Eur. J. Biochem. 204,847-856 (1992)

c,FEBS 1992

Multinuclear NMR spectroscopy of the cellular slime mold Polysphondylium pallidum Monitoring of the encystment and excystment processes Gerard KLEIN

’,Jean-Baptiste MARTIN’,

David A. COTTER and Michel SATRE

Departement de Biologie Moleculaire et Structurale/Laboratoire de Biologie Cellulaire, Centre d’Etudes Nucleaires de Grenoble, France DCpartemcnl de Biologie Moltculaire et Structurale/Laboratoirc de Rksondnce Magnetique en Biologie et Medecine, Centre d’Etudes Nucleaires de Grenoble, France Department of Biological Sciences, University of Windsor, Canada (Received September 26/0ctober 29, 1991) - EJB 91 1281

Polysphondylium pallidurn microcysts and amoebae have been investigated by ‘P- and naturalabundance proton-decoupled 3C-NMR spectroscopy. Microcysts have been found to contain as prominent metabolites a phosphomonoester, inositol hexakisphosphate (1.4 mM), two phosphodiesters [glycerophosphocholine (5.5 mM) and glycerophosphoethanolamine (2.6 mM)], as well as nucleoside triphosphates (3 mM) and polyphosphates (> 10 mM), the polyamines 1,3-diaminopropane (3.5 mM), putrescine (16 mM) and spermidine (3 mM) and the sugar trehalose (31 mM). In vivo 3LP-NMRhas shown that the level of nucleoside triphosphates in microcysts was maintained metabolically and that the pH of their cytosol, deduced from the chemical shift of cytosolic Pi was 7.2. The absence of trehalose, glycerophosphocholine and glycerophosphoethanolamine in P. pallidurn amoebae was the most remarkable difference from microcysts. Microcyst germination (excystment), induced by reduction of the ionic strength of the microcyst bathing medium, was monitored noninvasively by 31P-and 13C-NMR spectroscopy. The major modifications observed during excystment were the progressive disappearance of trehalose used as energy source, of glycerophosphocholine and glycerophosphoethanolamine used as membrane phospholipid precursors, and, finally, the appearance of NMR-visible polyphosphates and of cellobiose. As a mirror situation, P. pallidum amoebae responded to a high-ionic-strength stress by production of trehalose, glycerophosphocholine, and glycerophosphoethanolamine, and induction of an encystment process.

’

The cellular slime molds have been used extensively to study the basic processes of cellular differentiation as well as multicellular development (Bonner, 1967; Loomis, 1975; Raper, 1984). The Dictyostelid Polysphondylium pallidum may embark upon one of three alternative pathways; the choice is dependent upon specific environmental variables (Francis, 1979; Raper, 1984). When adequate moisture and light are available, thousands of myxamoebae may aggregate to produce a fruiting body consisting of dead stalk cells and dormant spores. If opposite mating types are present in the cell population, then the resulting aggregate may develop in the dark into thick-walled macrocysts in which zygotes are formed and meiosis takes place (Francis, 1975). The third pathway occurs when starved individual cells fail to aggregate but instead transform into microcysts under conditions of restricted moisture and light (Blascovics and Raper, 1957; Toama and Raper, 1967a; Cotter and Raper, 1968b). Microcyst formation can be induced in starving cells by addition of high concentrations of carbohydrates, KC1 and/or ammonia (Toama Correspondence to G. Klein, Laboratoire de Biologie Cellulaire, DBMS/BC, Centre d’Etudes Nucleaires, 85 X, F-38041 Grenoble Cedcx, France Abbreviations. GroPCho, glycerophosphocholine; GroPEtn, glycerophosphoethanolamine; GroPlnslSer, glycerophosphoinositol/ w i n e ; Imp6,inositol hexakisphosphatc.

and Raper, 1967a; Cotter and Raper, 1968b; Githens and Karnovsky, 1973; Lonski, 1976; Choi and O’Day, 1982). The mature globose microcyst is surrounded by a wall consisting of a dense inner and a somewhat looser outer layer, which is in contrast to the three layered wall of the mature spore (Hohl et al., 1970). The wall contains inclusions of cytoplasmic origin (Hohl et al., 1970) and has a composition (on a mass basis) of 28% cellulose, 35% protein, 21% lipids with the remainder consisting of undefined glucose polymers/ glycogen-like materials (Toama and Raper, 1967b). The protein component lies in the inner layer of the cyst wall while the carbohydrate components are distributed throughout both wall layers. The non-reducing carbohydrate trehalose has been detected in dormant microcysts by Githens and Karnovsky (1973); it may serve as an energy source for excystment (Tisa and Cotter, 1979). Microcysts do not require activation before germination as do spores of P. pallidurn. In fact, microcysts are heat-labile and require no added nutrients for excystment (Cotter and Raper, 1968b). Microcyst germination can be biochemically divided into a two-step process involving a swelling phase and an emergence phase (Cotter and Raper, 1968a; O’Day, 1974; O’Day et al., 1976; O’Day and Paterno, 1979). Upon a reduction in the salt concentration of the environment, the excystment process begins with the rehydration of the microcyst; it first involves enlargement of the polyvesicular bodies and

848

other vacuoles, the disappearance of the inner peripheral lining, and a general loosening of the wall texture within the first hour of germination (Hohl et al., 1970). These processes are independent of protein synthesis (Cotter and Raper, 1968b; O'Day, 1974). At 1.5 h into the germination process there is a striking, temporary dilatation of the cisternae of the endoplasmic reticulum, which is interpreted as the morphological expression of a burst of protein synthesis required for emergence of the myxamoeba (Hohl et al., 1970). At this time an increase in intramembrane particle density is observed, which continues until the level of particles is similar to that of the plasma membrane of the vegetative myxamoeba (Erdos and Hohl, 1980). Instead of splitting of the microcyst case the protoplast appears to dissolve a limited part of the wall through which it then escapes (Blascovics and Raper, 1957; Hohl et al., 1970). While the morphology and macromolecular events of microcyst formation and germination have been investigated, the underlying metabolism leading to the large changes in carbohydrate synthesis and degradation have been ignored. Since the developmental processes of encystment and excystment occur in simple buffer solutions they lend themselves to analysis by the non-destructive techniques of I3C- and 31PNMR spectroscopy. It is the purpose of this investigation therefore to use these techniques to analyze the metabolic shifts in carbon flow contributing to the above differentiation processes.

