For protein expression in Escherichia coli, the genes encoding TIP (TK0808) and Fen1 (TK1281) were PCR-amplified from T. kodakarensis genomic DNA using primers that include NdeI and SalI restriction enzymes in the forward and reverse primers, respectively (Supplementary Table S1). The amplified DNA was ligated with pET15b to introduce an in-frame His₆-tag at the N-terminus of each protein (25). All constructs were sequenced to insure the integrity of the sequence. The construction of the expression vectors for T. kodakarensis PCNA1, PCNA2, RFC, PolB and Methanothermobacter thermautotrophicus PCNA (mtPCNA) has been previously described (20,21,26,27).

Vectors that express the mutant forms of PCNA2 (PCNA2-A, K197A-Y200A-Y204A and PCNA2-E, K197E-Y200E-Y204E) were generated by site-specific mutagenesis using a QuikChange kit (Agilent Technologies) with plasmids that encode the wild-type proteins as templates, and oligonucleotides listed in Supplementary Table S1. The vector to express the monomeric form of PCNA1 (PCNA1m, E143K-V175D-I177K) was generated by site-specific mutagenesis using a QuikChange kit (Agilent Technologies) in two stages with plasmid that encode the wild-type protein as the first template, and the oligonucleotides listed in Supplementary Table S1. First the E143K mutant was introduced, followed by the introduction of the V175D-I177K double mutant. The vector to express the monomeric form of PCNA2 (PCNA2m, R75A-F111E-R143A) and mtPCNA (mtPCNAm, D141K-A173D-I175K) was synthesized and cloned into pET21a by Polepolar Technologies.

The pET-based vectors encoding the different proteins were transformed into BL21 DE3 Rosetta cells (Life Technologies) and protein expressions were induced at 37°C by the addition of 0.5 mM IPTG and further incubation for 3 h. The proteins were purified by adsorption to and elution from a Ni²⁺-column, as previously described for the PCNA proteins (20). All proteins were, at least, 95% pure as judged by polyacrylamide-sodium dodecyl sulphate (SDS) gel. Following purification, the proteins were aliquoted and stored at −80°C.

Aliquots of each experimental protein (100 μg) or protein mixtures (100 μg each) and Gel Filtration standards (Bio-Rad) were dissolved in 200 μl of 25 mM Tris–HCl (pH 8.0), 200 mM NaCl, 0.5 mM EDTA and 10% (v/v) glycerol and loaded onto a Superdex-200 column (HR10/30; GE Healthcare) pre-equilibrated in the same buffer. Fractions (250 μl) were collected from the column at a flow rate of 0.5 ml/min. The proteins present in aliquots (80 μl) of each fraction were separated by electrophoresis through a 15% (w/v) polyacrylamide-SDS gel and stained with Coomassie brilliant blue (R250).

PolB catalyzed elongation of singly primed M13 ssDNA was carried out in reaction mixtures (20 μl) containing 40 mM Tris–HCl (pH 8.0), 250 mM NaCl, 1.5 mM dithiothreitol (DTT), 100 μg/ml bovine serum albumin (BSA), 10 mM magnesium acetate, 2 mM ATP, 100 μM each of dCTP, dGTP, and dTTP, 20 μM [α-³²P]dATP (1.2 × 10⁴ cpm/pmol), 10 fmol of singly primed M13mp18, 440 fmol of RFC and 100 fmol of PolB and PCNA1, PCNA2 and TIP proteins were added as indicated in the figure legend. Reaction mixtures were incubated for 20 min at 70°C. Following incubation, reactions were treated with 2 μl of stop solution (containing 0.1 M ethylenediaminetetraacetic acid (EDTA), 5% SDS, 80 μg yeast tRNA, 20 μg proteinase K) and incubated for 20 min at 37°C. For quantitation, aliquots (4 μl) of reaction mixtures were removed, and DNA synthesis measured by adsorption to DE81 paper followed by washings and liquid scintillation counting. The mixture was adjusted to 3 M ammonium acetate and 70% ethanol and the DNA collected by centrifugation. Following in vacuo drying the pellets were dissolved in 0.05 M NaOH and 1 mM EDTA and separated by alkaline agarose electrophoresis.

