Targeting Cdc42 with the anticancer compound MBQ-167 inhibits cell polarity and growth in the budding yeast S. cerevisiae

ABSTRACT

The Rho GTPase Cdc42 is highly conserved in structure and function. Mechanical or chemical cues in the microenvironment stimulate the localized activation of Cdc42 to rearrange the actin cytoskeleton and establish cell polarity. A role for Cdc42 in cell polarization was first discovered in the budding yeast Saccharomyces cerevisiae, and subsequently shown to also regulate directional motility in animal cells. Accordingly, in cancer Cdc42 promotes migration, invasion, and spread of tumor cells. Therefore, we targeted Cdc42 as a therapeutic strategy to treat metastatic breast cancer and designed the small molecule MBQ-167 as a potent inhibitor against Cdc42 and the homolog Rac. MBQ-167 inhibited cancer cell proliferation and migration in-vitro, and tumor growth and spread in-vivo in a mouse xenograft model of metastatic breast cancer. Since haploid budding yeast express a single Cdc42 gene, and do not express Rac, we used this well characterized model of polarization to define the contribution of Cdc42 inhibition to the effects of MBQ-167 in eukaryotic cells. Growth, budding pattern, and Cdc42 activity was determined in wildtype yeast or cells expressing a conditional knockdown of Cdc42 in response to vehicle or MBQ-167 treatment. As expected, growth and budding polarity were reduced by knocking-down Cdc42, with a parallel effect observed with MBQ-167. Cdc42 activity assays confirmed that MBQ-167 inhibits Cdc42 activation in yeast, and thus, bud polarity. Hence, we have validated MBQ-167 as a Cdc42 inhibitor in another biological context and present a method to screen Cdc42 inhibitors with potential as anti-metastatic cancer drugs.

Introduction

Cdc42 (cell division control protein 42), a small Rho GTPase highly conserved in sequence and function, is established as a central regulator of polarization in cellular systems from fungi to human cells [1–6]. Localized activation of Cdc42 is an important step for eukaryotic cell polarization, and directs actin filaments to nucleate towards a determined site on the cell membrane [3,7–10]. Cdc42 is under robust spatial and temporal regulation; Cdc42 cycles between its active (GTP-bound) and inactive (GDP-bound) states in response to external signals, such as chemical gradients or physical contact that allow eukaryotic cells to orient themselves in space and respond to changes in their environment [1–3,6,11,12]. Cdc42 is negatively regulated by GTPase-activating proteins (GAPs) that are found in the cytoplasm maintaining global Cdc42 in its inactive form, and becomes locally activated by guanine nucleotide exchange factors (GEFs) that are released or activated upon certain stimuli [7,11,13]. Cell-surface receptors transduce such external stimuli by activating GEFs in their proximity. Consequently, Cdc42 becomes activated in a defined region of the inner-cellular membrane and induce the recruitment of the molecular machinery required for cellular polarization [1,2].

In eukaryotic cells, Cdc42 mediates events that cue the orientation of the cell, however the phenotypical manifestation of polarity varies depending on the species and cell type [2,14]. For example: Cdc42 is responsible for how epithelial cells orient themselves relative to other cells and differentially arrange their internal molecular components, in migrating fibroblasts and other mesenchymal cells for the establishment of directionality at the leading edge, in C. elegans zygotes for anterior-posterior intracellular organization, and in neurons for axonal growth [2]. In general, Cdc42 aids cells to asymmetrically distribute their molecular components and to generate coordinated responses to external stimuli depending on the tissue and/or their environment [1,2].

Cdc42 was first described in Saccharomyces cerevisiae, where it promotes cell polarity by directing the next budding event adjacent to the physical landmark left by the last cell division [3,15]. Cdc42 becomes activated by upstream Ras-like GTPase Rsr1 protein that in turn is activated by landmark proteins at the site of the last budding event [16,17]. Budding yeast cells share the same molecular mechanisms of polarization with other higher eukaryotes, including humans, and Cdc42 is pivotal for polarity determination, since it is conserved in the functional domains [18,19].

