,c,54 and the replacement of glycine with lysine for PNA conjugation would enable further modulation of the pH-dependent triplex stability of the conjugate hybrid. By varying the length of the linker PNA sequence or using lysine analogues with different alkylamino chains, one can tune the void space in nano-assemblies of the collagen-(PNA-GGG)₃ : DNA system. Tuning DNA-peptide conjugates to influence individual self-assembling features in a dynamic way is an emerging field.⁵⁵ The triplex–duplex system described herein is an early example of polyplexes in which two different types of helical structures are present in the same complex, without mutually affecting their individual stabilities. This adds to the repertoire of PNA-derived polyplexes, which we recently designed using a Janus PNA that can form fused duplexes and triplexes with enhanced stabilities in a synergistic manner.⁵⁶ The present collagen mimetic peptide-PNA conjugate with reversible switching properties through the assembly–disassembly process also has potential as a new emerging biomaterial in theranostics for targeting hybridization with damaged and denatured collagen, which are causes of pathologies associated with several diseases.^57,58

Experimental

General

N,N-Diisopropylethylamine (DIPEA), acetic anhydride, piperidine, trifluoroacetic acid (TFA), 1-hydroxy benzotriazole (HOBt), HBTU, and N,N,N′,N′-tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate were obtained from commercial sources with >99% purity. The solvents dichloromethane (DCM) and N,N-dimethyl formamide (DMF) used for peptide synthesis were dried by standard methods. Acetonitrile (HPLC grade 99.9%) was obtained from a commercial source and used without further treatment. The protected amino acid monomers 1–4 (Fmoc-Pro-OH, Fmoc-Hyp(OtBu)-OH, Fmoc-Lys(Mtt)-OH, and Fmoc-Gly-OH, respectively) were purchased from Sigma-Aldrich and PNA-G monomer [Fmoc-PNA(Bhoc)G-OH] purchased from ASM Research Chemicals (Part. No. 5004030) (SD3056). The rink amide resin (100–200 mesh) with a loading of 0.62 mmol g⁻¹ was procured from Novabiochem. The DNA sequences DNA 1 (5′-CCC TTT CCC-3′), DNA 2 (5′-GGG AAA GGG-3′) and DNA (T₉): 5′-TTT TTT TTT-3′ were purchased as a desalted product from Sigma-Aldrich and used as received. All spectroscopic data were plotted using the OriginPro 8.5.0 software.

Synthesis and purification of peptides

The peptides CMP-1, CMP-2, collagen-(PNA-GGG)₃ peptide (CMP-3) and CMP-4, as shown in Schemes 1 and 2, were prepared using a solid support employing Fmoc chemistry on rink amide resin following previously reported procedures.^46,47,58 The crude peptides collagen-(PNA-GGG)₃ (CMP-3) and CMP-4 were purified by reverse-phase (RP) HPLC using a Waters instrument with semi-prep column (21.2 × 250 mm, 10 μm particle size, 300 Å pore size, Phenomenex) using a binary solvent system. Solvent A: 95% water, 5% ACN and 0.1% TFA. Solvent B: 50% water, 50% ACN, 0.1% TFA flow rate (3 mL min⁻¹). Linear gradient 100% solvent A to 100% solvent B over 20 min and monitored at 220 and 260 nm. Retention time: CMP-4 (14.1 min) and collagen-(PNA-GGG)₃ peptide CMP-3 (12.7 min).

The peptides were characterized by MALDI-TOF in reflection mode using DHB as the matrix. CMP-4: mol. formula is C₁₂₂H₁₈₂N₃₂O₃₅ [M]⁺ calc. mass is 2655.34; observed, 2655.90. Collagen-(PNA-GGG)₃ peptide CMP-3: mol. Formula is C₂₁₂H₂₉₀N₉₄O₆₁ [M]⁺ calc. mass is 5128.24; observed, 5128.48.

Determination of peptide concentration: Concentration of PNA containing collagen-(PNA-GGG)₃ peptide (CMP-3) was calculated using the ε₂₆₀ value of the PNA monomer^59a (G = 11 700 M⁻¹ cm⁻¹) and the concentration of the CMP-4 was determined from the ε₂₅₇ of the phenylalanine residue^59b (Phe = 195 M⁻¹ cm⁻¹).

