T H E J O U R N A L OF BIOLOGICAI. CHEVIXTRY Vol. 258, No. 3. Issue of February 10, pp. 1678-1683. 1983 Printed Ln U. S. A.

Guanosine 5’-Triphosphate, 3”Diphosphate 5’-Phosphohydrolase PURIFICATION AND SUBSTRATE SPECIFICITY*

(Received for publication, July 1, 1982)

Akira Harat and Jose Sy From The Rockefeller Uniuersity, New York, New York 10021

The regulatory nucleotide guanosine 5’-diphosphate, 3”diphosphate (ppGpp) and its precursor guanosine 5’- triphosphate, 3”diphosphate (pppGpp) are accumu- lated during stringent response in bacterial cells. The enzyme pppGpp-5‘-phosphohydrolase, which catalyzes the conversion of pppGpp to ppGpp, was partially pu- rified from Escherichia coli. It has M, = 140,000 and an apparent K,,, of 0.11 m~ for pppGpp. It requires M&+ and a monovalent cation. NK+ is preferred over K+, while Na+ is inactive. The enzyme does not hydrolyze GTP, ATP, pppApp, or ppGpp. It is also not effectively inhibited by these nucleotides. pppGpp-fi’-phosphohy- drolase hydrolyzes the 3’-monophosphate analog pppGp equally well (apparent K,,, of 0.13 m), yielding the recently identified MS I11 nucleotide (ppGp). pppGpp-5’-phosphohydrolase does not have RNA 5’- terminal y-phosphatase activity; however, 5’-terminal phosphates are released by pppGpp-5’-phosphohydro- lase when the GTP-terminated RNA chains are first converted into oligonucleotides by RNase A treatment. pppGpp-5’-phosphohydrolase was found to actively hy- drolyze the dinucleotide fragment pppGpNp but ex- hibited very low activity toward longer chain frag- ments. The 3’-unphosphorylated dinucleotide pppGpN was, however, not hydrolyzed. The ability of pppGpp- 5’-phosphohydrolase to hydrolyze pppGpp, pppGp, and pppGpNp, but not pppG and pppGpN, indicates that pppGpp-5’-phosphohydrolase is rather nonspecific to- ward the 3’-OH substitutions of the substrates although a free, unsubstituted phosphate group at the 3”OH position is essential.

~

ppGpp‘ is a regulatory nucleotide found in all prokaryotic organisms tested so far (1). It functions as a signal during nutritional stresses to regulate the transcriptions of various genetic units as well as to control the rate of several key enzymes in various metabolic pathways (2-4). Eight genetic loci are currently known to be involved in the rapid accumu- lation and turnover of this nucleotide. The interconversion of guanosine nucleotides in the metabolic cycle of ppGpp is formed by a series of phosphoryl and pyrophosphoryl addi- tions and hydrolysis reactions

* This work was supported by United States Public Health Service Grant AI17080. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address, Meijo University, Nagoya, Japan. ‘ The abbreviations used are: ppGpp, guanosine 5’-diphosphate, 3”

diphosphate; pppGpp, guanosine 5’-triphosphate, ?-diphosphate; ppGp, guanosine 5’-diphosphate, 3’-monophosphate; pppGp, Guano- sine 5’-triphosphate, 3”monophosphate; ;, a 32P label in that position, PP,, inorganic pyrophosphate; PEI, polyethylenimine.

