- Email: [email protected]
Alterations of O-glycosylation, cell wall, and mitochondrial metabolism in Kluyveromyces lactis cells defective in KlPmr1p, the Golgi Ca2+-ATPase
BBRC Biochemical and Biophysical Research Communications 318 (2004) 1031–1038 www.elsevier.com/locate/ybbrc
Alterations of O-glycosylation, cell wall, and mitochondrial metabolism in Kluyveromyces lactis cells defective in KlPmr1p, the Golgi Ca2+-ATPase Francesca Farina,a,1 Daniela Uccelletti,a,1 Paola Goffrini,b Ronald A. Butow,c Claudia Abeijon,d and Claudio Palleschia,* a
Department of Developmental and Cell Biology, University of Rome “La Sapienza,” Piazza Aldo Moro 5, 00185 Rome, Italy Department of Genetics, Anthropology, Evolution, University of Parma, Parco Area delle Scienze 11/A, 43100 Parma, Italy c Department of Molecular Biology, University of Texas Southwestern Medical Center, 6000 Harry Hines Blvd., Dallas, TX 75390, USA Department of Molecular and Cell Biology, Goldman School of Dental Medicine, Boston University, 715 Albany Street, Boston, MA 02118, USA b
d
Received 5 April 2004 Available online 7 May 2004
Abstract In yeast the P-type Ca2þ -ATPase of the Golgi apparatus, Pmr1p, is the most important player in calcium homeostasis. In Kluyveromyces lactis KlPMR1 inactivation leads to pleiotropic phenotypes, including reduced N-glycosylation and altered cell wall morphogenesis. To study the physiology of K. lactis when KlPMR1 was inactivated microarrays containing all Saccharomyces cerevisiae coding sequences were utilized. Alterations in O-glycosylation, consistent with the repression of KlPMT2, were found and a terminal N-acetylglucosamine in the O-glycans was identified. Klpmr1D cells showed increased expression of PIRs, proteins involved in cell wall maintenance, suggesting that responses to cell wall weakening take place in K. lactis. We found over-expression of KlPDA1 and KlACS2 genes involved in the Acetyl-CoA synthesis and down-regulation of KlIDP1, KlACO1, and KlSDH2 genes involved in respiratory metabolism. Increases in oxygen consumption and succinate dehydrogenase activity were also observed in mutant cells. The described approach highlighted the unexpected involvement of KlPMR1 in energy-yielding processes. Ó 2004 Elsevier Inc. All rights reserved. Keywords: KlPMR1; Yeast; Heterologous arrays; O-glycans; Cell wall
In animal and plant cells cytosolic free calcium is a major signal-transducing element employed to induce a variety of responses from cell proliferation to cell differentiation. Calcium homeostasis is mainly regulated by P-type Ca2þ -ATPases. Genes encoding these Ca2þ ATPases of the early secretory pathway have been isolated from several organisms [1,2]; the pump encoded by PMR1 gene of Saccharomyces cerevisiae is perhaps the best characterized [3,4]. Pmr1p was localized to the medial Golgi compartment, where it was found to be important for functioning of the secretory pathway [5]. These studies showed that cells lacking functional *
Corresponding author. Fax: +39-064991-2132. E-mail address: [email protected] (C. Palleschi). 1 These authors contributed equally to this work. 0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.04.127
PMR1 exhibit defects in protein glycosylation, processing, sorting, and endoplasmic reticulum-associated protein degradation. In yeast, Pmr1p seems to be the major Ca2þ -ATPase contributing to the steady state free Ca2þ concentration in the endoplasmic reticulum. The amount of this cation decreases by 50% in the endoplasmic reticulum of pmr1 mutant cells, while the level of cytosolic calcium increases up to 16-fold despite a compensatory increase in the expression of the vacuolar Ca2þ -ATPase, PMC1 [6]. PMR1 also transports Mn2þ inside the lumen of the Golgi apparatus and appears to be important for the homeostasis of this ion. High concentrations of cytoplasmic Mn2þ are toxic and can interfere with Mg2þ binding sites on proteins. Defective Ty1 retrotransposition in a pmr1 mutant was recently shown to be due to Mn2þ inhibition of reverse
1032
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
transcriptase [7]. In addition, the low concentration of Mn2þ in the Golgi compartment of pmr1 mutant cells leads to defective N- and O-glycosylation [4]. An intriguing additional phenotype resulting from the inactivation of PMR1 is the enhanced release into the growth medium of secretory proteins; this characteristic has been exploited for the production of heterologous proteins [8] but the underlying mechanisms are still not known. We isolated the orthologous KlPMR1 gene of Kluyveromyces lactis and demonstrated its involvement in a variety of phenotypes including alterations in N-linked glycosylation and secretion as well as in cell wall morphogenesis [9]. The cell wall defects include a doubling of cell wall thickness and significant alterations in the glycan compositions; such changes are not detectable at all in the corresponding S. cerevisiae mutant. Moreover, the cell wall phenotypes observed in Klpmr1D cells are not just pleiotropic consequences of the glycosylation defects. In fact, the addition of external calcium almost reverts the Klpmr1D cell wall phenotypes but leaves completely unchanged the defective glycosylation. Significant differences in calcium homeostasis may therefore exist between S. cerevisiae and K. lactis. In addition, although K. lactis has been successfully used as alternative host for heterologous protein production [10], the secretion and glycosylation pathways in this yeast remain poorly understood [11–13]. The pleiotropic phenotypes observed in the Klpmr1 mutant cells prompted us to perform a genome-wide transcription profile analysis of the parental strain versus the mutant to uncover cellular links to the physiological role of KlPMR1. Heterologous S. cerevisiae genome arrays were used, due to the lack of the genome sequence for K. lactis. In this paper we report the involvement of K. lactis Ca2þ ATPase in the initiation and elongation of O-glycosylation. The unexpected presence of N -acetylglucosamine in the O-chains of this yeast was established. The increase in the amount of PIR proteins in the mutant cells indicates the presence of a cell wall integrity mechanism in K. lactis. We also found an unforeseen link between KlPMR1 and mitochondrial metabolism highlighted by increased oxygen consumption and transcriptional alteration of genes encoding for respiratory enzymes. Materials and methods Yeast strains and growth conditions. The strains used in this study were MW278-20C (MATa ade2, leu2, uraA), CPK1 (MATa ade2, leu2 uraA KlPMR1::KanR ), MG1/2 (MATa uraA, arg , lys Kþ pKD1þ ), and KL3 (MATa uraA, mnn2-2, arg Kþ pKD1þ ). Cells were grown on YPD (1% yeast extract, 1% bacto peptone, and 2% glucose) or YPD supplemented with 10 mM CaCl2 . Microarray analysis. Microarrays consisting of 6219 S. cerevisiae genes were prepared essentially as described by DeRisi et al. [14] and were based on PCR amplification of S288C yeast genomic DNA using gene-specific oligo pairs supplied by Research Genetics (Birmingham,
AL). RNA Isolation: cells were grown in YPD medium to an OD600 of 0.7–1.0. Total yeast RNA was prepared using the hot phenol method [15]. The poly(A)þ -mRNAs were prepared by using the Oligotex mRNA kit (Qiagen) according to the manufacturer’s instructions. Cy3- and Cy5-labelled cDNAs were prepared and hybridized to a microarray of 6219 S. cerevisiae genes. Replica experiments were carried out using independent cultures and with the opposite configuration of Cy3 and Cy5. The data were analyzed as described by Epstein et al. [16,17] using a cut-off of value of an average of 3-fold change in replicate hybridizations. O-linked carbohydrate analysis. The method of Haselbeck and Tanner [18] was followed for the isolation of total O-linked carbohydrates from [2-3 H]mannose (18 Ci/mmol; New England [NEN]) radiolabeled cells. Radiolabeled species in the supernatant fraction were subjected to thin-layer chromatography on Silica gel G plates (Merck). The chromatograms were treated with EN3 HANCE reagent (NEN) for fluorography and exposed to Kodak X-OMAT X-ray film at )70 °C. Labeling of cell wall proteins. Yeast cells were grown in 1 L YPD medium to an OD600 of 4–5 and labeled with NHS-LC-Biotin reagent as described in Mrsa et al. [19]. To visualize the biotin-labeled proteins, they were blotted to PVDF membrane (Bio-Rad) which was incubated in streptavidin–horseradish peroxidase conjugated solution (dilution 1:7500) and then developed using the ECL Kit (Amersham). Analysis of Hsp150/Pir2. The proteins in the culture medium of K. lactis strains were separated by SDS–PAGE using a 10% polyacrylamide gel and transferred to PVDF membranes (Bio-Rad). The membrane was incubated with antibody against S. cerevisiae Hsp150, a generous gift of Dr. M. Makarow (University of Helsinki, Finland), at a dilution of 1:2000 and then incubated with goat anti-rabbit IgG conjugated with peroxidase. Finally, the membrane was developed with the ECL detection kit (Amersham). Northern blot analysis. Total RNA of K. lactis strains was extracted by the hot phenol method [21]. The RNAs were quantified by absorption (OD260 ) and separated by denaturing agarose electrophoresis. Following electrophoresis the RNAs were transferred to nylon membranes and hybridized with 32 P-labeled random primed probes (Roche). All the probes were PCR amplified from K. lactis DNA genome (the sequences were kindly provided by Prof. Bolotin-Fukuhara, Paris). The PCR product of KlPMT2 was of 1.3 Kb using primers 50 -TTCTCGTTACCAACCGTTTG-30 and 50 -CACGGCAACCAAC ATAA-30 , of KlACS2 was of 3.0 Kb using primers 50 -TCTAAGTCA ATTCTTCAGCGG-30 and 50 -AAAGCTCCAGTGGTTTCGAGA30 , of KlPDA1 was of 1.5 Kb using primers 50 -TCAGGAACTCAGA GCTTCTTC-30 and 50 -CCATACCGTGAATGAAATGAA-30 , of KlIDP1 was of 900 bp using primers 50 -CGATTGCCATTGCCCTA AGT-30 and 50 -TGAAAGAGGCATGAGCGAAC-30 , of KlSDH2 was of 800 bp using primers 50 -TCTGCGATATCTACCTGGAT TC-30 and 50 -TATCCGTTTCCGTTAAGTTTTCAGA-30 , and of KlACO1 was of 800 bp using primers 50 -ATGTTGTCTGCTC GTGTTGC-30 and 50 -CGGATGTAGTGGCACCGATT-30 . Respiration and enzyme activity. The measurements of respiratory activity were performed according to Ferrero et al. [20] preparation of mitochondria and determination of succinate dehydrogenase activity (SDH) (EC 1.3.99.1) were carried out according to Lodi and Ferrero [21]. The SDH activity was expressed as nmol/min/mg protein.
