RESEARCH ARTICLE The toxic effects of yeast Ppz1 phosphatase are counteracted by subcellular relocalization mediated by its regulatory subunit Hal3 zquez, Antonio Casamayor and Joaquın Arin Marcel Albacar, Diego Vela ~o noma de Barcelona, Cerdanyola Institut de Biotecnologia i Biomedicina & Departament de Bioquımica i Biologia Molecular, Universitat Auto s, Spain del Valle Correspondence Overexpression of Saccharomyces cerevisiae protein phosphatase Ppz1 ~o, Institut de Biotecnologia i J. Arin strongly impairs cell growth. Ppz1 is negatively regulated by its subunit Biomedicina & Departament de Bioquımica i Hal3, and Hal3 overexpression fully counteracts the toxic effects derived Biologia Molecular, Universitat Auto noma de s, from high levels of the phosphatase. We show that Ppz1 localizes at the Barcelona, 08193, Cerdanyola del Valle Spain plasma membrane, and that co-expression of Hal3 recruits Ppz1 to internal Tel: +34 935811315 membranes (mostly vacuolar). This effect is not observed in a catalytically E-mail: joaquin.arino@uab.es impaired mutant of Ppz1. Disruption of intracellular trafficking by deletion of the ESCRT-0 component VPS27 abolishes both Hal3-mediated relocaliza- (Received 29 January 2022, revised 25 tion of Ppz1 and normalization of cell growth, without affecting Ppz1 protein February 2022, accepted 1 March 2022) levels. We propose that Hal3 counteracts the toxic effects caused by excess of Ppz1 not only by inhibiting its enzymatic activity but also by recruiting doi:10.1002/1873-3468.14330 the phosphatase to internal structures. Edited by Ivan Sadowski Keywords: intracellular trafficking; overexpression toxicity; protein phosphatase; yeast The Ppz protein phosphatases belong to the PP1 fam- to explain the sensitivity of this strain to cell wall ily of Ser/Thr phosphatases and are represented in stress situations [4,5,9,10]. Ppz1 appears also involved S. cerevisiae by two paralogs, Ppz1 and Ppz2 [1,2]. in translation initiation [11], in the regulation of ubiq- These proteins are characterized by a conserved cat- uitin turnover [12], and the dephosphorylation of the alytic C-terminal region (~ 340 residues, about 60% ubiquitin ligase adaptor Art1 [13]. identical to the bona fide S. cerevisiae PP1 catalytic Early reports demonstrated that even moderately subunit Glc7) that is preceded by a much more diver- higher than normal levels of Ppz1 cause a dramatic gent ~ 350 residues long N-terminal region [3,4]. Ppz1 cell growth arrest [9,14]. In fact, a more recent is the most functionally relevant enzyme, and it is genome-wide study showed that Ppz1 could be the involved in the maintenance of monovalent cation most toxic protein in S. cerevisiae [15]. Current evi- homeostasis. This occurs by regulating two different dence indicates that such toxicity derives from alter- processes, the repression of the expression of the Ena1 ation of multiple targets affecting intracellular Na+, K+-ATPase, crucial in the response to salt stress signalling [16,17], and the abnormal activation of the [5,6], and the inhibition of potassium uptake through Nha1 Na+,K+/H+ antiporter [18]. Remarkably, while the K+ high-affinity Trk1 and Trk2 transporters [7,8]. the phosphatase activity of Ppz1 is required for toxic- The increased K+ uptake observed in ppz1 cells, which ity [16], the N-terminal half of Ppz1 also contributes would lead to increased turgor pressure, was proposed to this effect [19]. Abbreviations EGFP, enhanced green fluorescent protein; ESCRT, endosomal sorting complex required for transport; PP1, protein phosphatase 1; Ppz1, serine/threonine protein phosphatase Z1. FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies 1 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Subcellular relocalization of Ppz1 phosphatase M. Albacar et al. Ppz1 is controlled by two regulatory subunits, Hal3 Growth tests in drops were done as in [16]. For growth and Vhs3 (Hal3 being the major regulator in vivo), assays in liquid medium, cultures were grown overnight to which play a regulatory role by interacting with the saturation. Dilutions were made to the indicated OD600 in catalytic C-terminal domain of the phosphatase and fresh medium and distributed by triplicate in honeycomb inhibiting its activity [9,20–23]. Notably, in S. cere- plates (Thermo Fisher, Waltham, MA, USA). A Bioscreen visiae and related fungi, Hal3 and Vhs3 are moonlight- C apparatus (Thermo Fisher) was used to monitor cell ing proteins that also play a role in the biosynthesis of growth at 28 °C by reading the OD600 every 30 min for 2 coenzyme A. This occurs by interaction with Cab3, a to 3 days. Plates were shaken at the maximum amplitude structurally related protein, to form an atypical hetero- setting for 7 min before each measure. trimeric phosphopantothenoylcysteine decarboxylase (PPCDC) enzyme [24]. It has been proposed that Hal3 DNA techniques, strains and plasmid has a higher propensity than Vhs3 to be released as a constructions monomer, thus being more able to interact with Ppz1 [22]. Escherichia coli DH5a cells were used as plasmid DNA host and were grown in LB medium supplemented with 50- In concordance with the inhibitory role of Hal3 on lgmL 1 ampicillin (when required), at 37 °C. Transforma- Ppz1 function, episomal expression of Hal3 eliminates tions of E. coli and standard recombinant DNA techniques the growth defect observed in cells expressing PPZ1 were performed as described [28]. S. cerevisiae cells were both from its own promoter in a multicopy plasmid [9] transformed by the lithium acetate method [29]. and from the powerful GAL1 promoter [16]. In spite Strain MAC003 (GAL1:PPZ1-EGFP) was constructed of some discrepancies among databases constructed