GDP ”.& GTP pppGpp 3- ppGpp GDP +P +PP. -P, -PP 1 2

In vivo studies have indicated that the regulation of the ppGpp metabolic cycle is at the two reactions involving py- rophosphates (reactions 2 and 4) . ppGpp and its precursor pppGpp are synthesized very rapidly during stringent re- sponse where restriction of stable RNA accumulation follows the starvation for a required amino acid. The enzyme catalyz- ing the synthesis of these nucleotides during stringent re- sponse is the translation product of the relA gene. The activity of this is greatly stimulated by the presence of a ribosome- mRNA complex containing a codon-specified uncharged tRNA. In addition to the ribosome-dependent enzyme, several bacilli have been found to contain a second enzyme which catalyzes a ribosome-independent synthesis of (p)ppGpp (5, 6). Both enzymes carry out a pyrophosphoryl transfer reaction with ATP as the donor and GTP or GDP as the acceptors. Since GTP is present in vivo a t a 20-fold higher concentration than GDP, pppGpp is assumed to be the primary product of these enzymes (7). The rapid in vivo turnover of ppGpp (reaction 4) is catalyzed by a Mn2’-dependent enzyme that releases the 3‘-pyrophosphate group as a unit (8, 9). This enzyme is the translation product of the spoT gene and its in vivo activity is linked to the availability of an energy source. An in vitro ATP-dependent system that stimulates the deg- radation of ppGpp has recently been discovered (10); however, the exact mechanism of this control is currently unknown. The conversion of pppGpp to ppGpp (reaction 3) was found to be a very rapid reaction (7,11,12). The half-life of pppGpp was variously estimated to be 6-13 s and, in contrast to the half-life of ppGpp, was not found to be appreciably affected by nutritional manipulations. Under in vitro conditions, elon- gation factors T and G have been shown to catalyze a ribo- some-dependent hydrolysis of pppGpp to ppGpp (13). The in vivo significance of these reactions is, however, uncertain. Recently, Somerville and Ahmed (14) reported a new class of mutants (gpp) with a much higher in vivo pppGpp to ppGpp ratio than normal cells and have found that the mutant cell- free extracts, when chromatographed on DEAE-cellulose, have a missing pppGpp hydrolytic activity. They showed that the pppGpp hydrolytic enzyme is probably a pppGpp-5‘-phos- phohydrolase since pppGpp was converted into a nucleotide chromatographically similar to ppGpp. The enzyme was, how- ever, not further characterized. We present here the purifica- tion and characterization of this enzyme and report here its substrate specificity.

EXPERIMENTAL PROCEDURES

Materials Escherichia coli JF1599 (thr, leu, his, argH, pyrE, spoT, thi, pGA1

(Tc’, pyrE+, SPOT’)) obtained from James D. Friesen (University of Toronto, Toronto, Canada), was grown in yeast extract, glucose, and

1678

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from

pppGpp-5'-phosphohydrolase 1679

phosphate medium (15) supplemented with the required nutrients. Ten-liter cultures were grown in a New Brunswick fermentor a t 37 "C and the cells were harvested when A m ,,,,, reached 4 to 5. CS411 (F-, leu, ara, proC, lacZ, hisF, cy&, str, spc, pyrE, metE, thi) and CS419 (same genotype as CS411 except met' and gpp-) obtained from Jonathan Gallant (University of Washington, Seattle), were grown in 1 liter of medium a t 37 "C in a new Brunswick rotatory shaking incubator and were harvested after an overnight incubation.

were prepared as previously descnbed (16). pppGp and ppGp were obtained from P-L Biochemicals and [Y-~'P]GTP was from New England Nuclear. E. coli RNA polymerase (type III), bovine pan- creatic RNase A (type 11-A), RNase T2, electrophoretically purified DNase, and yeast inorganic pyrophosphatase were purchased from Sigma Chemical Co. 5"[y-"PIpppGpp and 5"[y-32P]pppGp were pre- pared respectively by phosphorylation of ppGpp and ppGp with the coupled glyceraldehyde-3-phosphate dehydrogenase/3-phosphoglyc- erate kinase system and 32P, (17). Boric acid gel (0.1-0.4 mm) was obtained from Aldrich Chemical Co. and nucleotides were chromat- ographed in boric acid gel as described (8).