Results To identify genes regulated in response to inactivation of the KlPMR1 expression profiles of cells carrying the wild type and the null allele were compared. Due to the lack of comprehensive genome sequence data for K. lactis, we took advantage of the overall high similarity
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
1033
Table 1 Genes that show strongly altered expression in Klpmr1D mutant ORF
Gene name
Array (fold)
Gene product
YAL 023C YGR 088C YDL 066W YLL 041C YLR 153C YER 178W YKL 164C YJL 159W YKL 163W YKL 158C YPL 211W YOL 077C YNL 244C YGR 159C
PMT2 CTT1 IDP1 SDH2 ACS2 PDA1 PIR1 PIR2 PIR3 CIS3 NIP7 BRX1 SUI1 NRS1
)3.5 [)3] )4.5 )9 [)4.5] )3.5 [)2.5] 3.7 [2.5] 3.5 [1.5] 6.5 6 3.5 3.5 6.5 6 4 5.7
Mannosyltransferase responsible for initiation of O-glycosylation Cytosolic catalase T Isocitrate dehydrogenase (NADP+), mitochondrial Succinate dehydrogenase (ubiquinone) iron–sulfur protein (Ip) subunit Acetyl-CoA synthetase Pyruvato dehydrogenase complex E1-alpha subunit Protein required for tolerance to heat shock O-glycosylated cell wall protein required for tolerance to heat shock Protein with similarity to members of the PIR family proteins Cell wall protein with similarity to members of the PIR family Nucleolar protein required for efficient 60S ribosome subunit biogenesis Protein required for biogenesis of the 60S ribosomal subunit 16 kDa subunit of translation initiation factor eIF3 Nucleolar protein involved in processing 20S to 18S rRNA
Values in square brackets were obtained by homologous Northern hybridization.
of known K. lactis genes compared to S. cerevisiae ones and performed a heterologous microarray study. Labeled cDNA probes were generated from mRNA of wild type and Klpmr1D strains as described in Materials and methods and hybridized at 50 °C onto S. cerevisiae slides. Under these conditions we obtained the higher percentage of hybridization signals above background (about 60% of 6219 S. cerevisiae ORFs in all the performed experiments). Our attention was focused on those genes whose geometric mean expression level increased or decreased at least 3-fold in three independent experiments in Klpmr1D background. We identified 14 genes that are reported in Table 1. Among the genes that are down-regulated by the loss of the Ca2þ -ATPase was PMT2, a gene involved in the initiation of the O-glycosylation process [22]. It was puzzling to find such a strong defect in the step that initiates the O-linked sugar chains because their elongation is also impaired as discussed below. Several genes involved in the TCA cycle as well as in the oxidative phosphorylation were also down-regulated in Klpmr1D cells, although not all of them met the threshold to be included in Table 1. These expression profiles could be related to a reduction in the growth rate of the Klpmr1D cells. In fact, in rich medium, Klpmr1D cells showed a generation time of 3.5 h compared to 2 h of the wild type cells. Several genes were up-regulated in response to the loss of KlPMR1, such as those related to cell wall morphogenesis and genes involved in the assembly and processing of rRNA. O-glycosylated chains in K. lactis contain N-acetylglucosamine and are severely reduced in the Klpmr1D strain Among the genes down-regulated in Klpmr1D cells was PMT2 and this was also confirmed by homologous Northern hybridization (Fig. 1A). We therefore analyzed total O-glycosylated proteins; for this we labeled wild type and mutant cells with [2-3 H]mannose, released
their O-linked chains by b-elimination, and analyzed them by thin-layer chromatography (Fig. 1B). The oligosaccharide pattern from K. lactis wild-type strain showed a profile similar to that of the S. cerevisiae counterpart, where it is possible to observe species containing between one and five mannose residues (Fig. 1B, lanes 1 and 3). K. lactis pmr1D cells showed a partial block in their ability to extend O-linked mannan chains manifested by accumulation of M1 and reduction of M3, M4, and M5; the amount of M2 was quite similar to that of wild type (Fig. 1B, lane 2). We found a 4-fold decrease in pmol of b-eliminated mannose per mg wet weight in Klpmr1D cells as compared with wild type, indicating a severe reduction in the total O-mannosylated glycoproteins and an altered elongation pattern. The structure of O-linked chains has never been characterized in K. lactis. To gain insight into the structure of O-linked chains in wild type and Klpmr1D strains, we treated them with jack bean a-mannosidase (Fig. 2). The most abundant M2 as well as M3 chains were completely converted to free mannose, indicating a-linkages. Surprisingly, M5 chains were not digested by the exomannosidase, indicating either that the mannose at their non-reducing end is not in a-linkage or that other sugars may be present. M4 chains were only partially resistant, suggesting the existence of two different kinds of M4 chains in K. lactis (Fig. 2A, lanes 1 and 2). This is different from S. cerevisiae, which only has alinked mannose residues in its O-glycans and not other sugars (Fig. 2A, lanes 3 and 4). We hypothesized that the presence of an additional sugar such as N -acetylglucosamine accounts for the differences in K. lactis Ochains; the N-linked sugars of K. lactis mannoproteins have terminal a1–2 linked N -acetylglucosamine residue that is absent from the outer mannan chains of S. cerevisiae [23]. The gene encoding the K. lactis UDP-N acetylglucosamine Golgi transporter, MNN2, has been isolated and characterized [17]; we thus decided to take advantage of this knowledge to test our hypothesis. The
1034
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
Chains with 5 mannoses were completely absent in mnn2-2 cells strongly suggesting that the terminal sugar is N -acetylglucosamine (Fig. 2B, lanes 1 and 2). We also detected a partial reduction in the amount of chains with four sugar residues in the mutant that cannot add N acetylglucosamine. This suggests the existence of two types of four residue chains, one with four a-linked mannoses and another with three mannoses and a terminal N -acetylglucosamine (Fig. 2C). Digestion of Olinked chains synthesized by the mutant in the UDP-N acetylglucosamine transporter confirmed this hypothesis because the reduced amount of M4 produced was completely sensitive to a-mannosidase treatment (Fig. 2B, lane 3). Cell wall alterations in Klpmr1D cells
Fig. 1. O-glycosylation is altered in Klpmr1D cells. (A) Northern blot analysis of KlSDH2 in Klpmr1D cells. Total RNA was extracted from KlPMR1 and Klpmr1D strains (lanes 1 and 2, respectively) grown on YPD. The same amount of RNA (10 lg) was loaded on each lane. The ethidium bromide-stained gel of the autoradiogram is shown in the lower part of the panel. Densitometric quantification of mRNA was performed by the computer program Phoretix 1D and mRNA loading was normalized using the rRNA bands. (B) Total oligosaccharides from MW278-20C (Kl wt), CPK1 (Kl D), and BY4741 (Sc wt) strains were radiolabeled with [2-3 H]mannose. Alkali releasable saccharides were separated on thin layer chromatography (butanol, ethanol, water (5:3:2)) and subjected to autoradiography (50,000 cpm/lane). M1 mannose, M2 mannobiose, M3 mannotriose, M4 mannotetraose, and M5 mannopentose.
bulk of O-glycoproteins were analyzed from mnn2-2 cells carrying the mutated allele of this gene, which totally lacks N -acetylglucosamine in the outer mannan chains.
