A Regulatory Mutant of Hansenula polymorpha Exhibiting Methanol Utilization Metabolism and Peroxisome Proliferation in Glucose

Abstract

Mutant LGM-128 of Hansenula polymorpha harbors the recessive mutation glr2-1 which confers a complex pleiotropic phenotype, the major feature of which is the metabolically unnecessary induction of methanol utilization metabolism (C₁ metabolism) during growth on glucose, whether or not methanol is in the medium. Therefore, in this mutant, peroxisomes are formed and proliferate upon cultivation in glucose-containing media. In these media, LGM-128 shows induction levels of C₁ metabolism that are similar to those observed in methanol-containing media. This indicates that GLR2 controls the repression-derepression process stimulated by glucose and that the induction process triggered by methanol plays only a minor role in activating C₁ metabolism. Cultivating LGM-128 in methanol and then transferring it to glucose media revealed that active degradative processes occur, leading to the disappearance of C₁ metabolism. This observation suggests that, although stimulated by glucose, the two processes are controlled by elements which are, at least in part, distinct. Finally, glr2-1 does not affect ethanol repression, suggesting that in H. polymorpha the two repressing circuits are separated.

Peroxisomes are eukaryotic organelles with a central role in hydrogen peroxide metabolism. They are involved in various metabolic processes and contain one or more hydrogen peroxide-forming flavin oxidases, a catalase which decomposes the hydrogen peroxide to molecular oxygen and water, and a variety of other enzymes (more than 50 have been reported for mammalian peroxisomes [19, 25, 41]). Peroxisomal biogenesis and proliferation are highly regulated processes which are under the control of nuclear genes. They include the acquisition of membrane components (lipids and proteins), the import of matrix proteins and of various metabolites, and duplication of the organelles and their distribution during cell division. Genetic and biochemical studies are leading to the identification of different components of these processes, such as the signals that target matrix proteins to the peroxisome and the peroxines, a heterogeneous group of proteins, which appear to be essential for peroxisome assembly (3, 4, 20, 21, 23, 24, 36, 37, 47).

In yeasts, peroxisomes are inducible organelles responding mainly to nutritional stimuli, since they often contain enzymes involved in carbon or nitrogen utilization (39, 45, 46). In Hansenula polymorpha (syn. Pichia angusta) as in other facultative methylotrophic yeasts, when methanol is the only carbon source in a flask culture, the peroxisomes proliferate; concomitantly, methanol utilization metabolism (C₁ metabolism), three steps of which are peroxisomal, is induced (Fig. 1). The presence of other carbon sources, such as glucose or ethanol, abolishes both C₁ metabolism and the proliferation of C₁ peroxisomes, even in the presence of methanol (45). Other peroxisome proliferators in this yeast are certain nitrogen sources, such as methylamine, in accordance with the peroxisomal localization of some of the enzymes involved in their utilization (e.g., methylamine oxidase) (45).

FIG. 1.

Diagrammatic representation of C₁ metabolism. Xu5P, xylulose 5-phosphate.

So far, the mechanism of peroxisomal glucose repression in H. polymorpha has not been studied, although there have been sporadic reports describing glucose repression mutants (11, 26, 34). Some of the genes encoding the enzymes of C₁ metabolism have been cloned and characterized, such as the MOX gene, encoding the peroxisomal enzyme, methanol oxidase (16). Functional analysis of the MOX promoter has led to the identification of regions required for glucose repression of its transcription (22). It has also been shown that the DAS, CAT, and FMD genes, encoding three other major enzymes of C₁ metabolism, are also glucose repressed (9, 13, 22). Similarly, the H. polymorpha genes PER8 (PEX9), PER9 (PEX3), and PEX14 (PER10), all encoding peroxines required during methanol growth, are subject to transcriptional repression by glucose (1, 4, 15, 38).

