2.2.5 MASAJE INFANTIL
2.2.5.2 Beneficios del Masaje Infantil
activity
5.1 Summary
In addition to their lipid kinase activity, all class I PI 3-kinases possess a protein kinase activity. The best characterised substrate for this activity is the enzyme itself. More specifically, p i 10a predominantly phosphorylates its associated regulatory subunit p 8 5 a and it also autophosphorylates to a lesser extent. In contrast, the catalytic subunits p i 10(3 and p i 10Ô predominantly autophosphorylate. The class IB PI 3-kinase catalytic subunit p llO y also autophosphorylates. Additionally, evidence has been presented that two proteins, namely 1RS-1 and PDE3B, may serve as exogenous (i. e. not tightly associated with the enzyme) substrates for the protein kinase activity o f PI 3- kinase. The identification o f two novel potential protein substrates o f PI 3-kinase is described here. By employing in vitro kinase assays using recombinant proteins as the substrates, it is shown that the translational regulator 4EBP1 and the small GTPase H- Ras become phosphorylated by PI 3-kinases. This raises the possibility that these proteins could serve as physiological substrates for this activity. To further assess the likelihood that the protein kinase activity o f PI 3-kinase operates in a physiological context by phosphorylating substrates other than the enzyme itself, mutated versions o f PI 3-kinase were constructed in such a manner that the resulting mutant enzymes have a much less lipid phosphorylating capability, but they still retain their protein phosphorylating activity. The potential applications o f these mutants are discussed.
5.2 Introduction
All isoforms o f class I PI 3-kinase are dual specificity kinases possessing a serine kinase activity in addition to their lipid kinase activity. The protein kinase activity o f PI
3-kinases is manifested by their autophosphorylation. The p llO a catalytic subunit predominantly phosphorylates its associated regulatory subunit p 8 5 a and it also autophosphorylates to a lesser extent (Carpenter et a l, 1993b; Dhand et al., 1994b). In contrast, the catalytic subunits p i 10(3 and p i 106 predominantly autophosphorylate (Foukas et al., 2002; Vanhaesebroeck et al., 1999a). The class IB PI 3-kinase catalytic subunit p llO y also autophosphorylates (Stoyanova et al., 1997). Furthermore, evidence has been presented that some other proteins could be targets o f this activity. Reportedly, protein substrates for PI 3-kinase are IRS-1 (Lam et al., 1994; Tanti et al., 1996; Uddin et al., 1997) and PDE3B (Rondinone et al., 2000). However, only in the case o f p 8 5 a has the site definitively been identified, with phosphorylation occuring largely on Ser 608 (Dhand et al., 1994b).
In an effort to further define the role o f this protein kinase activity, we sought to identify other cellular substrates for the PI 3-kinase serine kinase by searching for proteins that participate in pathways regulated by PI 3-kinase activation, become phosphorylated in a wortmannin-sensitive manner and encompass phosphorylatable aminoacids surrounded by sequences similar to that o f the p 8 5 a Ser608 site (i. e. being embedded in an acidic environment). One such molecule is the translational regulator eukaryotic Initiation Factor 4E (eIF4E)-Binding Protein 1 (4EBP1), also known as Phosphorylated Heat and Acid Stable Protein I (PHAS I) (Gingras et al., 1999). Initiation is the rate-limiting step in the process o f translation (Gingras et al., 1999; Proud and Denton, 1997). Almost all eukaryotic mRNAs have the cap structure (7- methyl-GTP-N, where N is any nucleotide) at their 5’ termini. The eukaryotic Initiation Factor 4E is the mRNA cap-binding protein that associates with eIF-4G, thus forming the eIF-4F complex which facilitates the initiation o f translation. In resting cells, non- phosphorylated 4EBP1 binds and prevents eIF4E from taking part in the initiation complex. Upon agonist stimulation o f cells, 4EBP1 becomes phosphorylated in a wortmannin-sensitive manner. This phosphorylation occurs at multiple sites and causes
dissociation o f the eIF4E/4EBPl complex thus allowing the formation o f competent eIF-4F complexes.
Six phosphorylation sites have been identified on 4EBP1. Five o f them fall within a (Ser/Thr)Pro motif. One o f the (Ser/Thr)Pro sites, Ser82 is constitutively phosphorylated. The rest (Thr36, Thr45, Ser64, and Thr69) become phosphorylated after insulin stimulation o f cells in a wortmannin- and rapamycin-sensitive manner. Phosphorylation o f Ser64 depends on the previous phosphorylation o f all three threonine sites (Mothe-Satney et al., 2000). The sensitivity o f the above sites to rapamycin suggests that their phosphorylation is mediated by the mTOR pathway. As supporting evidence, it has been shown that mTOR itself as well as an mTOR- associated kinase directly phosphorylate these sites (Heesom and Denton, 1999). The sixth site, Seri 11, becomes phosphorylated in a rapamycin-insensitive, but yet wortmannin-sensitive manner. An insulin-stimulated Seri 11 kinase with an approximate molecular weight o f ISOkDa has been partially purified from primary rat adipocytes (Heesom et al., 1998). Furthermore, the Seri 11 site has also been shown to become phosphorylated by protein kinase CK2 in vitro (Fadden et al., 1998) and it also conforms to the recently reported consensus phosphorylation site o f the PI 3-kinase- related kinases (Kim et al., 1999), which phosphorylate 4EBP1 in vitro (Denning et al., 2001; Kim et al., 1999; Sarkaria et al., 1998).