Klpmr1D cells show defects in cell wall composition and morphology [9]. In the genome-wide analysis we found an increased expression of all members of the gene family encoding for PIR proteins (proteins with internal repeats), except YJL160c (Table 1); a class of cell wall proteins covalently linked to b-1,3-glucans [24]. Cell surface biotinylation of proteins was performed in K. lactis wild type and mutant cells; labeled proteins retained in the cell wall via non-covalent or S–S bridges were extracted from purified cell wall fractions by SDS treatment under reducing conditions. The PIR proteins were then extracted with NaOH and analyzed by electrophoresis after normalization per OD600 (Fig. 3A). These experiments show the presence of PIR-related proteins in K. lactis and their relative abundance in the Klpmr1D strain with respect to wild type, in agreement with the array data presented in Table 1. The most abundant PIR protein in S. cerevisiae, PIR2/HSP150, is a heat shock inducible protein that is only partially retained in the cell wall while most is secreted to the medium [25,26]. A Western blot of material in the media secreted from wild type and mutant cells was analyzed with a S. cerevisiae antiPIR2 antibody (Fig. 3B). The Klpmr1D strain secreted three times more protein than did the wild type strain; this is consistent with the observed increase in the content of PIR proteins. Altered expression of genes encoding for mitochondrial enzymes in the Klpmr1D strain Transcription of several genes involved in mitochondrial functions was altered in the Klpmr1D strain. Among them, KlIDP1 (isocitrate dehydrogenase) as well as KlSDH2 (a subunit of the succinate dehydrogenase) and KlACO1 (aconitase) were down-regulated in mutant cells, whereas KlPDA1 (alpha subunit of pyruvate dehydrogenase) and KlACS2 (acetyl CoA synthetase) expression was up-regulated. Since the
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
1035
Fig. 2. N -acetylglucosamine is present in K. lactis O-glycosydic chains. (A) Total O-linked chains from MW278-20C (Kl wt, lane 1) and BY4741 (Sc wt, lane 3) strains were treated with jack bean a-mannosidase (lanes 2 and 4, respectively), separated on thin layer chromatography as in Fig. 1. (B) Total O-linked chains from MG1/2 (Kl wt, lane 1) and KL3 (mnn2-2, lane 2) strains were separated by thin layer chromatography. (C) Schematic representation of O-linked chains in K. lactis.
array data come from heterologous hybridizations we decided to perform individual Northern blot analyses with K. lactis specific probes. Total RNA was extracted from cells grown in the same conditions utilized for the array experiments. The results showed an increase of 2.5- and 1.5-fold in the transcript levels of KlACS2 and KlPDA1 in the mutant versus wild type strain (Fig. 4A). These genes are functionally homologues to the corresponding S. cerevisiae genes [27,28]. In the same experiments we also observed a 4.5-, 2.5-, and a 2-fold reduction of KlIDP1, KlSDH2, and KlACO1 transcripts, respectively, in the Klpmr1D mutant. The KlACO1 gene was not included in Table 1 since it did not match the chosen threshold; how-
ever, its expression data were consistent in at least two out of three array experiments and were then analyzed by Northern (Fig. 4B). These independent experiments therefore confirmed the array data. On the basis of these results we analyzed the oxygen consumption in the parental and in the mutant strain. The respiratory activity of Klpmr1D mutant was 43 ll O2 /h/mg d.w. versus 30 of parental strain and antimycin A reduced this activity at the same level in both strains; the cytochrome content was, however, almost identical in the two strains (not shown). In addition, the SDH activity of the mutant was 1.6-fold higher than that measured in the parental strain (Fig. 5).
1036
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
Fig. 3. KlPIR-like proteins are increased in Klpmr1D cells. (A) Cell wall PIR-like proteins were extracted from biotinylated intact cells of MW278-20C (wt, lane 1) and CPK1 (D, lane 2) by 30 mM NaOH. The extracts were separated on SDS–PAGE, blotted and stained with streptavidin–horseradish peroxidase. Proteins corresponding to 40 OD600 were used per lane. (B) Culture medium amounts corresponding to OD600 0.1 of cells were taken from MW278-20C (wt, lane 1) and CPK1 (D, lane 2) liquid cultures, separated by SDS–PAGE, and probed with ScHsp150 antibody. Protein molecular weight standards are indicated alongside of both panels.
Discussion In the present work we have studied the physiological effects of KlPMR1 inactivation using heterologous genechip microarray technology. Because the K. lactis genome sequence is not yet completely available, we took advantage of the availability of S. cerevisiae microarrays. We obtained insight into the involvement of KlPMR1 in several cellular processes. The data obtained with the “heterologous” approach were validated by subsequent homologous Northern blots and biochemical experiments. In K. lactis, inactivation of KlPMR1 causes a reduction in mannose outer-chain extension of N-linked glycoproteins as well as defective cell wall morphogenesis [9]. We found defects in the initiation and extension of O-glycosylation in Klpmr1D strains. The unexpected down-regulation of PMT2, one of the genes involved in initiating O-linked glycosylation, prompted us to characterize K. lactis O-chains. This resulted in the first report on the presence of GlcNAc at the non-reducing end of the O-linked oligosaccharides in all fungi and yeasts studied to date. The PMT2 gene in S. cerevisiae encodes a mannosyltransferase responsible for initiation of
Fig. 4. Expression pattern of mitochondrial enzyme genes. Northern blot analyses of KlACS2 and KlPDA1 (A) KlIDP1, KlSDH2, and KlACO1 (B), in Klpmr1D cells. Total RNA was extracted from KlPMR1 and Klpmr1D strains (lanes 1 and 2, respectively) grown on YPD. The same amount of RNA (10 lg) was loaded on each lane. The ethidium bromide-stained gels of the autoradiogram are shown in the lower part of each panel. Densitometric quantification of mRNA was performed by the computer program Phoretix 1D and mRNA loading was normalized using the rRNAs bands.
Fig. 5. Succinate dehydrogenase activity in wild type and KlPmr1D strains. Enzyme activity was measured in mitochondria extracts prepared by exponentially growing cells and was expressed as nmol of substrate utilized per min per mg proteins. All values are means of three independent experiments. In no case the variation was higher than 15%.
O-glycosylation in the endoplasmic reticulum (ER) [22]. Analysis of bulk cell mannoproteins of Klpmr1D mutants showed that the normal elongation of O-chains
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
failed in this mutant. Elongation enzymes require manganese as a cofactor in S. cerevisiae and it is therefore likely that the elongation defect in Klpmr1D is due to the reduced availability of this ion in the Golgi apparatus. However, defects in the initiation of O-glycosylation in Klpmr1D were only discovered by our genome-wide analysis of gene expression. The molecular mechanism by which PMT2 is down-regulated is not known at the present time. We are tempting to speculate that because of the elongation defect the product of Pmt2p (M1) accumulates and product inhibition of the enzyme might occur. A feedback signaling would then reduce the expression of PMT2. It has been recently shown in S. cerevisiae that PMT2 is one of several genes co-repressed under different stress conditions, such as changes in temperature, oxidation, nutrients, pH, and osmolarity [29]. The molecular mechanism(s), however, are unclear. Cells carrying the inactivated allele of KlPMR1 have alterations in O-glycosylation and cell wall organization and could possibly undergo activation of signaling pathways similar to those induced by environmental stresses. A recent report demonstrates that Pmt2p is specifically responsible for the O-mannosylation of Mid2p, and that the glycosylation state of Mid2p strongly affects its functioning as a signaling element of cell wall stresses [30]. The loss of K. lactis Ca2þ -ATPase caused changes in the cell wall structure that could elicit a response, including the increase in the expression of PIR-related proteins. All members of this gene family, except YJL160c, were positively regulated by cell integrity signaling through activation of Mpk1 [24,31]. Our data indicate the presence of PIR related proteins in K. lactis and suggest the existence of cell wall integrity signaling pathways. The genome-wide analyses in Klpmr1D strains highlighted a genetic interaction between the Golgi Ca2þ -ATPase and mitochondrial metabolism. This is supported by the observed increase in the transcription of KlPDA1 and KlACS2 and down-regulation of KlSDH2, KlIDP1, and KlACO1 genes. In S. cerevisiae, Pmr1p is able to control Ca2þ homeostasis in the early secretory compartments; the absence of Pmr1p causes an increase in cytosolic Ca2þ [6]. Mitochondrial Ca2þ uptake following the rise of cytosolic Ca2þ concentration has been demonstrated in many mammalian cell types. The physiologic role of such mitochondrial Ca2þ uptake is mainly related to the control of energy metabolism, as supported by the observations that some dehydrogenases of TCA cycle as well as the ATP production are activated by Ca2þ [32]. We could speculate that similar events could take place in K. lactis cells: the absence of KlPmr1p would cause a rise in cytosolic Ca2þ followed by Ca2þ uptake by the mitochondria resulting in stimulation of energy metabolism. The increased mitochondrial metabolism,
1037
supported by the increasing in O2 uptake and SDH activity, would then activate a nucleo-mitochondria crosstalk ending in fine-tuning the transcription of genes encoding for mitochondrial enzymes of energy metabolism. It is worth noting that a genome-wide transcription analysis in S. cerevisiae has shown connections between calcium shortage, energetic metabolism, and cell-cycle progression [33]. An increase in the transcription of genes involved in ribosome assembly or in processing of rRNA was also found in Klpmr1D cells. Defects in several points of the secretory pathway that block protein transport downregulate ribosomal synthesis, through the repression of transcription of both rRNA and ribosomal protein genes in S. cerevisiae [34]. Although the mechanism is not yet understood, Klpmr1D shows both increased secretory capabilities [35] and up-regulation of genes involved in rRNA biosynthesis, suggesting that these cellular events are linked. In conclusion, this report highlights the involvement of K. lactis Ca2þ -ATPase in relevant cell processes like O-glycosylation, cell wall biogenesis and, mostly unexpected, in the functioning of mitochondria.