It is likely that most, if not all, of the genes encoding products involved in C₁ metabolism undergo glucose repression of their transcription and are—to a certain extent—coordinately regulated. To understand the regulation of these genes and to elucidate the mechanism of glucose repression acting upon C₁ metabolism and the peroxisomes, it is important to identify and characterize genetic factors involved in these processes. Here we describe a regulatory mutant of H. polymorpha in which C₁ metabolism and the proliferation of peroxisomes are repressed normally during growth on ethanol but are induced during growth on glucose, even in the absence of methanol. In this mutant, however, the activity of the peroxisomal degradation processes triggered by glucose after a shift from methanol to glucose are unaffected.

MATERIALS AND METHODS

Strains.

All strains are H. polymorpha homothallic haploids, derivatives of the wild-type strain NCYC-495. LGM-128 (met6-1 glr2-1) is a GLD1 (glucose-positive) segregant of mutant LGM-242.11.1 (leu1-1 glr2-1 gld1-1), screened in our laboratory (see results). M6 (met6) and L1 (leu1-1) were used for crosses and as control strains.

Media.

All YP media contained 1% yeast extract and 2% peptone. In addition, YPD contained 2% glucose and YPE contained 2% ethanol. All M media contained 0.2% Yeast Nitrogen Base without amino acids (Difco); when necessary, 0.006% leucine or 0.002% methionine was added. In addition, MD contained 2% glucose, ME contained 2% ethanol, and Mm contained 1% methanol. MDm contained 2% glucose plus 1% methanol. When necessary, 1.6% agar was added.

Cultures.

All cultures were grown in 250-ml flasks with 100 ml of medium, using an orbital incubator at 37°C and 220 rpm. For induction experiments, after 12-h precultivations in YPE, 10⁶ cells/ml were inoculated in MD, MDm, and Mm. When necessary, to prolong the exponential phase, aliquots of exponentially growing cells were reinoculated into the same medium. Cell number (direct count), enzyme activities, and MOX messenger levels were determined at fixed intervals. For repression experiments, after precultivations in Mm to mid-exponential phase, 10⁶ cells/ml were inoculated in MD. Cell number and enzyme activities were determined at fixed intervals.

Genetic methods.

Mutants with alcohol oxidase (AO) activities in media containing glucose were obtained by plating 10⁷ LGM 242.1 cells (unmutagenized) on plates containing minimal medium and methanol plus glucose and incubating at 37°C for 15 days. LGM-128 is a Glu⁺ Mgm⁺ Mog⁺ Gam⁺ segregant of one of those mutants (see results). Crosses and sporulations were on 2% malt extract plates.

Biochemical methods.

Crude extracts were obtained by glass bead vortexing in 50 mM potassium phosphate buffer, pH 7.0. Enzyme activities of methanol oxidase (AO; EC 1.1.3.13), catalase (CAT; EC 1.11.1.6), formate dehydrogenase (FDH; EC 1.2.1.2), and glucose phosphorylation were assayed by established procedures (2, 17, 42, 43). Protein concentrations were determined with a Bio-Rad kit (catalog no. 500-0006; Bio-Rad Laboratories, Richmond, Calif.) (standard, bovine serum albumin). Glucose and methanol concentrations in the cultures were determined by high-pressure liquid chromatography.

MOX mRNA levels.

RNA was extracted by the method of Schmitt et al. (31) and analyzed by Northern blotting, using formaldehyde as a denaturing agent (29). RNA was transferred to a Hybond-N⁺ membrane (Amersham, Amersham, United Kingdom) and hybridized under standard conditions with probes made by the random primer method (Sambrook et al. [29]) using DNA fragments derived from the MOX gene of H. polymorpha and from genes coding for the 25S rRNA of Saccharomyces cerevisiae. The latter probe was used as an internal loading control for the assessment of MOX mRNA levels.

Electron microscopy.

Whole cells were washed twice in distilled water, fixed in 1.5% (wt/vol) potassium permanganate for 20 min, washed three times, and postfixed in 1% uranyl acetate. After dehydration in a graded series of ethanol-water solutions, the specimens were embedded in Epon araldite. Ultrathin sections (ca. 30 nm thick) were stained in a solution of lead citrate for 40 s and observed by transmission electron microscopy (TEM) (80 kV; Philips CM12).