[3HlpppGpp, 3"[P-32PlpppG~~, and unlabeled PPPGPP and PPGPP

Methods

pppGpp-5'-phosphohydrolase Assay-Standard pppGpp-5"phos- phohydrolase assays were done in 25 pl containing 40 m~ Tris-OAc, pH 8.1, 1 m~ dithiothreitol, 10 mM Mg(OAc)z, 30 m~ NH~OAC, 0.2 mM EDTA, 0.3 mM [3H]pppGpp (1.2 X IO4 cpm). The reactions were carried out at 30 "C for 30 min and stopped by the addition of 1 p1 of HCOOH. The mixture was centrifuged and 5 p1 of the supernatant were applied to polyethylenimine-cellulose thin layer sheets. The hydrolytic products were resolved with 1.5 M KH2P04, pH 3.5. UV- absorbing areas corresponding to pppGpp and ppGpp were cut out, eluted a t room temperature with 0.35 ml of 1 M triethylammonium- HC03, pH 9, and counted with Hydrofluor.

One unit of enzyme was defined as the amount which converted 1 nmol of pppGpp into ppGpp per min under the above conditions. Specific activity was designated as the enzyme unit/mg of protein.

Purification ofpppGpp-5'-phosphohydrolase-Fifty g of E. coli J F 1599 were suspended with 100 ml of buffer (10 mM Tris-OAc, pH 7.8, 14 m Mg(OAc)?, 60 mM KOAc, 0.5 m~ EDTA, 1 m~ dithiothreitol) containing DNase at 1 pg/ml and disrupted by a French press a t 18,000 p.s.i. After centrifugation at 30,000 X g for 30 min, the super- natant (132 m l ) was collected and 15 g of (NH4)&304 were added. After standing for 30 min, the mixture was centrifuged at 10,OOO X g for 10 min and 21.2 g of (NH4)2S04 were added to the supernatant. The second ammonium sulfate precipitates were resuspended in 15 ml of buffer A (20 m~ Tris-HC1, pH 7.5, 2 mM Mg(OAc)a, 100 mM KC1,l m~ dithiothreitol) and dialyzed against buffer A. The dialysate was centrifuged at 250,000 X g for 60 min. The resulting supernatant (60 ml) was applied to a DEAE-cellulose column (2 X 28 cm) previ- ously equilibrated with buffer A. The column was washed with 50 ml of the same buffer and the proteins were eluted with a linear 0.10- 0.45 M KC1 gradient (total volume of 500 ml) in buffer A. The active fractions (90 ml) were pooled, dialyzed against 80 m~ phosphate buffer, pH 6.8, and applied to a column (2 X 7 cm) of hydroxylapatite (Bio-Rad) previously equilibrated with 80 m~ phosphate buffer, pH 6.8. The column was washed with 50 ml of the same buffer and the proteins were eluted with 200 m phosphate buffer. The active fractions (18.8 ml) were collected and dialyzed against 40 mM am- monium acetate buffer, pH 5.8, containing 2 n" Mg(OAc)z, 1 mM dithiothreitol. The precipitate produced during dialysis was removed by centrifugation at 10,000 X g for 10 min and the supernatant was again dialyzed against buffer A. The cleared dialysate was applied to a column (1.5 X 27 cm) of DEAE-Sephadex A-50 previously equili- brated with buffer A. The column was washed with 20 ml of buffer A and the proteins were eluted with a linear 0.10-0.40 M KC1 gradient (total volume of 300 ml) in buffer A. The active fractions (27.0 ml) were collected, concentrated to about 2 ml, and passed through a column (1.7 X 115 cm) of Sephadex G-150 previously equilibrated with buffer A. The proteins were eluted with the same buffer. The highly purified fractions obtained were concentrated and stored in liquid nitrogen until use.

Preparation of [y-32PJGTP-terminated RNA y-"'P-labeled RNA was synthesized by the method of Burgess (19)

using E. coli RNA polymerase. One ml of reaction mixture containing 40 mM Tris-OAc, pH 7.8, 10 n" Mg(OAc)z, 0.1 mM EDTA, 0.1 mM dithiothreitol, 0.4 mM potassium phosphate, pH 7.8, 150 mM KC1,0.2 n" UTP, 0.2 mM CTP, 0.2 mM ATP, 0.2 m~ [y-3ZP]GTP (0.34 mCi),

50 units of RNA polymerase, and 0.15 mg of calf thymus DNA was incubated at 37 "C for 1 h. The reaction was stopped by the addition of perchloric acid and the [y-32P]GTP-terminated RNA formed was purified by the method of Maitra and Hurwitz (20).

