OUP user menu

The Saccharomyces cerevisiaeYPR184w gene encodes the glycogen debranching enzyme

Marie Ange Teste, Brice Enjalbert, Jean Luc Parrou, Jean M. François
DOI: http://dx.doi.org/10.1111/j.1574-6968.2000.tb09410.x 105-110 First published online: 1 December 2000


The YPR184w gene encodes a 1536-amino acid protein that is 34–39% identical to the mammal, Drosophila melanogaster and Caenorhabditis elegans glycogen debranching enzyme. The N-terminal part of the protein possesses the four conserved sequences of the α-amylase superfamily, while the C-terminal part displays 50% similarity with the C-terminal of other eukaryotic glycogen debranching enzymes. Reliable measurement of α-1,4-glucanotransferase and α-1,6-glucosidase activity of the yeast debranching enzyme was determined in strains overexpressing YPR184w. The α-1,4-glucanotransferase activity of a partially purified preparation of debranching enzyme preferentially transferred maltosyl units than maltotriosyl. Deletion of YPR184w prevents glycogen degradation, whereas overexpression had no effect on the rate of glycogen breakdown. In response to stress and growth conditions, the transcriptional control of YPR184w gene, renamed GDB1 (for Glycogen DeBranching gene), is strictly identical to that of other genes involved in glycogen metabolism.

  • Glycogen
  • Debranching enzyme
  • Functional analysis
  • Stress
  • Saccharomyces cerevisiae

1 Introduction

Glycogen is a large branched polymer of glucose residues, which accumulates in yeast in response to nutrient limitation and to stress such as heat, osmotic and saline stress, and to nutrient starvation [13]. Resumption of growth of non-proliferating yeast cells or addition of uncouplers to resting cells is accompanied by a rapid mobilisation of glycogen [4]. This degradation occurs by the combined action of two enzymes: the glycogen phosphorylase and the glycogen debranching enzyme. The former (EC catalyses the sequential phosphorolysis of α-1,4-linked glucose units until two or three glucose units remain before a α-1,6-branch point [5,6]. The second enzyme eliminates the branch point in a two-step process that includes (i) the transfer of a maltotriosyl (or maltosyl) unit from the branch to an adjacent α-1,4-glucosyl chain by its oligo-1,4→1,4-glucanotransferase activity (EC, and (ii) the subsequent hydrolysis of the residual α-1,6-linked glucose residue by its α-1,6-glucosidase activity (EC A direct debranching enzyme termed isoamylase which can attack glycogen or amylopectin by hydrolysing α-1,6-linkages exists in bacteria but its occurrence in yeast is still controversial [7]. Debranching enzymes have been purified from several organisms including Escherichia coli[8], yeast [5,7] and mammals [911]. They all bear transferase and glucosidase activities on a single polypeptide with a molecular mass of 160–170 kDa. Various systems have been developed based on substrate specificity for measurement of the glucanotransferase, α-glucosidase or both activities of the debranching enzyme [12]. Evidence that the two reactions take place at two distinct catalytic sites was provided by enzymological studies, using inhibitors specifically affecting either the transferase or the glucosidase activity [13]. Same conclusions were given from a partial proteolysis of the purified enzyme with trypsin or chemical reactivity with carbodiimide, which resulted in the loss of the transferase activity without affecting α-glucosidase activity [10,14]. Furthermore, the presence of the four conserved consensus sequences of the α-amylase superfamily in the N-terminal sequence of the mammalian debranching enzyme [15] led the authors to suggest that the transferase activity may encompass the N-terminal part of the protein, leaving the α-glucosidase activity for the C-terminal part [16]. There is still no physical demonstration that the two activities of the debranching enzyme are functionally present in two distinct domains.

In yeast, most of the structural genes encoding enzymes involved in the biosynthesis and biodegradation of glycogen have been cloned and characterised (for a review, see [17]), except the gene encoding the debranching enzyme. The YPR184w gene sequence on the yeast chromosome 16 encodes a protein 39% identical to mammals glycogen debranching enzyme. In human, the deficiency in liver and muscle glycogen debranching enzyme causes the glycogen storage disease type III which results in a hyperaccumulation of branched glycogen [18]. In this report, we showed that the YPR184w open reading frame (ORF) encodes the yeast Glycogen DeBranching enzyme and that this gene, renamed GDB1, is essential for glycogen degradation in yeast.