by transforming strain ZCZ01 with the cassette GFP- from large-scale screening experiments (ref. [25] and HIS3MX6, following the methodology described in [30], https://yplp.yeastgenome.org/), it is generally accepted with the pYM28 vector and oligonucleotides S3 Ppz1 Ct that Ppz1 can be detected mainly at the cell periphery, Fluo Fw and S2 Ppz1 Ct Fluo Rev. The strain DBY746 likely attached to the plasma membrane because of its vps27 was made by transformation of wild-type DBY746 conserved myristoylated Gly2 [7,13,14,19,26]. With the cells with a 1.6 kbp vps27::kanMX4 fragment, amplified aim to better understand the mechanism by which with oligonucleotides VPS27-F and VPS27-R, from the Hal3 could neutralize Ppz1 toxicity, we overexpressed BY4741 deletant (EUROSCARF collection). The correct fluorescently tagged versions of Ppz1 and Hal3. To insertion of cassettes was assessed by colony PCR. Strains our surprise, we found that Hal3 was able to recruit used in this work are described in Table 1. Ppz1 to internal membranes (mostly vacuolar) and Generation of C-terminally GFP-tagged versions of Ppz1 that disruption of this intracellular trafficking abol- and Ppz1G2A expressed from plasmid YEplac195 (2-micron, ished Hal3-mediated normalization of cell growth. URA3) was described in [19]. The equivalent construct These results suggest that, in addition to the inhibition expressing the catalytically impaired R451L Ppz1 version of its enzymatic activity, removal of Ppz1 from the was made by replacing the PacI and Kpn2I fragment of plasma membrane contributes to counteract the toxic YEp195 Ppz1-GFP with the equivalent one from the effects caused by an excess of the phosphatase. pCM189 Ppz1-R451L construct (laboratory stock). The YEp195 Ppz1-mCherry vector was obtained by PCR ampli- fication of the PPZ1-mCherry fragment from strain Materials and methods Table 1. Yeast strains used in this work. Yeast strain growth and media Origin / Saccharomyces cerevisiae cells were grown at 28 °C in YPD Strain Name Genotype Reference (1% yeast extract, 2% peptone and 2% dextrose) or syn- DBY746 MATα ura3-52 leu2-3,112 D. Botstein thetic medium with the appropriate drop-out mix of amino his3-1 trp1-239 acids for plasmid selection when needed. The synthetic med- DBY746 vps27 DBY746 vps27::kanMX4 This work ium is composed of 0.17% yeast nitrogen base (YNB) with- DVS004 DBY746 HAL3-GFP This work out ammonium sulfate or amino acids, 0.5% ammonium DVS007 DBY746 PPZ1-mCherry This work sulfate, 2% glucose and 0.13% drop-out mix [27]. To culture DVS008 DBY746 HAL3-mCherry This work yeast cells carrying constructs under a GAL1-10 promoter, BY4741 MATa his3D1 leu2D0 met15D0 [43] glucose was replaced by 2% raffinose. For growing yeast ura3D0 ZCZ01 BY4741 pGAL1-10:PPZ1 [16] cells to be used in assays involving fluorescence detection, MAC003 ZCZ01 pGAL1-10:PPZ1-EGFP: This work media were prepared sterilizing the corresponding carbon HIS3MX6 source separately in order to reduce background signal. 2 FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies M. Albacar et al. Subcellular relocalization of Ppz1 phosphatase DVS007 using the oligonucleotides PPZ1_Fw_-470_KpnI (5 min at 1000 9 g), supernatants were discarded and pellets and EGFP_mCherry_Rv_HindIII. This fragment was sub- were resuspended in 100 lL of Lysis Buffer (50 mM Tris-HCl cloned in YEp195 by restriction (KpnI and HindIII) and pH 7.5, 150-mM NaCl, 10% Glycerol, 0.1% Triton X-100) sup- subsequent ligation. Strain DVS007, which carries a C- plemented with 1-mM fresh DTT and EDTA-free Protease terminally mCherry-tagged version of PPZ1, was made by Inhibitor Cocktail (Roche). Cell lysis was performed by adding transformation of the wild-type strain DBY746 with a 125 lL of Zirconia 0.5-mm beads (BioSpec, Bartlesville, OK, PPZ1-mCherry-kanMX4 cassette obtained by PCR amplifi- USA) and vigorously shaking the samples in a FastPrep (MP cation from the pFA6a-mCherry-kanMX4 plasmid using Biomedicals) at 5.5 ms 1 for 45 s (3 cycles, cooling the tubes S3_Ppz1_Ct_Fluo_Fw and S2_Ppz1_Ct_Fluo_Rev primers. on ice for 2 min between repeats). Twenty-five microlitres of Plasmid YEplac181-Hal3 was reported in [6]. YEp181 Lysis Buffer were added to the samples, and these were cen- Hal3-GFP and YEp181 Hal3-mCherry were prepared trifuged for 10 min at 650 9 g at 4 °C. After this, supernatants by PCR amplification from strain DVS004 (HAL3-GFP) were recovered, and the total protein was quantified by the or DVS008 (HAL3-mCherry) genomic DNA, respec- Bradford method (Sigma-Aldrich). tively, using oligonucleotides HAL3_Fw_-500_EcoRI and Protein extracts containing 40 µg of protein were mixed EGFP_mCherry_Rv_HindIII. The fragment was subcloned with 49 SDS/PAGE loading buffer (200 mM Tris–HCl (pH in YEp181 by restriction with EcoRI and HindIII. Strain 6.8), 8% (w/v) sodium dodecyl sulfate (SDS), 0.4% (w/v) DVS004 and DVS008 were made by the transformation of bromophenol blue, 40% (v/v) glycerol, 20% (v/v) b- strain DBY746 with a cassette amplified from plasmid pYM28 mercaptoethanol), so that the final concentration was 19, [30] or pFA6a-mCherry-kanMX6 [31], respectively, and primers heated at 95 °C for 5 min and resolved by SDS/PAGE. S3_Hal3_Ct_Fluo_Fw and S2_Hal3_Ct_Fluo_Rev. Oligonu- Proteins were transferred to a polyvinylidene difluoride cleotides used in this work are described in Table S1. (PVDF) membrane (Immobilon-P, Millipore, Burlington, Plasmid pRS315 mCherry-HDEL, a centromeric vector MA, USA) by semi-dry transfer using a TE77XP apparatus (LEU2 marker) containing the ER localization signal (Hoefer, Holliston, MA, USA). Ppz1 was detected using an tagged with mCherry, was a gift of Dr. Oriol Gallego anti-GST-Ppz1 