Rho GTPases are extensively studied because they are key regulators of a myriad of cellular processes, such as migration and invasion, cytoskeletal organization, transcriptional regulation, cell-cycle progression, apoptosis, vesicle trafficking, and cell-to-cell and cell-to-extracellular matrix adhesions [20]. Consequently, Cdc42 expression is upregulated ~ 5–20% in metastatic cancers, including common malignancies, such as melanoma, sarcoma, breast, bladder, pancreatic, cervical, and uterine carcinomas. Moreover, Cdc42 is not mutated in cancer, but is activated by oncogenic Ras and cell surface receptors, such as epidermal growth factor receptors (HER2 and EGFR) that ultimately activate Cdc42-specific GEFs [21–24]. Therefore, considering the central role of upstream effectors of Cdc42 in driving cancer progression associated with increased cancer cell survival, proliferation, and invasion/metastasis, the percentage of Cdc42 activated in invasive cancer is predicted to be much higher than the expressed levels.

Hyperactive Cdc42 promotes unregulated cell polarization events and forwards the process of cancer metastasis. Current treatments for advanced Stage IV (metastatic) breast cancer target overexpressed receptors, such as HER2 that signal to Cdc42 and its close homolog Rac, but their effectiveness is limited by therapy resistance. Thus, due to the critical need for improved therapeutics for advanced metastatic cancer, we focused on the Cdc24-GEF interaction as a viable target to prevent Cdc42-induced cell polarization, migration and metastasis [20,25–27].

We developed the small molecule compound MBQ-167 [20] (Suppl. Figure 1) as a derivative of EHop-016 [25–28], the first Rac/Cdc42 inhibitor developed by our group, which blocks the interaction of the GEF Vav with Rac more efficiently than with Cdc42. MBQ-167 interacts more closely with the GEF binding domain of both Cdc42 and Rac in-silico. Thus, MBQ-167 inhibits the activation of Rac and Cdc42 with equal efficacy. MBQ-167 inhibits Rac activation at an IC₅₀ of 100 nM and Cdc42 activation with an IC₅₀ of 78nM in metastatic human breast cancer cell lines. Moreover, in a HER2 positive breast cancer xerograft mouse model, MBQ-167 inhibited mammary tumor growth, angiogenesis, and metastasis [20].

Figure 1.

Knock-down of Cdc42 expression in the Tet-off Cdc42 strain.

Cdc42 protein expression was determined following exposure for 12 hours to the vehicle control (2% DMSO), 10μg/mL DOX, 100μM MBQ-167 or the combination of 10μg/mL DOX and 100μM MBQ-167 in control and Tet-Off Cdc42 yeast strains. Cells were lysed and subjected to western blotting for Cdc42. Representative western blots (top) and quantification of the Cdc42 (21kDa) band normalized by the PGK control (45kDa) band is shown for N = 3. (A) The wild-type BY4742 strain with no Tet-off system. (B) The Tet-off-Cdc42 strain. The expression of Cdc42 was not affected in the wild-type strain BY4742 but was significantly reduced in the mutant Tet-off-Cdc42 strain when exposed to DOX alone (p = 0.0003) or in combination with MBQ-167 (p < 0.0001).

A proposed mechanism of action of MBQ-167 is that by blocking the activation of Cdc42, it inhibits cell cycle progression, and impedes the polarization of cancer cells and thus, their ability to migrate and invade other tissues. Besides its clear anti-oncogenic properties, MBQ-167 is the only Cdc42 inhibitor currently reported to be effective at nano-molar (nM) concentrations [21], which makes this novel compound a promising candidate for further pre-clinical investigation for metastatic cancer. Herein, MBQ-167 is further validated as an inhibitor of Cdc42 and its associated cell polarity phenotype, using conditional knockdown of Cdc42 in haploid yeast S. cerevisiae for comparison.

The genome of S. cerevisiae was the first to be sequenced and many functional genomic studies have been conducted using this yeast strain, making it one of the most characterized and simplest eukaryotic systems for easy and cost-effective genetic manipulation [29–32]. Furthermore, most cell cycle and DNA repair regulatory genes, including Cdc42, are vastly conserved among eukaryotic cell systems [30,33]. Budding yeast also has the unique property of existing in both haploid and diploid states, making it a valuable tool for the validation of molecular targets and the elucidation of the mechanism of action of chemotherapeutics [29–31].

Previously, our group reported the use of the strain BY4742 with a tetracycline-controlled transcriptional activation (Tet-off) system in the promoter region of the yeast Cdc42 gene to control the expression of Cdc42 [20]. As reported, knock-down of Cdc42 resulted in non-polar budding events, and increased random bud emergence in mother cells [20,34]. Herein, we further exploit this model of Cdc42-mediated cell polarity regulation to validate the anti-metastatic drug MBQ-167 as a Cdc42 inhibitor. We also present a cost-effective and fast screening method to test Cdc42 inhibitors with the potential for development as anti-metastatic cancer therapeutics.

