Phosphatase, pseudo-phosphatase, or both? Understanding PRL oncogenicity

Summary Phosphatases of regenerating liver (PRL1–3) are among the most oncogenic protein phosphatases but their mechanism of action is poorly understood. Multiple substrates have been proposed as well as a non-catalytic function regulating magnesium transport. Our recent identification of a catalytically inactive PRL mutant that retains oncogenicity in a mouse model promises to resolve the question of whether PRLs act as phosphatases or pseudo-phosphatases in different cancer models.

retains the charge but is unable to catalyse the dephosphorylation reaction. When tested in cells, the PRL3 C104D mutant was able to repress CNNM magnesium efflux as effectively as wild-type PRL3. In a tail-vein assay of tumour metastasis, B16 cells transformed with the C104D mutant formed as many lung nodules as the wild type. 10 A complementary mutation, R138E, that blocks CNNM binding but not catalytic activity did not promote metastasis. These results strongly argue for the characterisation of PRLs primarily as pseudo-phosphatases.
This importance of CNNM binding has been largely ignored even though there is the widespread agreement that PRLs are poor enzymes. The extremely slow kinetics of PRLs is a direct consequence of the long lifetime of the phosphocysteine intermediate. In the absence of a mechanism to speed up hydrolysis of phosphocysteine, PRLs act in an almost stoichiometric fashion with one phosphatase molecule dephosphorylating one substrate. Studies in cells suggest that the lifetime of the phosphocysteine intermediate is several hours, which is consistent with the absence of a mechanism to promote dephosphorylation and increase activity. 5,10 Paradoxically, although the substrates are unknown, there is good evidence for a small amount of PRL activity in cells and tissues. Western blotting of endogenous PRLs shows that the majority of PRLs are phosphorylated on cysteine, which can only occur as a result of catalysis. 5 Indirect evidence of the importance of this phosphatase activity comes from the conservation of PRL catalytic site. Typically, pseudo-enzymes lose their catalytic activity through mutations in the active site, yet PRL phosphatase activity is conserved from tardigrades to humans. 10 When phosphorylated, PRLs are unable to bind CNNMs, which suggests that phosphorylation of the catalytic cysteine plays a role in controlling CNNM activity. The phosphorylation of PRLs changes in response to changes in magnesium in the cellular growth medium. 5 Magnesium deprivation leads to a slow decrease in phosphorylation while readdition leads to rapid rephosphorylation. 10 This naturally leads to a model for Mg 2+ homoeostasis in which PRL phosphorylation detaches PRLs from CNNMs to stimulate Mg 2+ efflux. Thus PRLs www.nature.com/bjc behave both as pseudo-phosphatases and phosphatases. They regulate CNNM proteins via direct binding but are themselves regulated through their catalytic activity.
Cysteine oxidation probably provides an additional level of regulation of the PRL-CNNM interaction. Little is known about the oxidation state in cells, but the purified PRL proteins are partially oxidised even in mildly reducing conditions. 6 The disulphide forms of PRLs show approximately 1000-fold weaker binding to CNNM proteins and, of course, are catalytically inactive. 5 An increase in reactive oxygen species would be expected to decrease CNNM binding and intracellular magnesium levels, possibly as a homoeostatic feedback loop to regulate mitochondrial metabolism.
Going forward, many questions remain to be answered. PRL expression levels are greatly increased in metastatic cancer and regenerating liver cells, but nothing is known about the signals that control PRL phosphorylation and oxidation. In cultured cells, PRL phosphorylation changes dramatically in response to magnesium availability but the mechanism and substrates involved are unknown. Finally, much more needs to be done to understand how changes in magnesium efflux lead to downstream effects and cancer metastasis. One point that is clear is that the PRL mutations that selectively impair phosphatase or pseudo-phosphatase activities will be valuable tools for future studies.

AUTHOR CONTRIBUTIONS
K.G. and H.M. contributed to the writing, revision, and approval of this manuscript.

ADDITIONAL INFORMATION
Ethics approval and consent to participate Not applicable.
Data availability Not applicable.

Competing interests The authors declare no competing interests.
Funding information This work was supported by the Natural Sciences and Engineering Research Council of Canada.
Note This work is published under the standard license to publish agreement. After 12 months the work will become freely available and the license terms will switch to a Creative Commons Attribution 4.0 International (CC BY 4.0).
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.