*Article* **The Cell Wall Integrity Receptor Mtl1 Contributes to Articulate Autophagic Responses When Glucose Availability Is Compromised**

**Sandra Montella-Manuel , Nuria Pujol-Carrion and Maria Angeles de la Torre-Ruiz \***

Cell Signalling in Yeast Unit, Department of Basic Medical Sciences, Institut de Recerca Biomèdica de Lleida (IRBLleida), University of Lleida, 25198 Lleida, Spain; sandra.montella@udl.cat (S.M.-M.); nuria.pujol@udl.cat (N.P.-C.)

**\*** Correspondence: mariaangeles.delatorre@udl.cat

**Abstract:** Mtl1protein is a cell wall receptor belonging to the CWI pathway. Mtl1 function is related to glucose and oxidative stress signaling. In this report, we show data demonstrating that Mtl1 plays a critical role in the detection of a descent in glucose concentration, in order to activate bulk autophagy machinery as a response to nutrient deprivation and to maintain cell survival in starvation conditions. Autophagy is a tightly regulated mechanism involving several signaling pathways. The data here show that in *Saccharomyces cerevisiae*, Mtl1 signals glucose availability to either Ras2 or Sch9 proteins converging in Atg1 phosphorylation and autophagy induction. TORC1 complex function is not involved in autophagy induction during the diauxic shift when glucose is limited. In this context, the *GCN2* gene is required to regulate autophagy activation upon amino acid starvation independent of the TORC1 complex. Mtl1 function is also involved in signaling the autophagic degradation of mitochondria during the stationary phase through both Ras2 and Sch9, in a manner dependent on either Atg33 and Atg11 proteins and independent of the Atg32 protein, the mitophagy receptor. All of the above suggest a pivotal signaling role for Mtl1 in maintaining correct cell homeostasis function in periods of glucose scarcity in budding yeast.

**Keywords:** cell wall integrity (CWI); Mtl1; autophagy; glucose; mitophagy; *Saccharomyces cerevisiae*

## **1. Introduction**

Carbon sources have a major impact on *Saccharomyces cerevisiae* metabolism and also affect longevity. Yeast uses a fermentative metabolism when the preferred carbon source, glucose, is abundant and ethanol, organic acids, and ATP are accumulated. Glucose limitation induces slowing of growth, contributing to the switch to respiratory metabolism, the hallmark of the diauxic shift, which along with other metabolic configurations prepares cells for the stationary phase and the process of chronological ageing [1]. Budding yeast chronological life span (CLS) forced by glucose restriction is also dependent on the availability of other nutrients in the culture media, such as amino acids or nitrogen among others [2–5]. In a simple way, glucose is a promotor of ageing through the activation of the Ras-cAMP-PKA pathway and TORC1-Sch9 [6–8] pathways, whereas amino acids operate through Gcn2 protein [9–12] as well as through activation of the TORC1 complex, upstream regulators of the Sch9 protein [13].

The TORC1 and Ras/PKA pathways are both negative and independent regulators of autophagy [13–16]. Sch9 downregulates autophagy independently and coordinately with the Ras/PKA pathway [17].

Ras proteins are essential for growth in fermentable carbon sources such as glucose. In that context, Ras proteins trigger the synthesis of cAMP and activation of the PKA pathway upon binding to the inhibitor Bcy1 protein. The absence of Ras proteins does not allow growth in non-fermentative carbon sources (see reviews [18,19]).

**Citation:** Montella-Manuel, S.; Pujol-Carrion, N.; de la Torre-Ruiz, M.A. The Cell Wall Integrity Receptor Mtl1 Contributes to Articulate Autophagic Responses When Glucose Availability Is Compromised. *J. Fungi* **2021**, *7*, 903. https://doi.org/ 10.3390/jof7110903

Academic Editors: María Molina and Humberto Martín

Received: 24 September 2021 Accepted: 23 October 2021 Published: 26 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The sucrose non-fermenting protein kinase, Snf1, is the yeast orthologue of mammalian AMPK, and is important for cell adaptation to glucose limitation. Snf1 is a key component of the main glucose repression pathway in yeast and controls genes involved in alternative carbon sources and metabolism. However, regulation of adaptation to glucose limitation is the main role of the SNF1 complex [20,21]. Snf1 also plays a role in autophagy [22]; its negative regulation is required to downregulate autophagy under certain conditions of nutritional limitation [23].

Mtl1 protein is a transmembrane protein cell wall mechano-sensor [24,25], part of the cell wall integrity pathway (CWI) with structural similarity to its paralogue Mid2 protein [26,27]. Mtl1 is also required to activate a stress response towards TORC1 and Ras/PKA signaling pathways under conditions of both oxidative stress and glucose starvation [28]. In addition, this cell wall receptor plays a role in Cyclin C localization and programmed cell death [29], as well as in the preservation of mitochondrial integrity and life span by regulating TORC1, Sch9, Slt2, and PKA [30].

The CWI is involved in sensing and transducing a wide variety of stimuli (see reviews [31,32]), including both nutritional and oxidative stresses [27,30,33]. In this pathway, a wide number of sensors are specialized in the detection of different stresses [25,34], which converge on the GTPase Rho1 to subsequently activate Pkc1 protein (see reviews [31,32]). Pkc1 in turn phosphorylates the Bck1 protein, the MAPKKK, thus activating the MAPKK module composed of Mkk1/Mkk2 proteins leading to dual phosphorylation and activation of the last member of the pathway, the MAPK Slt2 protein, whose double phosphorylation is impaired in the mutant *mtl1* [30].

Nutrient limitation strongly induces macroautophagy [23,35–37] to accomplish two main objectives; one is to detoxify and the second is to recycle components to build newly synthetized molecules.

Macroautophagy is a process in which several components (damaged or superfluous organelles, cytoplasmic elements, microorganisms . . . ) are engulfed within cytosolic double membrane vesicles named autophagosomes. The outer membrane of the phagosome fuses to the vacuolar membrane releasing a spherical body termed autophagic body, that is digested by hydrolases releasing the breakdown products back to the cytosol to be recycled by cells [38]. Morphological and structural characteristics of autophagy are highly conserved from yeast to humans.

The signaling network governing life span usually converges in the autophagic machinery [17,23,39]. In general, nutrient deprivation impinges on Atg13 dephosphorylation that triggers Atg1 kinase activity then leading to the formation of the complex Atg13/Atg1/Atg17/Atg29/Atg31 activating the autophagy process [40,41]. Atg7 is an E1-like enzyme essential for macroautophagy since it is part of the Atg8-Ibl conjugation system [42,43]. Mutants defective in autophagy display shorter life spans [44,45].

Autophagy entails nonselective engulfment of cytoplasmic components, but there are also other types of autophagy that selectively degrade specific cellular elements (see review [46]), such is the case of mitophagy. Mitophagy clears dysfunctional mitochondria and impinges on cellular function by promoting respiration proficiency during the process of ageing (see review [47]). Selective mitophagy requires the function of the Atg32 protein as a mitochondrial receptor and its binding to the adaptor Atg11, which interacts with the Atg8 protein in the phagophore inner surface (see review [48]). Atg33 is a yeast protein located at the outer mitochondrial membrane, its absence suppressed mitophagy in post-log cultures, however, its precise role in mitophagy is still controversial [49].

In this report, we show that bulk autophagy is highly induced during the transition to diauxic shift in a manner totally dependent on glucose and amino acids availability. The Mtl1 cell wall receptor is essential to sense glucose concentration and transmit the signal to both Ras2 and Sch9 to phosphorylate Atg1 and to activate the macroautophagic machinery. Gcn2 is the amino acid sensor. Both Mtl1 and Gcn2 operate independently of TORC1 in the signaling process leading to the activation of bulk autophagy. Moreover, Mtl1 is also

relevant for mitochondrial clearance dependent on Atg33 and the inactivation of either Ras2 or Sch9 in response to glucose exhaustion.

#### **2. Materials and Methods**

#### *2.1. Yeast Strains and Plasmids*

*Saccharomyces cerevisiae* strains are listed in Table 1. All the strains named GSL are derivatives of the CML128 background. New null mutants described in this study were obtained by a one-step disruption method that used the *Nat*Mx4 or *Kan*Mx4 cassettes [50]. Strains GSL197, 198, 199, 200, 201, 202, 226, 265, 279, 297, 352, 370, 382, 414, and 415 were constructed upon integration of plasmid pGFP–Atg8 (original name: pHab142), previously digested with Stu1, in the *URA3* gene. Strains GSL372 and 416 were constructed upon integration of plasmid pAtg1-HA previously digested with BstEII. The plasmid pAtg1-HA was obtained upon Atg1 cloning into the Pme1 and PstI sites of the integrative vector pMM351 [51].


**Table 1.** Yeast strains.

Plasmid descriptions are listed in Table 2. Each particular ORF was amplified by PCR from genomic DNA to be directionally cloned in the specific plasmid.


**Table 2.** Plasmids employed.

#### *2.2. Media, Growth Conditions and Reagents*

Yeasts were grown at 30 ◦C in SD medium (2% glucose, 0.67% yeast nitrogen base that lacked the corresponding amino acids for plasmid maintenance) plus amino acids [59].

Glucose depletion consisted of SD medium without glucose plus amino acids. Nitrogen depletion consisted of SD medium whose nitrogen base component was free of amino acids and ammonium sulphate plus amino acids. Amino acid depletion consisted of SD medium without adding amino acids. Glycerol medium (3% glycerol, 0.67% yeast nitrogen base that lacked the corresponding amino acids) plus amino acids. Sucrose medium (2% sucrose, 0.67% yeast nitrogen base that lacked the corresponding amino acids) plus amino acids. Fructose medium (2% fructose, 0.67% yeast nitrogen base that lacked the corresponding amino acids) plus amino acids.

Glucose was added as α-D-glucose monohydrate (Serva, Heidelberg, Germany 22720.01) at a final concentration of 2%; Amino acids were added at concentrations: 60 mg/mL Leucine, 20 mg/mL Histidine and 20 mg/mL Tryptophan. Nitrogen was added as Yeast Nitrogen Base w/o Amino Acids (Difco, Franklin Lakes, NJ, USA, 291940) at a final concentration of 0.67%. Sucrose was added as Sucrose (Sigma, Saint Louis, MI, USA, S0389) at a final concentration of 2% and Fructose was added as D-Fructose (Sigma, Saint Louis, MI, USA, 47740) at a final concentration of 2%.

We present a list of reagents detailing final concentrations in culture media and from which company they were purchased: N-Acetyl cysteine (NAC) 5 mM (Sigma, Saint Louis, MI, USA, A9165); FM4-64 30 µg/µL (Invitrogen, Waltham, MA, USA, T-3166); Rapamycin 200 ng/mL (Sigma, Saint Louis, MI, USA, R0395); ATP 200 mM (Sigma, Saint Louis, MI, USA, A1852); and Dihydroethidium (DHE) 50 µM (Sigma, Saint Louis, MI, USA, D7008). Cell cultures were exponentially grown at 600 nm [O.D600] of 0.6. Iron was added as ammonium iron (III) sulphate hexahydrate [NH4Fe(SO4)2·6H2O] (+Fe; Sigma, Saint Louis, MI, USA, F1543) at a final concentration of 10 mM.

#### *2.3. Vacuole and Dihydroethidium Staining*

For vacuole visualization, cells were stained with the fluorescent styryl dye FM4-64 (N- (3-triethylammoniumpropyl)-4-(p-diethylaminophenylhexatrienyl) pyridinium dibromide. To determine cellular oxidation, we used dihydroethidium (DHE). Both protocols were previously described by our group in [60].

#### *2.4. Cell Survival and Chronological Life Span*

To assay cell viability cells were grown to mid-log phase OD600:0.6 in SD medium supplemented with the required amino acids. Viability was registered through serial dilutions and plated by triplicate onto YPD plates.

We measured the chronological life span (CLS) in the different strains based on the survival of populations of non-dividing yeast cells according to [61]. The viability was

scored by counting the number of cells able to form colonies, CFU (colony-forming units). Cultures were started at an OD600:0.6. The same number of cells collected from each culture were plated in triplicated into YPD plates and allowed to grow at 30 ◦C for 3–4 days. CLS curves were plotted with the corresponding averages and standard deviations from three independent experiments.

#### *2.5. Protein Extraction and Immunoblot Analyses*

We followed an identical procedure as described in [23]. Total yeast protein extracts were prepared as previously described in [61]. The antibodies for Western blotting were as follows: anti-HA 3F10 (No. 12158167001; Roche Applied Science, Penzberg, Germany), used at a dilution of 1:2000 in 0.25% non-fat milk, and the corresponding secondary was goat anti-Rat IgG horseradish peroxidase conjugate (No. AP136P, Millipore, Burlington, MA, USA). Anti-GFP (No. 632381; Living Colors Mountain View, CA, USA) was used at a dilution of 1:2000 and anti-Phospho-glycerate kinase 1 (anti-PGK1) (459250, Invitrogen, Waltham, MA, USA) was used at a dilution 1:1200, both with the secondary antibody anti-Mouse horseradish peroxidase conjugate (LNA931v/AG, GE Healthcare, Chicago, IL, USA) and anti-Phospho-AMPKα (Thr172) (167253S, Cell Signalling, Danvers, MA, USA) at a dilution of 1:1000 with the secondary antibody anti-rabbit horseradish peroxidase conjugate (LNA934v/AG, GE Healthcare, Chicago, IL, USA). They were used as indicated by the manufacturers.

The protein–antibody complexes were visualized by enhanced chemiluminescence, using the Supersignal substrate (Pierce, Waltham, MA, USA) in a Chemidoc (Roche Applied Science, Penzberg, Germany).

For all the figures: We used anti-PGK1 to detect PGK1, selected as a loading control in all the Western blots shown in this study. For Western blots in this paper, we have selected representative samples.

#### *2.6. Autophagy Detection*

Autophagy progression is monitored through several complementary approaches, such as the immunological detection of GFP accumulation from GFP–Atg8 genomic fusion which is delivered to the vacuole to be degraded once autophagy is induced. The GFP moiety is very resistant to proteolysis compared to Atg8 which is rapidly degraded in the vacuole. Therefore, detection of free GFP processed from GFP–Atg8 is a very reliable tool to measure levels of complete autophagy through the autophagic flux [62], that is delivery and turnover of the cargo in the vacuole. Autophagic flux is the ratio of free GFP/GFP–Atg8+free GFP quantified upon Western blot detection by using anti-GFP antibody [23]. Another complementary approach is the microscopic observation of GFP accumulation in vacuoles. For all microscope panels, we have used a representative image of either log or one day samples in order to identify GFP–Atg8 localization. In general, both assays are sufficient as evidence of autophagy activity.

In some particular occasions we also use an alternative approach, consisting of measuring *pho8*∆60 enzymatic activity to determine nonspecific autophagy, as described by Noda and Klionsky [63] and modified by Guedes et al. [54].

#### *2.7. Glucose Determination*

We followed the directions detailed in [64].

#### *2.8. Statistical Analysis*

We followed the same procedure as described in Montella et al. [23]. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001; \*\*\* = 0.001 > *p* > 0.0001; \*\*\*\* = *p* > 0.0001.

#### **3. Results**

#### *3.1. Glucose, Amino Acids, Nitrogen and Iron Deprivation Determine the Induction of Bulk Autophagy during Diauxic Transition*

In a previous paper we demonstrated that autophagy is required for normal life span extension [23]. We wanted to determine which pathways are involved in autophagy regulation in the process of ageing. To start our analysis, we took daily samples from log phase (day 0) till day 15, following a standard CLS analysis. We used SD medium to avoid the addition of excess amino acids and thus did not affect the metabolism of the yeast cells. In Figure 1A we can observe that there is a great induction of autophagy and autophagic flux when cells reach the diauxic shift upon one day of growth, which is maintained and gradually decreases until day 6. The induction of autophagic flux is related with the phosphorylation of Atg1 protein, hence, Atg1 receives the starvation signal in order to induce autophagy during the diauxic and posdiauxic shifts, and also with the enzymatic activity determined by the *pho8*∆60 assay. Microscopic observation of the cultures confirmed the former results, since free GFP derived from GFP–Atg8 fusion protein was accumulated inside vacuoles which appear colored in green and surrounded by a red membrane stained by the fluorescent styryl dye FM4-64. In order to ascertain whether our results were compatible with bulk or selective autophagy we repeated this experiment in the mutants *atg7* (involved in general autophagy) and *atg11* (representing of selective autophagy) (Figures 1B and S1B). Our results demonstrate that free GFP liberated from GFP–Atg8 fusion protein and detected both in Western blot and in the fluorescence microscopy indicated bulk autophagy and was independent of any type of selective autophagy.

One day of culture in SD media is the transition between a fermentative to respiratory metabolism, the diauxic shift, a metabolic regulatory checkpoint determinant in the process of ageing. At this point, we could observe that glucose is nearly exhausted in the culture media (Figure 1C). Consequently, we performed refeeding experiments upon one day of culture and observed that only upon 6 h of glucose addition autophagy (determined by *pho8*∆60 specific activity, identification of free GFP by Western blot, and in vivo fluorescence identification of vacuolar accumulation of GFP) significantly decreased, concluding that a severe decrease in glucose concentration provoked the induction of autophagy, and that there is an increase in autophagy (Figures 1D and S1C). We carried out the same strategy with other nutrients which could also be limiting: amino acids, nitrogen and iron. Upon refeeding of iron, nitrogen, and amino acids, we observed that one-day refeeding did not provoke changes in autophagy (Figures 1D and S1C), however upon two days of culture we observed a clear descent in GFP accumulation caused by amino acid replenishment (Figure 1E,D). Both nitrogen and iron replenishment provoked a descent in autophagy upon two days of culture.


