|Home | About | Journals | Submit | Contact Us | Français|
Phosphatidylinositol-specific phospholipases C (PI-PLCs) catalyze the hydrolysis of phosphatidylinositols (PI) to inositol phosphates (IP) and diacyl glycerol, which is one of the most important steps in transmembrane signaling pathways. The catalytic mechanisms of the bacterial Ca2+-independent and mammalian Ca2+-dependent PI-PLC have been thoroughly investigated.1–3 The results of the studies performed so far indicate that both PI-PLCs catalyze the conversion of PI to IP in two chemical steps, with formation of 1,2-cyclic myo-inositol phosphate (1,2-IcP) as an intermediate (Scheme 1).
Recently, we have embarked on a study of the Ca2+-dependent PI-PLC from Streptomyces antibioticus (saPLC1)4 that we considered a simpler and more readily available model to study the mechanism of complex mammalian isozymes.5 Because of its homology with PI-PLCδ1, we initially assumed the analogous mechanism for this enzyme.5 We now report that saPLC1 catalyzes the hydrolysis of PI via a different mechanism involving formation of an unusual trans-1,6-cyclic myo-inositol phosphate (1,6-IcP), rather then the common cis-1,2-IcP.
While the wild-type saPLC1 does not release the cyclic intermediate at any detectable concentration, the reaction catalyzed by the H16A mutant clearly shows formation of the cyclic phosphate (Figure 1) giving rise to a 31P NMR signal at 18.1 ppm, followed by its hydrolysis to inositol 1-phosphate at 4.5 ppm, as the end product. The cyclic phosphate produced in this reaction is different from 1,2-IcP formed by the earlier investigated PI-PLCs, as shown by the addition of the standard IcP (Figure 1C). The presence of the large H-2 coupling to phosphorus (3J1H-31P = 20 Hz) is indicative of the cis-geometry of the five-membered ring in 1,2-IcP (Figure 1D). The lack of such large 1H-31P coupling (< 2 Hz)) together with the characteristic chemical shift indicate that the cyclic intermediate produced by saPLC1 is 1,6-IcP (Scheme 1) that we obtained earlier by chemical cyclization of inositol phosphoesters.6
In order to further characterize this intermediate we have prepared it by a combined chemical and biochemical approach. The treatment of myo-inositol S-benzyl phosphorothiolate with ammonium hydroxide produced a 3:1 mixture of 1,2- and 1,6-IcP.6 The mixture was ttreated with B. thuringiensis PI-PLC (btPLC) that resulted in the selective conversion of the cyclic phosphate at a higher field (cis-1,2-IcP) to inositol 1-phosphate, as expected (Figure 2A, middle). The remaining 1,6-cyclic phosphate was isolated and characterized by 1H, 31P NMR and MS (See Supplementary Materials). Similarly, the treatment of the mixture of myo-inositol 1,2- and 1,6-cyclic phosphate 4-phosphates with mammalian PLCδ1 resulted in the selective hydrolysis of 1,2-cyclic phosphate (Figure 2B). These experiments clearly confirm that neither of the two major types of enzymes does not process the 1,6-cyclic phosphate. In contrast, the treatment of (1,2+1,6)-IcP with saPLC1 resulted in the disappearance of the lower field signal to produce inositol 1-phosphate as a final product. We can estimate that the rate of the cleavage of 1,6-IcP by saPLC1 is at least 103-times higher than that of 1,2-IcP.
The formation of a five-membered cyclic transition state with cis-arrangement of the attacking OH nucleophile and the phosphoryl group has been a characteristic feature of phosphoryl transferases such as RNase A and PI-PLC. For RNase, this mechanism is favored by the flexibility of the ribose ring system and close juxtaposition of the attacking 2’-hydroxy group. As we have shown earlier,6 however, the cis-cyclization of inositol phosphodiesters is several hundred times slower than that of ribose phosphodiesters, and there is only a minor advantage for the cis- vs. trans-cyclization. Notwithstanding, all the previously investigated PI-PLC species adopted the same mechanistic paradigm as that of RNase A. Our study demonstrate for the first time that PI-PLC-catalyzed P-O ester bond cleavage reaction can proceed through an alternative, albeit somewhat less favorable, trans-cyclization process to form a 1,6-cyclic intermediate. It is quite possible that the natural target of saPLC1 could be a different isomer of an inositol phospholipid, such as scyllo-PI, for which only trans-cyclization is possible. The presence of scyllo-PI has been documented in both bacterial and plant systems,7–9 and scyllo-inositol itself is abundant in mammalian systems and other vertebrae,10 but to date no enzyme capable of cleaving scyllo-PI has been discovered.
This work was support by the grant from NIGMS GM 57568.