DNA sequence-specific polyamides alleviate transcription inhibition associated with long GAA.TTC repeats in Friedreich's ataxia

Abstract

The DNA abnormality found in 98% of Friedreich's ataxia (FRDA) patients is the unstable hyperexpansion of a GAA.TTC triplet repeat in the first intron of the frataxin gene. Expanded GAA.TTC repeats result in decreased transcription and reduced levels of frataxin protein in affected individuals. Beta-alanine-linked pyrrole-imidazole polyamides bind GAA.TTC tracts with high affinity and disrupt the intramolecular DNA.DNA-associated region of the sticky-DNA conformation formed by long GAA.TTC repeats. Fluorescent polyamide-Bodipy conjugates localize in the nucleus of a lymphoid cell line derived from a FRDA patient. The synthetic ligands increase transcription of the frataxin gene in cell culture, resulting in increased levels of frataxin protein. DNA microarray analyses indicate that a limited number of genes are significantly affected in FRDA cells. Polyamides may increase transcription by altering the DNA conformation of genes harboring long GAA.TTC repeats or by chromatin opening.

Fig. 1.

Polyamide structures, binding models, and nuclear localization in cell culture. (A) Structures of polyamides FA1, FA2, FA3, and FA4 (R, methyl) and their Bodipy conjugates (R, N-propylbutanamide-linked Bodipy FL). Polyamide structures are represented schematically as binding models. Filled and open circles are Im and Py rings, respectively; diamonds are β-alanine; and the semicircle with plus sign is dimethylaminopropylamine. Linear polyamides bind in a carboxyl-to-amino-terminal orientation with respect to the 5′-to-3′ sequence of their DNA target site (18). Mismatches formed with polyamides FA2 and FA4 are indicated with shaded boxes. (B) Deconvolution microscopy of unfixed lymphoid cells (GM15850, derived from an FRDA patient, and GM15851, derived from a healthy sibling) incubated with Bodipy conjugates of each of the indicated polyamides at 2 μM concentration in culture medium for 16 h before visualization, as described (44). (Scale bars, 10 μm.)

Fig. 2.

DNA-binding properties of the polyamides. Quantitative DNase I footprint analysis for polyamide binding to a radiolabeled PCR product containing a (GAA·TTC)₆ repeat sequence, labeled on the purine strand. DNA (at ≈20 pM) and polyamide were allowed to equilibrate for 16 h, with the indicated ranges of polyamide concentrations, before DNase digestion and gel analysis (25). The PhosphorImage of each gel is shown, with undigested DNA in the lane marked “−”; a G+A sequencing reaction of the same DNA is shown, along with DNase-treated DNA in the absence of polyamide (in the lane marked “0”). (A) FA1 targeting 5′-AAGAAGAAG-3′. (B) FA2, mismatch control for FA1. An excerpt of the DNA sequence cloned in pCR2.1 DNA, used to generate the PCR product for footprinting reactions is referenced between A and B.

Fig. 3.

Polyamide FA1 increases the levels of frataxin mRNA and protein in an FRDA lymphoid cell line. (A) Measurement of frataxin mRNA levels, relative to that of GAPDH, in cell lines derived from an unaffected individual (GM15851) and an FRDA patient (GM15850) by qRT-PCR (44). Polyamides FA1 and FA2 were included in the cell culture medium at the indicated concentrations, and frataxin and GAPDH mRNA were determined after 7 days, with media and polyamide replenished on days 3 and 5. Error bars are derived from the percent error of the average and standard deviation of the change in cycle threshold between frataxin and GAPDH for triplicate experiments, with triplicate qRT-PCR determinations for each experiment. (B) Effects of polyamides on frataxin protein in cultured lymphoid cells. FRDA or control cells were incubated as in A before Western blot analysis with antibody to human frataxin or actin. Equivalent amounts of total-cell-extract protein were loaded in each lane.

Fig. 4.

Microarray analysis of polyamide effects on global gene expression. Correlation of significant genes called from a class comparison of all arrays (n = 632 genes at P ≤ 0.005). Cells were cultured with and without the addition of FA1 at 1 μM or 2 μM or FA2 at 2 μM for 7 days before RNA extraction. Each graph represents the difference between the geometric mean of intensities (from an average of the logged robust multichip average data for each condition) (47) of untreated GM15851 cells (denoted 851 in the graphs) and untreated GM15850 cells (denoted 850) plotted versus the difference between the average from all GM15851 geometric means of intensities and each of the individual cell types and conditions. Thus, the untreated controls for GM15851 and GM15850 cells should give slopes for the least-squares-fit line approaching 0 and 1, respectively. Graphs for normal GM15851 cells are shown for the untreated control (A) and cells treated with FA1 at 1 μM (B) and 2 μM (C) and FA2 at 2 μM (D). Graphs for FRDA GM15850 cells are shown for the untreated control (E) and cells treated with FA1 at 1 μM (F) and 2 μM (G) and FA2 at 2 μM (H).

Fig. 5.

Effect of polyamide binding to plasmid DNA on sticky-DNA stability. (A) Illustration showing the assay for influence of polyamides on sticky-DNA conformation. (B) The capacity of a polyamide to disrupt the DNA·DNA-associated region in the sticky-DNA structure was revealed by the formation of linearized pRW4886 rather than the EcoNI-cleaved sticky DNA, which showed the absence of a perturbing influence of the ligands. The polyamides that had higher K_d values required higher concentrations to observe the disruption of the DNA·DNA-associated region of sticky DNA. The polyamides used were FA1 (green), FA2 (blue), FA3 (black), and FA4 (red).