**Figure 1.** *Cont.*

**Figure 1.** Sequential decreases of glucose and amino acids activates bulk autophagy during the diauxic shift in *Saccharomyces cerevisiae*. (**A**) wt cultures in which the fusion protein GFP–Atg8 or Atg1HA were integrated, were grown to log phase (OD600: 0.6) in SD medium at 30 °C. Aliquots were collected at the indicated times for total protein extraction and Western blot analysis. GFP–Atg8 was monitored using an anti-GFP antibody. We used anti-PGK1 to detect Pgk1 as a loading control. Microscopic observation of GFP–Atg8 was carried out using a fluorescence microscope. GFP vacuolar accumulation was also determined upon the use of the fluorescent dye FM4-64. Autophagic flux was calculated as the ratio of free GFP and total GFP–Atg8 in the samples. Total GFP–Atg8 was determined as the addition of the form GFP– Atg8 and the band corresponding to free GFP, as a result of Atg8 vacuolar degradation, both detected by Western blot. Enzymatic autophagy activity was measured by using the alkaline phosphatase assay in the strain BY4741*pho8* expressing a plasmid with the inactive Pho8 proenzyme targeted to the cytosol. Values of Atg1 proteins were determined upon Western blot analysis using anti-HA antibody. (**B**) *atg7* and *atg11* strains expressing the fusion protein GFP–Atg8 were grown in the same conditions as described in A. Autophagic flux and total Atg8 expression were determined as in (**A**). (**C**) Glucose content in the culture medium (%) was determined in wt cultures growing in SD media at 30 °C at the days indicated in the table for a total period of 15 days. Glucose in the sterile media, SD, at a final concentration of 2%, is the equivalent of 100% in the table. (**D**) wt cells bearing GFP–Atg8 in their genome were exponentially grown at OD600:0.6 at 30 °C in SD media and a sample was collected for analysis. Upon one day of culture, 2% glucose, amino acids (60 mg/mL leucine, 20 mg/mL histidine, and 20 mg/mL tryptophan), 0.67% nitrogen or 10 mM iron, were respectively added to the **Figure 1.** Sequential decreases of glucose and amino acids activates bulk autophagy during the diauxic shift in *Saccharomyces cerevisiae*. (**A**) wt cultures in which the fusion protein GFP–Atg8 or Atg1HA were integrated, were grown to log phase (OD600: 0.6) in SD medium at 30 ◦C. Aliquots were collected at the indicated times for total protein extraction and Western blot analysis. GFP–Atg8 was monitored using an anti-GFP antibody. We used anti-PGK1 to detect Pgk1 as a loading control. Microscopic observation of GFP–Atg8 was carried out using a fluorescence microscope. GFP vacuolar accumulation was also determined upon the use of the fluorescent dye FM4-64. Autophagic flux was calculated as the ratio of free GFP and total GFP–Atg8 in the samples. Total GFP–Atg8 was determined as the addition of the form GFP–Atg8 and the band corresponding to free GFP, as a result of Atg8 vacuolar degradation, both detected by Western blot. Enzymatic autophagy activity was measured by using the alkaline phosphatase assay in the strain BY4741*pho8*∆ expressing a plasmid with the inactive Pho8 proenzyme targeted to the cytosol. Values of Atg1 proteins were determined upon Western blot analysis using anti-HA antibody. (**B**) *atg7* and *atg11* strains expressing the fusion protein GFP–Atg8 were grown in the same conditions as described in A. Autophagic flux and total Atg8 expression were determined as in (**A**). (**C**) Glucose content in the culture medium (%) was determined in wt cultures growing in SD media at 30 ◦C at the days indicated in the table for a total period of 15 days. Glucose in the sterile media, SD, at a final concentration of 2%, is the equivalent of 100% in the table. (**D**) wt cells bearing GFP–Atg8 in their genome were exponentially grown at OD600:0.6 at 30 ◦C in SD media and a sample was collected for analysis. Upon one day of culture, 2% glucose, amino acids (60 mg/mL leucine, 20 mg/mL histidine, and 20 mg/mL tryptophan), 0.67% nitrogen or 10 mM iron, were respectively added to the cultures and samples were collected at 2, 6, and 24 h to perform *pho8*∆60 enzymatic assays. Autophagic flux and GFP–Atg8 total expression were determined as previously detailed in A. Enzymatic autophagic activity was determined as in (**A**). (**E**) As in (**D**), but results correspond to two days of growth. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001; \*\*\* = 0.001 > *p* > 0.0001; \*\*\*\* = *p* > 0.0001.

0.0001.

#### *3.2. Both Mtl1 and Gcn2 Control Autophagy Induction during Diauxic Transition 3.2. Both Mtl1 and Gcn2 Control Autophagy Induction during Diauxic Transition*

*J. Fungi* **2021**, *7*, x FOR PEER REVIEW 9 of 26

cultures and samples were collected at 2, 6, and 24 h to perform *pho8*60 enzymatic assays. Autophagic flux and GFP– Atg8 total expression were determined as previously detailed in A. Enzymatic autophagic activity was determined as in (**A**). (**E**) As in (**D**), but results correspond to two days of growth. Error bars in the histograms represent the standard

a Student's unpaired *t*-test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001; \*\*\* = 0.001 *p* 0.0001; \*\*\*\* = *p*

In this context, we decided to explore the signaling pathways that could be involved in signaling autophagy when cells age. In this context, we decided to explore the signaling pathways that could be involved in signaling autophagy when cells age.

We considered the possibility that amino acid and nitrogen depletion would provoke TORC1 inactivation at least partly. We analyzed TOR function by means of the readouts Rtg1, Sfp1, Msn2. Under TORC1 inactivation, Rtg1 and Msn2 were localized to the nucleus, whereas Sfp1 was located in the cytoplasm, confirming that Tor1 is not inactivated in our model (Figure 2A). In addition, we added rapamycin, a drug that specifically inactivates TORC1, to the previous cultures as a control, and we observed that Rtg1 and Msn2 localized to the nucleus and Sfp1 was located in the cytoplasm (Figure 2B). Moreover, rapamycin treatment caused a decrease in Atg13 phosphorylation and autophagy values calculated upon Pho8 assay did not increase, supporting the conclusion that TORC1 is not completely inactivated upon diauxic shift in the conditions of our study (Figure 2C). We considered the possibility that amino acid and nitrogen depletion would provoke TORC1 inactivation at least partly. We analyzed TOR function by means of the readouts Rtg1, Sfp1, Msn2. Under TORC1 inactivation, Rtg1 and Msn2 were localized to the nucleus, whereas Sfp1 was located in the cytoplasm, confirming that Tor1 is not inactivated in our model (Figure 2A). In addition, we added rapamycin, a drug that specifically inactivates TORC1, to the previous cultures as a control, and we observed that Rtg1 and Msn2 localized to the nucleus and Sfp1 was located in the cytoplasm (Figure 2B). Moreover, rapamycin treatment caused a decrease in Atg13 phosphorylation and autophagy values calculated upon Pho8 assay did not increase, supporting the conclusion that TORC1 is not completely inactivated upon diauxic shift in the conditions of our study (Figure 2C).

**Figure 2.** TORC1 is not inactivated during diauxic and posdiauxic shifts. (**A**) wt strains transformed with the plasmids Rtg1GFP, Sfp1GFP, or Msn2GFP respectively, were grown at 30 ◦C in SD media for the times indicated in the figures. Aliquots were collected for in vivo observation in the fluorescence microscope. Histograms represent the percentages of in vivo nuclear or cytoplasmic localization out of more than 1000 cells. (**B**) Cultures in A were treated with rapamycin (200 ng/mL) on day one of culture for 6 h and aliquots were collected for in vivo observation in the fluorescence microscope. Histograms are performed as in (**A**). (**C**) A wt strain expressing Atg13HA was exponentially grown at 30 ◦C in SD media. Rapamycin was added to the culture upon one day of growth at 200 ng/mL and samples were collected upon 6 and 24 h of exposure to the drug for total protein extraction, Western blot analysis, and identification of Atg13HA using anti-HA antibody. Autophagic enzymatic activity was determined through the alkaline phosphatase assay, as in (Figure 1A). Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001; \*\*\* = 0.001 > *p* > 0.0001.

However, when *TOR1* was deleted we observed that autophagy (free GFP detected in the Western blot and fluorescence microscope accumulated in vacuoles) was extended to longer times (Figures 3A and S2A), concomitant with longer life extension as described in [54]. Taken together, we conclude that TORC1 is not inactivated in our system during the transition between fermentative and respiratory metabolism, therefore it is not the main pathway responsible for the bulk autophagy induction. concentration during the diauxic shift and to transmission of this signal to the autophagy machinery. We have also detected similar results when using alternative and fermentative carbon sources such as sucrose or fructose (Figure S2F).

*J. Fungi* **2021**, *7*, x FOR PEER REVIEW 11 of 26

**Figure 3.** *Cont.*

**Figure 3.** Mtl1 and Gcn2 control autophagy induction during glucose and amino acid starvation. Growth conditions, total Atg8 determination and autophagic flux determined in: (**A**) *tor1* mutant expressing GFP–Atg8; (**B**) wt strain and *ras2* mutant expressing GFP–Atg8; and (**C**) *gcn2* mutant expressing GFP–Atg8, performed as described in Figure 1B. (**D**) *gcn2*  bearing GFP–Atg8 was exponentially grown at 30 °C in SD plus amino acids. Rapamycin (200 ng/mL) was added to the cultures upon 1 day of growth and samples were subsequently collected upon 2, 6 and 24 h of exposure to the drug. Aliquots were treated as in Figure 1B. (**E**) Growth conditions, total Atg8 and autophagic flux determinations in *mtl1* culture expressing the fusion protein GFP–Atg8 was performed as in Figure 1B. Identification of Atg1 protein in *mtl1* cultures transformed with the plasmid Atg1HA was performed upon Western blot analysis using the anti-HA antibody as in Figure 1A. (**F**) Percentage of Atg8 foci quantified in the experiments, described in Figures 1A and 3E, was calculated upon microscopic observation of more than 1000 cells. The axis label "% of GFP–Atg8 foci" refers to the percentage of cells with GFP– Atg8 foci. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p-*values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001. **Figure 3.** Mtl1 and Gcn2 control autophagy induction during glucose and amino acid starvation. Growth conditions, total Atg8 determination and autophagic flux determined in: (**A**) *tor1* mutant expressing GFP–Atg8; (**B**) wt strain and *ras2* mutant expressing GFP–Atg8; and (**C**) *gcn2* mutant expressing GFP–Atg8, performed as described in Figure 1B. (**D**) *gcn2* bearing GFP–Atg8 was exponentially grown at 30 ◦C in SD plus amino acids. Rapamycin (200 ng/mL) was added to the cultures upon 1 day of growth and samples were subsequently collected upon 2, 6 and 24 h of exposure to the drug. Aliquots were treated as in Figure 1B. (**E**) Growth conditions, total Atg8 and autophagic flux determinations in *mtl1* culture expressing the fusion protein GFP–Atg8 was performed as in Figure 1B. Identification of Atg1 protein in *mtl1* cultures transformed with the plasmid Atg1HA was performed upon Western blot analysis using the anti-HA antibody as in Figure 1A. (**F**) Percentage of Atg8 foci quantified in the experiments, described in Figures 1A and 3E, was calculated upon microscopic observation of more than 1000 cells. The axis label "% of GFP–Atg8 foci" refers to the percentage of cells with GFP–Atg8 foci. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001.

> *3.3. Mtl1 CWI Cell Wall Receptor Signals Glucose Concentration to the Autophagy Machinery in a Manner Partly Dependent on Intracellular ATP Levels* In previous reports it has been demonstrated that several nutritional stressors (nitrogen, amino acids, iron…) cause the induction of autophagy. In order to analyze the specificity that Mtl1 could play in macroautophagy regulation, we used different nutrient concentrations: (glucose: 0.5%, 0.1%, 0.05% and 0%; amino acids: 0.1% and 0%; nitrogen: 0.06%, 0.01% and 0%; and iron: 0%). The decrease of each of the nutrients, glucose, amino acids, iron, or nitrogen concentrations induced the activation of macroautophagy in wt cells (Figures 4A and S3A) along with corresponding Atg1 phosphorylation (Figure 4D). In addition, the absence of Gcn2 precluded autophagy in a manner only dependent on amino acid availability (Figures 4B and S3B), whereas the absence of Mtl1 specifically abolished the glucose deprivation dependent autophagy (Figures 4C and S3C), supported by a lack of Atg1 phosphorylation (Figure 4D). We demonstrated that glucose concentrations below 0.5% caused a clear induction of autophagy specifically mediated by Mtl1, as in *mtl1* mutants autophagy was not induced. We also explored a *ras2* mutant, since Ras2 is active in exponentially growing cells and becomes inactive as long as cells enter in respiratory metabolism and glucose becomes exhausted. *RAS2* deletion partially affected autophagy progression as compared to wt cultures (Figures 3B and S2B). Since other nutrients become depleted upon the diauxic shift, we took into consideration the Gcn2/eIF2alpha pathway which becomes activated under amino acid and other nutrient starvation [65]. The Gcn2 pathway is activated upon diauxic shifting, indicating the moment in which amino acid concentrations significantly decreased in the culture media. According to our results, and coincident with the above mentioned replenishment results, day 2 of growth should be the moment in which cells become starved for amino acids. We observed that upon the second day, autophagy disappears when Gcn2 is deleted, however the absence of Gcn2 the burst in autophagy observed upon 1 day of growth was not affected (Figures 3C and S2C). This result suggests that Gcn2 is required to induce macroautophagy upon two days of growth in SD medium, after the diauxic shift, probably due to a descent in the amino acid concentration one day after glucose starvation, suggesting that Gcn2 is not involved in autophagy signaling in response to carbon sources. Following this observation, we wanted to ascertain whether the regulatory function mediated by Gcn2 was dependent on TORC1. We treated *gcn2* mutant cultures on day 1 with rapamycin and observed induction of autophagy, suggesting that in the transition from fermentative to respiratory metabolism, TORC1 and *GCN2* are independent (Figures 3D and S2D). As expected, this conclusion is consistent with the previous observation that TORC1 function is not inactivated during the CLS experiment; as a consequence of that, this pathway is not relevant for autophagy induction during ageing in our experimental conditions.

Our results suggest that glucose is the principal nutrient in the culture medium we use, whose decrease causes autophagy induction during the diauxic shift (1 day of growth) therefore, we decided to analyze in more depth the role that Mtl1 could be playing in this process. Mtl1 is a cell wall receptor belonging to the CWI pathway, involved in glucose signaling during the diauxic shift and stationary phase [28,30]. Interestingly, in the absence of Mtl1, autophagy is undetectable by Western blot, with Atg1HA phosphorylation or in vivo GFP–Atg8 microscopic accumulation through all experiments, from day 1 to 15 (Figures 3E and S2E). In yeast, autophagy is initiated when the pre-autophagosome (PAS) structure is formed [66]. PAS can be detected in the fluorescence microscope as dotted

accumulations of Atg proteins next to the vacuole. In *mtl1* diauxic cultures, PAS can be detected, as also observed in wt strain, and they are significantly higher than in the *atg1* mutant (Figure 3F), suggesting that Mtl1 does not block the initiation of the autophagy complex. This result suggests that Mtl1 is essential for receiving the signal of glucose concentration during the diauxic shift and to transmission of this signal to the autophagy machinery. We have also detected similar results when using alternative and fermentative carbon sources such as sucrose or fructose (Figure S2F).

#### *3.3. Mtl1 CWI Cell Wall Receptor Signals Glucose Concentration to the Autophagy Machinery in a Manner Partly Dependent on Intracellular ATP Levels*

In previous reports it has been demonstrated that several nutritional stressors (nitrogen, amino acids, iron . . . ) cause the induction of autophagy. In order to analyze the specificity that Mtl1 could play in macroautophagy regulation, we used different nutrient concentrations: (glucose: 0.5%, 0.1%, 0.05% and 0%; amino acids: 0.1% and 0%; nitrogen: 0.06%, 0.01% and 0%; and iron: 0%). The decrease of each of the nutrients, glucose, amino acids, iron, or nitrogen concentrations induced the activation of macroautophagy in wt cells (Figures 4A and S3A) along with corresponding Atg1 phosphorylation (Figure 4D). In addition, the absence of Gcn2 precluded autophagy in a manner only dependent on amino acid availability (Figures 4B and S3B), whereas the absence of Mtl1 specifically abolished the glucose deprivation dependent autophagy (Figures 4C and S3C), supported by a lack of Atg1 phosphorylation (Figure 4D). We demonstrated that glucose concentrations below 0.5% caused a clear induction of autophagy specifically mediated by Mtl1, as in *mtl1* mutants autophagy was not induced. *J. Fungi* **2021**, *7*, x FOR PEER REVIEW 13 of 26

**Figure 4.** *Cont.*

**Figure 4.** Mtl1 signals glucose limitation to autophagy machinery. (**A**) wt cells expressing GFP–Atg8 were exponentially grown in SD media. Aliquots were taken, washed, and transferred to different minimum media containing: 0.5, 0.1, or 0.05% glucose; 0.1 or 0% amino acids; 0.06, 0.01, or 0% nitrogen or medium without iron (0%). Autophagic flux and Atg8 expression were determined as in Figure 1B. The same experiments as in A were carried out in (**B**) *gcn2* mutant cultures expressing GFP–Atg8 and in a (**C**) *mtl1* strain expressing GFP–Atg8. (**D**) Atg1HA protein was identified by Western blot using anti-HA antibody, as in Figure 1A. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p-*values from a Student's unpaired *t*-

> However, placing cells at 0% glucose, we did not observe free GFP in either the Western blot or accumulated in vacuoles. Hence, bulk autophagy was not induced in wt or

test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001; \*\*\* = 0.001 *p* 0.0001.

