Telomere Targeting Chimera Enables Targeted Destruction of Telomeric Repeat-Binding Factor Proteins

Abstract

Telomeres are naturally shortened after each round of cell division in noncancerous normal cells, while the activation of telomerase activity to extend telomere in the cancer cell is essential for cell transformation. Therefore, telomeres are regarded as a potential anticancer target. In this study, we report the development of a nucleotide-based proteolysis-targeting chimera (PROTAC) designed to degrade TRF1/2 (telomeric repeat-binding factor 1/2), which are the key components of the shelterin complex (telosome) that regulates the telomere length by directly interacting with telomere DNA repeats. The prototype telomere-targeting chimeras (TeloTACs) efficiently degrade TRF1/2 in a VHL- and proteosome-dependent manner, resulting in the shortening of telomeres and suppressed cancer cell proliferation. Compared to the traditional receptor-based off-target therapy, TeloTACs have potential application in a broad spectrum of cancer cell lines due to their ability to selectively kill cancer cells that overexpress TRF1/2. In summary, TeloTACs provide a nucleotide-based degradation approach for shortening the telomere and inhibiting tumor cell growth, representing a promising avenue for cancer treatment.

Graphical Abstract

INTRODUCTION

Normal cells undergo replicative senescence due to telomere shortening coupled with each cell cycle division, while cancer cells can gain infinite replication ability by activating telomerase, which extends telomere ends^1-6 (Scheme 1A). Hence, targeting the telomerase function serves as a straightforward and efficient approach to retard cancer deterioration.^1-6 In this respect, multiple therapies have been developed to target telomerase for cancer treatment, through either directly inhibiting telomerase activity by using small molecules^7,8 and oligonucleotide inhibitors^9-13 or indirectly disrupting the precise localization of the telomerase complex on the telomere by interrupting shelterin complex recruitment onto the TTAGGG repeats.¹⁴ However, the challenges are still inevitable, mainly restrained by the relatively low efficiency and adaptive drug resistance.^2,4 Imetelstat, which is the first nucleotide drug designed to target telomerase, is currently being investigated for clinical efficacy. Nevertheless, due to its side effects, further research is needed to fully evaluate its effectiveness and safety.^9-13 Therefore, there is still a need to develop potential strategies to overcome issues with undruggable telomerase. We, along with other groups, have developed nucleotide-based proteolysis-targeting chimera (PROTAC) platforms, which could degrade undruggable RNA-binding protein,¹⁵ transcription factors,^16-18 and G-quadruplex binding protein.¹⁹ Of note, nucleotide drugs have relatively low efficiency compared to small molecular inhibitors. Thus, the emerging PROTAC technology might provide a novel avenue to overcome this low-efficiency issue. However, it remains elusive whether such nucleotide-based PROTAC could be used for other DNA-binding proteins.

Scheme 1. Role of TRF1/2 in Maintaining Telomeres^a.

^a(A) Functional relationship between the telomere length and cellular senescence. (B) Structure analysis (PDB: 1W0T) to illustrate the interaction between TRF1 and telomere repeats (TTAGGG)n.

We propose the telomere-targeting chimera (TeloTAC) methodology to expand the application of DNA PROTAC to non-transcription factor proteins, telomeric repeat-binding factor 1/2 (TRF1/2). The TRF1/2 proteins are major components of the shelterin subcomplex and are responsible for specific binding with the TTAGGG repeats within the telomere, forming several critical hydrogen bonds between Lys421 (Lys488 for TRF2) with G4, Asp422 with C5′ (in the antisense strain, Asp489 for TRF2), and Arg425 with G5 (Arg492 for TRF2), as shown in Scheme 1B.^1,20-22 Given that TRF1/2 plays a bridging role in telomere–telomerase interactions, directly binding to the telomere end on one side and recruiting telomerase on the other, we designed the TeloTAC strategy using the TRF1/2 as a degron to induce the collapse of the shelterin complex and subsequently disrupt telomerase binding to the telomere end. TeloTAC presents a novel strategy to target the telomerase activity essential in cancer cells by degrading the structural proteins TRF1/2 instead of directly inhibiting the enzymatic activity of telomerase.

