Inositol polyphosphate–protein interactions: Implications for microbial pathogenicity

Abstract Inositol polyphosphates (IPs) and inositol pyrophosphates (PP–IPs) regulate diverse cellular processes in eukaryotic cells. IPs and PP–IPs are highly negatively charged and exert their biological effects by interacting with specific protein targets. Studies performed predominantly in mammalian cells and model yeasts have shown that IPs and PP–IPs modulate target function through allosteric regulation, by promoting intra‐ and intermolecular stabilization and, in the case of PP–IPs, by donating a phosphate from their pyrophosphate (PP) group to the target protein. Technological advances in genetics have extended studies of IP function to microbial pathogens and demonstrated that disrupting PP–IP biosynthesis and PP–IP‐protein interaction has a profound impact on pathogenicity. This review summarises the complexity of IP‐mediated regulation in eukaryotes, including microbial pathogens. It also highlights examples of poor conservation of IP–protein interaction outcome despite the presence of conserved IP‐binding domains in eukaryotic proteomes.


| INTRODUCTION
IPs and PP-IPs are produced by a series of sequentially acting IP kinases (IPKs). Using genetic and pharmacological approaches to modulate IP kinase (IPK) activity in conjunction with gel-, HPLC-and, more recently, mass spectrometry-based metabolic profiling strategies, the identification and role of IPs and PP-IPs were initially elucidated in mammalian cells and the model yeasts, Saccharomyces cerevisiae and Schizosaccharomyces pombe. IPK enzymatic function has also been confirmed in vitro using chemically synthesized substrates. These studies revealed that IPs and PP-IPs function in a diverse range of cellular processes including glucose homeostasis, insulin sensitivity and secretion, fat metabolism and cellular energy dynamics, growth factor signalling, phosphate homeostasis, vesicular trafficking, DNA damage and repair, chromatin remodelling, spermatogenesis, neuronal migration, neutrophil activity, aging, apoptosis and platelet function (Lee, Kim, Ahn, & Kim, 2020). Hence, it is not surprising that dysregulated IP and PP-IP biosynthesis in human cells is associated with numerous diseases including Huntington's disease (Ahmed et al., 2015), diabetes and obesity (Chakraborty et al., 2010) and cancer (Rao et al., 2015).
Advances in genome sequencing and genome manipulation technology have subsequently allowed investigation into the roles of IPs and PP-IPs in microbial pathogenicity. The most significant progress has been in: Cryptococcus neoformans, an AIDS-related fungal pathogen and the most common cause of fungal meningitis worldwide (Rajasingham et al., 2017); Candida albicans, the commensal and opportunistic nosocomial pathogen and most prevalent cause of fungal infections worldwide (Bongomin, Gago, Oladele, & Denning, 2017); Trypanosoma brucei and T. cruzi, the insect vectortransmitted protozoan parasites causing sleeping sickness and Chagas disease, respectively; the human immunodeficiency virus (HIV) which attacks the immune system and leads to acquired immunodeficiency syndrome (AIDS) and Clostridium difficile, a multidrug-resistant nosocomial bacterial pathogen that causes colitis. The importance of IPs and PP-PPs in these pathogens, as well as mechanistic insight into their mode of action at the molecular level, is discussed in more detail below.