Circular dichroism (CD) spectroscopy

The CD spectra of collagen-(PNA-GGG)₃ peptide (CMP-3), CMP-4, and CMP-3 + DNA 1 were recorded on a JASCO 1700 instrument by scanning from 320 nm to 200 nm at a scanning speed of 50 nm min⁻¹. Data were collected at 0.1 nm with a digital integration time (DIT) of 1s and data accumulated in three scans.

The reversible conformational switching of collagen-(PNA-GGG)₃ peptide (CMP-3) : DNA 1 and DNA 1 alone was conducted by changing the pH from 7.2 to 4.5 using a Horiba LAQUAtwin-pH-22 microvolume pH meter. Firstly, the CMP 3 : DNA 1 sample was prepared at pH 7.2, its CD spectrum was recorded, and then the pH of the sample was switched to pH 4.5 using the microvolume pH meter and the CD spectrum was again recorded. This process was repeated for each cycle, allowing 30 min between two measurements. The pH of the samples was controlled by the addition of 0.1 M HCl or 0.1 M NaOH solution.

CD melting study

All the samples were prepared in Tris buffer (10 mm, pH 7.2 or pH 4.5) and NaCl (10 mM) with peptides (20 μM). Variable temperature CD studies were carried out on a JASCO 1700 instrument in the temperature range of 5 °C–90 °C. Data were recorded at 5 °C intervals with a temperature ramping rate of 0.33 °C min⁻¹ and hold time at each temperature of 100 s.

Fluorescence study

The thioflavin-T binding assay for collagen-(PNA-GGG)₃ peptide (CMP-3) + DNA 1 was carried out by recording fluorescence emission spectra on a JASCO FP-8500 spectrofluorometer. A freshly prepared solution of ThT was added to the samples containing Tris buffer (10 mM, pH 7.2), NaCl (10 mM) to reach a final concentration of ThT of 50 μM and equilibrated for 30 min. The emission spectra were recorded with excitation at 412 nm and emission at 460 nm (bandwidth of 10 nm).

pH cycle and reversibility

The emission spectra of the sample prepared at pH 7.2 was initially recorded. Then, the pH of the sample was switched to pH 4.5 by the controlled addition of 0.1 M HCl and its emission spectrum recorded. The pH was reversed to 7.2 by the addition of 0.1 M NaOH solution and its fluorescence spectrum recorded again. This cycle was repeated 3–4 times with 30 min between two measurements in each cycle. The pH at each point of the cycle was monitored on a Horiba LAQUAtwin-pH-22 meter.

Ethidium bromide (Et-Br) displacement assay

The duplex DNA (DNA 1 : DNA 2, each 30 μM) was treated with EB (100 μM) and incubated at 4 °C for 1 h. The emission spectra were recorded in the range of 500–800 nm with excitation at λ 310 nm. The collagen-(PNA-GGG)₃ peptide (CMP-3) (0.33 equivalent, 2 μL) was incrementally added and incubated for 20–25 min after each addition, followed by recording of the emission spectra. The yield of collagen-(PNA-GGG)₃ : DNA 1 hybridization was calculated by treating the fluorescence intensity of EB in the DNA 1 : DNA 2 duplex as 100% (original state). Then, the loss of fluorescence intensity from the original state was used for calculating the percentage yield of collagen-(PNA-GGG)₃ : DNA 1 hybridization.

Gel electrophoresis

Native 10% PAGE was cast and run under 1xTAE buffer of pH 7.2 at 100 V for 70 min. The collagen-(PNA-GGG)x₃ peptide (CMP-3, 100 μM) and DNA 1 (50, 100, and 150 μM) samples were used together with 6x loading dye (5 μL) for visualizing the progress of the gel. Afterwards, the gel was stained with EB for 10 min, followed by washing with Milli-Q (MQ) water to remove EB. Then, the gel was stained with Coomassie blue solution for 2 h, followed by destaining by repeated washing with acetic acid-water. The gel was imaged using a BIORAD GelDox XR⁺ system. Buffer: Tris-acetate (40 mM) and EDTA (1 mM, pH = 7.2 or pH = 4.5). Loaded Dye: 2% bromophenol in 40% sucrose and 2% xylene cyanol. Staining solution: Coomassie brilliant blue R-250 (2 g L⁻¹) in methanol : glacial acetic acid : water (5 : 1 : 4). Destaining solution: methanol: glacial acetic acid: water (5 : 1 : 4).