Acknowledgments We thank M. Saliola, P. Robbins, and C. Hirschberg for helpful discussions. We also thank M. Makarow for the kind gift of antibodies. The excellent assistance of Mr. F. Castelli is acknowledged. This work was supported by MIUR, COFIN 2002 and by NIH Grants GM59773 to C.A and Grants CA77811 and grant I-0642 from The Robert A. Welch Foundation to R.A.B. D.U. is a recipient of a fellowship from the Pasteur Institute-Cenci Bolognetti Foundation.
References [1] M. Ghislain, A. Goffeau, D. Halachmi, Y. Eilam, Calcium homeostasis and transport are affected by disruption of cta3, a novel gene encoding Ca2þ -ATPase in Schizosaccharomyces pombe, J. Biol. Chem. 265 (1990) 18400–18407. [2] C.S. Park, J.-Y. Kim, C. Crispino, C.C. Chang, D.D.Y. Ryu, Molecular cloning of YlPMR1, a S. cerevisiae PMR1 homologue encoding a novel P-type secretory pathway Ca2þ -ATPase, in the yeast Yarrowia lipolytica, Gene 206 (1998) 107–116. [3] H.K. Rudolph, A. Antebi, G.R. Fink, C.M. Buckley, T.E. Dorman, J. LeVitre, S.D. Lance, J. Mao, D.T. Moir, The yeast secretory pathway is perturbed by mutations in PMR1, a member of a Ca2þ ATPase family, Cell 58 (1989) 133–145. [4] G. D€ urr, J. Strayle, R. Plemper, S. Elbs, S.K. Klee, P. Catty, D.H. Wolf, H.K. Rudolph, The medial-Golgi ion pump PMR1 supplies the yeast secretory pathway with Ca2þ and Mn2þ required for glycosylation, sorting, and endoplasmic reticulum-associated protein degradation, Mol. Biol. Cell 9 (1998) 1149–1162. [5] A. Antebi, G.R. Fink, The yeast Ca2þ -ATPase homologue, PMR1, is required for normal golgi function and localizes in a novel golgi-like distribution, Mol. Biol. Cell 3 (1992) 633–654. [6] J. Strayle, T. Pozzan, H.K. Rudolph, Steady-state free Cað2þÞ in the yeast endoplasmic reticulum reaches only 10 lM and is mainly
1038
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
controlled by the secretory pathway pump pmr1, EMBO J. 18 (1999) 4733–4743. E.C. Bolton, A.S. Mildvan, J.D. Boeke, Inhibition of reverse transcription in vivo by elevated manganese ion concentration, Mol. Cell 9 (2002) 879–889. M.M. Harmsen, M.I. Bruyne, H.A. Raue, Overexpression of binding protein and disruption of the PMR1 gene synergistically stimulate secretion of bovine prochymosin but not plant Thaumatin in yeast, Appl. Microbiol. Biotechnol. 46 (1996) 365–370. D. Uccelletti, P. Mancini, F. Farina, S. Morrone, C. Palleschi, Inactivation of the KlPMR1 gene of Kluyveromyces lactis results in defective cell wall morphogenesis, Microbiology 145 (1999) 1079–1087. W. Bao, H. Fukuhara, Secretion of human proteins from yeast: stimulation by duplication of polyubiquitin and protein disulfide isomerase genes in Kluveromyces lactis, Gene 272 (2001) 103– 110. C. Abeijon, P.W. Robbins, C.B. Hirschberg, Molecular cloning of the Golgi apparatus uridine diphosphate-N -acetylglucosamine transporter from Kluyveromyces lactis, Proc. Natl. Acad. Sci. USA 93 (1996) 5963–5968. D. Uccelletti, F. Farina, A. Morlupi, C. Palleschi, Mutants of Kluyveromyces lactis with altered protein glycosylation are affected in cell wall morphogenesis, Res. Microbiol. 150 (1999) 5–12. D. Bartkeviciute, K. Sasnauskas, Studies of yeast Kluyveromyces lactis mutations conferring super-secretion of recombinant proteins, Yeast 20 (2003) 1–11. J.L. DeRisi, V.R. Iyer, P.O. Brown, Exploring the metabolic and genetic control of gene expression on a genomic scale, Science 278 (1997) 680–686. M.E. Schmitt, T.A. Brown, B.L. Trumpower, A rapid and simple method for preparation of RNA from S. cerevisiae, Nucleic Acids Res. 18 (1990) 3091–3092. C.B. Epstein, J.A. Waddle, W.I. Hale, V. Dave, J. Thornton, T.L. Macatee, H.R. Garner, R.A. Butow, Genome-wide responses to mitochondrial dysfunction, Mol. Biol. Cell 12 (2001) 297–308. C.B. Epstein, W.I. Hale, R.A. Butow, Numerical methods for handling uncertainty in microarray data: an example analyzing perturbed mitochondrial function in yeast, Methods Cell Biol. 65 (2001) 439–452. A. Haselbeck, W. Tanner, O-glycosylation in Saccharomyces cerevisiae is initiated at the endoplasmic reticulum, FEBS Lett. 25 (1983) 335–338. V. Mrsa, T. Seidl, M. Gentzsch, W. Tanner, Specific labelling of cell wall proteins by biotinylation. Identification of four covalently linked O-mannosylated proteins of Saccharomyces cerevisiae, Yeast 13 (1997) 1145–1154. I. Ferrero, A.M. Viola, A. Goffeau, Induction by glucose of an antimycin-insensitive, azide-sensitive respiration in the yeast Kluyveromyces lactis, Antonie Leewenhoek 47 (1981) 11–24.
[21] T. Lodi, I. Ferrero, Isolation of the DLD gene of Saccharomyces cerevisiae encoding the mitochondrial enzyme D -lactate ferricytochrome c oxidoreductase, Mol. Gen. Genet. 238 (1993) 315–324. [22] M. Lussier, M. Gentzsch, A.M. Sdicu, H. Bussey, W. Tanner, Protein O-glycosylation in yeast. The PMT2 gene specifies a second protein O-mannosyltransferase that functions in addition to the PMT1-encoded activity, J. Biol. Chem. 270 (1995) 2770– 2775. [23] W.C. Raschke, C.E. Ballou, Characterization of a yeast mannan containing N-acetyl-D -glucosamine as an immunochemical determinant, Biochemistry 11 (1972) 3807–3816. [24] J.C. Kapteyn, H. Van Den Ende, F.M. Klis, The contribution of cell wall proteins to the organization of the yeast cell wall, Biochem. Biophys. Acta 1426 (1999) 373–383. [25] P. Russo, M. Simonen, A. Uimari, T. Teesalu, M. Makarow, Dual regulation of heat and nutrient stress of the yeast HSP150 gene encoding a secretory glycoprotein, Mol. Gen. Genet. 239 (1993) 273–280. [26] P. Russo, N. Kalkkimen, H. Sareneva, J. Paakkola, M. Makarow, A heat shock gene from Saccharomyces cerevisiae encoding a secretory glycoprotein, Proc. Natl. Acad. Sci. USA 89 (1992) 3671–3675. [27] A.M. Zeeman, M.A. Luttik, C. Thiele, J.P. van Dijken, J.T. Pronk, H.Y. Steensma, Inactivation of the Kluyveromyces lactis KlPDA1 gene leads to loss of pyruvate dehydrogenase activity, impairs growth on glucose and triggers aerobic alcoholic fermentation, Microbiology 144 (1998) 3437–3446. [28] A.M. Zeeman, H.Y. Steensma, The acetyl co-enzyme A synthetase genes of Kluyveromyces lactis, Yeast 20 (2003) 13–23. [29] H.C. Causton, C.T. Harbison, E. Kanin, E.G. Jennings, T.I. Lee, H.L. True, E.S. Lander, R.A. Young, Remodeling of yeast genome expression in response to environmental changes, Mol. Biol. Cell 2 (2001) 323–327. [30] B. Philip, D.E. Levin, Wsc1 and Mid2 are cell surface sensors for cell wall integrity signaling that act through Rom2, a guanine nucleotide exchange factor for Rho1, Mol. Cell. Biol. 21 (2001) 271–280. [31] U.S. Jung, D.E. Levin, Genome-wide analysis of gene expression regulated by the yeast cell wall integrity signalling pathway, Mol. Microbiol. 34 (1999) 1049–1057. [32] M. Brini, Cað2þÞ signalling in mitochondria: mechanism and role in physiology and pathology, Cell Calcium 34 (2003) 399–405. [33] L.J. Lombardia, M. Becerra, E. Rodriguez-Belmonte, N.C. Hauser, M.E. Cerdan, Genome-wide analysis of yeast transcription upon calcium shortage, Cell Calcium 32 (2002) 83–91. [34] K. Mizuta, J.R. Warner, Continued functioning of the secretory pathway is essential for ribosome synthesis, Mol. Cell. Biol. 14 (1994) 2493–2502. [35] D. Uccelletti, F. Farina, P. Mancini, C. Palleschi, KlPMR1 inactivation and calcium addition enhance secretion of nonhyperglycosylated heterologous proteins in Kluyveromyces lactis, J. Biotechnol. 109 (2004) 93–101.