Preparation of [y-3ZP]GTP-terminated Oligonucleotides [y-32P]GTP-terminated RNA (2.5 x IO' cpm) was exhaustively

digested with RNase A (2 mg) in 100 m~ Tris-HCl, pH 7.0, a t 30 "C for 30 min. A portion of the digest (4.6 X lo5 cpm) was then applied to a column (0.6 X 23 cm) of DEAE-Sephadex A-25 previously equilibrated with 20 mM Tris-HC1, pH 7.5,0.05 M NaC1,7 M urea. The column was washed with the starting buffer and the radioactive materials were eluted with a linear 0.05-0.4 M NaCl gradient (total 200 m l ) in Tris-urea buffer. The radioactive fractions were pooled and desalted by passing through a column of DEAE-cellulose previously equilibrated with triethylammonium bicarbonate.

Preparation of [y-'~'PJGTP-tt;rminated Dinucleotides (ippGpN

GppGpN was prepared by incubating E . coli RNA polymerase with DNA template in the presence of CTP, UTP, and a lower concentra- tion of GTP but omitting ATP. Under these conditions, a large amount of pppGpN was synthesized. Any 1ong:r chain polynucleo- tides made were then digested by RNase A to pppGpNp. One ml of reaction mixture containing 40 n" Tris-OAc, pH 7.8, 10 mM Mg(OAc)z, 0.1 mM EDTA, 0.1 mM dithiothreitol, 0.4 mM potassium phosphate, pH 7.8, 150 mM KC1,0.2 mM CTP, 0.2 mM UTP, 0.013 mM [y3'-P]GTP (0.13 mCi), 0.5 unit of inorganic pyrophosphatase, 30 units of RNA polymerase, and 0.15 mg of calf thymus DNA was incubated a t 37 "C for 1 h. The mixture was heated in a boiling water bath for 10 min and subsequently chilled in ice. Two hundred pl of 1%

30 min at 37 "C. The RNase A digest was diluted with 5 ml of 8 M RNase A were then added and the mixture was further incubated for

previously equilibrated with 20 mM Tris-HC1, pH 7.5,0.05 M NaCl, 7 urea and applied to a column (0.6 X 23 cm) of DEAE-Sephadex A-25

M urea. Radioactive dinucleotides were then separated as described in the preparation of [y-"PIGTP-terminated oligonucleotides. Ten pCi of GppGpN and 1.4 pCi of 6ppGpNp were obtained by this procedure.

and PPPGPNP)

Miscellaneous Methods

Nucleotide hydrolysis was determined as described for pppGpp-5'- phosphohydrolase assay except that [3H]GMP, ['HIGDP, [y-"'P] GTP, [y-32P]ATP, or [32P]PPi were used as substrate. For GTPase, ATPase, and inorganic pyrophosphatase assays, the polyethyleni- mine-cellulose plates were developed with 0.75 M KHZPO4, pH 3.5, and for GMP and GDP hydrolysis, with 0.5 M, KH2P04, pH 3.5. Protein concentrations were measured by the Bradford protein dye- binding method (18).

RESULTS

Table I summarizes the purification of pppGpp-5"phospho- hydrolase. The enzyme was purified approximately 370-fold; however, this is an underestimation since the crude extract of E. coli contains several nonspecific GTPases which can hy- drolyze pppGpp (13, 14). The final enzyme preparation was

TABLE I Summary of pppGpp-5'-phosphohydrolase purification . ~~~ ~~ . .

