A Near-Infrared Fluorogenic Pyrrole–Imidazole Polyamide Probe for Live-Cell Imaging of Telomeres

Abstract

Telomeres are closely associated with cellular senescence and cancer. Although some techniques have been developed to label telomeres in living cells for study of telomere dynamics, few biocompatible near-infrared probes based on synthetic molecules have been reported. In this study, we developed a near-infrared fluorogenic pyrrole–imidazole polyamide probe (SiR-TTet59B) to visualize telomeres by conjugating a silicon–rhodamine (SiR) fluorophore with a tandem tetramer pyrrole–imidazole polyamide targeting 24 bp in the telomeric double-stranded (ds)DNA. SiR-TTet59B was almost nonfluorescent in water but increased its fluorescence dramatically on binding to telomeric dsDNA. Using a peptide-based delivery reagent, we demonstrated the specific and effective visualization of telomeres in living U2OS cells. Moreover, SiR-TTet59B could be used to observe the dynamic movements of telomeres during interphase and mitosis. This simple imaging method using a synthetic near-infrared probe could be a powerful tool for studies of telomeres and for diagnosis.

Graphical Abstract

INTRODUCTION

Telomeres are specialized chromatin regions located at each end of linear chromosomes that play an important role in maintaining chromosome stability. In human, the TTAGGG repeat sequence is bounded by protein complexes called shelterins, which consist of TRF1, TRF2, POT1, RAP1, TIN2, and TPP1.^1,2 Generally, telomeres shorten with each cell division because of the end-replication problem.^3,4 When the telomere length reaches its lower limit, cells become senescent and the cell cycle is arrested irreversibly.^5,6 However, germ cells and stem cells avoid replicative senescence by elongating short telomeres using a telomerase complex composed of telomerase RNA (TR) and telomerase reverse transcriptase (TERT).^7,8 High telomerase activity is also observed in 85–90% of cancers, which contributes to their potential for abnormal proliferation.⁹ The remaining 10–15% of cancers use the alternative lengthening of telomere (ALT) mechanism to maintain telomere length by homologous recombination, which is characterized by heterogeneous telomere length, telomere clustering and extrachromosomal telomeric DNA.¹⁰ Dysfunctional telomeres induce the DNA repair response and cause telomere fusion.¹¹ This can lead to cancer and other diseases because the bridged chromosomes cannot be segregated correctly.¹² Because of the crucial relevance of telomeres to biological and pathological phenomena, techniques to visualize telomeres not only in fixed cells but also in living cells are in high demand for telomere studies and disease diagnoses.

To date, clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein (Cas) systems, transcription activator-like effectors (TALEs) and TRF1 or TRF2 fused with a green fluorescent protein (TRF1-GFP or TRF2-GFP) have been utilized for live-cell imaging of telomeres.^13–16 However, the methods based on plasmid DNA transfection are not suitable for high-throughput analysis and diagnostic use because of the time-consuming process. Although peptide nucleic acid (PNA) probes have been used previously,¹⁷ methods using synthetic molecules have rarely been demonstrated.

Pyrrole–imidazole polyamides (PIPs), which are composed of N-methylpyrrole and N-methylimidazole, are classified as synthetic DNA-binding molecules. PIPs bind to the minor groove of double-stranded (ds)DNA in a sequence-specific manner by antiparallel pyrrole/pyrrole pairs recognizing A/T or T/A base pairs and pyrrole/imidazole pairs recognizing C/G base pairs.^18–20

Unlike fluorescent in situ hybridization (FISH) methods which require denaturation of target DNA with heat and chemical treatment for their binding in fixed cells,²¹ PIPs can bind to DNA without such harsh treatments, thus maintaining chromatin structures. Utilizing these properties, PIPs and their conjugates to functional molecules have been used to regulate expression of genes of interest in nuclear and mitochondrial DNA.^22–25 Furthermore, fluorescently labeled PIP probes have been employed for visualization of specific regions in chromosomes. Previously, Boutorine and coworkers reported that PIP probes targeting the centromeric sequence yielded punctate labeling patterns in nuclei of living murine cells.^26,27 However, their signal-to-background ratio was low, presumably because the number of base pairs recognized by the PIP probes was small, which led to nonspecific binding. In addition to these centromere-targeting probes, PIP probes that target the telomeric TTAGGG repeat sequence have been developed.^28–33 Our group has reported a tetramethyl rhodamine-labeled tandem tetramer PIP probe (TAMRA-TTet59B, Figure S1), which recognizes 24 base pairs in the telomere sequence.³³ Although this probe can visualize telomeres in fixed cells with a higher signal-to-background ratio than shorter PIP probes, its application to living cells has not been examined.