antiserum [3], at a 1 : 250 dilution in TBST (Pompeu Fabra University, Barcelona). (Tris-buffered saline plus 0.1% Tween 20) supplemented with 5% fat-free powdered milk. Hal3 was detected using a polyclonal anti-Hal3 antibody [33] at a 1 : 500 dilution as Microscopy techniques above. In both cases, a 1 : 20 000 dilution of secondary Fluorescence microscopy observation of living cells was per- anti-rabbit IgG-horseradish peroxidase antibodies (GE formed as follows. Cells were grown overnight to saturation Healthcare, Chicago, IL, USA) was employed. For detec- in synthetic media, prepared as described in the Yeast tion of the GFP tag, a primary anti-GFP monoclonal anti- Strains and Media section. At this point, cells were diluted body (Invitrogen, #9F9.F9, 1 : 1000 dilution) followed by to OD600 = 0.2 in fresh media and growth was resumed at secondary anti-mouse IgG-horseradish peroxidase antibody 28 °C until OD600 = 0.6 was reached. One-mL dilutions at [(GE Healthcare), 1 : 10000 dilution] was used. OD600 = 0.12 were then prepared, and 300 µL loaded to a well of a µ-Slide 8 Well (Ibidi), previously coated with 10 µL Concanavalin A (1 mgmL 1) (Merck, Darmstadt, Ger- Results many). The µ-Slide 8 Well was incubated for 30 min at 28 °C to let cells settle to the bottom of the wells. Expression of Hal3 from a multicopy plasmid For the MAC003 strain, the medium contained 2% raffi- alters the subcellular localization of the native nose instead of glucose. When cells reached OD600 = 0.6, version of Ppz1 galactose was added to the culture at a final concentration of 2%, to induce Ppz1 overexpression. As documented in the Introduction section, previous Vacuolar staining with FM4-64 was done in exponentially reports described that Ppz1 was mostly associated with growing cells (OD600 = 0.6) essentially as in [32] except that the peripheral membrane. The ability to associate with synthetic medium was used instead of YPD. The incubation membranes was attributed to the myristoylable Gly at time with FM4-64 was 30 min. Cells were then deposited in position 2 [14], and recent work demonstrated that a µ-Slide 8 Well and observed under the microscope with a mutation of Gly2 to Ala results in disappearance of the Texas Red filter. Unless otherwise stated, pictures were taken phosphatase from the cell’s periphery and uniform dis- using a Nikon Eclipse TE2000-E fluorescence microscope. tribution inside the cell [13,19]. We tested here whether the phosphatase function of Ppz1 was relevant for its subcellular localization by constructing a GFP-tagged Immunoblot detection of Ppz1 and Hal3 version of Ppz1-R451L, a nearly inactive version of the Yeast protein extracts were prepared as follows. Cultures were phosphatase [14] whose overexpression barely affects grown until OD600 = 0.8. Cells were collected by centrifugation cell growth [16]. FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies 3 Subcellular relocalization of Ppz1 phosphatase M. Albacar et al. As seen in Fig. 1A, in our experimental setting alter at all the diffuse distribution of the G2A variant. (Ppz1 expressed from its own promoter in a multicopy In this case, co-expression of Hal3 led to higher levels plasmid), the native version of the phosphatase mostly of Ppz1-GFP as it can be deduced from the brighter localizes at the peripheral membrane in a somewhat image, and this was subsequently confirmed by immu- punctuated pattern, with minor presence in some inter- noblot analysis (data not shown). nal structures. The inactive version shares the periph- eric localization of the native Ppz1-GFP version, Overexpressed Ppz1 and Hal3 colocalize in although its presence in internal structures appears internal, mostly vacuolar membranes more accentuated. As previously described, when Gly2 is changed to Ala, Ppz1 no longer localizes in periph- The surprising behaviour of Ppz1 in cells overexpress- eral or internal membranes, but it becomes dispersed ing Hal3 prompted us to investigate the subcellular within the cell. These results indicate that the activity localization of this regulatory subunit. To this end, a of the phosphatase is not a crucial requirement for its GFP-tagged version of Hal3 was constructed and its subcellular localization and confirms the relevance of localization monitored in control cells or cells overex- Gly2 as a key structural requirement for Ppz1 intracel- pressing Ppz1. As shown in Fig. 2, in the absence of lular distribution. We then tested whether co- Ppz1 overexpression, Hal3 was evenly distributed in overexpression of Hal3 had any effect on Ppz1 intra- the cytosol. However, when Ppz1 was overexpressed, cellular localization. Remarkably, as shown in Fig. 1B, although part of the Hal3 signal remained diffusely the presence of Hal3 resulted in substantial removal of distributed in the cytosol, a significant signal was also Ppz1 from the cell periphery and in its accumulation detected in structures whose shape was reminiscent to in conspicuous intracellular structures, which was those observed for GFP-tagged Ppz1 (32.8% of moni- detectable in 53.3% of the cells monitored. This beha- tored cells). This result suggested that Hal3 might viour was much less evident (13.4%) in the Ppz1- actually escort Ppz1 to this particular subcellular loca- R451L version, whereas co-expression of Hal3 did not tion. To test this possibility, new constructs expressing mCherry-tagged versions of Ppz1 and Hal3 were pre- pared and colocalization experiments carried out. As (A) observed in Fig. 3, the distribution patterns of the phosphatase and its regulatory subunit were largely overlapping (> 95% of cells with signal for both dyes), suggesting a concurrence of both proteins at the same