Recognizing And Severing Lethal Dna-Protein Crosslinks

Saikat Setua

26th October, 2020


Replication and transcription are important for copying and transferring the information of DNA to an mRNA and further for decoding that information to make protein by translation mechanism in ribosomes. But replication and transcription are often blocked by some processes among which one of the most dangerous are DNA damages. The covalent cross-linkages between proteins and a strand of DNA, known as DNA-protein crosslinks (DPCs) are one of the most understudied and detrimental forms of DNA damage. These DPCs act as blockages to replication and transcription and if not repaired properly these can lead to mutations and cell death. Researchers have characterized an enzyme that breaks such crosslink bonds and illuminated how specifically it recognizes the sites of such damages.


The integrity of cellular DNA often faces challenges by a broad variety of genotoxic agents which lead to hazardous mutational lesions in the genome. That’s why cells maintain DNA repair systems continuously. In this field of genetics, common forms of DNA damages like oxidation and alkylation of DNA-nitrogenous-bases, wrong base-pairing, UV light-induced pyrimidine dimers have all been studied a lot but one category of DNA damages that remains poorly understood and explained is DNA crosslinks. Now both DNA-DNA crosslinks and DNA-protein crosslinks are harmful. DNA-DNA crosslinks may be within a single strand of DNA (intrastrand crosslink) or two opposing strands of DNA (interstrand crosslink). DPCs are formed when a nucleotide residue on DNA forms a covalent bond with a protein. Nucleotide residues can also form covalent bonds with small peptide chains (then the crosslink is called DNA-peptide crosslink or DPC). The repair of such DPCs appears to involve complex mechanisms dealing with a large number of protein factors. Among them, one of the most dedicated is the protease SPRTN which cleaves the covalent bond between the protein and DNA. Recently a team of researchers from Munich has tried to elucidate how SPRTN recognizes such crosslinks, which can differ significantly in structure. It appears that the enzyme utilizes such a way that it’s only activated in highly specific conditions.

Formation of DPCs

DPCs can be induced environmentally, therapeutically, or endogenously. Exposure to ionizing radiation, ultraviolet rays, and various transition metal ions including chromium and nickel from the environment can induce DPC formation. Some other carcinogens like crotonaldehyde, acrolein are also known to induce DPC formation. Ionizing radiation and chemical compounds that induce DPC or interstrand crosslinking are routinely used in chemotherapies. Each gray of ionizing radiation is thought to produce about 150 DPCs in the genome per cell. When the ionizing radiation attacks water or oxygen molecules in the cell, water and oxygen give rise to Reactive Oxygen Species (ROS) like hydroxyl radical and superoxide radical respectively. As 70% of the cell by mass is water, a large number of ROS get formed. Such ROS reacts with the nucleotides of DNA and makes such a change in the nucleotide that an amino acid of a nearby protein can react with that nucleotide to form a DPC.


Besides therapeutic radiation, some cancer drugs including nitrogen mustards, decitabine, and some platinum-based agents like cis-platin and trans-platin derivatives are also known as DPC inducer. Endogenously induced DPCs can be derived through enzymatic or non-enzymatic pathways. Certain DNA interacting enzymes that otherwise form transient covalent complexes with DNA can become entrapped onto DNA leading to the formation of enzymatically derived DPCs. This occurs most frequently with Topoisomerases upon inhibitor treatment. Besides these during base excision repair also DPCs can be formed. DNA bases often are subjected to hydrolysis, oxidation, and alkylation. The abasic sites formed by glycosylase, cleaving the damaged or wrong base can form covalent bonds being attacked by any nucleophilic center of proteins passing by.

For non-enzymatic pathways, reactive aldehydes play a certain role. Reactive aldehydes are often generated as a result of various metabolic processes in cells, such as amino acid metabolism, oxidative demethylation. When the carbonyl carbon of an aldehyde is sufficiently electrophilic, it can react with the nucleophiles in the surrounding environment. When such aldehyde comes in the vicinity of chromatin, it can react with the primary amine of a DNA base, producing an intermediate adduct which further can react with a primary amine of another nearby DNA base forming interstrand crosslink or it also can form an amide bond with a lysine or arginine residue of nearby protein forming a DPC. The best-studied DPC inducing aldehyde is formaldehyde. Studies have shown that cells have evolved dehydrogenase enzymes to metabolize aldehydes like formaldehyde and acetaldehyde, effectively detoxifying them.

Recognizing and cleaving DPCs

Proteolytic cleavage of DPCs is a way to repair these lesions. DPC specific proteases were discovered in budding yeast, Caenorhabditis elegans, and mammalian systems. The first proteolysis-based mechanism for DPC repair was characterized in budding yeast. The metalloprotease Wss1 in budding yeast promotes genome stability by specifically targeting DPC substrates, cleaving them into smaller adducts, and enabling replication past the lesion. Scientists have also identified metalloprotease SPRTN as a DPC cleaving protease. The SPRTN protein with a weight of 55-kDa bears the catalytic metalloprotease domain in its N-terminus. It’s evolutionarily related to Wss1 but SPRTN has relatively more modified domains than Wss1. While Wss1 possesses a small ‘ubiquitin-like’ modifier binding domain, SPRTN harbors a ubiquitin-binding domain in the C-terminal tail. In the C-terminus of SPRTN also there is a proliferating cell nuclear antigen (PCNA)-interacting protein motif, an SHP box required for binding to chaperone-like protein p97.