Results

Effects of Cdc42 knockdown and pharmacological inhibition on cell growth

The Tet-off-Cdc42 strain was created by the insertion of a Tetracycline (Tet)-controlled transcriptional activation system to regulate the expression of the CDC42 gene in S. cerevisiae wild-type R1158 haploid strain. The endogenous Cdc42 promoter was replaced with a kan^R-TetO₇-TATA cassette allowing for the binding of the tTA transactivator when supplemented with DOX, as described in [35–37]. This inducible Tet-off system for the knock-down of Cdc42 expression was validated by western blotting of Cdc42 from budding yeast liquid cultures incubated for 24 hours with vehicle control, 10μM DOX, 100μM MBQ-167, or the combination of 10μM DOX and 100μM MBQ-167. The relative Cdc42 expression was not affected by either treatment in the control BY4742 strain (Figure 1(A)). In the mutant Tet-off-Cdc42 strain, Cdc42 expression was significantly reduced by treatment with either 10μM DOX or combination of 10μM DOX and 100μM MBQ-167, where MBQ-167 had no independent effects on Cdc42 expression (Figure 1(B)).

We next assessed the phenotypic response of impairing Cdc42 either by genetic or pharmacological methods. Consistent with the previous results on Cdc42 expression (Figure 1), the wild-type BY4742 strain was sensitive only to MBQ-167 and not to DOX. MBQ-167 treatment inhibited the growth rate and maximum growth of the wildtype in a concentration dependent and statistically significant manner with ~ 40% inhibition in growth rate and maximum growth at 100μM with no significant response to the addition of DOX (Figure 2(A,(C,(E)) relative to the control. This observed non-significant effect between combining MBQ-167 and DOX versus the same concentration of MBQ-167 alone, confirms that the wildtype strain was not sensitive to DOX.

Figure 2.

Effects of knocking-down Cdc42 or MBQ-167 treatment on the growth of budding yeast.

Growth curves where obtained by diluting log phase growing yeast cultures (OD₆₀₀ = 0.2) in 500μL YPD containing the following treatments: Vehicle control (2% DMSO), 10μg/mL DOX, 100μg/mL DOX, 50μM MBQ-167, 100μM MBQ-167, 200μM MBQ-167, combination of 10μg/mL DOX and 50μM MBQ-167, or combination 10μg/mL DOX and 100μM MBQ-167. (A-B) Growth curves for the (A) wild-type BY4742 and (B) Tet-off-Cdc42 strains were obtained by taking OD₆₀₀ readings every 15 min for a total of 18 hours at 30ºC with low shaking between each measurement for 3 biological replicates. (C-D) Growth rates, relative to vehicle controls, calculated from the slope. (E-F) The Maximum Growth from the averaged values of the last three OD₆₀₀ measurements for each sample, relative to vehicle. N = 7 ± S.E.; ns = p > 0.05, * = p ≤ 0.05, ** = p ≤ 0.01, *** = p ≤ 0.001, **** = p ≤ 0.0001.

As expected, the growth of the mutant Tet-off-Cdc42 strain was significantly sensitive to Cdc42 knockdown via DOX treatment, with an ~ 40% decrease in growth rate and maximum growth relative to the control in response to 10μg/ml DOX, and an ~ 53% inhibition in response to 100 μg/ml DOX (Figure 2(B,(D,(F)). Thus, the growth rate and maximum growth of the Tet-off-Cdc42 strain was significantly reduced when exposed to DOX suggesting a positive correlation between Cdc42 expression and cell growth. Similar to the effect in the wildtype strain, 100μM MBQ-167 significantly reduced the growth rate and maximum growth of the Tet-off-Cdc42 strain by ~ 50% in the absence of DOX. When combined, 100μM MBQ-167 and 10μg/mL DOX in the Tet-off-Cdc42 strain, resulted in a significant ~ 60% reduction in the growth rate and maximum growth relative to the control. This effect was slightly greater than the inhibitory effects of 100μM MBQ-167 or 10μg/mL DOX alone (Figures 2(D,(F) and S2). At a lower concentration (50μM), MBQ-167 demonstrated a statistically significant ~ 25% decrease in growth in the Tet-off-Cdc42 strain, which was enhanced in combination with 10μg/mL DOX to a significant ~ 40–50% reduction in the growth rate and maximum growth relative to the control treatment (Figures 2(D,(F) and S2).