2 Materials and methods

2.1 Yeast strains, media and culture conditions

Strains KT1112 (Mata his3Δ200 leu2-3, 112 ura3-52), JF624 (Matα leu2-3, 112 ura3-52 his3 trp1Δ1 lys2-801) and its isogenic diploid strain JF626 were used as hosts for transformations. The haploid strain JF784 (Matα leu2-3, 112 ura3-52 his3 trp1Δ1 lys2-801 gph1::LEU2) is deleted for GPH1 encoding the yeast glycogen phosphorylase. A 3800-bp XhoI–XhoI fragment removed from YCpX:(gph1Δ:LEU2) [19] was used to disrupt GPH1. Yeast cells were grown at 30°C on YEP medium (1% yeast extract plus 2% bactopeptone) supplemented with 1% instead of 2% glucose, or on yeast nitrogen base (YNB) medium (0.17% of YNB without amino acids and ammonium sulphate plus 0.5% ammonium sulphate and 1% glucose) supplemented with 100 μg ml−1 of the auxotrophic requirement [20]. The YNB medium also contained 1% (w/v) succinic acid and 0.6% (w/v) NaOH to keep the pH around 5.8. To promote glycogen mobilisation, two types of experiments were carried out. A first experimental procedure was to dilute by 10 times a culture of late exponential yeast cells grown on YNB (optical density at 600 nm (OD600)∼2.0) into a fresh YNB medium and to collect 20-ml samples at different times for glycogen assays. The second experiment took the advantage that the addition of uncouplers to yeast cells stimulates glycogen degradation [4]. To this end, cultures of late exponential yeast cells were harvested by centrifugation and washed once with water. The cells were resuspended at OD600=20 into a nitrogen-free medium (50 mM Na-succinate, 5 mM KCl, 12 mM NaPi, pH 5.5) in the presence of glucose 2% (w/v). After 90 min of incubation at 30°C, carbonyl cyanide m-chlorophenyl hydrazone (CCCP) from a 20-mM stock solution made in methanol was added to the cells (0.2 mM final concentration), and 2-ml samples were collected by centrifugation (2 min at 3000×g) at different times for glycogen assay.

2.2 Plasmids and strains construction

The YPR184w or GDB1 gene was cloned as a 8440-bp StuI–KpnI fragment derived from cosmid 9705 (purchased from ATCC no. 71011) into pBSK to yield pMA2. A 3717-bp BglII–BglII which removed 80% of the GDB1 coding sequence was replaced by a 1.1-kbp BamHI (TRP1) fragment from YDpw [21] to yield pMA3. The 2640-bp NarI–KpnI fragment from this plasmid was integrated into the chromosomal locus of GDB1 in the diploid strain JF626 by classical yeast transformation procedure [22]. The haploid strain JF937 bearing the gdb1::TRP1 allele was obtained after sporulation and tetrads dissection. Construction of GDB1-lacZ fusion gene was made by PCR amplification of a 1014-bp fragment from the GDB1 promoter with the two following primers: GDB1-up 5′-CGCGGATCCCGCAGACCTACAACA and GDB1-dw 5′-ATGCCTGCAGGCAGTAATGATCTATT (the BamHI and PstI sites are underlined). The BamHI–PstI PCR fragment was inserted into the BamHI–PstI site of YIp356 [23] to yield pMA5. Recombinant PCR protocol [24] was applied in order to replace the unique STRE of the GDB1 promoter by a NheI site using four oligomers: GDB1-up and GDB1-dw as external primers, and Nhe15 (GCAAGCTAGCGAAATGGCCAATTCTTCTCG) and Nhe13 (TTGGCCATTTCGCTAGCTTGCTGCTGTTGTGACAAAAAAG) as internal primers (underlined is the NheI site which replaces the STRE). Insertion of the PCR fragment in YIp356 yielded YIpΔSTRE-gdb1. pMA5 and YIpΔSTRE-gdb1 were digested with NcoI and integrated at the ura3-52 locus in strain KT1112 by homologous recombination [22] to yield JF1200 and JF1201. Transformants were verified by Southern blot analysis (data not shown). For experiments of glycogen debranching enzyme overproduction, the PCR-amplified GDB1 coding sequence was cloned in frame to PGK1 promoter in pYGE2 [25]. The ORF was amplified using GDB-surex5 (5′-CGCTGACCCGGGATGAAATAGATCATTACTGCTACGT) and GDB-surex3 (5′-GGCGCCGGTACCTCAGGAATCATCTTCGTAGGCATC) (the restriction sites SmaI and KpnI are underlined). The amplified fragment was cloned into pBSK and a 4.4-kbp KpnI–KpnI fragment was recloned into the unique KpnI site of pYGE2 to yield pMA9.

