Radiotracer Studies of Nitrogen Metabolism in Cyanobacteria

14 Radiotracer Studies of Nitrogen Metabolism in Cyanobacteria

Downloaded by NANYANG TECHNOLOGICAL UNIV on June 8, 2016 | http://pubs.acs.org Publication Date: March 1, 1982 | doi: 10.1021/ba-1981-0197.ch014

J. C. MEEKS Department of Bacteriology, University of California, Davis, CA 95616 N has been used as a probe to examine the metabolic pathways and control of inorganic nitrogen assimilation in a number of cyanobacteria. Data derived from kinetic and inhibitor studies indicate that the primary pathway of NH + assimilation, whether metabolically derived from [ N]N , NO -, or NO -, or supplied exogenously, con sists of the sequential activities of glutamine synthetase and glutamate synthase in all dinitrogen-fixing species examined. Aerobic reduction of dinitrogen to ammonium occurs in heterocysts of those species that differentiate these spe cialized cells. In heterocysts [ N]N -derived NH + is rapidly assimilated into glutamine. However, the newly synthesized glutamine must be translocated to adjacent vegetative cells for transamidation to form glutamate. A portion of the synthesized glutamate is then returned to heterocysts for continued amidation. 13

13

4

13

2

13

3

13

2

13

2

13

4

V

egetative cells of certain filamentous cyanobacteria can differentiate to form thick-walled cells termed heterocysts and, i n species of Nostoc and Anabaena, the heterocysts form at semiregular intervals along the filaments (Figure 1 ) . Heterocysts have been shown to be the major (if not only) sites of dinitrogen fixation under aerobic growth conditions i n these cyanobacteria (1,2,3). The regulation of heterocyst formation is of interest because understanding the mechanisms that control the differentiation of heterocysts would be a major step toward establishing manipulations that could convert these cyanobacteria into biological "ammonia factories" for agricultural use (4). Moreover, differentiation processes i n the morphologically simple cyanobacteria provide a unique experimental system i n w h i c h to study the detailed biochemical mecha nisms that govern the formation of multicellular patterns ( 5 ) . 0065-2393/81/0197-0269$06.50/0 © 1981 American Chemical Society

Root and Krohn; Short-Lived Radionuclides in Chemistry and Biology Advances in Chemistry; American Chemical Society: Washington, DC, 1982.


270

SHORT-LIVED RADIONUCLIDES

Figure I. Photomicrograph of N^grown Anabaena cylindrica: h points to heterocyst; marker is 4 fiM Wolk and Quine (6) indicated that the semiregular spacing of heterocysts in the filaments occurs because heterocysts produce a dif fusible substance(s) that inhibits nearby vegetative cells from differen tiating into heterocysts. Since heterocysts are the major sites of aerobic dinitrogen fixation, it is possible that the inhibitory substance(s) produced by heterocysts is a nitrogenous compound. It is also well known that growth of heterocyst-forming cyanobacteria in the presence of ammonium and, in some cases, nitrate represses heterocyst formation (7,8). These observations indicate that the differentiation of vegetative cells into heterocysts is under control of the external nitrogen concentration and may be analogous to nitrogen control in certain enteric eubacteria (9,10). We have utilized the radioactive isotope of nitrogen, N (ft*, t = 10 min), to study nitrogen metabolism in intact filaments, to study inter actions between vegetative cells and heterocysts and, in particular, to attempt to identify the regulatory compound(s). In this chapter I shall review the results of studies, most of which were performed in collaboration with C. P. Wolk, J. Thomas, members of the staff of the Michigan State University cyclotron, and other colleagues, in which the initial products of assimilation of N-labeled dinitrogen, ammonium, nitrate, and nitrite by intact cyanobacteria (11-14) and by isolated heterocysts (I) were identified. Most of the experiments were performed using Anabaena cylindrica, which forms intercalary hetero cysts. Other cyanobacteria used were: A . variabilis and A. P C C 7120, which also form intercalary heterocysts; Cylindrospermum licheniforme, which normally forms heterocysts only at terminal positions; Plectonema boryanum, which does not form heterocysts but grows with N as the sole nitrogen source under microaerobic conditions; Gloeothece sp., a unicellular species that does not form heterocysts but grows slowly with N as the sole nitrogen source under the aerobic conditions; and Anacystis 1 8

1/2

13

2

2


14.