intracellular structures. The shape and number of the observed structures were compatible with the mem- branes of certain organelles such as vacuoles or (B) Fig. 1. Expression of Hal3 from a multicopy plasmid alters the subcellular localization of the native version of Ppz1. (A) DBY746 cells carrying the indicated combinations of plasmids were treated as described in Materials and Methods and observed at the Fig. 2. Hal3-GFP is cytosolic but can be detected in internal fluorescence microscope. The GFP signal was captured with a structures when Ppz1 is expressed from a multicopy plasmid. FITC filter. Excitation was performed during 3000 (panel A) or DBY746 cells were transformed with a multicopy Hal3-GFP vector 1500 ms (panel B). YEp-⌀ denotes the empty YEplac181 vector. plus an empty YEp vector (YEp-⌀) or the same vector carrying Bars (5 µm) are included for reference. The number of cells PPZ1, as indicated. The GFP signal was captured during 1000 ms. examined ranged from 127 to 341. At least 200 cells were examined. 4 FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies M. Albacar et al. Subcellular relocalization of Ppz1 phosphatase (A) Ppz1-GFP Hal3-mCherry merge Fig. 3. Ppz1 and Hal3 colocalize in internal structures in vivo. DBY746 carrying the following combination of plasmids: (A) Ppz1-GFP + Hal3-mCherry and (B) Ppz1- mCherry + Hal3-GFP were treated as described in Materials and Methods. GFP (B) Ppz1-mCherry Hal3-GFP merge and mCherry signal were analyzed using a FITC and a Texas Red filter respectively. Excitation was performed during 1500 ms for GFP and 2000 ms for mCherry. A minimum of 250 cells per condition was monitored. perinuclear endoplasmic reticulum (ER). To test the Ppz1-GFP does not decorate the entire vacuolar or possibility of Ppz1 being located at vacuolar mem- ER membranes, suggesting the existence of preferential branes, we stained these membranes with the lipophilic localization regions. Therefore, it can be concluded dye FM4-64 [34]. As shown in Fig. 4A, substantial that concomitant overexpression of Ppz1 and Hal3 colocalization of Ppz1-GFP with structures stained directs the phosphatase from the plasma membrane to with FM4-64 was observed, and this phenomenon was internal membranes (mostly the vacuolar membrane) identified in nearly half (44.2%) of monitored cells. and that Ppz1 is accompanied by its negative regula- Similarly, Ppz1-GFP was co-expressed with a plasmid tory subunit. carrying mCherry associated with the HDEL sequence, which promotes ER localization [35]. With this tool, Effect of the presence of Hal3 in the localization we could observe in some instances (10.6% of exam- of conditionally expressed Ppz1 ined cells after 2 h and 11.7% after 6 h of induction) colocalization of Ppz1-GFP with mCherry-HDEL con- The experiments presented above involve the co- taining membranes (Fig. 4B). It is worth noting that expression of Ppz1 and Hal3 from their own Ppz1-GFP FM4-64 merge (A) Fig. 4. Ppz1 mainly localizes in vacuolar membranes when cells carry a Hal3 multicopy plasmid and in some cases in the ER. (A) DBY746 cells carrying the plasmids YEp195 Ppz1-GFP and YEp181 Hal3 were treated with FM4-64 as described in Materials and Methods. Cell were then monitored in a confocal microscope (Leica TCS SP5). Two hundred Ppz1-GFP mCherry-HDEL merge fifty eight cells were analyzed. (B) (B) MAC003 cells transformed with plasmids YEp195 Hal3 and pRS315 mCherry-HDEL and cells observed after 2 h of galactose addition under a Nikon Eclipse TE2000-E fluorescence microscope. A total number of 218 cells were analyzed, and ER localization was observed in 10.6% of the population. FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies 5 Subcellular relocalization of Ppz1 phosphatase M. Albacar et al. promoters and a multicopy plasmid. It is known that the and that its level increased with time and remained high expression level achieved for Ppz1 in this experimental set- at least until 20 h after induction. Microscopic analysis of ting is relatively modest, due to copy number restriction these cultures (Fig. 5B) showed that, in the absence of [15]. We wanted to know whether the observed shift of Hal3 co-expression, Ppz1-GFP was already detectable (al- Ppz1 from the plasma membrane to internal membranes beit with a relatively weak signal) at the cell periphery could also be observed when Ppz1 was expressed at higher after 2 h of induction. The Ppz1 signal increased with time levels from a chromosomal copy and driven by the power- and largely maintained its peripheral distribution. Remark- ful GAL1 promoter. To this end, we created strain ably, when Hal3 was co-expressed, after 2 h of induction, MAC003, which is equivalent to the previously character- Ppz1 was already identifiable in internal structures in ized strain ZCZ01 [16,17], although in this case Ppz1 con- 16.7% of the cells. The intensity and abundance of these tains a C-terminal GFP tag. As shown in Fig. S1, the intracellular signals increased after 4 and 6 h of induction intensity of the toxic effect of GFP-tagged Ppz1 in strain (38.7% and 42.5% of cells, respectively), with lesser MAC003 is indistinguishable from that of strain ZCZ01. amounts of Ppz1 observable at the periphery. After 20 h Thus, strain MAC003 became an interesting tool because of induction, most Ppz1 was found associated with inter- it allowed following the fate of Ppz1 within the cell during nal structures (80.1% of cells), and little or no signal was the process of accumulation of the phosphatase. When the detected at the cell periphery. These results suggests that level of Ppz1-GFP accumulation in MAC003 cells was the presence of Hal3 allows removal of Ppz1 from the monitored by