• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Mol Pharmacol. Apr 2009; 75(4): 866–875.
Published online Dec 9, 2008. doi:  10.1124/mol.108.049809
PMCID: PMC2684929

Sphingolipids Function as Downstream Effectors of a Fungal PAQR[S with combining enclosing square]


The Izh2p protein from Saccharomyces cerevisiae belongs to the newly characterized progestin and adipoQ receptor (PAQR) superfamily of receptors whose mechanism of signal transduction is still unknown. Izh2p functions as a receptor for the plant PR-5 defensin osmotin and has pleiotropic effects on cellular biochemistry. One example of this pleiotropy is the Izh2p-dependent repression of FET3, a gene involved in iron-uptake. Although the physiological purpose of FET3 repression by Izh2p is a matter of speculation, it provides a reporter with which to probe the mechanism of signal transduction by this novel class of receptor. Receptors in the PAQR family share sequence similarity with enzymes involved in ceramide metabolism, which led to the hypothesis that sphingolipids are involved in Izh2p-dependent signaling. In this study, we demonstrate that drugs affecting sphingolipid metabolism, such as d-erythro-MAPP and myriocin, inhibit the effect of Izh2p on FET3. We also show that Izh2p causes an increase in steady-state levels of sphingoid base. Moreover, we show that Izh2p-independent increases in sphingoid bases recapitulate the effect of Izh2p on FET3. Finally, our data indicate that the Pkh1p and Pkh2p sphingoid base-sensing kinases are essential components of the Izh2p-dependent signaling pathway. In conclusion, our data indicate that Izh2p produces sphingoid bases and that these bioactive lipids probably function as the second messenger responsible for the effect of Izh2p on FET3.

The PAQR superfamily of proteins is composed of membrane-bound receptors (Tang et al., 2005) and includes two mammalian proteins that function as receptors for the hormone adiponectin, which represents a molecular link between obesity and type II diabetes. Other members of this family sense and respond to progesterone, making them excellent candidates for the long-sought-after receptors that mediate many of the rapid nontranscriptional effects of progesterone. Because of arguable similarity to G-protein-coupled receptors, it has been proposed that PAQRs function as a new class of GPCRs (Thomas et al., 2007). However, this assignment is highly debatable, and, in reality, almost nothing is known about how these proteins transduce signals. What is needed to explore this complicated question is a simple system with which to study the functionality of PAQR receptors.

We chose to use the model eukaryote, Saccharomyces cerevisiae, to study these receptors for several reasons. First, the S. cerevisiae genome encodes 4 proteins—Izh1p, Izh2p, Izh3p, and Izh4p—in the PAQR superfamily of receptors (Lyons et al., 2004), and the study of yeast orthologs as models for human proteins has been a remarkably fruitful method for characterizing poorly understood protein families (Foury, 1997). Second, the plant protein osmotin has been identified as an agonist for Izh2p (Narasimhan et al., 2005), allowing us to probe receptor activation in this organism. Third, the simplicity and tractability of S. cerevisiae as a genetic system make it ideal for mapping signal transduction pathways. Indeed, S. cerevisiae has already been successfully used to probe the mechanism of signal transduction by a variety of receptor types (McEwan, 2001).

To map a biochemical pathway, it is helpful to have a measurable phenotype associated with a functional receptor that can be exploited. While attempting to elucidate the physiological role of Izh2p, we discovered that Izh2p negatively regulates the expression of a gene involved in iron uptake called FET3 (Kupchak et al., 2007). The reason for Izh2p-dependent regulation of FET3 is still poorly understood; however, it is not critical to the goals of this study to understand why Izh2p regulates FET3. What is important is that we can use the effect of Izh2p on FET3 as a reporter for receptor activity and as a tool to map the pathway of signals emanating from this receptor.

We already used this system to partially map the signal-transduction pathway and determined that Izh2p overexpression represses FET3 by regulating transcription factors that exert opposing effects on a regulatory element in the FET3 promoter. In addition, we determined that cAMP-dependent protein kinase (PKA) was essential for these effects (Kupchak et al., 2007). Most importantly, however, we demonstrated that the PAQR-dependent signaling pathway in yeast does not involve Gα proteins, indicating that PAQRs are not likely to be GPCRs (Smith et al., 2008). Hence, we have already demonstrated the efficacy of this system for testing models of signal transduction and identifying novel players in the pathway. We also demonstrated that human PAQRs, when expressed in yeast, repress FET3 in response to activation by their respective agonists (Kupchak et al., 2007; Smith et al., 2008), thus demonstrating that the mechanism of signal transduction may be conserved in the PAQR family.

The question remains that if PAQRs are not GPCRs, then what is the chemical second messenger produced by these receptors? A thorough analysis of the sequence of PAQR receptors reveals significant similarity to a group of enzymes known as alkaline ceramidases. Ceramidases hydrolyze ceramides to generate free fatty acids and sphingoid bases, which are a class of bioactive aminolipids known to function as potent second messengers. Indeed, the ceramide-sphingoid base ratio is used as a rheostat in eukaryotic cells to regulate a variety of important physiological functions (Kobayashi and Nagiec, 2003). In this study, we demonstrate that Izh2p overexpression elicits the accumulation of sphingoid bases and that sphingoid bases are capable of independently activating the signaling pathway downstream of Izh2p. This accumulation requires de novo sphingoid base and ceramide biosynthesis. Moreover, we demonstrate that FET3 repression by Izh2p can be prevented by d-erythro-MAPP, an alkaline ceramidase inhibitor. Finally, we demonstrate that Izh2p-dependent FET3 repression requires a pair of sphingoid base-sensing kinases. Based on these findings, we propose a model in which Izh2p in particular, and perhaps PAQR receptors in general, generate sphingoid base second messengers.

