OUP user menu

Formation of volutin granules in Corynebacterium glutamicum

Srinivas Reddy Pallerla , Sandra Knebel , Tino Polen , Peter Klauth , Juliane Hollender , Volker F. Wendisch , Siegfried M. Schoberth
DOI: http://dx.doi.org/10.1016/j.femsle.2004.11.047 133-140 First published online: 1 February 2005

Abstract

Volutin granules are intracellular storages of complexed inorganic polyphosphate (poly P). Histochemical staining procedures differentiate between pathogenic corynebacteria such as Corynebacterum diphtheriae (containing volutin) and non-pathogenic species, such as C. glutamicum. Here we report that strains ATCC13032 and MH20-22B of the non-pathogenic C. glutamicum also formed subcellular entities (18–37% of the total cell volume) that had the typical characteristics of volutin granules: (i) volutin staining, (ii) green UV fluorescence when stained with 4′,6-diamidino-2-phenylindole, (iii) electron-dense and rich in phosphorus when determined with transmission electron microscopy and X-ray microanalysis, and (iv) 31P NMR poly P resonances of isolated granules dissolved in EDTA. MgCl2 addition to the growth medium stimulated granule formation but did not effect expression of genes involved in poly P metabolism. Granular volutin fractions from lysed cells contained polyphosphate glucokinase as detected by SDS–PAGE/MALDI-TOF, indicating that this poly P metabolizing enzyme is present also in intact poly P granules. The results suggest that formation of volutin is a more widespread phenomenon than generally accepted.

Keywords
  • Corynebacterium glutamicum
  • Inorganic polyphosphate
  • Volutin granules
  • Electron dispersive X-ray analysis
  • DAPI fluorescence
  • 31P NMR
  • Magnesium
  • Polyphosphate glucokinase
  • MALDI-TOF
  • DNA microarray

1 Introduction

Due to advances in novel enzymatic analytical methods, the study of inorganic polyphosphate (poly P) in prokaryotic and eukaryotic cells has become a re-emerging research topic in biotechnology, biochemistry and molecular biology [1]. Among the functions of poly P in cellular metabolism is its vital role in stress response and stationary-phase adaptation. Both stress response and stationary-phase adaptation also play a decisive role in the industrial production of amino acids with the aerobic Gram-positive rod, Corynebacterium glutamicum[2]. Yet though corynebacteria, viz., C. diphtheriae and C. xerosis, were among the first organisms to be studied with regard to some key poly P enzyme functions [1], very little is yet known on the poly P metabolism in C. glutamicum: (i) Zhang et al. [3] detected the gene sequence ppk2, coding for a novel poly P kinase, in the genome of this organism; (ii) Ishige et al. reported that the genes involved in the metabolism of pyrophosphate and polyphosphate were not affected by orthophosphate (Pi) starvation of C. glutamicum[4]; (iii) in a recent in vivo 31P NMR study, Lambert et al. [5] demonstrated the rapid accumulation of soluble intracellular poly P when oxygenated cell suspensions of C. glutamicum were spiked with glucose and Pi; this poly P pool of up to 300 mM (expressed in P-units) decreases within minutes under anoxic conditions with concomitant increase of phosphorylated glycolytic intermediates [5]; (iv) transmission electron microscopic (TEM) pictures of C. glutamicum showed intracellular electron-dense bodies that, without further experimental evidence, have been interpreted as granules consisting of poly P [6].

The prominent formation of metachromatic (poly P) granules in C. diphtheriae, as seen with special microscopic staining techniques, is commonly used as a diagnostic tool to distinguish this highly pathogenic strain from other corynebacteria [7]. In the latter, metachromatic granules are poorly, if ever, seen [7]. In continuation of our work [5], the studies reported here were directed towards answering the question: would C. glutamicum also form volutin granules? And if so, could their nature definitely be proven using a combination of several different analytical techniques?