Furthermore, in the course o f the present study, we found that the small GTPase Ras, a long-known binding partner o f class I PI 3-kinases, becomes phosphorylated by the protein kinase activity o f the latter. In vivo phosphorylation is a long-known, although overlooked, feature o f the H-Ras and K-Ras proteins. It has been reported that H-Ras possesses autokinase activity that strictly uses GTP as phosphoryl donor to autophosphorylate on Thr59 (Shih et al., 1982). Subsequently, it was shown that phorbol esters and permeable c-AMP derivatives stimulate K-Ras (with exon 4B) and and to a lesser extent H-Ras phosphorylation in intact cells (Ballester et al., 1987;
Saikumar et al., 1988). These phosphorylations were further demonstrated in vitro using purified PKA and PKC. However, a more recent study failed to detect stoichiometric phosphorylation o f H-Ras by PKA (Arimura et al., 1997). Furthermore, the phosphorylation sites have been mapped. Protein kinase C phosphorylates K-Ras (with exon 4B) on S e ri81 (Ballester et al., 1987) and H-Ras on S e ri77 (Jeng et al., 1987; Saikumar et al., 1988). These phosphorylation sites lie in the hypervariable region, which links the globular catalytic domain o f p21 Ras to the membrane- anchoring site at the C-terminus, a location suggesting that this phosphorylation may play a role in modulating transmembrane signaling.
As discussed above in detail, the best evidence that the PI 3-kinase protein kinase activity plays a direct role in downstream signalling comes from studies with mutant forms o f the class IB PI 3-kinase, which lack lipid kinase but retain protein kinase activity (Bondeva et al., 1998; Pirola et al., 2001). These mutants retained the ability to stimulate the MAP kinase pathway although the molecular targets o f the PI 3-kinase serine kinase were not identified (Bondeva et al., 1998). The strategy followed for the construction o f these mutants involved whole domain swapping between different members o f the PI 3-kinase family. This approach results in enzymes with grossly mutated catalytic domains and this might interfere with proper recognition o f their physiological protein substrates. Substitution o f specific aminoacids responsible for lipid substrate recognition could be a potentially less interfering strategy compared to domain swapping since it requires only subtle changes. The residues critical for such strategies have been identified by the recently reported crystal structure o f p i lOy, which identified the aminoacids that play a role in lipid substrate recognition (Walker et al., 1999). According to the model proposed in that study, the Lys973 interacts with the 5’ phosphate o f PtdIns-4,5-P2- Therefore, mutation o f this residue would presumably render the enzyme incapable o f phosphorylating PtdIns-4,5-P2, which is thought to be its physiological substrate.
In the following section, the identification o f two novel potential substrates for the PI 3-kinase protein kinase activity is described. Also, it is demonstrated that certain point mutations within the activation loop o f p i 10a and p i lOp abolish the lipid but not the protein kinase activity. These mutants could be used as a tool for the identification o f protein substrates for PI 3-kinase in vivo.
5.3 Results
5.3.1 Class lA PI 3-kinase phosphorylates 4EBP1
in v i t r o in
a wortmannin-sensitive manner
To test the possibility that PI 3-kinase could phosphorylate 4EBP1, class lA PI 3- kinase was immunoprecipitated from HEK 293 lysates using p85 antibodies. After extensive washing, the immunoprecipitates were subjected to an in vitro kinase assay with recombinant 4EBP1 and [y-^^P]ATP as the substrates. It was found that recombinant 4EBP1 is indeed phosphorylated by a kinase activity present in p85 immunoprecipitates (Fig. 5.1 A). This activity was largely inhibited with lOOnM wortmannin consistent with the phosphorylation being caused by the serine kinase activity o f a member o f the PI 3-kinase superfamily. Furthermore, the ability o f PI 3- kinase to phosphorylate 4EBP1 when the latter is in complex with eIF4E was tested. To this end, in vitro kinase assays were performed using p85 immunoprecipitates as a source o f PI 3-kinase and either free 4EBP1 or 4E/4EBP1 complex as a substrate in the presence or the absence o f lOOnM wortmannin. As it can be seen in Fig. 5. IB, 4EBP1 was phosphorylated with the same efficiency either when fi-ee or when in complex with eIF4E.
5.3.2
The
features
of
the
protein
kinase
activity
phosphorylating 4EBP1 are consistent with those for p85a
Ser608 autophosphorylation
It has been reported that both lipid kinase and protein kinase activities o f PI 3-kinase are inhibited by nM concentrations o f wortmannin (Woscholski et al., 1994b). Therefore the dose-response o f 4EBP1 phosphorylation to wortmannin was tested. PI 3-kinase immunoprecipitated from HEK 293 cell lysates using p85 antibodies was preincubated in the presence o f wortmannin at various concentrations for 15min and then a kinase assay with recombinant 4EBP1 as a substrate was performed. As it can be seen in Fig. 5.2 A, wortmannin potently inhibited 4EBP1 phosphorylation by p85 immunoprecipitates with an IC50 o f approximately 40nM, which is relatively close to the IC50 for the PI 3-kinase serine kinase activity reported to be 17.1 nM (Woscholski et al., 1994b). Furthermore, the divalent cation requirement for 4EBP1 phosphorylation by p85 immunoprecipitates was assessed by performing in vitro kinase assays in the presence o f either Mn^^ or Mg^^. It was found that phosphorylation was strictly dependent on the presence o f Mn^^ and this could not be substituted by Mg^^ (Fig. 5.2B). This is also in keeping with the previously reported cation dependence for p85 autophosphorylation which was also found to be strictly dependent on the presence o f Mn^^ (Carpenter et al., 1993b; Dhand et al., 1994b). Therefore, the features o f the kinase activity phosphorylating 4FBP1 are consistent with the phosphorylation being caused by the serine kinase activity o f class lA PI 3-kinase.