To achieve the targeted degradation of TRF1/2, we designed several prototype TeloTAC molecules by conjugating the TRF1/2-binding motif (TTAGGG) with the von Hippel–Lindau (VHL) ligand. The TeloTAC is constructed via the copper-free strain-promoted azide–alkyne cyclo-addition (SPAAC) reaction²³ between azide-modified oligonucleotide N₃-(TTAGGG)₃ and bicyclooctyne (BCN)-containing VHL ligands with linkers of different types and lengths, using a similar strategy as described previously.¹⁷ Upon TeloTACs entering the cells, the oligonucleotide moiety directly binds to TRF1/2 proteins, while the VHL ligand moiety recruits the VHL E3 ligase to facilitate the ubiquitination and subsequent degradation of TRF1/2 (Scheme 2).

Scheme 2. TeloTAC Induces TRF1/2 Ubiquitination and Subsequent Degradation by Recruiting an E3 Ligase.

RESULTS AND DISCUSSION

Design of TRF-ODN as TRF1/2 Ligands.

We first designed a ligand consisting of three TTAGGG tandem repeats that bind to the TRF1/2 proteins, named TRF-ODN. To create TRF-ODN, a single-stranded DNA oligomer (5′-GGGTTAGGGTTAGGGTTATTTTAACCCTAACCCTAACCC-3′) was synthesized by inserting a “-TTT-linker”, which forms a stem-loop structure that stabilizes the DNA oligomer, as previously described^1,20-22 (Figure 1A). To evaluate the binding affinity of TRF-ODN to TRF1/2 proteins, we carried out the streptavidin-biotin pull-down assay using the biotinylated DNA oligomer as a bait¹⁷ (Figure 1B,C). As expected, the biotin TRF-ODN (oligodeoxynucleotide) bound with both TRF1 and TRF2, which could be competitive with free TRF-ODN, indicating the specificity of binding (Figure 1D,E).

Figure 1.

Specific binding of TRF-ODN with TRF1/2 proteins. (A) ODN-binding sites of TRF1/2, and the biotin-TRF-ODN sequence. (B) Schematic protocol of the biotin-TRF-ODN pull-down assay. (C) PAGE analysis of the annealed TRF-ODN. (D) TRF1-ODN binds to FLAG-TRF1 extracted from HEK293 cells, monitored with immunoblotting using FLAG and TRF1 antibodies. (E) TRF-ODN binds to Myc-TRF2 extracted from HEK293 cells, monitored with immunoblotting using Myc and TRF2 antibodies.

TeloTAC Induces TRF1/2 Degradation in Multiple Cancer Cell Lines.

Using TRF-ODN as a ligand, we constructed PROTACs for TRF1/2, hereafter referred to as TeloTAC degraders, by conjugating them onto the VHL E3 ligand through the SPAAC reaction¹⁷ (Figure 2A). Given the importance of the linker type and length that determine the right position between E3 and the substrate in the PROTAC ternary structure,^24-26 we synthesized a series of VHL ligands, namely, compounds 1a–18a¹⁷ as previously described by coupling a bicyclooctyne group with diverse linkers onto VHL1.²⁷ After conjugation of compounds 1a–18a onto azide-modified TRF-ODN for 16 h at room temperature to attain the TeloTAC degrader 1b–18b, the click efficiency was monitored by PAGE, and most of these compounds achieved >80% of yields except 7a–10a (Figure 2A,B). In HeLa cells, treatment with compounds 15b–17b (5 μg/mL, 24 h) resulted in a significant degradation (Figure 2C), while compounds 1b, 2b, and 6b–10b showed relatively weaker effects. The weak effects of 1b and 2b could be attributed to their short linker length, while the increased hydrophobicity derived from the long alkane chain in 6b–10b could also have contributed to the weaker effects. These results suggest that optimizing linker lengths and amphipathicity is crucial for degradation efficiency. This also explains why the 3b–5b work less efficiently than other alkane degraders, due to an insufficient length and higher hydrophobicity. As for compounds 11b–13b, the lower efficiency may be due to the linker conformations, which might cause undesired positioning between VHL E3 ligase and TRF.¹⁷

Figure 2.