2 Materials and methods

2.1 Bacterial strains, growth conditions, and staining procedures

C. glutamicum strains ATCC 13032 and MH20-22B [8] were grown at 30 °C on CGIII complex medium [9], CGXII minimal medium [9] with 0.03 g l−1 protocatechuic acid [4] (and, for strain MH20-22B, with 0.3 g l−1 leucine), or Loeffler's agar medium [10] as indicated. A non-pathogenic strain of C. diphtheriae was grown on Loeffler agar at 37 °C.

For volutin staining, cells were mounted on slides and subjected to both Neisser and toluidine staining [11].

2.2 Fluorescence microscopy, electron microscopy and electron dispersive X-ray analysis

For microscopic observation, a culture aliquot diluted to 0.1 OD600 or extracted polyphosphate granules were stained with 4′,6-diamidino-2-phenylindole (DAPI, 5 μg ml−1 final concentration) and were retained by filtration on 25 mm diameter 0.2 μm GTBP isopore membrane filters (Millipore, Schwalbach) and washed twice with 4 ml sterile water. The wet filter was placed on a microscope slide into a 30-μl drop of water, a drop of water was applied on the filter surface that was then sealed with a coverslip and a drop of immersion oil. Fluorescence microscopy setup and image analysis were done as described [12]. Staining with DAPI renders DNA blue and poly P green when cells or isolated poly P granules were irradiated with UV light of 330–380 nm [13].

C. glutamicum MH20-22B was grown in liquid CGXII medium and in CGXII plus 10 mM MgCl2 at 30 °C for 70.5 h. Two millilitres culture aliquots each were suspended in 2 ml cacodylate buffer [14], centrifuged, and the pellets mixed with liquid agarose (3%, w/w) and cut into 1 mm3 blocks after solidification. For transmission electron microscopy (TEM 400, Philips, Eindhoven), the agarose blocks were postfixed with 1% (w/v) OsO4, and contrasted with uranyl acetate and lead citrate as described [15]. For energy dispersive X-ray analysis (EDX analysis system, EDAX, Mahwah, NJ), samples were neither postfixed nor contrasted.

2.3 Extraction of poly P granules and 31P NMR analysis

Cells of C. glutamicum ATCC 13032 grown in CGXII minimal media with 100 mM MgCl2 were harvested after 24 h by centrifugation for 5 min at 4 °C and 3500g. Washing of the pellets, cell breakage by French-press, and isolation of the granule fraction was done as described [16]. The resulting white material was resuspended in bi-distilled water, followed by adding 5 μg ml−1 of DAPI and observed under the fluorescence microscope as described above.

The polyphosphate granules were dissolved in 0.1 ml of 1 M EDTA (pH 8), 0.2 ml of D2O and the final volume adjusted to 1 ml by using fresh bi-distilled water. Standards of Pi and poly P (“P68” with polymerization from 10 to 40, BK Giulini Chemie, Ladenburg) were prepared having final concentrations of 10 mM, in terms of Pi. These standards were prepared in a final volume of 3 ml fresh bi-distilled water containing 0.3 ml of 1 M EDTA (pH 8), and 0.6 ml D2O. The samples were analyzed with 31P NMR for poly P as described [5].

2.4 Analysis of the proteins in polyphosphate granules by SDS–PAGE and MALDI-TOF-MS

About 200 μl of polyphosphate granular preparation was mixed with 4X NuPAGE LDS buffer (Invitrogen, Karlsruhe). Then SDS–PAGE was done as described [9]. Proteins were identified on gel by comparing the molecular weights of marker protein (see blue prestained standard). Spots of interest were excised, digested with trypsin and the peptide masses were determined by MALDI-TOF-MS [17]. Peptide mass lists were used to search a tryptic digest database of 3746 predicted C. glutamicum proteins provided by Degussa (Halle).

2.5 Global gene expression analysis

Generation of C. glutamicum DNA microarrays, total RNA preparation, cDNA synthesis, DNA microarray hybridization, and gene expression analysis were performed as described elsewhere [4,9,18]. Strain ATCC 13032 was used for this experiment.