Alterations of O-glycosylation, cell wall, and mitochondrial metabolism in Kluyveromyces lactis cells defective in KlPmr1p, the Golgi Ca2+-ATPase Francesca Farina,a,1 Daniela Uccelletti,a,1 Paola Goffrini,b Ronald A. Butow,c Claudia Abeijon,d and Claudio Palleschia,* a
Department of Developmental and Cell Biology, University of Rome “La Sapienza,” Piazza Aldo Moro 5, 00185 Rome, Italy Department of Genetics, Anthropology, Evolution, University of Parma, Parco Area delle Scienze 11/A, 43100 Parma, Italy c Department of Molecular Biology, University of Texas Southwestern Medical Center, 6000 Harry Hines Blvd., Dallas, TX 75390, USA Department of Molecular and Cell Biology, Goldman School of Dental Medicine, Boston University, 715 Albany Street, Boston, MA 02118, USA b
d
Received 5 April 2004 Available online 7 May 2004
Abstract In yeast the P-type Ca2þ -ATPase of the Golgi apparatus, Pmr1p, is the most important player in calcium homeostasis. In Kluyveromyces lactis KlPMR1 inactivation leads to pleiotropic phenotypes, including reduced N-glycosylation and altered cell wall morphogenesis. To study the physiology of K. lactis when KlPMR1 was inactivated microarrays containing all Saccharomyces cerevisiae coding sequences were utilized. Alterations in O-glycosylation, consistent with the repression of KlPMT2, were found and a terminal N-acetylglucosamine in the O-glycans was identified. Klpmr1D cells showed increased expression of PIRs, proteins involved in cell wall maintenance, suggesting that responses to cell wall weakening take place in K. lactis. We found over-expression of KlPDA1 and KlACS2 genes involved in the Acetyl-CoA synthesis and down-regulation of KlIDP1, KlACO1, and KlSDH2 genes involved in respiratory metabolism. Increases in oxygen consumption and succinate dehydrogenase activity were also observed in mutant cells. The described approach highlighted the unexpected involvement of KlPMR1 in energy-yielding processes. Ó 2004 Elsevier Inc. All rights reserved. Keywords: KlPMR1; Yeast; Heterologous arrays; O-glycans; Cell wall
In animal and plant cells cytosolic free calcium is a major signal-transducing element employed to induce a variety of responses from cell proliferation to cell differentiation. Calcium homeostasis is mainly regulated by P-type Ca2þ -ATPases. Genes encoding these Ca2þ ATPases of the early secretory pathway have been isolated from several organisms [1,2]; the pump encoded by PMR1 gene of Saccharomyces cerevisiae is perhaps the best characterized [3,4]. Pmr1p was localized to the medial Golgi compartment, where it was found to be important for functioning of the secretory pathway [5]. These studies showed that cells lacking functional *
Corresponding author. Fax: +39-064991-2132. E-mail address: [email protected] (C. Palleschi). 1 These authors contributed equally to this work. 0006-291X/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.04.127
PMR1 exhibit defects in protein glycosylation, processing, sorting, and endoplasmic reticulum-associated protein degradation. In yeast, Pmr1p seems to be the major Ca2þ -ATPase contributing to the steady state free Ca2þ concentration in the endoplasmic reticulum. The amount of this cation decreases by 50% in the endoplasmic reticulum of pmr1 mutant cells, while the level of cytosolic calcium increases up to 16-fold despite a compensatory increase in the expression of the vacuolar Ca2þ -ATPase, PMC1 [6]. PMR1 also transports Mn2þ inside the lumen of the Golgi apparatus and appears to be important for the homeostasis of this ion. High concentrations of cytoplasmic Mn2þ are toxic and can interfere with Mg2þ binding sites on proteins. Defective Ty1 retrotransposition in a pmr1 mutant was recently shown to be due to Mn2þ inhibition of reverse
1032
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
transcriptase [7]. In addition, the low concentration of Mn2þ in the Golgi compartment of pmr1 mutant cells leads to defective N- and O-glycosylation [4]. An intriguing additional phenotype resulting from the inactivation of PMR1 is the enhanced release into the growth medium of secretory proteins; this characteristic has been exploited for the production of heterologous proteins [8] but the underlying mechanisms are still not known. We isolated the orthologous KlPMR1 gene of Kluyveromyces lactis and demonstrated its involvement in a variety of phenotypes including alterations in N-linked glycosylation and secretion as well as in cell wall morphogenesis [9]. The cell wall defects include a doubling of cell wall thickness and significant alterations in the glycan compositions; such changes are not detectable at all in the corresponding S. cerevisiae mutant. Moreover, the cell wall phenotypes observed in Klpmr1D cells are not just pleiotropic consequences of the glycosylation defects. In fact, the addition of external calcium almost reverts the Klpmr1D cell wall phenotypes but leaves completely unchanged the defective glycosylation. Significant differences in calcium homeostasis may therefore exist between S. cerevisiae and K. lactis. In addition, although K. lactis has been successfully used as alternative host for heterologous protein production [10], the secretion and glycosylation pathways in this yeast remain poorly understood [11–13]. The pleiotropic phenotypes observed in the Klpmr1 mutant cells prompted us to perform a genome-wide transcription profile analysis of the parental strain versus the mutant to uncover cellular links to the physiological role of KlPMR1. Heterologous S. cerevisiae genome arrays were used, due to the lack of the genome sequence for K. lactis. In this paper we report the involvement of K. lactis Ca2þ ATPase in the initiation and elongation of O-glycosylation. The unexpected presence of N -acetylglucosamine in the O-chains of this yeast was established. The increase in the amount of PIR proteins in the mutant cells indicates the presence of a cell wall integrity mechanism in K. lactis. We also found an unforeseen link between KlPMR1 and mitochondrial metabolism highlighted by increased oxygen consumption and transcriptional alteration of genes encoding for respiratory enzymes. Materials and methods Yeast strains and growth conditions. The strains used in this study were MW278-20C (MATa ade2, leu2, uraA), CPK1 (MATa ade2, leu2 uraA KlPMR1::KanR ), MG1/2 (MATa uraA, arg , lys Kþ pKD1þ ), and KL3 (MATa uraA, mnn2-2, arg Kþ pKD1þ ). Cells were grown on YPD (1% yeast extract, 1% bacto peptone, and 2% glucose) or YPD supplemented with 10 mM CaCl2 . Microarray analysis. Microarrays consisting of 6219 S. cerevisiae genes were prepared essentially as described by DeRisi et al. [14] and were based on PCR amplification of S288C yeast genomic DNA using gene-specific oligo pairs supplied by Research Genetics (Birmingham,
AL). RNA Isolation: cells were grown in YPD medium to an OD600 of 0.7–1.0. Total yeast RNA was prepared using the hot phenol method [15]. The poly(A)þ -mRNAs were prepared by using the Oligotex mRNA kit (Qiagen) according to the manufacturer’s instructions. Cy3- and Cy5-labelled cDNAs were prepared and hybridized to a microarray of 6219 S. cerevisiae genes. Replica experiments were carried out using independent cultures and with the opposite configuration of Cy3 and Cy5. The data were analyzed as described by Epstein et al. [16,17] using a cut-off of value of an average of 3-fold change in replicate hybridizations. O-linked carbohydrate analysis. The method of Haselbeck and Tanner [18] was followed for the isolation of total O-linked carbohydrates from [2-3 H]mannose (18 Ci/mmol; New England [NEN]) radiolabeled cells. Radiolabeled species in the supernatant fraction were subjected to thin-layer chromatography on Silica gel G plates (Merck). The chromatograms were treated with EN3 HANCE reagent (NEN) for fluorography and exposed to Kodak X-OMAT X-ray film at )70 °C. Labeling of cell wall proteins. Yeast cells were grown in 1 L YPD medium to an OD600 of 4–5 and labeled with NHS-LC-Biotin reagent as described in Mrsa et al. [19]. To visualize the biotin-labeled proteins, they were blotted to PVDF membrane (Bio-Rad) which was incubated in streptavidin–horseradish peroxidase conjugated solution (dilution 1:7500) and then developed using the ECL Kit (Amersham). Analysis of Hsp150/Pir2. The proteins in the culture medium of K. lactis strains were separated by SDS–PAGE using a 10% polyacrylamide gel and transferred to PVDF membranes (Bio-Rad). The membrane was incubated with antibody against S. cerevisiae Hsp150, a generous gift of Dr. M. Makarow (University of Helsinki, Finland), at a dilution of 1:2000 and then incubated with goat anti-rabbit IgG conjugated with peroxidase. Finally, the membrane was developed with the ECL detection kit (Amersham). Northern blot analysis. Total RNA of K. lactis strains was extracted by the hot phenol method [21]. The RNAs were quantified by absorption (OD260 ) and separated by denaturing agarose electrophoresis. Following electrophoresis the RNAs were transferred to nylon membranes and hybridized with 32 P-labeled random primed probes (Roche). All the probes were PCR amplified from K. lactis DNA genome (the sequences were kindly provided by Prof. Bolotin-Fukuhara, Paris). The PCR product of KlPMT2 was of 1.3 Kb using primers 50 -TTCTCGTTACCAACCGTTTG-30 and 50 -CACGGCAACCAAC ATAA-30 , of KlACS2 was of 3.0 Kb using primers 50 -TCTAAGTCA ATTCTTCAGCGG-30 and 50 -AAAGCTCCAGTGGTTTCGAGA30 , of KlPDA1 was of 1.5 Kb using primers 50 -TCAGGAACTCAGA GCTTCTTC-30 and 50 -CCATACCGTGAATGAAATGAA-30 , of KlIDP1 was of 900 bp using primers 50 -CGATTGCCATTGCCCTA AGT-30 and 50 -TGAAAGAGGCATGAGCGAAC-30 , of KlSDH2 was of 800 bp using primers 50 -TCTGCGATATCTACCTGGAT TC-30 and 50 -TATCCGTTTCCGTTAAGTTTTCAGA-30 , and of KlACO1 was of 800 bp using primers 50 -ATGTTGTCTGCTC GTGTTGC-30 and 50 -CGGATGTAGTGGCACCGATT-30 . Respiration and enzyme activity. The measurements of respiratory activity were performed according to Ferrero et al. [20] preparation of mitochondria and determination of succinate dehydrogenase activity (SDH) (EC 1.3.99.1) were carried out according to Lodi and Ferrero [21]. The SDH activity was expressed as nmol/min/mg protein.