RESULTS

Isolation and characterization of LGM-128.

Mutant LGM-128 of H. polymorpha harbors the recessive pleiotropic mutation glr2-1, which determines three characteristic phenotypes: Mog⁺ (AO activities in media containing glucose), Mgm⁺ (AO activities in media containing glucose plus methanol) and Gam⁺ (growth in media containing glucosamine plus methanol). glr2 mutants were obtained in our laboratory as a result of a large-scale screen which yielded more than 200 mutants characterized by significant AO activities in media containing glucose or glucose plus methanol. In particular, the strategy adopted to obtain LGM-128, employed a glucose-negative (Glu⁻) strain, harboring the gld1-1 mutation, which causes reduced activities of various glycolytic enzymes (unpublished data). This strain (LGM 242.1) exhibits wild-type C₁ metabolism and wild-type glucose repression (Mgm⁻ Mog⁻) and does not grow on methanol- plus glucose-containing minimal medium plates, since the glucose repression mechanisms impede the activation of the C₁ metabolism. Therefore, we selected our Mgm⁺ mutants on this medium. After 15 days of incubation, several colonies were obtained on each plate. Of these, some had simply reverted to the Glu⁺ phenotype and retained wild-type glucose repression. Others were still Glu⁻ and showed altered glucose repression of the C₁ metabolism (Mgm⁺). Interestingly, some of these mutants showed also Mog⁺ and Gam⁺ phenotype. One of these mutants, LGM-242.11.1 (leu1 gld1-1 Mgm⁺ Mog⁺ Gam⁺), was crossed with M6 (met6 GLD1 Mgm⁻ Mog⁻ Gam⁻), and diploids were selected onto minimal medium plates. All resulting diploids showed wild-type glucose growth (Glu⁺), as well as good glucose repression of the C₁ metabolism (Mog⁺). They were sporulated, and Glu⁺ Mgm⁺ Mog⁺ Gam⁺ segregants were isolated, indicating that the Glu⁺ trait could be segregated independently of the Mog⁺ phenotype. One of these segregants was backcrossed to L1 three times to obtain segregant LGM-128 (leu1 Glu⁺ Mgm⁺ Mog⁺).

LGM-128 was crossed with M6 (met6 Mgm⁻ Mog⁻ Gam⁻), and diploids were selected onto minimal medium plates. They were sporulated, and random analyses of ca. 600 spore populations revealed a 1:1 ratio of Mgm⁺ and Mgm⁻ phenotypes from each population. Thirty selected spore products were examined further. They revealed that in all cases the Mgm⁺, Mog⁺, and Gam⁺ phenotypes cosegregated. Taken together, these data indicated that the phenotype of LGM-128 was due to a single recessive mutation inherited in a Mendelian fashion. Since this mutation was able to complement the glr1 mutation (Gam⁺ Mgm⁺) previously described (34), we named it glr2. The analysis and genetic characterization of more mutants obtained from our large-scale screen is under way.

Induction of the C₁ metabolism in the presence of glucose.

In order to study the induction kinetics of C₁ metabolism in mutant LGM-128, we set up various flask cultures using glucose, methanol, or combinations of these two carbon sources. In preliminary experiments, we had determined that cultivating LGM-128 in ethanol-containing media would repress C₁ metabolism, as shown by the absence of three marker activities—AO, CAT, and FDH. Therefore, LGM-128 was precultivated in YPE and then transferred to the experimental media. The results of these experiments are shown in Fig. 2 to 4 and show that in all media the rates of growth and of methanol utilization of LGM-128 are lower than in the control.

FIG. 2.

H. polymorpha L1 (control) and mutant LGM-128 in Mm (with 1% methanol). (Left) Graphs showing growth (A), methanol concentration (percent) (B), AO activity (in units/milligram of protein) (C), CAT activity (ΔA₂₄₀/milligram of protein) (D), and FDH activity (in units/milligram of protein) (E) of strains L1 (○) and LGM-128 (□). (Right) TEM survey of cells at different culture stages. M, mitochondria; N, nucleus; P, peroxisome; V, vacuole. Bars = 0.5 μm.