Three oligonucleotides were used to generate the substrate for Fen1 assays. H7 (5′-CTTCAATCGGCTCAGACCGAGCAGAATTCTATGTGTTTACCAAGCGCTG-3′) was labeled at the 5′-end with ³²P using [γ-³²P]ATP and polynucleotide kinase. Following the labeling the DNA was hybridized to two additional oligonucleotides, H2 (5′-CAGCGCTTGGTAAACACATAGAATTCTGCTCGGTCTCTCGGCAGATTCTAGAAATCGACGCTAGCAAGTGAC-3′) and H5 (5′-GTCACTTGCTAGCGTCGATTTCTAGAATCTGCCGAG-3′). The substrate was purified from a polyacrylamide gel as previously described (28).

Unless otherwise noted, Fen1 nuclease assays were performed in 20 μl reactions containing 20 fmol substrates, 25 mM Tris–HCl (pH 8.0), 10 mM MgCl₂, 125 μg/ml BSA, and proteins as indicated in the figure legends. Reactions were incubated at 60°C for 1 h. Following incubation, reactions were stopped by adding 20 μl of 95% formamide, 10 mM EDTA and 0.1× Tris-Borate-EDTA (TBE) followed by incubation at 100°C for 2 min. The reaction products were separated on 20% polyacrylamide–8 M urea gels in 1× TBE followed by visualization and quantification using phosphorimaging (GE Healthcare).

Thermococcus kodakarensis strains were grown in artificial sea water (ASW) with 5 g/l each of yeast extract (Y) and tryptone (T), and 2 g/l sulfur (S°) at 85°C with the growth of cultures measured by an increase in optical density at 600 nm (OD_{600 nm}) as previously described (29). Plasmids pOSU0808A, pOSU0808B and pOSU0808D were created using standard molecular biology techniques and maintained in E. coli (29). Transformation of T. kodakarensis cells (30) and strain construction for markerless deletion of the TK0808 gene were performed as previously described (29). Strains were constructed by transformation of T. kodakarensis TS559 (29), with transformants grown on media lacking agmatine for plasmid incorporation, and then counter-selected on media containing 6-methylpurine to generate a strain containing the deletion of TK0808 with no selectable marker (29). The TK0808 open reading frame (ORF) contains 195 bp. The THH2 strain (ΔTK0808) was generated by deleting the first 189 bp of the TK0808 ORF. The last 6 bp of the TK0808 ORF overlap with the TK0807 gene and were thus retained in the chromosome and not deleted.

Sequences encoding TIP were amplified via PCR with a 5′ primer that contained the promoter for histone B from M. thermautotrophicus (P_hmtB). The resultant amplicon was cloned into pTS543 (31) at a unique NotI restriction site using Clonetech In-Fusion^® HD cloning.

The genomic organization of strain THH2 was confirmed via Southern blots of Acc65I and BglII-digested genomic DNA. The Acc65I and BglII restriction fragments that hybridized to a digoxigenin (DIG)-labeled amplicon probe, PCR-generated from within TK0808 (probe B/C) and a flanking region (probe E/F) (Figure 5), were identified using anti-DIG antibodies coupled to alkaline phosphatase as previously described (32).

Thermococcus kodakarensis strains were grown overnight in 5 ml ASW YT S° medium, and strains lacking pTHH6 were supplemented with 1 mM agmatine sulfate. Growth at 85°C of triplicate cultures was monitored at OD_{600 nm} of 1:100 dilutions of inoculated cultures.

Strains were exposed to DNA damaging agents essentially as previously described (33). Briefly, aliquots (5 μl) of serial dilutions of cells sampled from growing cultures of T. kodakarensis TS559 and THH2 (ΔTK0808) were plated, under anaerobic conditions, on ASW-S° medium containing all 20 amino acids and 1 mM agmatine, plus or minus 0.1% methyl methanesulfonate (MMS) or 100 μg/ml mitomycin C. The colonies present on each plate, after 60 h incubation at 85°C, were counted. Similarly, aliquots of dilutions of the same cultures were spotted onto ASW-S° plates containing all 20 amino acids and 1 mM agmatine and were then exposed to ultraviolet (UV) irradiation (model UVH-46550, International Biotechnologies Inc.) for 1 min before the plates were incubated anaerobically at 85°C for 60 h to allow colony formation.