Materials and Methods

Yeast Strains, Plasmids, and Materials. Unless specifically noted, all wild-type and single mutant strains were purchased from Euroscarf (Frankfurt, Germany) and are in the BY4742 background (MATα; lys2Δ; ura3Δ; his3Δ; leu2Δ; kanMX4 deletion cassette). The ypc1Δydc1Δ double mutant (MATα; ade2; his3; leu2; trp1; ura3; can1; ypc1::LEU2; ydc1::TRP1) and the isogenic W303-1A wild type (MATα; ade2; his3; leu2; trp1; ura3; can1) background strains were a gift from Dr. Howard Riezman at the University Basel, Switzerland (Schorling et al., 2001). FET3-lacZ was used as a FET3 promoter/lacZ reporter plasmid and was described previously (Kupchak et al., 2007). IZH2 and YPC1 were cloned into pRS316-GAL1 via gap repair as described previously (Lyons et al., 2004). Primers for cloning the YPC1 gene into pRS316-GAL1 by gap repair are shown in Supplemental Data 1. Fumonisin B1, myriocin, d-erythro-MAPP, and N,N-dimethylsphingosine were purchased from Cayman Chemical (Ann Arbor, MI). Phytosphingosine, dihydrosphingosine, and C2-phytoceramide were purchased from Avanti Polar Lipids (Alabaster, AL). Stearylamine and thaumatin were purchased from Sigma-Aldrich (St. Louis, MO). NBD-C12-phytoceramide and NBD-dodecanoic acid were provided by Dr. Yusuf Hannun. In this article, we discuss many yeast genes with which a reader who is not in the field may not be familiar. To make it easier to read this article, we have included a brief description of these genes and their human homologs in Supplemental Table 1.

Growth Conditions and Biochemical Assays. For β-galactosidase and ferroxidase assays, cells were pregrown in synthetic defined medium with 2% glucose as the carbon source and reinoculated into low iron medium (LIM) containing galactose to induce the expression of genes driven by the GAL1 promoter. Iron deficiency and iron repletion were generated by adding FeCl3 back to LIM at either 1 μM or 1 mM final concentrations, respectively. Galactose (2%) was used to fully induce GAL1-driven genes, whereas modest overexpression was achieved by adding 0.05% galactose and 1.95% raffinose to the medium. Cells were allowed to grow to mid-log phase in LIM, and β-galactosidase or ferroxidase assays were performed as described previously (Kupchak et al., 2007). For individual experiments, each data point was done in triplicate, and the error bars represent ± 1 S.E.

Fresh stocks of phytosphingosine (PHS), dihydrosphingosine, stearylamine, d-erythro-MAPP, N,N-dimethylsphingosine (N,N-DMS), fumonisin B1 (FB1), and myriocin were made by dissolution in ethanol (d-erythro-MAPP, myriocin, N,N-DMS) methanol (FB1), or 1:1 CHCl3/MeOH (PHS, dihydrosphingosine, and stearylamine). For all experiments, equal volumes of the appropriate solvent were added to untreated cells to control for vehicle effects. For all cases except myriocin, the drug was added upon reinoculation of cells into LIM. In the case of myriocin, cells were treated with the drug for 1 h before β-galactosidase assay. ED50 values were obtained using the web-based BioDataFit software using sigmoidal curve-fitting (available at http://www.changbioscience.com/stat/ec50.html).

Deletion of the multidrug resistance transporter, Yor1p, causes cells to be hypersensitive to the antifungal drug fumonisin B1 (FB1) because of the inability to export the toxic compound (Mao et al., 2000a). For FB1 sensitivity experiments, yor1Δ cells were grown in synthetic defined medium containing 2% galactose. FB1 (from DMSO stocks) was added to a final concentration of 100 μM. Cells not treated with FB1 were treated with an equivalent volume of DMSO vehicle. For each experiment, the growth of three independent colonies of each strain was tested. Error bars show ± 1 S.E.

Sphingolipid Extraction and Analysis. Overnight cultures of yeast were inoculated into synthetic defined medium containing 2% galactose, allowed to grow to mid-log phase, and harvested at OD600 = 0.8. Cells treated with myriocin, FB1 (from DMSO stocks), or C2-phytoceramide (from ethanol stock) were so-treated for 30 min before extraction. Untreated controls were exposed to an equal volume of either DMSO or ethanol. Sphingolipids were extracted using a modified Bligh-Dyer protocol (Bligh and Dyer, 1959). Cell pellets from 30 ml of culture were vortexed with 3 ml of CHCl3/MeOH (1:2) to extract total lipids, after which 0.8 ml of water was added. The suspension was transferred to a screw-capped glass tube and left at 4°C overnight. Cellular debris was pelleted at 3000 rpm for 10 min. The supernatant was transferred to a new glass tube and vortexed with 2 ml of CHCl3/H2O (1:1). After 30 min, the mixture was centrifuged for 5 min at 3000 rpm to obtain a clean-phase separation. The organic phase was quantitatively transferred to a new glass vial and dried down using a nitrogen evaporator. Dried lipids were dissolved in 300 μl of CHCl3, of which 100 μl was transferred to another tube for phosphate determination. A 0.8-ml portion of 0.125 M methanolic KOH was added to the remaining 200 μl, and the mixture was heated at 37°C for 75 min. After cooling to room temperature, 1400 μl of CHCl3 (to make a 2:1 ratio to MeOH), 200 μl of 0.3 M HCl, and 400 μl of a 1 M NaCl/5% glycerol solution were added. After centrifugation for 10 min at 3000 rpm, the organic phase was transferred to a new tube, dried as before, and stored at -20°C until ready for HPLC analysis.

Sphingolipids were labeled with ortho-phthalaldehyde and quantified by HPLC using established protocol (Merrill et al., 2000) with l-threo-dihydrosphingosine as an internal standard to control for variations in extraction efficiency. Lipid pellets were dissolved in 100 μl of MeOH. Fresh ortho-phthalaldehyde reagent (50 μl) was added to 50 μl of lipid extract solution and allowed to react for 20 min in the dark at room temperature. Column buffer (300 μl; 90:10 MeOH/5 mM KHPO4, pH 7.0) was added and loaded onto the HPLC. Total inorganic phosphate was determined using a modification of a published method (Merrill et al., 2000). Dried extracts of total lipids were dissolved in 0.6 ml of ashing buffer (10 N H2SO4/70% HClO4/H2O, 9:1:40) and heated at 160°C until 100 μl remained. To this was added 0.9 ml of H2O, 0.5 ml 0.9% (w/v) ammonium molybdate, and 0.2 ml of 9% (w/v) ascorbic acid with vortexing between each addition. This mixture was incubated for 30 min at 45°C, and inorganic phosphate was determined by measuring absorption at 820 nm. For individual experiments, each data point was done in duplicate, and the error bars represent ± 1 S.E. Representative experiments are shown. Paired Student's t tests were performed using Excel (Microsoft, Redmond, WA) only when the difference between two means was small or significance is not easily inferable by error bars.