In humans, mutations that reduce the activity of the enzyme SPRTN are associated with a high incidence of liver cancer in early life and markedly accelerate the aging process. In mammalian cells, knockdown of SPRTN with siRNA resulted in hypersensitivity toward formaldehyde and persistence of DPCs on DNA. These results suggest that SPRTN is involved in the

DPC repair system. 

The first author of the study about SPRTN as a DPC repairer, Hannah K. Reinking raised the question that how SPRTN such specifically cleaves only the crosslinked proteins, while leaving surrounding chromatin proteins unharmed. Because depending on the protein and the DNA subunit involved, the structure of the crosslink can vary widely.

Findings of Reinking, Sattler and Stingele

As an answer to the raised question about the striking specificity of SPRTN, recently Hannah K. Reinking associated with Michael Sattler and Julian Stingele all from Munich have shown some results from their research. Reinking and colleagues constructed models consisting of proteins attached to defined positions within DNA strands and examined whether the SPRTN protease could repair them in the test-tube.

SPRTN has been reported to be efficiently activated by DNA oligonucleotides, whether they were single or double-stranded. But a strikingly different scenario was found using long circular DNA for activation of SPRTN. Single-stranded DNA (ssDNA) circles were found to activate SPRTN much more strongly than double-stranded DNA (dsDNA) circles. Denaturation of dsDNA circles into ssDNA by heating and snap-cooling on ice restores their activation potential. But when dsDNA was linear or simple double-stranded oligonucleotide instead of being circular it had well-efficiency of activating SPRTN.


To explain that they tested taking PCR-generated dsDNA fragments. For each reading of the strength of activation of SPRTN they took shorter dsDNA fragments than the previous but maintained a total amount of DNA constant. To tell easily they just fragmented the DNA into a greater number of parts in every reading relative to the previous. Strikingly they found that the shorter the dsDNA fragment, the more strongly it activates SPRTN under high-salt conditions. From this experiment, they concluded that when shorter fragments were used, the number of dsDNA ends increased. So, it was found that SPRTN cleaves DPCs at dsDNA ends.


In the structure of SPRTN, between the C-terminus and N-terminal protease domain, there’s a basic DNA binding region (BR) of low complexity that bears several positively charged amino acids. A recent crystal structure of an N-terminal SPRTN fragment revealed an unexpected Zinc binding domain (ZBD) immediately after the protease domain and preceding the BR region. The ZBD domain is speculated to constitute an ssDNA binding DNA. Now specific amino acid replacement in the ZBD domain showed reduced auto-cleaving in SPRTN variants. Similarly, an SPRTN variant with amino acid replacements in the BR domains shows a comparable reduction in activity. This experiment showed that both ZBD and BR domain contributes to binding with DNA. Later to prove the structural contribution of ZBD and BR for DNA binding they did Nuclear Magnetic Resonance (NMR) spectroscopy analysis which showed clearly that SPRTN contains two recognition domains. The ZBD domain binds with the single-stranded and BR domain to the double-stranded DNA. As a conclusion from Stingele, the enzyme uses a modular system for substrate recognition. The possibility of the enzyme being active is only when both the ZBD and BR domains are engaged. And DNA in which double-stranded and single-stranded regions are in close proximity is often found in the vicinity of crosslinks.



These results from Reinking, Sattler and Stingele are of great clinical relevance. The action of many chemotherapeutic drugs depends on their ability to force crosslinks with DNA. Since tumor cells divide more frequently than non-malignant cells, they are particularly sensitive to this type of DNA damage. Such DNA repair enzymes like SPRTN are therefore of great interest as potential drug targets for use in case of personalized cancer therapies and also the agents that inhibit the SPRTN protease could eventually be employed to boost the efficacy of chemotherapy.



  1. James A. Swernberg, (2010, December 16). Endogenous versus exogenous DNA adducts.  From

  2. Sharon Baker, Michael Weinfeld, (2005, March 2). DNA-Protein crosslinks: their induction, repair, and biological consequences.  From

  3. Julian Stingele, Stefan Jentsch, (2015, July 1). DNA-protein crosslink repair. From

  4. Yves Pommier, (2006, August 4). Repair of Topoisomerase I mediated DNA damage. From

  5. Rajendra Prasad, Julie K. Horton and colleagues,  (2014, April 25). Suicidal crosslinking of PARP-1 to AP site intermediates in cells undergoing BER. From

  6. Julian Stingele, Michael S. Schwarz and colleagues, (2014, July 3). A DNA dependent protease involved in DPC repair. From

  7. Jaime Lopez-Mosqueda, (2016, November 16). SPRTN is a mammalian DNA binding metalloprotease that resolves DPCs. From

  8. Julian Stingele, Roberto Bellelli and colleagues, (2016, October 27). Mechanism and regulation of DPC repair by the DNA dependent metalloprotease SPRTN. From

  9. Hannah K. Reinking, Michael Sattler, Julian Stingele and colleagues, (2020, August 26). DNA structure specific cleavage of DPCs by the SPRTN protease. From

  10. DNA repair: Locating and severing lethal links. (2020, August 27). Retrieved October 01, 2020, from

About the Author

20200819_194632 - SAIKAT SETUA.jpg

Saikat is a MSC. Biotech student from St. Xavier's College, Kolkata