For H/D exchange (HDX) mass spectrometry (MS) analyses, the protein stock solution was diluted in PBS buffer (20 mM sodium phosphate, 500 mM NaCl, pH 7.6) to prepare a 15 μM final analytical concentration and equilibrated at 4°C for 2 h. The PCNA–TIP complex was prepared by mixing PCNA and TIP at a ratio of 1:1 (PCNA monomer:TIP monomer) and equilibrated at 4°C for 16 h.

HDX was conducted on a HDX PAL robot (LEAP Technologies, Carrboro, NC, USA). Protein solutions (5 μl) were diluted into 21 μl D₂O buffer (20 mM sodium phosphate, 500 mM NaCl, pD 7.6) at 3°C. This temperature was chosen to slow the H/D exchange rates so that the observed rates better fit the dynamic range of the instrument. At selected times (30 s, 5, 20, 40 and 60 min) the HDX was quenched by mixing with 40 μl of 3 M urea, 1% trifluoroacetic acid (TFA) at 1°C. The quenched solution was injected into an on-line pepsin-digestion device for 3 min. The digested protein solution was trapped on a C18 guard column (1.0 mm diameter, 5 μm, Grace Discovery Sciences). The peptide mixture was separated with a C18 analytical column (1.0 mm diameter × 5 cm length, 1.9 μm, Hypersil GOLD, Thermo Scientific) via a Dionex Ultimate 3000 UPLC with a 9.5 min gradient operated with a binary mixture of solvents A and B at 50 μl/min flow rate. The gradient settings were: 5–35% solvent B in 3 min, 35–70% solvent B in 5 min, 70–100% solvent B in 0.5 min and isocratic flow at 100% solvent B for 0.5 min, then returned to 5% solvent B in 0.5 min. Solvent A was water containing 0.1% formic acid, and solvent B was 80% acetonitrile and 20% water containing 0.1% formic acid. All LC connection lines and valves were housed in the refrigerated compartment of the HDX PAL at 2°C. Peptides were analyzed on a Thermo LTQ Orbitrap Elite (Thermo Fisher, San Jose, CA, USA). The instrument settings were: spray voltage, 3.7 kV; sheath gas flow rate, 25 (arbitrary units); capillary temperature, 270°C. Three replicates were obtained for each ion-exchange time point.

Peptides of PCNA (PCNA1, PCNA2, PCNA2-A mutant, PCNA2-E mutant) and TIP were identified using tandem MS (MS/MS) on the Thermo LTQ Orbitrap Elite. One full mass spectral acquisition triggered six scans of MS/MS [precursor ion is activated by collision-induced dissociation (CID)] whereby the most abundant precursor ions were sequenced. Peptide identification was achieved by submitting Thermo RAW files to both MASCOT (Matrix Science, Oxford, UK) and MassMatrix database search engines. The MASCOT settings were: enzyme, none; MS tolerance, 20 ppm; MS/MS tolerance, 0.6 Da; maximum number of missed cleavages, 3; peptide charge of 1+, 2+ and 3+. The MassMatrix settings were: enzyme, nonspecific; precursor ion tolerance, 10 ppm; product ion tolerance, 0.8 Da; minimum pp score, 5.0. All identifications of peptides were manually confirmed. From mass spectra obtained during HDX-MS experiments the centroid of each deuterated peptide envelope and the relative deuterium uptake by each peptide were calculated with HDX Workbench (34). Corrections for back exchange were made by considering the values of 80% deuterium content of the exchange buffer and an estimated 70% deuterium recovery. Paired t-tests, computed by the Prism 5.01 software package (35), were used in verifying the deuterium uptake differences. The quality of the HDX data fixed the threshold for statistical significance at P < 0.05.

TIP protein, encoded by TK0808, was co-isolated with His₆-PCNA1 (TK0535) from T. kodakarensis cell lysates (24). To determine if the two proteins also interacted in vitro, recombinant TIP and PCNA1 were mixed and the products were examined by size exclusion chromatography. As previously reported (21,22), PCNA1 protein (29.1 kDa) elutes as a trimeric complex (Figure 1A; elution peak in fraction 55). TIP (9.8 kDa) alone eluted as a monomer (Figure 1C, elution peak in fraction 67). When incubated together, TIP and PCNA1 interacted to form a complex, as shown by the elution of TIP in earlier fractions (Figure 1D, elution peak in fraction 58) and elution of PCNA1 in later fractions (Figure 1D, elution peak in fraction 58). Each TIP molecule binds to one PCNA monomer as judged by low resolution X-ray structure of the complex (unpublished observation).