Preliminary Ceramidase Assays. Ceramidase activity was determined by the release of NBD-labeled fatty acid from the fluorescent substrate NBD-C12-phytoceramide. In brief, 20 μl of total membrane preps were resuspended in buffer B (25 mM Tris-HCl at three different pH levels, 8.56, 9.4, and 7.0, and 0.5 mM CaCl2) and mixed with 20 μl of substrate dissolved in buffer B with 0.4% Nonidet P-40. After incubation at 30°C for 60 min, the reactions were stopped by boiling for 5 min and dried in a heating block at 80°C for 20 min in the dark. Reaction mixtures were dissolved in 30 μl of chloroform/methanol [2:1 (v/v)]. A 25-μl portion of each sample was applied to a silica gel 60A TLC plate (Whatman, Clifton, NJ) and resolved by a solvent system composed of chloroform/methanol/4.2 N ammonium hydroxide [90:30:0.5 (v/v)]. NDB-C12-phytoceramide and NBD-dodecanoic acid were run as controls. Fluorescent lipids were detected by scanning the TLC plate in a Storm 860 Chromatoscanner (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK), operated in the fluorescent mode, with an excitation wavelength of 450 nm and an emission of wavelength of 525 nm.

Sequence Analysis. Multiple sequence alignments were produced using ClustalX (Thompson et al., 1997) with default parameters. Alignments shown in Supplemental Figs. Figs.11 to to33 were performed manually. Hydropathy plots were first generated with TopPredictII 1.2 (Claros and von Heijne, 1994) using the Kyte-Doolittle algorithm and default parameters. Data were downloaded into a spreadsheet, and the hydropathy values were aligned based on the multiple sequence alignment produced by ClustalX. Average hydropathy across the entire set of aligned proteins was then calculated. Sequences used for these comparisons are listed in the Supplemental Data 1.

Fig. 1.
Thaumatin represses FET3 expression via Izh2p. In A and B, cells were grown in iron-limited LIM containing 0.05% galactose to modestly overexpress IZH2 driven by the GAL1 promoter. Cell surface ferroxidase activities are plotted as a percentage of ...
Fig. 3.
Involvement of sphingoid bases in Izh2p-dependent signaling. Cells were grown in iron-deficient LIM containing 2% galactose to induce GAL1 driven genes. A, Steady-state levels of phytosphingosine in WT (BY4742) cells carrying either empty expression ...


Izh2p Activation by PR-5 Defensins. Before communicating the results, it is important to recap what is known about the link between Izh2p and FET3. FET3 encodes a ferroxidase involved in iron uptake whose expression is induced by iron limitation. We discovered previously that Izh2p overexpression inhibits the ability to induce FET3 expression during iron limitation and that this effect is not seen in strains lacking an isoform of PKA (Tpk2p) or either of a pair of transcriptional repressors called Nrg1p and Nrg2p (Kupchak et al., 2007).

These published experiments involved the overexpression of the Izh2p receptor, not its activation. We interpreted the data to mean that Izh2p, like many other signaling proteins, has basal signaling activity. Herein, we confirm that Izh2p overexpression and Izh2p activation are mechanistically similar at low concentrations of agonist. Because Izh2p is activated by the plant protein osmotin, it follows that osmotin should repress FET3 in an Izh2p-dependent manner when Izh2p is expressed at low levels. Unfortunately, osmotin is not commercially available. Thaumatin, on the other hand, is a PR-5 defensin from the plant Thaumatococcus danielli that is sold as an industrial sweetener (Min et al., 2004). A pair-wise comparison of the osmotin and thaumatin mature peptides using Blast 2 Sequences (Tatusova and Madden, 1999) reveals an expect value of 10-59, confirming that the two proteins are closely related. Therefore, we decided to test whether thaumatin could substitute for osmotin as an activator of Izh2p.

When thaumatin is added to wild-type cells carrying only the genomic copy of IZH2, there is a dose-dependent repression of FET3 as shown by a reduction in cell-surface ferroxidase activity (Fig. 1A, □). The ED50 value for this effect is approximately 3.6 μM. When Izh2p dosage is reduced by deleting the IZH2 gene, FET3 expression is desensitized to the effects of thaumatin (ED50 ~8.1 μM; Fig. 1A, [filled square]). The dosage of IZH2 can be increased by transforming wild-type cells with a plasmid bearing the IZH2 gene driven by the GAL1 promoter. When these cells are grown in 2% galactose, Izh2p is fully induced, and FET3 is constitutively repressed (Kupchak et al., 2007). However, when these cells are grown in 0.05% galactose/1.95% raffinose, Izh2p expression is increased above normal but is not high enough to constitutively affect FET3. Under these latter conditions, in which Izh2p levels are slightly elevated, FET3 expression is sensitized to the effects of thaumatin (ED50 ~0.1 μM; Fig. 1A, [open triangle]) Moreover, the repression of FET3 caused by 500 nM thaumatin in cells carrying slightly elevated levels of Izh2p did not occur in strains lacking Tpk2p, Nrg1p, or Nrg2p (Fig. 1B).

These findings indicate that Izh2p partially mediates the effects of thaumatin on FET3. This is consistent with the fact that the antifungal effects of osmotin, another PR-5 defensin, are only partially mediated by Izh2p (Narasimhan et al., 2005) and with a model in which Izh2p functions as a general PR-5 defensin receptor rather than as a specific osmotin receptor. More importantly, these data confirm that Izh2p overexpression and Izh2p activation repress FET3 by similar mechanisms. However, they also show that we cannot use thaumatin treatment as a model for Izh2p activation because thaumatin clearly has Izh2p-independent effects. For this reason, unless otherwise noted, we use full Izh2p overexpression (in 2% galactose) instead of activation as a model to isolate the effects of this receptor on downstream signaling pathways.

The PAQR Family of Receptors Shares Similarity with Alkaline Ceramidases. The PAQR family can be divided into three classes based on sequence alignments; however, the entire family is unified by the presence of a core of seven transmembrane domains (TMs) (Fig. 2A) and three highly conserved, albeit short, amino acid motifs (Lyons et al., 2004). A multiple sequence alignment of the regions containing these motifs can be found in Supplemental Figs. Figs.1, 1, ,2,2, and and3.3. Motif A consists of Ex2-3Nx3H and can be found N-terminal to TM1. Motif B consists of an Sx3H motif at the end of TM2 and an aspartic acid residue at the beginning of TM3. Motif C consists of an Hx3H motif in the loop region between TM6 and TM7. Figure 2C shows the predicted topology of the PAQR family and the location of these motifs.