3 Results and discussion

C. glutamicum is the most widely used microorganism for the biotechnological production of amino acids [2]. In industrial fermentations increases in osmotic pressure due to product accumulation are detrimental to bacterial growth and formation of product [19]. While strategies employed by C. glutamicum to overcome this situation have been studied (see [19] and references cited therein), a possible role of poly P in the stress responses of this organism has been neglected. An essential place for poly P in the regulation of responses to nutritional deficiencies, environmental stresses and survival in the stationary phase has been established in Escherichia coli, [20]. E. coli mutants that fail to produce poly P are more susceptible to osmotic stress [20].

Though some corynebacteria were among the first organisms to be studied with regard to poly P and its enzymes, knowledge on the occurrence and role of poly P in C. glutamicum is scarce (see Section 1). In consequence, we have previously established that this organism is able to accumulate up to 300 mM of poly P (expressed in P-units) under aeration, with concomitant decrease of poly P within minutes under anoxic conditions [5]. In the present paper, we show that C. glutamicum is able to form poly P granules, a capability hitherto ascribed chiefly to pathogenic corynebacteria species of medical importance [7]. To this end, we applied staining, fluorescence, and TEM/EDX techniques:

3.1 Metachromatic bodies

Subcellular poly P rich granules have been termed “metachromatic” bodies due to a special color reaction of dried cells stained with basic dyes such as toluidine blue or Neisser stains [10,11]. To obtain good visibility of stained granules, cells are usually grown on special media, such as Loeffler medium.

Thus, strain C. glutamicum MH20-22B was grown not only on CGIII agar medium and CGIII plus 100 mM MgCl2, but also on Loeffler's medium as indicated. Samples for both Neisser and toludine staining were taken after 48 h of growth. A non-pathogenic strain of C. diphtheriae served as control. This organism was grown on Loeffler agar at 37 °C. Neisser staining gave more consistent results than toluidine staining. C. glutamicum MH20-22B showed Neisser stained volutin granula on all three media, though granule formation was most pronounced on CGIII with 100 mM MgCl2. Fig. 1(B) shows a typical picture of volutin globules formed under these conditions. Globules were comparable in number and size to those of C. diphtheriae (Fig. 1(A)).

1

Microscopic evidence for poly P granules in C. glutamicum. Formation of metachromatic granules in C. diphtheriae (A) and in C. glutamicum MH20-22B for comparison. C. glutamicum was grown on CGIII agar plus 100 mM MgCl2 at 30 °C for 48 h, C. diphtheriae on Löffler medium agar at 37 °C for 48 h. Cells were mounted on slides and stained according to Neisser (see Section 2). (C, D) Influence of growth conditions on DAPI fluorescence of granula in C. glutamicum MH20-22B. Cells were grown in CGIII medium (C), and in CGIII with 100 mM MgCl2 (D). (E) DAPI fluorescence of C. glutamicum ATCC13032 after 35 h growth in CGXII minimal medium supplemented with 100 mM MgCl2. (F) Typical DAPI fluorescent aggregate in purified granular fraction of cells depicted in (E). For details see text. Bars represent 5 μm each.

However, there have been reports that compounds such as poly-β-hydroxybutyric acid or heteroglycans may also stain heterochromatically [21,22]. Therefore we further investigated the nature of the metachromatic granules in C. glutamicum.

3.2 Intracellular electron dense and phosphorus-rich granules in C. glutamicum

Fig. 2 shows the TEM ultrastructure of C. glutamicum MH20-22B cells grown on CGIII-medium with 10 mM MgCl2. The large intracellular electron-dense granules (Fig. 2, arrows (a)) resemble those in C. glutamicum MG [6] or in Micrococcus lysodeikticus[23]. They were tentatively assigned to volutin granules. Occasionally, equally large yet less electron-dense globules could be observed (Fig. 2, arrows (b)). They resembled the granular centers in M. lysodekticus[24].

2

TEM ultrastructure of C. glutamicum MH20-22B. Cells harvested after 70 h of growth in CGIII medium with 10 mM MgCl2. Samples were fixed with OsO4, but not contrasted. Arrows mark electron-dense globules (a) and less electron-dense “spongy” large globules (b), respectively. Bar represents 0.5 μm.

