Chemical approaches to discover the full potential of peptide nucleic acids in biomedical applications

Abstract

Peptide nucleic acid (PNA) is arguably one of the most successful DNA mimics, despite a most dramatic departure from the native structure of DNA. The present review summarizes 30 years of research on PNA's chemistry, optimization of structure and function, applications as probes and diagnostics, and attempts to develop new PNA therapeutics. The discussion starts with a brief review of PNA's binding modes and structural features, followed by the most impactful chemical modifications, PNA enabled assays and diagnostics, and discussion of the current state of development of PNA therapeutics. While many modifications have improved on PNA's binding affinity and specificity, solubility and other biophysical properties, the original PNA is still most frequently used in diagnostic and other in vitro applications. Development of therapeutics and other in vivo applications of PNA has notably lagged behind and is still limited by insufficient bioavailability and difficulties with tissue specific delivery. Relatively high doses are required to overcome poor cellular uptake and endosomal entrapment, which increases the risk of toxicity. These limitations remain unsolved problems waiting for innovative chemistry and biology to unlock the full potential of PNA in biomedical applications.

Keywords: PNA; antisense; chemical modifications; diagnostics; peptide nucleic acid.

Figure 1

Structure of DNA and PNA.

Figure 2

PNA binding modes: (A) PNA–dsDNA 1:1 triplex; (B) PNA–DNA–PNA strand-invasion triplex; (C) the Hoogsteen and Watson–Crick parts are linked together in a bis-PNA; (D) shortening the Hoogsteen part and extending the Watson–Crick part of the bis-PNA creates a tail-clamp PNA (tcPNA); (E) and (F) single and double invasion using only Watson–Crick hydrogen bonding; (G) Janus-wedge triple helix.

Figure 3

Structure of P-form PNA–DNA–PNA triplex from reference [41]. (A) view in the major groove and (B) view in the minor groove.

Figure 4

Structures of backbone-modified PNA.

Figure 5

Structures of PNA having α- and γ-substituted backbones.

Figure 6

Structures of modified nucleobases in PNA to improve Hoogsteen hydrogen bonding to guanine and adenine. R¹ denotes DNA, RNA, or PNA backbones.

Figure 7

Proposed hydrogen bonding schemes for modified PNA nucleobases designed to recognize pyrimidines or the entire Hoogsteen face of the Watson–Crick base pairs. R¹ denotes DNA, RNA, or PNA backbones.

Figure 8

Modified nucleobases to modulate Watson–Crick base pairing and chemically reactive crosslinking PNA nucleobases. R¹ denotes DNA, RNA, or PNA backbones.

Figure 9

Examples of triplets formed by Janus-wedge PNA nucleobases (blue). R¹ denotes DNA, RNA, or PNA backbones.

Figure 10

Examples of fluorescent PNA nucleobases. R¹ denotes DNA, RNA, or PNA backbones.

Figure 11

Endosomal entrapment and escape pathways of PNA and PNA conjugates.

Figure 12

(A) representative cell-penetrating peptides (CPPs), (B) conjugation designs and linker chemistries.

Figure 13

Proposed delivery mode by pHLIP-PNA conjugates (A) the transmembrane section of pHLIP interacting with lipid bilayer, (B) low surface pH leads to partial protonation of negative residues triggering interfacial helix formation and deeper partitioning into lipid bilayer, and (C) the transmembrane helix formation and release of PNA into cytosol by disulfide cleavage.

Figure 14

Structures of modified penetratin CPP conjugates with PNA linked through either disulfide (for study in HeLa pLuc705 cells) or thioether bonds (for study in cultured mdx mouse myotubes or mouse model).

Figure 15

Chemical structure of C₉–PNA, a stable amphipathic (cyclic-peptide)–PNA conjugate.

Figure 16

Structures of PNA conjugates with a lipophilic triphenylphosphonium cation (TPP–PNA) through (A) thioether and (B) cleavable disulfide linkage; (C) PNA–R₉ conjugates with lipids, phospholipids and cleavable lipids.

Figure 17

Structures of (A) chloesteryl–PNA, (B) cholate–PNA and (C) cholate–PNA(cholate)₃.

Figure 18

Structures of PNA–GalNAc conjugates (A) (GalNAc)₂K, (B) triantennary (GalNAc)₃, and (C) trivalent (T-γ-GalNAc)₃.

Figure 19

Vitamin B₁₂–PNA conjugates with different linkages.

Figure 20

Structures of (A) neomycin B, (B) PNA–neamine conjugate, and (C) PNA–neosamine conjugate.

Figure 21

PNA clamp (red) binding to target DNA containing a mixture of sequences (A) PNA binds with higher affinity to the perfectly matched wild-type sequence while binding to the mutant containing as few as one mismatch is weaker. Once elongation begins, the perfectly matched complex stalls the polymerase inhibiting elongation while the mismatched complex dissociates allowing for elongation to continue; (B) LNA probes (blue) can also out compete PNA/DNA complexes mismatched allowing for sequence selective detection of mutant alleles; (C) NAVIGATER uses DNA-guided Argonaute to selective degrade wild-type oligos to enrich the mutant population increasing the sensitivity of PCR clamping.

Figure 22

Rolling circle amplification using PNA openers (red) to invade a dsDNA target forming a P-loop. A padlock DNA probe (blue) can bind to the DNA liberated by the PNA openers. Ligase circularizes the padlock DNA resulting in an earring complex which acts as a primer for DNA polymerase. The resulting rolling circle amplification product (orange) can then be isolated or detected in solution.

Figure 23

Molecular beacons containing generic fluorophores (Fl) and quenchers (Q) recognizing a complementary oligonucleotide. (A) PNA/DNA chimeras [244] (PNA in red, DNA in blue) and (B) PNA [245] with self-complementary stems were originally used to ensure close proximity of the fluorophore and quencher; (C) stemless beacons [246] lack partially self-complementary sequences instead relying on PNA aggregation to keep the fluorophore and quencher in proximity; (D) two complementary PNAs can also be used to ensure the proximity in dsPNA beacons.

Figure 24

(A) Light-up fluorophores such as thiazole orange display fluorescence enhancement upon binding to a target oligo. In the free, single-stranded state, thiazole orange has a low fluorescence quantum yield as a result of collisional quenching with solvent upon excitation. (B) Thiazole orange can be tethered to PNA either at the terminus [151] or (C) through modified base pairs [150]. Modifying PNA at a nucleobase position with thiazole orange, typically referred to as forced intercalation (FIT) probes also results in sequence specific fluorescence enhancement. (D) FIT probes can be coupled in a FRET system with NIR-667 dye [249].

Figure 25

Templated fluorogenic detection of oligonucleotides using two PNAs. (A) Templated FRET depends on hybridization of PNAs to adjacent positions on the target sequence to bring the donor and acceptor in proximity. Templated reactions such as (B) Staudinger reaction or (C) conjugate addition of thiols can be used to turn on fluorescence of a caged pro-fluorophore. (D) Photochemical templated reactions target an immolative linker which both tethers and quenches a pro-fluorescent molecule.

Figure 26

Lateral flow devices use a streptavidin labeled strip on nitrocellulose paper to anchor a capture PNA (red). The target oligonucleotide (blue) and the detection PNA probe (red) are then allowed to run the length of the strip. If the target is present, it will act as a hybridization scaffold bringing the two PNAs in proximity. This allows for either simple ligation (A) or fluorogenic ligation (B) which generates an optical signal allowing for detection of the target.