271

Nitrogen Metabolism in Cyanobacteria

MEEKS

nidulans (syn. Synechococcus P C C 6301 [15]), a unicellular species that does not form heterocysts or grow with N as the sole nitrogen source aerobically or microaerobically. 2

Experimental Preparation of Cyanobacteria for Labeling. Details of the fermenter culture of cyanobacteria have been presented elsewhere (1,11-14). In all cases a short time before exposure to N the cyanobacteria were harvested and concentrated to 27 ftg of chlorophyll a per mL by centrifugation and were incubated aerobically or microaerobically in the absence of N prior to use, depending on the experiment and species. In some experiments, L-methionineDL-sulfoximine (MSX; Sigma Chemical Company, St. Louis, Missouri),. O-diazoacetyl-L-serine (azaserine; Calbiochem, La Jolla, California), or aminooxy acetate (AOA; Sigma) were added singly or in combination to the cyanobacterial suspensions at the time of resuspension to a concentration of each of ImM ( [ N ] N experiments) or 2mM ( N H or N 0 " experiments; final concen tration ImM upon dilution with the aqueous label). Generation of, and Labeling with, N . N was generated by one of two reactions. At the Michigan State University cyclotron, N was generated by the C(p,n) N nuclear reaction with beams of 11-MeV protons at a bom bardment current of 1-4 fiA for 20-30 min (16,17). The target was 18.6-mg amorphous carbon, 97 at. % C (Monsanto Research Corporation, Mound Laboratory, Miamisburg, Ohio). At the University of California, Davis, Crocker Nuclear Laboratory cyclotron, N was generated by the 0 ( p , a ) N nuclear reaction with beams of 20-MeV protons at bombardment current of 20 fiA for 20 min. The target was glass-distilled water in a 60-mL recirculation target system (18). [ N ] N and N H were generated from the irradiated C target; N 0 ~ was purified and N 0 " generated from the irradiated H 0 target. [ N ] N was generated by subjecting the irradiated target, together with 0.18 mg K N 0 , to Dumas combustion in a Coleman Nitrogen Analyzer, Model 29 (Coleman Instruments Corporation, Maywood, Illinois) supplemented with an additional postheater. Carrier C 0 and oxides of nitrogen were removed with a liquid nitrogen trap, and a Toeppler mercury pump was used to com press the [ N ] N into 1.0-mL evacuated vials (16,17). N H was generated by acid digestion (approximately 250°C, 10 min) of the irradiated target with concentrated sulphuric acid saturated with potas sium dichromate. The solution was made alkaline with saturated sodium borate and 40% (w/v) NaOH, and oxides of nitrogen were reduced with saturated silver sulfate. N H was then distilled under vacuum and collected in a liquid nitrogen trap (1,13). N 0 " was purified from contaminating N 0 " and N H in a two-step procedure modified from (19). N 0 " was first oxidized to N 0 " by 1% (v/v, final concentration) H 0 with the solution pH adjusted to 2.0 with formic acid. The solution was evaporated to dryness, removing most of the hydrogen peroxide and formic acid. The residue, containing N 0 " and N H , was resuspended with distilled water and the pH adjusted to 10.0. The sample was evaporated to a dryness a second time to remove contaminating N H . The final residue was resuspended with buffer to a final pH of 7.0^-7.5, and 50 fig of catalase was added to remove any residual H 0 .