It has been reported that Mg is present in poly P granules [22,23]. Therefore, the elemental composition of large granules and cytoplasmic areas in C. glutamicum strain MH20-22B was analyzed by electron dispersive X-ray analysis (EDX). Several samples and cells were analyzed. Fig. 3 shows a representative spectrum. The large electron-dense bodies accumulated phosphorus 10-fold in comparison to the cytoplasm. Magnesium, and to a lesser extent aluminum, appeared to be slightly enriched in the dark large globules, while chloride (and silicium) were equally distributed between granules and background (Fig. 3).

3

Electron dispersive X-ray analysis of C. glutamicum MH20-22B. Unstained and unfixed cells taken from CGIII-medium were examined for the presence of electron-dense granules. The elemental compositions of a large granule (inset, TEM, g) and a cytoplasmic area (inset, c) were analyzed by electron dispersive X-ray analysis. Bar represents 0.5 μm.

It has been reported that Mg ions stimulated the formation of volutin granules in Chlorella cells [22]. Therefore, we studied the effect of Mg on granule formation in C. glutamicum.

3.3 MgCl2 stimulates formation of volutin granules in C. glutamicum

We used fluorescence microscopy to find out whether (i) the C. glutamicum granules showed the typical green fluorescence of intracellular DAPI-poly P complexes (in contrast, DAPI-DNA fluoresces blue), and (ii) whether Mg in the growth medium would intensify the green (poly P) fluorescence in cells cells. Fig. 1(D) shows that addition of 100 mM MgCl2 to the growth medium increased the green fluorescence in cells of C. glutamicum MH20-22B considerably (compare to Fig. 1(C)). A comparable effect of MgCl2 was observed with C. glutamicum ATCC 13032 (not shown). Thus, Mg ions stimulated the poly P formation in C. glutamicum.

Besides poly P, protein is a major component of volutin granules in other microorganisms [23]. The pyrophosphate and poly P rich acidocalcisomes detected in Agrobacterium tumefaciens and Rhodospirillum rubrum[25] also contain protein. To obtain further evidence that the electron-dense and P-rich granules in C. glutamicum contained both poly P and protein, we isolated and analyzed these granules.

3.4 31P NMR, SDS–PAGE and MALDI-TOF of granules extracted from cells

Granular fractions from disrupted cells were obtained as described in Section 2. Fig. 1((E) and (F)) shows cells with intracellular granules before Fig. 1(E), and the granular fraction after extraction Fig. 1(F). Both micrographs were taken with DAPI as fluorescence marker. Volutin granules (green) in intact cells appeared predominantly at the cell poles, while the remainder showed the typical blue fluorescence of DAPI–DNA complexes. The volume occupied by the volutin granules was estimated to 18–37% of the total cell volume, taking dimensions of granules and cells from Fig. 1(E).

Granules isolated from these cells (see Section 2), aggregated to larger agglomerizations, as seen in Fig. 1(F). They resembled in size and appearance those aggregates shown in TEM pictures of volutin granular fractions from cells of M. lysodeikticus[23,24].

NMR spectroscopy was used to corroborate that poly P was a major constituent in volutin granules. They were dissolved in EDTA buffer ([23], and Section 2) and subjected to 31P NMR. Fig. 4(C) shows a representative spectrum of these preparations. The resonance at about −20 ppm is typical for long-chain poly P (core phosphates of poly P; see [5] for details). The same signal was also observed in the poly P standard (Fig. 4(B)). A comparison with an orthophosphate standard (Fig. 4(A)) shows that some orthophosphate was also present in the granule preparation (Fig. 4(C)).

4

31P NMR spectrum of dissolved poly P granules extracted from C. glutamicum ATCC 13032. (A) Orthophosphate; (B) spectrum of synthetic poly P (P68, see Section 2); (C) dissolved granular fraction.