Vol- Protein Total ac- Specific ac- ume tivity" tivity"

ml mg units Crude extract 132 5,196 (21,219) (4.08) (NH4)zS04 fraction 60 1,634 14,175 8.68 DEAE-cellulose 99 316 11,756 37.2

Hydroxylapatite 18.8 80.4 6,049 75.2

Dialysis at pH 5.8 21.1 18.5 4,373 236 DEAE-Sephadex A-50 27.0 3.76 1,748 465

Sephadex G-150 21.7 0.875 1,324 1,513

fraction

fraction

fraction

The total activity and specific activity of the crude extract are in parentheses to indicate that these activities include those contributed by other pppGpp hydrolytic enzymes.

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from

1680 pppGpp-5’-phosphohydrolase

found to be free of GTPase and ATPase activities. I t is rather unstable a t 4 “C but can be stored in liquid N2. An apparent M , = 140,000 for pppGpp-5“phosphohydrolase was found by gel filtration (Fig. 1). Although the enzymic activities coin- cided fairly well with the protein profile in gel filtration, the enzyme preparation was not electrophoretically pure and, on sodium dodecyl sulfate-polyacrylamide gel electrophoresis, contained four major bands and several minor bands. A pH optimum of 9 was found using either Tris-acetate or borate- KCl-NaOH buffers with the borate buffer showing 50% more activity at the pH optimum. An apparent K, of 0.11 mM for pppGpp was found under our assay condition. pppGpp-5‘- phosphohydrolase activities were determined in extracts ob- tained from CS411 (gpp’) and CS419 (gpp-). The specific activity of pppGpp-5’-phosphohydrolase at the (NH4)2S04 purification step (Table I) was 12.3 units/mg of protein for CS411 (gpp’), a value similar to that obtained with the JF1599 strain, while that of the gpp mutant CS419 was t0.8 unit/mg of protein. This result indicates that the pppGpp-5’-phospho- hydrolase activity purified and characterized here is the gpp gene product.

Ionic Reguirements-pppGpp-5‘-phosphohydrolase has an absolute requirement for Mg2+ (Fig. 2A ). The optimum Mg” concentration was above 1 m. In addition to the Mg” requirement, the enzyme also has a requirement for a mono- valent cation (Fig. 2B). NH,’ was the most effective, while K+ was a poor substitute and Na’ was ineffective.

Hydrolysis of pppGpp-Somerville and Ahmed (14), in their identification of pppGpp-5‘-phosphohydrolase, showed the hydrolysis of 3’-[a-”’P]pppGpp into a nucleotide that was chromatographically indistinguishable from ppGpp in the thin layer chromatographic system of Cashel and Kalbacher (21) and, therefore, suggested that the enzyme is a 5’-phospho- monohydrolase. However, a 3’-monophosphate hydrolysis cannot be completely ruled out under this condition since pppG$, the product of a 3’-monophosphate hydrolysis, has a mobility similar to that of ppGpp in this thin layer chromat- ographic system. To distinguish the two modes of hydrolysis, we have prepared pppGpp with ‘lP label separately at the 5” y ($ppGpp) and at the 3‘4 positions (pppGp$). Hydrolysis of the pppGpp yielded inorganic ”‘Pi while hy$olysis of the pppGpp yielded the expected labeled ppGpp (Fig. 3). No further hydrolysis ofppGp5 was found. The spa“ amount a2Pi detected during the hydrolysis of thepppGpp was due to the

2.5

2.0 ’ >” 1 \ pppGpyose I .\ I

1.5 t \ 5

1 .o- 4.0 4.5 5.0 5.5

log M.W. FIG. 1. Molecular weight of pppGpp-5’-phosphohydrolase.

The molecular weight of pppGpp-5”phosphohydrolase (pppGpp-use) was estimated by Sephadex G-150 ( 1 . i X 115 cm) column chromatog- raphy. pppGpp-5”phosphohydrolase was assayed as described under “Methods.” Protein molecular weight standards (1, cytochrome c; 2, chymotrypsinogen; 3. egg albumin; 4, bovine serum albumin; 5, cata- lase) were chromatographed in the same column in a parallel run.