Excystment was monitored by placing 25 pl of the suspension on a glass slide and observing 200 objects with a Zeiss phase microscope at a magnification of 400. Excystment was considered complete when a myxamoeba was observed which was free of its microcyst case. A population of young dormant microcysts normally germinates within 3-4 h when suspended in dilute buffer at room temperature (Hohl et al., 1970). Induction of amoeba encystment was performed by a method modified from Cotter and Raper (1968b); i.e. myxdmoebae were washed and resuspended in 120 mM KIPi pH 6.5 at 23°C.

MATERIALS AND METHODS

13C-and "P-NMR spectroscopy of HC104 extracts from P.pallidurn amoebae and microcysts

Organism and growth conditions Polysphondyfium pullidurn, strain WS-320 (ATCC 44843) was obtained from the American Type Culture Collection (Rockville, MD). Vegetative myxamoebae were grown in the axenic culture medium described by Watts and Ashworth (1970) containing 0.25 g/l dihydrostreptomycin and 18 g/l maltose as carbon source. Cell cultures between 3 -6 x lo6 cells/ml were harvested by centrifugation. The spent axenic medium was removed and the cells were resuspended in 40 mM potassium phosphate (K/Pi) pH 6.5 to avoid osmotic shock. The cells were centrifuged a second time and resuspended in 20 mM K/Pi pH 6.5. After a third centrifugation, the cells were resuspended in 10 mM K/Pi pH 6.5 (13C-NMR experiments) or 10 mM potassium 4-morpholineethane sulfonate (K/Mes) pH 6.5 ( 31P-NMRexperiments). In a few experiments, amoebae were grown either in suspension in 10 mM K/Pi pH 6.5 in the presence of Escherichia coli Bjr as food source or in a lecithin/ milk axenic medium (Sussman, 1987). Similar results were obtained. To study the excystment process, massive quantities of microcysts were required. These were produced by growing P. pallidum together with E. coli B/r on plates of glucoselsalt agar in the dark (Cotter and Raper, 1968b). After a 7-day incubation at 23 "C, the plates containing dormant micrmysts were flooded with cold distilled HzO; the microcysts were dislodged with a glass spreader and washed by centrifugation.

Excystment and encystment processes

Microcysts were resuspended in 10 mM K/Pi or 10 mM K/Mes buffer at 23 "C to induce excystment (germination).

I n vivo NMR spectroscopy of amoebae and microcysts

Amoebae or microcysts were suspended in 7 ml buffer (see legends to figures for specific details) and aerobiosis was maintained by a steady bubbling of oxygen (20 ml/min) through three glass capillaries. NTP levels and cytosolic pH reached their equilibrium values in less than 5 min of oxygenation. I n vivo proton-decoupled 13C- and "P-NMR spectra of microcysts and amoebae were recorded at 23 C on a Bruker AM400 spectrometer equipped with a 15-mm-diameter ' 3C, 31Pdual probe operating at 100.6 MHz for 13Cand 162 MHz for 31P.Acquisition conditions for 31Pand for 13C were as described previously (Martin et al., 1987; Klein et al., 1990).

HC104 extracts for I3C- and 3'P-NMR were prepared as described (Martin et al., 1987). They were performed on suspensions oxygenated for 15 min at 22°C in 0.1 M K/Pi or K/Mes pH 6.5 for microcysts, 20 mM Na/Pi, K/Pi or Na/Mes pH 6.5 for amoebae, before addition of ice-cold HC104 to a final concentration of 1.3 M. Proton-decoupled 13C- and 31P-NMR spectra of HC104 extracts were recorded on a Bruker WM250 spectrometer equipped with a 10-mm-diameter probe operating at 62.9 MHz for I3C and 101.2 MHz for "P (Satre et al., 1986; Klein et al., 1990). Intracellular metabolite concentrations were determined by a two-step method as follows. Relative amounts of metabolites (I3C-NMR) were measured in fully relaxed (20-s repetition time) proton-decoupled spectra in which the nuclear Overhauser enhancement was avoided by irradiating 'H during the acquisition time (Klein et al., 1990). An internal calibration was then performed by addition of known amounts of authentic compounds to those already present in the extract (trehalose in microcyst extracts, putrescine and 1,3diaminopropane for amoeba1 extracts). The increase in the area of the spiked compound was used to convert relative amounts into absolute amounts. Similarly, relative concentrations of phosphorylated metabolites were derived from the intensity of the respective resonance lines in fully relaxed proton-coupled spectra (20-s repetition time). Inorganic phosphate content of the extract was determined by the method of Fiske and SubbaRow (1925) and was used to derive absolute amounts of all other phosphorylated compounds. The spiking method gave identical results. Intracellular metabolite concentrations were calculated using the mean cellular volumes indicated below.