2.3 Biochemical and analytical procedures

The glycogen debranching enzyme was assayed by four different methods. A quick assay is based on the increase in the iodine staining intensity with limit dextrin [26]. Samples (210 μl) were incubated at 30°C with 0.1% (w/v) of glycogen phosphorylase limit dextrin prepared according to Lee et al. [5] in 50 mM citrate sodium pH 6.0 in a total volume of 0.3 ml. After 1 h of incubation, the reaction was mixed with 5 μl of an iodine staining solution (0.2% I2–2% KI) and 5 μl of HCl 2 M, and the OD was read at 550 nm. The combined activity of the glucanotransferase and α-1,6-glucosidase was measured by the release of glucose from glycogen phosphorylase limit dextrin as described in [5]. The glucanotransferase activity of the debranching enzyme was determined at 30°C in 50 mM Na-citrate pH 6.0 in the presence of 50 mM maltopentaose as the substrate. The different oligosaccharides generated by this activity were separated and quantified by high performance anionic exchange chromatography with pulsed amperometric detection (HPAEC-PAD, Dionex, Sunnyvale, CA, USA) according to [27]. Finally, the amylo-1,6-glucosidase activity was estimated by the incorporation of [U-14C]glucose into glycogen following the procedure of Hers et al. [28], with the exception that the buffer was Na-citrate 50 mM, pH 6.0 and the final concentration of [U-14C]glucose in the assay was 25 mM (specific activity of 2000 dpm nmol−1).

For partial purification of the debranching enzyme, 1 g of yeast cells (wet weight) grown to stationary phase on YNB was resuspended in 4 ml of ice-cold solution containing 50 mM Tris–Cl, pH 7.5, 1 mM EDTA, 1 mM phenylmethyl sulphonate fluoride and 10 μg ml−1 of a protease cocktail inhibitors (complete EDTA free protease inhibitors from Roche Diagnostic). The cells were disrupted with 4 g glass beads (0.4–0.5 mm diameter) by vigorous shaking on a vortex for five 30-s periods at 1-min intervals. The extract was centrifuged for 5 min at 3000×g. The nucleic acids were precipitated by addition to the supernatant of protamine sulphate (purchased from Sigma) at a final concentration of 0.1% (w/v). After 10 min of incubation on ice, the crude extract was centrifuged for 30 min at 20 000×g. The supernatant was filtered through a minisart filter (0.2 μm porosity) and applied onto a Resource 15 Q column (FPLC system from Pharmacia Biotech) equilibrated with 50 mM Tris–Cl, pH 7.5. After washing the column with 5 ml of Tris–Cl solution, a continuous gradient of NaCl (0–500 mM) made in this buffer solution was applied in a FPLC system. Fractions of 1 ml were collected and the iodine staining method was used to identify the presence of debranching enzyme activity. Protein concentrations were determined by the method of Bradford [29] using bovine serum albumin as a standard. Cell samples and assays of glycogen and β-galactosidase were performed as described [2].