FIG. 4.

Hansenula polymorpha L1 (control) and mutant LGM-128 in MDm (with 2% glucose and 1% methanol). (Left) Graphs showing growth (A), glucose and methanol concentrations (percent) (B), AO activity (in units/milligram of protein) (C), CAT activity (ΔA₂₄₀/milligram of protein) (D), and FDH activity (in units/milligram of protein) (E) of strains L1 (○) and LGM-128 (□). •, methanol concentration in L1 cultures; ▪, methanol concentration in LGM-128 cultures. (Right) TEM survey of cells at different culture stages. P, peroxisome; V, vacuole; N, nucleus. Bars = 0.5 μm.

In M medium with 1% methanol (Mm), both LGM-128 and the control show progressive inductions of AO, CAT, and FDH activities, as typically found during adaptation to methanol substrates (Fig. 2). Although the maximum AO and FDH activities are comparable to those of the control, maximum CAT activity in the mutant is twofold higher than in the control (Fig. 2D). All maxima are found in the exponential phases (mid-exponential for the control; late exponential for LGM-128).

As expected, in M medium with 2% glucose as the sole carbon source (MD), the control strain shows that C₁ metabolism is repressed. In contrast, in the same medium, LGM-128 shows clear inductions of AO, CAT, and FDH activities, indicating that C₁ metabolism is active (Fig. 3). These marker activities are of the same magnitude as during methanol growth. When these activities are maximal (ca. 50 h), the residual glucose in the culture is still high (ca. 1.8%), thus excluding the possibility that the observed values are due to derepression triggered by low glucose, as sometimes observed in strains of this species.

FIG. 3.

H. polymorpha L1 (control) and mutant LGM-128 in MD (with 2% glucose). (Left) Graphs showing growth (A), glucose concentration (percent) (B), AO activity (in units/milligram of protein) (C), CAT activity (ΔA₂₄₀/milligram of protein) (D), and FDH activity (in units/milligram of protein) (E) of strains L1 (○) and LGM-128 (□). (Right) TEM Survey of cells at different culture stages. M, mitochondria; N, nucleus; P, peroxisome; V, vacuole. Bars = 0.5 μm.

As expected, in M medium with 2% glucose plus 1% methanol (MDm), the control strain shows a sequential utilization of the two carbon sources, methanol being utilized only when glucose concentrations in the medium are low (Fig. 4). In contrast, after a lag phase of variable duration, LGM-128 shows a phase of methanol utilization, indicated by a significant appearance of AO, CAT, and FDH activities, and by the concomitant disappearance of methanol from the medium. During this phase, glucose is not utilized, as indicated by its constant concentration in the medium. Only when methanol concentration is halved, does glucose utilization start. During this phase, glucose and methanol are utilized at the same time, and the AO activity is maximal (residual glucose, 1.9% [Fig. 4]). The values observed for two of the marker activities under study, namely, AO and FDH, are higher than in glucose media. The third marker activity (CAT) is lower than in glucose media. When methanol disappears, the induction levels of AO, CAT, and FDH decline.

Peroxisome induction in glucose-containing media.

Since the induction of C₁ metabolism is associated with peroxisome proliferation in H. polymorpha, we analyzed the ultrastructure of mutant LGM-128 to investigate whether peroxisomes were present in glucose-containing media. TEM of ultrathin cell sections reveals that, as expected, in L1 (control) grown in ethanol (not shown), glucose (Fig. 3), or glucose plus methanol (Fig. 4), peroxisome proliferation was absent. Also, when LGM-128 is cultivated in ethanol media, peroxisome proliferation is absent (not shown). When transferred to medium containing methanol without glucose, both strains manifest typical peroxisome proliferation (Fig. 2). However, mutant LGM-128 also shows abundant peroxisome proliferation in media containing glucose or glucose plus methanol (Fig. 3 and 4). These peroxisomes are similar to those observed in the control. Their number and morphology vary with the growth medium and the culture stage.

Regulation of MOX mRNA levels by carbon source.