| Cryptococcus neoformans
Enzymes involved in IP and PP-IP biosynthesis were identified in C.
cerevisiae are similar and less complex compared to the mammalian pathway ( Figure 1). Two IP 3 kinase homologues were identified in C.
F I G U R E 1 Diagram of the biosynthesis pathways of soluble IPs in humans, S. cerevisiae and C. neoformans, colour-coded to differentiate each pathway. Dashed line: only occurs in absence of Ipk1 activity neoformans: Arg1 and Arg2, which share 22% and 15% identity, respectively, with Arg82 and 17% identity with each other (Lev et al., 2013). Both Arg1 and Arg2 contain a conserved PDKG motif essential for the catalytic activity of IP 3 kinases. However, only Arg1 has IP 3 kinase activity in vivo (Lev et al., 2013), and the physiological relevance of Arg2 remains to be elucidated. In contrast to yeast cells, conversion of IP 3 to IP 5 in mammalian cells occurs via several routes involving four different IPKs, with inositol polyphosphate multikinase (IPMK) involved in all pathways (Figure 1). The synthetic redundancy in the early steps of the pathway, coupled with a low IPK sequence homology, highlights early steps in the IPK pathway as attractive targets for anti-fungal drug development, due to the minimised chance of off-target effects on the human host (Lev et al., 2019;Li et al., 2016b).
Like the Vip1/Asp1 homologues in S. cerevisiae/S. pombe, Asp1 of C. neoformans has the features of a bi-functional enzyme, with an Nterminal ATP grasp domain responsible for the IPK activity and a Cterminal histidine acid phosphatase domain (Lev et al., 2015). Vip1/ Asp1 homologues in model yeast control the levels of 1-PP-IP 5 and 1,5-PP 2 -IP 4 (IP 8 ) via their synthesis (IPK domain) and destruction (acid phosphatase domain) and inositol pyrophosphate levels can be skewed in either direction by mutating individual catalytic sites (Dollins et al., 2020). Both the bi-functionality of Asp1 and its ability to synthesize the 1-PP-IP 5 isoform remain to be demonstrated in C.
neoformans. Breakdown of inositol pyrophosphate can also occur via the diphosphoryl inositol polyphosphate phosphatase (DIPP) class of pyrophosphatase (Safrany et al., 1999). However, DIPPs are more promiscuous and hydrolyse nucleotide dimers and polyphosphates (Lonetti et al., 2011). DIPP homologues have been identified in C. neoformans but remain to be characterized.
Disrupting IP biosynthesis in C. neoformans leads to dramatically altered transcriptional profiles, numerous cellular defects and loss of virulence in mouse infection models (Lev et al., 2013;Lev et al., 2015;Li et al., 2016a;. The absence of IP 7 (isoform 5-PP-IP 5 ) has a greater impact on cellular function than absence of IP 6 or IP 8 (Lev et al., 2015;Li et al., 2016a). Cryptococcal 5-PP-IP 5 deficiency coincides with a 37 C growth defect, defective mitochondria, inability to utilise alternative carbon sources, compromised cell wall integrity, diminished melanisation and reduced mannoprotein exposure at the cell surface (Lev et al., 2015). The latter coincides with a failure to elicit a strong immune response in vivo and in vitro. The polysaccharide capsule, which is a major virulence factor and diagnostic marker of this pathogen, was also altered in 5-PP-IP 5 -deficient C. neoformans, being larger and more mucoid (Lev et al., 2015). Furthermore, the phosphate (PHO) signalling pathway failed to become activated in 5-PP-IP 5 -deficient C. neoformans when cellular phosphate levels declined as discussed below. Thus, 5-PP-IP 5 is required for C. neoformans to respond to host stress, undergo metabolic adaptation to the host environment and acquire phosphate. Infection with 5-PP-IP 5deficient C. neoformans is asymptomatic, and the pathogen cannot disseminate from the lungs to the brain (Lev et al., 2015).
Inositol is important for the development and pathogenicity of C.
neoformans and is a precursor for the synthesis of the Plc1 substrate, PIP 2 and hence the generation of IP 3 . Liao et al. (2018) showed that PP-IP biosynthesis fine-tunes inositol acquisition to maintain inositol homeostasis in C. neoformans. In contrast to S. cerevisiae, they found that PP-IP biosynthesis is dispensable for de novo synthesis of inositol in C. neoformans, consistent with the role of PP-IPs in inositol metabolism in C. neoformans, being distinct from that of S. cerevisiae.
The combined loss of IP 4-7 in the arg1Δ mutant results in a more exacerbated 37 C growth defect, reduced capsule size, enhanced recognition by phagocytes, thickened cell walls and enlarged vacuoles . Interestingly, the invasion-promoting enzyme, phospholipase B1 (PLB1), was excessively N-linked glycosylated, and this coincided with a blockage in PLB1 secretion. In contrast to infection with the 5-PP-IP 5 -deficient kcs1Δ mutant, infection with the IP 4-7deficient arg1Δ mutant was cleared in a mouse infection model .

