Of yeast and men

Summary

The evolution of PtdIns(3,4,5)P₃ synthesis

Since the identification of the lipid second messenger phosphatidylinositol (3,4,5)-trisphosphate (PtdIns(3,4,5)P₃; Traynor-Kaplan et al, 1988; Auger et al, 1989), this molecule has become a paradigm for how phosphoinositides (PIs) mediate complex signalling events in cells. PtdIns(3,4,5)P₃ can regulate chemotaxis, cell survival, translation and cytoskeletal rearrangement (Cantley, 2002), the deregulation of which contributes to the pathology of diseases such as diabetes, and breast and prostate cancer.

PtdIns(3,4,5)P₃ is a phosphorylated form of the PI phosphatidylinositol (PtdIns). All PIs have two fatty-acid chains that anchor them to membranes and a hydrophilic inositol head group that is exposed to the aqueous cytosol (Fig 1A). This head group can be phosphorylated at the 3-, 4- and 5-positions and so far, a total of seven phosphorylated derivatives of PtdIns have been reported (Fig 1A).

Figure 1.

Pathways of PtdIns(3,4,5)P₃ synthesis. (A) Phosphatidylinositol (PtdIns) is the basic building block of phosphorylated phosphoinositides (PIs). The inositol head group can be phosphorylated in the 3-, 4- and 5-positions to generate derivatives of PtdIns, which are listed on the right. (B) Schematic for receptor-driven PtdIns (3,4,5)-trisphosphate (PtdIns(3,4,5)P₃) production. Agonist stimulation (tyrosine kinase or G-protein-coupled receptors) activates class I PI 3-kinase to induce phosphorylation of PtdIns(4,5)P₂ on the 3-position of the inositol group. (C) The proposed new pathway for PtdIns(3,4,5)P₃ production in Schizosaccharomyces pombe. Fission yeast enzymes are shown in blue. Their mammalian counterparts are depicted in turquoise. PIP 5-K, phosphatidylinositol-4-phosphate 5-kinase; PTEN, phosphatase and tensin homologue.

After the treatment of cells with certain agonists, the transient synthesis of PtdIns(3,4,5)P₃ is initiated by class I PI 3-kinases. These lipid kinases are normally inactive; however, agonist stimulation increases their ability to phosphorylate the 3-position of PtdIns(4,5)P₂ to generate PtdIns(3,4,5)P₃ (Hawkins et al, 1992). The receptor-driven activation of class I PI 3-kinases can give rise to greater than a tenfold increase in the cellular levels of PtdIns(3,4,5)P₃ (which rarely approach 0.1% of total PI levels, making its detection difficult—even for the more hardened lipid biochemists among us). In turn, subsets of proteins that specifically recognize and bind to PtdIns(3,4,5)P₃ (mediated by specialist lipid-binding modules such as pleckstrin homology (PH) domains) are then recruited to the membrane, which often leads to their activation and further transduction of the agonist signal (Fig 1B; Lemmon, 2003).

Recently, PtdIns(3,4,5)P₃ has stepped into the scientific limelight once again, this time not because of its biological function but intriguingly because of the way it is synthesized in fission yeast. A recent article in the Journal of Cell Biology (Mitra et al, 2004) has revealed for the first time that Schizosaccharomyces pombe can generate the 3-phosphorylated lipid PtdIns(3,4,5)P₃. This is exciting because researchers have assumed that yeast are incapable of producing PtdIns(3,4,5)P₃. This dogma has arisen as PtdIns(3,4,5)P₃ has never been found in yeast, and yeast do not have a class I PI 3-kinase, which suggests that they lack one of the essential reaction steps necessary to generate PtdIns(3,4,5)P₃.

The paper by Mitra and co-workers begins with the identification of ptn1, a homologue of the mammalian phosphatase and tensin homologue (PTEN) gene. PTEN is a lipid phosphatase (Maehama et al, 1998), which dephosphorylates the 3-position of PtdIns(3,4,5)P₃ to generate PtdIns(4,5)P₂. When the authors tested the lipid phosphatase activity of Ptn1 in vitro, they found that it also dephosphorylated PtdIns(3,4,5)P₃. This led Mitra and colleagues to question why S. pombe would have maintained a functional PtdIns(3,4,5)P₃ phosphatase in its genome if not to regulate PtdIns(3,4,5)P₃ levels.

In mammalian cells, disruption of PTEN activity leads to elevated PtdIns(3,4,5)P₃ levels and surprisingly, examination of the phosphoinositide levels in an S. pombe strain lacking ptn1 (ptn1Δ) revealed a dramatic increase in PtdIns(3,4)P₂ and PtdIns(3,4,5)P₃. This finding is noteworthy in itself; however, the authors went on to postulate that S. pombe synthesizes PtdIns(3,4,5)P₃ through a pathway that is different to the one used by mammalian cells after growth factor stimulation.

