ITPK1 (Inositol Tetrakisphosphate 1-Kinase)
Historical Background: Evolution of the Concept that ITPK1 Is a Cell-Signaling Enzyme
A 6-kinase activity that phosphorylates Ins(1,3,4)P3 to Ins(1,3,4,6)P4 (EC 22.214.171.124) was originally described in 1987 by two separate groups working independently (Balla et al. 1987; Shears et al. 1987). The gene encoding the 46 kDa 6-kinase was cloned almost 10 years later (Wilson and Majerus 1996). The enzyme also has a less active 5-kinase activity towards Ins(1,3,4)P3 (Shears 1989; Wilson and Majerus 1996), and so was generally referred to as an “inositol 1,3,4-trisphosphate 5/6-kinase.” This prompted the HGNC (the HUGO Gene Nomenclature Committee; http://www.genenames.org/) to accept that the corresponding gene could be named as ITPK1. Subsequently, the enzyme was found to also add a 1-phosphate to Ins(3,4,5,6)P4 (Yang and Shears 2000). This was a particularly significant observation from a cell signaling perspective, as Ins(3,4,5,6)P4 had already been established to regulate chloride flux across the plasma membrane (Vajanaphanich et al. 1994; Xie et al. 1998). The number of physiological processes regulated by this effect of Ins(3,4,5,6)P4 has since expanded significantly. In parallel, the HGNC-approved name of the gene has been changed to “inositol-tetrakisphosphate 1-kinase”; it appears to have been determined that this updated description could be accommodated without incurring the inconvenience of altering the original ITPK1 designation.
Mammalian genomes encode a single ITPK1 gene. In plants there are often four to six ITPK1 genes that encode different isoforms. The reason for this diversity in plants is not well understood, but likely it reflects metabolic specialization of ITPK1 for particular substrates, and/or the capability (or not) to perform phosphotransferase activity.
Human ITPK1 is comprised of 12 exons on chromosome 14. The mRNA transcripts of ITPK1 were found to be ubiquitously expressed in tissues with highest expression levels in the brain and the heart (Wilson and Majerus 1996). Alternative splicing at the 5′ untranslated region is possible, although the significance is unknown (Yang and Shears 2000).
A protein kinase activity was found to be associated with recombinant human ITPK1 that had been expressed in insect cells (Wilson et al. 2001). The latter study proposed that the protein kinase activity was performed by ITPK1 itself. This hypothesis potentially added substantial significance to the signaling capacity of ITPK1. However, subsequent work (Qian et al. 2005) demonstrated conclusively that this protein kinase activity is performed by a separate, contaminating protein. When freed of this contaminant, ITPK1 does not exhibit any protein kinase activity (Qian et al. 2005). The conclusion that ITPK1 cannot be a protein kinase was reinforced when the crystal structure of the human ITPK1 was solved (Chamberlain et al. 2007): peptides cannot be modeled into its active site.
ITPK1 and the De Novo Synthesis of Inositol Phosphates
ITPK1 has a purely metabolic function; its Ins(1,3,4)P3 6-kinase activity generates Ins(1,3,4,6)P4 as a precursor for the synthesis of Ins(1,3,4,5,6)P5, as well as InsP6, and ultimately the inositol pyrophosphates. Indeed, overexpression of ITPK1 in mammalian cells leads to increased levels of Ins(1,3,4,6)P4, Ins(3,4,5,6)P4, InsP5, and InsP6, whereas depletion of ITPK1 by RNAi results in decreased levels of these products (Verbsky et al. 2005). Nevertheless, the IPMK-mediated phosphorylation of Ins(1,4,5)P3 to Ins(1,3,4,5,6)P5 may by-pass ITPK1 in some cell types (Fig. 2). Yeasts and Drosophila exclusively use the IPMK-dependent pathway, since they lack an ITPK1 gene.
Mouse embryos that, through a gene-trap approach, were made hypomorphic for ITPK1 expression were found to exhibit neural tube defects at high frequency (Wilson et al. 2009). It is therefore intriguing that maternal ITPK1 expression has been found to be reduced in human populations with a high risk of neural tube defects during embryonic development (Guan et al. 2014). These pathological effects might reflect reductions in levels of InsP6 (Guan et al. 2014), but attenuation of Ins(3,4,5,6)P4-signaling might also make a contribution.
Structure of ITPK1
The crystal structure of two orthologs of ITPK1 has been solved – those from humans (Chamberlain et al. 2007) and Entamoeba histolytica (Miller et al. 2005). These data show that the ATP is sandwiched between two sets of four-stranded, antiparallel β-sheets, a structure commonly described as an ATP-grasp fold. In the human enzyme, it has been estimated that only 5% of bound nucleotide is solvent accessible, suggesting that there are substantial conformational changes whenever ATP is bound and ADP is released. However, for efficient operation of the phosphotransferase cycle of ITPK1, it is expected that the enzyme tightly binds nucleotide, so that neither ATP nor ADP is released into the bulk phase during this series of reactions (Fig. 1).