Fig. 2.
Hydropathy plots of PAQRs receptors and alkaline ceramidases. In A and B, hydropathy plots for individual proteins were generated and aligned (broken lines). An average hydropathy plot for all proteins plotted in A and B was generated (solid line). ...

Searches of the National Center for Biotechnology Information database using the iterative PSI-BLAST program (Altschul and Koonin, 1998) revealed that PAQRs display distant similarity with a family of proteins known as alkaline ceramidases (AlkCer), although expect values for pair-wise comparisons between PAQRs and AlkCers are always lower than the default cutoff value for significance. Like the PAQR family, the AlkCer family is unified by the presence of 7 TMs (Fig. 2B) and three short amino acid motifs that are nearly identical in sequence and position with those that unify the PAQRs. In fact, the only difference between the consensus sequences that unify the AlkCer family from those that unify the PAQR family is the substitution of a histidine in motif A of the PAQRs with an asparagine in the alkaline ceramidases (Fig. 2C and Supplemental Figs. 1-3).

Because Izh2p bears sequence similarity with enzymes in the alkaline ceramidase family, we decided to test whether overexpression of a known alkaline ceramidase would have an effect on FET3 similar to that of Izh2p overexpression. Figure 2D shows that overexpression of a yeast alkaline ceramidase, Ypc1p, repressed FET3-lacZ. As with Izh2p overexpression or activation with thaumatin, the repression of FET3-lacZ by Ypc1p overexpression required Tpk2p and the Nrg1p/Nrg2p transcriptional repressors (Fig. 2D).

Accumulation of Sphingoid Base in Response to Izh2p Overproduction. Alkaline ceramidases produce sphingoid bases from the hydrolysis of ceramides. This led us to hypothesize that sphingoid bases may be produced when Izh2p is overexpressed and play a role in the repression of FET3. To test this hypothesis, we measured the effect of Izh2p on steady-state accumulation of the primary sphingoid base in yeast-PHS. Empty expression vector was used as a negative control and Ypc1p, whose overexpression is known to enhance PHS accumulation (Mao et al., 2000a), was used as positive control. Overexpression of either Izh2p or Ypc1p caused an increase in the mean steady-state levels of PHS in a wild type strain grown in synthetic defined medium (Fig. 3A). We used a paired t test to determine the significance of the difference between the control mean and the Izh2p overexpression mean and found that the increase, although small, was significant to >95% confidence.

It follows that if Izh2p or Ypc1p overexpression represses FET3 by increasing PHS levels, then the addition of exogenous PHS should bypass the need for Izh2p and directly repress FET3. Indeed, the addition of 100 μM PHS repressed FET3-lacZ (Fig. 3B). The addition of a similar lipid (stearylamine) at the same concentration had no effect on FET3-lacZ, suggesting that the repression of FET3 by PHS was not due to a nonspecific effect of the lipid on membranes. As with other treatments, the repression of FET3-lacZ by exogenous PHS required Tpk2p and the Nrg1p/Nrg2p transcriptional repressors (Fig. 3C).

Thus, it is clear that Izh2p increases steady-state levels of PHS and that PHS is capable of mediating the repression of FET3 in a manner that requires the same signaling proteins as Izh2p overexpression. These findings suggest that any treatment causing PHS to accumulate would have a similar effect on FET3 expression. One simple way to elicit PHS accumulation is by preventing its catabolism. This can be achieved by adding N,N-DMS, a specific inhibitor of the first step (phosphorylation) of sphingoid base degradation (Welsch et al., 2004). Alternatively, the prevention of ceramide synthesis also results in PHS accumulation. This can be achieved by adding FB1, a specific inhibitor of the acyl-CoA-dependent ceramide synthases (Wu et al., 1995). Both N,N-DMS and FB1 cause a dose-dependent repression of FET3-lacZ in wild-type cells carrying empty expression vector (Fig. 3D).

The Effect of Izh2p Requires Ceramide Biosynthesis but Does Not Require Alkaline Ceramidases. The data in the previous sections suggest that the effect of Izh2p on FET3 is mediated by PHS, which raises the question of how Izh2p affects PHS levels. One possibility is that Izh2p stimulates an endogenous ceramidase enzyme and thus enhances ceramide catabolism. Yeast only possess two known ceramidases (Ypc1p and Ydc1p), both of which belong to the alkaline ceramidase family (Mao et al., 2000b). The logical experiment is to determine whether Izh2p overexpression still elicits PHS accumulation in a strain lacking Ypc1p and Ydc1p (ypc1Δydc1Δ). Indeed, Izh2p overexpression still elicits significant PHS accumulation in this double-mutant strain (Fig. 4A). Overexpression of Ypc1p in ypc1Δydc1Δ is used as a positive control. A paired t test comparing the control and Izh2p overexpressor means indicates that the increase in PHS elicited by Izh2p is small but significant to >95% confidence. These results indicate that Izh2p does not increase PHS by stimulating the activity of known ceramidases.

Fig. 4.
Izh2p-dependent signaling does not require endogenous alkaline ceramidases. A, steady-state accumulation of phytosphingosine in ypc1Δydc1Δ double-mutant cells isogenic to WT (W303-1A) carrying either empty expression vector or the ...

Another possibility is that Izh2p stimulates PHS biosynthesis or prevents its incorporation into ceramides. To distinguish between these possibilities, we performed two important experiments. First, we measured the effect of short-term FB1 treatment on Izh2p-dependent PHS accumulation in the ypc1Δydc1Δ strain. Treatment with FB1 prevents the incorporation of newly made PHS into ceramides by inhibiting the CoA-dependent ceramide synthases. Although FB1 treatment (at a concentration that maximally represses FET3-lacZ) causes a large accumulation of PHS in all strains (data not shown), there is no further increase in PHS accumulation in cells overexpressing Izh2p relative to those carrying empty expression vector. This is evident in Fig. 4B, which plots the ratio of PHS accumulation in Izh2p-overexpressing cells over PHS accumulation in cells carrying empty expression vector. The second experiment measured the effect of the antifungal drug, myriocin, under the same conditions. Myriocin is a specific inhibitor of serine palmitoyltransferase and, as such, stops de novo sphingoid base biosynthesis (Yamaji-Hasegawa et al., 2005). Figure 4B shows that short-term treatment with myriocin also prevented the accumulation of PHS caused by Izh2p overexpression. Moreover, myriocin caused a dose-dependent alleviation of the repression of FET3-lacZ caused by Izh2p overexpression (Fig. 4C). Note that the concentration of myriocin that maximally represses Izh2p-dependent FET3-lacZ repression fully suppresses PHS accumulation caused by Izh2p overexpression. (At very high concentrations of myriocin, we saw a decrease in FET3-lacZ even in cells carrying empty expression vector. This effect is probably due to the fungicidal properties of myriocin at these concentrations.) Other means of repressing FET3-lacZ, such as YPC1 overexpression, treatment with 100 μM PHS, or the addition of 1 mM Fe3+ to the grown medium, were unaffected by short-term treatment with myriocin (Fig. 4C).