As discussed above, the T. kodakarensis genome encodes for two PCNA proteins (21,22) with similar overall structures (20). It was reported that PCNA2 is present at low levels in vivo [<50 molecules per cell (21,22)], which may explain the inability to detect PCNA2–TIP interaction in vivo. Nevertheless, the ability of TIP to interact with PCNA2 was also evaluated. PCNA2 protein (29.3 kDa) elutes as a trimeric complex (Figure 1B; elution peak in fraction 55) as previously reported (21,22). Similar to PCNA1, TIP also forms a complex with PCNA2, as shown by the elution of TIP in earlier fractions than TIP alone (Figure 1E, elution peak in fraction 52), but with significantly lower affinity.

Homologues of TIP have been identified only in the genomes of Thermococcales. This may suggest that only PCNA proteins from these organisms can bind TIP. Therefore, as a control, a PCNA homologue from a different euryarchaeon, M. thermautotrophicus, was used. As previously reported (27), the mtPCNA protein (30.4 kDa) elutes from the sizing column as a trimer (Figure 1F; elution peak in fraction 55). The presence of TIP did not affect this elution profile (compare Figure 1G to Figure 1F) and the elution profile of TIP was not altered by the presence of mtPCNA (compare Figure 1G to Figure 1C).

As shown in Figure 1D and E, TIP binds to both PCNA1 and PCNA2. However, in comparison to PCNA1, only a small fraction of TIP co-eluted with PCNA2 (compare Figure 1D to Figure 1E), suggesting a weaker interaction between TIP and PCNA2 than with PCNA1. It was previously shown that PCNA2 forms a more stable trimer than PCNA1 (20,21). It is thus possible that TIP requires an exposed interface on the PCNA ring in order to bind. As PCNA2 is a stable trimer in solution, at any given time only a small fraction of the rings are open and therefore only a small subset of the interfaces are exposed and available to bind TIP. This may contribute to the observed weak binding between TIP and PCNA2 (Figure 1E).

To evaluate this possibility, a monomeric form of PCNA2 (PCNA2m) was generated using the PCNA2 structure as a guide (PDB: 3LX2) (20). Three residues in the interfaces between the protomers were mutated (R75A–F111E–R143A) to generate the monomer PCNA2m. TIP binds poorly to PCNA2 as shown by the small amount of TIP–PCNA2 complex formed (Figure 2D, fraction 49). PCNA2m (Figure 2B) and TIP (Figure 2C) are both monomeric (elution peaks in fraction 61 and fraction 70, respectively). When incubated together, TIP and PCNA2m interacted to form a stable complex as shown by the elution of both proteins as a complex (Figure 2E, elution peak in fraction 58). Taken together with the results presented in Figure 1 and the previous data on the stability of PCNA1 and PCNA2 (20,21), these findings suggest that TIP needs an exposed interface for PCNA binding. However, TIP does not interact with any monomeric PCNA as it does not bind the monomeric form of mtPCNA (mtPCNAm, Supplementary Figure S1).

PCNA proteins bind and stimulate the activity of many proteins involved in nucleic acid metabolic processes. The best understood function for PCNA is its role as the processivity factor for DNA polymerases. While the processivity of replicative polymerases alone is low, the addition of PCNA dramatically increases processive synthesis (8). TIP binding to PCNA may regulate PCNA functions. Therefore, the effect of TIP on PCNA stimulation of PolB activity was evaluated using a singly primed M13 template.

As previously reported (20,21), the elongation of a primed ssM13 template by PolB was detected only in the presence of RFC and either PCNA1 (Figure 3 compare lanes 7–9 to lane 1) or PCNA2 (Figure 3 compare lanes 3 and 4 to lane 1). The presence of TIP, however, inhibited the stimulatory effect of both PCNA proteins (Figure 3 compare lanes 5 and 6 to lane 4; lanes 10–13 to lane 8; and lanes 14 and 15 to lane 7). TIP, without PCNA, did not affect PolB activity (Figure 3 lane 2). It was previously shown that higher concentrations of PCNA1, in comparison to PCNA2, are needed to achieve a similar stimulation of PolB (20,21). This is likely due to the lower stability of PCNA1 trimeric rings (20). Therefore, the concentrations of PCNA1 used in the experiments shown in Figure 3 are higher than those used for PCNA2.