Granular fractions obtained from M. lysodeikticus contained protein [23]. To see whether proteins were also present in the C. glutamicum granules, an SDS–PAGE analysis of the particulate fraction was done. Fig. 5 shows that several protein bands ranging from 19 kDa to about 150 kDa were observed. The arrow in Fig. 5 marks the most prominent band. To identify this protein, small spots were excised and the tryptic digests subjected to MALDI-TOF analysis. Peptide mass lists were used to search a tryptic digest database of 3746 predicted C. glutamicum proteins (see Section 2), and the band at ∼29 kDa was assigned to a putative poly P glucokinase (PPGK, EC 2.7.1.63, encoded by NCgl1835). Thus, our findings reveal for the first time a poly P metabolizing enzyme, PPGK, in poly P granules.

5

SDS–PAGE of proteins in volutin granules extracted from C. glutamicum ATCC13032. S, sample; M, molecular mass marker. Arrow marks band at ∼29 kDa from which samples for MALDI-TOF analysis were taken (see text).

Docampo and coworkers find pyrophosphatase (PPase) in acidocalcisomes of A. tumefaciens and R. rubrum[25]. We cannot rule out that one of the other proteins in the C. glutamicum granules (Fig. 5) represents PPase which is present in this organism [26]. However, this enzyme in C. glutamicum is soluble, while the acidocalcisomal PPases in A. tumefaciens and R. rubrum are membrane-bound H-types. Yet there remains a striking coincidence: (i) C. glutamicum cells did form polar granules (Fig. 1(e), 2), and a poly P metabolizing enzyme, viz., PPGK, was located in the granules (Fig. 5 and Section 3); (ii) the poly P rich acidocalcisomes in A. tumefaciens also show in the poles of the cells and contain a PPase involved in P metabolism [13]; (iii) Brevibacterium lactofermentum ATCC 13869 (a close relative if not identical with C. glutamicum) also accumulates PPase at the cell poles ([26]; granule formation has not been reported yet).

In M. lysodeikticus, membrane-like structures trap poly P particles [23]. Acidocalcisomes in A. tumefaciens and R. rubrum are also surrounded by a membrane of unknown composition [13,25]. Some of the less electron-dense globular subcellular structures in C. glutamicum (Fig. 2, arrows b) resemble the granular centers in M. lysodekticus[24] or the “spongy” acidocalcisomes in A. tumefaciens and R. rubrum[13,25]. Currently we do not know whether a membrane also surrounds the similar structures in C. glutamicum.

Finally, we tried to answer the following question: it has been shown above that MgCl2 stimulated poly P formation in C. glutamicum strains MH20-22B and ATCC 13032, but did MgCl2 also affect expression of genes involved in poly P metabolism?

3.5 Global gene expression in C. glutamicum under high magnesium conditions

We identified genes differentially expressed in response to growth in the presence of high MgCl2 concentrations as follows: cells of C. glutamicum ATCC 13032 were precultured in CGIII medium for 12 h and then inoculated into CGXII minimal medium (Section 2) with normal, low, MgCl2 concentrations (1 mM MgCl2) or with high MgCl2 (101 mM), to a starting OD600 of 0.5. Cells from 12 h cultures were then transferred into fresh Mg-low/high media, harvested after 7 h of growth (OD600 of 4–5) and used for DNA microarray analyses (see Section 2). Under high-Mg conditions, operons encoding putative uptake systems for Mn, Zn, iron siderophores and alkanesulfonates showed increased expression (RNA ratios higher than 3). Three genes showed decreased expression (RNA ratios lower than 1/3). These genes code for a transcriptional AraC-like regulator, a putative ATPase, and a putative secreted hydrolase (Table 1). The expression of the poly P metabolism genes ppgK, ppk2A, ppk2B, ppx1 and ppx2[3,27] did not change significantly.