1 3

2

13

13

2

4

+

1 3

1 3

3

1 3

1 3

13

13

1 3

1 3

13

13

1 3

2

4

13

+

13

1 3

1 3

3

16

2

2

2

3

2

13

1 3

4

2

+

1 3

1 3

3

1 3

3

1 3

2

2

1 3

4

+

1 3

2

3

2

1 3

3

1 3

1 3

2

2


4

+

3

272


N 0 " was synthesized fro m N 0 " by reduction in a cadmium-copper column and purified from N H by vacuum evaporation at pH 10.0 (20). Fixation of [ N ] N by 250 fiL of concentrated cyanobacterial suspension took place in the light (ca. 400 ft-c from an incandescent lamp) at room tem perature (ca. 23°C) under an atmosphere of A r : N : C 0 (97:2:1, vol. ratio) in 1.0-mL Reactivials (Pierce Chemical Company, Rockford, Illinois) fitted with stopcocks. The suspensions were stirred with a small triangular stirring bar set in the cone of the vial. A short length of tubing led from the stopcock through a serum stopper to a centrifuge tube containing 1.0 mL of methanol. The reaction was initiated by addition of the cyanobacterial suspension and at prescribed times the stopcock was opened, the reaction stopped by injection of the suspension into the methanol, and the unfixed [ N ] N then removed by flushing with cylinder gas (II, 12,14). Assimilation of N H , N 0 " , and N 0 " by the suspended cyanobac teria took place in the light, generally in 15-mL conical centrifuge tubes under air. Reactions were initiated by combining 50 or 250 (iL of N solution with 250 /xL of cyanobacterial suspension and were terminated by mixing the result ing suspension with 4 volumes of methanol. The methanolic suspensions in all cases were mixed for 1 min on a vortex micromixer to complete extraction and centrifuged at 1000 X g for 1 min. The supernatant solution was decanted and subjected to analysis. The supernatant methanolic extract was treated in one of two ways (I, 12,13): (i) The fluid was subjected sequentially to vacuum distillation at pH 10.0 to recover free ammonia and then to steam distillation in strong alkali to recover amide nitrogen-derived ammonia. Radioactivity in the distillates and in a portion of the original solution was determined by scintillation spectros copy and was corrected for decay to a standard time, (ii) Alternatively, in preparation for electrophoresis, the fluid was dried under vacuum at 50°C, the residue dissolved in 200 /iL of 80% methanol and redried as before, and the final residue dissolved in 50 /iL of 80% methanol to which were added 5 /xL of a standard amino acid solution. The final methanolic solution was spotted on a 2 X 10 mm area parallel to the short axis of 5 X 20 cm glass plate bearing a 0.1-mm layer of cellulose (E. Merck and Company, Darmstadt, West Germany). The lipid-soluble substances in the dried spotted material were displaced by ascending chromatography in the short axis of the plate in chloroform/methanol (3:1, vol. ratio or—for extracts of isolated heterocysts— 1.3:1, vol. ratio). The plate was then dried, sprayed with buffer, and subjected to electrophoresis at 3000 V for 1.5-12 min using a high-voltage electrophoresis apparatus (Model Q l l SAE 3202, Shandon Scientific Company, London, England; or custom-built). Buffers used were with 70mM sodium borate, pH 9.2 and, on occasion, 750mM formic acid, pH 2.0. After electrophoresis, the dis tribution of N on the thin-layer plates was determined with a radiochromatogram scanner (Model 7201, Packard Instruments Company, Downers Grove, Illinois). Radioactive amino acids were identified by comigration with stable and C-labeled amino acids during electrophoresis (1,11,12). The radio activity in amino acids was quantitated by integration of peaks in radioscans, with corrections applied for decay to a standard time. An example of onedimensional separation and radioelectrophoretogram scanning following fixa tion of [ N ] N for varying periods of time is shown in Figure 2. Aspartate and glutamate migrate very closely during electrophoresis at pH 9.2. Thus radioactivity from aspartate generally appeared first as a shoulder on, then as a composite with, glutamate as one major peak. In most instances, radio activity in this region was calculated as a combined glutamate plus aspartate 13

13

2

1 3

13

4

3

+

2


2

2

13

1 3

4

+

13

3

13

2

2

1 3

1 3

14

13

2


MEEKS


273


14.

