Unpair to Repair or Degrade-Structure Sensing Nucleases

Lead Research Organisation: University of Sheffield
Department Name: Chemistry

Abstract

Genetic information is stored within deoxyribonucleic acid (DNA) in almost all organisms. DNA is composed of two polymer strands that are entwined to form a twisted ladder, known as the double helix. The rungs of the ladder are made up of a DNA alphabet A, C, T and G. These alphabet pieces, known as nucleotides, pair with each other according to special rules. A will pair only with a T in the opposite strand, and G only with a C. Thus each strand of the DNA molecule serves as a template to specify the sequence of nucleotides during duplication, or replication, of the complementary strand. When DNA becomes damaged, or is temporarily formed into a non-double helical structure (e.g. during replication), genetic information can be lost with life threatening consequences.

To reinstate the genome to its double-helical form it is essential that it is precisely cut at high speed, removing the aberrant portion of DNA. The protein catalysts (enzymes) that mediate fast cutting reactions on DNA are called nucleases. However, like a child playing with scissors or knives, potentially nucleases are risky molecules to have in cells. Uncontrolled and imprecise cutting of DNA could lead to destruction of genetic information. Instead nucleases must carry out reactions with high precision in the correct place and on the right DNA molecule. Understanding how structure-sensing high precision nucleases function is important. Failure of any one of them to act, or the risk of action in the wrong place, is life threatening.

Nucleases restore DNA back to its double helical form by cutting the DNA on one strand, sensing where the correct double helical duplex structure still exists. One group of these enzymes, known as the 5'-nuclease superfamily, look similar but act in many different situations on different incorrect DNA structures. We discovered that one member of the 5'-nuclease superfamily called FEN1 senses duplexes and carries out vital high precision cutting by testing the ability of the ends of double helical DNA to unpair. Specifically, FEN1 makes two nucleotides at the end of one strand of the intact duplex unravel so it reaches the nuclease active site and can get cut. This end of duplex sensing mechanism is known as double nucleotide unpairing (DNU). Unlike ends of duplexes, normal double helical DNA cannot undergo DNU and so is protected from nuclease damage. Moreover, the position of the active site relative to the nuclease bound duplex, controls which DNA strand is cut and precisely where it is incised.

We plan to study the way FEN1 and other superfamily members carry out DNU. We have some evidence that the ability to carry out DNU is very precisely controlled by the 5'-nuclease superfamily in a way related to each enzyme's biological function. For FEN1, which is an essential nuclease during replication in all life forms, we will test if DNU ability is related to correct target DNA recognition. We also plan to explore whether the DNU mechanism is universal by testing another family member EXO1. We will ask whether controlling the ability to effect DNU by interaction with other proteins can be used as an on-off switch to control EXO1 so that it only acts when needed. Using the DNU apparatus may also be important to cut another nucleic acid known as ribonucleic acid (RNA). RNA is a single stranded molecule, but can fold up to contain double helical regions by base-pairing with itself. Unlike DNA, RNA is targeted for destruction by nucleases when it has served its purpose. We will ask whether DNU is used to deal with duplex regions in RNA molecules by an enzyme called XRN1. We will also ask whether preventing DNU is a way we can therapeutically prevent nuclease action with small molecules. Elevated amounts of nucleases are present in rapidly dividing cancer cells and some are also required for viral infection. Thus understanding how to interfere with DNU could lead to new treatments for human diseases.

Technical Summary

Flap endonuclease-1 (FEN1), essential for lagging-strand DNA replication in all organisms, is the prototypical member of the 5'-nuclease superfamily. FEN1 processes Okazaki fragments removing 5'-flaps by catalysing phosphodiester hydrolysis one nucleotide (nt) into a duplex. We explain this reaction specificity by a novel double nucleotide unpairing (DNU) mechanism. DNU of the duplex end allows the scissile bond to contact catalytic metal ions and in hFEN1 requires a structural motif known as the helical cap. We propose this mechanism is universal for the superfamily. Here, we aim to understand the role of DNU in FEN-family catalysis, with implications for substrate selection, regulation of these enzymes in vivo and therapeutic inhibition strategies.