In order to examine the regulation of MOX mRNA levels by carbon source, we grew strains LGM-128 and L1 as described above for the induction experiments and harvested the cultures at time zero and at different culture stages. At time zero, i.e., after precultivation in ethanol medium, MOX mRNA was undetectable (Fig. 5A, lanes 1 and 2). Growth in methanol-containing medium resulted in induction of MOX transcript, both in the mutant and in the control. In the early exponential growth phase, the levels of transcription are low (Fig. 5A, lanes 7 and 8); strong MOX mRNA signals are seen in the late exponential phase (Fig. 5B, lanes 5 and 6). These data confirm that in LGM-128, MOX is activated in response to methanol and repressed in response to ethanol in the same manner as in the control strain. The MOX mRNA level of LGM-128 in methanol medium was assessed by densitometry to be about twofold lower than that of the control strain. Following growth in medium containing glucose and methanol, detectable levels of MOX mRNA were present only in mutant LGM-128, presumably resulting from partial alteration of the normal glucose repression of MOX (Fig. 5B, lanes 3 and 4). The MOX mRNA level of LGM-128 in medium containing glucose and methanol was only slightly lower than that in methanol-containing medium; MOX mRNA was undetectable during the early stages of culture (Fig. 5A, lane 6). These observations parallel the activation of C₁ metabolism in cultures grown in methanol plus glucose previously found and suggest that the glr2-1 mutation eliminates a great part of glucose repression of MOX transcription. We also expected to detect MOX mRNA following cultivation of LGM-128 in glucose medium, as the enzymatic and electron microscopial experiments had also revealed the activation of C₁ metabolism under these conditions. We found that these activations were very slow, so that MOX mRNA was detected only upon prolonged exponential growth in MD medium, such as that obtained by repeated transfers of exponentially growing cells to fresh glucose medium (Fig. 5C, lane 1). The reason behind this difference in the temporal expression of MOX mRNA and AO activity when LGM-128 is growing in glucose remains obscure.

FIG. 5.

Northern blots showing an analysis of the regulation of MOX transcription by carbon source in H. polymorpha L1 (control) and mutant LGM-128. (A) Lane 1, L1 preculture (YPE, 12 h); lane 2, LGM-128 preculture (YPE, 12 h); lane 3, L1 in 2% glucose grown to exponential phase (10 h); lane 4, LGM-128 in 2% glucose grown to exponential phase (24 h); lane 5, L1 in 2% glucose plus 1% methanol grown to exponential phase (10 h); lane 6, LGM-128 in 2% glucose plus 1% methanol grown to exponential phase (24 h); lane 7, L1 in 1% methanol grown to exponential phase (80 h); lane 8, LGM-128 in 1% methanol grown to exponential phase (100 h). (B) Lane 1, LGM-128 in 2% glucose grown to late exponential phase (40 h); lane 2, L1 in 2% glucose grown to late exponential phase (18 h); lane 3, LGM-128 in 2% glucose plus 1% methanol grown to late exponential phase (40 h); lane 4, L1 in 2% glucose plus 1% methanol grown to late exponential phase (18 h); lane 5, LGM-128 in 1% methanol grown to late exponential phase (250 h); lane 6, L1 in 1% methanol grown to late exponential phase (200 h). (C) Lane 1, LGM-128 in 2% glucose grown to late exponential phase (40 h) after precultivation (10 h) in the same medium; lane 2, L1 in 2% glucose grown to late exponential phase (18 h) after precultivation (10 h) in the same medium.

Peroxisome degradation in LGM-128.

When methanol-grown H. polymorpha is transferred to glucose-containing media, the peroxisomes become unnecessary. Inactivation-degradation processes are induced, and the peroxisomes are rapidly destroyed. Although the molecular nature of these processes is still largely obscure, it has been shown that a selective (i.e., targeted), autophagic mechanism occurs, in which most peroxisomes are individually encapsulated in multiple membranous layers and then fused to the vacuole (reference 46 and references therein).