TeloTAC targets TRF1 and TRF2 for degradation in a VHL- and proteasome-dependent manner. (A) Schematic illustration of the SPAAC reaction between BCN-modified VHL ligands 1a–18a with different linker types and lengths and azide-modified DNA (N₃-TRF-ODN) for the synthesis of TeloTACs 1b–18b. (B) PAGE analysis of the click efficiency. The SPAAC reaction products (1 μg) were subjected to 20% PAGE at 100 V for 1 h. (C) Screening of the TeloTAC that degrades TRF2 and TRF1 in HeLa cells. HeLa cells were treated with 5 μg/mL TeloTAC (1b–18b) for 24 h, and the cells were harvested for western blotting analysis of TRF1 and TRF2. (D) Competition experiments of TRF-ODN, VH032, and MG132 toward 15b in degrading TRF2. HeLa cells were treated with 15b (5 μg/mL) with or without TRF-ODN (5 μg/mL), VH032 (5 μM), or MG132 (2.5 μM) for 24 h, and the cells were harvested for western blotting analysis of TRF1 and TRF2. (E) Targeted degradation of TRF1 and TRF2 by the TeloTAC in the A431 cell line. TeloTACs 6b, 14b, and 15b mediated TRF1 and TRF2 degradation. A431 cells were treated with 5 μg/mL TeloTACs 6b, 14b, and 15b for 24 h and then harvested for western blot analysis of TRF1 and TRF2. (F) TeloTAC 6b, 14b, and 15b treatment led to the death of the A431 cancer cell. A431 cancer cells expressing mCherry were treated with TRF-ODN, 6b, 14b, and 15b TeloTAC in 5 μg/mL for 24 h, and then cell imaging was acquired for analyzing cell numbers. NC, negative control.

Apart from TRF2, the 18 TeloTACs showed almost the same tendency of degradation efficiency toward TRF1 (Figure 2C). Moreover, the degradation of TRF2 by 15b (5 μg/mL) was blocked by free TRF-ODN (5 μg/mL), demonstrating that the specific degradation of TRF is likely in a TRF-ODN-dependent manner (Figure 2D). Furthermore, TeloTAC 15b-mediated degradation of TRF2 was blocked by the proteasome inhibitor MG132 (2.5 μM) or the VHL ligand VH-032 (5 μM), indicating that TeloTAC-mediated TRF degraded in TRF-ODN-, VHL-, and proteosome-dependent manners (Figure 2D). In addition to HeLa cells, TeloTACs 6b, 14b, and 15b were also effective in degrading TRF1 and TRF2 in A431 and MDA-MB-231 cells (Figures 2E and S6, other different kinds of cancer cell lines are shown in Figure S7). As a result, TeloTACs 6b, 14b, and 15b (at a concentration of 5 μg/mL for 24 h) suppressed the proliferation of A431 cells (as shown in Figure 2F), indicating a promising anticancer effect of these TeloTAC compounds. Moreover, we sought to investigate whether the degradation of TRF1 and TRF2 could affect the telomerase’s physiological function. To achieve this purpose, the telomere length was monitored using qPCR assay as described before.²⁸ The results showed that the telomere length was significantly shortened after treatment with 14b and 15b in MDA-MB-231, A431, and HeLa cells (Figure S8).

TeloTAC-Mediated TRF1/2 Degradation Leads to Selective Anticancer Activity.

