5. DESENVOLUPAMENT DE LA PROPOSTA DIDÀCTICA
5.2. Proposta didàctica: dos models d’itineraris didàctics
5.2.2. L’itinerari per Son Peretó: Adaptem Son Peretó
Results of equilibrium cleavage suggested that some mutants at H289, including glutamine, are partially defective in ligation since they accumulate large amounts of covalent intermediate. Ligation activity has been measured directly, independent of cleavage in Cre, vaccinia
topoisomerase (Woodfield, Cheng, Shuman & Burgin, 2000)and Flp (Whiteson et al., 2007) by
using a synthetically modified DNA substrate containing a tyrosine analog attached to the 3' end of a scissile phosphate. The tyrosine analog pNP (para-nitro-phenol) with a pKa of 7.1 was esterified to the 3' phosphate of either a TS or BS substrate to mimic an activated tyrosyl leaving group during ligation. The modified DNA substrate was radiolabeled and annealed to
complementary and adjacent strands, forming a three stranded loxP substrate containing a nick
on either the TS or BS at the scissile phosphate as depicted in Fig. 3.11. Cre catalyzes the ligation of the two adjacent strands by release of the pNP without the need for Y324 (this work) (Woodfield et al., 2000). Fig. 3.11 shows ligation of Cre mutants using TS and BS pNP
substrates at pH 7.5. At this pH, ligation does not require a proton donor to assist the pNP leaving group. The 60 minute reactions were quenched with SDS and digested with proteinase K to remove covalently attached Cre and then separated by denaturing PAGE. The unligated modified 36mer (BS) or 26mer (TS) substrate containing pNP is a single band with greater mobility than the ligated 64mer loxP product. A second band appears above the unreacted TS substrate for
mutants that contain Y324 but not those containing the Y324F mutation indicating that this species is the proteinase K digested product of a phosphotyrosine covalent intermediate. The product of the ligation reaction is an intact loxP site, so active Cre mutants are able to cleave the
ligated loxP product and generate covalent intermediate. However, K201A, which is inactive for
cleavage, accumulates significant amounts of covalent intermediate with the ligation
The observed covalent intermediate could be due to the direct attack of the pNP by tyrosine (Fig. 3.11B). If pNP assumes a position in the active site similar to O5' of an intact loxP site, then
tyrosine is able to attack the pNP directly in a mechanism similar to cleavage. K201A is able to generate covalent intermediate because the pKa of pNP is much lower than 5'-OH, and as has been determined with 5'-bridging phosphorothiolate substrates (Ghosh et al., 2005a; Krogh &
Shuman, 2000), the leaving group does not require a proton donor. K201A does not generate any appreciable ligated loxP product, consistent with its role in activating the incoming 5'-OH. Any Cre mutants containing Y324 could exploit this secondary pathway to ligated products
complicating analysis. For this reason, any mutants studied by this assay should also mutate Y324.
WT Cre and a selection of mutants were tested for BS and TS ligation of pNP modified substrates under conditions unfavorable (0.2 nM substrate) and favorable (50 nM substrate) for synapsis formation. Ligation product was normalized for each mutant to the maximum value measured with each substrate. The data shown in Fig. 3.11 is consistent with strand preference observations made for Cre cleavage, where WT Cre and Y324F are highly active on the TS half site in the absence of synapsis, and only become active at the BS half site in conditions favoring synapsis (Ghosh et al., 2005a). Congruent with this observation, the A36V mutant, which is
defective in synapsis is robust at ligation of the TS and deficient in BS ligation at both substrate concentrations. In general, Cre appears less active on the BS ligation substrate than on the TS substrate. However, time course reactions comparing both substrates demonstrate that ligation is nearly complete at sixty minutes, indicating that 100% conversion to product will not occur on the BS substrate (data not shown). Perhaps an artifact during synthesis of the BS-pNP generated a fraction of unreactive substrate.
H289Q also appears to be more active on the TS ligation substrate than on the BS ligation substrate. If the H289Q generated covalent intermediate observed in the TS reaction is from ligation product that was then a substrate for H289Q cleavage, then the actual extent of ligation
for H289Q is nearly double that reported in Fig. 3.11. However, since the H289Q/Y324F double mutant does not generate this additional ligation product, either Y324 is aiding ligation or the covalent intermediate generated by direct attack on pNP by Y324. Although H289Q and H289Q/Y324F have similar extents of ligation, the H289A and H289A/Y324F mutants are significantly different. The H289A mutant generates ligation product on the order of H289Q, however the H289A/Y324F double mutant abolishes ligation activity. Perhaps the hydroxyl group of Y324 aids in stabilization of phosphate in the absence of H289. Alternatively, H289A may have greater ligation activity on the covalent intermediate than with the pNP substrate. The disruption of ligation observed with H289A and H289Q, particularly the double mutants with Y324F, support a role for H289 in ligation. However, since pNP does not require an acid to donate a proton at physiological pH, these experiments are unable to determine if H289 is acting as a proton donor in ligation.