Transmission electron microscopy (TEM)

To observe CMP-3 (5 μL) and DNA 1 alone or CMP-3 + DNA 1 mixture, the sample was dropped onto a freshly glow-discharged carbon-covered grid (400 mesh) for two minutes. Then, two successive rinsing steps were performed by dropping 5 μL of buffer on the grid and leaving in contact for 4 min and 6 min, respectively. Finally, the grid was negatively stained with 5 μL of uranyl acetate (2% in water) for one minute. The excess water was removed by using a filter paper between steps. The grids were observed at 200 kV with a Tecnai G2 (FEI) microscope. Images were acquired with an Eagle 2k (FEI) CCD camera.

pH-Dependent nanostructure switching

Images were first recorded at pH 7.2, and then the pH of the sample was switched to pH 4.5, and again images were recorded. The pH was controlled by the addition of 0.1 M HCl and 0.1 M NaOH solution. Before applying the samples to the grid, they were kept at room temperature for 30 min.

Author contributions

Shahaji More – Idealised the project, carried out all synthesis, purification, analysis (spectra), gel electrophoresis and pH cycling experiments. Marc Schumtz – preparation of samples and all TEM experiments. Loïc Jierry – helpful discussion on assembly–disassembly processes and resources. Krishna Ganesh – Conceptualization of the project and analysis of results and resources.

Data availability

The data supporting this article are included as part of the ESI.† This includes HPLC, MALDI-TOF, CD spectra, gel electrophoresis, TEM images and fluorescence spectra of collagen peptide, PNA conjugate and their assemblies.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

SHM gratefully acknowledges CNRS France, IISER Pune, IISER Tirupati for fellowships. KNG acknowledges SERB Govt of India, Science Chair grant NSC/2022/000007. We thank ICS, France for facilities for characterization and microscopy platform.