Effects of Cdc42 knockdown and pharmacological inhibition on bud polarity

Since Cdc42 activity is essential for the determination of cell polarity, the polarity of wildtype and mutant strains were evaluated following vehicle or MBQ-167 treatment. Cell polarity was determined by staining budding “scars” with calcofluor to visualize chitin redistribution, as described in [38]. Since budding yeast cells exhibit spatial memory for bud replacement, in the axial budding of haploid cells, bud scars are contiguous with each other in a proximate parallel series indicative of the maintenance of cell polarity. This axial division pattern is characteristic of polar budding in the haploid S. cerevisiae, where all divisions occur towards the same pole of the cell [1,16,39]. Cells with budding scars arranged randomly or in opposite poles of the cell are considered non-polar (bipolar budding pattern is only characteristic in diploid yeast) [16,39].

Accordingly, when the haploid yeast strains were stained for chitin, following vehicle treatment, 71.2% of the Tet-off-Cdc42 strain and 90.0% of the wild-type BY4742 strain were classified with an axial budding pattern, thus maintaining the establishment of polarity during multiple cell cycles. Treatment with 100μM MBQ-167 reduced the percentage of cells with axial budding pattern to 23.3% in the Tet-off-Cdc42 strain and to 26.0% in the wild-type BY4742 strain (Figure 3(B,(C)). This ~ 80% inhibition in polarity was also reflected in the Tet-off-Cdc42 strain when Cdc42 expression was knocked down by exposure to 10μg/mL DOX (21.0% axial budding cells). Cell polarity was even more severe (13.3%) in the presence of combined MBQ-167 and Cdc42 knockdown, where polarity was inhibited by ~ 87% (Figure 3(A,(C)). Additionally, the group treated with the combination of MBQ-167 and DOX demonstrated a number of cells that were not dividing (no buds) and appeared to be larger in size (Figure 3(A)).

Figure 3.

Effects of knocking-down Cdc42 or MBQ-167 treatment on budding yeast cell polarity.

(A) Morphology and polarity phenotype of yeast cells (left, BY4742 wildtype strain; right, Tet-off Cdc42 strain) treated with DOX, MBQ-167, or combination were determined by staining bud scars with 1μg/mL calcofluor white. Images were taken at constant exposure and magnification (100x, bar = 5µm). (B,C) The percentage of polar cells as quantified for 100 cells per experiment for 3 biological replicates. Cells with localized, clear and consecutive budding scars were considered as polar and quantified blinded. (B) Wildtype BY4742 strain treated with vehicle or 100 μM MBQ-167. (C) Tet-off-Cdc42 strain treated with 10µg/mL DOX, 100µM MBQ-167, or a combination of 10µg/mL DOX (p < 0.0001) and 100µM MBQ-167 (p < 0.0001). N = 3 ± S.E.

A factorial analysis of variance (two-way ANOVA) was performed to compare how the Cdc42 knockdown or pharmacological inhibition of Cdc42 activity with MBQ-167 affected the growth and the polarity phenotype of budding yeast cells with the Tet-off system (Figure S2). Bonferroni’s postdoc test was performed to conduct a multiple comparison analysis between each experimental group for the growth curves and budding polarity experiments. Results revealed a significant interaction between treatments to generate the observed phenotypic responses of growth rate), maximum growth, and budding polarity inhibition (Figures 2, 2S, 3 and Table 1).

Table 1.

Bliss independence analysis of additivity.

Bliss Additivity Score = (Ea+ Eb)-(EaEb)
Combination	Growth Rate	Max Growth	Polarity Inhibition
Mix1	0.518	0.474
Mix2	0.720	0.678	0.903
Synergy Score = Bliss Additivity Score – Eab
Combination	Growth Rate	Max Growth	Polarity Inhibition
Mix1	−0.015	0.105
Mix2	0.132	0.111	0.091

Mix1 = MBQ-167 50μM + DOX 10μg/mL; Mix2 = MBQ-167 100μM + DOX 10μg/mL; Bliss Additivity Score reflects the expected relative effect of the combinations of MBQ-167 and DOX, as calculated by using the observed relative effects of individual MBQ-167 (E_a) and DOX (E_b). The Synergy Score subtracts the observed relative effect of each combination (E_ab) to the expected effect (bliss additivity score).