3 Results and discussion

3.1 Sequences comparison with debranching enzymes from various organisms

In Saccharomyces cerevisiae databases (see for instance the YPD database), the YPR184w is assigned as an ORF encoding a putative glycogen debranching enzyme, because it encodes an 174 845 kDa protein that exhibits between 34 and 39% identity with Drosophila melanogaster, Caenorhabditis elegans and mammals (Homo sapiens and rabbit) debranching enzymes. This identity is distributed all over the 1536-amino acid protein sequence but two distinct regions with interesting features may be distinguished in the sequence. The N-terminal half of the protein (750 amino acids) possesses the four consensus sequences commonly found in the α/β barrel domain of α-amylase and α-glucanotransferase [15], suggesting that this region may encompass the oligo-1,4→1,4-glucanotransferase activity of the debranching enzyme. The 800 residues at the C-terminus may correspond to a second functional domain, since it shows 50% similarity with the C-terminal part of the debranching enzymes from other eukaryotic species. According to Liu et al. [16], this domain could bear the amylo-1,6-glucosidase activity of the debranching enzyme. However, all enzyme hydrolysing α-1,6-glucosidic linkages also display the four regions that are conserved among the α-amylase superfamily [15], suggesting that the N-terminal of the protein is required for the glucosidase activity. The only pertinent difference at the protein sequence level between specific α-1,4-amylases and amylo-1,6-glucosidases is at the level of the consensus II (Fig. 1) in which two amino acids (K and H) are not conserved in these latter enzymes [30]. A potential function of the C-terminal domain of the debranching enzyme could be in the binding of glycogen as suggested by Chen and colleagues [31]. This suggestion was based on a short sequence identity between two segments at the C-terminal region of the human muscle debranching enzyme and the glycogen-binding region of glycogen phosphorylase. However, these segments were not found in the yeast protein. Preliminary experiments showed the deletion of a 555-bp SphI fragment in the C-terminal part of the YPR184w gene abolished both activities, indicating that this C-terminus is required for both activities of the debranching enzyme. Finally, it is interesting to point out that S. cerevisiae, C. elegans and mammals debranching enzyme bear little similarity to E. coli and starch R-debranching enzyme [32,33], possibly because the structure of the glucose polymers is less ramified in E. coli and in starch than that in yeast and mammals.

Figure 1

Predicted amino acid sequence alignment of the yeast with other debranching enzymes. Conserved sequence in the α-amylase superfamily. Enzymes are numbered from the N-terminal end. Invariable residues are in boldface. AMYpp (α-amylase pig pancreas), CGTbm (cyclodextrin glycosyltransferase of Bacillus macerans), ISAps (isoamylase from Pseudomonas amyloderamosa), GDEhs (glycogen debranching enzyme from human liver), GDErb (glycogen debranching enzyme from rabbit), GDBce (glycogen debranching enzyme from C. elegans), GDBdm (glycogen debranching enzyme from D. melanogaster) and GDBsc (glycogen debranching enzyme from S. cerevisiae).

3.2 Determination of glycogen debranching activity in wild-type and GDB1 overexpressing strains

The debranching enzyme can be assayed by several methods, which illustrates α-glucosidase [27], transferase [6] or both activities [5]. Since wild-type cells contained very low debranching activity (about 0.4 nmol [U-14C]glucose incorporated in glycogen per hour and per mg proteins for the amylo-1,6-glucosidase assay), we developed a partial purification of this enzyme by loading 15 mg proteins from crude extracts on an anionic exchange column followed by elution with a NaCl gradient. The assay by the iodine staining method identified debranching activity in peak fractions eluting at ca. 280 mM NaCl (fractions 43–47). These fractions contained about four times more activity when the same amount of protein of a crude extract from a strain overexpressing GDB1 was loaded on the column, whereas they were absent in the extract from a gdb1 null mutant (Fig. 2). The iodine assay was confirmed by quantitative measurement of glucose released from glycogen phosphorylase dextrin limit (not shown). Moreover, α-1,6-glucosidase and transferase activities were determined on fraction no. 45 from the overproducer strain. The α-glucosidase activity incorporated 10.5 nmol [U-14C]glucosyl units in glycogen per min per ml of fraction. The incubation with 50 mM maltopentaose for 1 h at 30°C resulted in a higher production of maltotriose (74 nmol ml−1 of fraction) and maltoheptaose (63 nmol ml−1) than maltose (25 nmol ml−1) and maltooctaose (25 nmol ml−1), which confirmed that the yeast debranching enzyme preferentially transfers two glucose units on an adjacent linear chain [5,6].