849

PolyP

(

Grc Cho uNTP GroPEtn

I

r

1

li

1

I

1

1

I

1

1 -25

I

3

l

l

l

l

i

2

l

l

l

,

l

l

l

1

Chemical shifl (ppm)

I

. I . ,

0

-

-22 Chemical shlfi (ppm) -21

Fig. 1. 3'P-NMR spectrum of a HC104 extract from P.pallidurn aerobic microcysts. The HC104 extract was prepared from a total of 1.\ x 10'O microcysts as described under Materials and Methods and adjusted to pH 8.2 before NMR measurements on a WM250 Bruker spectrometer. This proton-decoupled spcctrum (A) was the sum of 7200 free-induction decays with 4-s interpulse delays. Specific portions of the spectrum are shown on expanded scales in (B) and (C). Enlargment (B) is a portion of a spectrum acquired without broad-band proton decoupling, showing the fine structure of GroPCho and GroPEtn and the characteristic splitting of the InsP6 peaks into doublets. Peak assignments were as follows: PME, phosphomonoesters; Pi, inorganic phosphate; Imp6, inositol hexakisphosphate; GroPCho, glycerophosphocholine; GroPEtn, glycerophosphoethanolamine; GroPlns/Ser, glycerophosphoinositol/serine;a, p or yNTP, unresolved a-P, j - P or y-P resonances of nucleoside triphosphates; a or BNDP, unresolved a-P or /?-P resonances of nucleoside diphosphates: UDP sugars, UDP-Glc, UDP-GlcNAc or UDP-Gal; PolyP, ccntral P atoms of polyphosphate chains; ATP/GTP, p-P resonances of purine nucleoside triphosphates; CTP/UTP, 8-P resonances of pyrimidine nucleoside triphosphates. PME were identified as phosphocholine, phosphoethanolamine and, tentatively. as nucleoside monophosphates.

8 50

Cell size was determined on a Coulter counter ZM equipped with a Channelizer C256. Microcysts had a mean cell volume of 91 10 pm3 ( n = 10) and axenically growing amoebae a mean volume of 500 60 pm3 (n = 6).

Table 1. Intracellular concentrations of metabolites in microcysts and amoebae of P. pallidurn. Two independent determinations were performed on extracts from both microcysts and amoebae. n.d., not detected. Very-long-chain polyphosphates are NMR-invisible in vivo and might be partially lost during the HC104 extraction as precipitates. Polyphosphate content represents thus a minimal estimate in microcysts. Free Mg2+ (mean SD, n = 6) was derived from in vivo spectra as described by Gupta et al. (1983).

RESULTS

Compound

Cell size

31P-and I3C-NMR spectra of P. pallidurn microcysts A typical 31P-NMR spectrum of HC104 extracts of aerobic microcysts from P. pallidurn is shown in Fig. 1. Identifications in these extracts were made on the basis of several criteria, i. e. comparison with the compounds previously identified in D.discoideurn amoebae and spores (Satre et al., 1986; Martin et al., 1987; Klein et al., 1988) and direct spiking of the extracts with authentic compounds. Phosphorylated compounds identified in the spectrum of microcyst extracts (Fig. 1A) were, starting from the low-field side of the spectrum, several phosphomonoesters, inositol hexakisphosphate (InsP6), inorganic phosphate (Pi), the phosphodiesters GroPCho, GroPEtn, GroPlnslSer, which accounted for the resonance lines at 0, 0.6 and 1.1 ppm, the diphosphodiesters NAD(H) and NADP(H), UDP sugars, the various resonance lines of nucleoside diphosphates and triphosphates (- 5.6, - 10.8 and -21.3 ppm) and central phosphates of polyphosphates (-21.8 ppm). Intracellular metabolite concentrations are given in Table 1. Interestingly, P. pallidurn microcysts contained very high InsP, levels (1.4 mM). The phospholipid-derived phosphodiesters GroPCho (5.5 mM), GroPEtn (2.6 mM) and GroPlnslSer (0.9 mM) were present in the relative ratio 1/0.5/ 0.35. Nucleoside diphosphates and triphosphates amounted to 0.2 mM and 3.0 mM, respectively, and the purine/pyrimidine ratio was close to 3.0. The average chain length of NMRvisible polyphosphates ( 211 mM) can be estimated from the intensity of the resonances near -5 ppm arising from their two terminal phosphates relative to the resonance of the central P atoms at -21.8 ppm. Terminal residues of polyphosphate chains were barely visible in the 31P-NMR spectrum, an indication of a chain length higher than 50 phosphate atoms. Two phosphomonoesters were identified as phosphocholine (3.6 ppm) and phosphoethanolamine (4.0 ppm), both present at approximately 0.5 mM. Fig. 2 shows an in vivo 31P-NMR spectrum of an oxygenated suspension of P. pullidurn microcysts. Major resonances broadened by the intracellular environment corresponded to those identified as sharp lines in the HC104 extract, except for the polyphosphate resonance at - 22 ppm, which was almost undetectable in the in vivo spectrum. A possible explanation for this discrepancy is that microcysts have long-chain polyphosphate polymers with correlation times such that line widths are broadened beyond detectability under in vivo conditions, possibly by formation of insoluble complexes with divalent cations. Nucleoside diphosphates and triphosphates resonated at -5.5, -10.6 and -19.4 ppm. The fraction of nucleoside triphosphates NTP/(NDP + NTP) was close to 65 _+ 10% (mean _+ SD, n = 3) corresponding to an NTP/NDP ratio of 2, a value in close agreement with previous data found with D . discoideum amoebae and spores (Satre et al., 1986; Klein et al., 1990). Part of the nucleoside triphosphates appeared thus NMR-invisible in vivo, as in the HCIO4 extract the NTP/NDP ratio was about 15 (Table 1). A contribution of terminal polyphosphate chains to the yNTP- PNDP peak