To determine if the observed inhibitory effects that DOX and MBQ-167 had on growth and budding polarity were both via targeting Cdc42 (downregulation or inhibition), we used the bliss independence approach [40–43]. The observed inhibitory responses of the combined treatments (MBQ-167 50μM or 100μM and 10μg/mL DOX) was very similar to the one expected by adding the observed inhibitory responses generated independently by each treatment (Figures 2 and S2), as indicated by the combination index. Synergy scores calculated from both experiments did not exceed the threshold for synergy of more than 0.15, indicating that the expected combined response (bliss independence score) was similar to the combined observed phenotypic responses in growth and budding polarity experiments. Moreover, the additivity score for growth rate, maximum growth, and polarity inhibition was < 1 for all categories, which also indicates no additivity (Table 1). Hence, the non-synergistic nature of the combined treatments suggests that MBQ-167 is potentially acting via inhibition of Cdc42 activation (or expression).

Inhibition of yeast Cdc42 activation by MBQ-167

To validate the Cdc42 inhibitory effect of MBQ-167, suggested by the growth and polarity experiments, we performed an affinity precipitation assay to isolate the active Cdc42-GTP present in cell lysates, as described in [44]. Relative to the groups exposed to the vehicle, when exposed to 100μM or 200μM MBQ-167, the amount of active Cdc42 was significantly reduced in both wildtype and Tet-off-Cdc42 budding yeast strains. This effect was concentration dependent, where 100μM MBQ-167 inhibited Cdc42 activation by about 30–40%, while 200μM MBQ-167 inhibited Cdc42 activation by ~ 70%, which parallels the observed reduction in growth (53% inhibition) and bud polarity (75–80%) seen with MBQ-167 treatment in both wild type and Cdc42 mutant strains (Figure 4). Therefore, the effect of MBQ-167 in reducing yeast cell growth and polarity is consistent with its ability to inhibit Cdc42 activation.

Figure 4.

Inhibition of Cdc42 activation in budding yeast by MBQ-167.

Pulldowns of active GTP-bound Cdc42 in (A) wild type, (B) Tet-Off Cdc42 strain following vehicle, or 100 or 200mM MBQ-167 for 24 h. Top panels show representative western blots. Bottom panels show quantification of positive bands from western blots of relative Cdc42 activation calculated by dividing the integrated density of active Cdc42 from the integrated density of total Cdc42 from the same cell lysates. N = 4 ± S.E., p < 0.0001.

Discussion

The objective of this study was to demonstrate the specificity of our recently characterized Rac/Cdc42 inhibitor MBQ-167 [20] as a specific inhibitor for Cdc42 activation, using yeast as a model system. As previously described [35–37], the conditional knock-down of Cdc42 expression in haploid budding yeast, through the use of a titratable promoter, can be used to investigate the function of Cdc42, an essential gene for cell growth. This strategy was used for a comparative analysis of the phenotype obtained from genetically reduced Cdc42 expression versus the inhibitory effect of a small molecule compound designed to block Cdc42 activation.

To determine the relationship between the cell polarity phenotype and the associated effect on cell division and growth of budding yeast, the morphology of cell polarization in wildtype and Cdc42 knocked down cells were assayed by evaluating the budding pattern. Under normal circumstances, most haploid yeast cells will organize new budding events to occur next to where the last cell division occurred or towards the same pole of the mother cell where the last bud emerged; this coordination is dependent on establishment and maintenance of cell polarity by Cdc42, and is known as the axial budding pattern [15,16]. Localized activation of Cdc42 adjacent to the last cell division event is known to be a key step for cells to maintain their polarity by orienting themselves in space and direct the polymerization of actin fibers towards the site where the new bud (daughter cell) will emerge [3,45,46]. As expected, knock-down of Cdc42 expression decreased the number of polar cells showing an axial-budding pattern. In both strains of budding yeast tested, MBQ-167 had a similar effect as the knock-down mutant group, resulting in more cells with random or non-organized bud “scars”. When Cdc42 knock-down cells were treated with MBQ-167, the reduction of polarity or axial budding was greater than under individual conditions of reduced Cdc42 expression or inhibition of Cdc42 activation.