Testing the Ceramidase Activity of Izh2p. One explanation for these data is that Izh2p functions as an alkaline ceramidase that hydrolyzes newly made ceramides. The simplest experiment to test this model would be to purify Izh2p and measure its ability to hydrolyze ceramide substrates. However, attempts to purify and reconstitute Izh2p have not yet been successful. To get around this problem, we measured ceramidase activity in total membrane preparations from cells carrying either empty vector or the Ypc1p and Izh2p ovexpression plasmids. The substrate for this assay was phytosphingosine that had been N-acylated with fluorescently labeled dodecanoic acid. After exposing the substrate to membrane preps for 1 h at various pH values, we did not see any bands on the TLC plate that corresponded to the free fatty acid in the Izh2p-overexpressing lane (data not shown). Although this result is difficult to reconcile with the proposed ceramidase function for Izh2p, it is important to note that we have serious reservations about the validity of this experiment that will be addressed under Discussion.

Despite the negative result from the in vitro ceramidase assays, we decided to determine whether d-erythro-MAPP, an inhibitor of alkaline ceramidases (Szulc et al., 2008), could inhibit Izh2p-dependent repression of a FET3-lacZ reporter construct. Consistent with our model, d-erythro-MAPP alleviated the effects of both Izh2p overproduction (in 2% galactose) and activation (in 0.05% galactose + 500 nM thaumatin) on FET3-lacZ expression with ED50 values of ~1 and 5 nM, respectively (Fig. 5A). Moreover, d-erythro-MAPP also potently inhibited Ypc1p-dependent FET3 repression (ED50 ~150 pM). d-erythro-MAPP had no effect on FET3-lacZ in strains carrying empty expression vector or in cells in which FET3-lacZ is repressed by exogenous PHS, indicating that PHS functions downstream of d-erythro-MAPP in the pathway. In addition, the repression of FET3-lacZ by high levels of iron was not alleviated by d-erythro-MAPP, indicating that ceramidase inhibition does not result in general induction of FET3 (Fig. 5A).

Fig. 5.
Possible ceramidase function for Izh2p. A, β-galactosidase activity generated by the FET3-lacZ reporter is shown as a percentage of activity seen in untreated WT (BY4742) cells carrying the empty expression vector. Cells were grown in iron-deficient ...

Izh2p Confers Fumonisin B1 Resistance. The acyl-CoA-dependent ceramide synthases are sensitive to FB1, and treatment with this drug results in decreased accumulation of complex sphingolipids (Cowart and Obeid, 2007). Because complex sphingolipids are essential for yeast, the sensitivity of yeast to FB1 has been used to identify genes that can bypass the need for acyl-CoA-dependent ceramide synthesis in the production of ceramides. In fact, this screen led to the discovery of the alkaline ceramidases, which have FB1-independent ceramide synthase or “reverse ceramidase” activity (Mao et al., 2000a). In Fig. 5B, we show that Izh2p overexpression resulted in the same level of resistance to FB1 as does Ypc1p overexpression, suggesting that it, too, can overcome the ceramide synthase defect.

Izh2p- and Sphingoid Base-Dependent FET3 Repression Requires Pkh1p and Pkh2p. If sphingoid base is the second messenger generated by Izh2p, what transduces the message further downstream? Intriguingly, the IZH1 and IZH4 genes are divergently transcribed from the PKH1 and PKH2 genes, respectively (Lyons et al., 2004). The organization of these genes has long led us to suspect the involvement of the PKH gene products in IZH-dependent signaling. PKH1 and PKH2 genes encode a pair of partially redundant kinases, and deletion of either PKH1 or PKH2 resulted in the loss of repression of FET3-lacZ in response to Izh2p overexpression, Ypc1p overexpression, or treatment with exogenous PHS (Fig. 6A). Ypc1p overexpression and PHS treatment still exerted small effects in the pkh1Δ and pkh2Δ mutants, presumably because of the redundant functions of Pkh1p and Pkh2p. Because the pkh1Δpkh2Δ double mutation is lethal (Casamayor et al., 1999), we could not measure the effect of losing both kinases on Izh2p-dependent FET3 repression. Pkh1p and Pkh2p were also essential for FET3-lacZ repression caused by thaumatin activation of slightly elevated levels of Izh2p (in 0.05% galactose) (Fig. 6B). It must be noted that FET3-lacZ and cell surface ferroxidase activities responded normally to iron in the pkh1Δ and pkh2Δ strains, indicating that these mutations did not result in a generalized inability to repress FET3 (data not shown).

Fig. 6.
Yeast phosphoinositide-dependent kinases are involved in Izh2p-dependent signaling. A, β-galactosidase activity from FET3-lacZ is plotted as a percentage of activity seen in untreated WT cells carrying empty expression vector as a control. ...


Data from this and other published works (Kupchak et al., 2007; Smith et al., 2008) allow us to draw some important conclusions about how Izh2p might work as a receptor. For example, the ED50 value for thaumatin-dependent FET3 repression is decreased by increasing the expression of Izh2p. This can be explained by a synergism between the Izh2p-dependent and Izh2p-independent pathways of signal transduction. However, these data also fit the spare receptors model (Zhu, 1993), which would suggest that the amount of signal produced by Izh2p exceeds the capacity of the downstream signaling pathway. In other words, there are either more Izh2p receptors or more second messengers produced by Izh2p than there are effectors to carry the signal further downstream.