References

1. (a) Y. F. Lee, A. Nomoto and E. Wimmer, Prog. Nucleic Acid Res. Mol. Biol., 1976, 19, 89–96; (b) J. B. Flanegan, R. F. Pettersson, V. Ambros, M. J. Hewlett and D. Baltimore, Proc. Natl. Acad. Sci. U. S. A., 1977, 74, 961–965; (c) E. A. Carusi, Virology, 1977, 76, 380–394; (d) D. M. Rekosh, W. C. Russel, A. J. Bellet and A. J. Robinson, Cell, 1977, 11, 283–295; (e) J. M. Hermoso and M. Salas, Proc. Natl. Acad. Sci. U. S. A., 1980, 77, 6425–6428.
2. G. Stubbs, S. Warren and K. Holmes, Nature, 1977, 267, 216–221.
3. (a) J. L. Osterhage and K. L. Friedman, J. Biol. Chem., 2009, 284, 16061–16065; (b) B. Albert, A. Johnson, J. Lewis, D. Morgan, M. Raff, K. Roberts and P. Walter, Molecular biology of cell, Gerland Science (Taylor and Francis), 5th edn, 2008.
4. (a) R. D. Kornberg, Science, 1974, 184, 868–871; (b) D. E. Olins and A. L. Olins, Mol. Cell Biol., 2003, 4, 809–814; (c) A. Yonath, Angew. Chem., Int. Ed., 2010, 49, 4340–4354.
5. C.-H. Tung, Bioconjugate Chem., 2000, 11, 605–618.
6. K. Lu, Q.-P. Duan, L. Ma and D.-X. Zhao, Bioconjugate Chem., 2010, 21, 187–202.
7. E. E. Watson and N. Winssinger, Biomolecules, 2022, 12, 1523–1534.
8. R. Freeman, M. Han, Z. Alvarez, J. L. Lewis, J. R. Wester, N. Stephanopoulos, M. T. McClendon, C. Lynsky, J. M. Godbe, H. Sangji, E. Luijten and S. I. Stupp, Science, 2018, 362, 808–813.
9. F. Praetorius and H. Dietz, Science, 2017, 355, 1283.
10. (a) B. A. Rosenzweig and A. D. Hamilton, Angew. Chem., Int. Ed., 2009, 48, 2749–2751; (b) P. Ghosh and A. D. Hamilton, J. Am. Chem. Soc., 2012, 134, 13208–13211.
11. (a) M. Humenik and T. Scheibel, ACS Nano, 2014, 8, 1342–1349; (b) B. A. R. Williams, K. Lund, Y. Liu, H. Yan and J. C. Chaput, Angew. Chem., Int. Ed., 2007, 46, 3051–3054.
12. (a) C. Lou, M. C. Martos-Maldonado, C. S. Madsen, R. P. Thomsen, S. R. Midtgaard, N. J. Christensen, J. Kjems, P. W. Thulstrup, J. Wengel and K. J. Jensen, Nat. Commun., 2016, 7, 12294–12302; (b) C. Lou, N. J. Christensen, M. C. Martos-Maldonado, S. R. Midtgaard, M. Ejlersen, P. W. Thulstrup, K. K. Sorensen, K. J. Jensen and J. Wengel, Chem. – Eur. J., 2017, 23, 9297–9305.
13. T. Jiang, T. A. Meyer, C. Modlin, X. Zuo, V. P. Conticello and Y. Ke, J. Am. Chem. Soc., 2017, 139, 14025–14028.
14. M. L. Daly, S. J. Klawa and R. Freeman, New functions emerging from peptide-DNA materials, in Peptide-based Biomaterials, ed. M. O. Guler, Royal Society of Chemistry, London, 2020, 459–486.
15. T. MacCulloch, A. Buchberger and N. Stephanopoulos, Org. Biomol. Chem., 2019, 17, 1668–1682.
16. N. Stephanopoulos, Bioconjugate Chem., 2019, 30, 1915–1922.
17. L. Roglin and O. Seitz, Org. Biomol. Chem., 2008, 6, 3881–3887.
18. M. D. Shoulders and R. T. Raines, Annu. Rev. Biochem., 2010, 78, 929–958 and appropriate references cited therein.
19. (a) L. E. Bretscher, C. L. Jenkins, K. M. Taylor, M. L. DeRider and R. T. Raines, J. Am. Chem. Soc., 2001, 123(4), 777–778; (b) I. R. Babu and K. N. Ganesh, J. Am. Chem. Soc., 2001, 123, 2079–2080; (c) M. Umashankara, M. V. Sonar, N. D. Bansode and K. N. Ganesh, J. Org. Chem., 2015, 80, 8552–8560; (d) C. Siebler, R. S. Erdmann and H. Wennemers, Chimia, 2013, 67, 891.
20. (a) J. A. Fallas, L. E. R. O'Leary and J. D. Hartgerink, Chem. Soc. Rev., 2010, 39, 3510–3527; (b) S. A. H. Hulgan and J. D. Hartgerink, Biomacromolecules, 2022, 23, 1475–1489.