**Figure 4.** Mtl1 signals glucose limitation to autophagy machinery. (**A**) wt cells expressing GFP–Atg8 were exponentially grown in SD media. Aliquots were taken, washed, and transferred to different minimum media containing: 0.5, 0.1, or 0.05% glucose; 0.1 or 0% amino acids; 0.06, 0.01, or 0% nitrogen or medium without iron (0%). Autophagic flux and Atg8 expression were determined as in Figure 1B. The same experiments as in A were carried out in (**B**) *gcn2* mutant cultures expressing GFP–Atg8 and in a (**C**) *mtl1* strain expressing GFP–Atg8. (**D**) Atg1HA protein was identified by Western blot using anti-HA antibody, as in Figure 1A. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p-*values from a Student's unpaired *t*test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001; \*\*\* = 0.001 *p* 0.0001. **Figure 4.** Mtl1 signals glucose limitation to autophagy machinery. (**A**) wt cells expressing GFP–Atg8 were exponentially grown in SD media. Aliquots were taken, washed, and transferred to different minimum media containing: 0.5, 0.1, or 0.05% glucose; 0.1 or 0% amino acids; 0.06, 0.01, or 0% nitrogen or medium without iron (0%). Autophagic flux and Atg8 expression were determined as in Figure 1B. The same experiments as in A were carried out in (**B**) *gcn2* mutant cultures expressing GFP–Atg8 and in a (**C**) *mtl1* strain expressing GFP–Atg8. (**D**) Atg1HA protein was identified by Western blot using anti-HA antibody, as in Figure 1A. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001; \*\*\* = 0.001 > *p* > 0.0001; \*\*\*\* = *p* > 0.0001.

However, placing cells at 0% glucose, we did not observe free GFP in either the Western blot or accumulated in vacuoles. Hence, bulk autophagy was not induced in wt or However, placing cells at 0% glucose, we did not observe free GFP in either the Western blot or accumulated in vacuoles. Hence, bulk autophagy was not induced in wt or *mtl1* strains, as previously described by [67] (Figures 5A and S4A). These authors attributed this result to the sudden lack of ATP required for the autophagy machinery. We added ATP to wt and *mtl1* cultures completely depleted of glucose and observed that whereas in wt cultures autophagy induction was high, in the *mtl1* mutant the autophagy response was only partially restored (Figures 5B and S4B). Identical results were obtained during the diauxic shift in *mtl1* cultures when ATP was added to exponentially growing cells (Figures 5C and S4C). These results led us to the conclusion that the absence of *MTL1* provoked ATP starvation when glucose concentration drops below a threshold. In the former experiment, we can observe that autophagy induction occurs when glucose concentration drops from 2% to 0.5% in wt cultures (Figures 4A and S3A). However, in *mtl1* cultures, any decrease below 2% aborted autophagy (Figures 4C and S3C).

In order to ascertain the contribution of mitochondrial ATP to autophagy response during the diauxic shift and a reduction in glucose concentrations in exponentially growing cultures, we analyzed a *rho0* mutant lacking mitochondrial DNA. We observed that the absence of mitochondrial DNA did not preclude the induction of bulk autophagy upon one day of culture (Figures 5D and S4D) and upon glucose concentration reduction (Figures 5E and S5E). These results suggest that the functional role that Mtl1 plays in the autophagy response to glucose availability is not only linked to ATP accumulation.

*mtl1* strains, as previously described by [67] (Figures 5A and S4A). These authors attributed this result to the sudden lack of ATP required for the autophagy machinery. We added ATP to wt and *mtl1* cultures completely depleted of glucose and observed that whereas in wt cultures autophagy induction was high, in the *mtl1* mutant the autophagy response was only partially restored (Figures 5B and S4B). Identical results were obtained during the diauxic shift in *mtl1* cultures when ATP was added to exponentially growing cells (Figures 5C and S4C). These results led us to the conclusion that the absence of *MTL1* provoked ATP starvation when glucose concentration drops below a threshold. In the former experiment, we can observe that autophagy induction occurs when glucose concentration drops from 2% to 0.5% in wt cultures (Figures 4A and S3A). However, in *mtl1* 

In order to ascertain the contribution of mitochondrial ATP to autophagy response during the diauxic shift and a reduction in glucose concentrations in exponentially growing cultures, we analyzed a *rho0* mutant lacking mitochondrial DNA. We observed that the absence of mitochondrial DNA did not preclude the induction of bulk autophagy upon one day of culture (Figures 5D and S4D) and upon glucose concentration reduction (Figures 5E and S5E). These results suggest that the functional role that Mtl1 plays in the autophagy response to glucose availability is not only linked to ATP accumulation.

cultures, any decrease below 2% aborted autophagy (Figures 4C and S3C).

**Figure 5.** Mtl1 signals a decrease in glucose concentration to the autophagy machinery in a manner not fully dependent on ATP production by mitochondria. (**A**) wt and *mtl1* cells expressing GFP–Atg8 grown in SD media were transferred to minimum media devoid of glucose (0% glucose) to determine autophagic flux and Atg8 expression as in Figure 1B. (**B**) Upon transference to minimum media without glucose, ATP (at 200 mM final concentration) was added to the cultures described in (**A**) and samples were collected at the indicated times for determination of autophagy as in (**A**). (**C**) ATP (200 mM) was added to *mtl1* cultures growing in SD minimum medium for one day at the diauxic shift, and samples were collected at 4 and 8 h for autophagic flux and Atg8 expression determination, as in Figure 1B. (**D**) A *rho0* mutant expressing GFP–Atg8 was grown (as in Figure 1B) for autophagy determination. (**E**) *rho0* cells exponentially growing in SD were washed and transferred to several media containing different glucose concentrations to analyze autophagy, as in Figure 4A. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p-*values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001. *3.4. Both Ras2 or Sch9 Suppress mtl1 Deficiency in Bulk Autophagy Activation in all Metabolic*  **Figure 5.** Mtl1 signals a decrease in glucose concentration to the autophagy machinery in a manner not fully dependent on ATP production by mitochondria. (**A**) wt and *mtl1* cells expressing GFP–Atg8 grown in SD media were transferred to minimum media devoid of glucose (0% glucose) to determine autophagic flux and Atg8 expression as in Figure 1B. (**B**) Upon transference to minimum media without glucose, ATP (at 200 mM final concentration) was added to the cultures described in (**A**) and samples were collected at the indicated times for determination of autophagy as in (**A**). (**C**) ATP (200 mM) was added to *mtl1* cultures growing in SD minimum medium for one day at the diauxic shift, and samples were collected at 4 and 8 h for autophagic flux and Atg8 expression determination, as in Figure 1B. (**D**) A *rho0* mutant expressing GFP–Atg8 was grown (as in Figure 1B) for autophagy determination. (**E**) *rho0* cells exponentially growing in SD were washed and transferred to several media containing different glucose concentrations to analyze autophagy, as in Figure 4A. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001.

PKA during the diauxic shift or upon glucose depletion [30].

We next tried to identify the pathway or pathways with which Mtl1 is connected to

We demonstrated that Mtl1 mediates Bcy1 activating phosphorylation through TORC1 downregulation thus leading to PKA activation [30]. Therefore, and taking into consideration the hypothesis that *mtl1* mutants could cause impairment of the glucose signal through PKA, we also analyzed the overexpression of Bcy1, the PKA inhibitor, in both wt and *mtl1* strains. Overexpression of Bcy1 prolonged autophagy induction in wt cultures for longer times (Figures 6A and S5A) as compared to the wt empty strain (Figures 1A and S1A). However, Bcy1 overexpression did not restore autophagy in the *mtl1* mutant (Figures 3E, 6A, S2E and S5A). In previous papers, we observed clear impairment of Slt2 phosphorylation upon both oxidative stress and glucose deprivation in *mtl1* mutants [30]. Consistent with this information, we used a plasmid overexpressing the Pkc1 protein and a second plasmid bearing the *BCK1-20* allele which keeps Slt2 kinase constitutively activated, as Bck1 is the MAPKKK of the CWI pathway. Results depicted in Figures 3E, 6A, S2E and S5A demonstrate that the lack of bulk autophagy activation observed in the *mtl1* mutant, as a result of a decrease in glucose concentration during diauxic

found that Mtl1 is negatively related to both Tor1 and Ras2 in response to oxidative stress and glucose starvation [53]. In addition, Mtl1 is also negatively related to Sch9, Slt2, and

*Conditions that Imply Reduced Glucose Levels*

#### *3.4. Both Ras2 or Sch9 Suppress mtl1 Deficiency in Bulk Autophagy Activation in All Metabolic Conditions That Imply Reduced Glucose Levels*

We next tried to identify the pathway or pathways with which Mtl1 is connected to the signaling process converging on the autophagy machinery. In a previous study, we found that Mtl1 is negatively related to both Tor1 and Ras2 in response to oxidative stress and glucose starvation [53]. In addition, Mtl1 is also negatively related to Sch9, Slt2, and PKA during the diauxic shift or upon glucose depletion [30].

We demonstrated that Mtl1 mediates Bcy1 activating phosphorylation through TORC1 downregulation thus leading to PKA activation [30]. Therefore, and taking into consideration the hypothesis that *mtl1* mutants could cause impairment of the glucose signal through PKA, we also analyzed the overexpression of Bcy1, the PKA inhibitor, in both wt and *mtl1* strains. Overexpression of Bcy1 prolonged autophagy induction in wt cultures for longer times (Figures 6A and S5A) as compared to the wt empty strain (Figures 1A and S1A). However, Bcy1 overexpression did not restore autophagy in the *mtl1* mutant (Figures 3E, 6A, S2E and S5A). In previous papers, we observed clear impairment of Slt2 phosphorylation upon both oxidative stress and glucose deprivation in *mtl1* mutants [30]. Consistent with this information, we used a plasmid overexpressing the Pkc1 protein and a second plasmid bearing the *BCK1-20* allele which keeps Slt2 kinase constitutively activated, as Bck1 is the MAPKKK of the CWI pathway. Results depicted in Figure 3E, Figure 6A, Figures S2E and S5A demonstrate that the lack of bulk autophagy activation observed in the *mtl1* mutant, as a result of a decrease in glucose concentration during diauxic transition, is not caused by the lack of Slt2 kinase activity, since neither Pkc1 overexpression nor the *BCK1-20* allele restored the lack of autophagy induction in the *mtl1* mutant. We next decided to check Ras2 deletion in *mtl1*, since the GTPase is activated in the presence of glucose and is responsible for the synthesis of cAMP when glucose is the carbon source. Ras2 deletion in the *mtl1* mutant provoked the activation of bulk autophagy during the diauxic shift; in fact, the distribution levels of autophagy were equivalent between *ras2* and *mtl1ras2* (Figure 3B, Figure 6A, Figures S2B and S5A respectively) strains, suggesting that Mtl1 signals to Ras2 inactivation upon glucose starvation and diauxic transition signaling to activate bulk autophagy. Lastly, given the association between Mtl1 and Sch9 kinase [30], we also analyzed whether both proteins were also related to autophagy signaling. Deletion of *SCH9* restored autophagy in the *mtl1* mutant during the diauxic shift (Figures 6A and S5A). This result suggests a connection of Mtl1 with Sch9 towards autophagy. In order to corroborate the glucose specificity of these responses, exponentially grown cultures of *mtl1ras2* and *mtl1sch9* along with the corresponding controls were assayed for bulk autophagy response upon glucose starvation (Figures 6B and S5B). As expected, deletion of *RAS2* or *SCH9* suppressed the lack of autophagy induction in the absence of *MTL1*.

Snf1 is an AMPK family member which is highly conserved in eukaryotes. When glucose is exhausted at the beginning of the diauxic shift, Snf1 becomes activated to trigger a wide response of regulating activators and repressors to trigger respiratory metabolism (see review [21]). In order to detect a possible defect in Snf1 activation in the *mtl1* mutant, we analyzed Snf1 phosphorylation in Thr20 residues on the activation loop of the catalytic subunit in samples of wt*, mtl1, ras2, ras2mtl1, sch9* and *mtl1sch9*. In Figure 6C, it can be observed that Snf1 is correctly and similarly phosphorylated in all strains, concluding that *mtl1* defects in autophagy during the transition to stationary phase and upon glucose depletion are not a principal consequence of the lack of Snf1 activation as a response to glucose limitation.

In summary, our results suggest that either *RAS2* or *SCH9* deletion reverted the lack of autophagy in the *mtl1* mutant, suggesting that Mtl1 receives the signal of decreased glucose concentration and connects to both Ras2 and Sch9 inactivation, a mechanism that converges on macroautophagy induction. We have also observed similar results when the carbon source is either sucrose or fructose (Figure S2F).

carbon source is either sucrose or fructose (Figure S2F).

to glucose limitation.

transition, is not caused by the lack of Slt2 kinase activity, since neither Pkc1 overexpression nor the *BCK1-20* allele restored the lack of autophagy induction in the *mtl1* mutant. We next decided to check Ras2 deletion in *mtl1*, since the GTPase is activated in the presence of glucose and is responsible for the synthesis of cAMP when glucose is the carbon source. Ras2 deletion in the *mtl1* mutant provoked the activation of bulk autophagy during the diauxic shift; in fact, the distribution levels of autophagy were equivalent between *ras2* and *mtl1ras2* (Figures 3B, 6A, S2B and S5A respectively) strains, suggesting that Mtl1 signals to Ras2 inactivation upon glucose starvation and diauxic transition signaling to activate bulk autophagy. Lastly, given the association between Mtl1 and Sch9 kinase [30], we also analyzed whether both proteins were also related to autophagy signaling. Deletion of *SCH9* restored autophagy in the *mtl1* mutant during the diauxic shift (Figures 6A and S5A). This result suggests a connection of Mtl1 with Sch9 towards autophagy. In order to corroborate the glucose specificity of these responses, exponentially grown cultures of *mtl1ras2* and *mtl1sch9* along with the corresponding controls were assayed for bulk autophagy response upon glucose starvation (Figures 6B and S5B). As expected, deletion of *RAS2* or *SCH9* suppressed the lack of autophagy induction in the absence of *MTL1*.

Snf1 is an AMPK family member which is highly conserved in eukaryotes. When glucose is exhausted at the beginning of the diauxic shift, Snf1 becomes activated to trigger a wide response of regulating activators and repressors to trigger respiratory metabolism (see review [21]). In order to detect a possible defect in Snf1 activation in the *mtl1* mutant, we analyzed Snf1 phosphorylation in Thr20 residues on the activation loop of the catalytic subunit in samples of wt*, mtl1, ras2, ras2mtl1, sch9* and *mtl1sch9*. In Figure 6C, it can be observed that Snf1 is correctly and similarly phosphorylated in all strains, concluding that *mtl1* defects in autophagy during the transition to stationary phase and upon glucose depletion are not a principal consequence of the lack of Snf1 activation as a response

In summary, our results suggest that either *RAS2* or *SCH9* deletion reverted the lack of autophagy in the *mtl1* mutant, suggesting that Mtl1 receives the signal of decreased glucose concentration and connects to both Ras2 and Sch9 inactivation, a mechanism that converges on macroautophagy induction. We have also observed similar results when the

**Figure 6.** Both Ras2 and Sch9 suppress *mtl1* deficiency in autophagy signaling upon glucose concentration decreasing. **(A**) *mtl1ras2, sch9, mtl1sch9, slt2, mtl1slt2,* wt+pBcy1, *mtl1*+pBcy1, wt+p*BCK1-20*, *mtl1*+p*BCK1-20*, wt+pPkc1\* and *mtl1*+pPkc1\* strains expressing GFP–Atg8 were grown at 30°C in SD media for 15 days. Samples were taken to determine autophagic flux and total Atg8 expression as described in Figure 1B. (**B**) Strains *ras2, mtl1ras2, sch9,* and *mtl1sch9* expressing GFP–Atg8 were exponentially grown in SD media to be subsequently transferred to minimum medium containing the indicated concentration of glucose. Samples were taken to determine autophagy as in Figure 4A. (**C**) wt samples (from Figure 4A) and *mtl1* samples (from Figure 4C) along with *mtl1, ras2, mtl1ras2, sch9* and *mtl1sch9* samples (from Figure 6B) were used for Western blot analysis and AMPK1 detection by using anti-AMPK1-P antibody. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p-*values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001. **Figure 6.** Both Ras2 and Sch9 suppress *mtl1* deficiency in autophagy signaling upon glucose concentration decreasing. (**A**) *mtl1ras2, sch9, mtl1sch9, slt2, mtl1slt2,* wt+pBcy1, *mtl1*+pBcy1, wt+p*BCK1-20*, *mtl1*+p*BCK1-20*, wt+pPkc1\* and *mtl1*+pPkc1\* strains expressing GFP–Atg8 were grown at 30 ◦C in SD media for 15 days. Samples were taken to determine autophagic flux and total Atg8 expression as described in Figure 1B. (**B**) Strains *ras2, mtl1ras2, sch9,* and *mtl1sch9* expressing GFP–Atg8 were exponentially grown in SD media to be subsequently transferred to minimum medium containingthe indicated concentration of glucose. Samples were taken to determine autophagy as in Figure 4A. (**C**) wt samples (from Figure 4A) and *mtl1* samples (from Figure 4C) along with *mtl1, ras2, mtl1ras2, sch9* and *mtl1sch9* samples (from Figure 6B) were used for Western blot analysis and AMPK1 detection by using anti-AMPK1-P antibody. Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001.