immunoblot (Fig. 5A), we observed that the plasma membrane as soon as the level of the phosphatase phosphatase was undetectable prior induction, as expected may become dangerous for the cell (see Discussion). It is for GAL-driven expression in cells growing on raffinose, worth noting that the presence of Ppz1 at these internal (A) Fig. 5. Time course of Ppz1-GFP (B) expression and localization in the MAC003 strain. MAC003 cells transformed with an empty YEplac195 plasmid (YEp-⌀) or with the same vector carrying HAL3 (YEp-Hal3) were grown until saturation, and cultures were processed for Ppz1 overexpression. (A) Cultures were processed for SDS/ PAGE electrophoresis (8% acrylamide gels) and transferred to membranes. Ppz1- GFP was detected by mean of anti-GFP antibodies. No bands below the 70 kDa marker were observed. Ponceau red staining is shown to monitor loading and transfer efficiency. (B) Monitoring of Ppz1- GFP signal in cultures prepared as above at the fluorescence microscope at different times upon induction. The number of cells documented ranged from 144 to 328. Exposition was performed during 3000 ms (YEp-⌀) and 1500 ms (YEp-Hal3). 6 FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies M. Albacar et al. Subcellular relocalization of Ppz1 phosphatase structures is not a transient event, as it is observed even Vps27 disturbs the observed relocalization of Ppz1 and 20 h after induction of the expression. Hal3 when both proteins are co-expressed. Prompted by these results, we monitored growth in wild-type and vps27 cells overexpressing Ppz1-GFP and in the pres- Mutation of VPS27 disrupts Ppz1-Hal3 ence or absence of Hal3 overexpression. As shown in colocalization in internal structures and impairs Fig. 7A, the vps27 mutation by itself did not signifi- Hal3-mediated growth improvement cantly affect growth of cells carrying the empty plas- Overexpression of Hal3 from a multicopy plasmid is mids, while expression of Ppz1-GFP from the able to fully eliminate the growth defect of cells over- multicopy plasmid moderately impaired growth of the expressing Ppz1 from its own promoter in high-copy wild-type strain. It is worth noting that expression of number [9] or from the GAL1 promoter (reference [16] Ppz1 from its own promoter, even from a multicopy and Fig. S1). Therefore, we wanted to test whether the plasmid, results in less dramatic effects on growth than ability to normalize cell functions was related to the when the phosphatase is expressed from the strong shift in subcellular localization. To this end, we used a GAL promoter (compare with Fig. S1). Interestingly, vps27 deletion mutant, which lacks the Vps27 protein, overexpression of Ppz1 in the vps27 background a component of the Endosomal Sorting Complex resulted in a growth defect much more dramatic than Required for Transport (ESCRT) [36]. Wild-type that observed in the wild-type strain. In addition, DBY746 and vps27 cells were transformed with while in the wild-type background overexpression of YEp195 Ppz1-GFP plus either YEp181 or YEp181 Hal3 fully counteracted the impact of Ppz1 overexpres- Hal3. As presented in Fig. 6, co-expression of Ppz1 sion, the growth defect was barely attenuated in the and Hal3 in the wild-type strain shows Ppz1 associated case of the vps27 strain. It is important to note that, with the mentioned internal membranes. In contrast, as deduced from the immunoblot analysis presented in these were rarely seen in the vps27 background (4.1% Fig. 7B, overexpression of Hal3 allows a much higher of cells), while more Ppz1 could be detected at the cell accumulation of Ppz1, but the levels of Ppz1 and the periphery. In addition, in the vps27 strain, Hal3 proportion between the amounts of Ppz1 and Hal3 in showed a diffuse cytoplasmatic localization, with mini- wild-type and vps27 cells are very much alike. This mal localization at internal structures when Ppz1 was implies that the strongly impaired growth of vps27 co-expressed (3.6% of cells). Therefore, the lack of cells expressing Ppz1 is not due to an abnormal (A) Ppz1-GFP (B) Ppz1-GFP + Hal3 3s 1.5 s WT 3s 1.5 s vps27 1s (C) Hal3-GFP (D) Hal3-GFP + Ppz1 Fig. 6. Mutation of VPS27 disrupts 1s 1s colocalization of Ppz1 and Hal3 in internal structures. DBY746 (WT) and vps27 cells carrying the combination (A) Ppz1-GFP + WT empty vector, (B) Ppz1-GFP + Hal3, (C) empty vector + Hal3-GFP and (D) Ppz1 + Hal3-GFP were grown until saturation in 1s 1s the appropriate selective medium and treated as described in Material and vps27 Methods. Pictures are representative of 257 to 343 documented cells. The time of exposure is shown for each image. FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies 7 Subcellular relocalization of Ppz1 phosphatase M. Albacar et al. (A) (B) WT vps27 Hal3 1.6 Ppz1-GFP kDa 1.4 Ppz1- 130 - GFP Cell growth (OD600) 1.2 1.0 70 - Ponceau 0.8 55 - staining 0.6 0.4 70 - Hal3 55 - 0.2 70 - 0.0 Ponceau 0 5 10 15 20 25 30 35 55 - staining h Fig. 7. Mutation of VPS27 aggravates Ppz1 toxicity and impairs Hal3-mediated growth rate normalization. (A) Wild-type strain DBY746 (WT, empty symbols) and vps27 (filled symbols) were co-transformed with the empty vectors (circles), Ppz1-GFP plus an empty vector (triangles) or Ppz1-GFP plus the YEp-Hal3 plasmid (squares). Cells were processed for growth determination in liquid culture at an initial OD600 of 0.02. Values are the mean  SEM of 5 to 12 independent cultures. (B) Immunodetection of Ppz1 and Hal3 levels in DBY746 (WT) and vps27 cells. Whole yeast extracts were prepared as described in Materials and Methods (cultures carrying Ppz1-GFP were grown and assayed in duplicate). Two gels (10% acrylamide) were loaded in parallel; one was processed for