Fig. 4. Distinct proteolysis kinetics of PL and CRY. Representative 7% SDS–PAGE gels stained with SYPRO Ruby Red that illustrate the proteolytic pattern of PL (A) or CRY (B) by PK as a function of reaction time (E:S = 1:750). Proteolysis time (hours) is given above each lane; asterisks denote minor cleavage products as in Fig. 2. The concentration of parent protein (fraction of initial concentration on logarithmic scales, averages of three experiments with standard errors) as a function of proteolysis time for PL and CRY is shown in (C). Lines are fits to first order exponential decays (t ? 0.5 h), or linear equations (t ? 1 h). Inset shows same data plotted on a linear scale; lines are fits to a double first order exponential decays.Figure optionsDownload full-size imageDownload as PowerPoint slideThe biphasic decay kinetics, evident predominantly for PL, indicate reduced rate of primary cleavage as daughter
CP-724714 accumulate during proteolysis. The simplest explanation is product inhibition, through interaction with either the protease or the parent protein. The former leads to cleavage of the daughter proteins (Fig. 4), and could account for reduced cleavage of parent PL. However, if significant, reaction of CRY daughter proteins should likewise compete for its primary cleavage. Alternatively, interaction of the daughter proteins or peptide fragments with the parent protein may inhibit primary cleavage by PK. Since this inhibition is significant only for PL, it may be dominated by interaction at its unique cleavage site in the N-terminal α/β domain. Consequently, we propose that the first order reaction seen for both CRY and PL at times ?0.5 h predominantly reflects cleavage in their recognition loops.DiscussionCorrelating proteolysis kinetics with recognition loop dynamicsThe initial kinetic data presented here reveal faster recognition loop cleavage and lack of a competing reaction site in CRY, as compared to PL. The observed rate constants include the Michaelis–Menton kinetics for enzyme reaction with a denatured polypeptide segment, kint = kcat/KM[E] ( Scheme 1), since the KM of PK [26] is in the range of 300–600 μM, the protein concentration is 1 μM, and the cleavable “substrate” concentration will be significantly lower [20] and [25]. As we have not identified the exact cut site within the recognition loops, we have not measured kint [20]. However, given that PK is largely non-specific, significant variations in kint are not expected. Consistent are results from proteolysis under denaturing conditions (0.5% SDS): Lacking the constraints of their secondary and tertiary structures, both parent proteins are cleaved much more rapidly, with kobs ～ 1 × 10?3 s?1, similar for PL and CRY within experimental error. Greater access to cleavable recognition loop conformations in CRY likely contributes significantly to its 10-fold higher reactivity than PL under native conditions.It is possible that PL and CRY proteolyze only when denatured [22] and [23], and that increased access to reactive conformations in CRY reflects lower global stability. Detailed investigations of the thermodynamic stability of these proteins will be required to address this fully. However, our current results are not consistent with a proteolytic mechanism that requires global unfolding. First, reaction of largely unfolded protein with the non-specific PK should generate many fragments [27], rather than the specific primary cleavage products observed. Second, these proteins precipitate significantly if denatured. When CRY or PL is incubated with BSA for the same time periods, we detect little loss (?10%) of soluble parent protein by UV–Vis spectroscopy or electrophoresis. Differences in accessibility of the recognition loop are also not readily ascribed to variations in the static structures of PL and CRY: The proteins possess 60% identical or closely related residues, >90% homology in secondary structure, and
base flipped conformation [14] and [17]. However, the binding affinity of CRY for a CPD in ssDNA regions is closely matched to PL, as required for their equivalent repair efficiency [13] and [14]. We propose that heightened recognition loop dynamics in CRY may contribute to its compromised base flipping. Base flipping enzymes access nucleosides by bending the DNA duplex, and rotating the target nucleoside(s) near 180°[15] and [16]. For PL, the activation barrier is estimated at ～5–7 kcal/mol [28], and base flipping is presumably facilitated by the ～30° kink in the DNA duplex induced by the CPD [29]. Induced-fit binding modes, engaging interactions with the flipped out dimer and complementary strand, likely also contribute significantly to the affinity of PL for its dsDNA substrate, but require ordering, and loss of conformational entropy, within the recognition loop. The more dynamic recognition loop of CRY may render this entropic penalty too large, and deactivate base flipping. By providing first data on distinct recognition loop dynamics in these proteins, our results offer a testable model for compromised base flipping by CRY.AcknowledgmentsWe thank the Natural Sciences and Engineering Research Council of Canada and the Michael Smith Foundation for Health Research for financial support.Appendix A. Supplementary dataSupplementary data. Help with PDF filesOptionsDownload file (7725 K)