*3.5. Mtl1 is Required for Mitochondrial Degradation Dependent on Atg33 and Independent of* 

We wondered whether Mtl1 involvement in autophagy regulation would be related to any carbon source, not only to glucose availability. To answer this question, we decided to analyze a non-fermentative carbon source, glycerol, that forces cells to directly enter into respiratory metabolism. Mtl1 was clearly not involved in the detection of glycerol

similar levels and patterns of autophagy (Figures 7A and S6A). In a previous paper, we described that Mtl1 presented uncoupled respiration that provoked mitochondrial dysfunction and ROS accumulation [30]. In order to ascertain whether oxidative stress would be the cause of the autophagy problem, we added the antioxidant NAC to both wt and *mtl1* diauxic cultures (Figures 7B and S6B). In order to demonstrate that NAC was exerting its antioxidant function, samples were collected and stained with dihydroethidium (DHE) for in vivo visualization of cellular oxidation in the fluorescent microscope (Figure S6C). Our results indicate that oxidative stress is not the cause of autophagy impairment during

diauxic shift in the *mtl1* mutant (Figures 7B and S6B,C).

*Atg32 during Chronological Ageing*

## *3.5. Mtl1 Is Required for Mitochondrial Degradation Dependent on Atg33 and Independent of Atg32 during Chronological Ageing*

We wondered whether Mtl1 involvement in autophagy regulation would be related to any carbon source, not only to glucose availability. To answer this question, we decided to analyze a non-fermentative carbon source, glycerol, that forces cells to directly enter into respiratory metabolism. Mtl1 was clearly not involved in the detection of glycerol concentration linked to autophagy activity, since in both wt and *mtl1* cells we detected similar levels and patterns of autophagy (Figures 7A and S6A). In a previous paper, we described that Mtl1 presented uncoupled respiration that provoked mitochondrial dysfunction and ROS accumulation [30]. In order to ascertain whether oxidative stress would be the cause of the autophagy problem, we added the antioxidant NAC to both wt and *mtl1* diauxic cultures (Figures 7B and S6B). In order to demonstrate that NAC was exerting its antioxidant function, samples were collected and stained with dihydroethidium (DHE) for in vivo visualization of cellular oxidation in the fluorescent microscope (Figure S6C). Our results indicate that oxidative stress is not the cause of autophagy impairment during diauxic shift in the *mtl1* mutant (Figures 7B and S6B,C). *J. Fungi* **2021**, *7*, x FOR PEER REVIEW 18 of 26

**Figure 7.** Mtl1 is not deficient in bulk autophagy in respiratory conditions. (**A**) wt and *mtl1* cultures were grown in minimum medium SGly (containing glycerol as unique carbon source) plus amino acids at 30 °C to stationary phase for 6 days. Samples were collected at the indicated times to identify macroautophagy as described in Figure 1B. (**B**) wt and *mtl1* cultures in SD medium growing to 1 day were treated with N-Acetyl cysteine (NAC) 5 mM for 8 and 11 h. Samples were collected for autophagy determinations as in (**A**). Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001. **Figure 7.** Mtl1 is not deficient in bulk autophagy in respiratory conditions. (**A**) wt and *mtl1* cultures were grown in minimum medium SGly (containing glycerol as unique carbon source) plus amino acids at 30 ◦C to stationary phase for 6 days. Samples were collected at the indicated times to identify macroautophagy as described in Figure 1B. (**B**) wt and *mtl1* cultures in SD medium growing to 1 day were treated with N-Acetyl cysteine (NAC) 5 mM for 8 and 11 h. Samples were collected for autophagy determinations as in (**A**). Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001.

There is an alternative possibility, that if the problem of *mtl1* is mitochondrial function, we would expect to detect severe deficiencies in mitophagy. For this purpose, we analyzed mitophagy in cells expressing a fusion of mitochondrial matrix protein to GFP, Idp1–GFP, transformed in either wt or *mtl1* strains [68]. We monitored the vacuole clipping of the fusion protein Idp1–GFP. The identification of free GFP with anti-GFP antibody in a Western blot would reveal the existence of mitophagy, since GFP is very resistant to degradation [69]. Mitophagy studies are usually carried out in respiratory carbon sources or alternatively in stationary cultures. When wt and *mtl1* cells were grown in SGly to the stationary phase, we observed mitophagy in both strains (Figures 8A and S7A) as opposed to *atg32* and *atg11* mutants, in which mitophagy was undetectable (Figures 8B and S7B). Atg32 is a mitochondrial outer protein required to initiate mitophagy as a selective type of autophagy (see review [70]). Atg11 is a critical protein for selective autophagy; it is essential in selective and non-selective autophagy processes (see review [71]). From the results shown in Figures 8B and S7B, we conclude that the *mtl1* mutant does not have any defects regarding mitophagy dependent on Atg32. There is an alternative possibility, that if the problem of *mtl1* is mitochondrial function, we would expect to detect severe deficiencies in mitophagy. For this purpose, we analyzed mitophagy in cells expressing a fusion of mitochondrial matrix protein to GFP, Idp1–GFP, transformed in either wt or *mtl1* strains [68]. We monitored the vacuole clipping of the fusion protein Idp1–GFP. The identification of free GFP with anti-GFP antibody in a Western blot would reveal the existence of mitophagy, since GFP is very resistant to degradation [69]. Mitophagy studies are usually carried out in respiratory carbon sources or alternatively in stationary cultures. When wt and *mtl1* cells were grown in SGly to the stationary phase, we observed mitophagy in both strains (Figures 8A and S7A) as opposed to *atg32* and *atg11* mutants, in which mitophagy was undetectable (Figures 8B and S7B). Atg32 is a mitochondrial outer protein required to initiate mitophagy as a selective type of autophagy (see review [70]). Atg11 is a critical protein for selective autophagy; it is essential in selective and non-selective autophagy processes (see review [71]). From the results shown in Figures 8B and S7B, we conclude that the *mtl1* mutant does not have any defects regarding mitophagy dependent on Atg32.

We also checked mitophagy during the diauxic shift and stationary phase in cultures grown in SD with glucose as the only carbon source. We observed mitochondrial degra-

ures 8C and S7C). However, this particular mitochondrial degradation was undetected in each of the *atg1*, *atg7* or *atg11* strains (Figures 8C and S7C). This particular mitophagy-like activity was not dependent on Atg32, since we observed similar results in both wt and *atg32* strains (Figures 8C and S7C). We discarded the possibility that our results reflected bulk autophagy, since *atg11* cultures did not demonstrate defects in bulk autophagy during the diauxic shift nor stationary phase (Figures 8C and S7C). It has been reported in yeast that Atg33 is a mitophagy mitochondrial outer membrane protein [68] required for the stationary phase. We observed a deficiency in mitophagy when we analyzed the *atg33* mutant (Figures 8C and S7C). More interesting was the finding that the *mtl1* mutant was as deficient as *atg11* and *atg33* in Idp1 mitophagy during the diauxic shift and stationary phase (Figures 8C and S7C). The three mutants turned out to have shorter chronological life spans than the corresponding wt (Figure 8D). Our results suggest that yeast cultures in SD demonstrate mitochondrial degradation in the vacuole upon the diauxic shift and

during the stationary phase through a selective autophagy process independent of Atg32 but dependent on the Atg1, Atg7, Atg11 and Atg33 proteins. We also demonstrate that Mtl1 plays a relevant role in initiating this mechanism; one potential target would be

We decided to check whether the absence of mitochondrial degradation during the chronological life span of *mtl1* was also alleviated by either *RAS2* or *SCH9* deletion, and obtained equivalent results to those described for bulk autophagy, inactivation of *RAS2* or *SCH9* restored mitophagy-like degradation during the diauxic shift to *mtl1* mutants

Atg33 that will have to be investigated further.


**Figure 8.** *Cont.*

**Figure 8.** Mtl1 is needed for specific mitochondrial degradation during stationary phase. (**A**) wt and *mtl1* cultures transformed with the plasmid Idp1–GFP were grown in SGly media plus amino acids at 30 °C. Samples were taken at the indicated times to determine mitophagic flux. Mitophagic flux was calculated as the ratio between free GFP and total Idp1–GFP detected in the Western blot. (**B**) *atg32* and *atg11* cultures transformed with Idp1–GFP were grown in SGly medium plus amino acids at 30°C. Samples were collected at the indicated times to determine mitophagic flux as in (**A**). (**C**) wt, *mtl1, atg1, atg7, atg11, atg32* and *atg33* strains bearing the plasmid Idp1–GFP were grown in SD media at 30 °C for 15 days in continuous shacking. Mitophagic flux was determined as in A. (**D**) Chronological life span curves for wt, *mtl1, atg1, atg7, atg11, atg32* and *atg33* strains cultured in SD media plus amino acids at 30 °C. Samples were taken at the indicated times to determine CLS, as described in the Materials and Methods. Numerical data regarding maximum life span (the day when cultures reach 10% survival) and average life span (the day at which 50% survival was recorded) for each strain is depicted. (**E**) *ras2, mtl1ras2, sch9* and *mtl1sch9* mutants, transformed with Idp1–GFP were treated as in (**C**) and mitophagic flux was determined as in (**A**). Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. **Figure 8.** Mtl1 is needed for specific mitochondrial degradation during stationary phase. (**A**) wt and *mtl1* cultures transformed with the plasmid Idp1–GFP were grown in SGly media plus amino acids at 30 ◦C. Samples were taken at the indicated times to determine mitophagic flux. Mitophagic flux was calculated as the ratio between free GFP and total Idp1–GFP detected in the Western blot. (**B**) *atg32* and *atg11* cultures transformed with Idp1–GFP were grown in SGly medium plus amino acids at 30 ◦C. Samples were collected at the indicated times to determine mitophagic flux as in (**A**). (**C**) wt, *mtl1, atg1, atg7, atg11, atg32* and *atg33* strains bearing the plasmid Idp1–GFP were grown in SD media at 30 ◦C for 15 days in continuous shacking. Mitophagic flux was determined as in A. (**D**) Chronological life span curves for wt, *mtl1, atg1, atg7, atg11, atg32* and *atg33* strains cultured in SD media plus amino acids at 30 ◦C. Samples were taken at the indicated times to determine CLS, as described in the Materials and Methods. Numerical data regarding maximum life span (the day when cultures reach 10% survival) and average life span (the day at which 50% survival was recorded) for each strain is depicted. (**E**) *ras2, mtl1ras2, sch9* and *mtl1sch9* mutants, transformed with Idp1–GFP were treated as in (**C**) and mitophagic flux was determined as in (**A**). Error bars in the histograms represent the standard deviation (SD) calculated from three independent experiments. Significance of the data was determinate by *p*-values from a Student's unpaired *t*-test denoted as follows: \* = 0.05 > *p* > 0.01; \*\* = 0.01 > *p* > 0.001.

follows: \* = 0.05 *p* 0.01; \*\* = 0.01 *p* 0.001. **4. Discussion** Our results point to a situation by which gradual glucose depletion activates bulk autophagy, and for this response Mtl1 activity is essential. Unlike in wt cells, in the single mutant *mtl1* there was no a detectable response towards autophagy unless either *RAS2* or *SCH9* were deleted. Deletion of *RAS2* reverted the *mtl1* phenotype regarding bulk autophagy, which points to the importance of glucose availability and the switch from respiratory to fermentative metabolism, suggesting that during that transition the Ras2 pathway must be not active and Mtl1 is the connector between glucose and Ras2 activity. This is supported by the observation that these results also extended to other fermentative sugars (Figure S2F). Interestingly, the signal to autophagy in the models of glucose deprivation did not flow to TORC1, nor to the PKC1 pathway or PKA, but directly to the autophagy machinery to phosphorylate the Atg1 protein. Accordingly, some authors [2] already observed that TORC1 does not seem to play a principal function in glucose starvation. According to former studies [67], the abrupt transition from 2% glucose to 0% glucose does not activate macroautophagy, because for this mechanism, ATP is essential and the mentioned transition causes cells to be suddenly exhausted for ATP. This is understandable, as cells are transferred from a culture containing high glucose concentrations to a We also checked mitophagy during the diauxic shift and stationary phase in cultures grown in SD with glucose as the only carbon source. We observed mitochondrial degradation as free GFP derived from Idp1–GFP accumulated in vacuoles in wt cultures (Figures 8C and S7C). However, this particular mitochondrial degradation was undetected in each of the *atg1*, *atg7* or *atg11* strains (Figures 8C and S7C). This particular mitophagy-like activity was not dependent on Atg32, since we observed similar results in both wt and *atg32* strains (Figures 8C and S7C). We discarded the possibility that our results reflected bulk autophagy, since *atg11* cultures did not demonstrate defects in bulk autophagy during the diauxic shift nor stationary phase (Figures 8C and S7C). It has been reported in yeast that Atg33 is a mitophagy mitochondrial outer membrane protein [68] required for the stationary phase. We observed a deficiency in mitophagy when we analyzed the *atg33* mutant (Figures 8C and S7C). More interesting was the finding that the *mtl1* mutant was as deficient as *atg11* and *atg33* in Idp1 mitophagy during the diauxic shift and stationary phase (Figures 8C and S7C). The three mutants turned out to have shorter chronological life spans than the corresponding wt (Figure 8D). Our results suggest that yeast cultures in SD demonstrate mitochondrial degradation in the vacuole upon the diauxic shift and during the stationary phase through a selective autophagy process independent of Atg32 but dependent on the Atg1, Atg7, Atg11 and Atg33 proteins. We also demonstrate that Mtl1 plays a relevant role in initiating this mechanism; one potential target would be Atg33 that will have to be investigated further.

Significance of the data was determinate by *p-*values from a Student's unpaired *t*-test denoted as

culture without glucose nor other carbon source. However, during the transition from fermentative to respiratory metabolism, the decrease in glucose concentration occurs gradually. In wt cells, autophagy was activated already when the glucose concentration reached a value of 0.5% (27.77 mM), and reached maximum values when glucose levels decreased to 0.05% (0.13 mM), whereas in *mtl1* mutant autophagy was never induced. We hypothesized that this could have occurred because there might be a threshold for cells to We decided to check whether the absence of mitochondrial degradation during the chronological life span of *mtl1* was also alleviated by either *RAS2* or *SCH9* deletion, and obtained equivalent results to those described for bulk autophagy, inactivation of *RAS2* or *SCH9* restored mitophagy-like degradation during the diauxic shift to *mtl1* mutants (Figures 8C,E and S7C,D).

#### sense glucose levels (or other alternative fermentable sugars), which would activate au-**4. Discussion**

tophagy to obtain energy and nutrients probably linked to the induction of respiratory metabolism. Mtl1 could be the sensor for this threshold, and if unable to switch properly Our results point to a situation by which gradual glucose depletion activates bulk autophagy, and for this response Mtl1 activity is essential. Unlike in wt cells, in the

single mutant *mtl1* there was no a detectable response towards autophagy unless either *RAS2* or *SCH9* were deleted. Deletion of *RAS2* reverted the *mtl1* phenotype regarding bulk autophagy, which points to the importance of glucose availability and the switch from respiratory to fermentative metabolism, suggesting that during that transition the Ras2 pathway must be not active and Mtl1 is the connector between glucose and Ras2 activity. This is supported by the observation that these results also extended to other fermentative sugars (Figure S2F). Interestingly, the signal to autophagy in the models of glucose deprivation did not flow to TORC1, nor to the PKC1 pathway or PKA, but directly to the autophagy machinery to phosphorylate the Atg1 protein. Accordingly, some authors [2] already observed that TORC1 does not seem to play a principal function in glucose starvation.

According to former studies [67], the abrupt transition from 2% glucose to 0% glucose does not activate macroautophagy, because for this mechanism, ATP is essential and the mentioned transition causes cells to be suddenly exhausted for ATP. This is understandable, as cells are transferred from a culture containing high glucose concentrations to a culture without glucose nor other carbon source. However, during the transition from fermentative to respiratory metabolism, the decrease in glucose concentration occurs gradually. In wt cells, autophagy was activated already when the glucose concentration reached a value of 0.5% (27.77 mM), and reached maximum values when glucose levels decreased to 0.05% (0.13 mM), whereas in *mtl1* mutant autophagy was never induced. We hypothesized that this could have occurred because there might be a threshold for cells to sense glucose levels (or other alternative fermentable sugars), which would activate autophagy to obtain energy and nutrients probably linked to the induction of respiratory metabolism. Mtl1 could be the sensor for this threshold, and if unable to switch properly could consequently be unable to induce autophagy. In a previous study we observed that the *mtl1* mutant accumulates higher cAMP levels than wt cells. We believe that since the *mtl1* mutant has high cyclic AMP accumulation both in exponential or stationary cultures [28], cells would be depleted of ATP, with consequent Ras2 deletion potentially compensating for that depletion, thus avoiding the accumulation of cyclic AMP in *mtl1*. Nonetheless, this hypothesis was not sustained, since other nutrient stresses were capable of activating bulk autophagy in *mtl1* exponential cells (Figure 4C). It could be argued that during the diauxic shift, glucose reduction forces the switch from fermentative to respiratory metabolism and in these circumstances the main ATP source would be mitochondrial. Since in the absence of Mtl1 the ratio of cAMP/ATP would be higher than in wt cells, this would generate a signal of glucose starvation leading to the blockade of autophagy. However, our results demonstrate that the absence of mitochondrial DNA, and consequently the absence of ATP production through respiration metabolism (*rho0* mutant) did not preclude autophagy induction in glucose reduction conditions. Consequently, we also discarded this second hypothesis.