Figure 3.11
Ligation of pNP modified TS and BS loxP DNA substrates by Cre
A. Bottom strand ligation of labeled 36mer DNA substrate (S) containing a 3'pNP, to an adjacent unlabeled 30mer to form a full 66mer loxP BS (P). B. Top strand ligation of labeled 25mer DNA substrate (S) containing 3' pNP, to an adjacent 40mer to form a full 66mer loxP TS (P). Mutants containing Y324 were found to accumulate Cre covalent intermediate via a 3'-phosphotyrosine linkage, which when digested with proteinase K runs just above the substrate on the gel (Cov pTYR). Covalent intermediate can form via two distinct mechanisms. The first involves direct attack of pNP phosphate by Y324 and the second involves Y324 mediated cleavage of the intact loxP product resulting from ligation. (A and B) Reactions were performed in conditions favorable (50nM substrate) and unfavorable for synapsis (0.2 nM substrate). The 60 minute reactions were quenched and separated by UREA Page. Reactions for WT, Y324F and H289Q are the averaged value of three experiments with associated standard deviation.3.4
Discussion
Fifteen amino acid substitutions at 289 are tolerated by Cre as measured by recombination
in vivo and in vitro, which prompted further study to understand the function and reason for
conservation in YRs. The cleavage activity of the purified H289 mutants correlated well with the observed in vitro recombination activity. Studies of this histidine residue in YRs have suggested a
role in ligation that is primarily based on the accumulation of covalent intermediate by the H305L in Flp (Pan et al., 1993b; Parsons et al., 1988; Serre & Jayaram, 1992), H240L in XerC and
H244L in XerD (Arciszewska et al., 2000; Cornet et al., 1997),and H234L in TnpI (Vanhooff et al.,
2009). The ligation deficiency of H305Q in Flp was observed directly using synthetic substrates (Whiteson et al., 2007). All nineteen H289 mutants in Cre were studied for cleavage of an intact loxP site under equilibrium conditions. The Gln, Leu and Met mutants all accumulated covalent
intermediate at levels above WT, but this does not appear to be a general property, as the other H289 mutants accumulated covalent intermediate at levels similar to or less than WT Cre. Accumulation of covalent intermediate under equilibrium conditions suggests that the rate of cleavage has increased relative to the rate of ligation. Since cleavage by H289Q, H289L and H289M is slower than WT Cre, a defect in ligation is mostly likely responsible for the buildup of covalent intermediate. These three mutants are the most competent in suicide cleavage, but several other mutants that do not accumulate covalent intermediate such as the alanine and asparagine are nearly as active.
Leu, Met and Gln are good structural mimics of histidine. Amino acid substitutions at 289 with branch points at Cβ are disruptive to activity. This is evident by comparing the activities of Leu, Ile and Val as well as Ser and Thr. Modeling suggests that smaller amino acids such as alanine, asparagine and serine leave enough space to accommodate a water molecule in a similar position as Nε of histidine as observed in transition state structure of TopIB (Davies et al.,
space occupied by histidine and may prevent solvent from efficiently assisting in ligation. Structures of the H289A and H289Q mutants done in collaboration with Kushol Gupta in the Van Duyne group are consistent with this hypothesis and are shown in Fig. 3.12. with statistics presented in Table. 3.3. Clear density for a well ordered water molecule in H289A is observed in the same space as the histidine in the active subunit. As expected, glutamine occupies the same general space as histidine, capable of providing stabilizing hydrogen bonds to the scissile phosphate, but unable to participate in acid/base catalysis. The density for the glutamine is well- defined, with no evidence of density corresponding to solvent in this region.
Despite a possible defect in ligation, H289Q is the most active of all the histidine mutants, which underscores the importance of stabilizing hydrogen bonds to the scissile phosphate. The pH profiles examining cleavage demonstrated that assistance of a general base is not required to activate the tyrosine for cleavage. This may be due in part to the differences in the pKa of tyrosine (~10) compared to the 5'-hydroxyl group (~15), and is consistent with observations from model reactions of nucleophilic substitution at phosphodiester linkages that show a modest dependence on the basicity of the nucleophile (βnuc ~ 0.2-0.3) but a large dependence on the basicity of the
leaving group (βlg ~ -0.8) (Cassano, Anderson & Harris, 2004; Kirby & Younas, 1970a; Kirby &
Younas, 1970b; Maegley, Admiraal & Herschlag, 1996). Therefore, the dependence on a general acid to assist the leaving group O5' is more important than a base to aid in the activation of the tyrosine nucleophile. During ligation, the histidine in YRs or water in TopIB would serve as an acid to assist in protonating the tyrosine leaving group. Experiments using modified tyrosine peptides were used to argue that H305 in Flp recombinase is acting as a general base in cleavage and an acid in ligation (Whiteson et al., 2007). In these experiments, Flp H305Q was rescued for
cleavage using a modified flourotyrosine peptide with a lower pKa than tyrosine, however Flp H305A was not rescued. Histidine is perhaps the only amino acid that could provide an important contribution to stabilization of the transition state and also be able to participate in acid/base
catalysis. Although water can substitute in acid/base catalysis, other amino acids at this position are either unable to stabilize the transition state or sterically exclude water from effectively participating in catalysis.