DISTANCE ( c m ) Journal of Biological Chemistry

Figure 2. Scan of radioactivity from N in electrophoretograms of com pounds extracted from A. cylindrica with 80% methanol after (A) 120, (B) 60, and (C) 20s of fixation of [ N]N (12): O, origin; L, lipids dis placed by chromatography prior to electrophoresis. Distance is measured from the negative (left) end of the plate. The standard amino acids were visualized by spraying the thin-layer plate with acidic ninhydrin solution after radioelectrophoretogram scanning. 13

13

B



274


Journal of Biological Chemistry

Figure 3. Two-dimensional scan of radioactivity from N in two-dimen sional electrophoretogram (positive end of the plate to the right) and chromatogram of N-hbeted compounds extracted from A. cylindrica with 80% methanol after 900 s of assimilation of NHf (13): (%) points labeled with radioactivity. Peak areas of radioactivity were localized by drawing isorads on the printout at (- • •) 3, ( ) 9, ( ) 24, (—) 63, ana (••"••) disintegrations per area element. The mean background during the 20-min scan was approximately 0.08 counts per area element. 13

13

13

peak. To determine the respective radioactive contribution of these two amino acids in this region, in certain experiments the plates were scanned and sprayed with ninhydrin solution, and then the ninhydrin positive areas ascribed to aspartate and to glutamate were excised, eluted with water, and N radioac tivity in the eluents were determined by scintillation spectroscopy. To obtain additional verification of the identity of the N-labeled amino acids, the methanolic extracts from certain N H experiments were subjected to two-dimensional separation and scanning. In these experiments the concen trated methanolic solution was spotted on a 2 X 5 mm area parallel to the long axis of a 5 X 20 cm thin-layer cellulose glass plate. Following initial chroma tography (along the short axis of the plate) and electrophoresis, (at pH 9.2 for 9 mm), the thin-layer plate was dried and subjected to ascending chromatogra phy in the short direction of the plate in phenol/water (3:1, v/v) equilibrated with 3% (v/v) aqueous N H O H for 17 min. The plate was scanned in two dimensions using the scanner described by Markam et al. (21). The rays arising from a given area element were added. After the scan was completed, the sums were printed in a corresponding two-dimensional array. The standard amino acids were visualized with ninhydrin and their positions in the twodimensional scan determined by reference to points labeled with radioactivity. An example of this separation and scanning is shown in Figure 3. 1 3

13

1 3

4

+

4

Results and Discussion Pathways of Assimilation of N by Intact Cyanobacteria* F I X A TION OF [ N ] N . The first stable product of fixation of [ N ] N by 1 3

13

2

13


2

14.

275


MEEKS

intact Anabaena cylindrica was N H (Figure 4). After 1.25 s of fixation approximately 61% of the total N recovered in methanolic extracts distilled as N H . The fraction of N H in extracts then declined until, after 15 s of fixation, it accounted for less than 20% of the total radio activity. The radioactivity in the ammonium ion pool, once apparently saturated, remained low; the N-labeling of the amide nitrogen increased linearly for at least up to 15 s (Figure 4) and probably to 60 s (12). Incubation for up to 120 s with [ N ] N resulted in incorporation of N into three to six organic constituents (depending on the species) by the five N -fixing cyanobacteria examined (12,14). These radioactive constituents were tentatively identified as glutamine, glutamate, aspartate, citrulline, alanine, and arginine by their comigration with stable standard amino acids during electrophoresis at p H 9.2 (Figure 2). In A. cylindrica the identity of certain of these amino acids was also determined by coelectrophoresis at p H 2.0 and p H 9.2 with C-labeled amino acids and by two-dimensional coelectrophoresis and cochromatography followed in each dimension by radioscanning (12). Moreover, after 60 s of fixation by A. cylindrica, up to 87% of the N radioactivity that coelectrophoresed with stable glutamine distilled as amide nitrogen, and approxi mately 14% could be recovered as a-amino nitrogen (12). The initial organic products of fixation of [ N ] N were glutamine and glutamate in all species examined (Figure 5). After 15 s of fixation, 1 3