Amount in microcysts

amoebae

mM Trehalose Putrescine Diaminopropane Spermidine InsP, Pi GroPInslSer GroPEtn GroPCho Phosphoethanolamine Phosphocholine NTP NDP Polyphosp hates Free Mg2

21.5 -34 15 -11 3 -4 2 -4 1.2 -1.6 2.6 -2.8 0.8 -1.0 2.5 -2.1 5.4 -5.1 0.5 -0.1 0.2 -0.4

2.5 -3.5 0.15 -0.2 210.5 -13 0.24 0.04

2.5 - 3 12 -16 1.5 -2.5 z 1.5 0.5 -0.6 2.4 -2.6 n.d. 0.2 -0.3 0.4 -0.5 n.d. n.d. 1 -1.5 0.1 -0.15

n.d. 0.32 f0.1

at - 5.5 ppm could lead to an overestimation of NDP content and thus to a reduced in vivo NTP/NDP ratio. Such an hypothesis is not likely as no corresponding central polyphosphate chains were detectable at -22 ppm. The p-P resonance characteristic of NTP at - 19.4 ppm strictly required aerobic conditions to be present. It disappeared upon anaerobiosis and was restored upon subsequent reoxygenation. At the same time, cytosolic pH, estimated from the chemical shift of cytosolic Pi, acidified from pH 7.2 (6 = 1.98 ppm) in aerobic conditions to pH 6.8 (6 = 1.55 ppm), when nitrogen was bubbled into the cell suspension. Another argument for cytosolic acidification upon anaerobiosis could be derived from the resonance line of phosphomonoesters, mainly a mixture of phosphoethanolamine and phosphocholine, which is shifted from 3.8 ppm to 3.2 ppm. Realkalinization of the cytosol accompanied return to aerobic conditions. The chemical shifts of the various resonance lines of nucleoside triphosphates are dependent upon nucleotides complexing with magnesium (Gupta et al., 1983). The data derived from the in vivo 31P-NMR spectra of P. pullidurn microcysts indicated that 84 & 2% (n = 6) of total nucleoside triphosphates were in a complexed form and that the level of free MgZf was 0.24*0.04mM, taking into account a dissociation constant of 50 pM for NTP-Mg. The sharp resonances in the region at -0.05 ppm and 0.5 ppm arose from the presence of the large amounts of GroPCho and GroPEtn, identified in the HC104 extracts (see Fig. 1). The sharpness of the resonance lines was linked to their insensitivity to pH and to rapid tumbling of the free molecules. Carbon metabolites were identified in the 13C-NMRspectrum of an HC104 extract of P. pallidurn microcysts (Fig. 3); signal assignments are given in Table 2. The most abundant metabolites detected in an HC104 extract from P. palli-

851 GroPCho

I InrP6

0 2

NAD(P)

I aNTP aNDP

pNTP

Polyp

A __v1

L 10

I 5

1

I

I

0

-5

-10

I

-15

I

-20

1

-25

Chemtcal Shtft (ppm)

Fig. 2. I n vivo 3'P-NMR spectra of an aerobiosis-anaerobiosis sequence on P. pallidurn microcysts. 12-day-old microcysts were collected, washed and resuspended in 0.1 M K/Mes pH 6.5, 6% (by vol.) 2 H 2 0 ,at a cell density of 1.25 x 10' cells/ml and a 15-min 31P-NMR spectrum of the aerobic microcyst suspension recorded (A) on an AM400 Bruker spectrometer. Nitrogen was then bubbled into the cell suspension and, after a 15-min adaptation period, a IS-min "P-NMR spectrum (B) was taken. Oxygenation was then reinstated and after a 15-min purging, a 31PNMR spectrum taken (C). Arrows in (B) correspond to the positions of cytosolic Pi and fi-NTP of the aerobic spectrum. Peak assignments were as in Fig. 1

triphosphates. The pH of the cytosol, deduced from the chemical shift of cytoplasmic Pi, was 7.4 k 0.05 (n = 6). The ratio of purine/pyrimidine nucleoside triphosphates, determined on an HC104 extract, was 3, thus unchanged as compared to P. pullidurn microcysts. Free magnesium concentration was cdculated to be 0.32 & 0.1 mM (n = 6), a value very close to that in P . pallidurn microcysts (see above) and D.discoideum amoebae and spores (Satre and Martin, 1985; Klein et al., 1988).The in vivo 13C-NMRspectrum of P.pullidum amoebae (Fig. 4B) was dominated by the contributions from phospholipid acyl chains and protein side chains, resonating at 15 - 50 ppm (methylene aliphatic carbons of fatty acid chains), 130 ppm (olefinic carbons) and 175 ppm (carbonyls 31P-and 13C-NMR spectra of P. pallidurn amoebae of phospholipids). When compared to microcysts, trehalose, In vivo 31P-and 13C-NMRspectra of P.pallidurn amoebae GroPCho and GroPEtn levels were drastically decreased in (Fig. 4A and B) were very similar to those of D. discoideum axenic amoebae. The residual fraction of these compounds in amoebae (Satre et al., 1986; Martin et al., 1987; Klein et al., amoebae might be accounted for by a small percentage of 1990). The 31P-NMR spectrum (Fig. 4A) showed the typical microcysts ( 5 - loo/,) present among the population of broad cluster around Oppm, with peaks of phospho- amoebae. The high amounts of glycogen observed in D. dismonoesters, cytosolic Pi,Imp6 and external Pi and the three cnideurn amoebae (Klein et al., 1990) were not found in P. broad resonance lines characteristic of nucleoside di- and pallidurn amoebae.

durn microcysts (Table 1) were the disaccharide trehalose (30.8 mM), the three polyamines: 1,3-diaminopropane (3.5 mM), putrescine (1 6 mM) and spermidine (3 mM), as well as the phosphodiesters GroPCho and GroPEtn (about 5 and 3 niM respectively), already detected by 3'P-NMR spectroscopy (Table 1). All the above resonance lines could be identified in the in vivo "C-NMR spectra of P.pallidurn microcysts, besides the contributions from phospholipid and protein chains at 1550 ppm and around 130 ppm and 175 ppm (not shown).