Figure 2

Elution profile of debranching enzyme from a Resource 15 Q column chromatography. Proteins (about 15 mg) prepared from wild-type (○), gdb1 mutant (●) or GDB1 overexpressing mutant (▪) were loaded on a Resource 15 Q column. Activity of glycogen debranching was determined by iodine staining method as described in Section 2.

3.3 Physiological effect of deletion and overexpression of GDB1

As shown in Fig. 3A, the transfer of glycogen-containing cells from stationary phase to a fresh growth medium triggered the degradation of the polysaccharide. As expected, this event was completely abolished in a gph1 mutant, and interestingly, it was also prevented in a gdb1 null mutant. Reintroduction of the wild-type GDB1 in the mutant strain restored normal glycogen degradation (not shown). Glycogen mobilisation can also be triggered by the addition of uncouplers, such as CCCP to yeast cells incubated in a nitrogen-free glucose medium [4]. This CCCP-induced glycogen mobilisation was abolished upon deletion of GPH1 or GDB1 (Fig. 3B). Overexpression of GDB1 gene under the control of a PGK1 promoter had no consequence on the rate of glycogen degradation (Fig. 3B), supporting the notion that the rate-limiting step in this process is determined by glycogen phosphorylase [17]. Taken together, these results demonstrate that the in vivo degradation of glycogen requires the presence of both glycogen phosphorylase and glycogen debranching enzyme. It is possible that the inability of glycogen phosphorylase alone to reduce the glycogen content is due to the fact that the α-1,6-branched points of the glycogen molecule restrict the action of glycogen phosphorylase at the polymer periphery.

Figure 3

Effect of loss of GDB1 function on glycogen degradation upon resuspension of stationary phase yeast in fresh growth medium (A) or in response to addition of CCCP to suspension of yeast (B). Strains used in this experiment were JF624 (WT; ○), JF937 (gdb1Δ::TRP1; ●); JF784 (gph1Δ::LEU2; ▲) and JF624 transformed with pMA9 (GDB1 under the PGK1 promoter, ▪). CCCP was added at a final concentration of 0.2 mM.

3.4 Expression of GDB1 during the diauxic growth on glucose and in response to various stress

Previous works demonstrated a coordinate regulation at the level of transcription of genes involved in glycogen biosynthesis and biodegradation under various nutritional and stress conditions ([3], and reference therein). Using an integrated GDB1-lacZ fusion gene, it is shown that the increase in β-galactosidase activity from this construct during diauxic growth on glucose strictly followed the transcriptional patterns found with the other genes involved in reserve carbohydrates metabolism (Fig. 4 and [3]). As in the case of GSY2[34], the removal of the single STRE element located at −191 bp upstream of the ATG codon of the GDB1 ORF did not prevent its growth-related transcriptional induction although reducing by 4-fold the β-galactosidase activity from GDB1-lacZ construct. In contrast, the 3-fold increase of GDB1-lacZ expression in response to a temperature shift from 25 to 37°C and its 2-fold increase in response to the addition of 0.3 M NaCl were abolished upon elimination of the STRE element (results not shown). These results strengthen the absolute necessity of STRE element for gene activation in response to stress conditions, while the coordinated – STRE independent – transcriptional induction during the diauxic growth on glucose [34,35] is indicative of a singular, not yet identified, sensing mechanism coordinating growth and cell division.

Figure 4

Expression patterns of GDB1-lacZ and ΔSTRE-gdb1-lacZ translational fusion during the diauxic growth on glucose. Strains JF1200 (GDB1-lacZ gene fusion, A) and JF1201 (ΔSTRE-gdb1-lacZ construct, B) were grown on YEP in the presence of 1% glucose. Symbols are: cell growth (●), β-galactosidase (♦).


This work was supported in part by the Research program ‘Génie des Procédés chimiques, physiques et Biotechnologiques PIRGPBIO12 of the CNRS’ and by the EU project BIO-CT95-0132 from the Framework IV project to J.F. We thank Mr. Escalier for his help with the use of the FPLC apparatus and Miss Groussac for her help with the HPAEC apparatus.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
View Abstract