4

+

1 3

1 3

1 3

3

4

+

13


13

2

1 3

2

14

1 3

13

2

FIXATION TIME (s) JOURNAL OF BIOLOGICAL CHEMISTRY

Figure 4. Time course of incorporation of N into pools of N H and amide nitrogen (12). After fixation of [ N]N for 1.25, S, and IS s, the suspensions of A. cvlindrica were extracted and the methanolic extract subjected sequentially (O) to vacuum distillation at pH 10.0 to recover free NH and then (%) to steam distillation in the presence of alkali to recover amide-derived NH . The ordinate represents cpm of NH * or [ N]amide nitrogen in the fixation vial, corrected to the time of the start of fixation, and normalized to equal amounts of [ N]N (in /xCi) in the 13

13

13

4

+

2

S

13

13

S

13

13

g


U


276


100

20

Fixation

time (s)

60

100

Figure 5. Time course of the distribution of N in amino acids ex tracted after fixation of [ N]N for 15, 60, and 120 s by (A) A. cylindrica, (B) A. variabilis, (C) C. licheniforme, (D) P. boryanum, and (E) Gloeothece sp (12,14). The radioactivity of amino acids in methanolic extracts, subjected to electrophoresis, was quantitated by integration of peaks in radioscans, with corrections applied for decay: (A) glutamine, (O) gluta mate plus aspartate, (X) citrulline (or, in the case of C. licheniforme, alanine). 1S

13

t

radioactivity associated with glutamine accounted for 77% (Gleoeihece sp.) to 88% (A. cylindrica) of the total N recovered in organic products. The remainder of the N recovered after 15 s of fixation was associated with glutamate. The fraction of N in glutamine decreased and that in glutamate increased during longer incubation periods until, after 60-120 s of fixation, glutamate was more highly radioactive than glutamine in all five species. In A . cylindrica, after 120 s of fixation, radioactivity was additionally associated with aspartate, citrulhne, and arginine (Figure 2). Neither aspartate nor arginine were detectably radioactive after 120 s of 1 3

1 8

1 8


14.

277


MEEKS

fixation by the other four species. However, N-labeled citrulhne was observed in A. variabilis, Plectonema boryanum, and Gleoethece sp. within 60-120 s of fixation. Radioactivity was detected in alanine but not citrulline after 60-120 s fixation by Cylindrospermum licheniforme (14). When C-labeled carbamyl phosphate was added to methanolic extracts and was recovered in thin-layer electrophoretograms, no N radioactivity was detected to be associated with this compound after 60 s of fixation of [ N ] N by A. cylindrica (12). When A. cylindrica was exposed to [ N ] N for 15 s, the gas phase evacuated and then replaced with a mixture of N / C 0 (99:1, v/v), radioactivity associated with glutamine decreased while that associated with glutamate correspondingly increased as a fraction of the total N extracted with methanol (Figure 6). The change in the fraction of N in glutamine and glutamate was more rapid during the chase period than in typical time-course experiments (compare Figures 5 and 6). The time-course and pulse-chase experiments indicate that glutamine is the initial organic product of assimilation of [ N]N -derived NH4 , presumably by direct amidation of glutamate. Then glutamate becomes labeled, presumably by a sequential amination of a-ketoglutarate. To verify that the labeling of these amino acids occurs in a sequential 13

14

1 3


13

2

13

2

2

2

1 3

1 3

13

13

2

+

1.0

CHASE PERIOD (s) Journal of Biological Chemistry

Figure 6. Time course of the distribution of N in amino acids extracted from A. cylindrica when fixation of [ N]N for IS s was followed by evacuation of the label and then continued fixation of N /CO* (99:1, v/v) for 45, 105, and 285 s (12). The radioactivity of amino acids in metha nolic extracts, subjected to electrophoresis, was quantitated by integration of peaks in radioscans, with corrections applied for decay: (%) glutamine, (O) glutamate (plus a small fraction of aspartate after 45 s of chase), (A; citrulline, (M) arginine; and (D) an unknown substance that migrated between arginine and alanine. 13