anti-Ppz1 detection (upper panel); and the other for anti-Hal3 detection (lower panel). Ponceau red staining is shown to verify loading and transfer efficiency. accumulation of the phosphatase or to insufficient suggesting that myristoylation of Gly2 would be neces- expression levels of the regulatory subunit. Therefore, sary for the relocalization process. these results indicate that the lack of vps27 clearly We observed that, whereas Hal3 accompanied Ppz1 worsens the toxic phenotype derived from Ppz1 over- to these internal structures, significant amounts of expression and that, in the absence of this essential this regulatory subunit were still detected at the cyto- component of the ESCRT pathway, the presence of sol. This could be explained by Hal3’s moonlighting Hal3 cannot modify the intracellular localization of function in CoA biosynthesis : a portion of Hal3 Ppz1 nor alleviate the growth defect caused by the molecules would remain cytosolic as part of the excess of phosphatase. PPCDC complex [24,37]. Using strain MAC003, we also observe (Fig. 5) that Ppz1 is recruited to internal structures shortly after its expression is induced (2 h). Discussion It should be noted that our recent work with the We confirm in this work that, upon overexpression, equivalent strain ZCZ01 demonstrated that 2 h of most Ppz1 can be detected at the cell periphery and induction is sufficient to provoke dramatic changes in that the myristoylable Gly2 is a requisite for such the transcriptome and phosphoproteome of the cell, peripheral localization. We also demonstrate that, in and a significant arrest in cell proliferation [16–18]. contrast, the phosphatase function of Ppz1 is not Since co-expression of Hal3 fully normalizes cell essential for its localization. More importantly, our growth, it could be assumed that the toxic effects of results indicate that in cells co-expressing Ppz1 and Ppz1 are neutralized as soon as they appear by Hal3, most Ppz1 is removed from the cell periphery means of Hal3-mediated relocalization of the phos- and recruited in internal membranes. This is not phatase. observed for the catalytically impaired Ppz1 version, Because most known targets for Ppz1, such as Trk1/ suggesting that relocalization of Ppz1 is triggered by Trk2 and Nha1 (and perhaps Pma1), are located at the excess of Ppz1 activity. Similarly, the Ppz1G2A ver- the plasma membrane, one might assume that removal sion remains fully cytosolic when Hal3 is co-expressed, of Ppz1 from the cell periphery should be sufficient for 8 FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies M. Albacar et al. Subcellular relocalization of Ppz1 phosphatase counteracting toxicity. However, the observation that, Hse1, Vps27 composes the ESCRT-0 complex, which when expressed from plasmid pCM190 (a tetO- is the first and necessary step for progression in the regulatable vector), mutation of Gly2 to Ala leads only ESCRT pathway. This conserved pathway is involved to a slight attenuation of toxicity [19], indicates that in diverse cellular functions, such as eukaryotic cell very relevant targets for Ppz1 toxicity are not located abscission, exosome secretion, or macroautophagy and at the plasma membrane. Among such targets could microautophagy, but it has a recognized role in traf- be diverse cytosolic protein kinases (i.e. Rck2, Tda1, ficking of membrane proteins to vacuole for degrada- Npr1, Hog1 or Snf1) whose phosphorylation state tion [36,38,39]. Our results clearly indicate that the (and likely activation state) was found significantly removal of Ppz1 from the plasma membrane requires altered upon short-time induction of Ppz1 expression the ESCRT pathway. However, the current evidence [17]. Therefore, one could assume that relocalization argues against the excess of Ppz1 being directed to the of Ppz1 to internal membranes would be a strategy vacuole for degradation because (a) even after a long not only to remove Ppz1 from the plasma membrane, period of Ppz1 overexpression, the phosphatase is but also to avoid excessive amounts of cytosolic phos- mainly detected in the vacuolar membrane but very phatase. Because, as far as we know, Ppz1 has not rarely in the vacuolar lumen (Fig. 5) and (b) co- been identified by proteomic analysis in vacuolar mem- expression of Hal3 does not lead to a decrease in the branes of cells grown under standard conditions [25], amount of Ppz1, but instead, it allows for the accumu- it could be postulated that relocalization of Ppz1 lation of the phosphatase [see [16] and Fig. 5A and would be, rather than an ordinary regulatory mecha- 7B]. It could be speculated that Ppz1 may be trapped nism, an exceptional safeguard instrument against the at the vacuolar (and ER) membrane by its N- toxic effects of Ppz1. This concept fits well with the myristoylated N-terminus. There are some examples of observation that the inactive version of the phos- yeast proteins that translocate to the vacuolar mem- phatase (Ppz1R451L), which is expressed at higher levels brane as part of their normal function. For instance, than native Ppz1 but has virtually no effect on cell Saccharomyces cerevisiae Sch9 (a member of the AGC growth [16], is not internalized by co-expression of family of protein kinases) displays a dynamic localiza- Hal3 (Fig. 1). It is unknown what would be the trigger tion, shuttling from cytosol to the vacuolar membrane for the association of Ppz1 and Hal3 leading to the where it could be phosphorylated and activated by the accumulation of both proteins at internal membranes. TORC1 complex leading to specific modulation of the However, we showed very recently that cells overex- Sch9 branch of TORC1 signalling [40,41]. Another pressing Ppz1 