Results To identify genes regulated in response to inactivation of the KlPMR1 expression profiles of cells carrying the wild type and the null allele were compared. Due to the lack of comprehensive genome sequence data for K. lactis, we took advantage of the overall high similarity
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
1033
Table 1 Genes that show strongly altered expression in Klpmr1D mutant ORF
Gene name
Array (fold)
Gene product
YAL 023C YGR 088C YDL 066W YLL 041C YLR 153C YER 178W YKL 164C YJL 159W YKL 163W YKL 158C YPL 211W YOL 077C YNL 244C YGR 159C
PMT2 CTT1 IDP1 SDH2 ACS2 PDA1 PIR1 PIR2 PIR3 CIS3 NIP7 BRX1 SUI1 NRS1
)3.5 [)3] )4.5 )9 [)4.5] )3.5 [)2.5] 3.7 [2.5] 3.5 [1.5] 6.5 6 3.5 3.5 6.5 6 4 5.7
Mannosyltransferase responsible for initiation of O-glycosylation Cytosolic catalase T Isocitrate dehydrogenase (NADP+), mitochondrial Succinate dehydrogenase (ubiquinone) iron–sulfur protein (Ip) subunit Acetyl-CoA synthetase Pyruvato dehydrogenase complex E1-alpha subunit Protein required for tolerance to heat shock O-glycosylated cell wall protein required for tolerance to heat shock Protein with similarity to members of the PIR family proteins Cell wall protein with similarity to members of the PIR family Nucleolar protein required for efficient 60S ribosome subunit biogenesis Protein required for biogenesis of the 60S ribosomal subunit 16 kDa subunit of translation initiation factor eIF3 Nucleolar protein involved in processing 20S to 18S rRNA
Values in square brackets were obtained by homologous Northern hybridization.
of known K. lactis genes compared to S. cerevisiae ones and performed a heterologous microarray study. Labeled cDNA probes were generated from mRNA of wild type and Klpmr1D strains as described in Materials and methods and hybridized at 50 °C onto S. cerevisiae slides. Under these conditions we obtained the higher percentage of hybridization signals above background (about 60% of 6219 S. cerevisiae ORFs in all the performed experiments). Our attention was focused on those genes whose geometric mean expression level increased or decreased at least 3-fold in three independent experiments in Klpmr1D background. We identified 14 genes that are reported in Table 1. Among the genes that are down-regulated by the loss of the Ca2þ -ATPase was PMT2, a gene involved in the initiation of the O-glycosylation process [22]. It was puzzling to find such a strong defect in the step that initiates the O-linked sugar chains because their elongation is also impaired as discussed below. Several genes involved in the TCA cycle as well as in the oxidative phosphorylation were also down-regulated in Klpmr1D cells, although not all of them met the threshold to be included in Table 1. These expression profiles could be related to a reduction in the growth rate of the Klpmr1D cells. In fact, in rich medium, Klpmr1D cells showed a generation time of 3.5 h compared to 2 h of the wild type cells. Several genes were up-regulated in response to the loss of KlPMR1, such as those related to cell wall morphogenesis and genes involved in the assembly and processing of rRNA. O-glycosylated chains in K. lactis contain N-acetylglucosamine and are severely reduced in the Klpmr1D strain Among the genes down-regulated in Klpmr1D cells was PMT2 and this was also confirmed by homologous Northern hybridization (Fig. 1A). We therefore analyzed total O-glycosylated proteins; for this we labeled wild type and mutant cells with [2-3 H]mannose, released
their O-linked chains by b-elimination, and analyzed them by thin-layer chromatography (Fig. 1B). The oligosaccharide pattern from K. lactis wild-type strain showed a profile similar to that of the S. cerevisiae counterpart, where it is possible to observe species containing between one and five mannose residues (Fig. 1B, lanes 1 and 3). K. lactis pmr1D cells showed a partial block in their ability to extend O-linked mannan chains manifested by accumulation of M1 and reduction of M3, M4, and M5; the amount of M2 was quite similar to that of wild type (Fig. 1B, lane 2). We found a 4-fold decrease in pmol of b-eliminated mannose per mg wet weight in Klpmr1D cells as compared with wild type, indicating a severe reduction in the total O-mannosylated glycoproteins and an altered elongation pattern. The structure of O-linked chains has never been characterized in K. lactis. To gain insight into the structure of O-linked chains in wild type and Klpmr1D strains, we treated them with jack bean a-mannosidase (Fig. 2). The most abundant M2 as well as M3 chains were completely converted to free mannose, indicating a-linkages. Surprisingly, M5 chains were not digested by the exomannosidase, indicating either that the mannose at their non-reducing end is not in a-linkage or that other sugars may be present. M4 chains were only partially resistant, suggesting the existence of two different kinds of M4 chains in K. lactis (Fig. 2A, lanes 1 and 2). This is different from S. cerevisiae, which only has alinked mannose residues in its O-glycans and not other sugars (Fig. 2A, lanes 3 and 4). We hypothesized that the presence of an additional sugar such as N -acetylglucosamine accounts for the differences in K. lactis Ochains; the N-linked sugars of K. lactis mannoproteins have terminal a1–2 linked N -acetylglucosamine residue that is absent from the outer mannan chains of S. cerevisiae [23]. The gene encoding the K. lactis UDP-N acetylglucosamine Golgi transporter, MNN2, has been isolated and characterized [17]; we thus decided to take advantage of this knowledge to test our hypothesis. The
1034
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
Chains with 5 mannoses were completely absent in mnn2-2 cells strongly suggesting that the terminal sugar is N -acetylglucosamine (Fig. 2B, lanes 1 and 2). We also detected a partial reduction in the amount of chains with four sugar residues in the mutant that cannot add N acetylglucosamine. This suggests the existence of two types of four residue chains, one with four a-linked mannoses and another with three mannoses and a terminal N -acetylglucosamine (Fig. 2C). Digestion of Olinked chains synthesized by the mutant in the UDP-N acetylglucosamine transporter confirmed this hypothesis because the reduced amount of M4 produced was completely sensitive to a-mannosidase treatment (Fig. 2B, lane 3). Cell wall alterations in Klpmr1D cells
Fig. 1. O-glycosylation is altered in Klpmr1D cells. (A) Northern blot analysis of KlSDH2 in Klpmr1D cells. Total RNA was extracted from KlPMR1 and Klpmr1D strains (lanes 1 and 2, respectively) grown on YPD. The same amount of RNA (10 lg) was loaded on each lane. The ethidium bromide-stained gel of the autoradiogram is shown in the lower part of the panel. Densitometric quantification of mRNA was performed by the computer program Phoretix 1D and mRNA loading was normalized using the rRNA bands. (B) Total oligosaccharides from MW278-20C (Kl wt), CPK1 (Kl D), and BY4741 (Sc wt) strains were radiolabeled with [2-3 H]mannose. Alkali releasable saccharides were separated on thin layer chromatography (butanol, ethanol, water (5:3:2)) and subjected to autoradiography (50,000 cpm/lane). M1 mannose, M2 mannobiose, M3 mannotriose, M4 mannotetraose, and M5 mannopentose.
bulk of O-glycoproteins were analyzed from mnn2-2 cells carrying the mutated allele of this gene, which totally lacks N -acetylglucosamine in the outer mannan chains.