The fact that Izh2p can overwhelm downstream effectors helps explain our finding that Izh2p (as well as several human PAQRs; Kupchak et al., 2007; Smith et al., 2008) constitutively transduce signals when expressed at high levels but require agonist when expressed at low levels. One interpretation of this finding is that there is a “signal transduction repressor” found in excess of Izh2p and that this repressor can be dissociated by agonist binding at low receptor concentrations or titrated away by receptor overproduction. However, the spare receptors model argues against there being such a “signal repressor” because its existence would imply a signaling bottleneck at the receptor and not downstream of it. An alternative interpretation is that, under normal conditions, Izh2p (and perhaps other PAQRs) constitutively produces very low quantities of a second messenger that are too low to activate downstream effectors. When Izh2p is activated or overexpressed, the net production of messenger increases above the threshold necessary to saturate downstream effectors.

What remains to be determined is the chemical identity of this second messenger and, to achieve this, we decided to map the signal transduction pathway using the repression of FET3 as a reporter of receptor activity. This is where the use of yeast as a model system has been particularly fruitful. Our previous studies showed that Izh2p regulated FET3 by affecting a competition between the Msn2p/Msn4p transcriptional coactivators and the Nrg1p/Nrg2p transcriptional repressors, which have opposing effects on an element in the FET3 promoter (Kupchak et al., 2007).

The identification of these transcription factors allowed us to begin mapping the signal transduction pathway from the bottom up. Msn2p/Msn4p are negatively regulated by PKA, which causes their translocation into the cytoplasm (Garmendia-Torres et al., 2007). Furthermore, Nrg1p/Nrg2p are negatively regulated by nuclear AMPK (Vyas et al., 2001), and active PKA causes AMPK to relocalize to the cytoplasm (Hedbacker et al., 2004). Thus, the balance of the competition between Nrg1p/Nrg2p and Msn2p/Msn4p can be tipped toward Nrg1p/Nrg2p, and therefore FET3 repression, by simply activating PKA. Not surprisingly, PKA, Nrg1p and Nrg2p are all essential for FET3 repression caused by either Izh2p overexpression, Izh2p activation, Ypc1p overexpression, or PHS treatment. A model for this signal transduction pathway is shown in Fig. 7, although it must be noted that much more work is required to confirm this model.

Fig. 7.
A model for Izh2p-dependent signaling in yeast cells. Solid lines indicate that a particular molecule or protein acting upon another. Broken lines indicate the effects. Arrows indicate positive effects, whereas bars indicate negative effects.

The problem with this bottom-up approach to mapping the Izh2p-dependent signaling pathway is that it does not connect PKA to Izh2p. To make this connection, it is necessary to map the pathway from the top down. Unfortunately, Izh2p is not known to interact physically or genetically with other proteins, making it difficult to develop models for how it might activate PKA. However, the fact that Izh2p resembles a group of proteins with a known enzymatic activity gave us a starting point. Although it is uncertain whether the similarities between the PAQR and alkaline ceramidase families are the result of random chance, common ancestry, or convergent evolution, it is intriguing that overexpression of Ypc1p has the same effect on FET3 as Izh2p overexpression or activation. It is also intriguing that both Izh2p activation and Ypc1p overexpression have been shown to induce apoptosis in yeast (Narasimhan et al., 2005; Aerts et al., 2008), indicating that the functional link between Izh2p and Ypc1p is not merely an artifact of our reporter system.

Because alkaline ceramidases produce PHS, we hypothesized that Izh2p also produces PHS. Indeed, Izh2p overexpression does cause a small but significant increase in steady-state levels of PHS. It is important to note that the relatively small increase in total PHS caused by Izh2p is likely to be a reflection of Izh2p's, affecting a distinct pool of PHS that may be small compared with bulk PHS. Indeed, the fact that myriocin inhibits Izh2p-dependent but not Ypc1p-dependent FET3 repression suggests that Izh2p and Ypc1p affect different pools of PHS.

The inhibition of Izh2p-dependent signaling by myriocin suggests that Izh2p stimulates de novo PHS biosynthesis. However, the fact that FB1 can prevent Izh2p-dependent PHS accumulation suggests that the pool of PHS that is influenced by Izh2p must first be incorporated into ceramides. By corollary, this finding also suggests that these newly generated ceramides must be hydrolyzed to produce the PHS we measured. We know, however, that Izh2p-dependent accumulation of PHS does not require the known endogenous ceramidases Ypc1p and Ydc1p. This leaves us with the proposition that Izh2p is stimulating some ceramidase activity that is independent of the known enzymes. Of course, the similarities between the PAQRs and alkaline ceramidases make it enticing to speculate that Izh2p itself possesses an intrinsic ceramidase activity.

We have attempted to test this idea by assaying ceramidase activity in membrane from cells overexpressing Izh2p and found no increase in the ability to hydrolyze a fluorescently labeled synthetic ceramide substrate. However, it must be stated that it is extremely difficult to interpret this result. For example, it is possible that the synthetic substrate we chose is simply not a good substrate for Izh2p. After all, there are dozens of ceramide permutations that differ in head group, fatty acid, and sphingoid base. The lack of reactivity in our experiment could merely be due to the bulky fluorescent label or to the absence of the correct head group or fatty acid. Moreover, it is possible that Izh2p requires some soluble cofactor or subunit that did not purify with our membranes. Consequently, this line of experimentation is premature, and what is needed is a sphingolipidomic study to identify particular ceramide species that may be altered by Izh2p overexpression to narrow the search for potential substrates.

For now, we have presented some circumstantial evidence that Izh2p may, indeed, function as an alkaline ceramidase. First, Izh2p-dependent repression of FET3 was potently inhibited by d-erythro-MAPP, a known inhibitor of alkaline ceramidases. It should be noted, however, that there are concerns about the specificity of d-erythro-MAPP, and it is possible that this drug exerts its effects on Izh2p-dependent signaling by an off-target mechanism. Second, Izh2p rescued the FB1 sensitivity of a yor1Δ strain, suggesting that this protein has “reverse ceramidase” or ceramide synthase activity, a hallmark of alkaline ceramidases. Taken together, these results suggest that Izh2p may possesses ceramidase activity, although much more work needs to be done before this assertion can be proven. This line of questioning will be the focus of future research.