Biological Significance of ITPK1-Mediated Cell Signaling
The direct role of ITPK1 in signal transduction stems from its ability to couple PLC activation to the synthesis of Ins(3,4,5,6)P4, which inhibits a CaMKII-activated chloride flux across the plasma membrane of many different cell types (Ganapathi et al. 2013; Mitchell et al. 2008; Xie et al. 1998). In this way, Ins(3,4,5,6)P4 may regulate salt and fluid secretion from epithelial cells (Xie et al. 1998) and migration of aortic smooth muscle cells (Ganapathi et al. 2013). In hippocampal neurons, an Ins(3,4,5,6)P4-regulated chloride conductance contributes to the overall regulation of the synaptic efficacy in generating action potentials. Therefore, Ins(3,4,5,6)P4 has the potential to affect electrical excitability in neurons via regulating chloride fluxes and hence directly influencing neuronal development (Mitchell et al. 2008).
Ins(3,4,5,6)P4 not only regulates chloride-fluxes at the plasma membrane, it also inhibits flux across intracellular vesicles such as insulin granules and endosomes (Mitchell et al. 2008; Renström et al. 2002). In these vesicles, the chloride conductance may serve as a charge neutralization shunt that supports acidification of these vesicles. Intravesicular pH helps determine the fate of vesicular trafficking whether for recycling of plasma membrane proteins or exocytosis. The inhibition of chloride influx into vesicles by Ins(3,4,5,6)P4 leads to a more alkaline compartment (Mitchell et al. 2008). In the case of insulin granules in pancreatic β-cells, their alkalinization by Ins(3,4,5,6)P4 reduces insulin secretion (Renström et al. 2002). Perhaps abnormally elevated cellular Ins(3,4,5,6)P4 levels might contribute to hyperglycemia-dependent refractoriness of β-cells which typifies type-2 diabetes.
As for the mechanism by which Ins(3,4,5,6)P4 regulates transmembrane chloride currents, research into this complex subject has stalled of late. There is clear evidence that the ClC-3 protein is involved, but its precise role remains unclear (Ganapathi et al. 2013; Mitchell et al. 2008). ClC-3 exhibits an outwardly rectifying chloride current that is activated by CaMKII and exhibits time-dependent inactivation at positive potentials (Xie et al. 1998). This current is inhibited by Ins(3,4,5,6)P4 in a highly specific manner: The 3,4,5,6-phosphate configuration, and the 1-hydroxyl group, are both essential to the inhibitory activity. One possibility is that Ins(3,4,5,6)P4 directly blocks ClC-3 conductance, but that simple idea is confounded by the unexplained observation that the inhibitory activity of Ins(3,4,5,6)P4 is dependent upon protein phosphatase activity (Xie et al. 1998).
There is further evidence that the role of ClC-3 is even more complex: Of particular note are experiments in which the patch clamp technique was used to record whole-cell chloride currents in murine aortic smooth muscle cells (Ganapathi et al. 2013). The chloride current was activated by using calcium-buffers to elevate intracellular free calcium. (The calcium concentration used in such experiments has to be carefully chosen: above 1 μM, calcium attenuates the inhibition by Ins(3,4,5,6)P4 (Ho et al. 2001)). In cultured smooth muscle cells derived from wild-type mice, the activated chloride current was inhibited by either pharmacological blockade of CaMKII or Ins(3,4,5,6)P4 (Ganapathi et al. 2013). Neither the activation of chloride current by CaMKII, nor the Ins(3,4,5,6)P4 inhibition, were observed in cells derived from mice in which the ClC-3 gene had been disrupted (Ganapathi et al. 2013). This is a clear demonstration that ClC-3 mediates the effects of CaMKII and Ins(3,4,5,6)P4 upon chloride current. However, this is a chloride current that exhibits time-dependent activation at hyperpolarizing potentials (Ganapathi et al. 2013); this is kinetically different from the CaMKII-activated current that appears to be directly carried by ClC-3 (see above), and is instead the exact electrophysiological fingerprint of TMEM16A, a well-characterized calcium-activated chloride channel described initially by three separate laboratories in 2008 (Ferrera et al. 2010). One interpretation of these data is that TMEM16A is regulated by ClC-3. That is, ClC-3 brings CaMKII-activation and Ins(3,4,5,6)P4-inhibition to the TMEM16A-mediated chloride current. The possibility that ClC-3 is a channel regulator (as well as a channel) appears not to have been considered previously.
The synthesis and metabolism of Ins(3,4,5,6)P4 are solely regulated by a single multifunctional kinase and phosphotransferase, ITPK1. This enzyme dynamically couples the cell surface receptor activated PLC hydrolysis of PtdIns(4,5)P2 to the cellular levels of Ins(3,4,5,6)P4. This is a biologically significant event because Ins(3,4,5,6)P4 regulates a specific pathway for transmembrane chloride conductance that is involved in numerous cellular functions including synaptic efficacy, epithelial salt and fluid secretion, insulin release from pancreatic β-cells, and inflammatory responses.