Could the glr2-1 mutation determine alterations of these degradative processes triggered by glucose? We set up time course experiments in which LGM-128 and L1 cells were pregrown in methanol-containing media and then shifted to 2% glucose-containing media. At fixed intervals, cell samples were collected, AO, CAT, and FDH activities were determined, and TEM sections were prepared. Since any reduction of activity is the result of two causes, an active degradation-inactivation process and a dilution effect due to cell growth, we checked the growth rates after the shift and found that the control strain and mutant LGM-128 are comparable in the interval under study. As shown in Fig. 6, the three marker activities under observation undergo strong reductions both in the control strain and in LGM-128, with kinetics that are comparable in the two strains. After 6 h, AO activity is reduced by ca. 90%, CAT activity is reduced by ca. 80%, and FDH activity is reduced by ca. 60%.

FIG. 6.

Decreases in enzymatic activities after H. polymorpha L1 (control) and LGM-128 grown in medium containing 1% methanol were transferred to medium containing 2% glucose. (Top) Graphs showing AO activity (in units/milligram of protein) (A), CAT activity (ΔA₂₄₀/milligram of protein) (B), and FDH activity (in units/milligram of protein) (C) of strains L1 (○) and LGM-128) (□). (Bottom) TEM survey of cells of H. polymorpha L1 (control) and LGM-128 shortly after transfer from 1% methanol media to 2% glucose media. N, nucleus; P, peroxisome; V, vacuole. Bars = 0.5 μm.

After the disappearance of these activities, however, the control strain and the mutant behaved differently. Whereas in the former the three activities remained absent, in the latter they reappeared with time course kinetics similar to those observed in glucose-containing media after shift from ethanol precultures. These data were mirrored by the electron microscopical observations of the peroxisomes (Fig. 6). Indeed, at time zero, both L1 and LGM-128 are characterized by large peroxisomes which, following the shift, undergo degradation. Probabily, L1 peroxisomes are more often degraded by taking up vacuolar vesicles (as typically observed during initial stages of adaptation [Fig. 6]), while LGM-128 peroxisomes usually migrate inside a large autophagic vacuole (a process typically found during late stages of adaptation [46] [Fig. 6]). In L1 cells, after 3 h, the peroxisomes are surrounded by many vacuolar vesicles, probably originating from fragmentation of the vacuole. The vesicles are strictly associated with the peroxisomes which are now surrounded by multiple membranes. At this stage, vacuole-peroxisome fusions can be observed, and at 5 h, peroxisome degradation has proceeded significantly (Fig. 6). Cells are now characterized by a single, large vacuole with electron-dense material in it, which is likely to be what remains of peroxisomes. At 7 h, no cells of the L1 population show peroxisomes (not shown). In the LGM-128 population, the pattern is more complex. Here, intact peroxisomes together with peroxisomes at all stages of degradation seem to coexist during the hours following the shift. After 1 h, peroxisomes surrounded by multiple membranes can be found. These peroxisomes are often surrounded by the vacuole. In the following hours many peroxisomes are found inside the vacuole; these peroxisomes show different stages of degradation. During these stages, small, spherical peroxisomes are often observed in the cytoplasm (Fig. 6). We believe that these peroxisomes are newly formed organelles, resulting from glr2-1 derepression of the C₁ metabolism in glucose media. At 11 and 13 h, in LGM-128, these newly formed peroxisomes are larger and greater in number. Sometimes the vacuole is still large and harbors electron-dense formations; sometimes the vacuoles are small, as typically found in populations with no active degradation processes. After 23 h, LGM-128 shows one or a few large peroxisomes, and no visible degradation processes, as observed during growth of this mutant in glucose (not shown).

Glucose phosphorylation in mutant LGM-128.

Since sugar kinase mutants may determine alterations in some glucose repression processes in S. cerevisiae (14, 18, 28, 30), we tested whether the glr2-1 lesion resulted in altered glucose phosphorylation activities. Our results indicate that the control strain and the mutant have lower activities in methanol-containing media and higher activities in glucose-containing media, as expected from the different glycolytic flows that characterize these two metabolisms (Table 1). However, in glucose media, LGM-128 shows activities that are ca. 40% lower than those of the control strain. Whether this fact is the direct effect of glr2-1 or whether it is a secondary effect of this mutation (e.g., consequence of metabolic flux alterations or of a general metabolic depression) is difficult to say. During growth in glucose plus methanol, the glucose phosphorylation activities of LGM-128 are only slightly higher than those observed in methanol medium, while those of the control are similar to those recorded in glucose. These observations are consistent with the fact that, in the interval under observation, L1 is utilizing glucose while LGM-128 is utilizing methanol.