Here, we report a new telomere-targeting PIP probe (SiR-TTet59B, Figure 1 and S1) optimized for live-cell imaging, which has advantages over previous TAMRA-TTet59B in terms of near-infrared absorption/emission and high fluorogenicity. SiR-TTet59B was synthesized by conjugating a near-infrared silicon-rhodamine (SiR) fluorophore³⁴ to a tandem tetramer PIP (TTet59B) with the highest specificity for telomeres.³³ For live-cell imaging, probes that have excitation and emission wavelengths in the near-infrared region are desirable because of their low phototoxicity and autofluorescence. Moreover, fluorogenicity is also favorable to reduce background signals from unbound probes. Among the SiR probes developed so far,^34–38 some probes are reported to exhibit remarkable fluorescence turn-on properties on binding to their cellular targets.^34–37

Figure 1.

Structure of SiR-TTet59B and a schematic representation of fluorogenic recognition of the telomere sequence.

Consistent with the previous reports, SiR-TTet59B showed a dramatic increase in fluorescence intensity on binding to the telomeric DNA sequence. By using a delivery reagent, we successfully visualized telomeres in living U2OS cells, which are a representative human ALT cell line. To the best of our knowledge, this is the first demonstration of telomere visualization with PIP probes in living human cells.

RESULTS AND DISCUSSION

Design and synthesis of a near-infrared telomere-targeting PIP probe.

As a PIP moiety, we employed a tandem tetramer PIP (TTet59B), which recognizes 24 base pairs in the telomeric TTAGGG repeat sequence, because it showed the highest specificity for telomeres in our previous study.³³ TTet59B was synthesized using a machine-assisted Fmoc solid-phase synthetic method as reported previously.^29–31,33 SiR fluorophore was then coupled to TTet59B to produce SiR-TTet59B (Scheme S1). Analytical HPLC profiles and MALDI-TOF-MS spectra are shown in Figure S2.

Spectroscopic analysis of PIP probes.

To examine its spectroscopic properties, we measured the emission and absorbance spectra of SiR-TTet59B and the previously reported TAMRA-TTet59B in the absence or presence of double-stranded oligodeoxynucleotides (dsODNs) (Figures 2 and S3, respectively). The used dsODNs comprised 32 base pairs of TTAGGG, TCAGGG, and TCAGAG repeat sequences, which respectively have no mismatches, 4 base-pair mismatches and 8 base-pair mismatches within the estimated binding site. Remarkably, SiR-TTet59B was almost nonfluorescent in the absence of dsODNs. However, it showed a dramatic increase, 31.0-fold, in the maximum fluorescence intensity by the addition of the telomeric TTAGGG repeat sequence (Figure 2A). SiR fluoropfore is known to be in equilibrium between a nonfluorescent spirolactone form and a fluorescent zwitterion form, depending on the dielectric constant of the surrounding environment.³⁴ Previous studies suggested that SiR fluorophore of SiR probes forms a nonfluorescent form in water because of the hydrophobic environment offered by the aggregation of SiR probes or by the nonspecific binding to hydrophobic surfaces.^34,35 Based on the studies, it is suggested that the hydrophobic environment generated by intermolecular or intramolecular interaction of SiR-TTet59B maintains SiR fluorophore in its nonfluorescent spirolactone form. These interactions are disrupted on the binding of the PIP moiety to the telomeric sequence, which transforms the dye to its fluorescent zwitterionic form. Supporting the mechanism for fluorescence increase, a blue shift and increase in absorbance maxima were observed after the addition of the telomeric TTAGGG repeat sequence (Figure 2B and Table S1). Compared with SiR-TTet59B, TAMRA-TTet59B was fluorescent even in the absence of ODNs and showed a much smaller fluorescence increase (2.42-fold; Figure S3A). This might be because TAMRA is unlikely to shift to its nonfluorescent spirolactone form in a hydrophobic environment.³⁴ Although SiR-TTet59B was also nonfluorescent in the presence of the TCAGAG repeat sequence, its maximum fluorescence intensity in the presence of the TCAGGG repeat sequence reached about 80% of that in the presence of the TTAGGG repeat sequence. This ratio was higher than that of TAMRA-TTet59B (Figures 2A and S3A), presumably because the hydrophobic interaction would be disrupted even by the weaker interaction between the PIP moiety and TCAGGG repeat sequence, because of the steric configuration of dimethyl silicon.

