Specific targeting of insect and vertebrate telomeres with pyrrole and imidazole polyamides

Abstract

DNA minor groove-binding compounds (polyamides) that target insect and vertebrate telomeric repeats with high specificity were synthesized. Base pair recognition of these polyamides is based on the presence of the heterocyclic amino acids pyrrole and imidazole. One compound (TH52B) interacts uniquely and with excellent specificity (K(d) = 0.12 nM) with two consecutive insect-type telomeric repeats (TTAGG). A related compound, TH59, displays high specificity (K(d) = 0.5 nM) for tandem vertebrate (TTAGGG) and insect telomeric repeats. The high affinity and specificity of these compounds were achieved by bidentate binding of two flexibly linked DNA-binding moieties. Epifluorescence microscopy studies show that fluorescent derivatives of TH52B and TH59 stain insect or vertebrate telomeres of chromosomes and nuclei sharply. Importantly, the telomere-specific polyamide signals of HeLa chromosomes co-localize with the immunofluorescence signals of the telomere-binding protein TRF1. Our results demonstrate that telomere-specific compounds allow rapid estimation of relative telomere length. The insect-specific compound TH52 was shown to be incorporated rapidly into growing Sf9 cells, underlining the potential of these compounds for telomere biology and possibly human medicine.

Fig. 1. Schematic structure of telomere compounds used in this study. N-methylpyrrole rings are represented by open circles. The flexible, hydrophilic linker (8-amino-3,6-dioxaoctanoic acid) is indicated by a boxed L. The C-terminal dimethylaminopropylamide (Dp) is represented by a plus sign. A diamond represents β-alanine. N-methylimidazole is represented by a black circle. The γ-turn monomer (R-2,4,-diaminobutyric acid) is indicated by a curved line. The solid wedge attached to the plus sign (in the γ-turn monomer) represents the amino substitution at C2. Chemical structures of the symbols used are shown at the bottom.

Fig. 2. DNase I footprinting experiments of the telomere-specific compounds. Ligand concentrations (in nM) are indicated at the top of each lane. The letter G refers to a G nucleotide cleavage reaction. (A) The monomeric hairpin H64 or its homodimer TH52A preferably binds TTAGG repeats (the bottom C-rich strand that contains AATCC is labeled) but also protects several mismatch sequences on the vector at higher concentrations. TH52A possesses increased affinity but unimproved specificity. (B) TH52B possesses excellent specificity for TTAGG repeats. Again, the bottom C-rich strand is shown. Unlike H64 and TH52A, high specificity is retained even at a concentration several hundred times that required for protection. The position of the band that is not protected in the CCTAA repeats by TH52B corresponds to the adenine base positioned on the linker. (C) Footprint of TH52B on the top G-rich strand of the same probe as in (A) and (B), showing that TH52B protects TTAGG repeats at subnanomolar concentrations. (D) Insertion of an unpaired imidazole in compound TH57 coupled to the linker generates binding preference for TTAGGG repeats over TTAGG repeats. The bottom C-rich strand is labeled. (E) When an imidazole is also inserted at the same position in the C-terminal hairpin (left), the resulting compound (TH59) has affinity similar to TH57 and binds TTAGG and TTAGGG repeats equally well. (F) The TH59 ‘monomer’ H60 shows lower specificity for telomeric repeats than the tandem hairpin TH59, but similar affinity.

Fig. 3. Putative binding model of TH59. Proposed binding model for the complex of TH59 with 5′-AGGGTTAGGGTT-3′. Circles with two dots represent lone pairs of electrons on N3 of purines and O2 of pyrimidines at the edges of the bases. Circles containing an H represent the N2 hydrogen of guanine. Putative bifurcated hydrogen bonds to the amide NHs are illustrated by dashed lines.

Fig. 4. Staining of insect-type telomere repeats (TTAGG) with TH52B-T. Chromosomes or nuclei prepared from Sf9 and HeLa cells were co-stained with TH52B-T (red) and DAPI (blue). Note that TH52B-T sharply highlights red foci in Sf9 (B) but not HeLa nuclei (C). The two images (B and C) were obtained under identical conditions and are shown on an identical intensity scale. TH52B-T also stains the ends of chromosomes, as observed in metaphase spreads derived from Sf9 cells (A). A number of non-telomeric signals can also be noted (black and white inset). Scale bars represent 5 µm.

Fig. 5. Staining of vertebrate telomeric repeats (TTAGGG) with TH59-T and co-localization with TRF1. Chromosomal material prepared from vertebrate cell lines was stained with polyamide TH59-T (red) and DAPI (blue). Note that TH59-T sharply highlights red foci in HeLa cell nuclei (A) and metaphase chromosomes derived from Indian Muntjac (B), X.laevis (C) or HeLa cells (D). (E) TH59-T foci co-localize with green telomere spots revealed by indirect immunofluorescence with TRF1-specific antibodies. The black and white inset shows the green and red signals separately. Scale bars represent 5 µm.

Fig. 6. Quantitative aspect of vertebrate telomere signals. Metaphase chromosome spreads were prepared from human lymphocytes and HeLa-L cells, a HeLa subclone with long telomeres. Subregions of metaphase spreads are shown derived from human lymphocytes (A) and HeLa-L cells (B). The two images (A and B) were obtained under identical conditions and are shown on the same intensity scale. For quantification, the telomere foci were contoured appropriately (indicated) and the total integrated signal intensities were then determined for entire spreads. The total telomere signal intensity distribution obtained for different chromosomal spreads is shown in Figure 7. The fractional amount of polyamide bound at telomeres relative to that encompassing the chromosomal body was determined. An example of this is shown for HeLa-L chromosomes (B, C and D). The telomere spots contoured in (B) were extracted to yield (D). The image lacking telomere spots was then contoured again using a lower threshold level so as to encompass the chromosomal bodies (C). (C) and (D) are shown on a strongly enhanced intensity scale to visualize the staining of chromosomal bodies. The total integrated signal derived from TH59-T encompassing either telomeres (D) or chromosomal bodies (C) was then obtained and corrected for general background level. Scale bar represents 5 µm.

Fig. 7. Distribution of telomere signals of different chromosomal spreads. Metaphase chromosome spreads were prepared from human lymphocytes and two closely related HeLa subclones (HeLa-L and HeLa-S) that differ with regard to telomere length. The bar graph shows the intensity distribution of the telomere signals obtained for the different chromosomal spreads. Note that the distribution of the telomere signals is strongly skewed toward higher intensity for telomeres of HeLa-L chromosomes as compared with those obtained from HeLa-S and lymphocytes. The weight averages of these distributions were also determined and are shown in the inset.

Fig. 8. TH52B-T rapidly highlights telomere foci of Sf9 cells in vivo. TH52B-T and Hoechst 33258 were added to the media of growing Sf9 cells at different concentrations and periods of incubation. Cells were then fixed by cold methanol and examined by fluorescence microscopy to determine the strength of the TH52B-T telomere signals and that of Hoechst 33258 (general DNA stain). (A) The average telomere signal intensity obtained after 16 h incubation in media at the indicated concentrations of TH52B-T. (B) A time course of incorporation. The averages of the telomere signal intensity of TH52-T (telomere foci, triangles), Hoechst 33258 (general stain, diamonds) and Texas Red alone (background, circles) are shown. Note that specific foci were already observed after a 15 min exposure of Sf9 cells to TH52B-T.