Excessive downregulation of Cdc42 expression or activity compromises the efficiency and frequency of cell division, thus explaining the observed reduction in growth rate and maximum growth under conditions of Cdc42 knock-down or pharmacological inhibition. As reported previously, using temperature-sensitive Cdc42 yeast S. cerevisiae mutants, conditional downregulation of Cdc42 expression halted budding, but the cells continued to grow in size [15]; as observed here when knocking down Cdc42 or exposing the yeast cells to MBQ-167. Moreover, Cdc42 activity has been shown to be crucial for the frequency, establishment, and maintenance of bud emergence within a cell cycle [13,34,47]. Therefore, the observed reduction of growth and the appearance of multiple, random and un-defined buds, could be attributed to the loss of Cdc42 activity.

Data presented herein, on the effect of Cdc42 on budding yeast growth, show that genetic knock-down of Cdc42 or treatment with the Cdc42 inhibitor MBQ-167 has similar effects on the growth and polarity of budding yeast. When cells with Cdc42 knockdowns were treated with MBQ-167, these inhibitory effects were significantly enhanced but not in an additive or synergistic manner. The non-synergistic interaction between treatments was validated by the bliss independence scores, since the observed inhibitory phenotypic responses were not significantly greater than the one predicted by the low synergy scores (< 0.15) [41,43]. This suggests that the interaction observed between Cdc42 knockdown and MBQ-167 treatment are probably due to both modes targeting the same protein, and can be interpreted as a possible competitive interaction between the treatments [40,42,48]. Therefore, taken together, our analyzes suggest that MBQ-167 directly targets Cdc42; however, the significant increase in the response of the cells when MBQ-167 is administered to cells with Cdc42 knockdown may indicate potential off-target effects, or that the Cdc42 knockdown with the Tet-off system was not complete and MBQ-167 may be inhibiting the residual wildtype Cdc42.

To directly validate the inhibition of Cdc42 by MBQ-167, we performed a pull-down assay for active Cdc42-GTP in the presence and absence of MBQ-167 and show that MBQ-167 inhibits the activity of Cdc42 in a concentration dependent manner. As hypothesized, similar concentrations of MBQ-167 reduced the growth and polarity of budding yeast S. cerevisiae in wild type and Cdc42 knock-down strains.

As shown by our in-silico [20] and binding studies using a Cdc42 nucleotide free mutant form (G15A) (data not shown), MBQ-167 is predicted to interact closely with GEF interactive switch I and II domains of Cdc42. This prediction was validated in human breast cancer cells by demonstrating direct inhibition of Cdc42 activation with an IC₅₀ of 78 nM, which is a pharmacologically relevant concentration for Cdc42 inhibition. In this study, we validated the direct inhibition of Cdc42 by MBQ-167 in budding yeast, where Cdc42 plays a central regulatory role in bud polarity and cytokinesis, and thus cell growth. Therefore, we conclude that MBQ-167 acts as a specific inhibitor of Cdc42 activation in eukaryotic cells and has potential for future development as a targeted inhibitor to treat advanced cancers with Cdc42 dysregulation. The experimental strategies described herein using budding yeast, can also be used as a fast and inexpensive method for the pre-clinical screening of putative Cdc42 inhibitors.

Materials and methods

Treatments

MBQ-167 (PCT/US17/29921 patent application) and its derivatives were synthetized by our medicinal chemist collaborators as previously described and dissolved in DMSO to obtain a 10mM stock solution [20]. Doxycycline hyclate (Sigma cat# D9891) was dissolved in deionized water for a stock solution of 5mg/mL. Treatments for experiments were prepared by diluting stocks in autoclaved Yeast Extract-Peptone-Dextrose (YPD) Broth (BD cat#242820).

Budding yeast strains

Saccharomyces cerevisiae wild-type BY4742 haploid strains served as controls. Knock-down of Cdc42 was achieved by insertion of a tetracycline (tet)-controlled transcriptional activation system to control the expression of the CDC42 gene. The mutant strain R1158 was obtained from Dharmacon’s yeast Tet-Promoters Hughes Collection (yTHC), generated from the wild-type MAT, a haploid strain BY4741 by one-step integration of the tTA transactivator controlled by a CMV promoter at the URA3 locus [35,37]. To control the expression of the CDC42 gene, its endogenous promoter was replaced with a kan^R-TetO₇-TATA cassette allowing for the binding of the repressor protein, tTA transactivator, by supplementing the media with DOX (ED₅₀ = 10µg/mL) as described in [35–37]. The mutant R1158 strain containing the tetracycline (tet)-regulated promoter for the knock-down of Cdc42 is referred here as the Tet-off-Cdc42 strain. To validate the knocking-down of Cdc42 with DOX western blots of Cdc42 were performed from budding yeast incubated (30ºC) for 12 hours in the vehicle control (2% DMSO), 10μg/mL DOX, 100μM MBQ-167 or the combination of 10μg/mL DOX and 100μM MBQ-167 (Figure 1).