Figure 2.

(A) Emission and (B) absorbance spectra of 3.0 µM SiR-TTet59B in the absence or presence of 9.0 µM dsODNs with 32 base-pair lengths of TTAGGG, TCAGGG, or TCAGAG repeat sequences. The spectra were measured in 10 mM sodium cacodylate, 10 mM NaCl and 1% dimethyl sulfoxide (DMSO) buffer. The estimated binding site and mismatched base pairs are shown in bold and red, respectively.

Visualization of telomeres in living human cells.

We then applied SiR-TTet59B to live-cell imaging. It is known that PIPs penetrate the cellular membrane and localize in nuclei.³⁹ Therefore, we incubated U2OS cells with SiR-TTet59B without any other reagents. However, no foci were observed in the nuclei; rather, the fluorescence seemed to be observed mainly in lysosomes (Figure 3A and S4). This result suggested that SiR-TTet59B was incorporated into cells by endocytosis and accumulated in lysosomes without endosomal escape. To further examine the cellular uptake mechanism, we suppressed endocytosis by incubating cells at 4 °C. In this experiment, TAMRA-TTet59B was used instead of SiR-TTet59B because of its much stronger fluorescence emission in the absence of the telomeric DNA sequence. As shown in Figure S5, 4 °C incubation significantly reduced the cytoplasmic signals, supporting the cellular uptake mechanism of the PIP probes by endocytosis. The much larger structures of the PIP probes in this study compared with most PIPs²⁴ could prevent its translocation to the nucleus.

Figure 3.

Live cell imaging with SiR-TTet59B in U2OS cells. (A) Live cell images of cells treated with 1.0 µM SiR-TTet59B and 0.3% DMSO without any delivery reagent. Scale bar, 10 µm. (B) Telomere visualization in interphase (upper) and mitosis (lower) with SiR-TTet59B and Endo-Porter. The cells were incubated with 1.0 µM SiR-TTet59B, 6.0 µM Endo-Porter and 0.3% DMSO at 37 °C for 24 h. For the imaging of mitotic cells, the cells were incubated for a further 7 h after the addition of 100 nM nocodazole to arrest cells in early mitosis. Images show the maximum intensity projections of 5 slices (interphase) or one focal plane (mitosis). Enlarged images of a mitotic chromosome from another focal pane are shown in inset. Scale bar, 5 µm.

To resolve this problem, we utilized Endo-Porter, a weak-base amphiphilic peptide, as a delivery reagent.^25,40 This peptide induces the release of compounds from endosomes in acidic conditions by a proton-sponge effect. By simply adding Endo-Porter to media, we could observe numerous clear foci in nuclei of almost all cells observed (Figure 3B and S6A). Especially, the foci in mitotic cells were observed at the ends of chromosomes where telomeres exist. Although cytoplasmic signals observed even after the addition of Endo-Porter seemed to be mainly due to SiR-TTet59B still remaining in lysosomes, some foci were observed in regions other than lysosomes (Figure S6B). This suggests that SiR-TTet59B is able to visualize extrachromosomal telomeric DNA in the cytoplasm.^10,32 The cytotox icity study based on WST assay showed that the treatment with SiR-TTet59B and Endo-Porter is not toxic and does not affect cell proliferation in the concentration and treatment time of this study (Figure S7).

Examination of the telomere-staining ability in living cells.

To confirm the specificity of SiR-TTet59B for telomeres, immunostaining experiments were performed. Living U2OS cells treated with SiR-TTet59B were fixed and subjected to immunostaining with anti-TRF2 antibody. As shown in Figure 4A, the signals for TRF2 and SiR-TTet59B overlapped. Taken together with the results shown in Figure 3B, this result confirmed the specific visualization of telomeres in living U2OS cells. Although living A549 cells, which are one of the human telomerase-positive cell lines, also showed clear telomere signals, a few nonspecific foci were observed inside the nuclei of most cells (Figure S8). These nonspecific foci were not observed in U2OS cells and seem to be easily discernable from telomere foci because of their much larger size. In addition to the immunostaining for TRF2, we also performed Telomere dysfunction-Induced Foci (TIFs) assay^2,41 using anti-γH2AX antibody to investigate whether SiR-TTet59B affects telomere function. Consistent with the fact that ALT cells have spontaneous TIFs,⁴² U2OS cells showed the overlap of SiR-TTet59B foci with γH2AX foci. In contrast, these foci seemed not to overlap in A549 cells (Figure S9). Therefore, it is suggested that the binding of SiR-TTet59B to telomeric DNA does not cause telomere dysfunction.