| Candida albicans
Recent studies on IPK knockout strains in the opportunistic fungal pathogen, C. albicans, have begun to elucidate the role of IPs and PP-IPs in cellular function (Li, Zhang, et al., 2017;Peng, Yu, Liu, Ma, & Li, 2020;Zhu et al., 2020). However, as no metabolic profiling was performed, the roles of specific IPs and PP-IPs are only putative.
Homozygous deletion of the putative IP 5 kinase-encoding gene, IPK1, to create ipk1ΔΔ, resulted in dysfunctional mitochondria, which coincided with down-regulation of genes involved in mitochondrial function, particularly those associated with oxidative phosphorylation (Zhu et al., 2020). The ipk1ΔΔ mutant also had a fitness defect when grown on standard laboratory media and was hypersensitive to anti-fungal drugs, oxidising agents, cell wall perturbing agents and macrophageinduced killing and was attenuated for virulence in a mouse dissemination model (Zhu et al., 2020). The results implicate the importance of IP 6-8 for cellular functions required for pathogenicity.
The same group also evaluated the role of Kcs1, Vip1 and Ipk2 in C. albicans by creating the corresponding deletion mutants (Li, Zhang, et al., 2017;Peng et al., 2020). They found that Vip1 plays a more important role than Kcs1 in regulating energy metabolism but without damaging mitochondria. Specifically, they found that growth of the homozygous deletion mutant vip1ΔΔ, but not kcs1ΔΔ, was reduced in glucose-containing medium, and that this coincided with an upregulation in glycolysis and down-regulation in mitochondrial function.
The glycolysis-skewed metabolism was compensated for by an accumulation of lipid droplets . Only the vip1ΔΔ mutant accumulated cell wall chitin and exhibited plasma membrane leakage which eventually led to death. Neither the vip1ΔΔ nor the kcs1ΔΔ mutants were tested for virulence in animal models. The results suggest that, in contrast to C. neoformans (Lev et al., 2015) and S.
A homozygous IPK2 deletion mutant could not be created in C.
albicans, presumably because complete loss of IPK2 function is lethal (Li, Zhang, et al., 2017). Instead, a conditional knock-down approach was used to limit IP 4-8 biosynthesis. Similar to C. neoformans, IP deficiency impacted numerous cellular functions and coincided with altered gene expression and secretion (Li, Zhang, et al., 2017). Interestingly, virulence traits such as the secretion of hydrolytic enzymes involved in nutrient acquisition, invasion of host tissues and hyphal development were enhanced (Li, Zhang, et al., 2017). Enhanced hyphal development coincided with an increase in the expression of hypha-specific genes and transport of hypha-specific factors. Changes in Ca 2+ homeostasis were also observed in a IP 4 -8 -deficient conditional knock down strain, which would have elevated IP 3 . This is consistent with the hypothesis that IP 3 influences the activity of calcium channels in the vacuole as discussed below. Given that animal studies were unable to be conducted with the conditional knock down mutant, phenotypes observed in vitro were unable to be correlated with virulence in animal models. The results suggest that loss of IP 4-8 has a more dramatic impact on cellular function in C. albicans than loss of IP 6-8 .