We will investigate whether control of DNU is related to biological substrate recognition by hFEN1. We will explore protein-DNA interactions required to activate hFEN1-DNU apparatus with implications for biological specificity. Extending our studies to other superfamily members, we will test whether DNU is a universal mechanism. We will ask whether DNU mediated by hEXO1 requires the helical cap and specific protein motifs that stabilize this. We will determine whether prevention of DNU through interactions with other hEXO1 specific domains in its extended C-terminus and/or protein modification is used to regulate EXO1 activity, thereby forming the basis of auto-inhibition strategies that are also used by other 5'-nuclease superfamily members. The processive 5'-exoribonucleases (XRNs) also conserve the FEN1 protein architecture and active site. We will probe whether XRNs use DNU to deal with RNA secondary structure. In addition, we will begin to unravel how processivity is achieved by the RNA hydrolyases testing the possibility that this is related to helical cap structural integrity. Finally, we will investigate strategies that could be universally applied to inhibit the superfamily by preventing DNU.

Planned Impact

The impact of this work is directed towards:

(1) Maintaining and enhancing the skills base of the biological and bioteconology workforce, by training and developing two PDRAs ensuring that the skills base exists for innovation and science that benefits the economy and society.

(2) Pharmaceutical and biotechnology companies developing novel targets for drug design and seeking to inhibit 5'- superfamily members. FEN1 is already a target for therapeutic intervention for a number of companies, including some located within the UK. Other members of the FEN1 protein family are also potential drug targets. Ultimately such efforts rely on molecular understanding and effective high throughput assays, and several aspects of this proposal are therefore highly pertinent to inhibitor development and strategies towards inhibitor design. Ultimately society will benefit if new therapeutic approaches become available.

(3) Biotechnologists seeking to develop tools for mutation detection or gene expressions. e.g. FENs are an integral part of the Taqman technology. Enzymes with subtly different activities could be useful for developing new assays and procedures.

(4) XRN1 is commercially available for use in molecular biology (New England Biolabs). Enzymes with subtly different activities could be useful for specific applications and understanding the substrate preferences of this enzyme in more detail will assist those seeking to use XRN1 in their experiments.
 
Description (1) Studies on FEN1.
Substrate and Protein Requirements for DNA Conformational Change Our first work package was to study the substrate and protein requirements for binding DNA in a bent conformation and transfer to the active site. These studies revealed that binding of DNA to hFEN1 in a bent conformation did not require active site metal ions or the presence of conserved active site residues or a 5'-flap. Notably, when a 5'-flap was present, but prevented from threading under the helical cap by conjugation with streptavidin, the substrate could still be bound in a bent conformation (albeit with lower affinity) but could not transfer to the active site. In addition, we determined that placement of the scissile phosphate diester in the active site required the presence of divalent metal ions, a Watson-Crick base pair at the terminus of the reacting duplex, and the intact secondary structure of the enzyme helical cap. Optimal positioning of the scissile phosphate additionally required active site conserved residues Y40, D181 and R100 and a reacting duplex 5'-phosphate. These studies suggest a FEN1 reaction mechanism where junctions are bound, 5'-flaps are threaded (when present), and finally the substrate is transferred onto active site metals initiating cleavage and highlight an important role for the reacting duplex 5'-phosphate.
This work is now published:
SI Algasaier, J. C. Exell, I. A. Bennet, M. J. Thompson, V. J. B. Gotham, S. J. Shaw, T. D. Craggs, L. D. Finger and J. A. Grasby (2016). "DNA and Protein Requirements for Substrate Conformational Changes Necessary for Human Flap Endonuclease-1-catalyzed Reaction." Journal of Biological Chemistry 291(15): 8258-8268.

3'-Flap Specificity We have also initiated studies on the relationship between 3'-flap recognition and transfer of the scissile phosphodiester bond to the active site. Systematic mutation of FEN1 3'-flap binding residues located >25 Å away form the active site has revealed a key role for L53. The mutated protein L53A prevents DNA conformational changes and reduces maximal single-turnover rates by approximately 1,000-fold. Although L53A does bind DNA less well than the WT protein, KD is raised by less than a factor of 10. Substrates lacking the 3'-flap nucleotide or the entire 3'-flap strand are also inhibited in DNA conformational changes that are indicative of movement into the active site and have reduced reaction rates. This work confirms our original hypothesis that correct substrate recognition is linked to function via active site access. We are currently writing this work up.