Yeast cultures

Colonies of BY4742 and Tet-off-Cdc42 strains were grown in agar plates of YPD media, incubated at 30ºC for 2 days and stored at 4ºC up to one month. A colony from each strain was grown overnight in YPD liquid media at 30ºC with shaking before each experiment. The strain Tet-off-Cdc42 was grown in YPD supplemented with 300μg/mL geneticin (Sigma-Aldrich, Inc. cat#G8168) to select for cells resistant to kanamycin conferred by the integration of kan^R-TetO₇-TATA cassette in the genome. Absorbance at 600nm (OD₆₀₀) was monitored to follow the growth of yeast cells in liquid culture (standard spectrophotometer with 1cm diameter of cuvette). Before each assay, each liquid culture was diluted to an OD₆₀₀ = 0.1 – 0.2 and incubated for 1–2 hours until they entered the exponential phase of growth (OD₆₀₀ = 0.4 – 0.6); where they are growing at their fastest rate. Their concentration was adjusted to approximately OD₆₀₀ = 0.2 (1.02x10⁸ cell/mL) prior each experiment.

Growth curves

Growth curves of mutant Tet-off-Cdc42 and wild-type BY4742 strains were performed using 48-well plates (Thermo-Fisher, Inc. cat#08-772-3D) filled with 500μL of YPD liquid media supplemented with the correspondent treatment. Some wells contained only YPD media without treatment or cells to detect for possible contamination and/or correct for the evaporation of the media [49]. About 5.1x10⁶ cells were seeded in each well by transferring 50μL from the liquid culture [50]. Each strain was exposed to different conditions in technical triplicates: Vehicle control (2% DMSO), 10μg/mL DOX, 100μg/mL DOX, 50μM MBQ-167, 100μM MBQ-167, 200μM MBQ-167, combination of 10μg/mL DOX and 50μM MBQ-167, or combination 10μg/mL DOX and 100μM MBQ-167. A microplate spectrophotometer (Bio-Rad Benchmark Plus) was used to monitor the growth of yeast while incubating at 30ºC. The instrument was programed to read the optical density at wave length 600nm (OD₆₀₀) every 15 min for 18 hours, and 10 min of low shaking before each measurement [49–51]. To generate the growth curves, the OD₆₀₀ reading at each time point within a treatment group was divided by their correspondent initial OD₆₀₀ (OD/OD_t=0) to normalize each group to be compared [49]. Additionally, to correct for factors such as media evaporation, OD₆₀₀ readings of wells with only YPD media (no yeast cells) were subtracted from the OD₆₀₀ values of each group at the same time point [51,52]. The growth rate was determined from the slope of the exponential phase of each curve, and the maximum growth from the averaged last time point values of each curve [50].

Calcofluor staining of budding “scars”

To determine the polarity of budding yeast we used the fluorophore Calcofluor White (Sigma-Aldrich, Inc. cat#18909) which allowed the staining of chitin depositions that after cell division are left in the cell wall of the mother cell outlining the last site of cell division [16,38,53]. First, budding yeast cultures were exposed for 12 hours to 2% DMSO (vehicle control), 10μg/mL DOX, 100μM MBQ-167, or the combination of 10μg/mL DOX and 100μM MBQ-167. Then, each yeast sample was incubated with YPD media containing calcofluor (50μg/mL) for 1 hour and images were taken by fluorescent microscopy (see below). Quantification of polar cells was done by quantifying the number of polar cells per 100 cells counted in the images taken by a blinded researcher. Cells with defined, localized, and consecutive “bud scars” were counted as polar and diffused and spread scars as non-polar, as described in [16,39].

Fluorescent microscopy

Images where taken using DAPI filter with a magnification of 100x. The same exposure time was used for all the images.