| Trypanosoma species
The parasite T. brucei is transmitted between vertebrates, including humans, by the tsetse fly and causes African sleeping sickness. Similar to C. albicans, a conditional knock-down approach was used in T.
brucei to study IP function, and the results demonstrated that almost every IP conversion step is essential for parasite growth and infectivity (Cestari, Haas, Moretti, Schenkman, & Stuart, 2016). For example, IP 3 -mediated calcium homeostasis is essential for growth and infectivity of T. brucei. (Huang, Bartlett, Thomas, Moreno, & Docampo, 2013).
In a recent study, Mantilla, Amaral, et al. (2021) dissected the contribution of IPs and PP-IPs to the life cycle stages of T. cruzi (epimastigotes, cell-derived trypomastigotes and amastigotes). Their combined use of reverse genetics and liquid chromatography mass spectrometry revealed the presence of IP 6 , IP 7 and IP 8 . These species were not detected previously by HPLC analyses of cell lysates containing products of exogenously administered, radio-labelled inositol, and suggest that IP 6-8 is derived from an endogenous source of inositol. The kinases involved in IP synthesis, TcIPMK, TcIP5K and TcIP6K, were also identified. In contrast to T. brucei, the TcIPMK knockout strain was viable; hence, the TcIPMK gene is dispensable in T. cruzi epimastigotes. However, TcIPMK was critical for virulence of the infective stages. The detection of highly phosphorylated IPs in TcIPMK knockout cells suggests that endogenous inositol is utilized for their synthesis. In contrast to T. brucei, TcIP5K was essential for survival of T. cruzi epimastigotes, consistent with the critical importance of IP 6-8 . In another recent study, Mantilla, Amaral, et al. (2021) revealed 5-IP 7 -regulated processes in the two proliferative stages of T. cruzi, which is discussed in the IP mechanism section below on pyrophosphorylation.
IPs also contribute to T. cruzi pathogenicity through the biosynthesis of glycosylphosphatidylinositol (GPI) membrane anchors, which attach variant surface glycoproteins (VSG). The periodic switching of VSG in Trypanosoma species helps the parasite evade clearance by the host immune system (Cestari & Stuart, 2015). Another parasite genus of medical importance, Leishmania, also utilises GPI anchors to tether surface molecules to the plasma membrane (Forestier, Gao, & Boons, 2014). GPI anchors are synthesized from glucose, with glucose 6-phosphate being converted to inositol-3-phosphate by the inositol-3-phosphate synthase, Ino1 (Figure 1). Inositol-3-phosphate is dephosphorylated by inositol monophosphatase, generating myo-inositol. Inositol is utilised by phosphatidylinositol (PI) synthase to produce PI, which is preferentially used for the synthesis of GPI anchors (Martin & Smith, 2006).

| Human Immunodeficiency Virus
HIV is a retrovirus with only a small genome and does not encode IP biosynthetic machinery. However, the HIV virus uses IPs synthesized by the host cell to promote viral replication. Recent studies have demonstrated a critical role for host-derived IP 6 as an inter-molecular stabilising agent involved in both the maturation and replication of the HIV virion. The role of IPs as intermolecular glue is discussed in more detail below. It has also been shown that IP 5 can substitute for IP 6 as the intermolecular stabilising agent when conversion of IP 5 to IP 6 is blocked in infected cells. Specifically, host-derived IP 6 plays a role in capsid assembly. IP 6 binding increases HIV-1 capsid stability from minutes to hours and promotes DNA accumulation inside intact structures during reverse transcription Marquez et al., 2018). Conversely, HIV within IP 6 -deficient cells produces unstable capsids and fewer virions, while virions that fail to bind sufficient IP 6 are poorly infectious and fail to replicate in primary cells (Mallery et al., 2019).

| Pyrophosphorylation-dependent effects of PP-IPs on protein target function
Inositol pyrophosphates (PP-IPs) are high-energy phosphate metabolites that can donate the terminal phosphate of their pyrophosphate moiety to a pre-phosphorylated serine or threonine on a target protein in a β-phosphoryl transfer reaction (Bhandari et al., 2007).
Pyrophosphorylation is an unusual modification in that it is enzymeand ATP-independent. It also requires Mg 2+ as a cofactor and 'priming' by a protein kinase, commonly casein kinase 2 CK2. 5-PP-IP 5 affinity chromatography and mass spectrometry have been used to identify the IP 7 interactome in S. cerevisiae and more recently in T.  (Chanduri et al., 2016). Pyrophosphorylation is also prevalent in the nucleolus. In yeast, pyrophosphorylated proteins include the ribosomal chaperone, SRP40 and NSR1, which is involved in ribosome assembly and export (Bhandari et al., 2007;Saiardi, Bhandari, Resnick, Snowman, & Snyder, 2004). Despite the growing number of pyrophosphorylated proteins being discovered, the effect of pyrophosphorylation on many of these target proteins remains to be elucidated.