TABLE 1.

Glucose phosphorylation activities in H. polymorpha mutant LGM-128 and control strain L1 during cultivation in different carbon sources

Carbon source	Strain	Glucose phosphorylation activity (U/mg of protein)
		12 h	24 h	36 h	48 h
2% Glucose	L1	0.297	0.266	0.259	0.265
	LGM-128	0.169	0.178	0.166	0.156
1% Methanol	L1	0.031	0.022	0.029	0.029
	LGM-128	0.032	0.025	0.027	0.028
2% Glucose + 1% methanol	L1	0.307	0.296	0.289	0.285
	LGM-128	0.075	0.076	0.068	0.095

DISCUSSION

Glucose repression is a process common to many heterotrophic microorganisms. In glucose-grown cells, the expression of a large number of genes, often involved in the utilization of other carbon sources, is repressed. In S. cerevisiae, glucose repression is a global regulatory system controlling various metabolic processes, including carbon source utilization, gluconeogenesis, and mitochondrial biogenesis. Genetic analysis has led to the identification of a variety of genes required either for glucose repression or for derepression (release from glucose repression). Whereas repression genes act negatively on the expression of glucose-repressible genes, the derepression genes have a positive action on the same genes (reviewed in references 14, 27, and 40). In S. cerevisiae, peroxisome proliferation is induced by oleic acid, a carbon source alternative to glucose. The transcription of the FOX genes, encoding the peroxisomal enzymes of the oleate utilization pathway, is repressed by glucose and induced by oleate (see reference 35 and references therein). A region required for glucose repression of FOX1 transcription has been identified by Wang et al. (47). The well-characterized SSN6 (CYC8) gene of S. cerevisiae is required for glucose repression of FOX3 (12), whereas SNF1, SNF4, and ADR1 have been found to be involved in derepressing FOX2, FOX3, and the peroxine-encoding gene PAS1 (32, 33).

In this report we describe LGM-128, a pleiotropic mutant of H. polymorpha (a yeast with academic and commercial interest [references 8 and 44 and references therein]) with profound alterations in the control of peroxisome proliferation (reference 44 and references therein) and of C₁ metabolism (reference 8 and references therein). In this mutant, peroxisomes are formed and proliferate upon cultivation in glucose-containing media, whether or not methanol is present. We would like to draw attention to the fact that simple cultivation on glucose media brings about the induction described above without methanol precultivation; in fact, precultivations were done in ethanol media—a condition for full repression of C₁ metabolism also in the mutant. As shown by Northern analyses, after inoculating the glucose media, the transcription of MOX is activated; this fact parallels the induction of C₁ metabolism and of the peroxisomes. These observations rule out the possibility that peroxisomes originated in the preculture are maintained in the mutant by virtue of altered inactivation or degradation processes, as was apparent for other H. polymorpha mutants (26). Additionally, as discussed below, precultivating our mutant in methanol and then inoculating it in glucose media resulted in active degradation processes leading to the disappearance of C₁ metabolism. However, these processes were rapidly followed by new induction phenomena and by the activation, under glucose conditions, of C₁ metabolism.