The inhibitory effect of TIP on PCNA-dependent PolB activity could be due to the binding of TIP to PCNA or the binding of TIP to PolB. The ability of TIP to bind PolB and the effect of TIP on PolB–PCNA interactions were evaluated using size exclusion chromatography. TIP (9.8 kDa) alone eluted as a monomer (Supplementary Figure S2C, elution peak in fraction 67). PolB (89.6 kDa) alone also eluted as a monomer (Supplementary Figure S2D, elution peak in fraction 55). The elution profile did not change when TIP and PolB proteins were incubated together (compare Supplementary Figure S2I to S2C and S2D). PCNA1 (29.1 kDa) and PCNA2 (29.3 kDa) eluted as homotrimers (Supplementary Figure S2A and B; elution peaks in fraction 55). When PCNA1 or PCNA2 were incubated with PolB, they formed a complex (Supplementary Figure S1E and G; elution peaks in fraction 49). When TIP was added to the complex, however, the interactions between PCNA1 and PCNA2 with PolB were inhibited (Supplementary Figure S2F and H) likely due to the PCNA–TIP interactions. These results suggest that the inhibitory effect of TIP on PCNA-dependent processivity of PolB occurs by inhibiting the interactions between PCNA and PolB.

In a number of organisms PCNA stimulates the activity of Fen1, a structure-specific nuclease involved in Okazaki fragment maturation (36–38). Fen1 removes the flap structure that contains the RNA primer prior to gap filling by the polymerase and subsequent ligation of the two adjacent Okazaki fragments (36,37). Although the effect of PCNA on Fen1 activity has been shown for other organisms, it has not yet been reported for T. kodakarensis. Therefore, the effects of PCNA1 and PCNA2 on Fen1 activity were evaluated using a Fen1 assay (shown schematically in Figure 4A). The experiments were performed in the absence of RFC using the ability of PCNA to thread onto duplex DNA as was previously reported (39). Fen1 alone has weak nuclease activity (Figure 4B lanes 1 and 7) and both PCNA1 (Figure 4B lanes 2–5, also see panel C) and PCNA2 (Figure 4B lanes 8–11, also see panel C) stimulate the activity of Fen1 in a concentration-dependent manner. The concentrations of PCNA used to examine Fen1 stimulation (Figure 4) were higher than those used to stimulate PolB (Figure 3) and therefore at the concentrations tested both PCNA proteins stimulate Fen1 activity to a similar extent (Figure 4B and C). As expected, neither PCNA1 nor PCNA2 alone possess nuclease activity (Figure 4B lanes 6 and 12).

Next, the effect of TIP on Fen1 stimulation by each PCNA protein was evaluated. PCNA-dependent Fen1 activity was inhibited in the presence of TIP in a concentration-dependent manner (Figure 4D compare lanes 3–7 to lane 2; and lanes 10–14 to lane 9, also see panel E). Although both PCNAs stimulate Fen1 activity to a similar extent (Figure 4C), TIP is more efficient in inhibiting the effect of PCNA2 than PCNA1 (Figure 4E). Why TIP appears to be more efficient in inhibiting PCNA2 is not clear especially in light of the stronger binding of TIP to PCNA1 (Figure 1). However, as gel-filtration is not an equilibrium technique, it is possible that under equilibrium conditions TIP binds PCNA2 better than PCNA1. Nevertheless, as with the effect of TIP on PolB activity, the binding of TIP to PCNA inhibits Fen1 activity.

PCNA is an essential protein and plays critical roles in many nucleic acid transactions. Therefore, one could expect that TIP, which regulates PCNA activity, may also be essential for T. kodakarensis viability. To test this hypothesis, we constructed a T. kodakarensis strain deleted for the gene encoding TIP, TK0808. Thermococcus kodakarensis THH2 (ΔTK0808) was obtained and the genome structure of the mutant confirmed by diagnostic PCR (Figure 5B) and Southern blots (Figure 5C). The growth of the THH2 strain was indistinguishable from the wild-type strain (Figure 5D), suggesting that TIP is not essential for T. kodakarensis viability under normal growth conditions. As the PCNA protein also plays an essential role in DNA repair (3), the THH2 strain was evaluated for its sensitivity to the DNA damaging agents including UV, MMS and mitomycin C. No substantial differences were detected between the wild-type and ΔTK0808 strains (data not shown).