suffer a drop in internal pH, likely due example could be the PpAtg9 protein from Pichia pas- to the combination of an hyperactivation of Nha1 and toris, which localizes in structures near the plasma the inhibition of the Pma1 proton pump [18]. It has membrane during methanol growth and, on glucose- been proposed that the interaction between Ppz1 and induced macropexophagy, it translocates to perivacuo- Hal3 could be modulated by pH, in a way that acidifi- lar structures [42]. At this point, it is not clear why cation would increase Ppz1-Hal3 interaction [7]. There- Ppz1 is maintained in these internal structures and not fore, a plausible scenario would be that the abnormal directed to degradation, although a possibility could intracellular acidification would lead to an increased be the recycling of Ppz1 when its levels decrease below interaction between Hal3 and Ppz1 that would firstly the toxic threshold. result in direct inhibition of the Ppz1 activity and then in the sequestering of the phosphatase away from its Acknowledgements targets. The notion that Hal3-mediated relocalization of The technical assistance of Montserrat Robledo is Ppz1 to internal membranes is crucial for toxicity safe- acknowledged. We thank Oriol Gallego (Pompeu guard is reinforced by the observation that mutation Fabra University, Barcelona) for plasmid pRS315 of VPS27 eliminates Ppz1 relocalization, worsens the mCherry-HDEL, Ivana Malcova (Institute of Microbi- cell growth defect of cells overexpressing Ppz1 and lar- ology, CAS, Hungary) for plasmid pFA6a-mCherry- gely abolishes the beneficial effect of co-expression of kanMX6 and Cesar Roncero (Universidad de Sala- Hal3. We have no reason to believe that elimination of manca, Spain) for inspiring comments on the use of Vps27 might affect the ability of Hal3 to directly inter- the vps27 mutant. This research was funded by grant act with Ppz1 and to inhibit its phosphatase activity. BFU2017-82574-P to JA and AC from the Ministerio Therefore, it must be assumed that the inhibition de Economıa, Industria y Competitividad (Spain). MA of Ppz1 activity could be necessary but not sufficient and DV were recipient of PhD fellowships from the to counteract Ppz1-mediated toxicity. Together with PIF UAB program (Spain). FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies 9 Subcellular relocalization of Ppz1 phosphatase M. Albacar et al. Author contributions activated protein kinase pathway to increases in internal turgor pressure caused by loss of Ppz protein AC and JA conceived and supervised the study; MA phosphatases. Eukaryot Cell. 2004;3:100–7. carried out most of the experiments; DV provided new 11 De Nadal E, Fadden RPP, Ruiz A, Haystead T, Ari~ no tools and reagents; MA, AC and JA analysed data; JA J. A role for the Ppz Ser/Thr protein phosphatases in wrote the manuscript and all authors made manuscript the regulation of translation elongation factor 1Balpha. revisions. J Biol Chem. 2001;276:14829–34. 12 Lee S, Tumolo JM, Ehlinger AC, Jernigan KK, Qualls- Histed SJ, Hsu P-C, et al. Ubiquitin turnover and Data accessibility endocytic trafficking in yeast are regulated by Ser57 The data that support the findings of this study are phosphorylation of ubiquitin. Elife. 2017;6:76. available in Figs. 1 to 7 and the supplementary mate- 13 Lee S, Ho H-C, Tumolo JM, Hsu P-C, MacGurn JA. rial of this article (Table S1 and Fig. S1). Additional Methionine triggers Ppz-mediated dephosphorylation of Art1 to promote cargo-specific endocytosis. J Cell Biol. data, such as sets of micrographs for specific experi- 2019;218:977–92. ments, are available from the corresponding author 14 Clotet J, Posas F, De Nadal E, Arino J. The NH2- [Joaquin.Arino@uab.es] upon reasonable request. terminal extension of protein phosphatase PPZ1 has an essential functional role. J Biol Chem. 1996;271:26349– 55. References 15 Makanae K, Kintaka R, Makino T, Kitano H, Moriya 1 Ari~no J, Velazquez D, Casamayor A. Ser/Thr protein H. Identification of dosage-sensitive genes in phosphatases in fungi: Structure, regulation and Saccharomyces cerevisiae using the genetic tug-of-war function. Microb Cell. 2019;6:217–56. method. Genome Res. 2013;23:300–11. 2 Offley SR, Schmidt MC. Protein phosphatases of 16 Calafı C, Lopez-Malo M, Velazquez D, Zhang C, Saccharomyces cerevisiae. Curr Genet. 2019;65:41–55. Fernandez-Fernandez J, Rodrıguez-Galan O, et al. 3 Posas F, Casamayor A, Morral N, Ari~ no J. Molecular Overexpression of budding yeast protein phosphatase cloning and analysis of a yeast protein phosphatase Ppz1 impairs translation. Biochim Biophys Acta Mol with an unusual amino-terminal region. J Biol Chem. Cell Res. 2020;1867:118727. 1992;267:11734–40. 17 Velazquez D, Albacar M, Zhang C, Calafı C, L opez- 4 Lee KS, Hines LK, Levin DE. A pair of functionally Malo M, Torres-Torronteras J, et al. Yeast Ppz1 redundant yeast genes (PPZ1 and PPZ2) encoding type 1- protein phosphatase toxicity involves the alteration of related protein phosphatases function within the PKC1- multiple cellular targets. Sci Rep. 2020;10:15613. mediated pathway. Mol Cell Biol. 1993;13:5843–53. 18 Albacar M, Sacka L, Calafı C, Velazquez D, 5 Posas F, Camps M, Ari~ no J. The PPZ protein Casamayor A, Ari~ no J, et al. The Toxic Effects of Ppz1 phosphatases are important determinants of salt Overexpression Involve Nha1-Mediated Deregulation of tolerance in yeast cells. J Biol Chem. 1995;270:13036–41. K+ and H+ Homeostasis. J Fungi (Basel, Switzerland). 