To further investigate the antiproliferation activity of the TeloTAC, we performed cell growth curves, colony formation, and MTT assays in multiple cancer cell lines, including MDA-MB-231, A431, and HeLa cells. First, 5 μg/mL TRF-ODN, 14b, and 15b suppressed the cell proliferation of MDA-MB-231, A431, and HeLa cells (Figure 3A). For comparison, we also treated these cell lines with 2 μM telomerase inhibitor RHPS4^29-31 and BIBR1532.^8,32-34 Notably, the inhibitory effect of TeloTACs 14b and 15b were stronger than TRF-ODN and two inhibitors (Figure 3A). Next, the colony formation assay was performed to evaluate the effect of TeloTAC on cancer cell transformation abilities in vitro. Consistent with the results of cell viability assays (Figure 3A), TeloTAC 14b and 15b dramatically inhibited the colony formation potential (Figure 3B), whereas TRF-ODN had relatively low efficiency. These results suggest that TRF-ODN might act as a DNA decoy to suppress the TRF function, while TeloTAC leads to the destruction of TRF protein, thus being more potent than TRF-ODN.

Figure 3.

TeloTAC-mediated TRF1/2 degradation leads to cell death in tumor cells. (A) TeloTACs 14b or 15b suppressed the proliferation of MDA-MB-231, A431, and HeLa cells. Cells were treated in 5 μg/mL 14b or 15b for 3 days, and the cell number was counted daily. (B) TeloTACs 14b or 15b inhibit the colony formation of HeLa cells. HeLa cells were treated with 5 μg/mL TeloTACs 14b and 15b, and the colony number was quantified 2 weeks after the treatment. (C) IC₅₀ of TeloTAC 15b. HeLa cells were treated with the indicated dose of 15b for 72 h, and the cell viability was measured (by CCK-8 assay). (D) Coculture model of HeLa (labeled in green by GFP) and LF1 (labeled in red by mCherry) for the specific killing analysis of 15b; 5 μg/mL 15b and 2 μM RHPS 4 were added into the cocultured cells. After 24 h, the imaging was performed on a 10× microscope with both green and red channels, 100 μm.

Given that cancer and noncancerous normal cells have different dependencies on telomerase for survival, we further determine the effect of TeloTACs 14b and 15b in two additional normal cell lines, LF1 and MCF10A We found that TeloTACs 14b and 15b were effective in degrading TRF1/2 (Figures S10 and S12) but had no noticeable effects on cell proliferation (Figures S9 and S11). In line with this notion, TeloTAC 15b had a half maximal inhibitory concentration (IC₅₀) of approximately 0.2, 0.3, and 0.39 μM on MDA-MB-231, A431, and HeLa cancer cells, respectively, while the IC₅₀ of 15b on noncancerous normal cells was about 10-fold higher (Figures 3C and S13). On the other hand, the IC₅₀ of the telomerase inhibitors RHPS4 and BIBR1532 was about 100-fold higher than that of 15b on cancer cells (Figures 3C and S13). To further confirm the selectivity of TeloTAC on cancer cells rather than normal cells, we also utilized a coculture system³⁵ consisting of GFP-labeled HeLa cells (cancer cells) and mCherry-labeled LF1 cells (noncancerous normal cells) to show the specific effect of TeloTAC 15b on cancer cells (Figure 3D). Figure 3D shows that TeloTAC 15b selectively killed cancer cells but not noncancerous normal cells, whereas the telomerase inhibitor RHPS4 showed nonselective cytotoxic effects on both types. Together, these results demonstrated that 15b exhibited selective anticancer activity toward cancer cells while sparing normal cells, providing a therapeutic window for the TeloTAC.

TeloTAC Rewires Apoptosis and Senescence Signaling.