8 52 Table 2. Assignments of resonances for the proton-decoupled naturalabundance I3C-NMR spectra of P. pallidurn amoebae and microcysts.

33 32 29 26

KO,!

Peak

11

PPm 23.00 24.10 24.80 25.80 26.90 31.40 37.60 37.60 40.00 40.10 41 .oo

12 13 14

45.40 47.90 54.70

15 16

54.80 60.30

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

60.90 60.90 61.44 61.60 63.20 63.20 67.00 67.40 70.32 70.50 70.87 71.50 71.90 72.16 72.16 72.90 73.40 74.03 74.82 75.14 75.55 76.41 16.16 79.36 79.36 92.31 94.20 96.60 103.30

20

,

1 2 3 4 5 6 7 8 9

I

I

9.10

10

I

1

100

80

1

1

J

60

40

20

Chemical shift (ppm)

Fig. 3. Proton-decoupled natural-abundance I3C-NMR spectrum of qn HCIO4 extract of P. pallidurn microcysts. An HCIO4 extract of aerobic P. pallidurn microcysts (7.2 x lo9 microcysts) was prepared as described under Materials and Methods. A total of 5400 free-induction decays was accumulated on a WM250 Bruker spectrometer with interpulse delays of 4 s. Assignments for the numbered resonances are given in Table 2.

Excystment of P . pallidurn microcysts as followed by 31Pand I3C-NMR Transitions between the two states, amoebae and microcysts, can easily be triggered by adjusting the ionic strength of the incubation medium; further experiments were focused on the evolution of phosphorus and carbon metabolites during the excystment and encystment processes. The evolution of the major phosphorylated metabolites detected in the 31P-NMRprofile of an oxygenated suspension of microcysts was followed as a function of time after reduction of the ionic strength of the incubation medium to 10 mM K/Mes pH 6.5 (Fig. 5). After a short latency, GroPCho and GroPEtn were used up linearly with time until after 3 h, by which time 80% was consumed. Cytosolic pH increased with time for 80-90 min from 7.2 to reach the value observed in axenic amoebae (pH 7.4). After that time limit (80 - 90 min), NMR-invisible polyphosphates were extensively broken down into NMR-visible shorter polymers. As excystment was followed in K/Mes buffer in this 31P-NMR experiment, NMR-visible polyphosphates were likely to arise from degradation of long-chain NMR-invisibile polyphosphate polymers, and not from neosynthesis. Polyphosphates produced during excystment were extracellular, as they were fully removed by washing the newborn amoebae. The NTP/ (NTP + NDP) ratio and free Mg2+both remained stable during excystment, at levels similar to that of starting microcysts. The overall excystment process followed by I3C-NMR is illustrated by a difference spectrum between the amoeba and the microcyst stages (Fig. 6A) and the kinetics of evolution of the metabolite concentrations shown in Fig. 6 B . Trehalose, GroPCho and GroPEtn were used during the excystment,

Chemical shift

44 45

Metabolite

C5 spermidinc C2,3 putrescine C2,6 spermidine C2 diaminopropane C3 glutamine C4 glutamine C1,3 diaminopropane C1 spermidine C1,4 putrescine C7 spermidine CH2NH2 GroPEtn C3 spermidine C4 spermidine (CH,), N GroPCho C2 glutamine CH2CHzN GroPCho C 6 j cellobiose C6ct cellobiose C6’ cellobiose C6 trehalose CH2CH2NH2GroPEtn C H 2 0 H GroPCho, GroPEtn CHIN Gro PCho CH20PGroPCho,GroPEtn C4’ cellobiose C4 trehalose CSz cellobiose CHOH GroPCho, GroPEtn C2 trehalose C3r cellobiose C2a cellobiose CS trehalose C3 trehalose C2’ cellobiose C2D cellobiose C3p cellobiose C5P cellobiose C3’ cellobiose C5’ cellobiose C4p cellobiose C4a cellobiose Cla cellobiose C1 trehalose C l p cellobiose C1‘ cellobiose

__

1 11 111

IV V Vl VII VIII IX X XI XI1 XI11

14.8 23.4 25.2 27.8 30.1 34.3 40.1 54.7 62.8 67.0 71.5 128.5 130.0 173.2

o(CH,) fatty acids w-1(CH2-CH3)fatty acids CH2-CH2-CO-fatty acids -CH = CH-CHI-CH, fatty acids -(CH2)n-fatty acids -CH2-CHI-CO- fatty acids -CH2-NH2PtdEtn (CH3)3-N PtdCho Cl glycerol (ester) C3 glycerol (ester) C2 glycerol (ester) CH = CH CH = CH-CH,-CH = CH, CH = CH-CHZ-CH CH R-0-C-OR

853 cyt Pi + lnsP6

'

InsP6

i

I

ext Pi

f

PNDP WTP

aNDP aNTP

I

I

I

Y "

I

I

1

I

10

5

0

J

I 180

'

.