| IP competition with phosphoinositides for binding to PH domains
Despite their low abundance, phosphoinositides play key roles in regulating cellular function by tethering specific types of cellular proteins to the membrane to control their compartmentalisation and often their activity. One example is the pleckstrin homology (PH) domaincontaining proteins, which are recruited to and regulated by phosphoinositides, including PIP 2 and PIP 3 (Hammond & Balla, 2015).
Phosphoinositide-PH domain protein interactions regulate diverse functions including signalling, cytoskeletal organisation, vesicular trafficking, phospholipid processing and glucose homeostasis (Lenoir, Kufareva, Abagyan, & Overduin, 2015  . This competition would prevent Akt phosphorylation by membrane kinases. IP-based analogues are currently being pursued as inhibitors, particularly in the case of phosphoinositide 3-kinase (PI3K)-mediated signalling pathways, which have a well-established role in cancer development and progression (Maffucci & Falasca, 2020 Ferguson, 2000). Alternatively, excess IP 3 could inhibit the activity of its membraneassociated progenitor, Plc1, via a feedback inhibition loop, since the major substrate of Plc1 in C. neoformans is the phosphoinositide, PIP 2 (Lev et al., 2013). Support for the latter is that both the arg1Δ and plc1Δ mutants share all the phenotypes that are absent in IP 7 -deficient strains and accumulate PIP 2 (Lev et al., 2013).

| IP competition in mRNA decapping
The 5-PP-IP 5 isomer of IP 7 has recently been demonstrated to regulate mRNA stability and the dynamics of P-bodies, which are purportedly the sites for sequestration and storage of mRNAs away from the translating pool and where mRNA decay occurs . By modulating 5-PP-IP 5 levels genetically and pharmacologically, this study demonstrated that 5-PP-IP 5 competes with 5 0 -capped mRNA for hydrolysis by NUDT3, a DIPP1 which dephosphorylates all PP-IPs, including 5-PP-IP 5 , and thereby impacts cellular mRNA transcript levels. The study also reported that P-body abundance changed in accordance with the 5-PP-IP 5 -modulated levels of NUDT3-regulated mRNA transcripts.

| IP roles in calcium homeostasis
A C. albicans conditional mutant, predicted to have reduced levels of IP 4-8 and an excess of IP 3 , has an elevated level of Ca 2+ (Li, Zhang, et al., 2017). It is known that IP 3 spikes in higher eukaryotes in response to external stimuli and that IP 3 binds to IP 3 -gated calcium channels in the endoplasmic reticulum (ER) to trigger a transient influx of Ca 2+ into the cytosol (Foskett, White, Cheung, & Mak, 2007). However, the source of the elevated intracellular Ca 2+ in C. albicans is not known.
Several studies demonstrated an IP 3 -dependent increase in cytosolic calcium in S. cerevisiae. However, IP 3 -gated ER calcium channel orthologues have not been identified in fungal genomes (Alzayady et al., 2015;Tisi et al., 2004) suggesting that the source of the intracellular Ca 2+ is not the ER. In yeast, the vacuole, rather than the ER, is the most important calcium storage compartment, with Ca 2+ homeostasis regulated by a Ca 2 + -ATPase (Pmc1), a Ca 2+ /H + exchanger (antiport) (Vcx1) and a calcium channel homologue of the transient receptor potential channels (Yvc1) (Palmer et al., 2001). Yvc1 mediates Ca 2+ efflux from the vacuole to the cytoplasm under conditions of stress (Denis & Cyert, 2002;Zhou et al., 2003). Evidence obtained in S.
cerevisiae using an IPK2/YVC1 double mutant suggests that IP 3 could interact directly, or indirectly, with Yvc1 to control its opening and trigger Ca 2+ signalling in the cytosol (Bouillet et al., 2012). Whether IP 3 has the same function in fungal pathogens remains to be elucidated.