4or A

101- 1 . I I I I I I

0 1 2 3 4 5 6 mM Mq2+

40r B

30t 0

Salt (mM) FIG. 2. M%+ and monovalent cation requirements of

pppGpp-5’-phosphohydrolase. Sephadex G-150-purified enzyme which was dialyzed against buffer A minus M e and NH + Ions ’ was used in these assays. Standard pppGpp-5”phosphohydrolase assays were done as described under “Methods” using 0.18 pg of the dialyzed enzyme per assay. A, Mg’+ ion concentrations were varied as indi- cated. B, NHI’ (O), K+(O), and Na’ (X) concentrations were varied as indicated.

t Pi

t PPi

+ PPGPP

PPPGPP

1 2 3 4 5 FIG. 3. Hydrolysis of CppGpp and pppGpIf. Twenty-five pl of

standard assay mixture containing 0.15 unit of pppGpp-5’-phospho- hydrolase were incybated with 6ppGpp (150,000 cpm/reaction), lanes 1 and 2, or pppGpp (68,000 cpm/reaction), lanes 3 and 4. At 0 time (lanes 1 and 3) and after 30 rnin (lanes2 and 4 ) at 30 “C, the reactions were stopped by HCOOH addition and 5 pI were applied to PEI- cellulose plates. The chromatograms were developed with 1.5 M KHzPO,, pH 3.5, and radioautographed. Lane 5 contained ”Pi and “‘PPi standards.

presence of some 5’-[y-32P]pppGpp. Some exchange of “P label occurred during the preparation of these compounds using the stringent factor-ribosome-mRNA-uncharged tRNA system (22) between [y-”PIATP and GTP. The result pre- sented in Fig. 3 c o n f i s and proves that pppGpp-5”phospho- hydrolase is indeed a 5’-phosphomonohydrolase.

Substrate Specificity-The substrate specificity of pppGpp- 5‘-phosphohydrolase on mononucleotides was tested. The en- zyme did not hydrolyze [“HIGDP, [“HIGMP, [y-‘”PIGTP, [ y

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from

pppGpp-5'-phosphohydrolase 1681

"PIATP, or ["'PIPPi. pppApp and ppApp, the adenosine analogs of the guanosine 5',3"polyphosphates, were also not hydrolyzed (data not shown). The ability of these nucleotides to inhibit the activity of pppGpp-5'-phosphohydrolase was also tested (Table 11). Except for the slight inhibition shown by ppGpp and pppApp, all nucleotides tested, as well as PPi and Pi, were ineffective as inhibitors when used a t a 10-fold higher concentration than pppGpp.

The enzyme was found to readily hydrolyze6ppGp. This activity required a monovalent cation with NH4' preferred over K' (Table 111). The enzyme has an apparent K,,, of 0.13

TABLE I1 Effect of various compounds on pppGpp-5'-phosphohydrolase

activity Standard assay mixtures containing 0.1 unit of pppGpp-5'-phos-

phohydrolase were used with the indicated compound added a t 3.0 m. 49% of pppGpp was hydrolyzed in the control reaction and this value was normalized as 100%.

Compound added Activity I

None 100 PPGPP 87 GTP 101 GDP 105 GMP 119 ATP 110 PPPAPP 92 PPAPP 99 PP, 102 P, 111

mM for pppGp, a value very similar to pppGpp. The ability of the enzyme to rapidly hydrolyze pppGp suggests that the specificity at the 3' end of the substrate may not be very stringent and that compounds containing other substitutions at the 3'-phosphoryl group may also serve as substrate. Since the 5'-terminal nucleotide of an RNA chain starting with a guanosine residue (pppGp(Np),,) can be considered as an analog of pppGpp, it was interesting to find out if pppGpp-5'- phosphohydrolase would hydrolyze the 5"terminal phosphate of an RNA chain.