Next, we attempted to delineate the TeloTAC-induced cell death mechanism using cancer cells and noncancerous normal cell lines. Previous reports show that TERT, TRF1, and TRF2 are highly expressed and essential for the proliferation of cancer cells, while noncancerous normal cells do not rely on telomerase for survival.^1-6,36,37 These findings advocate that TRF1 and TRF2 are promising anticancer targets. To this end, we compared the levels of expression of TRF1/2 mRNA and protein levels in multiple cancer and noncancerous normal cells and found that cancer cells exhibited relatively higher levels of TRF1/2 mRNA and protein than noncancerous normal cells (Figure 4A). Since TeloTAC significantly shortened the telomere length in cancer cells (Figure S8), we next investigated whether TeloTAC treatment affects cellular senescence and apoptosis signaling pathways in cancer and noncancerous normal cells. As a result, treatment with TeloTAC (6b, 14b, and 15b) increased the levels of senescence markers p53 and p27 in cancer cells but not in noncancerous normal cells (Figure 4B). Additionally, TeloTAC triggered apoptotic caspase activations in cancer cells but not in normal cells, as evidenced by the increased levels of cleaved PARP1 and Caspase 3 (Figure 4C). Thus, our data suggest that TeloTAC-triggered TRF1/2 degradation and subsequent telomere shortening rewire the senescence and apoptosis signaling pathways to suppress the proliferation of cancer cells but not normal cells (Figure 4D).

Figure 4.

Mechanism study of TRF-ODN, 6b, 14b, and 15b rewired cell signaling on cancer cells compared to noncancerous normal cells. (A) Western blot analysis of whole-cell lysates derived from the indicated cell lines (left) and the Kaplan–Meier survival curves of breast cancer patients based on TERF2 mRNA expression (right). (B) Redirected senescence pathway in cancer cells caused by TeloTAC-mediated TRF1/2 degradation. p53 cannot be detected in some cell lines. (C) Rewired apoptosis pathway in cancer cells resulted from TeloTAC-mediated TRF1/2 degradation. (D) Reported apoptosis and senescence mechanism of cancer and noncancerous normal cells, as well as the proposed mechanism of the TeloTAC in cancer cell death.

CONCLUSIONS

In conclusion, tumor cells are addicted to high telomerase activity to maintain telomeres and thereby avoid cellular senescence, which represents one of the cancer cell hallmarks.^1-6 In this study, we developed a proof-of-concept TeloTAC to selectively degrade TRF1 and TRF2 and perturb the telomerase’s activity, thus providing a new approach to target telomeres in cancer cells. The prototype TeloTAC 15b effectively degraded TRF1 and TRF2, resulting in significantly higher cytotoxicity in cancer cells compared to normal cells. Hence, this TeloTAC method is valuable for a novel cancer-targeting PROTAC strategy. However, due to the nature of these nucleotide oligomer-based designs, the TeloTAC has several limitations, especially in future clinical translation. First, the TeloTAC has a relatively large molecular weight, usually over 10,000 Da, which might cause cellular permeability issues, and thus, a special delivery approach might be needed. In this regard, a nucleotide medicine delivery approach could be also applied to the TeloTAC.³⁸ Second, the native nucleotide backbone of the TeloTAC is vulnerable to intracellular nuclease, which leads to the degradation of the TeloTAC. This shortcoming could be overcome by replacing the native nucleotide with a modified one, such as phosphorothioate modification. Further optimization of the molecular structure is needed to achieve more efficient degradation. Last, the selective degradation of TRF1 and TRF2 is difficult to be achieved in the present design, as these two proteins share highly identical binding sites toward telomere DNA. This might be solved by establishing a DNA library to screen the best binder for each protein, which will provide a much more applicable tool for DNA-induced degradation in the future. Overall, the methodologies outlined in this study will provide an advantageous approach to drug discovery for telomeres, potentially overcoming the challenges of undruggable and drug-resistant targets. The TeloTAC demonstrates an original design to target critical oncogenic proteins while sparing noncancerous normal cells and provides a proof of concept that TERT, TRF1, and TRF2 function as promising cancer-specific targets, which shows potential to be developed into broad applications in the future.