| IP roles as intermolecular stabilisers in HIV pathogenicity
In addition to regulating protein activity, IPs function as intermolecular stabilisers within multiprotein complexes. One example is during HIV pathogenesis where the replicating virus utilises IP 6 from the host cell to stabilise its capsid and promote the assembly and maturation of infectious virions (Dick, Mallery, Vogt, & James, 2018;Mallery et al., 2018;Mallery et al., 2019;Marquez et al., 2018). IP 6 specifically interacts with two lysine residues (K158 and K227) in the immature Gag hexamer and assists in driving the formation of the immature lattice in HIV. This once again demonstrates the importance of electrostatic interactions between IPs and positively charged residues in target proteins. The highly conserved immature lattice lysine rings, K158 and K227, and mature capsid charged ring (e.g., R18) across diverse lentiviruses suggest that IP 6 is essential for lentiviral replication in general. Furthermore, Azevedo, Burton, Ruiz-Mateos, Marsh, and Saiardi (2009) showed that IP 7 (5-PP-IP 5 )-mediated pyrophosphorylation of AP3B1, a clathrin-associated protein complex required for HIV-1 Gag release from HeLa cells, modulates AP3B1 interaction with a motor protein of the kinesin superfamily, Kif3A, which is also required for HIV-1 Gag release, and consequently affects release of HIV-1 virus-like particles.

