A Stapled Peptide Probe of Replication Protein A’s Protein-Protein Interactions
By: Carol A. Rouzer, VICB Communications
Published: February 12, 2014
The innovative incorporation of an unnatural amino acid yields a stapled peptide that binds tightly to RPA’s protein-protein interaction domain.
DNA damage is a hallmark of cancer, and many cancer treatments, including radiation therapy and most chemotherapeutic agents, inflict additional DNA damage. Consequently, DNA damage response pathways are often activated in cancer cells and may contribute to the development of resistance to treatment. This has led to the hypothesis that inhibition of the DNA damage response and subsequent DNA repair may be an effective therapeutic target for some cancers. To test this hypothesis requires probe molecules that disrupt DNA damage response and repair pathways in intact cancer cells. Now, VICB members Stephen Fesik, Walter Chazin, David Cortez, and Alex Waterson report the discovery of a potent stapled peptide inhibitor of replication protein A, a key player in the DNA damage response [A. O. Frank, et al. (2014) J. Med. Chem., published online February 3, doi:10.1021/jm401730y].
RPA is a single-stranded DNA (ssDNA) binding protein comprising 70 kDa, 32 kDa, and 14 kDa subunits (Figure 1). The protein contains four oligonucleotide binding (OB) folds, designated A, B, C, and D, that bind ssDNA in decreasing order of affinity. RPA binds to ssDNA that is exposed during replication, DNA damage, and DNA repair. It protects the DNA from digestion by nucleases and prevents hairpin formation. The N-terminal domain of RPA’s 70 kDa subunit (RPA70N) also contains an OB fold that weakly binds ssDNA. This fold contains a basic cleft that serves primarily as a docking site for multiple proteins. Thus, RPA also provides a focal point for the assembly of DNA replication and repair complexes. Binding of proteins, such as ATRIP, RAD9, MRE11, and p53, to RPA70N is critical to the DNA damage response and DNA repair. Thus, blocking these interactions may have a therapeutic benefit in cancer. However, it is likely that complete inhibition of all of RPA’s functions would be highly toxic. This led the VICB team to search for probe molecules that block only RPA70N-mediated protein-protein interactions. Their efforts have already produced small molecules with submicromolar affinity that bind to this domain [A. O. Frank, et al. (2013) J. Med. Chem., published online October 22, doi:10.1021/jm401333u]. In this most recent work, they report the discovery of a potent stapled peptide that binds to the basic cleft of RPA70N with high affinity.
Figure 1. Ribbon model of the structure of RPA domains showing the portions of the molecule for which crystal structure data are available. The dotted lines connecting these domains indicate flexible linkers that are the source of the dynamic architecture of the complete protein. Image provided by Walter Chazin.
Prior studies had shown that peptides derived from ATRIP, RAD9, and MRE11 all bind competitively to the basic cleft of RPA70N. X-ray crystallographic data revealed that the region of p53 that interacts with RPA70N’s basic cleft is folded into a helix, suggesting that a stable helical peptide based on the sequence of the RPA70N binding partners should efficiently block these protein-protein interactions. The investigators chose to start with the sequence of the ATRIP-derived peptide (Table 1), since it had the highest affinity of those studied to date.
To improve the binding affinity of the ATRIP peptide, the researchers began by replacing each of the amino acids with alanine. The results indicated that the D1, F2, T3, D6, L7, E9, L10, D11, and L13 amino acids were all important contributors to binding, as substitution of any of these residues reduced affinity substantially. In contrast, substitution of E8 or T12 with alanine increased affinity slightly, while substitution of D5 had essentially no effect.
Table 1. Sequences of RPA70N-binding peptides. Colored amino acids indicate substitutions made during peptide optimization. Z indicates 3,4-dichlorophenylalanine. X indicates α-(4-pentenyl)alanine. Values are provided for the binding affinity (Kd) of peptides to RPA obtained from a fluorescence polarization anisotropy assay. Ac- indicates an N-terminal acetyl group. FITC- indicates an N-terminal fluorescein group. -NH2 indicates a C-terminal amide.
The p53-derived RPA70N binding peptide contains a WF sequence at positions 10 and 11 (Table 1). These amino acids form a hydrophobic binding core, and the researchers hypothesized that introduction of this core into the ATRIP peptide would lead to an increase in affinity. This proved to be the case, as the resulting Peptide-19, which also included an alanine substitution at position 12, exhibited a 24-fold increase in potency (Table 1). Addition of a fluorescein tag at the peptide’s N-terminus (Peptide-20) increased the potency an additional 2.5-fold.
These initial optimization efforts produced a peptide with submicromolar affinity for RPA70N. However, the net charge of Peptide-20, at -6, did not bode well for delivery into cells, which is usually favored for neutral or positively charged peptides. Attempts to reduce the charge by substituting acidic residues with alanine or asparagine produced Peptide-29, with a desirable net charge of -2, but with a markedly reduced (43-fold) affinity (Table 1).
The ATRIP-derived binding peptide was 15 amino acids long. However, structural studies had suggested that interaction with the basic cleft of RPA70N did not require a peptide of that length. Therefore, the investigators experimented with removing residues from the C-terminus, as these had not been shown to be required for binding affinity. The results yielded Peptide-31, with 13 amino acids and a slight improvement in affinity over Peptide-20 (Table 1). This change did not, however, solve the problem of high negative charge since only uncharged amino acids were eliminated.