13

g

2


278


reaction, a number of experiments were performed using specific inhib itors of enzymes (12). When A. cylindrica was incubated with [ N ] N in the presence of methionine sulfoximine, an inhibitor of glutamine synthetase, little or no N was detected to be associated with glutamate, glutamine, or any other organic product, and the amount of N H distilled from methanolic extracts increased nearly 18-fold over that in extracts from cells incubated in the absence of the inhibitor (12). Fixa tion of [ N ] N in the presence of azaserine, an inhibitor of glutamine amide transfer reactions, resulted in extensive accumulation of N radio activity in glutamine alone (12). ASSIMILATION O F N H . Studies on the assimilation of exogenous N H by N - and NH -grown cultures were initiated to determine whether the pathways of assimilation of exogenous ammonium differed from the pathway operative during N fixation. Metabolic products of assimilation of N H after 900 s of incubation included glutamine, glutamate, aspartate, citrulline, alanine, and arginine. These constituents were tentatively identified by their comigration with stable amino acids standards during electrophoresis at p H 9.2, sometimes followed by chromatography in an orthogonal direction and subsequent two-dimensional scanning (e.g., Figure 3). N in certain amino acids, such as aspartate and glutamate was also measured by elution of the ninhydrin reacting material on the thin-layer plate and by scintillation spectroscopy. Moreover, a major fraction of the N radioactivity asso ciated with glutamine after incubation periods of 15 s or less could, in the case of each species, be distilled as amide nitrogen. After 1 s of assimilation of N H , 77 (Anacystis nidulans) to 97% (C. licheniforme and A. cylindrica) of the N in organic compounds extracted with methanol was associated with glutamine (Figure 7). In A. cylindrica, approximately 88% of the N associated with glutamine after 3 s of assimilation distilled as amide nitrogen (13). The second major radioactive product of assimilation of N H was glutamate in all species examined. In all cases the fraction of label in glutamine decreased and that in glutamate increased with longer incubation periods. There were some species-dependent variations in the extent of the shift in the fractions of N in glutamine and glutamate, but only in the case of A. variabilis did the fraction of label in glutamate exceed that in glutamine within 120 s of incubation, as occurred during the fixation of [ N ] N (compare Figures 5 and 7). Aspartate was detectably radioactive in all the species as of 120 or 900 s of assimilation and, in A . cylindrica, was equal to the radioactivity in glutamate after 900 s of assimilation. Alanine was detectably radio active within 1 s of assimilation by A. variabilis, C. licheniforme, and A. nidulans and within 15 s of assimilation by Gleoethece sp., but it was 13

2

1 3

1 3


13

4

+

2

1 3

1 3

1 3

4

+

2

4

4

+

+

2

1 3

4

+

1 3

1 3

1 3

4

+

1 3

1 3

1 3

4

+

1 3

13


2


14.

MEEKS

279


Assimilation time ( s )

Figure 7. Time course of the distribution of N in amino acids extracted after assimilation of NHj+ for 1,15,120, and 900 s by (A) A. cylindrica, (B) A. variabilis, (C) C. licheniforme, (D) P. boryanum, (E) Cloeothece sp., and (F) A. nidulans (12,14). The radioactivity of amino acids in metha nolic extracts, subjected to electrophoresis, was quantitated by integration of peaks in radioscans, with corrections applied for decay: (A) glutamine, (O) glutamate plus aspartate, (X) citrulline plus alanine, (D) arginine plus other compounds. See text for species-dependent variations of N in glutamate plus aspartate and citrulline plus alanine. 13

13

13

detected only as a shoulder on the citrulline peak after 120 and 900 s in P. boryanum and was detected only in experiments with inhibitors and high exogenous N H concentrations in A. cylindrica. Radioactivity in citrulline was detected within 360 to 900 s of assimilation by A. cylindrica, P. boryanum, and Gleoethece sp., but it was observed only as a shoulder on the alanine peak after 900 s of assimilation by A. variabilis, C. licheni forme, and A. nidulans. Low levels of radioactivity were observed associated with arginine after 900 s of assimilation, (i.e., < 7 - l l % of the total in A. variabilis and A. cylindrica, and

Radiotracer Studies of Nitrogen Metabolism in Cyanobacteria

Recommend Documents