Figure 4.

(A) Colabeling of telomeres with SiR-TTet59B and anti-TRF2 antibody. The living U2OS cells treated with SiR-TTet59B were subjected to immunostaining. Scale bar, 10 µm. (B) SiR-TTet59B foci in U2OS cells. The cells treated with SiR-TTet59B were fixed before image acquisition. Maximum intensity projections of the z-stack images were used for analysis. (n = 59).

Next, we counted telomere foci in interphase nuclei to investigate the effectiveness of telomere staining in living U2OS cells. As previously reported, telomeres move rapidly, and some telomeres can be in close proximity to each other.^13,16,43 For precise analysis, the telomere-stained cells were fixed and z-stack images were acquired. In this experiment, some cells showed much more foci than the maximum of 70 seen in previous studies where TALE and PNA were used to label telomeres (Figure 4B).^15,17 Although these results should not be directly compared because the number of chromosomes is highly altered in U2OS cells,⁴⁴ our results indicated that SiR-TTet59B visualizes telomeres effectively.

Telomere length can be a potential biomarker for cancer diagnosis and cell-proliferation ability. We next stained telomeres in HeLa S3 and HeLa 1.3 cells, whose telomere lengths are 2–10 kb and ~23 kb, respectively.^45,46 Visible foci were brighter in HeLa 1.3 than in HeLa S3, reflecting the difference of their telomere length (Figure S10).This result suggested the possibility that SiR-TTet59B can stain telomeres in their length-dependent manner.

Observation of telomere dynamics in mitosis and interphase.

Finally, we used SiR-TTet59B to observe telomere dynamics. During mitosis, telomeres show the most dynamic movement for chromosome segregation. Because the population of mitotic cells is generally small, we synchronized cells in early mitosis using nocodazole, which is a reversible inhibitor of microtubule polymerization. By time-lapse imaging, the progress of cells from prometaphase to cytokinesis was successfully monitored (Figure 5A and Movie S1). The signals remaining after chromosome decondensation revealed that the SiR-TTet59B could be used to observe telomeres during mitosis and dynamic changes in chromatin structure. During the observation, we noticed that extrachromosomal signals increased gradually. So far, it has been reported that lysosomal macromolecules can be converted to autofluorescent molecules with wide-range emission through oxidative modification within lysosomes.⁴⁷ We speculate that SiR-TTet59B remaining in lysosomes generates reactive oxygen species from repeated laser irradiation and promotes the generation of the lysosomal autofluorescent molecules.

Figure 5.

Observation of telomere dynamics. (A) Time-lapse imaging of mitosis. U2OS cells arrested with nocodazole in prometaphase were released after washing and observed. The Z-stack images were acquired every 8 min and the maximum intensity projections are shown. Some telomere and extrachromosomal foci in a first image are indicated by white and green arrows, respectively. Scale bar, 5 µm. (B) High-speed time-lapse imaging of U2OS cells in interphase. The cells treated with SiR-TTet59B were observed without nuclear staining. The images were acquired every 194 msec on average for a total of 1000 frames. White arrows indicate merging or merged foci. Scale bar, 2 µm.

To observe the relatively small but rapid movement of interphase telomeres, high time-resolution imaging, which often suffers from a low signal-to background ratio, is required. The near-infrared excitation and emission wavelengths and fluorogenicity of SiR-TTet59B were expected to be useful for the high time-resolution analysis because these properties lead to a reduction in the level of background signals. High-speed time-lapse imaging (every 194 msec on average) captured the moment that two separated foci merged, suggesting the association of telomeres in ALT cells (Figure 5B and Movie S2).^17,48 This result implied that the binding of SiR-TTet59B to telomeric dsDNA did not prevent telomere association.

CONCLUSION

We have developed a telomere-targeting near-infrared fluorogenic PIP probe, SiR-TTet59B, which offers advantages over previously reported TAMRA-TTet59B in terms of low phototoxicity and background signals. Using SiR-TTet59B, we demonstrated specific and effective visualization of telomeres in living U2OS cells. Moreover, SiR-Tet59B could be used to observe telomere dynamics in mitotic and interphase cells. The telomere-staining method described here is simple and not based on plasmid DNA transfection. Therefore, it has the potential for versatile applications. Furthermore, because PIPs are tunable to target any DNA sequence by changing the arrangement of pyrrole and imidazole, this study can be expanded to develop near-infrared fluorogenic probes that target other specific regions of interest in the genome.