The inability of the team to reduce the charge of the peptide without a substantial loss of affinity posed a serious problem. To solve this problem, they turned to results from their earlier work directed toward the discovery of a small molecule inhibitor of RPA70N protein-protein interactions. These studies, which started with the identification of small molecule “fragments” that bind to RPA70N, revealed a hydrophobic binding hotspot in the basic cleft. Optimization of fragments that bind in this hotspot indicated particularly high affinity for compounds containing a 3,4-substituted phenyl ring. Crystallography data showed that the phenylalanine corresponding to position 11 of the p53 peptide also occupies this hotspot (Figure 2), suggesting that the comparably placed phenylalanine in the test peptides would also bind there. The researchers hypothesized that substitution of that phenylalanine with the unnatural amino acid 3,4-dichlorophenylalanine should increase the peptide’s affinity for RPA70N. In support of this hypothesis, the proposed substitution, yielding Peptide-33, increased the affinity of Peptide-31 by over 10-fold (Table 1), and X-ray crystal data confirmed binding of the 3,4-dichlorophenylalanine moiety in the hydrophobic hotspot (Figure 2).
Figure 2. (Left) X-ray co-crystal structure of RPA70N with p53 (red helix) and a 3,4-dichloro-phenyl fragment that binds to the RPA70N hydrophobic hot spot. (Right) X-ray co-crystal structure of RPA70N and Peptide-33 (blue), showing that the 3,4-dichlorophenylalanine moiety binds in the hydrophobic hot spot, which is also the site of binding of the comparably placed phenylalanine residue of the p53 peptide (red). Image reproduced with permission from A. O. Frank, et al. (2014) J. Med. Chem., published online February 3, doi:10.1021/jm401730y, copyright 2014 American Chemical Society.
The use of helical peptides as probes in intact cells is hampered by their physical and metabolic instability. A highly effective method to stabilize the helical structure and reduce susceptibility to proteolytic degradation is to create a hydrocarbon “staple,” linking one amino acid to a second amino acid three, four, or seven residues down the chain (Figure 3). This is accomplished by placing two α-(alkenyl)alanine residues at the desired distance apart in the peptide. An olefin ring-closing metathesis reaction joins the two pentenyl moieties to form the hydrocarbon staple (Figure 3). Based on the alanine substitution data, the investigators decided to place these residues at the D5 and E9 positions. The resulting stapled peptide (Peptide-34) was nearly as potent as the parent Peptide-31, but its charge (-4) was still too high for use in cells. Once again, substitutions at D1 and/or D6 to lower the charge (Peptides-36 and 37, Table 1) resulted in an over 10-fold reduction in affinity. However, substitution of 3,4-dichlorophenylalanine at position 11 resulted in a marked recovery of potency. The resulting stapled peptide (Peptide-39) boasted a charge of -2 and a 690-fold increase in binding affinity when compared to the original ATRIP peptide
Figure 3. Stabilization of helical peptides with hydrocarbon staples. (Top) Two α-alkene-substituted alanines are placed 3, 4, or 7 positions apart in the helical chain and then linked using an olefin ring-closing metathesis (RCM) reaction to form a hydrocarbon staple that links the two amino acids. (Bottom) In this study, α-(4-pentenyl)alanine residues were placed at the i and i+4 positions (substituting amino acids D5 and E9 in the peptide). Top image reproduced by permission from Macmillan Publishers Ltd, from Y.-W. Kim, et al. (2011) Nature Protocols, 6, 761, copyright 2011.
Fluorescence polarization anisotropy binding assays demonstrated that Peptide-39 binds to the basic cleft of RPA70N, but not to the OB folds of RPA70A/B or RPA70C/32D/14. Consistent with these findings, electrophoretic mobility shift assays demonstrated that Peptide-39 does not interfere with the binding of ssDNA to RPA. Incubation of U2OS human osteosarcoma cells with Peptide-39 resulted in uptake of the peptide in a pattern consistent with endocytosis, as expected. The fluorescein label was visible diffusely in the cytosol, and some nuclear staining was also observed (Figure 4). Peptides-36 and 37, with a charge of -3 and -2, respectively, also entered the cells, while Peptide-34, with a charge of -4, was excluded.
Figure 4. Uptake of stapled peptides by U2OS cells. Peptides with a -2 charge (Peptides-37 and 39) or a -3 charge (Peptide-36) were taken up by the cells, but Peptide-34, with a -4 charge was excluded. Image reproduced with permission from A. O. Frank, et al. (2014) J. Med. Chem., published online February 3, doi:10.1021/jm401730y, copyright 2014, American Chemical Society.
The results confirm that Peptide-39 binds with high affinity to the basic cleft of RPA70N, while not interfering with the binding of ssDNA to RPA. Uptake of the peptide by cells supports the hypothesis that Peptide-39 will be a valuable probe of RPA function in both cancer and normal cells. The findings also demonstrate the value of combining conventional peptide optimization strategies with approaches more frequently used in small molecule drug discovery. In particular, the introduction of an unnatural amino acid based on the results of fragment-based drug discovery data was key to obtaining the required high binding affinity in a peptide with acceptable negative charge.