Protein extraction

BY4742 and Tet-off-Cdc42 strains were incubated for 24 h at 30ºC in 30 mL of liquid YPD media containing the vehicle or treatments. After they reached an OD > 1.0 a pellet of yeast cells was obtained by centrifugation (8,000 g) for 10 min at 4ºC. The supernatant was discarded, and cells where lysed by adding 1.0mL of ice-cold glass beads (Sigma cat# G8772) and 1.0mL of lysis buffer [25mM Tris-HCl (pH 7.5), 1mM DTT, 30mM MgCl₂, 2M NaCl, 0.5% (v/v) Nonidet P-40, leupeptin (1μg/ml), aprotinin (1μg/ml), and 1mM PMSF] [54]. Then vortexed three times for 1 min maintaining samples at 4ºC, followed by centrifugation (8,000 g) for 10 min to obtain a pellet with of glass beads and cell debris. The supernatant or protein extract was transferred to a new 2.0mL centrifuge tube, then 5μL from each extract was aliquoted to be used for protein quantification using the DC Protein Assay (Bio-Rad cat#500-0112). Samples were stored at −80ºC.

Pulldown assays

The Cdc42 activity assay consisted of a pull-down of the active Cdc42-GTP using the P21-binding domain (PBD) of PAK fused to glutathione S-transferase (GST) agarose beads (Cell Biolabs, Inc. cat# STA-411) incubated for 1 hour with 1.5mg of protein from each lysate at 4ºC as described previously [20]. The same lysis buffer used for protein extraction was used to incubate the beads with the lysates, and the washing buffer [25mM Tris-HCl (pH 7.6), 1mM DTT, 30mM MgC12, 40mM NaCI, 1% (v/v) Nonidet P-40] [54]. Agarose Beads contained the human p21-binding domain (PBD) of the Cdc42 downstream effector p21-activated kinase (PAK), where active Cdc42-GTP is expected to selectively bind the PBD-GST complex. The orthologue of the Cdc42 effector PAK in budding yeast is Ste20, which is highly conserved in the PBD, thus allowing for measuring Cdc42 activation in yeast as demonstrated by others [18,55].

Western blots

Western blots of total cell lysates or pull-downs were preformed to evaluate changes in the total expression and activity of Cdc42. PGK was used as a control to normalize each sample signal. Nitrocellulose membranes and the Odyssey system (LI-COR, Inc.) used for the western blots following standard procedures. Primary antibodies: Cdc42 (ABCAM cat#2466S) and PGK (Invitrogen cat#459250). Secondary antibodies: IRDye® 800CW anti-Rabbit and IRDye® 680RD anti-Mouse (LI-COR, Inc. cat# 926–32211/926–68070).

Data analysis and statistics

Each experiment was conducted with 3 to 7 biological replicates. Data was represented relative to the average of vehicle controls. Statistical probability was calculated from ANOVAs with a confidence interval of 95% and significant differences at P ≤ 0.05. A factorial analysis of variance (two-way ANOVA) was also performed to compare the effects of Cdc42 knockdown or pharmacological inhibition of Cdc42 activity with MBQ-167. Bonferroni’s postdoc test was performed to conduct a multiple comparison analysis between each experimental group.

The bliss independence approach was used to assess possible additivity between knockdown and pharmacological inhibition by MBQ-167 [40–43]. The expected inhibitory effect of the combination of DOX and MBQ-167 was calculated as in [43]. The independent observed effects of DOX (E_DOX) and MBQ-167 (E_MBQ) were used to obtain the expected additive effect or Bliss additivity score = [(1-E_DOX)+(1-E_MBQ)]-(E_DOXE_MBQ). A synergy score was obtained by subtracting the observed effect of the combined treatments (E_DOX+MBQ) to the Bliss additivity score: Synergy score = Bliss additivity score – E_DOX,MBQ. In this method, high synergy scores (> 0.15) are considered as indicative of synergistic or non-additive interactions, while a score of ~ 1 signifies additivity.

Funding Statement

This work is partially supported by grants from the Susan G Komen for the Cure, the Puerto Rico Science and Technology Research Trust, NIH/NIGMS SC3GM084824, NIH/NCI U54 CA096297/09629300 (S.D.), NIH/NIGMS P20GM103475 (JRRM), NIH/NIGMS-RISE R25 GM061838 (MJR), and Linda and Jack Gill Endowment (BR).

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplementary data for this article can be accessed here.