21. (a) Y. Xu and M. Kirchner, Bioengineering, 2021, 8, 1–24; (b) J.-D. Malcor and F. Mallein-Gerin, Acta Biomater., 2022, 148, 1–21.
22. J. Banerjee and H. S. Azevedo, Interface Focus, 2017, 7, 20160138.
23. K. Strauss and J. Chmielewski, Curr. Opin. Biotechnol., 2017, 46, 34–41.
24. (a) V. A. Patil and K. S. Masters, Bioengineering, 2020, 7, 163–183; (b) T. Luo and K. L. Kiick, Bioconjugate Chem., 2017, 28, 816–827.
25. (a) D. E. Przybyla and J. Chmielewski, J. Am. Chem. Soc., 2008, 130, 12610–12611; (b) D. E. Przybyla, C. M. R. Perez, J. Gleaton, V. Nandwana and J. Chmielewski, J. Am. Chem. Soc., 2013, 135, 3418–3422.
26. (a) P. E. Nielsen, M. Egholm, R. H. Berg and O. Buchardt, Science, 1991, 254, 1497–1500; (b) M. Egholm, O. Buchardt, L. Christensen, C. Behrens, S. M. Freier, D. A. Driver, R. H. Berg, S. K. Kim, B. Norden and P. E. Nielsen, Nature, 1993, 365, 566–568; (c) P. Wittung, P. E. Nielsen, O. Buchardt, M. Egholm and B. Norden, Nature, 1994, 368, 561–563; (d) L. Betts, J. A. Josey, J. M. Veal and S. R. Jordon, Science, 1995, 270, 1838–1841; H. Knudsen and P. E. Nielsen, Nucleic Acids Res., 1996, 24, 494–500; (e) K. K. Jensen, H. Orum, P. E. Nielsen and B. Norden, Biochemistry, 1997, 36, 5072–5077; (f) P. Wittung, P. Nielsen and B. Norden, Biochemistry, 1997, 36, 7973–7979; (g) T. Bentin, H. J. Larsen and P. E. Nielsen, Biochemistry, 2003, 42, 13987–13995.
27. (a) K. N. Ganesh and V. A. Kumar, Acc. Chem. Res., 2005, 38, 404–412; (b) K. N. Ganesh and V. A. Kumar, Curr. Top. Med. Chem., 2007, 7, 715–726; (c) C. Soprarpprom and T. Vilaivan, RSC Chem. Biol., 2022, 3, 648–697.
28. (a) C. S. Swenson, A. Velusamy, H. S. Argueta-Gonzalez and J. M. Heemstra, J. Am. Chem. Soc., 2019, 141, 19038–19047; (b) H. S. Argueta-Gonzalez, C. S. Swenson, G. Song and J. M. Heemstra, RSC Chem. Biol., 2022, 3, 1035–1043; (c) H. S. Argueta-Gonzalez, C. S. Swenson, K. J. Skowron and J. M. Heemstra, ACS Omega, 2023, 8(40), 37442–37450.
29. (a) A. M. Hansen, G. Bonke, C. J. Larsen, N. Yavari, P. E. Nielsen and H. Franzyk, Bioconjugate Chem., 2016, 27(4), 863–867; (b) U. Tsylents, I. Siekierska and J. Trylska, Eur. Biophys. J., 2023, 52, 533–544.
30. (a) K. Gehring, L.-J. Leroy and M. Gueron, Nature, 1993, 363, 561–565; (b) C. Kang, I. Berger, C. Lockshin, R. Ratliff, R. Moyzis and A. Rich, Proc. Natl. Acad. Sci. U. S. A., 1994, 91, 11636–11640; (c) J.-L. Mergny, L. Lacroix, X. Han, L.-J. Leroy and C. Helene, J. Am. Chem. Soc., 1995, 117, 8887–8898; (d) M. Gueron and L.-J. Leroy, Curr. Opin. Struct. Biol., 2000, 10, 326–331.
31. (a) D. Liu and S. Balasubramanian, Angew. Chem., Int. Ed., 2003, 42, 5734–5736; (b) Y. Wang, X. Li, X. Liu and T. Li, Chem. Commun., 2007, 4369–4371.
32. D. Li and D. L. Elbert, J. Pept. Res., 2002, 60, 300–303.
33. M. G. Venugopal, J. A. M. Ramshaw, E. Braswell, D. Zhu and B. Brodsky, Biochemistry, 1994, 33, 7948–7956.
34. Y. Feng, G. Melacini, J. P. Taulane and M. Goodman, J. Am. Chem. Soc., 1996, 118, 10351–10358.
35. A. V. Persikov, J. A. M. Ramshaw, A. Kirkpatrick and B. Brodsky, Biochemistry, 2000, 39, 14960–14967.
36. A. V. Persikov, X. Yujia and B. Brodsky, Protein Sci., 2004, 13, 893–902.
37. Y. Zhang, R. M. Malamakal and D. M. Chenoweth, J. Am. Chem. Soc., 2015, 137, 12422–12425.
38. R. S. Erdmann and H. Wennemers, J. Am. Chem. Soc., 2010, 132, 13957–13959.