New Structures and Models Explaining FEN1 Substrate Selection We have also continued our structural collaboration with John Tainer (MD Anderson, Texas), specifically seeking to answer a long-standing and controversial question on FEN1 reaction specificity-how FEN1 confines its reactions to discontinuous DNAs that have free 5'-termini. This specificity is essential during replication and repair as multiple ss-ds junctions are revealed. Our new structures show FEN1 threads 5' flaps through a hole formed over its active site by the helical cap (above pink) that would necessarily exclude the reaction of continuous DNAs. We have shown that mutation of the residues implicated in 'phosphate steering' reduces reaction rate and reaction site specificity in vitro, and through collaboration with Sergei Mirkin (Tufts University, MA) we have studied the effects of these mutations on genomic instability in vivo. Notably, a D86N hFEN1 substrate structure, detailed in the application, also confirms threading and provides the first hFEN1 structure with DNA substrate in contact with catalytic metal.

This work is now published:
Tsutakawa, SE, Thompson, MJ, Arvai, AS Neil, AJ, Shaw, SJ, Algasaier, SI, Kim, JC, Finger, LD, Jardine, E, Gotham, VJB, Sarker, AS, Her, MZ, Rashid, F, Hamdan, SM, Mirkin, SM, Grasby JA and Tainer JA (2017) Phosphate steering by flap endonuclease 1 promotes 5´-flap specificity and incision to prevent genome instability Nature Communications 8, 15855 doi: 10.1038/ncomms15855

(2) EXO1.

We have cloned, overexpressed and set up an assay for a truncated variant hEXO1 that has allowed determination of catalytic parameters under both single and multiple turnover conditions. We have demonstrated that EXO1 threads its substrates under the helical cap, like FEN1. This is an important results at it unifies the mode of interaction with substrates across the superfamily. Further experiments are planned on the full-length protein.
This work is now published:
Shaw, S.J., Finger, L.D. and Grasby J.A. (2017) Human Exonuclease 1 Threads 5'-Flap Substrates through Its Helical Arch Biochemistry, 56, 3704-3707

(3) XRN1.
We have cloned, overexpressed, purified and assayed XRN1 including employing it in studies for FEN1 inhibitors (see below). Further experiments addressing XRN1 function are planned for the rest of the grant.
(4) Inhibitors.

We planned to investigate the mode of action of FEN1 inhibitors and to ascertain whether this inhibition strategy could be applied across the superfamily. Our work on mode of action has demonstrated that the current FEN1 inhibitors bind to active site magnesium ions. The mode of inhibition varied depending on the compound's sidechain, with smaller sidechains leading to competitive inhibition with substrate and larger sidechains to mixed. Since the latter case predicts that inhibitor can bind to both FEN1 and FEN1-DNA in the presence of Mg2+, we verified this by monitoring the formation of FEN1-DNA complexes in the presence of inhibitor by both fluorescence polarisation and FRET. All compounds blocked substrate access to the active site, either by excluding DNA binding, or by allowing the DNA to bind but not transfer to the active site. With the proteins from 2 & 3 we were able to investigate whether either XRN1 or EXO1 was inhibited by the same compounds. EXO1 was inhibited with a similar IC50 to FEN1, but XRN1 was not. The FEN1 inhibitors have also been shown to be engage with FEN1 in cells and to cause apoptosis. However, cellular EC50s were significantly higher than in vitro IC50s and we have investigated the possible reasons for this including FEN1 in its cellular context bound to PCNA. Notably, cell lines deficient in FEN1 synthetic lethal partners were much more sensitive to inhibitors, suggesting future avenues for FEN1-directed drugs in personalised medicine or combination therapies.

This work is now published:

J. Exell, MJ Thompson, L. D. Finger, S. Shaw, J. Debreczeni, T. Ward, C. McWhirter, C. Siöberg, D. Molina, W. Abbott, C. Jones, WJ. Nissink, S. Durant and J.A Grasby (2016). "Cellularly active N-hydroxyurea FEN1 inhibitors block substrate entry to the active site." Nature Chemical Biology 12, 815-821.
Exploitation Route The work is relevant to biochemists, cell biologists and molecular biologists trying to understand DNA repair and to those attempting to target the DNA damage response as an oncology treatment.
Sectors Healthcare,Pharmaceuticals and Medical Biotechnology