In humans, ATP increases autophagic flux in some cell lines but not in others [72]. Moreover, the addition of ATP to *mtl1* cultures only partly restored at a low degree of autophagy both in the diauxic transition and upon partial glucose depletion. This supports the model by which Mtl1 function couples glucose starvation to Ras activity and autophagy induction.

In agreement with the studies of [67], we observed that after full reduction of glucose, no nutritional stress could provoke bulk autophagy induction (rapamycin, amino acids, nitrogen, or iron).

In this study, we also present a nutritional model by which Gcn2 detects the signal of amino acid deprivation connecting to bulk autophagy in a manner independent of TORC1 activity. Hence, our data demonstrates that both Mtl1 and Gcn2 are key factors that induce bulk autophagy during the starvation process involving firstly glucose and secondly amino acid deprivation that occurs during the transition from fermentative to respiratory metabolism.

The observation that the *mtl1* mutant grows in the presence of glycerol as the only carbon source at similar rates to wt cells, indicates that mitochondrial function is sufficient

to support the ATP requirements in this condition. Moreover, we also observed that Mtl1 did not participate in the induction of mitophagy in glycerol cultures dependent on Atg32 (Figure 8A,B).

We observed that either Mtl1 or Atg11 (involved in selective autophagy), are required to degrade mitochondria during transition to the stationary phase in synthetic media containing glucose as the only carbon source. This degradation process depends on the autophagy machinery, since in the absence of Atg7 or Atg1 it does not take place. Our data are consistent with a previous paper in which the authors showed that autophagy in response to carbon starvation requires Atg11 as a scaffold protein for the PAS [67]. Nevertheless, the mitophagy-like process that occurred in media containing glucose is independent of Atg32 but dependent on Atg33 (Figure 8C). This is not unexpected, since Atg33 was characterized as an autophagy protein specific to *Saccharomyces cerevisiae* whose role was linked to the induction of mitophagy during the stationary phase [68]. Whether or not the mitochondrial degradation that we observed during the stationary phase is a type of mitophagy dependent of Atg11 and Atg33 but independent of Atg32, should be further analyzed in future studies. Consequently, the availability of glucose as a carbon source has specific responses regarding autophagy in which Mtl1 is directly involved.

Snf1 is an AMPK orthologous to the mammalian AMP-kinase (reviewed in [73]) whose activity has been reported to be required in response to glucose starvation [21,74] to downregulate autophagy in the stationary phase [23], and is also required to induce autophagy in that context [67]. Snf1p is a catabolic regulator that is activated by an increase in the ADP/ATP ratio [75]. Nonetheless, in our studies Snf1, both in *mtl1* mutant or in wt cells, is highly activated during the diauxic shift when glucose levels are reduced and this activity is high during the stationary phase, therefore we cannot attribute lack of Snf1 activity to the defects observed in the *mtl1* mutant.

The conclusion to our data is that the transition from high to low concentrations of glucose triggers the connection between Mtl1 and autophagy and is not fully dependent on mitochondrial function and ATP accumulation. Another argument to support our hypothesis is that in the *mtl1* mutant, the presence of PAS is clearly detectable in the diauxic transition and in the presence of low glucose concentrations (Figure 3F), as opposed to that observed by previous authors [67] in the absence of ATP.

The linkage between Mtl1 and Ras2 has been previously described [28]. Our data suggest that Ras2 is the key regulator of bulk autophagy and the autophagy of mitochondria during the stationary phase. Whether or not this connection is related to mitochondria is at this moment unknown, since direct evidence for a regulation of mitochondria by Ras via cAMP–PKA is absent. Moreover, it is not unlikely that the Ras protein acts independently of adenylate cyclase and cAMP according to [76]. PKA is one of the effectors of Ras2 [77,78], and is also involved in autophagy regulation. However, concerning Mtl1 signaling in autophagy, PKA does not appear to be required as an intermediary molecule. Moreover, constitutive activation of the RAS-cAMP signaling pathway confers resistance to rapamycin [79–81]. This would explain why in an *mtl1* mutant, rapamycin does not provoke the induction of autophagy in the diauxic shift (not shown), taking into consideration the hypothesis that Mtl1 helps in the switch from fermentative to respiratory metabolism through the Ras/cAMP pathway.

Sch9 is a kinase effector of the TORC1 pathway [82]. In addition, Sch9 has been described as acting in a different pathway than PKA in glucose response. This is in agreement with our observations that Bcy1 overexpression or *SLT2* deletion did not suppress the lack of autophagy observed in the *mtl1* mutant. Our results are in line with the conclusion that TORC1, PKA, and Sch9 independently regulate autophagy during growth [17,83]. Deletion of Sch9 is sufficient to activate autophagy [17]. Here, we observed that the *sch9* mutant suppressed the lack of both bulk autophagy and mitophagy of *mtl1* during the diauxic shift and in conditions of low glucose, restricting the signal of glucose availability. Sch9 is also implicated in the selective autophagic degradation of ribosomes, mitochondria, and peroxisomes [84,85]. Although we still do not have a certain interpretation of the fact that

Sch9 regulates ribosomal gene transcription, and that ribosomal biogenesis is one of the mechanisms requiring more ATP consumption, one interpretation would be that diminishing the level of ribosomal biogenesis also diminishes ATP consumption, perhaps favoring the induction of autophagy. The molecular mechanisms underlying the involvement of cAMP in the induction of autophagy related to nutritional starvation conditions deserves future investigation.

Our findings suggest that the Mtl1 cell wall receptor of the CWI pathway is a glucose sensor required to activate both bulk autophagy and Atg33-Atg11 mitophagy in response to glucose concentration decreases. Activation occurs through either Ras2 or Sch9 inactivation converging on Atg1 phosphorylation.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/jof7110903/s1, Figure S1. Sequential descent of glucose and amino acids activates bulk autophagy during the diauxic shift in *Saccharomyces cerevisiae*. Figure S2. Mtl1 and Gcn2 control autophagy induction during glucose and amino acid starvation. Figure S3. Mtl1 signals glucose limitation to the autophagy machinery. Figure S4. Mtl1 signals decreases in glucose concentration to the autophagy machinery in a manner not fully dependent on ATP production by mitochondria. Figure S5. Both Ras2 and Sch9 suppress *mtl1* deficiency in autophagy signaling upon glucose concentration descent. Figure S6. Mtl1 is not deficient in bulk autophagy in respiratory conditions. Figure S7. Mtl1 is needed for specific mitochondrial degradation during stationary phase.

**Author Contributions:** Conceptualization, M.A.d.l.T.-R.; Methodology, M.A.d.l.T.-R., N.P.-C., S.M.-M.; Software, M.A.d.l.T.-R., N.P.-C., S.M.-M.; Validation M.A.d.l.T.-R., N.P.-C., S.M.-M.; Formal Analysis N.P.-C., S.M.-M., M.A.d.l.T.-R.; Investigation, M.A.d.l.T.-R., N.P.-C., S.M.-M.; Resources, M.A.d.l.T.-R.; Data Curation, S.M.-M., N.P.-C., M.A.d.l.T.-R.; Writing-Original Draft Preparation, M.A.d.l.T.-R.; Writing-Review & Editing, S.M.-M., N.P.-C., M.A.d.l.T.-R.; Visualization, M.A.d.l.T.-R., N.P.-C., S.M.-M.; Supervision, M.A.d.l.T.-R.; Project Administration, M.A.d.l.T.-R.; Funding Acquisition, M.A.d.l.T.-R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Plan Nacional de I+D+I of the Spanish Ministery of Economy, Industry and Competitiveness (BIO2017-87828-C2-2-P).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We want to thank D. Abeliovitch (The Hebrew University of Jerusalem Cell Biology, Freiburg, Germany) for kindly sending us the plasmids pGFP–Atg8 and pIdp1–GFP. We want to acknowledge Inmaculada Montoliu and Roser Pané for their technical support. The research described in this publication was partly supported by the Plan Nacional de I+D+I of the Spanish Ministry of Economy, Industry and Competitiveness (BIO2017-87828-C2-2-P). Sandra Montella is funded by a fellowship from the Catalan Government (Spain).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Regulation of Pkc1 Hyper-Phosphorylation by Genotoxic Stress**

**Li Liu <sup>1</sup> , Jiri Veis 2,3, Wolfgang Reiter 2,4 , Edwin Motari 1,†, Catherine E. Costello <sup>5</sup> , John C. Samuelson 1,6 , Gustav Ammerer <sup>2</sup> and David E. Levin 1,6,\***


**Abstract:** The cell wall integrity (CWI) signaling pathway is best known for its roles in cell wall biogenesis. However, it is also thought to participate in the response to genotoxic stress. The stress-activated protein kinase Mpk1 (Slt2, is activated by DNA damaging agents through an intracellular mechanism that does not involve the activation of upstream components of the CWI pathway. Additional observations suggest that protein kinase C (Pkc1), the top kinase in the CWI signaling cascade, also has a role in the response to genotoxic stress that is independent of its recognized function in the activation of Mpk1. Pkc1 undergoes hyper-phosphorylation specifically in response to genotoxic stress; we have found that this requires the DNA damage checkpoint kinases Mec1 (Mitosis Entry Checkpoint) and Tel1 (TELomere maintenance), but not their effector kinases. We demonstrate that the casein kinase 1 (CK1) ortholog, Hrr25 (HO and Radiation Repair), previously implicated in the DNA damage transcriptional response, associates with Pkc1 under conditions of genotoxic stress. We also found that the induced association of Hrr25 with Pkc1 requires Mec1 and Tel1, and that Hrr25 catalytic activity is required for Pkc1 hyperphosphorylation, thereby delineating a pathway from the checkpoint kinases to Pkc1. We used SILAC mass spectrometry to identify three residues within Pkc1 the phosphorylation of which was stimulated by genotoxic stress. We mutated these residues as well as a collection of 13 phosphorylation sites within the regulatory domain of Pkc1 that fit the consensus for CK1 sites. Mutation of the 13 Pkc1 phosphorylation sites blocked hyper-phosphorylation and diminished *RNR3* (RiboNucleotide Reductase) basal expression and induction by genotoxic stress, suggesting that Pkc1 plays a role in the DNA damage transcriptional response.

**Keywords:** Hrr25; Mec1; Tel1; Pkc1; hydroxyurea; UV irradiation

#### **1. Introduction**

The cell wall integrity (CWI) signaling pathway of the budding yeast *Saccharomyces cerevisiae* has been well characterized with regard to its regulation by cell wall stress [1–4]. This pathway regulates biosynthesis of cell wall polymers, organization of the actin cytoskeleton, exocytosis, and the protein kinase C 1 (Pkc1)-mediated stress-activated protein kinase (SAPK) cascade through activation of the small GTPase, Rho1. The SAPK cascade is a linear pathway comprised of Pkc1, a MEKK (Bck1), a pair of redundant MEKs (Mkk1/2) and a SAPK (Mpk1/Slt2). The activation of Mpk1, in response to cell wall stress or hyperactivation of upstream pathway components, drives transcription in support of cell wall biogenesis [5–9] through the SRF-like transcription factor Rlm1 [10,11] and the cell cycle

**Citation:** Liu, L.; Veis, J.; Reiter, W.; Motari, E.; Costello, C.E.; Samuelson, J.C.; Ammerer, G.; Levin, D.E. Regulation of Pkc1 Hyper-Phosphorylation by Genotoxic Stress. *J. Fungi* **2021**, *7*, 874. https://doi.org/10.3390/jof7100874

Academic Editors: María Molina and Humberto Martín

Received: 11 September 2021 Accepted: 13 October 2021 Published: 17 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

transcription factor SBF [12–15]. Moreover, loss-of-function mutants in the SAPK cascade display cell lysis defects that are suppressed by external osmotic support [2], highlighting the central role of this signaling pathway in the maintenance of cell wall integrity.

Although the CWI pathway is best understood for its essential role in maintaining the structural integrity of the cell wall during growth, morphogenesis, and in response to cell wall stressors, there is evidence that this pathway is also important for survival of genomic stress. Defects in any component of the CWI pathway, from the cell surface sensors to Mpk1, cause hyper-sensitivity to a variety of DNA damaging agents [16–20]. The CWI pathway SAPK Mpk1 is phosphorylated by the DNA damage checkpoint kinases Mec1 and Tel1 in response to treatment with methylmethane sulfonate (MMS) [21] or caffeine [22]. The Mec1 and Tel1 protein kinases are orthologs of mammalian ATR and ATM, respectively [23], and have overlapping but distinct functions in the maintenance of yeast genome integrity. Tel1 signals the presence of double-strand breaks specifically [24], whereas Mec1 signals the presence of a variety of DNA damage types, including double-strand breaks [25]. Mec1 and Tel1 phosphorylate and activate the checkpoint effector kinases Chk1 and Rad53, which mediate cell cycle arrest and gene expression in support of DNA repair [26–29].

In addition to phosphorylation of Mpk1, Soriano–Carot et al. [30] detected a DNA damage-induced hyper-phosphorylation of Pkc1. They suggested the existence of a reciprocal regulatory circuit in which Pkc1 was required to activate the DNA damage checkpoint and the DNA damage checkpoint kinase Tel1 was required for the phosphorylationinduced Pkc1 band-shift in response to DNA damage. However, we found that the DNA damage checkpoint is activated normally in *pkc1*∆ mutants from various strain backgrounds [31]. Additionally, there is a bifurcation of pathway outputs at Pkc1, first revealed as a 100-fold increase in mitotic recombination frequency in pkc1 mutants that is not observed in mutants of pathway components below Pkc1 [32]. This is significant because mitotic recombination is the principal mode of double-stranded break repair of DNA in yeast [33]. Moreover, Pkc1 phosphorylates and activates CTP synthetase directly [34], revealing a role in nucleotide metabolism.

Finally, Mpk1 is activated by genotoxic stress through a pathway that does not require the activation of Pkc1 or the other protein kinases that function above Mpk1, but involves ubiquitination and degradation of Msg5, the protein phosphatase that normally maintains Mpk1 in a low activity state [31,35]. Intriguingly, Mpk1 activated in response to genotoxic stress does not drive cell wall stress transcription, suggesting that its activation in this context has a different function. These observations have led to the proposal that Pkc1 plays important roles in the response to genotoxic stress that are separate from its function in the activation of the CWI SAPK cascade [30], but these roles have not been elucidated. We hypothesize that the CWI pathway plays multiple unrecognized roles in the response to DNA damage—some driven by Mpk1, others by another pathway branch from Pkc1. In this study, we establish the pathway through which Pkc1 is hyper-phosphorylated in response to DNA damage and identify several sites within the Pkc1 regulatory domain whose phosphorylation is stimulated under conditions of genotoxic stress. Mutation of these sites impacts DNA damage-regulated gene expression.

#### **2. Materials and Methods**

#### *2.1. Strains, Growth Conditions, and Transformations*

The *S. cerevisiae* strains used in this study were derived from the EG123 background [36], the RDK2669 background (M. Smolka), the W303 background (J.C. Igual), or the Research Genetics background BY4742 (Research Genetics, Inc.; Huntsville, AL, USA) and are listed in Table 1.


**Table 1.** Yeast strains used in this study.

Yeast cell cultures were grown in YPD (1% Bacto yeast extract, 2% Bacto Peptone, 2% glucose) or minimal selective medium, SD (0.67% Yeast nitrogen base, 2% glucose) supplemented with the appropriate nutrients to select for plasmids, which are listed in Table 2. Yeast cells were transformed according to Geitz et al. [38]. Sorbitol (0.5 M) was used to prevent cell lysis for pkc1∆ strains. Cell wall stress was induced by treatment with calcofluor white (CFW, 40 µg/mL; Millipore Sigma, Burlington, MA), or by heat shock at 39 ◦C. Genotoxic stress was induced by treatment with hydroxyurea (HU; 250 mM, except where indicated otherwise; Fisher Scientific, Waltham, MA) or ultraviolet light (UV; 150 J/m<sup>2</sup> for Pkc1 band-shift, or 250 J/m<sup>2</sup> for viability assay). UV irradiation was carried out using an Analytik Jena UVP Crosslinker (Fisher Scientific). Cultures for viability tests were pelleted and resuspended in phosphate-buffered saline (PBS), dispersed on the surface of an empty petri dish for irradiation prior to dilution and plating. Cultures for Pkc1 band-shift were incubated in YPD for an additional 2 h for recovery after irradiation. To inhibit Hrr25 catalytic activity, 5 µM or 10 µM of PP1 analog IV, PP1 analog II (1NM PP1), or PP1 analog (Millipore Sigma) were used in agar plates. For Hrr25 inhibition in culture, PP1 analog IV (20 µM) was added to cultures at the time of addition of HU and incubated for 4 h.




**Table 2.** *Cont.*

#### *2.2. Chromosomal Deletions and Strain Construction*

A *sml1*∆::*TRP1 mec1*∆::*HIS3 tel1*∆::*URA3* strain (DL4277) was generated as a meiotic segregant of DL3954 (*MAT***a**/α MBS115 *SML1*/*sml1*∆::*TRP1 MEC1*/*mec1*∆::*HIS3 TEL1*/*tel1*∆::*URA3*). An *sml1*∆::*TRP1 mec1*∆::*KanMX tel1*∆::*URA3* strain (DL4503) was created by homologous recombination of *mec1*∆::*KanMX* at the *MEC1* locus in an *sml1*∆::*TRP1 tel1*∆::*URA3* strain (DL3951). The *mec1*∆::*KanMX* allele was amplified by PCR from genomic DNA of yeast strain (DL3952) using Phusion high-fidelity DNA polymerase (Thermo Fisher Scientific, San Jose, CA). Transformants were selected for antibiotic G418 resistance and validated by genomic PCR across both integration junction sites. Endogenous tagging of Pkc1 with HTBeaq [43] was achieved by transformation of BY4741 with HB-tagging cassettes amplified from plasmid pWR268 [43], resulting in yeast strain JV826.