While the structures of both PCNA proteins from T. kodakarensis are known (20), the structure of TIP is not. One approach to understand the structural properties of these two partners is by studying their conformational dynamics in solution. Measurements of peptide-level HDX kinetics can potentially reveal protein dynamics in solution with a spatial resolution of 6–10 residues (40). To attain such resolution, proteolysis of the subject proteins during the HDX-MS measurements must efficiently produce a distribution of peptides of overlapping sequence and leave negligible levels of the parent protein. Immobilized pepsin digested PCNA2 efficiently, but failed to adequately digest PCNA1 for study (Supplementary Figure S3); hence, our HDX-MS studies focused on the PCNA2 system. Proteolytic digestion of native PCNA2 and TIP with pepsin yielded HDX information for 98% of the PCNA2 backbone and 99% of the TIP backbone (Supplementary Figure S4).

The higher order structure of TIP is unknown. As shown by the HDX-MS exchange rate heat map of TIP (Figure 6A), amide groups of TIP, especially in the C-terminal half of the molecule, rapidly exchange to contain over 80% deuterium within the first 30 s after dilution in D₂O. Such fast conformational dynamics are in accord with the view that TIP is mostly unstructured, as has also been suggested by NMR analysis (data not shown).

When the dynamic structure of PCNA2 based on our HDX-MS kinetic data was mapped onto the crystal structure of PCNA2 (PDB: 3LX2) (20), it showed that the interior–helical regions near the trimeric interfaces are structurally stable (Figure 6B), as they exchanged an extremely low amount of deuterium over durations ranging from seconds to hours. On the other hand, the outer helices and loop regions, which are closed to the trimeric interfaces, are relatively more flexible, as their amide sites exchanged substantial levels of deuterium on a time scale of 20 min at 3°C (Figure 6B). The exterior, long–loop regions were less dynamic as their amide sites accumulated deuterium on the time scale of hours. These HDX-MS results are in accord with the quaternary structure of the PCNA2 trimer (Figure 6B).

The HDX-MS data and protein dynamics of PCNA2 can be compared with those reported previously for the β-subunit of E. coli DNA polymerase III (41). While these species serve analogous cellular functions, they differ in that PCNA2 forms a homotrimer in solution, while the β-subunit forms a homodimer (4,17). At long exchange time (>2 h) HDX-MS data observed for the β-subunit exhibits EX1 exchange kinetics patterns (41). In contrast, our data exhibits only EX2 kinetics patterns for PCNA2 at the longest HDX time (1 h) [these mechanisms are outlined in (42)]. We note that the β-subunit data were obtained at 25°C and the data for PCNA2 were measured at 3°C; thus, we cannot rule out temperature as a controlling variable for the differing kinetic behavior. Also, given that PCNA and β-subunit are different molecules, they may manifest different structural dynamics.

HDX-MS has the potential to reveal binding interfaces in the protein complex by detecting changes in deuterium uptake rates among backbone amides in response to the formation of the protein complex (43–45). Pairwise comparisons of PCNA2 (PCNA2 trimer alone) and PCNA2 complexed to TIP (PCNA2–TIP complex) and TIP and TIP complexed to PCNA2 provide information regarding the interfaces located on PCNA2 and TIP that stabilize the TIP–PCNA2 complex.

The HDX-MS results for the PCNA2–TIP complex (PCNA monomer:TIP monomer = 1:1) show that peptide region 187–204 of native PCNA2 (Figure 7A and Supplementary Figure S5), located outside of the ring structure, exchanges significantly less deuterium (P ≤ 0.01, where any value of P < 0.05 is considered significant) upon TIP binding. The diminished exchange rate exhibited by peptide region 187–204 in native PCNA2 suggests that these amides are involved in PCNA2–TIP binding. We also observed minor diminished D-uptake rates of peptide region 243–259 of native PCNA2 upon TIP binding (data not shown). However, we regard this observation more cautiously because the 243–259 region comprises mostly the His-tag; thus, the observed decrease in solvent accessibility at this region may be caused by non-specific interactions. Interestingly, peptide regions 34–38 and 104–110 of native PCNA2 exchange more deuterium upon TIP binding (Figure 7A and Supplementary Figure S5), indicating diminished protection factors. Because these amides are located at the trimer interfaces, the diminished protection factors suggest that binding to TIP facilitates the dissociation of PCNA2 trimer.