Klpmr1D cells show defects in cell wall composition and morphology [9]. In the genome-wide analysis we found an increased expression of all members of the gene family encoding for PIR proteins (proteins with internal repeats), except YJL160c (Table 1); a class of cell wall proteins covalently linked to b-1,3-glucans [24]. Cell surface biotinylation of proteins was performed in K. lactis wild type and mutant cells; labeled proteins retained in the cell wall via non-covalent or S–S bridges were extracted from purified cell wall fractions by SDS treatment under reducing conditions. The PIR proteins were then extracted with NaOH and analyzed by electrophoresis after normalization per OD600 (Fig. 3A). These experiments show the presence of PIR-related proteins in K. lactis and their relative abundance in the Klpmr1D strain with respect to wild type, in agreement with the array data presented in Table 1. The most abundant PIR protein in S. cerevisiae, PIR2/HSP150, is a heat shock inducible protein that is only partially retained in the cell wall while most is secreted to the medium [25,26]. A Western blot of material in the media secreted from wild type and mutant cells was analyzed with a S. cerevisiae antiPIR2 antibody (Fig. 3B). The Klpmr1D strain secreted three times more protein than did the wild type strain; this is consistent with the observed increase in the content of PIR proteins. Altered expression of genes encoding for mitochondrial enzymes in the Klpmr1D strain Transcription of several genes involved in mitochondrial functions was altered in the Klpmr1D strain. Among them, KlIDP1 (isocitrate dehydrogenase) as well as KlSDH2 (a subunit of the succinate dehydrogenase) and KlACO1 (aconitase) were down-regulated in mutant cells, whereas KlPDA1 (alpha subunit of pyruvate dehydrogenase) and KlACS2 (acetyl CoA synthetase) expression was up-regulated. Since the
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
1035
Fig. 2. N -acetylglucosamine is present in K. lactis O-glycosydic chains. (A) Total O-linked chains from MW278-20C (Kl wt, lane 1) and BY4741 (Sc wt, lane 3) strains were treated with jack bean a-mannosidase (lanes 2 and 4, respectively), separated on thin layer chromatography as in Fig. 1. (B) Total O-linked chains from MG1/2 (Kl wt, lane 1) and KL3 (mnn2-2, lane 2) strains were separated by thin layer chromatography. (C) Schematic representation of O-linked chains in K. lactis.
array data come from heterologous hybridizations we decided to perform individual Northern blot analyses with K. lactis specific probes. Total RNA was extracted from cells grown in the same conditions utilized for the array experiments. The results showed an increase of 2.5- and 1.5-fold in the transcript levels of KlACS2 and KlPDA1 in the mutant versus wild type strain (Fig. 4A). These genes are functionally homologues to the corresponding S. cerevisiae genes [27,28]. In the same experiments we also observed a 4.5-, 2.5-, and a 2-fold reduction of KlIDP1, KlSDH2, and KlACO1 transcripts, respectively, in the Klpmr1D mutant. The KlACO1 gene was not included in Table 1 since it did not match the chosen threshold; how-
ever, its expression data were consistent in at least two out of three array experiments and were then analyzed by Northern (Fig. 4B). These independent experiments therefore confirmed the array data. On the basis of these results we analyzed the oxygen consumption in the parental and in the mutant strain. The respiratory activity of Klpmr1D mutant was 43 ll O2 /h/mg d.w. versus 30 of parental strain and antimycin A reduced this activity at the same level in both strains; the cytochrome content was, however, almost identical in the two strains (not shown). In addition, the SDH activity of the mutant was 1.6-fold higher than that measured in the parental strain (Fig. 5).
1036
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
Fig. 3. KlPIR-like proteins are increased in Klpmr1D cells. (A) Cell wall PIR-like proteins were extracted from biotinylated intact cells of MW278-20C (wt, lane 1) and CPK1 (D, lane 2) by 30 mM NaOH. The extracts were separated on SDS–PAGE, blotted and stained with streptavidin–horseradish peroxidase. Proteins corresponding to 40 OD600 were used per lane. (B) Culture medium amounts corresponding to OD600 0.1 of cells were taken from MW278-20C (wt, lane 1) and CPK1 (D, lane 2) liquid cultures, separated by SDS–PAGE, and probed with ScHsp150 antibody. Protein molecular weight standards are indicated alongside of both panels.
Discussion In the present work we have studied the physiological effects of KlPMR1 inactivation using heterologous genechip microarray technology. Because the K. lactis genome sequence is not yet completely available, we took advantage of the availability of S. cerevisiae microarrays. We obtained insight into the involvement of KlPMR1 in several cellular processes. The data obtained with the “heterologous” approach were validated by subsequent homologous Northern blots and biochemical experiments. In K. lactis, inactivation of KlPMR1 causes a reduction in mannose outer-chain extension of N-linked glycoproteins as well as defective cell wall morphogenesis [9]. We found defects in the initiation and extension of O-glycosylation in Klpmr1D strains. The unexpected down-regulation of PMT2, one of the genes involved in initiating O-linked glycosylation, prompted us to characterize K. lactis O-chains. This resulted in the first report on the presence of GlcNAc at the non-reducing end of the O-linked oligosaccharides in all fungi and yeasts studied to date. The PMT2 gene in S. cerevisiae encodes a mannosyltransferase responsible for initiation of
Fig. 4. Expression pattern of mitochondrial enzyme genes. Northern blot analyses of KlACS2 and KlPDA1 (A) KlIDP1, KlSDH2, and KlACO1 (B), in Klpmr1D cells. Total RNA was extracted from KlPMR1 and Klpmr1D strains (lanes 1 and 2, respectively) grown on YPD. The same amount of RNA (10 lg) was loaded on each lane. The ethidium bromide-stained gels of the autoradiogram are shown in the lower part of each panel. Densitometric quantification of mRNA was performed by the computer program Phoretix 1D and mRNA loading was normalized using the rRNAs bands.
Fig. 5. Succinate dehydrogenase activity in wild type and KlPmr1D strains. Enzyme activity was measured in mitochondria extracts prepared by exponentially growing cells and was expressed as nmol of substrate utilized per min per mg proteins. All values are means of three independent experiments. In no case the variation was higher than 15%.
O-glycosylation in the endoplasmic reticulum (ER) [22]. Analysis of bulk cell mannoproteins of Klpmr1D mutants showed that the normal elongation of O-chains
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
failed in this mutant. Elongation enzymes require manganese as a cofactor in S. cerevisiae and it is therefore likely that the elongation defect in Klpmr1D is due to the reduced availability of this ion in the Golgi apparatus. However, defects in the initiation of O-glycosylation in Klpmr1D were only discovered by our genome-wide analysis of gene expression. The molecular mechanism by which PMT2 is down-regulated is not known at the present time. We are tempting to speculate that because of the elongation defect the product of Pmt2p (M1) accumulates and product inhibition of the enzyme might occur. A feedback signaling would then reduce the expression of PMT2. It has been recently shown in S. cerevisiae that PMT2 is one of several genes co-repressed under different stress conditions, such as changes in temperature, oxidation, nutrients, pH, and osmolarity [29]. The molecular mechanism(s), however, are unclear. Cells carrying the inactivated allele of KlPMR1 have alterations in O-glycosylation and cell wall organization and could possibly undergo activation of signaling pathways similar to those induced by environmental stresses. A recent report demonstrates that Pmt2p is specifically responsible for the O-mannosylation of Mid2p, and that the glycosylation state of Mid2p strongly affects its functioning as a signaling element of cell wall stresses [30]. The loss of K. lactis Ca2þ -ATPase caused changes in the cell wall structure that could elicit a response, including the increase in the expression of PIR-related proteins. All members of this gene family, except YJL160c, were positively regulated by cell integrity signaling through activation of Mpk1 [24,31]. Our data indicate the presence of PIR related proteins in K. lactis and suggest the existence of cell wall integrity signaling pathways. The genome-wide analyses in Klpmr1D strains highlighted a genetic interaction between the Golgi Ca2þ -ATPase and mitochondrial metabolism. This is supported by the observed increase in the transcription of KlPDA1 and KlACS2 and down-regulation of KlSDH2, KlIDP1, and KlACO1 genes. In S. cerevisiae, Pmr1p is able to control Ca2þ homeostasis in the early secretory compartments; the absence of Pmr1p causes an increase in cytosolic Ca2þ [6]. Mitochondrial Ca2þ uptake following the rise of cytosolic Ca2þ concentration has been demonstrated in many mammalian cell types. The physiologic role of such mitochondrial Ca2þ uptake is mainly related to the control of energy metabolism, as supported by the observations that some dehydrogenases of TCA cycle as well as the ATP production are activated by Ca2þ [32]. We could speculate that similar events could take place in K. lactis cells: the absence of KlPmr1p would cause a rise in cytosolic Ca2þ followed by Ca2þ uptake by the mitochondria resulting in stimulation of energy metabolism. The increased mitochondrial metabolism,
1037
supported by the increasing in O2 uptake and SDH activity, would then activate a nucleo-mitochondria crosstalk ending in fine-tuning the transcription of genes encoding for mitochondrial enzymes of energy metabolism. It is worth noting that a genome-wide transcription analysis in S. cerevisiae has shown connections between calcium shortage, energetic metabolism, and cell-cycle progression [33]. An increase in the transcription of genes involved in ribosome assembly or in processing of rRNA was also found in Klpmr1D cells. Defects in several points of the secretory pathway that block protein transport downregulate ribosomal synthesis, through the repression of transcription of both rRNA and ribosomal protein genes in S. cerevisiae [34]. Although the mechanism is not yet understood, Klpmr1D shows both increased secretory capabilities [35] and up-regulation of genes involved in rRNA biosynthesis, suggesting that these cellular events are linked. In conclusion, this report highlights the involvement of K. lactis Ca2þ -ATPase in relevant cell processes like O-glycosylation, cell wall biogenesis and, mostly unexpected, in the functioning of mitochondria.