I

160

1

I 140

1

I

1

-5 -1 0 Chemical shift (ppm)

I 120

I

1 100

I

1 80

I

1

60

I

I

I

-1 5

-20

-25

, . I 40

' . I 20

, . I 0

Chemical shift (ppm)

Fig.4. In vivo "P- and I3C-NMR spectra of P.pallidurn axenic amoebae. (A) "P-NMR spectrum of a suspension of acrobic P. pullldum amoebae. Data were accumulated from a total of 2.3 x lo9 amoebae oxygenated in 7 ml 40 mM Na/Mes pH 5.3, 6% (by vol.) 'H20. This proton-decoupled spectrum, recorded on an AM400 Bruker spectrometer was the sum of 6000 free-induction decays with 0.6-s interpulse delays. Peak assignments were as in Fig. 1 . (B) Proton-decoupled natural-abundance 13C-NMR spectrum of a suspension of aerobic P. pallidurn amoebae. A total of 4.1 x lo9 axenic amoebae collected a t the early plateau phase of the growth curve (2.9 x 10' cells/ml) was suspended in 17 mM Na/K/Pi pH 6 5 containing 6% (by vol.) 'HzO and transferred to a 15-mm-diameter NMR tube bubbled with oxygen. The NMR spectrum was recorded on an AM400 Bruker spcctrometer and was the sum of 7200 free-induction decays with 0.4-s interpulse delays. Assignments for the numbered resonances are given in Table 2.

8 54 6

g

, 30

34-41

I -

A

4

a g

3

0

60

120

180

240

300

360

Time (min)

Fig.5. Evolution of phosphorylated metabolites during excystment of P. pallidurn microcysts. A total of 9.6 x lo9 P. pallidurn microcysts was resuspended in 10 mM K/Mes pH 6.5, 6 % (by vol.) 'HzO, to trigger excystment; 31P-NMR spectra were recorded as a function of time on an AM400 Bruker spectrometer. The areas of the resonance peaks of GroPCho, GroPEtn and polyphosphates were measured in the successive spectra and normalized relative to the concentrations of the metabolites in microcysts at time zero. (0)GroPCho; ( 0 ) GroPEtn; (A)polyphosphates.

43 33 100

110

and a mixture of sugars produced. Consumption of trehalose followed the same kinetics as for GroPCho and GroPEtn and was almost complete in 3 - 4 h. One of the sugars produced was identified unambiguously as the p-1,Cglucobiose (cellobiose) from the chemical shifts of its resonance lines and by direct spiking with authentic cellobiose. Several resonance lines remained unidentified but, from their chemical shifts, it could be inferred that they arose from a sugar closely related to cellobiose, as for example gentiobiose (~-1,6-glucobiose). These sugars appeared as extracellular degradation products from the cellulose/glycogen-likewall after 60 - 90 min as they were removed by a washing step, and were probably linked to the emergence process.

As encystment of amoebae into microcysts was triggered by 3 20 mM K/Pi,the measurement of GroPCho and GroPEtn evolution was not feasible by 31P-NMR because of the overlapping of external phosphate and phosphodiester resonance lines. Slight variations of extracellular pH precluded use of difference spectra to follow phosphodiester variations during encystment. No short NMR-visible polyphosphate intermediates could be detected, but PPi was formed after 90 min up to 0.2 mM at the plateau (4- 5 h into encystment). In contrast, in a I3C-NMR difference spectrum (Fig. 7A), the formation of trehalose during encystment and, to a lesser extent, of fatty acids and of the two soluble phosphodiesters, GroPCho and GroPEtn, was clearly demonstrated. The amount of trehalose started to increase as soon as the ionic strength of the medium was increased (Fig. 7B) and the phenomenon was almost complete after 5 - 6 h. The situation in P. pallidurn was thus different from that found in D. discoideum amoebae in which no significant accumulation of trehalose was observed when an osmotic stress was imposed (Ternesvari, Cotter, Klein, Martin and Satre, unpublished results). It should also be remembered that D . discoideum amoebae do not encyst in such conditions.

90 ao 70 Chemical Shift (ppm)

50

60

35

30 25

20 L

15

0

10 5

0 0

Encystment of P.pallidurn amoebae as followed by 31Pand I3C-NMR

29

60

1 2 0 180 240 Time (min)

300

360

Fig. 6. Evolution of carbon metabolites during excystment of P.pallidurn microcysts. (A) A total of 1 x 10" P.pnllidum microcysts was washed in 0.1 M K/PipH 6.5 and resuspended in 10 mM K/Pi pH 6.5,6% (by vol.) *HzO,to trigger excystment. 13C-NMR spectra were recorded as a function of time on an AM400 Bruker spectrometer. A difference spectrum between the 8 - 10-h period (corresponding to nascent amoebae) and the 0-2-h period (microcyst stage) into excystment is shown. Positive peaks correspond to metabolites produced during the excystment process whereas negative peaks correspond to the compounds which are consumed. Assignments for the numbered resonances are given in Table 1. (B) The areas of the C1 resonance line of trehalose (resonance line 43) and the C1' resonance line of cellobiose (resonance line 45) were measured in the successive spectra, normalized relative to that of trehalose at time zero and converted into absolute Concentrations using the trehalose concentration in microcysts (Table 1). The levels of trehalose (0)and cellobiose ( 0 ) were plotted as a function of time.