| IP roles as allosteric regulators and intermolecular stabilisers in transcriptional and cell cycle regulation
Crystal structure analysis revealed that IP 6 binds to the catalytic domain of human ADAR2, an RNA editing enzyme (Macbeth et al., 2005). IP 6 is buried within the enzyme core and contributes to the protein fold and is required for enzyme activity. IP 6 was also found to be essential for deamination of adenosine 37 of tRNA ala by ADAT1 in vivo and in vitro (Macbeth et al., 2005). and they mediate ubiquitylation of numerous proteins (Lin et al., 2020;Scherer et al., 2016). Crystal structure analysis revealed that IP 6 binding to a cognate pocket formed by conserved lysine resi-  et al., 2007); 5-PP-IP 5 is the dominant activator of polyphosphate synthesis by the vacuolar transporter chaperone in S. cerevisiae (Gerasimaite et al., 2017;Wild et al., 2016), and 5-PP-IP 5 stimulates Na + /phosphate cotransport by Pho91 in T. brucei (Potapenko et al., 2018). Both 5-PP-IP 5 and IP 8 have been implicated in regulating XPR1-driven phosphate efflux in human cells Wilson, Jessen, & Saiardi, 2019). However, using a number of different strategies including liposome-mediated delivery of metabolically resistant phosphate-carbon-phosphate (PCP) analogues of PP-IPs into cells,  found that the hierarchy of importance favours IP 8 over 5-PP-IP 5 and 1-PP-IP 5 in human cells. A recent study has shed light on the regulatory role of PP-IPs in phosphate homeostasis in the fungal pathogen C. neoformans. C. neoformans activates its phosphate (PHO) signalling pathway via the transcription factor Pho4, which is essential for its pathogenicity . This is consistent with the pathogen experiencing phosphate deprivation during infection. In S. cerevisiae and C. neoformans, PHO pathway activation depends on Pho81, a cyclin-dependent protein kinase (CDK) inhibitor containing an SPX domain (Desmarini et al., 2020;Toh-e et al., 2015). Pho81 forms a trimeric complex with the CDK, Pho85, and its associated cyclin, Pho80, which directs Pho85 to phosphorylate its substrate, Pho4. Pho81-Pho80-Pho85 forms a complex in S. cerevisiae and C. neoformans irrespective of phosphate status (Desmarini et al., 2020;Schneider, Smith, & O'Shea, 1994). During phosphate deprivation, Pho81 inhibits Pho85 and prevents it from phosphorylating Pho4. Unphosphorylated Pho4 is retained in the nucleus where it promotes transcription of genes involved in phosphate acquisition . In S. cerevisiae the IP 7 isoform, 1-PP-IP 5 , allosterically regulates the SPX domain of Pho81 to trigger PHO pathway activation (Lee et al., 2007). In contrast, the 5-PP-IP 5 isoform regulates PHO pathway activation in C.
neoformans (Desmarini et al., 2020). In this study, a conserved lysine surface cluster, K 221,224,228 , was identified in the SPX domain of Pho81 and demonstrated to be important for binding 5-PP-IP 5 (Desmarini et al., 2020). Similar to PHO4 deletion, disrupting 5-PP-IP 5 interaction with the SPX domain of Pho81 prevented PHO pathway activation in phosphate-starved C. neoformans but led to avirulence, rather than attenuated virulence, in a mouse infection model (Desmarini et al., 2020). The reduction in virulence is consistent with the IP 7 -regulated CDK complex having functions that extend beyond the regulation of phosphate homeostasis.
It was proposed that 1-PP-IP 5 has a role in phosphate sensing in S. cerevisiae, as its levels increase during phosphate deprivation (Lee et al., 2007). In contrast, the levels of 5-PP-IP 5 decrease in C. neoformans during phosphate deprivation (Desmarini et al., 2020).
Despite this reduction, the remaining 5-PP-IP 5 functions as intermolecular 'glue' to stabilise the association of Pho81 with Pho85 and its cyclin Pho80. Furthermore, 5-PP-IP 5 binding-defective Pho81 and native Pho81 are degraded during phosphate starvation and IP 7 deficiency, respectively (Desmarini et al., 2020). This suggests that Pho81 stability is dependent upon its association with Pho80-Pho85. Further evidence of the differing roles of IP 7 isoforms in S. cerevisiae and C.
cerevisiae, the hierarchy of PP-IP importance in phosphate homeostasis favours 5-PP-IP 5 in C. neoformans. Taken together, the literature reveals that even in relatively closely related species, IP 7 -target protein duos are not conserved and have evolved to produce a different regulatory outcome for the same function.
It is tempting to speculate that the multiple phosphorylation sites displayed by IP 7 to components of a CDK complex provide a biological alternative to multisite phosphorylation by various kinases. The latter has been proposed for the interaction of the CDK inhibitor, Cip1 (the putative homologue of mammalian p21), with Cdk1 and the cyclin Cln2 to promote cell cycle progression through G1 in S. cerevisiae (Chang et al., 2017). Cip1 becomes phosphorylated at three positions by the kinase Hog1 under hyperosmotic stress, and this phosphorylation is hypothesised to strengthen Cip1 and Cdk1-G1 cyclin interaction and induce transitory cell cycle arrest. Although IP 7 has a proven role in stabilising the phosphate responsive CDK complex in C. neoformans, it cannot be ruled out that additional kinase-induced phosphorylation of Pho81 contributes to CDK complex stabilisation. For example, in S. cerevisiae, Pho81 is a substrate of its own binding partner, Pho85.

| CONCLUSIONS
Similar to mammalian cells and model yeast, dysregulated IP biosynthesis in pathogenic fungi and parasites leads to numerous cellular defects and major changes in the transcriptional profile. Studies in mammalian cells and yeast have provided much needed mechanistic insight into the biological roles of IPs and the outcomes of IP-protein interactions. A major theme that has emerged is that IPs and PP-IPs are key components of multisubunit complexes where they function as intermolecular glue to stabilise the complex and/or modulate complex activity. Via these associations, they are involved in the regulation of a diverse range of cellular functions. These studies have paved the way for elucidation of the roles IPs play as intermolecular stabilisers in microbial pathogens and promotion of an understanding that, despite the presence of conserved IP-binding domains in eukaryotic proteomes, IP-target protein duos are not conserved and have evolved to produce a different regulatory outcome, even in closely related species. Recent advances in methods used to concentrate and detect IP target proteins using stable, synthetic and conjugated IP analogues and to detect pyrophosphorylation targets using sophisticated mass spectrometry, will expedite the discovery of more novel IP targets and IP-target interaction in cellular function.

ACKNOWLEDGEMENT
We thank Stephen Schibeci for proof-reading this review. We also acknowledge the work of others, which could not be included due to space restrictions.