Plasmid-borne "gatekeeper" alleles of *HRR25* were tested for function initially in an *hrr25*∆ strain maintained with a plasmid-borne, *GAL1*-controlled *HRR25degron* (DL4290) [37], which is viable only in galactose-containing medium. Later experiments with these mutant alleles were conducted in an *hrr25*∆ strain without the *HRR25degron*. To create *hrr25*∆::*HPHMX4* in W303 (DL4515), a plasmid bearing *HRR25-HA* (p3484; pRS425-*HRR25- HA*) was first introduced into DL4206 (*MAT***a** W303 *ade2 trp1 leu2 his3 ura3 can1*). The *HPHMX4* gene was amplified by PCR from plasmid p3064 (pAG32-*RGC2*) using primers with extensions homologous to the 50 and 30 ends of *HRR25* and used to transform DL4206 pRS425-*HRR25-HA* to hygromycin B resistance. Gene replacement was validated by PCR analysis across both integration junctions. The plasmid-borne *HRR25* in DL4515 was replaced with other alleles of *HRR25* carried on pRS313 or pRS423 using plasmid shuffle.

#### *2.3. Plasmid Construction and Mutagenesis*

The *HRR25* gene was epitope-tagged on its C-terminus with 3xHA and expressed from its natural promoter. The promoter region of *HRR25* (from position −979) and its entire coding sequence was amplified from genomic yeast DNA (DL2772) by high-fidelity PCR polymerase (Phusion) using primers designed with a HindIII site (upstream) and a NotI site (downstream), and subcloned into pRS425-*3HA*-*ADH1<sup>T</sup>* (p3149) to generate pRS425- *HRR25-HA* (p3484). To create pRS425-*hrr25-*∆*404-HA* (p3538), the promoter region of *HRR25* (from position −979) and its C-terminally truncated coding sequence was amplified from pRS425-*HRR25-HA* (p3484) by high-fidelity PCR polymerase (Phusion) using primers designed with a HindIII site (upstream) and a NotI site (downstream) and subcloned into pRS425-*3HA-ADH1<sup>T</sup>* (p3149).

To express HA-tagged *HRR25* from *HIS3*-marked plasmids pRS313-*3HA-ADH1<sup>T</sup>* (p3504) and pRS423-*3HA-ADH1<sup>T</sup>* (p3544), first required subcloning of the *3HA-ADH1<sup>T</sup>* sequence from pRS425-*3HA-ADH1<sup>T</sup>* (p3149) into pRS313 or pRS423 [39] at SmaI and

SacI sites. Plasmid pRS313-*HRR25-HA* (p3545) was next created by subcloning a Sal1- Not1 fragment containing the promoter region of *HRR25* (from position −979) and its entire coding sequence from pRS425-*HRR25-HA* (p3484) into pRS313-*3HA-ADH1<sup>T</sup>* (p3504). Plasmid pRS313-*hrr25-*∆*404-H*A (p3546) was created similarly. Plasmid pRS423-*HRR25- HA* (p3547) was generated by subcloning the Sal1-Not1 fragment containing *HRR25* into pRS423-*3HA-ADH1<sup>T</sup>* (p3544).

The *HRR25* gene was also epitope-tagged on its C-terminus with GFP and expressed from its natural promoter. A PCR fragment containing the GFP coding sequence was amplified from pRS425-*GFP* (p1202) with primers designed with Sma1 and Not1 sites (upstream) and a SacII site (downstream) and cloned into pRS423 to create GFP-tagging vector pRS423-*GFP* (p3560). An Apa1-Not1 fragment containing the promoter region of *HRR25* (from position −979) and its entire coding sequence digested from pRS423-*HRR25-HA* (p3547) was subcloned into pRS423-*GFP* (p3560) at the Apa1-Not1 sites to generate pRS423- *HRR25-GFP* (p3562). Plasmid pRS423-*hrr25-*∆*404-GFP* (p3567) was similarly created from pRS313-*hrr25-*∆*404-HA* (p3546).

Point mutations in *PKC1* were constructed initially in high copy plasmid YEp351- *PKC1-HA* (p813) for use in band-shift assays. However, for phenotypic analyses, we expressed mutant alleles of *PKC1-HA* from a centromeric plasmid. To create pRS314-*PKC1- HA*, *pkc1-3A-HA*, and *pkc1-S/T13A-HA*, *PKC1-HA* (from position -950) with its 30 region (640 bp) were PCR amplified from YEp351-*PKC1-HA* (p813), YEp351-*pkc1-S3A-HA* (p3619) or YEp351-*pkc1-S/T13A-HA* (p3612), respectively, by PrimeSTAR, Max DNA Polymerase (Takara Bio USA, Mountainview, CA) using primers designed with a SalI site (upstream) and with a SacII site (downstream) and cloned into pRS314 (p118).

All point mutations in *HRR25* or *PKC1* were created using QuickChange or QuikChange Lightning Site-Directed Mutagenesis Kits (Agilent Technologies, Santa Clara) according to manufacturer's instructions. All DNA sequences derived from PCR mutagenesis were confirmed by sequence across the entire gene.

#### *2.4. Protein Extraction*

Protein extraction was carried out as described previously for co-immunoprecipitation [44], or using the rapid boiling method [45] for direct immunoblot experiments.

#### *2.5. β-galactosidase Measurements*

Measurements of β-galactosidase activity from *RNR3-lacZ* (p2947) expression experiments were conducted in triplicate and carried out as described in Zhao et al. [46].

#### *2.6. Dephosphorylation Assay*

Protein extracts were prepared from *sml1*∆ (DL3950) cells carrying PKC1-HA (p813), treated with 250 mM HU or untreated, by bead-beating in lysis buffer (10% glycerol, 20 mM Hepes pH 7.5, 150 mM NaCI, 0.5% triton X-100 and Mini protease inhibitor cocktail (Milipore Sigma), without phosphatase inhibitors, followed by centrifugation. The dephosphorylation reactions were performed using Lambda protein phosphatase (New England Biolabs, Ipswich, MA) according to manufacturer's protocol. Briefly, 10 µg of protein extract was incubated with 400 units of Lambda protein phosphatase in Lambda protein phosphatase buffer for 30 min at 30 ◦C. Phosphatase reactions were stopped by adding an equal volume of 2× SDS sample buffer and the resulting samples were boiled for 3 min prior to SDS-PAGE and immunoblot analysis.

#### *2.7. Co-Immunoprecipitation*

To detect association of Pkc1 with Hrr25 in wild-type cells, a plasmid expressing Pkc1-HA under the control of its own promoter (p813) was co-transformed with plasmids expressing either GFP-Hrr25 (p3357) or GFP alone (p3358) into DL100. Transformants were grown to mid-log phase in selective medium and starved for methionine for two hours to induce expression of GFP-Hrr25 or GFP, which were expressed under the control of

the *MET25* promoter. Cultures were then treated with 250 mM HU for 4 h. To test the effect of *MEC1* and *TEL1* loss on the association of Pkc1 with Hrr25, a plasmid expressing Hrr25-GFP under the control of its own promoter (p3562) was co-transformed with p813 into strains DL3950 *(sml1*∆) and DL4503 (*sml1*∆ *mec1*∆ *tel1*∆). To test the association of Pkc1-HA with truncated Hrr25 (Hrr25-∆404-GFP), plasmid p813 was transformed into strains bearing a chromosomal deletion of *HRR25* complemented by plasmids expressing either Hrr25-GFP (DL4541) or Hrr25-∆404-GFP (DL4542). Transformants were grown to mid-log phase in selective medium and treated with 250 mM HU for 4 h. Protein extracts were made as described previously [44]. Extracts (100 µg of protein) were incubated with 10µL of GFP-trap agarose beads (Chromotek, Munich, Germany) at 4 ◦C for 2 h and the samples were washed with IP buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% Triton) three times and boiled in SDS-PAGE buffer.

#### *2.8. SDS-PAGE Electrophoresis and Immunoblot Analysis*

Proteins were separated by SDS-PAGE (10% or 4–20% Mini-PROTEAN TGX gels, BioRad, Hercules, CA, USA) followed by immunoblot analysis using mouse monoclonal α-HA (BioLegend, San Diego, CA, USA) or α-GFP (Millipore Sigma) at a dilution of 1:10,000 and polyclonal rabbit α-Rad53 (Abcam) antibodies at a dilution of 1:2000 to detect Pkc1-HA or -GFP, Hrr25-HA or -GFP and Rad53, respectively. Secondary goat anti-mouse (Jackson ImmunoResearch, Westgrove, PA, USA) and donkey anti-rabbit (GE Healthcare, Chicago, IL, USA) antibodies were used at a dilution of 1:10,000. SDS-PAGE gels used to detect Pkc1 band-shifts were 10% and those used for Pkc1 association with Hrr25 were 4–20%.

#### *2.9. Mass Spectrometric Analysis of Pkc1 Co-Immunoprecipitates*

Pkc1-GFP was expressed from plasmid p2062 (gift of M. Cyert) in strain DL100 and isolated using GFP-trap agarose. Pkc1-GFP immunoprecipitates were run on SDS-PAGE gels prior to in-gel digestion of the contents of the entire lane with trypsin. After desalting, LC-MS/MS was performed using a nanoAcquity ultra-performance liquid chromatography (UPLC) capillary system (Waters Corp., Milford, MA, USA), coupled to an LTQ-Orbitrap hybrid mass spectrometer (Thermo Fisher Scientific) equipped with a TriVersa NanoMate ion source (Advion, Ithaca, NY, USA). Sample concentration and desalting were performed online using a nanoAcquity UPLC trapping column (180 µm by 20 mm; packed with 5-µm, 100-Å-pore-size Symmetry C<sup>18</sup> material; Waters Corp.) at a flow rate of 15 µL/min for 1 min. Separation was accomplished on a nanoAcquity UPLC capillary column (100 µm by 100 mm; packed with 1.7-µm, 130-Å-pore-size bridged ethyl hybrid [BEH] C<sup>18</sup> material; Waters Corp.). A linear gradient of A and B buffers (buffer A, 3% ACN–0.1% formic acid [FA]; buffer B, 97% ACN–0.1% FA) from 7% to 45% buffer B over 124 min was used at a flow rate of 0.5 µL/min to elute peptides into the mass spectrometer. Columns were washed and re-equilibrated between LC-MS/MS experiments. Electrospray ionization was carried out at 1.7 kV using the NanoMate, with the LTQ heated capillary set to 150 ◦C.

Mass spectra were acquired in the orbitrap mass analyzer in the positive-ion mode over the range of *m*/*z* 300 to 2000 at a resolution of 60,000 @ *m*/*z* 400. Mass accuracy after internal calibration was within 4 ppm. Simultaneously, tandem MS spectra were acquired using the LTQ for the five most abundant, multiply charged species in the mass spectrum with signal intensities of >8000 signal/noise levels. With MS/MS collision energies set at 35%, and helium used as the collision gas, MS/MS spectra were acquired over a range of *m*/*z* values dependent on the precursor ion. Dynamic exclusion was set such that MS/MS data for each species were acquired a maximum of twice. All spectra were recorded in profile mode for further processing and analysis. Xcalibur software was used for MS and MS/MS data analysis, while peptide and protein assignments were conducted using Mascot to search against the *ScerevisiaeOrfTrans* database.

#### *2.10. Mass Spectrometric Analysis of Pkc1 Phospho-Sites*

Histidine-biotin tandem affinity purifications of Pkc1-HTBeaq were based on methods described elsewhere [43], with modifications below. Stable isotope labeling using amino acids in cell culture (SILAC) [47] was achieved as described previously [43]. Cells expressing Pkc1 C-terminally tagged with a HTBeaq tag [43] were grown to mid-logarithmic phase (OD600 nm = 2.0), treated with 200 mM HU for 4 h, harvested by filtration, and rapidly deep frozen in liquid N2. In-solution digestion with trypsin and enrichment of phosphorylated peptides using TiO<sup>2</sup> was performed as described previously [43].

For these mass spectrometry analyses, a Q Exactive HF Orbitrap (Thermo Fisher Scientific) mass spectrometer was used with the following settings: Peptides were separated applying an increasing organic solvent (acetonitrile) gradient from 2.5% to 40% in 0.1% formic acid over 60 (YPD) or 120 (SILAC) minutes at a flow rate of 275 nl/min. MS1 resolution was set to 70,000 @ *<sup>m</sup>*/*<sup>z</sup>* 200, AGC 3 <sup>×</sup> <sup>10</sup><sup>6</sup> . MS2 resolution was set to 17,500 @ *<sup>m</sup>*/*<sup>z</sup>* 200, AGC 1 <sup>×</sup> <sup>10</sup><sup>5</sup> , 500 ms max. ion injection time (IIT). The mass spectrometer was configured to pick the twelve most abundant precursor ions for data-dependent MS2 scans, applying HCD for fragmentation with a normalized collision energy (NCE) of 27. Dynamic exclusion time was set to 30 s. MS raw files were processed using MaxQuant (Max Planck Institute of Biochemistry, Munich) [48] software version 1.5.2.8 using standard settings, except for the following modifications. Spectra were searched against the *Saccharomyces* Genome Database (http://www.yeastgenome.org/ accessed on 8 October 2015) containing 6717 entries (3 February 2011), including a list of 248 common laboratory contaminants as well as reversed versions of all sequences. The enzyme specificity was set to trypsin. A maximum of two missed cleavages was allowed. Phosphorylation of serine, threonine, and tyrosine residues, oxidation of methionine, and deamidation of asparagine were set as variable modifications. For stable isotope labeling using amino acids in cell culture (SILAC)–labeled samples, Lys6 and Arg6 were additionally selected. Carbamidomethylation of cysteine was searched as a fixed modification. A maximum of five modifications per peptide was allowed. The false discovery rate for peptide, protein, and site identification was set to 1%. All files were searched together. Minimum delta score for modified peptides was set to 6. The MS phospho-proteomics data have been deposited at the zenodo repository (https://zenodo.org/ accessed on 10 September 2021) and can be accessed via https://doi.org/10.5281/zenodo.5552273.

#### *2.11. Notes on Reproducibility*

All immunoblots, and co-IPs, were reproduced at least once in independent experiments with representative images shown.

#### **3. Results and Discussion**

Pkc1 undergoes a phosphorylation-induced band-shift in response to DNA damage [30]. We found that the Pkc1 band-shift was induced specifically in response to DNA damage by alkylating agent methylmethane sulfonate (MMS), by dNTP depletion by hydroxyurea (HU) treatment, or by UV irradiation, but not in response to cell wall stress treatments calcofluor white (CFW) or elevated growth temperature (Figure 1a). We confirmed that this band-shift results from hyper-phosphorylation by treatment with Lambda protein phosphatase (Figure 1b). We also found that it is dependent on the partially redundant DNA damage checkpoint kinases Mec1 and Tel1 (Figure 1c). This is in contrast to the findings of Soriano-Carot, et al. [30], who found that Tel1 was uniquely responsible for the Pkc1 band-shift. In any case, we found that both of the known effector kinase targets of the checkpoint kinases, Rad53 and Chk1 [49,50], were dispensable for the Pkc1 band-shift (Figure 1d). These results suggest that Mec1 and Tel1 either phosphorylate Pkc1 directly, or act on another protein kinase that phosphorylates Pkc1. Thus, the DNA damage checkpoint kinases, Mec1 and Tel1, regulate hyper-phosphorylation of Pkc1 in response to genotoxic stress.