View this table:
1

Global gene expression changes in C. glutamicum ATCC 13032 under high magnesium conditions

Gene id.aGene id.bAnnotationExpression ratioc
cg0040NCgl0025Putative secreted protein downstream of cg0040/NCgl002611.3
cg0041NCgl0026Putative zinc ABC uptake system, binding protein11.1
cg0042NCgl0027Putative zinc ABC uptake system, permease2.2
cg0043NCgl0028Putative zinc ABC uptake system, ATPase1.5
cg1120NCgl0943Transcriptional regulator, AraC family0.3
cg1377NCgl1174Putative alkanesulfonates ABC uptake system, permease4.0
cg1379NCgl1175Putative alkanesulfonates ABC uptake system, ATPase2.0
cg1380NCgl1176Putative alkanesulfonates ABC uptake system, binding protein1.6
cg1671NCgl1422Putative membrane-associated GTPase0.1
cg1737NCgl1482Aconitase (EC 4.2.1.3)3.0
cg1832NCgl1564Putative iron siderophores uptake system, permease1.5
cg1833NCgl1565Putative iron siderophores uptake system, binding protein3.5
cg1834NCgl1566Putative iron siderophores uptake system, ATPase1.4
cg1930NCgl1646Putative secreted hydrolase0.2
cg2911NCgl2539Putative manganese ABC uptake system, binding protein14.6
cg2912NCgl2540Putative manganese ABC uptake system, ATPase4.3
cg2913NCgl2541Putative manganese ABC uptake system, permease4.4
cg3107NCgl2709Zinc-dependent alcohol dehydrogenase (EC 1.1.1.1)0.3
  • In addition to ORFs showing RNA level changes of at least 3-fold or not more than 1/3, all ORFs of putative operons are shown.

  • aGenBank Accession No. BX927147.

  • bGenBank Accession No. NC003450.

  • cmRNA levels of cells: grown with 100 mM Mg2+/control.

Thus, an influence of high MgCl2 concentrations on the expression of poly P synthesizing genes was excluded. Stress dependent formation of poly P [1] from high concentrations of Mg or Cl seems unlikely: growth was not inhibited by 100 mM MgCl2, and C. glutamicum tolerates high concentrations of chloride [28]. Therefore, a more direct role of Mg2+ ions in poly P complex formation or on the activity of Mg-dependent poly P enzymes seems likely: Mg2+ is one of the constituents of poly P granules/acidocalcisomes in several bacteria, and in C. glutamicum (as shown in Fig. 3), and stabilizes granule structure during isolation [13,23,25]. Possibly, a similar mechanism was responsible in C. glutamicum for enhanced granule formation during growth in high-Mg medium.

In conclusion, we revealed for the first time a poly P metabolizing enzyme, PPGK, in poly P granules, we also showed that the high concentrations of MgCl2 enhanced granule formation in C. glutamicum. Our results indicate that substantial formation of poly P granules in bacteria may not be restricted to the classical “volutin” organisms, such as C. diphtheriae, Acinetobacter johnsonii, and others. Work of deleting/overexpressing genes that code for key poly P metabolizing enzymes and the effects on the amino acid production in C. glutamicum is in progress. It would also be tempting to investigate whether poly P occurs in high amounts in the related C. ammoniagenes that is an organism for the biotechnological production of the P-containing metabolites XMP and IMP [29]).

Acknowledgements

We would like to thank Hermann Sahm for continuous support, and Tina Zeller, Doris Rittmann and Christina Mack (IBT-1, Forschungszentrum Jülich) and Maike Meindorf (Institut für Hygiene und Umweltmedizin, RWTH Aachen) for skillful technical assistance. We thank Günter Hollweg (Elektronenmikroskopische Einrichtung der Medizinischen Fakultät/Institut für Pathologie der RWTH Aachen) for TEM and EDX analyses and Rudolf Lütticken and Marlies Breuer-Werle (Institut für Medizinische Mikrobiologie der RWTH-Aachen) for a non-pathogenic strain of Corynebacterium diphtheriae and technical instructions. We thank F. Wahl (BK Giulini Chemie, Ladenburg) for a gift of poly P P68, Vijayalakshmi Subramaniam (IBT-1, Forschungszentrum Jülich) for help with light-microscopic photographs, and Jörn Kalinowski (Institut für Genomforschung, Universität Bielefeld), for helpful discussions. We are indebted to Ilan Friedberg (Department of Cell Research and Immunology, Tel Aviv University) for electron microscopic pictures of M. lysodeiktikus.

References

  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].
View Abstract