6 Ruiz A, Yenush L, Ari~ no J. Regulation of ENA1 Na+- 2021;7:1010. ATPase gene expression by the Ppz1 protein 19 Calafı C, Lopez-Malo M, Albacar M, Casamayor A, phosphatase is mediated by the calcineurin pathway. Ari~no J. The N-terminal region of yeast protein Eukaryot Cell. 2003;2(5):937–948. phosphatase Ppz1 is a determinant for its toxicity. Int J 7 Yenush L, Merchan S, Holmes J, Serrano R. pH- Mol Sci. 2020;21:1–16. Responsive, Posttranslational Regulation of the Trk1 20 Mu~ noz I, Sim on E, Casals N, Clotet J, Ari~no J. Potassium Transporter by the Type 1-Related Ppz1 Identification of multicopy suppressors of cell cycle Phosphatase. Mol Cell Biol. 2005;25:8683–92. arrest at the G1 - S transition in Saccharomyces 8 Yenush L, Mulet JM, Ari~ no J, Serrano R. The Ppz cerevisiae. Yeast. 2003;20:157–69. protein phosphatases are key regulators of K+ and pH 21 Ruiz A, Mu~ noz I, Serrano R, Gonzalez A, Simon E, homeostasis: implications for salt tolerance, cell wall Arino J. Functional characterization of the integrity and cell cycle progression. EMBO J. Saccharomyces cerevisiae VHS3 gene: a regulatory 2002;21:920–9. subunit of the Ppz1 protein phosphatase with novel, 9 De Nadal E, Clotet J, Posas F, Serrano R, Gomez N, phosphatase-unrelated functions. J Biol Chem. Ari~no J. The yeast halotolerance determinant Hal3p is an 2004;279:34421–30. inhibitory subunit of the Ppz1p Ser/Thr protein 22 Abrie JA, Molero C, Ari~ no J, Strauss E. Complex phosphatase. Proc Natl Acad Sci USA. 1998;95:7357–62. stability and dynamic subunit interchange modulates 10 Merchan S, Bernal D, Serrano R, Yenush L. Response the disparate activities of the yeast moonlighting of the Saccharomyces cerevisiae Mpk1 mitogen- proteins Hal3 and Vhs3. Sci Rep. 2015;5:15774. 10 FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies M. Albacar et al. Subcellular relocalization of Ppz1 phosphatase 23 Abrie JAA, Gonzalez A, Strauss E, Ari~ no J. Functional 36 Hurley JH. The ESCRT complexes. Crit Rev Biochem mapping of the disparate activities of the yeast Mol Biol. 2010;45:463–87. moonlighting protein Hal3. Biochem J. 2012;442:357–68. 37 Olzhausen J, Moritz T, Neetz T, Sch€ uller H-J. 24 Ruiz A, Gonzalez A, Munoz I, Serrano R, Abrie JA, Molecular characterization of the heteromeric coenzyme Strauss E, et al. Moonlighting proteins Hal3 and Vhs3 A-synthesizing protein complex (CoA-SPC) in the yeast form a heteromeric PPCDC with Ykl088w in yeast Saccharomyces cerevisiae. FEMS Yeast Res. CoA biosynthesis. Nat Chem Biol. 2009;5:920–8. 2013;13:565–73. 25 Wiederhold E, Veenhoff LM, Poolman B, Slotboom 38 Laidlaw KME, MacDonald C. Endosomal trafficking DJ. Proteomics of Saccharomyces cerevisiae Organelles. of yeast membrane proteins. Biochem Soc Trans. Mol Cell Proteomics. 2010;9:431–45. 2018;46:1551–8. 26 Venturi GM, Bloecher T, Williams-Hart T, Tatchell K. 39 Pavlin MR, Hurley JH. The ESCRTs – converging on Genetic interactions between GLC7, PPZ1 and PPZ2 mechanism. J Cell Sci. 2020;133:240333. in Saccharomyces cerevisiae. Genetics. 2000;155:69–83. 40 Jorgensen P, Rupes I, Sharom JR, Schneper L, 27 Adams A, Gottschling DE, Kaiser CA, Stearns T. Broach JR, Tyers M. A dynamic transcriptional Methods in yeast genetics : a Cold Spring Harbor network communicates growth potential to ribosome Laboratory course manual Cold. Suffolk County, NY, synthesis and critical cell size. Genes Dev. USA: Spring Harbor Laboratory Press. 1998. 2004;18:2491–505. 28 Sambrook J, Russell DW. Molecular Cloning: A 41 Novarina D, Guerra P, Milias-Argeitis A. Vacuolar Laboratory Manual. Cold Spring Harbor, NY: Cold Localization via the N-terminal Domain of Sch9 is Spring Harbor Laboratory Press; 2001. Required for TORC1-dependent phosphorylation and 29 Ito H, Fukuda Y, Murata K, Kimura A. downstream signal transduction. J Mol Biol. Transformation of intact yeast cells treated with alkali 2021;433:167326. cations. J Bacteriol. 1983;153:163–8. 42 Chang T, Schroder LA, Thomson JM, Klocman AS, 30 Janke C, Magiera MM, Rathfelder N, Taxis C, Reber Tomasini AJ, Strømhaug PE, et al. PpATG9 encodes a S, Maekawa H, et al. A versatile toolbox for PCR- novel membrane protein that traffics to vacuolar based tagging of yeast genes: New fluorescent proteins, membranes, which sequester peroxisomes during more markers and promoter substitution cassettes. pexophagy in Pichia pastoris. Mol Biol Cell. Yeast. 2004;21:947–62. 2005;16:4941–53. 31 Malcova I, Farkasovsky M, Senohrabkova L, Vasicova 43 Brachmann CB, Davies A, Cost GJ, Caputo E, Li J, P, Hasek J. New integrative modules for multicolor- Hieter P, et al. Designer deletion strains derived from protein labeling and live-cell imaging in Saccharomyces Saccharomyces cerevisiae S288C: a useful set of cerevisiae. FEMS Yeast Res. 2016;16:56. strains and plasmids for PCR-mediated gene 32 Gonzalez A, Ruiz A, Serrano R, Arino J, Casamayor A. disruption and other applications. Yeast. 1998;14:115– Transcriptional profiling of the protein phosphatase 2C 32. family in yeast provides insights into the unique functional roles of Ptc1. J Biol Chem. 2006;281:35057–69. 33 Ferrando A, Kron SJ, Rios G, Fink GR, Serrano R. Supporting information Regulation of cation transport in Saccharomyces cerevisiae the salt tolerance gene HAL3. Mol Cell Biol. Additional supporting information may be found 1995;15:5470–81. online in the Supporting Information section at the end 34 Baggett J, Shaw J, Sciambi C, Watson H, Wendland B. of the article. Fluorescent labeling of yeast. Curr Prot Cell Biol. Fig. S1. Characterization of strain MAC003, overex- 2003;20:1–4. pressing a chromosomally encoded GFP-tagged ver- 35 Pelham HR. Evidence that luminal ER proteins are sion of Ppz1. sorted from secreted proteins in a post-ER Table S1. Oligonucleotides employed in this work. compartment. EMBO J. 1988;7:913–8. FEBS Letters (2022) ª 2022 The Authors. FEBS Letters published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies 11