In the meantime, our data also provide an explanation for how sphingoid bases might affect PKA activity, thus connecting our top-down and bottom-up approaches to mapping the signal-transduction pathway. We demonstrated that the kinases (Pkh1p and Pkh2p) are essential for Izh2p- and sphingoid base-dependent repression of FET3. These kinases are homologs of human PDK1 (Inagaki et al., 1999), which functions as a master regulator of downstream kinases in the AGC kinase family (Komander et al., 2005). The potential involvement of the yeast PDK1 homologs is intriguing for several reasons. First, PDK1, Pkh1p, and Pkh2p are directly activated by sphingoid bases (Friant et al., 2001; Liu et al., 2005). Second, PDK1 homologs in both humans and fungi have been shown to activate PKA (Cheng et al., 1998; Tang and McLeod, 2004). Third, yeast Pkh1p was found to physically interact with an isoform of PKA in a two-hybrid screen (Ho et al., 2002). Fourth, there is evidence from mammalian cells suggesting that sphingoid base activates PKA by mimicking the positive effects of cAMP on its regulatory subunit (Ma et al., 2005).

Hence, we propose a model in which Izh2p or a yet-to-be identified protein regulated by Izh2p catabolize newly made ceramides to generate sphingoid base. The accumulation of sphingoid base activates the yeast homologs of PDK1, which in turn activate PKA. The activation of PKA then regulates FET3 expression through its ability to sequester AMPK and Msn2p/Msn4p in the cytoplasm. Of course, more work is necessary to unequivocally confirm this model; however, it is consistent with our data and with what is known about the signaling proteins involved. It is also intriguing to note that sphingolipids have been shown to function downstream of both adiponectin and progesterone receptors (Strum et al., 1995; Kase et al., 2007), suggesting that the human PAQRs also transduce signals by the same mechanism. Future studies will focus on determining whether human PAQRs also produce sphingoid base second messengers.

Supplementary Material

[Data Supplement]


We thank the University of Florida Department of Chemistry for material support of this research.


This work was supported by the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases [Grant R21-DK074812]; the National Institutes of Health National Institute of General Medical Sciences [Grant GM63265]; and the Department of Veteran's Affairs Merit Review Entry Program.

N.Y.V. and B.R.K. contributed equally to the preparation of the manuscript.

ABBREVIATIONS: PAQR, progestin and adipoQ receptor; IZH, implicated in zinc homeostasis; d-erythro-MAPP, d-(2-myristoylamino)-3-phenylpropanol; FET, ferric iron transporter; PHS, phytosphingosine; GPCR, G-protein-coupled receptor; PKA, cAMP-dependent protein kinase; LIM, low iron medium; N,N-DMS, N,N-dimethylsphingosine; FB1, fumonisin B1; DMSO, dimethyl sulfoxide; HPLC, high-performance liquid chromatography; TLC, thin-layer chromatography; TM, transmembrane domain; AlkCer, alkaline ceramidase; AMPK, AMP-activated protein kinase; PDK1, phosphoinositide-dependent kinase 1; WT, wild type.

[S with combining enclosing square]The online version of this article (available at http://molpharm.aspetjournals.org) contains supplemental material.