Although both PCNA1 and PCNA2 bind TIP, mtPCNA does not (Figure 1). After identifying the region at 187–204 in PCNA2 as a potential binding interface in TIP binding, this region was compared in the three PCNAs. PCNA1 and PCNA2 share 59% identity in this region (10/17 residues), mtPCNA shares 23% identity (4/17 residues) with PCNA1 and shares 27% identity (5/18 residues) with PCNA2 (Supplementary Figure S6). The low similarity between mtPCNA and PCNA1 and PCNA2 in the region required for TIP binding may explain the inability of mtPCNA to bind TIP.

Based on the HDX-MS results, region 187–204 of PCNA2 was selected for site-directed mutagenesis to examine its role in regulating PCNA2–TIP interaction. Three residues in this region, K197, Y200 and Y204, were mutated to Ala (Figure 8B, PCNA2-A) or Glu (Figure 8B, PCNA2-E). First, the influence of these mutations on the regulation of PCNA by TIP was evaluated using the Fen1 activity assay (Figure 8A). Both PCNA2-A and PCNA2-E stimulate the PCNA-dependent activity of Fen1 (Figure 8C compare lanes 3, 9 and 15 to lane 2; also see panel D) indicating that these mutations did not substantially affect the interactions between PCNA and Fen1. When TIP was added to the reactions, the stimulatory effect of PCNA2-A on Fen1 activity was inhibited to a similar extent as observed with wild-type PCNA protein (Figure 8C compare lanes 10–14 to lanes 4–8; also see panel D). However, the inhibitory effects of TIP on PCNA2-E were substantially reduced (Figure 8C compare lanes 16–20 to lanes 4–8; also see panel D) suggesting a weaker interaction between TIP and the PCNA mutant. However, on size exclusion chromatography analysis TIP did not bind to either PCNA2-A or PCNA2-E (Supplementary Figure S7). We also attempted to use several peptides derived from TIP to inhibit PCNA-dependent Fen1 stimulation. However, none of the peptides either alone or in combination inhibited the activity.

The influence of the mutations in PCNA on the TIP–PCNA interactions was then evaluated using HDX-MS. The PCNA2-A mutant showed significant decreases in rate of deuterium uptake not only in region 187–204, but also in regions 39–48 and 218–229 (Figure 7A and Supplementary Figure S8). These regions are composed of short loops that are geometrically close to each other (blue regions in Figure 7), suggesting potential binding interfaces. The HDX-MS results for the PCNA2-A mutant suggest that these mutations strengthen the PCNA2–TIP interactions. Stronger interaction between the two proteins is supported by the observation that TIP has a stronger inhibitory effect on PCNA2-A stimulation of Fen1 at low TIP concentrations (Figure 8D). Surprisingly, the PCNA2-E mutant showed significant decrease of deuterium uptake only in region 39–48 (Figure 7A and Supplementary Figure S9). The absence of a deuterium uptake difference in region 187–204 underscores the importance of this region in modulating the PCNA2–TIP interactions.

Turning to an examination of the differential deuterium uptake of TIP and the TIP–PCNA2 complex, we noted that peptide regions 22–28 and 55–58 of the TIP–PCNA2 complex (Figure 9A and Supplementary Figure S10A) show slow deuterium uptake rates in the native PCNA2–TIP complex, suggesting potential binding interfaces on TIP in TIP–PCNA2 interaction. Moreover, the HDX kinetics data show that the deuterium uptake of peptide regions 22–28 and 55–58 of TIP in the presence of native PCNA2 eventually reaches the same level as that observed with TIP alone, indicating the flexible nature of the TIP structure. Not surprisingly, our results for the PCNA2-A mutant–TIP complex exhibit a similar trend of HDX kinetics in regions 22–28 and 55–58 of TIP (Figure 9B and Supplementary Figure S10B) as compared with the native PCNA2–TIP complex (Figure 9A). This similarity is in keeping with the formation of a stable PCNA2-A–TIP complex. In contrast, TIP alone and PCNA2-E mutant–TIP complex exhibit no significant difference in deuterium uptake behaviors (Figure 9C and Supplementary Figure S10C), indicating that the PCNA2-E mutant interaction with TIP is weak. In summary, our HDX analyses of PCNA2–TIP complexes show the importance of region 187–204 in regulating the PCNA2–TIP interactions.