**Figure 1.** A phosphorylation band‐shift in Pkc1 requires the *MEC1* and *TEL1* DNA damage check‐ point genes. (**a**) A Pkc1 band‐shift is induced specifically by genotoxic stressors. Wild‐type cells (DL3950; *sml1*Δ) expressing Pkc1‐HA (from p813) were exposed to genotoxic stress (250 mM HU for 4 h, 0.04% MMS for 2 h, or 150 J/m2 UV with a 2 h recovery period), cell wall stress (40 μg/mL CFW for 1 h, or heat shock at 39 °C for 1 h), or untreated (Con). Extracts were subjected to SDS‐PAGE and immunoblot analysis for Pkc1‐HA; (**b**) Phosphorylation is responsible for the HU‐induced band‐ shift. Wild‐type cells (DL3950) expressing Pkc1‐HA were either treated with HU as above, or un‐ treated. Extracts were treated with Lambda protein phosphatase (LPP), as described in Methods, prior to immunoblot analysis; (**c**,**d**) The HU‐induced Pkc1 band‐shift requires *MEC1* and *TEL1*, but not the checkpoint genes that they regulate (*RAD53* or *CHK1*). Cultures were treated with HU as above prior to immunoblot analysis for Pkc1‐HA. Strains are DL3950 (*sml1*Δ), DL3951 (*sml1*Δ *tel1*Δ), DL3952 (*sml1*Δ *mec1*Δ), DL4277 (*sml1*Δ *mec1*Δ *tel1*Δ), DL3953 (*sml1*Δ *rad53*Δ), DL2772 (Res. Gen. WT), and DL4286 (Res. Gen. *chk1*Δ). The *sml1*Δ mutation is required to suppress the lethality of the *mec1*Δ and *rad53*Δ mutations. **Figure 1.** A phosphorylation band-shift in Pkc1 requires the *MEC1* and *TEL1* DNA damage checkpoint genes. (**a**) A Pkc1 band-shift is induced specifically by genotoxic stressors. Wild-type cells (DL3950; *sml1*∆) expressing Pkc1-HA (from p813) were exposed to genotoxic stress (250 mM HU for 4 h, 0.04% MMS for 2 h, or 150 J/m<sup>2</sup> UV with a 2 h recovery period), cell wall stress (40 µg/mL CFW for 1 h, or heat shock at 39 ◦C for 1 h), or untreated (Con). Extracts were subjected to SDS-PAGE and immunoblot analysis for Pkc1-HA; (**b**) Phosphorylation is responsible for the HU-induced band-shift. Wild-type cells (DL3950) expressing Pkc1-HA were either treated with HU as above, or untreated. Extracts were treated with Lambda protein phosphatase (LPP), as described in Methods, prior to immunoblot analysis; (**c**,**d**) The HU-induced Pkc1 band-shift requires *MEC1* and *TEL1*, but not the checkpoint genes that they regulate (*RAD53* or *CHK1*). Cultures were treated with HU as above prior to immunoblot analysis for Pkc1-HA. Strains are DL3950 (*sml1*∆), DL3951 (*sml1*∆ *tel1*∆), DL3952 (*sml1*∆ *mec1*∆), DL4277 (*sml1*∆ *mec1*∆ *tel1*∆), DL3953 (*sml1*∆ *rad53*∆), DL2772 (Res. Gen. WT), and DL4286 (Res. Gen. *chk1*∆). The *sml1*∆ mutation is required to suppress the lethality of the *mec1*∆ and *rad53*∆ mutations.

#### *3.1. Genotoxic Stress Induces Hrr25 Association with Pkc1*

*3.1. Genotoxic Stress Induces Hrr25 Association with Pkc1* Next, we took a proteomics approach to identify protein kinases that associate with Pkc1 in response to HU‐induced genotoxic stress. Cells expressing Pkc1‐GFP were treated with 250 mM HU for 4 h, or were untreated, and Pkc1‐GFP was immunoprecipitated from extracts and subjected to mass spectrometric (MS) analysis to identify co‐precipitating proteins. Proteins that were found associated with Pkc1 only after HU treatment, or only without HU treatment are presented in Supplemental Tables S1 and S2, respectively. Alt‐ hough numerous non‐specific associations were detected in both the untreated and treated samples, a minor signal from the DNA damage and repair kinase Hrr25 (HO and radiation repair) [51] was detected only in the HU‐treated sample. *HRR25* encodes an iso‐ form of casein kinase 1 (CK1) that has been implicated in the repair of DNA double strand breaks [51] and is required for the transcriptional induction of the ribonucleotide reduc‐ tase genes *RNR2* and *RNR3* in response to DNA damage [52]. Many physical interactors have been identified for Pkc1, but Hrr25 is not among them (*Saccharomyces* Genome Data‐ base). Therefore, we validated the physical association between Pkc1 and Hrr25 by co‐ immunoprecipitation (co‐IP). Not only did we detect a stable interaction between Pkc1‐ Next, we took a proteomics approach to identify protein kinases that associate with Pkc1 in response to HU-induced genotoxic stress. Cells expressing Pkc1-GFP were treated with 250 mM HU for 4 h, or were untreated, and Pkc1-GFP was immunoprecipitated from extracts and subjected to mass spectrometric (MS) analysis to identify co-precipitating proteins. Proteins that were found associated with Pkc1 only after HU treatment, or only without HU treatment are presented in Supplemental Tables S1 and S2, respectively. Although numerous non-specific associations were detected in both the untreated and treated samples, a minor signal from the DNA damage and repair kinase Hrr25 (HO and radiation repair) [51] was detected only in the HU-treated sample. *HRR25* encodes an isoform of casein kinase 1 (CK1) that has been implicated in the repair of DNA double strand breaks [51] and is required for the transcriptional induction of the ribonucleotide reductase genes *RNR2* and *RNR3* in response to DNA damage [52]. Many physical interactors have been identified for Pkc1, but Hrr25 is not among them (*Saccharomyces* Genome Database). Therefore, we validated the physical association between Pkc1 and Hrr25 by co-immunoprecipitation (co-IP). Not only did we detect a stable interaction between Pkc1- HA and Hrr25-GFP, but the interaction was reproducibly induced 5–10-fold in response to treatment with HU (Figure 2a). We detected no non-specific binding of Pkc1-HA to either GFP alone (Figure 2a), or to the GFP-trap beads (Supplemental Figure S1). Hrr25

HA and Hrr25‐GFP, but the interaction was reproducibly induced 5–10‐fold in response

was recruited similarly to Pkc1 in response to ultraviolet (UV) light exposure (Figure 2b),

We next tested the role of Mec1 and Tel1 in the induced association of Hrr25 with Pkc1. Here, we found that, not only are the DNA damage checkpoint kinases required for

suggesting that this is part of the generalized response to genotoxic stress.

*J. Fungi* **2021**, *7*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/jof

was recruited similarly to Pkc1 in response to ultraviolet (UV) light exposure (Figure 2b), suggesting that this is part of the generalized response to genotoxic stress. *HRR25* strain (Figure 3a). This suggests that Hrr25 catalytic activity is, indeed, required for the Pkc1‐bandshift observed in response to genotoxic stress.

the Pkc1 phosphorylation band‐shift, but they are also required for the induced associa‐ tion of Hrr25 with Pkc1 (Figure 2c), suggesting that Mec1 and Tel1 induce the Pkc1 phos‐ phorylation band‐shift indirectly by driving its association with Hrr25. Therefore, we sought to test this idea using mutants in *HRR25*. Because the *HRR25* gene is essential for viability, we constructed two analog‐sensitive "gatekeeper" alleles [53] of *HRR25* (*hrr25‐ I82A* and *hrr25‐I82G*). These were introduced on plasmids into an *hrr25*Δ strain that is maintained with a galactose‐inducible form of *HRR25* and tested for their sensitivity to a collection of protein kinase inhibitory ATP‐analogs in glucose‐containing medium (YPD). We determined that the *hrr25‐I82A* allele conferred optimum growth sensitivity to PP1 analog IV, which did not appreciably inhibit the growth of the strain expressing wild‐type *HRR25* (Supplemental Figure S2). We next introduced the *hrr25‐I82A* allele to an *hrr25*Δ strain by plasmid shuffle (See Materials and Methods, Section 2) to test the importance of this protein kinase in the HU‐induced Pkc1 band‐shift. The strain expressing the *hrr25‐ I82A* allele along with an epitope tagged form of Pkc1, was subjected to inhibition of Hrr25, together with HU treatment to induce genotoxic stress. Pkc1 failed to display an HU‐in‐ duced band‐shift in the *hrr25‐I82A* strain in the presence of inhibitor, in contrast to the

*J. Fungi* **2021**, *7*, x FOR PEER REVIEW 10 of 17

**Figure 2.** HU treatment induces association of Pkc1 with Hrr25. (**a**) Wild‐type cells (DL100) co‐ex‐ pressing Pkc1‐HA (from p813) and Hrr25‐GFP (from p3357) or GFP (from p3358) were treated with 250 mM HU for 4 h. Hrr25‐GFP or GFP was immunoprecipitated (IP) from extracts and samples were tested by immunoblot analysis for co‐IP of Pkc1‐HA. Input Pkc1‐HA from extracts is shown at bottom; (**b**) UV treatment induces association of Pkc1 with Hrr25. Wild‐type cells (DL100) co‐ expressing Pkc1‐HA (from p813) and Hrr25‐GFP (from p3357) or GFP (from p3358) were treated with UV light (150 J/m2) and returned to culture for 2 h post‐irradiation prior to extract preparation. Hrr25‐GFP or GFP was immunoprecipitated (IP) from extracts and treated as above; (**c**) *MEC1* and *TEL1* are required for the HU‐induced association of Hrr25 with Pkc1. Cultures co‐expressing Pkc1‐ HA and Hrr25‐GFP (from p3562) were treated with HU as above and processed for co‐IP of Pkc1‐ HA with Hrr25‐GFP. Strains were DL3950 (*sml1*Δ) and DL4277 (*sml1*Δ *tel1*Δ *mec1*Δ). Molecular mass markers (in kDa) are shown on the right. **Figure 2.** HU treatment induces association of Pkc1 with Hrr25. (**a**) Wild-type cells (DL100) coexpressing Pkc1-HA (from p813) and Hrr25-GFP (from p3357) or GFP (from p3358) were treated with 250 mM HU for 4 h. Hrr25-GFP or GFP was immunoprecipitated (IP) from extracts and samples were tested by immunoblot analysis for co-IP of Pkc1-HA. Input Pkc1-HA from extracts is shown at bottom; (**b**) UV treatment induces association of Pkc1 with Hrr25. Wild-type cells (DL100) co-expressing Pkc1-HA (from p813) and Hrr25-GFP (from p3357) or GFP (from p3358) were treated with UV light (150 J/m<sup>2</sup> ) and returned to culture for 2 h post-irradiation prior to extract preparation. Hrr25-GFP or GFP was immunoprecipitated (IP) from extracts and treated as above; (**c**) *MEC1* and *TEL1* are required for the HU-induced association of Hrr25 with Pkc1. Cultures co-expressing Pkc1-HA and Hrr25-GFP (from p3562) were treated with HU as above and processed for co-IP of Pkc1-HA with Hrr25-GFP. Strains were DL3950 (*sml1*∆) and DL4277 (*sml1*∆ *tel1*∆ *mec1*∆). Molecular mass markers (in kDa) are shown on the right.

*J. Fungi* **2021**, *7*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/jof We next tested the role of Mec1 and Tel1 in the induced association of Hrr25 with Pkc1. Here, we found that, not only are the DNA damage checkpoint kinases required for the Pkc1 phosphorylation band-shift, but they are also required for the induced association of Hrr25 with Pkc1 (Figure 2c), suggesting that Mec1 and Tel1 induce the Pkc1 phosphorylation band-shift indirectly by driving its association with Hrr25. Therefore, we sought to test this idea using mutants in *HRR25*. Because the *HRR25* gene is essential for viability, we constructed two analog-sensitive "gatekeeper" alleles [53] of *HRR25* (*hrr25-I82A* and *hrr25-I82G*). These were introduced on plasmids into an *hrr25*∆ strain that is maintained with a galactose-inducible form of *HRR25* and tested for their sensitivity to a collection of protein kinase inhibitory ATP-analogs in glucose-containing medium (YPD). We determined that the *hrr25-I82A* allele conferred optimum growth sensitivity to PP1 analog IV, which did not appreciably inhibit the growth of the strain expressing wild-type *HRR25* (Supplemental Figure S2). We next introduced the *hrr25-I82A* allele to an *hrr25*∆ strain by plasmid shuffle (See Materials and Methods, Section 2) to test the

importance of this protein kinase in the HU-induced Pkc1 band-shift. The strain expressing the *hrr25-I82A* allele along with an epitope tagged form of Pkc1, was subjected to inhibition of Hrr25, together with HU treatment to induce genotoxic stress. Pkc1 failed to display an HU-induced band-shift in the *hrr25-I82A* strain in the presence of inhibitor, in contrast to the *HRR25* strain (Figure 3a). This suggests that Hrr25 catalytic activity is, indeed, required for the Pkc1-bandshift observed in response to genotoxic stress. *J. Fungi* **2021**, *7*, x FOR PEER REVIEW 11 of 17

**Figure 3.** An analog‐sensitive form of Hrr25 shows that its catalytic activity is required for the HU‐induced Pkc1 band‐ shift. (**a**) An *hrr25*Δ strain complemented by plasmid‐borne *HRR25* (DL4527) or *hrr25‐182A* (DL4528; encoding an analog‐ sensitive form) and expressing Pkc1‐HA (from p813) was treated simultaneously with HU (250 mM) and/or PP1 analog IV (20 μM) for 4 h. Extracts were processed for immunoblot analysis of Pkc1‐HA; (**b**) A C‐terminal truncation of Hrr25 lacking three potential Mec1/Tel1 phosphorylation sites associates normally with Pkc1. An *hrr25*Δ strain complemented by plasmid‐borne *HRR25‐GFP* (DL4541) or *hrr25‐*Δ*404* (DL4542) and expressing Pkc1‐HA (from p813) was treated with HU (250 mM for 4 h). Hrr25‐GFP was immunoprecipitated from extracts and tested for co‐IP of Pkc1‐HA by immunoblot analysis. Molecular mass markers (in kDa) are on the right; (**c**) A C‐terminal truncation of Hrr25 does not impact the HU‐ induced Pkc1 band‐shift. An *hrr25*Δ strain complemented by plasmid‐borne *HRR25* (DL4527) or *hrr25‐*Δ*404* (DL4555) and expressing Pkc1‐HA was treated with HU as above and processed for immunoblot analysis of Pkc1‐HA (**d**) A mutant form of *HRR25* lacking three potential Mec1/Tel1 phosphorylation sites does not impact the HU‐induced Pkc1 band‐shift. An *hrr25*Δ strain complemented by plasmid‐borne *HRR25* (DL4527) or *hrr25‐3A* (DL4556) and expressing Pkc1‐HA was **Figure 3.** An analog-sensitive form of Hrr25 shows that its catalytic activity is required for the HU-induced Pkc1 band-shift. (**a**) An *hrr25*∆ strain complemented by plasmid-borne *HRR25* (DL4527) or *hrr25-182A* (DL4528; encoding an analog-sensitive form) and expressing Pkc1-HA (from p813) was treated simultaneously with HU (250 mM) and/or PP1 analog IV (20 µM) for 4 h. Extracts were processed for immunoblot analysis of Pkc1-HA; (**b**) A C-terminal truncation of Hrr25 lacking three potential Mec1/Tel1 phosphorylation sites associates normally with Pkc1. An *hrr25*∆ strain complemented by plasmidborne *HRR25-GFP* (DL4541) or *hrr25-*∆*404* (DL4542) and expressing Pkc1-HA (from p813) was treated with HU (250 mM for 4 h). Hrr25-GFP was immunoprecipitated from extracts and tested for co-IP of Pkc1-HA by immunoblot analysis. Molecular mass markers (in kDa) are on the right; (**c**) A C-terminal truncation of Hrr25 does not impact the HU-induced Pkc1 band-shift. An *hrr25*∆ strain complemented by plasmid-borne *HRR25* (DL4527) or *hrr25-*∆*404* (DL4555) and expressing Pkc1-HA was treated with HU as above and processed for immunoblot analysis of Pkc1-HA (**d**) A mutant form of *HRR25* lacking three potential Mec1/Tel1 phosphorylation sites does not impact the HU-induced Pkc1 band-shift. An *hrr25*∆ strain complemented by plasmid-borne *HRR25* (DL4527) or *hrr25-3A* (DL4556) and expressing Pkc1-HA was treated with HU as above and processed for immunoblot analysis of Pkc1-HA.

treated with HU as above and processed for immunoblot analysis of Pkc1‐HA. Hrr25 possesses three potential Mec1/Tel1 sites (S/T‐Q) [21], which all reside within the C‐terminal tail of Hrr25 (residues S405, S438, and T453). Therefore, we generated a truncated version of Hrr25 that is missing the C‐terminal domain from residues 405–494 (*hrr25‐*Δ*404*), which removes all three potential Mec1/Tel1 sites. This allele complemented the lethality of the *hrr25*Δ mutation, therefore we tested it for HU‐induced association with Pkc1 and for the HU‐induced Pkc1 band‐shift. The truncated form of Hrr25 was re‐ cruited normally to Pkc1 in response to HU treatment (Figure 3b) and retained its ability to drive the Pkc1 band‐shift (Figure 3c). However, we considered the possibility that the Hrr25 C‐terminal region might carry an auto‐inhibitory domain, truncation of which Hrr25 possesses three potential Mec1/Tel1 sites (S/T-Q) [21], which all reside within the C-terminal tail of Hrr25 (residues S405, S438, and T453). Therefore, we generated a truncated version of Hrr25 that is missing the C-terminal domain from residues 405–494 (*hrr25-*∆*404*), which removes all three potential Mec1/Tel1 sites. This allele complemented the lethality of the *hrr25*∆ mutation, therefore we tested it for HU-induced association with Pkc1 and for the HU-induced Pkc1 band-shift. The truncated form of Hrr25 was recruited normally to Pkc1 in response to HU treatment (Figure 3b) and retained its ability to drive the Pkc1 band-shift (Figure 3c). However, we considered the possibility that the Hrr25 C-terminal region might carry an auto-inhibitory domain, truncation of which could activate Hrr25 independently of Mec1/Tel1. To address this possibility, we mutated the three potential Mec1/Tel1 sites within this domain to Ala residues, yielding *hrr25-3A*, and tested the influence of this allele on the HU-induced Pkc1 band-shift. As with the truncated allele, the *hrr25-3A* mutant was able to mediate the HU-induced band-shift normally

could activate Hrr25 independently of Mec1/Tel1. To address this possibility, we mutated

truncated allele, the *hrr25‐3A* mutant was able to mediate the HU‐induced band‐shift nor‐ mally (Figure 3d). Therefore, we conclude that Mec1 and Tel1 likely regulate HU‐induced Hrr25 association with Pkc1 and the Pkc1 band‐shift by means other than direct phos‐ phorylation of Hrr25. Because neither Rad53, nor Chk1, the known effector kinases of Mec1 and Tel1, were required forthe HU‐induced Pkc1 band‐shift, itremains unclear how the sensor kinases regulate the Hrr25 action on Pkc1. It is possible that Mec1 and Tel1 act on Pkc1, rather than on Hrr25, to regulate the association of these protein kinases. Never‐ theless, our results strongly suggest that both Hrr25 and Pkc1 are indirect Mec1/Tel1 ef‐

*J. Fungi* **2021**, *7*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/jof

*3.2. Identification of Pkc1 Phospho‐Sites in Response to HU Treatment*

fectors.