39. Y. Krishnan-Ghosh, E. Stephens and S. Balasubramanian, J. Am. Chem. Soc., 2004, 126, 5944–5945.
40. (a) N. K. Sharma and K. N. Ganesh, Chem. Commun., 2005, 4330–4332; (b) Y. Krishnan-Ghosh, E. Stephens and S. Balasubramanian, Chem. Commun., 2005, 5278–5280.
41. (a) L. Shi, P. Peng, Y. Du and T. Li, Nucleic Acids Res., 2017, 45, 4306–4314; (b) J. Kypr, I. Kejnovska, D. Renciuk and M. Vorlickova, Nucleic Acids Res., 2009, 37, 1713–1725.
42. (a) W. C. Tse and D. L. Boger, Acc. Chem. Res., 2004, 37, 61–69; (b) P. Wittung, P. E. Nielsen and B. Norden, J. Am. Chem. Soc., 1996, 118, 7049–7054; (c) D. R. Jain, A. V. Libi, M. Lahiri and K. N. Ganesh, J. Org. Chem., 2014, 79, 9567–9577.
43. J. L. Mergny and L. Lacroix, Oligonucleotides, 2003, 13, 515–537.
44. M. Jasinki, J. Miszkiewicz, M. Feig and J. Trylska, J. Phys. Chem. B, 2019, 123, 8168–8177.
45. T. H. Steinberg, Methods Enzymol., 2009, 463, 541–563.
46. B. R. Madhangopal, S. H. More, N. D. Bansode and K. N. Ganesh, ACS Omega, 2020, 5, 2181–21795.
47. S. H. More and K. N. Ganesh, Pept. Sci., 2019, e24140.
48. M. Biancalana and S. Koide, Biochim. Biophys. Acta, 2010, 1804, 1405–1412.
49. S. Verma, V. Ravichandiran and N. Ranjan, Biochimie, 2021, 190, 111–123.
50. X. Zeng, G. Liu, W. Tao, Y. Ma, X. Zhang, F. He, J. Pan, L. Mei and G. Pan, Adv. Funct. Mater., 2017, 27, 1605985.
51. Z. Gong, X. Liu, B. Zhou, G. Wang, X. Guan, Y. Xu, J. Zhang, Z. Hong, J. Cao, X. Sun, Z. Gao, H. Lu, X. Pan and J. Bai, Colloids Surf., B, 2021, 202, 111673.
52. S. Deng, C.-X. Cui, L. Duan, L. Hu, X. Yang, J.-C. Wang, L.-B. Qu and Y. Zhang, ACS Omega, 2021, 6, 553–558.
53. A. Tafech and A. Stephanou, Biology, 2024, 13, 225.
54. J. Egli, C. Siebler, B. Maryasin, R. S. Erdmann, C. Bergande, C. Ochsenfeld and H. Wennemers, Chem. – Eur. J., 2017, 23, 7938–7944.
55. (a) E. R. Taylor, A. Sato, I. Jones, P. G. Gudeangadi, D. M. Beal, J. A. Hopper, W.-F. Xue, M. R. Reithofer and C. J. Serpell, Chem. Sci., 2023, 14, 196–202; (b) P. R. Na Bhuket, Y. Li and Y. S. M. Yu, Acc. Chem. Res., 2024, 57, 1649–1657.
56. (a) M. K. Gupta, B. R. Madhanagopal, D. Datta and K. N. Ganesh, Org. Lett., 2020, 22, 5255–5260; (b) P. Bhingardeve, B. R. Madhanagopal and K. N. Ganesh, J. Org. Chem., 2020, 85, 13680–13693; (c) M. K. Gupta, B. R. Madhanagopal and K. N. Ganesh, J. Org. Chem., 2021, 86, 414–428.
57. (a) P. Peng, Y. Du, J. Zheng, H. Wang and T. Li, Angew. Chem., Int. Ed., 2019, 58, 1648–1653; (b) X. Hu, F. Li, S. Wang, F. Xia and D. Ling, Adv. Healthcare Mater., 2018, 1800359.
58. (a) H. Wahyudi, A. A. Reynolds, Y. Li, S. C. Owen and S. M. Yu, J. Controlled Release, 2016, 240, 323–331; (b) X. Li, Q. Zhang, S. M. Yu and Y. Li, J. Am. Chem. Soc., 2023, 145, 10901–10916.
59. (a) Applied Biosystems’ PNA Quantification Calculator: https//www.appliedbiosystems.com/support/concentration.cfm; (b) Molar extinction coefficient value of Phenylalanine: https://www.biosyn.com/tew/How-is-Extinction-Coefficient-Determined-for-Proteins.aspx.

---

Footnotes

† Electronic supplementary information (ESI) available: HPLC, MALDI-TOF, CD spectra, gel electrophoresis, TEM images and fluorescence spectra of collagen peptide, PNA conjugate and their assemblies. See DOI: https://doi.org/10.1039/d4bm00955j

‡ New Chemistry Unit, Jawaharlal Nehru Center for Advanced Scientific Research (JNCASR), Jakkur, Bengaluru 560064, India.

---