Microcysts formed on agar plates when germinated at high concentration in NMR experiments were capable of redifferentiating into microcysts (without growth) inside the NMR equipment if subjected to high salt solutions consisting of 120 mM K/Pi. These second generation microcysts would germinate readily when washed and placed in 10 mM K/Pi solutions.

29 28

20

I

3 3'(3 2

I

110

,

I

100

.

l

l

.

1

90

80

0

.

I

855

.

I

.

I

1 ' .

.

70 60 50 Chomlcal rhltt (ppm)

40

60

300

1 2 0 180 2 4 0 Time (min)

1

30

.

1

,

20

360

Fig. 7. Evolution of carbon metabolites during encystment of P. pallidurn amoebae. (A) A total of 4.4 x 1 O9 P.pallidurn amoebae was collected in 40 mM K/Pi pH 6.5, washed in that buffer and resuspended at 5 x 10' cells/ml in 0.12 M K/Pi pH 6.5, 6% (by vol.) 'HzO, to trigger encystment. I3C-NMR spectra werc recorded as a function of lime on a n AM400 Bruker spectrometer. A difference spectrum between the 31 5 - 375-min and 15- 75-min periods into encystment is shown. Peaks correspond to the metabolites produced during the encystment process. Assignments for the numbered resonances are given in Table 2. (B) The height of the C1 resonance line of trehalose (resonance line 43) was measured in the successive spectra, normalized relative to the maximal height extrapolated at infinite time and converted into absolute concentrations using the trehalose concentration in microcysts (Table I). The level of trehalose ( 0 )was plotted as a function of time.

Examination of microcyst resistance properties While the microcyst is not as resistant to environmental extremes as the spore, it nevertheless is reported to be partially resistant to desiccation and heat (Cotter and Raper, 1968a; Raper, 1984). We found that microcysts formed on the surface of glucose/salt agar plates were resistant to the desiccation which resulted when the plates were air-dried in the dark; after 14 months, the majority of such microcysts still germinated within 12 h upon rehydration. Likewise, the microcysts formed in liquid culture were viable for several months and would germinate when washed and incubated in 10 mM K/Pi or K/Mes buffers. While spores of P . pallidurn tolerate heat at 45°C for 30 min, this treatment severely damages microcysts. Mature microcysts exposed to heat were not killed, but the lag phase before amoebae began to emerge was extended for up to 8 h. As observed above in Fig. 2, the NTP level in dormant and germinating microcysts was maintained by an active metabolism. The reimposition of O2 resulted in the reappearance of the NTP peak and the resumption of germination. Dormant microcysts were resistant to anaerobiosis while germinating microcysts became progressively more sensitive to an-

aerobiosis as they approached the emergence stage (data not shown).

DISCUSSION In this work, we have analyzed metabolites in P . pallidurn amoebae and microcysts, together with their variations during the transitions between these two states by 'P- and protondecoupled natural-abundance I3C-NMRspectroscopy. NMR analysis of P . pallidurn spores was not performed because of the large quantity of material needed for NMR analysis (about 1 ml packed cell volume). Spore heads of P. pallidurn are very small and the whorls of lateral branches hinder efficient collection of spores. The major components detected in microcysts were Pi, nucleoside diphosphates and triphosphates, the polyamines diaminobutane, diaminopropane and spermidine, InsP6, the phosphodiesters GroPCho and GroPEtn and trehalose. The latter three compounds were not present in amoebae, as the low concentrations measured in amoebae could be correlated to the presence in the axenic culture of a small percentage of microcysts.