(Figure 3d). Therefore, we conclude that Mec1 and Tel1 likely regulate HU-induced Hrr25 association with Pkc1 and the Pkc1 band-shift by means other than direct phosphorylation of Hrr25. Because neither Rad53, nor Chk1, the known effector kinases of Mec1 and Tel1, were required for the HU-induced Pkc1 band-shift, it remains unclear how the sensor kinases regulate the Hrr25 action on Pkc1. It is possible that Mec1 and Tel1 act on Pkc1, rather than on Hrr25, to regulate the association of these protein kinases. Nevertheless, our results strongly suggest that both Hrr25 and Pkc1 are indirect Mec1/Tel1 effectors.

#### *3.2. Identification of Pkc1 Phospho-Sites in Response to HU Treatment*

Hrr25 is an ortholog of mammalian casein kinase 1 delta (CK1δ) [54]. Two phosphorylation site motifs for this class of protein kinase have been described [55–58]. These protein kinases require either a priming phosphorylation at position −3 relative to the target S/T site (i.e., p-S/TXXS/T) or an acidic residue at position −3 relative to the target site (i.e., D/EXXS/T).

We conducted both a shotgun mass spectrometric (MS) analysis and a quantitative SILAC MS analysis of HU-induced phosphorylation sites on Pkc1 [43]. These analyses identified many phosphorylation sites on Pkc1, several of which had not been described previously (i.e., T570, T626, S666, T779, S781, and T785; Supplemental Figure S3). Among the phosphorylation sites identified in our SILAC MS experiments, only three appeared to be upregulated in response to HU treatment: S2, S577, and S657 (Supplemental Table S3). Of these, both S577 and S657 are potential Hrr25 phosphorylation sites because S577 has a priming phosphorylation site (p-S574) and S657 is positioned three residues beyond E654 (Figure 4a). Therefore, we started our mutational analysis with these three sites. The individual mutations did not display a detectable impact on the HU-induced Pkc1 bandshift (not shown), so we created a triple mutant. The *pkc1-3A* (*pkc1-S2A, S577A, S657A*) allele was able to complement the lethality of a *pkc1*∆ mutant for growth on rich medium in the absence of osmotic support. Therefore, we examined its behavior in response to genotoxic stress. The HU-induced Pkc1 band-shift was somewhat diminished in the *pkc1- 3A* mutant (Figure 4b), revealing a cumulative effect of these phosphorylation sites. We next tested this mutant for its sensitivity to HU and UV treatment. However, this mutant did not display increased sensitivity to either treatment (Figure 4c). Therefore, we considered a larger collection of Pkc1 phosphorylation sites.

Among the phosphorylation sites identified in our analyses, together with those established previously [21,59–61], we identified 13 phosphorylation sites within Pkc1 that fit either of the two CK1 consensus sequences. Figure 4a shows a map of these sites, all of which reside within the N-terminal regulatory domain of Pkc1. However, most of these sites (11) are clustered in an area without recognizable regulatory features between residues 577 and 804. We mutated all of these phosphorylation sites to alanine residues in groups and in series. We were only able to detect an impact on the HU-induced Pkc1 band-shift once at least nine of these CK1 residues were mutated (data not shown). Blocking the HUinduced band-shift completely required mutation of all 13 sites in the *pkc1-S/T13A* mutant (Figure 4d). The UV-induced band-shift was also blocked in the *pkc1-S/T13A* mutant (Figure 4e). The DNA damage checkpoint was activated normally in the *pkc1-S/T13A* mutant, as judged by a strong Rad53 band-shift (Figure 4d,e). The *pkc1-S/T13A* allele also complemented the null mutant for growth in the absence of osmotic support. Therefore, we tested the sensitivity of this strain to genotoxic stress. However, like the *pkc1-3A* mutant, the *pkc1-S/T13A* mutant did not display enhanced sensitivity to either HU or UV treatment (Figure 4c), suggesting that the biological impact of Pkc1 hyper-phosphorylation during genotoxic stress is too subtle to detect by this viability assay.

ered a larger collection of Pkc1 phosphorylation sites.

Hrr25 is an ortholog of mammalian casein kinase 1 delta (CK1δ) [54]. Two phosphor‐ ylation site motifs for this class of protein kinase have been described [55–58]. These pro‐ tein kinases require either a priming phosphorylation at position −3 relative to the target S/T site (i.e., p‐S/TXXS/T) or an acidic residue at position −3 relative to the target site (i.e.,

We conducted both a shotgun mass spectrometric (MS) analysis and a quantitative SILAC MS analysis of HU‐induced phosphorylation sites on Pkc1 [43]. These analyses identified many phosphorylation sites on Pkc1, several of which had not been described previously (i.e., T570, T626, S666, T779, S781, and T785; Supplemental Figure S3). Among the phosphorylation sites identified in our SILAC MS experiments, only three appeared to be upregulated in response to HU treatment: S2, S577, and S657 (Supplemental Table S3). Of these, both S577 and S657 are potential Hrr25 phosphorylation sites because S577 has a priming phosphorylation site (p‐S574) and S657 is positioned three residues beyond E654 (Figure 4a). Therefore, we started our mutational analysis with these three sites. The individual mutations did not display a detectable impact on the HU‐induced Pkc1 band‐ shift (not shown), so we created a triple mutant. The *pkc1‐3A* (*pkc1‐S2A, S577A, S657A*) allele was able to complement the lethality of a *pkc1*Δ mutant for growth on rich medium in the absence of osmotic support. Therefore, we examined its behavior in response to genotoxic stress. The HU‐induced Pkc1 band‐shift was somewhat diminished in the *pkc1‐ 3A* mutant (Figure 4b), revealing a cumulative effect of these phosphorylation sites. We next tested this mutant for its sensitivity to HU and UV treatment. However, this mutant did not display increased sensitivity to either treatment (Figure 4c). Therefore, we consid‐

D/EXXS/T).

*J. Fungi* **2021**, *7*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/jof **Figure 4.** CK1 phosphorylation sites within the Pkc1 regulatory domain are responsible for the genotoxic stress-induced Pkc1 band-shift. (**a**) Phosphorylation sites within the Pkc1 regulatory domain mutated in this study. Phosphorylated residues are marked in red. The three residues mutated in the *pkc1-3A* allele (S2, S577, and S657) are marked by boxes and were identified by SILAC MS as increased in phosphorylation state in response to HU treatment. Two of these residues (S577 and S657) are within consensus CK1 phosphorylation sites, with either a priming phospho-Ser at position −3 (S577) or an acidic residue (Asp) at position −3 (S657). Other phosphorylated residues that reside within CK1 consensus sites are also indicated and were mutated in the *pkc1-S/T13A* allele. Known regulatory elements, Rho-binding domains (HR1), calcium/lipid-binding domain (C2), and Cys-rich domain (CRD) are also shown. The catalytic domain is C-terminal to the regulatory domain and starts at residue 824; (**b**) HU-induced phosphorylation band-shift of the Pkc1-3A mutant. Plasmids were *PKC1-HA* (p813) and *pkc1-3A-HA* (p3619) (**c**) The *pkc1-3A* and *pkc1-S/T13A* mutants do not show increased sensitivity to genotoxic stress. Serial 10-fold dilutions of cultures grown to mid-log phase in YPD were spotted onto plates (left to right) with or without HU. Cultures treated with UV were similarly diluted and spotted onto YPD plates. Plates were incubated at 25 ◦C for two days. Plasmids were *PKC1-HA* (p3623), *pkc1-3A* (p3624), and *pkc1-S/T13A* (p3625); (**d**) HU-induced phosphorylation band-shift of the Pkc1-S/T13A mutant and Rad53. Plasmids were *PKC1-HA* (p813) *pkc1-S/T13A* (p3612); (**e**) UV-induced phosphorylation band-shift of the Pkc1-S/T13A mutant and Rad53. Plasmids were *PKC1-HA* (p813) and *pkc1-S/T13A* (p3612). Strain DL1021 (*pkc1*∆) was used for experiments shown in (**b**,**d**,**e**). Strain DL376 (*pkc1*∆) was used for the experiment in panel (**c**); (**f**) HU-induced *RNR3-lacZ* expression is diminished in a *pkc1*∆ mutant, and in the *pkc1-3A* and *pkc1-S/T13A* mutants. Strain DL376 (*pkc1*∆) was co-transformed with p*RNR3-lacZ* (p2947) and *PKC1-HA* (p3623), *pkc1-3A* (p3624), *pkc1-S/T13A* (p3625), or vector alone (p118). Cells were cultured in the presence of 0.5 M sorbitol for osmotic support (pair on left), or in the absence of sorbitol. Cultures were treated for 4 h with 250 mM HU and β-galactosidase activity was measured from extracts. The *hrr25-I82A* mutant (DL4528; right) was treated with HU plus or minus PP1 analog IV (20 µM) for 4 h. Each value is the mean and standard deviation from three independent cultures. Pair-wise *p*-values for HU-treated and untreated samples were calculated using student t-test and were all at least *p* ≤ 0.00001, except the HU-treated *PKC1* and *pkc1-3A* pair, which was *p* = 0.0012. An additional p-value of *p* < 0.00001 was obtained for the *hrr25-I82A* mutant for HU-treated samples, with and without analog IV.

*HRR25* has been implicated in the transcriptional response to DNA damage, most notably in the induction of the *RNR3* gene [52], but its role has not been clearly established. The cell cycle transcriptional regulatory factor SBF, comprised of Swi4 and Swi6, is similarly important for this transcriptional response and Hrr25 associates with and phosphorylates Swi6 in vitro [52]. However, *RNR3* induction in response to DNA damage was also shown to be largely under the control of the Mec1-Rad53-Dun1 pathway through the Crt1 transcriptional repressor, which is hyper-phosphorylated in response to genotoxic stress [26]. Therefore, we asked if the expression of an *RNR3-lacZ* reporter was influenced by *PKC1*. We found that both basal *RNR3-lacZ* expression and its induction by HU treatment were strongly diminished in a *pkc1*∆ mutant grown in the presence of osmotic support (Figure 4f). We also examined the effect of the *pkc1-3A* and *pkc1-S/T13A* alleles. *RNR3-lacZ* basal expression and induction were only modestly diminished in the *pkc1-3A* mutant, but were more strongly impaired in the *pkc1-S/T13A* mutant (Figure 4f). Although the basal and induced levels of *RNR3-lacZ* expression were reduced in each of these mutants, the relative induction was retained for all (approximately 6 to 8-fold). Finally, we tested the analog-sensitive *hrr25-I82A* mutant for induction of *RNR3-lacZ* expression and found that, as anticipated, inhibitor treatment strongly diminished HU induction (Figure 4f). These results suggest that Hrr25 may regulate the transcriptional response to DNA damage in part through phosphorylation of Pkc1 (Figure 5). This pathway would be independent of the Mec1-Rad53-Dun1 pathway, because Pkc1 hyper-phosphorylation in response to genotoxic stress does not require Rad53. *J. Fungi* **2021**, *7*, x FOR PEER REVIEW 14 of 17 basal and induced levels of *RNR3‐lacZ* expression were reduced in each of these mutants, the relative induction was retained for all (approximately 6 to 8‐fold). Finally, we tested the analog‐sensitive *hrr25‐I82A* mutant for induction of *RNR3‐lacZ* expression and found that, as anticipated, inhibitor treatment strongly diminished HU induction (Figure 4f). These results suggest that Hrr25 may regulate the transcriptional response to DNA dam‐ age in part through phosphorylation of Pkc1 (Figure 5). This pathway would be independ‐ ent of the Mec1‐Rad53‐Dun1 pathway, because Pkc1 hyper‐phosphorylation in response to genotoxic stress does not require Rad53.

**Figure 5.** Proposed contribution of Hrr25 and Pkc1 to DNA damage‐induced transcription. Pathway from Mec1 and Tel1 through Hrr25 and Pkc1 is added to the pathway established by Huang et al. [26]. Dashed arrows suggest indirect regu‐ lation. **Figure 5.** Proposed contribution of Hrr25 and Pkc1 to DNA damage-induced transcription. Pathway from Mec1 and Tel1 through Hrr25 and Pkc1 is added to the pathway established by Huang et al. [26]. Dashed arrows suggest indirect regulation.

#### **4. Conclusions 4. Conclusions**

We can draw several conclusions from this study. First, the checkpoint kinases Mec1 and Tel1 regulate hyper‐phosphorylation of Pkc1 under conditions of genotoxic stress by inducing the association of CK1 homolog Hrr25 with Pkc1. This happens through a mech‐ anism that does not require the phosphorylation of Hrr25 by Mec1 or Tel1 and suggests that Hrr25 and Pkc1 are indirect effectors of the checkpoint kinases. Second, a large col‐ lection of CK1 phosphorylation sites contribute to the genotoxic stress‐induced Pkc1 band‐shift. Finally, CK1 phosphorylation site mutants in Pkc1 are partially deficient in *RNR3‐lacZ* basal and DNA damage‐induced expression, suggesting that Pkc1 hyper‐ phosphorylation by Hrr25 contributes to this response. We can draw several conclusions from this study. First, the checkpoint kinases Mec1 and Tel1 regulate hyper-phosphorylation of Pkc1 under conditions of genotoxic stress by inducing the association of CK1 homolog Hrr25 with Pkc1. This happens through a mechanism that does not require the phosphorylation of Hrr25 by Mec1 or Tel1 and suggests that Hrr25 and Pkc1 are indirect effectors of the checkpoint kinases. Second, a large collection of CK1 phosphorylation sites contribute to the genotoxic stress-induced Pkc1 band-shift. Finally, CK1 phosphorylation site mutants in Pkc1 are partially deficient in *RNR3-lacZ* basal and DNA damage-induced expression, suggesting that Pkc1 hyperphosphorylation by Hrr25 contributes to this response.

**Supplementary Materials:** The following are available online at www.mdpi.com/xxx/s1, Table S1: Pkc1 MS with HU only, Table S2: Pkc1 MS without HU only, Table S3: SILAC MS Pkc1 with and without HU, Figure S1: Binding of Hrr25 to Pkc1. Figure S2: Sensitivity of *hrr25* "gatekeeper" mu‐ tants to growth inhibition by inhibitory ATP analogs, Figure S3: Pkc1 phosphorylation sites identi‐ fied in this study. **Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10 .3390/jof7100874/s1, Table S1: Pkc1 MS with HU only, Table S2: Pkc1 MS without HU only, Table S3: SILAC MS Pkc1 with and without HU, Figure S1: Binding of Hrr25 to Pkc1. Figure S2: Sensitivity of *hrr25* "gatekeeper" mutants to growth inhibition by inhibitory ATP analogs, Figure S3: Pkc1 phosphorylation sites identified in this study.

**Author Contributions:** D.E.L., L.L., W.R. and E.M. contributed to the design of the experimental approach and the interpretation of data. Conceptualization, D.E.L. and W.R.; Methodology, D.E.L.,

L.L., W.R. and E.M.; Writing—Review & Editing, L.L., J.V., W.R., E.M., C.E.C., J.C.S., G.A. and D.E.L.; Supervision, Project Administration, and Funding Acquisition, D.E.L., G.A., C.E.C. and J.C.S.

**Funding:** This work was supported by grants from the NIH (R01 GM48533 to DEL; R01 GM129324 to JCS; and R24 GM134210 and S10 RR020946 to CEC). GA, JV, and WR were supported by the FWF

**Data Availability Statement:** The MS phospho‐proteomics data have been deposited at the zenodo

Austrian Science Fund Special Research Program F34.

All authors have read and agreed to the published version of the manuscript.

repository (https://zenodo.org/) and can be accessed via 10.5281/zenodo.5102666.

*J. Fungi* **2021**, *7*, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/jof

**Author Contributions:** D.E.L., L.L., W.R. and E.M. contributed to the design of the experimental approach and the interpretation of data. Conceptualization, D.E.L. and W.R.; Methodology, D.E.L., L.L., W.R., G.A. and E.M.; Software, J.V., W.R. and E.M.; Validation and Formal Analysis, D.E.L., L.L., W.R. and E.M.; Data Curation, L.L., J.V. and E.M.; Writing—Original Draft Preparation, D.E.L., L.L., W.R. and E.M.; Writing—Review & Editing, L.L., J.V., W.R., E.M., C.E.C., J.C.S., G.A. and D.E.L.; Supervision, Project Administration, and Funding Acquisition, D.E.L., G.A., C.E.C. and J.C.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by grants from the NIH (R01 GM48533 to DEL; R01 GM129324 to JCS; and R24 GM134210 and S10 RR020946 to CEC). GA, JV, and WR were supported by the FWF Austrian Science Fund Special Research Program F34.

**Data Availability Statement:** The MS phospho-proteomics data have been deposited at the zenodo repository (https://zenodo.org/ accessed on 10 September 2021) and can be accessed via 10.5281/ zenodo.5102666.

**Acknowledgments:** We thank David Hollenstein for general MS-related support, and Martha Cyert, Scott Emr, Juan Carlos Igual, Marcus Smolka, Stephen Elledge, and Gerhard Paravicini for yeast strains and plasmids.

**Conflicts of Interest:** The authors declare no conflict of interests. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**

