G-quadruplex formation at the 3' end of telomere DNA inhibits its extension by telomerase, polymerase and unwinding by helicase

Abstract

Telomere G-quadruplex is emerging as a promising anti-cancer target due to its inhibition to telomerase, an enzyme expressed in more than 85% tumors. Telomerase-mediated telomere extension and some other reactions require a free 3' telomere end in single-stranded form. G-quadruplex formation near the 3' end of telomere DNA can leave a 3' single-stranded tail of various sizes. How these terminal structures affect reactions at telomere end is not clear. In this work, we studied the 3' tail size-dependence of telomere extension by either telomerase or the alternative lengthening of telomere (ALT) mechanism as well as telomere G-quadruplex unwinding. We show that these reactions require a minimal tail of 8, 12 and 6 nt, respectively. Since we have shown that G-quadruplex tends to form at the farthest 3' distal end of telomere DNA leaving a tail of no more than 5 nt, these results imply that G-quadruplex formation may play a role in regulating reactions at the telomere ends and, as a result, serve as effective drug target for intervening telomere function.

Figure 1.

Extension of telomere by telomerase depends on the size of single-stranded tail at the 3′ side of the farthest distal G-quadruplex on telomere overhang. A telomere tail of less than four T₂AG₃ repeats (0–23 nt) will stay in single-stranded form. Those with tails long enough but unable to form G-quadruplex can be extended (top) while others without or with too short tails may not be extended (bottom).

Figure 2.

Effect of 3′ tail size on the extension of G-quadruplex substrate by telomerase assayed by the TRAP-G4 method. (A) G-quadruplex and control substrates used. In the control substrates, the 5′ distal GG was mutated to TT to prevent G-quadruplex formation. (B) Telomerase products obtained with G-quadruplex substrates (top image) and control substrates (bottom image). (C) Ratio of telomerase activity on G-quadruplex over control substrate as a function of 3′ tail size. Reaction buffer contained 63 mM K⁺.

Figure 3.

Effect of 3′ tail size on the invasion and extension by Klenow polymerase of G-rich telomeric oligonucleotide with G-quadruplex into telomere plasmid. (A) Oligonucleotides used. Lowercase ‘a’ indicates mutation to fix G-quadruplex formation at desired positions (blue letters). Underlined region is for stacking hybridization with a Cy3-labeled probe to visualize uptake on gel. (B) G-quadruplex formation in three representative oligonucleotides (0, 10, 20) revealed by reduced cleavage of guanines (blue diamonds) in the (TTAGGG)₄ region (blue dotted box) in DMS footprinting in K⁺ versus in Li⁺ solution. G-quadruplex forms in K⁺ but not in Li⁺ solution. (C) Gels from top to bottom show plasmid stained by ethidium bromide (EB), uptake of G-rich telomere DNA (labeled with Cy3-probe) and extension by Klenow polymerase (labeled by Cy5-ddGTP incorporation), respectively. The two major bands in EB staining represent the relaxed (RL) and super coiled (SC) plasmid, respectively. The lane ‘P-’ was the same of the 20 nt lane, but without plasmid. (D) Oligonucleotide uptake in relaxed (RL) and super coiled (SC) plasmid normalized by the amount of plasmid. (E) Quantitation of oligonucleotide extension relative to uptake as a function of 3′ tail size. Reaction buffer contained 150 mM K⁺.

Figure 4.

Effect of 3′ tail size on the invasion and extension by Klenow polymerase of G-rich telomeric oligonucleotide without G-quadruplex into telomere plasmid. (A) Oligonucleotides used, which are the same as those in Figure 3 except that the (TTAGGG)₄ region was mutated into (AGTCAT)₄ to abolish formation of G-quadruplex. (B) Gels from top to bottom show plasmid stained by ethidium bromide (EB), uptake of G-rich telomere DNA (labeled with Cy3-probe) and extension by Klenow polymerase (labeled by Cy5-ddGTP incorporation), respectively. The two major bands in EB staining represent the relaxed (RL) and super coiled (SC) plasmid, respectively. The lane ‘P-’ was the same of the 20 nt band, but without plasmid. (C) Uptake of oligonucleotide normalized by the amount of plasmid and ΔG of oligonucleotide hybridization with plasmid (calculated using the online Hetero Dimer tool of OligoAnalyzer at http://www.idtdna.com/). (D) Oligonucleotide extension normalized by the amount of plasmid. Reaction buffer contained 150 mM K⁺.

Figure 5.

Effect of 3′ tail size on the unwinding of G-quadruplex by BLM helicase. (A) Duplex substrate flanked by a 3′ (T₂AG₃)₄ tail capable of forming G-quadruplex in the presence of K⁺. (B) Scheme showing expected effect of G-quadruplex formation on the unwinding of duplex. Unwinding of duplex releases the probe labeled with a fluorescent dye tetramethylrhodamine (TMR). The dye is quenched in duplex by the guanine at the opposite strand and becomes fluorescent upon duplex unwinding. (C) Unwinding kinetics of duplex at various concentrations of K⁺. Li⁺ was supplied to make the final concentration of mono cations to 150 mM. (D) G-quadruplex substrates with single-stranded 3′ tails of 0, 2, 4, 6, 8 and 10 nt and unwinding scheme. Unwinding of G-quadruplex followed by a duplex leads to increase in TMR fluorescence as a result of reporter releasing. (E) Unwinding kinetics of G-quadruplex carrying a 3′ tail of indicated size. Percent of unwinding was calculated as 100 × ΔF/ΔF_max, where ΔF is the increase in fluorescence at a given time and ΔF_max the difference of fluorescence between the reporter in the bound and fully released form (20).

Figure 6.

Measurement of K_F for G-quadruplex formed by G₃(T₂AG₃)₃ on the 3′ end of single-stranded G-rich overhang in a ssDNA/dsDNA hybrid construct by the IDH method (21). (A) Scheme of reactions. A 5-carboxyfluorescein (FAM, star) labeled C-rich DNA probe (purple) was hybridized to the G(T₂AG₃)₂T₂AG₂ region in the G-quadruplex-forming (G21) and the non-G-quadruplex forming (TT-G21) overhangs, respectively, in 150 mM K⁺ solution. Hybridization quenched the fluorescence of the FAM at the 5′ end of the C-rich DNA. (B) Representative hybridization curves from which the dissociation constant K_Dq and K_D was derived, respectively. The K_F was obtained as K_Dq/K_D − 1 (21) and expressed as the mean of three measurements.

Figure 7.

Biological implications of telomere ending structure. Preferential formation of G-quadruplex at the very 3′ end of G-rich overhang is inhibitory to telomere processing that requires a free single-stranded 3′ tail, such as telomere extension by (A) telomerase or (B) the alternative lengthening of telomeres (ALT) and (C) unwinding by activities like helicases (schemes at left). G-quadruplex disruption by proteins may remove the inhibition to permit telomere end processing (schemes at right). The counteracting of G-quadruplex formation and unfolding by proteins may regulate the accessibility of the 3′ telomere end in various processes.