  • Aerts AM, Zabrocki P, François IE, Carmona-Gutierrez D, Govaert G, Mao C, Smets B, Madeo F, Winderickx J, Cammue BP, and Thevissen K (2008) Ydc1p ceramidase triggers organelle fragmentation, apoptosis and accelerated ageing in yeast. Cell Mol Life Sci 65 1933-1942. [PubMed]
  • Altschul SF and Koonin EV (1998) Iterated profile searches with PSI-BLAST-a tool for discovery in protein databases. Trends Biochem Sci 23 444-447. [PubMed]
  • Bligh EG and Dyer WJ (1959) A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37 911-917. [PubMed]
  • Casamayor A, Torrance PD, Kobayashi T, Thorner J, and Alessi DR (1999) Functional counterparts of mammalian protein kinases PDK1 and SGK in budding yeast. Curr Biol 9 186-197. [PubMed]
  • Cheng X, Ma Y, Moore M, Hemmings BA, and Taylor SS (1998) Phosphorylation and activation of cAMP-dependent protein kinase by phosphoinositide-dependent protein kinase. Proc Natl Acad Sci U S A 95 9849-9854. [PMC free article] [PubMed]
  • Claros MG and von Heijne G (1994) TopPred II: an improved software for membrane protein structure predictions. Comput Appl Biosci 10 685-686. [PubMed]
  • Cowart LA and Obeid LM (2007) Yeast sphingolipids: recent developments in understanding biosynthesis, regulation, and function. Biochim Biophys Acta 1771 421-431. [PMC free article] [PubMed]
  • Foury F (1997) Human genetic diseases: a cross-talk between man and yeast. Gene 195 1-10. [PubMed]
  • Friant S, Lombardi R, Schmelzle T, Hall MN, and Riezman H (2001) Sphingoid base signaling via Pkh kinases is required for endocytosis in yeast. EMBO J 20 6783-6792. [PMC free article] [PubMed]
  • Garmendia-Torres C, Goldbeter A, and Jacquet M (2007) Nucleocytoplasmic oscillations of the yeast transcription factor Msn2: evidence for periodic PKA activation. Curr Biol 17 1044-1049. [PubMed]
  • Hedbacker K, Townley R, and Carlson M (2004) Cyclic AMP-dependent protein kinase regulates the subcellular localization of Snf1-Sip1 protein kinase. Mol Cell Biol 24 1836-1843. [PMC free article] [PubMed]
  • Ho Y, Gruhler A, Heilbut A, Bader GD, Moore L, Adams SL, Millar A, Taylor P, Bennett K, Boutilier K, et al. (2002) Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature 415 180-183. [PubMed]
  • Inagaki M, Schmelzle T, Yamaguchi K, Irie K, Hall MN, and Matsumoto K (1999) PDK1 homologs activate the Pkc1-mitogen-activated protein kinase pathway in yeast. Mol Cell Biol 19 8344-8352. [PMC free article] [PubMed]
  • Kase H, Hattori Y, Jojima T, Okayasu T, Tomizawa A, Suzuki K, Banba N, Monden T, Satoh H, Akimoto K, et al. (2007) Globular adiponectin induces adhesion molecule expression through the sphingosine kinase pathway in vascular endothelial cells. Life Sci 81 939-943. [PubMed]
  • Kobayashi SD and Nagiec MM (2003) Ceramide/long-chain base phosphate rheostat in Saccharomyces cerevisiae: regulation of ceramide synthesis by Elo3p and Cka2p. Eukaryot Cell 2 284-294. [PMC free article] [PubMed]
  • Komander D, Kular G, Deak M, Alessi DR, and van Aalten DM (2005) Role of T-loop phosphorylation in PDK1 activation, stability, and substrate binding. J Biol Chem 280 18797-18802. [PubMed]
  • Kupchak BR, Garitaonandia I, Villa NY, Mullen MB, Weaver MG, Regalla LM, Kendall EA, and Lyons TJ (2007) Probing the mechanism of FET3 repression by Izh2p overexpression. Biochim Biophys Acta 1773 1124-1132. [PMC free article] [PubMed]
  • Liu K, Zhang X, Sumanasekera C, Lester RL, and Dickson RC (2005) Signalling functions for sphingolipid long-chain bases in Saccharomyces cerevisiae. Biochem Soc Trans 33 1170-1173. [PubMed]
  • Lyons TJ, Villa NY, Regalla LM, Kupchak BR, Vagstad A, and Eide DJ (2004) Metalloregulation of yeast membrane steroid receptor homologs. Proc Natl Acad Sci U S A 101 5506-5511. [PMC free article] [PubMed]
  • Ma Y, Pitson S, Hercus T, Murphy J, Lopez A, and Woodcock J (2005) Sphingosine activates protein kinase A type II by a novel cAMP-independent mechanism. J Biol Chem 280 26011-26017. [PubMed]
  • Mao C, Xu R, Bielawska A, and Obeid LM (2000a) Cloning of an alkaline ceramidase from Saccharomyces cerevisiae. An enzyme with reverse (CoA-independent) ceramide synthase activity. J Biol Chem 275 6876-6884. [PubMed]
  • Mao C, Xu R, Bielawska A, Szulc ZM, and Obeid LM (2000b) Cloning and characterization of a Saccharomyces cerevisiae alkaline ceramidase with specificity for dihydroceramide. J Biol Chem 275 31369-31378. [PubMed]
  • McEwan IJ (2001) Bakers yeast rises to the challenge: reconstitution of mammalian steroid receptor signalling in S. cerevisiae. Trends Genet 17 239-243. [PubMed]
  • Merrill AH, Caligan TB, Wang E, Peters K, and Ou J (2000) Analysis of sphingoid bases and sphingoid base 1-phosphates by high-performance liquid chromatography. Methods Enzymol 312 3-9. [PubMed]
  • Min K, Ha SC, Hasegawa PM, Bressan RA, Yun DJ, and Kim KK (2004) Crystal structure of osmotin, a plant antifungal protein. Proteins 54 170-173. [PubMed]
  • Narasimhan ML, Coca MA, Jin J, Yamauchi T, Ito Y, Kadowaki T, Kim KK, Pardo JM, Damsz B, Hasegawa PM, et al. (2005) Osmotin is a homolog of mammalian adiponectin and controls apoptosis in yeast through a homolog of mammalian adiponectin receptor. Mol Cell 17 171-180. [PubMed]
  • Schorling S, Vallée B, Barz WP, Riezman H, and Oesterhelt D (2001) Lag1p and Lac1p are essential for the Acyl-CoA-dependent ceramide synthase reaction in Saccharomyces cerevisiae. Mol Biol Cell 12 3417-3427. [PMC free article] [PubMed]
  • Smith JL, Kupchak BR, Garitaonandia I, Hoang LK, Maina AS, Regalla LM, and Lyons TJ (2008) Heterologous expression of human mPRα, mPRβ and mPRγ in yeast confirms their ability to function as membrane progesterone receptors. Steroids 73 1160-1173. [PMC free article] [PubMed]
  • Strum JC, Swenson KI, Turner JE, and Bell RM (1995) Ceramide triggers meiotic cell cycle progression in Xenopus oocytes. A potential mediator of progesterone-induced maturation. J Biol Chem 270 13541-13547. [PubMed]
  • Szulc ZM, Mayroo N, Bai A, Bielawski J, Liu X, Norris JS, Hannun YA, and Bielawska A (2008) Novel analogs of d-e-MAPP and B13. Part 1: synthesis and evaluation as potential anticancer agents. Bioorg Med Chem 16 1015-1031. [PMC free article] [PubMed]
  • Tang Y and McLeod M (2004) In vivo activation of protein kinase A in Schizosaccharomyces pombe requires threonine phosphorylation at its activation loop and is dependent on PDK1. Genetics 168 1843-1853. [PMC free article] [PubMed]
  • Tang YT, Hu T, Arterburn M, Boyle B, Bright JM, Emtage PC, and Funk WD (2005) PAQR proteins: a novel membrane receptor family defined by an ancient 7-transmembrane pass motif. J Mol Evol 61 372-380. [PubMed]
  • Tatusova TA and Madden TL (1999) BLAST 2 Sequences, a new tool for comparing protein and nucleotide sequences. FEMS Microbiol Lett 174 247-250. [PubMed]
  • Thomas P, Pang Y, Dong J, Groenen P, Kelder J, de Vlieg J, Zhu Y, and Tubbs C (2007) Steroid and G protein binding characteristics of the seatrout and human progestin membrane receptor alpha subtypes and their evolutionary origins. Endocrinology 148 705-718. [PubMed]
  • Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, and Higgins DG (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 25 4876-4882. [PMC free article] [PubMed]
  • Vyas VK, Kuchin S, and Carlson M (2001) Interaction of the repressors Nrg1 and Nrg2 with the Snf1 protein kinase in Saccharomyces cerevisiae. Genetics 158 563-572. [PMC free article] [PubMed]
  • Welsch CA, Roth LW, Goetschy JF, and Movva NR (2004) Genetic, biochemical, and transcriptional responses of Saccharomyces cerevisiae to the novel immunomodulator FTY720 largely mimic those of the natural sphingolipid phytosphingosine. J Biol Chem 279 36720-36731. [PubMed]
  • Wu WI, McDonough VM, Nickels JT Jr, Ko J, Fischl AS, Vales TR, Merrill AH Jr, and Carman GM (1995) Regulation of lipid biosynthesis in Saccharomyces cerevisiae by fumonisin B1. J Biol Chem 270 13171-13178. [PubMed]
  • Yamaji-Hasegawa A, Takahashi A, Tetsuka Y, Senoh Y, and Kobayashi T (2005) Fungal metabolite sulfamisterin suppresses sphingolipid synthesis through inhibition of serine palmitoyltransferase. Biochemistry 44 268-277. [PubMed]
  • Zhu BT (1993) The competitive and noncompetitive antagonism of receptor-mediated drug actions in the presence of spare receptors. J Pharmacol Toxicol Methods 29 85-91. [PubMed]

Articles from Molecular Pharmacology are provided here courtesy of American Society for Pharmacology and Experimental Therapeutics
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...