Yeast have only one PI 3-kinase homologue, Vps34, which regulates vesicle transport to the vacuole (Takegawa et al, 1995). When a strain lacking vps34 was crossed with ptn1Δ, the authors found that levels of PtdIns(3,4,5)P₃ and PtdIns(3,4)P₂ decreased. Unlike class I PI 3-kinases, Vps34 can only phosphorylate PtdIns to generate PtdIns(3)P in vitro, indicating that it is not the 3-phosphorylation of PtdIns(4,5)P₂, but the synthesis of PtdIns(3)P that is the important step in S. pombe PtdIns(3,4,5)P₃ synthesis. Mitra and colleagues then turned their attention to the kinases that might phosphorylate the 4- and 5-positions of the inositol head group. The lipid kinase Fab1 can phosphorylate PtdIns(3)P at the 5-position; however, its deletion did not affect the PtdIns(3,4)P₂ or PtdIns(3,4,5)P₃ production in ptn1Δ cells, indicating that there was no role for Fab1 in PtdIns(3,4,5)P₃ synthesis. Next, they examined Its3, the S. pombe homologue of the mammalian phosphatidylinositol-4-phosphate 5-kinase (PIP 5-K) family, which are key enzymes in the regulation of PtdIns(4,5)P₂ levels. When a strain retaining only 10% of wild-type PIP 5-K activity was crossed with ptn1Δ, levels of PtdIns(4,5)P₂ were lowered as expected. Surprisingly, PtdIns(3,4)P₂ and PtdIns(3,4,5)P₃ levels were also dramatically attenuated. As PtdIns(4,5)P₂ is not a substrate for Vps34, this suggests a novel role for Its3 in PtdIns(3,4,5)P₃ synthesis. (It is possible that yeast have a PtdIns(4,5)P₂ 3-kinase that is yet to be identified or that PtdIns(4,5)P₂ can allosterically regulate PtdIns(3,4,5)P₃ synthesis.) In vitro, mammalian PIP 5-Ks can sequentially phosphorylate PtdIns(3)P at the 4- and 5-positions to generate PtdIns(3,4,5)P₃ (Zhang et al, 1997), and this led Mitra and colleagues to propose that PtdIns(3)P generated by Vps34 is converted to PtdIns(3,4,5)P₃ by Its3 (Fig 1C).

What is the function of PtdIns(3,4,5)P₃ in yeast? Although the ptn1Δ strain displayed vacuolar defects and was more susceptible to osmotic stress than wild-type yeast, neither of these observations were directly linked to elevated PtdIns(3,4)P₂ and PtdIns(3,4,5)P₃ levels. Importantly, the authors demonstrated that at least two S. pombe proteins can bind PtdIns(3,4,5)P₃ in vitro, with the intracellular localization of one of them showing some dependency on the presence of Ptn1. This raises the possibility that PtdIns(3,4,5)P₃-dependent signalling networks exist in yeast, analogous to those described in metazoan systems.

So, what are the consequences of these findings? The fact that the PIP 5-K-driven PtdIns(3,4,5)P₃ synthesis pathway is present in yeast indicates that, historically, this was the original pathway that organisms evolved to generate PtdIns(3,4,5)P₃. Class I PI-3-kinases evolved much later, which suggests that the growth-factor-driven pathway found in Caenorhabditis elegans and mammalian cells is a 'new kid on the block'. If so, then what became of the PIP 5-K-driven PtdIns(3,4,5)P₃ pathway in mammalian cells? Intriguingly, whereas growth-factor-stimulated PtdIns(3,4,5)P₃ synthesis is dependent on class I PI 3-kinases in mammalian cells, we recently showed that certain stress agonists, such as H₂O₂, increase PtdIns(3,4,5)P₃ levels in vivo by the PIP 5-K-mediated 5-phosphorylation of PtdIns(3,4)P₂ (Halstead et al, 2001).

Finally, on a clinical note, PTEN (a tumour suppressor) deletion in human cancers leads to the activation of cell survival pathways through the upregulation of PtdIns(3,4,5)P₃ levels. In light of these new findings that connect PTEN with PIP 5-K-driven PtdIns(3,4,5)P₃ synthesis, it would be interesting to dissect how PtdIns(3,4,5)P₃ is generated in these cancer cells.

At the moment, this new, but ancient, biochemical pathway for PtdIns(3,4,5)P₃ synthesis seems to be an 'outsider' in the field of cell biology. Hopefully, with further study, it too can find a place of its own.