Additional data from our studies indicate that excystment P ol~~sphondylium is another genus of the Dictyosteliaceae family, besides Dictyostelium, to contain millimolar InsP6 is an ATP-dependent strictly aerobic process, and that new concentrations, both in its amoeba and microcyst forms. Be- born amoebae, emerged from microcysts inside the N M R sides the plant kingdom where InsP, is found in high concen- spectrometer, are readily able to form second-generation trations in seeds and in tubers (Kime et al., 1982; Delfini et microcysts. Encystment of amoebae triggered by high salt al., 1985), Dictyostelids appear thus rather exceptional for concentrations corresponded to the mirror image of excysttheir soluble InsP, content, 1 - 2 orders of magnitude higher ment. As an immediate response to osmotic stress, P. pallidurn than in mammalian cell lines (Downes and Macphee, 1990). amoebae synthesized trehalose and hydrolyzed a fraction of Even though its concentrations are different in the two living their phospholipids into fatty acids, GroPCho and GroPEtn. stages by a factor of about 2, InsP6 does not seem to be used Synthesis of trehalose induced by high ionic strength seems during excystment; the difference might be linked to a later limited to amoebae able to encyst, as no trehalose was formed metabolite concentration adjustment. The function of such in D. discoideum amoebae in high salt solutions. high amounts of InsP, in slime molds is still unknown and its This work was supported by grants from the Commissariat a elucidation a challenge for further research. The presence of high polyamine concentrations is another constant among the I’Energie Atomique (Dkpurtement de Biologie Molficuluire et members of this family of slime molds. The three polyamines StructuralelBiolo~ieCellulaire) and from the Centre National de lu present in Dicfyosfeliurn, putrescine (1 ,Cdiaminobutane), 1,3- Recherche Scientifique i Unit; de Recherche AssociCe 1130). diaminopropane and spermidine, were also detected in Poljqdiondylium, both in microcysts and in amoebae; no major variations were measured during the transitions be- REFERENCES tween these living stages. North and Murray (1980) reported that the unusual polyamine 1,3-diaminopropane was present Blascovics, J. C. & Raper, K. B. (1957) B i d . Bull. 113, 58-88. J. T. (1967) The cellular slime molds, 2nd cdn, pp. 1-205, in D.discoideum and D.mucoroides regardless of the food Bonner,Princeton University Press, Princeton NJ. bacterium (Escherichia coli or Klehsiella pneumoniae), but ab- Choi, A. H. C. & O’Day, D. H. (1982) DEV.Biol.92,356-364. sent in P. pallidum, P . violaceum and D. purpureum. The differ- Cotter, D. A. & Rapcr, K. B. (1968a) J . Bucteriol. 96, 1680-1689. ence with our results may be due to growth conditions or to Cotter, D. A. & Raper, K. B. (1968b) J . Bucteriol. 96, 1690-1695. a strain difference. Cotter, D. A . & Raper, K . B. (1970) Dev. Biol.22, 112- 128. A major finding of our studies on Polysphondylium Delfini, M., Angelini, R., Bruno, F., Conti, F., Di Cocco, M. E., Giuliani, A. M. & Manes, F. (1985) Cell. Mol. Biol. 31, 385microcysts was the observation of large amounts of two 389. phosphodiesters, GroPCho and GroPEtn, and of a storage sugar, trehalose. As far as these compounds are concerned, P. Downes, C. P. & Macphee, C. H. (1990) Eur. J . Biochem. lY3,1-18. G. W. & Hohl, H. R. (1980) Cyiobios 29, 7-16. pallidurn microcysts are very similar to D. discoideum spores. Endos, Fiske, C. H . & SubbaRow, Y. (1925) J . Biol. Chem. 66, 375-400. The major difference between microcysts and spores resides Francis, D. (1975) J . Gen. Microbiol. 89, 310-318. in the presence of tremendous amounts of glutamine (70 mM) Francis, D. W. (1979) DifSerenfiation 15, 187-192. and glutamate (20 mM) in spores (Klein et al., 1990), whereas Githens, S. & Karnovsky, M. L. (1973) J . Cell Bid. 58, 522-535. these amino acids are below the limit of I3C-NMR detection, Gupla, R. K., Gupta, P., Yushok, W. D. & Rose, Z. B. (1983) Biochem. Biophys. Res. Commun. 117, 210-216. which was about 3 mM with microcysts. This difference might be linked to the pathways leading to the formation of spores Hohl, H . R.. Miura-Santo, L. Y. & Cotter, D. A. (1970) J . Cell Sci. 7,285 - 306. and microcysts, as spores differentiate in a starvation-induced Kime, M. J., Ratcliffe, R. G., Williams, R. J. P. & Loughman, B. C. process and microcyst formation is induced by high salt. (1982) J . EXP.Bot. 33, 656-669. From the time course of metabolite evolution followed by Klein, G., Cotter, D. A., Martin, J. B., Bof, M. & Satre, M. (1988) 31 P- and 3C-NMR spectroscopy, excystment of P. pallidum Biochemistry 27, 8199-8203. can be divided into two distinct phases: (a) an early phase, Klein, G.. Cotter, D. A,, Martin, J. B. & Satre, M. (1990) Eur. J . which starts as soon as the salt concentration of the solution is Biochem. 193, 135-142. reduced, characterized by the rapid consumption of trehalose, Lonski, J. (1976) Dev. Biol. 51, 158-165. GroPCho and GroPEtn and by the adjustment of cytosolic Loomis, W. F. (1975) Dictyostelium discoideum. A developmentul system, pp. 1-214, Academic Press, New York. pH to that of emerging amoebae; (b) a later phase, starting 60 - 90 min after induction of the excystment process, which Martin, J. B., Foray, M. F., Klein, G. & Satre, M. (1987) Biochim. Biophys. Acta 931, 16-25. is characterized by a shortening of very-long-chain polyphosphates and a partial hydrolysis of the cellulosicwall yield- North, M. J. & Murray, S. (1980) F E M S Lett. 9, 271 -274. O’Day, D. H. (1974) Dev. B i d . 36,400-410. ing cellobiose. There is a good correlation between these two O’Day, D. H. & Paterno, G. D. (1979) Arch. Microbiol. 121, 231 phases and the previously described swelling and emergence 234. phases (Hohl et al., 1970). O’Day, D. H., Gwynne, D. I. & Blakey, D. H. (1976) Exp. Cell Res. In agreement with previous data (Tisa and Cotter, 1979) 97,359-365. and as described for D . discoideum spores (Klein et al., 1990), Raper, K. B. (1984) The Dictyostelids, pp. 1-453, Princeton University Press, Princeton, NJ. the non-reducing carbohydrate trehalose serves as a carbon source during excystment. Fatty acids produced by sequential Satre, M., Klein, G. & Martin, J. B. (1986) Biochimie (Puris) 68, 1253-1261. hydrolysis of phospholipids into GroPCho and GroPEtn durM. & Martin, J. B. (1985) Biochem. Biophys. Res. Commun. ing encystment may ensure a complementary source of energy Satre, 132, 140-146. for later excystment (Cotter and Raper, 1970; Klein et a]., Sussman, M. (1987) Methods Cell Biol. 28, 9-29. 1988), in addition to that derived from sugar metabolism. The Tisa, L. S. & Cotter, D. A. (1979) Curr. Microbiol. 3, 33-35. decrease of both GroPCho and GroPEtn during excystment Toama, M. A. & Raper, K. B. (1967a) J . Bucteriol. 94, 1 143- 1149. is probably linked to an increased membrane phospholipid Toama, M. A. & Raper, K. B. (1967b) J . Bucteriol. Y4, 11 50- 1153. Watts, D. J. &Ashworth, J. M. (1970) Biochem. J . 119, 171-174. synthesis.