To our knowledge, LGM-128 is the first glucose repression mutant of H. polymorpha which has been shown to be due to a single recessive gene. It is clear that GLR2 is part of a negative regulatory response to glucose of the expression of C₁ metabolism and related peroxisomes, since in our glr2 mutant the syntheses of three C₁ marker enzymes and the proliferation of peroxisomes appear to be highly active. It has been suggested that C₁ metabolism and the peroxisomes undergo two distinct controlling processes: methanol induction and glucose repression-derepression. The evidence supporting this view consists of the fact that, although derepression of peroxisomal matrix enzymes and of peroxisome biogenesis can be observed in the absence of methanol and in the presence of low glucose, their synthesis reaches maximum levels only upon addition of methanol (5–7, 10). The finding that our glr2-1 mutant supports C₁ metabolism and elaborates peroxisomes in glucose in the absence of methanol indicates that in the expression of these enzymes, methanol induction plays only a minor role. This discrepancy between our results and the classical view (which attributes significant weight to methanol induction) remains to be explained; strain differences, as well as dissimilar experimental conditions, may account for it, at least in part. The phenotype conferred by the glr2-1 mutation suggests that glucose repression is the main controlling mechanism over the genes involved in C₁ metabolism. This is suggested by the fact that during growth in glucose media without methanol, LGM-128 shows induction levels that are similar to those observed in methanol media. A direct repressing action may be mediated by a repressor protein, such as the putative repressor of MOX (FB2) described by Pereira (22). In addition to this mechanism, others can be active, so that glucose repression also controls the induction process. This control can be achieved in at least two ways. First, control could be achieved by inhibiting the formation of a putative inducer molecule. In the case of methanol induction, this is unlikely to be achieved by impeding the intake of this alcohol, which occurs by membrane diffusion. Whichever the inducer molecule, the inhibition can be achieved by modifying this molecule or by impeding its formation. Second, it has been suggested that the MOX promoter, and possibly other genes encoding proteins involved in C₁ metabolism, require the putative transcriptional activator MBF1 (10). This positive regulator activates MOX transcription only in the absence of glucose and is therefore another likely target for glucose repression (or inhibition). With glucose repression acting directly and indirectly, it is possible to envisage the existence of mutants like LGM-128, showing expression levels of the C₁ pathway and of peroxisome proliferation in glucose media that are nearly maximal.

In S. cerevisiae, glucose phosphorylation and HXK2 (hexokinase PII) take part in glucose repression, and many hxk2 mutants fail to repress various genes controlled by glucose repression (14, 18, 28, 30). Our data suggest that the glr2-1 mutant shows minor reductions of glucose-phosphorylating activity. Whether this fact is the direct effect of glr2-1 or whether it is a secondary effect is difficult to say. We believe that in glucose media this is the effect of a general depression of the metabolism of this mutant, but in glucose-plus-methanol media it could be due to the fact that during the stages of culture considered, this mutant is growing at the expense of methanol, with large peroxisome proliferation.

It is interesting to note that, as with other H. polymorpha regulatory mutations (26), glr2-1 does not affect ethanol repression, suggesting that in this yeast the two repressing circuits are distinct. Whatever the nature of GLR2p, our experiments reveal that this protein does not play a major role in peroxisome degradation triggered by glucose. Indeed, our mutant shows rapid and effective degradation processes. This observation suggests that, although stimulated by glucose, the two processes are controlled by elements which are, at least in part, distinct. Indeed, glucose-triggered peroxisome degradation serves to quickly eliminate unnecessary peroxisomes and therefore needs to respond to the presence of glucose as quickly as possible. Our findings suggest that glucose signals may be detected and processed in different ways, depending on whether glucose comes suddenly after methanol cultivation or not. In our mutant, the signal following glucose-to-methanol shift is still processed correctly, whereas the one triggering the glucose repression of MOX and of peroxisome proliferation is not. In S. cerevisiae, it has been demonstrated that glucose repression can be divided into two temporal stages. The early-response stage, acting within 1 h after the shift causes a rapid decrease in the mRNA level of invertase, even in some mutants known to affect glucose repression. The late-response stage, instead, is achieved only if Hxk2p is functioning and maintains the repression during the following hours (30).

Mutants like LGM-128 are instrumental in the genetic analysis of the regulatory network acting upon C₁ metabolism and peroxisome biogenesis. Experiments are under way to isolate the GLR2 gene by complementation and learn more about its structure and function. This understanding is important not only to improve the commercial applications of H. polymorpha but also to shed further light on the mechanism of glucose repression and peroxisome biogenesis and proliferation.