Acknowledgments We thank M. Saliola, P. Robbins, and C. Hirschberg for helpful discussions. We also thank M. Makarow for the kind gift of antibodies. The excellent assistance of Mr. F. Castelli is acknowledged. This work was supported by MIUR, COFIN 2002 and by NIH Grants GM59773 to C.A and Grants CA77811 and grant I-0642 from The Robert A. Welch Foundation to R.A.B. D.U. is a recipient of a fellowship from the Pasteur Institute-Cenci Bolognetti Foundation.
References [1] M. Ghislain, A. Goffeau, D. Halachmi, Y. Eilam, Calcium homeostasis and transport are affected by disruption of cta3, a novel gene encoding Ca2þ -ATPase in Schizosaccharomyces pombe, J. Biol. Chem. 265 (1990) 18400–18407. [2] C.S. Park, J.-Y. Kim, C. Crispino, C.C. Chang, D.D.Y. Ryu, Molecular cloning of YlPMR1, a S. cerevisiae PMR1 homologue encoding a novel P-type secretory pathway Ca2þ -ATPase, in the yeast Yarrowia lipolytica, Gene 206 (1998) 107–116. [3] H.K. Rudolph, A. Antebi, G.R. Fink, C.M. Buckley, T.E. Dorman, J. LeVitre, S.D. Lance, J. Mao, D.T. Moir, The yeast secretory pathway is perturbed by mutations in PMR1, a member of a Ca2þ ATPase family, Cell 58 (1989) 133–145. [4] G. D€ urr, J. Strayle, R. Plemper, S. Elbs, S.K. Klee, P. Catty, D.H. Wolf, H.K. Rudolph, The medial-Golgi ion pump PMR1 supplies the yeast secretory pathway with Ca2þ and Mn2þ required for glycosylation, sorting, and endoplasmic reticulum-associated protein degradation, Mol. Biol. Cell 9 (1998) 1149–1162. [5] A. Antebi, G.R. Fink, The yeast Ca2þ -ATPase homologue, PMR1, is required for normal golgi function and localizes in a novel golgi-like distribution, Mol. Biol. Cell 3 (1992) 633–654. [6] J. Strayle, T. Pozzan, H.K. Rudolph, Steady-state free Cað2þÞ in the yeast endoplasmic reticulum reaches only 10 lM and is mainly
1038
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
F. Farina et al. / Biochemical and Biophysical Research Communications 318 (2004) 1031–1038
controlled by the secretory pathway pump pmr1, EMBO J. 18 (1999) 4733–4743. E.C. Bolton, A.S. Mildvan, J.D. Boeke, Inhibition of reverse transcription in vivo by elevated manganese ion concentration, Mol. Cell 9 (2002) 879–889. M.M. Harmsen, M.I. Bruyne, H.A. Raue, Overexpression of binding protein and disruption of the PMR1 gene synergistically stimulate secretion of bovine prochymosin but not plant Thaumatin in yeast, Appl. Microbiol. Biotechnol. 46 (1996) 365–370. D. Uccelletti, P. Mancini, F. Farina, S. Morrone, C. Palleschi, Inactivation of the KlPMR1 gene of Kluyveromyces lactis results in defective cell wall morphogenesis, Microbiology 145 (1999) 1079–1087. W. Bao, H. Fukuhara, Secretion of human proteins from yeast: stimulation by duplication of polyubiquitin and protein disulfide isomerase genes in Kluveromyces lactis, Gene 272 (2001) 103– 110. C. Abeijon, P.W. Robbins, C.B. Hirschberg, Molecular cloning of the Golgi apparatus uridine diphosphate-N -acetylglucosamine transporter from Kluyveromyces lactis, Proc. Natl. Acad. Sci. USA 93 (1996) 5963–5968. D. Uccelletti, F. Farina, A. Morlupi, C. Palleschi, Mutants of Kluyveromyces lactis with altered protein glycosylation are affected in cell wall morphogenesis, Res. Microbiol. 150 (1999) 5–12. D. Bartkeviciute, K. Sasnauskas, Studies of yeast Kluyveromyces lactis mutations conferring super-secretion of recombinant proteins, Yeast 20 (2003) 1–11. J.L. DeRisi, V.R. Iyer, P.O. Brown, Exploring the metabolic and genetic control of gene expression on a genomic scale, Science 278 (1997) 680–686. M.E. Schmitt, T.A. Brown, B.L. Trumpower, A rapid and simple method for preparation of RNA from S. cerevisiae, Nucleic Acids Res. 18 (1990) 3091–3092. C.B. Epstein, J.A. Waddle, W.I. Hale, V. Dave, J. Thornton, T.L. Macatee, H.R. Garner, R.A. Butow, Genome-wide responses to mitochondrial dysfunction, Mol. Biol. Cell 12 (2001) 297–308. C.B. Epstein, W.I. Hale, R.A. Butow, Numerical methods for handling uncertainty in microarray data: an example analyzing perturbed mitochondrial function in yeast, Methods Cell Biol. 65 (2001) 439–452. A. Haselbeck, W. Tanner, O-glycosylation in Saccharomyces cerevisiae is initiated at the endoplasmic reticulum, FEBS Lett. 25 (1983) 335–338. V. Mrsa, T. Seidl, M. Gentzsch, W. Tanner, Specific labelling of cell wall proteins by biotinylation. Identification of four covalently linked O-mannosylated proteins of Saccharomyces cerevisiae, Yeast 13 (1997) 1145–1154. I. Ferrero, A.M. Viola, A. Goffeau, Induction by glucose of an antimycin-insensitive, azide-sensitive respiration in the yeast Kluyveromyces lactis, Antonie Leewenhoek 47 (1981) 11–24.
[21] T. Lodi, I. Ferrero, Isolation of the DLD gene of Saccharomyces cerevisiae encoding the mitochondrial enzyme D -lactate ferricytochrome c oxidoreductase, Mol. Gen. Genet. 238 (1993) 315–324. [22] M. Lussier, M. Gentzsch, A.M. Sdicu, H. Bussey, W. Tanner, Protein O-glycosylation in yeast. The PMT2 gene specifies a second protein O-mannosyltransferase that functions in addition to the PMT1-encoded activity, J. Biol. Chem. 270 (1995) 2770– 2775. [23] W.C. Raschke, C.E. Ballou, Characterization of a yeast mannan containing N-acetyl-D -glucosamine as an immunochemical determinant, Biochemistry 11 (1972) 3807–3816. [24] J.C. Kapteyn, H. Van Den Ende, F.M. Klis, The contribution of cell wall proteins to the organization of the yeast cell wall, Biochem. Biophys. Acta 1426 (1999) 373–383. [25] P. Russo, M. Simonen, A. Uimari, T. Teesalu, M. Makarow, Dual regulation of heat and nutrient stress of the yeast HSP150 gene encoding a secretory glycoprotein, Mol. Gen. Genet. 239 (1993) 273–280. [26] P. Russo, N. Kalkkimen, H. Sareneva, J. Paakkola, M. Makarow, A heat shock gene from Saccharomyces cerevisiae encoding a secretory glycoprotein, Proc. Natl. Acad. Sci. USA 89 (1992) 3671–3675. [27] A.M. Zeeman, M.A. Luttik, C. Thiele, J.P. van Dijken, J.T. Pronk, H.Y. Steensma, Inactivation of the Kluyveromyces lactis KlPDA1 gene leads to loss of pyruvate dehydrogenase activity, impairs growth on glucose and triggers aerobic alcoholic fermentation, Microbiology 144 (1998) 3437–3446. [28] A.M. Zeeman, H.Y. Steensma, The acetyl co-enzyme A synthetase genes of Kluyveromyces lactis, Yeast 20 (2003) 13–23. [29] H.C. Causton, C.T. Harbison, E. Kanin, E.G. Jennings, T.I. Lee, H.L. True, E.S. Lander, R.A. Young, Remodeling of yeast genome expression in response to environmental changes, Mol. Biol. Cell 2 (2001) 323–327. [30] B. Philip, D.E. Levin, Wsc1 and Mid2 are cell surface sensors for cell wall integrity signaling that act through Rom2, a guanine nucleotide exchange factor for Rho1, Mol. Cell. Biol. 21 (2001) 271–280. [31] U.S. Jung, D.E. Levin, Genome-wide analysis of gene expression regulated by the yeast cell wall integrity signalling pathway, Mol. Microbiol. 34 (1999) 1049–1057. [32] M. Brini, Cað2þÞ signalling in mitochondria: mechanism and role in physiology and pathology, Cell Calcium 34 (2003) 399–405. [33] L.J. Lombardia, M. Becerra, E. Rodriguez-Belmonte, N.C. Hauser, M.E. Cerdan, Genome-wide analysis of yeast transcription upon calcium shortage, Cell Calcium 32 (2002) 83–91. [34] K. Mizuta, J.R. Warner, Continued functioning of the secretory pathway is essential for ribosome synthesis, Mol. Cell. Biol. 14 (1994) 2493–2502. [35] D. Uccelletti, F. Farina, P. Mancini, C. Palleschi, KlPMR1 inactivation and calcium addition enhance secretion of nonhyperglycosylated heterologous proteins in Kluyveromyces lactis, J. Biotechnol. 109 (2004) 93–101.