The Hydrolysis of pppGpNp-RNA chains with 5'-[y-"'Pp] GTP termini were synthesized by using E. coli RNA polym- erase and calf thymus DNA as template. This substrate was

TABLE 111 The hydrolysis of pppGp by pppGpp-5'-phosphohydrolase

Standard assay mixtures were used except that ippGp (0.3 mM, 25.000 cpm) was used as the substrate and NH,' and K' ion concen- trations were varied as indicated. Two-tenths unit of pppGpp-5'- phosphohydrolase were used per assay.

Additions pppGp hydrolyzed rn .u Q

NH.,' 0 5 1 75 5 91

0 5 1 18 5 42

K'

B

A

0 IO 20 30

Minutes 1 2 3 4 5 FIG. 4. pppGpp-5'-phosphohydrolase hydrolysis of RNase A-digestzd [y-32P]GTP-terminated RNA. A,

time course of hydrolysis. Fifty pI of standard reaction mixtures containing pppGpp (96,000 cpm; 0) or RNase A- digested [y-32P]GTP-terminated RNA (48,000 cpm; X) were incubated with 1.8 units of pppGp-5'-phosphohydrolase at 30 "C. Aliquots of 5 p1 were withdrawn a t 5-min intervals, mixed with 1 pl of 206 HCOOH, and then chromatographed on PEI-cellulose thin layer sheets. '"Pi formed was resolved from the nucleotides by developing in 0.5 M KHzPO~, pH 3.5. Per cent hydrolysis corresponds to the percentage of input counts hydrolyzed to '"P,. B, 25 pl of standard reaction mixtures containing RNase A-digested [y-'"PIGTP-terminated RNA (30,000 cpm) were incubated a t 30 "C for 30 min with I , no enzyme; 2, 10 units of RNase Tz; 3, 10 units of RNase T? and 0.9 unit of pppGpp-5"phosphohydrolase; and 4,O.g unit of pppGpp-5'-phosphohydrolase. One pl of HCOOH was added at the end of incubation and 5 p1 of each sample were applied to PEI-cellulose thin laye! sheets. After devL %ng with 1.0 M KHzP04, pH 3.5, the plate was radioautographed. Lane 5 contained standard pppGp.

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from

1682 pppGpp-5’-phosphohydrolase

found not to be hydrolyzed by pppGpp-5‘-phosphohydrolase (data not shown). However, hydrolysis of the y-”>P label into Pi was found when the RNA chains were first cleaved by

RNase A (Fig. 4A) into oligonucleotides. The phosphatase reaction was dependent on the presence of NH,’ and was completely inhibited by the presence of 1 mM pppGpp, indi- cating that pppGpp-5’-phosphohydrolase was the enzyme re- sponsible for this reaction. The rate of hydrolyzing the oligo- nucleotide termini was considerably slower than that of pppGpp hydrolysis. The oligonucleotide termini obtained by RNase A digestion showed, on PEI-cellulose chromatography, three major spots with trailing of radioactivity toward the origin (Fig. 4B, lane 1). The major spots were sensitive to pppGpp-5’-phosphohydrolase digestion while the trailing ma- terials were not digested (Fig. 4B, lane 4). All the radioactive material was converted to ;ppGp when the RNase A digest RNA chains were further treated with RNase T2, indicating that all ”P labels were at the 5’ termini. The resulting ;ppGp was completely hydrolyzed by pppGpp-5’-phospho- hydrolase. The presence in the chromatogram of trailing materials that were resistant to pppGpp-5’-phosphohydrolase indicated that long chain oligonucleotides were not hydrolyzed by pppGpp-5’-phosphohydrolase. To determine the chain length of the oligonucleotides sensitive to pppGpp-5’-phos- phohydrolase, the RNase A-digested fraction was chromato- graphed on a DEAE-cellulose column in the presence of 7 M urea (Fig. 5A). The majority of the phosphate-labeled nucleo- tides were found to have a -7 charge. There was also a minor amount of -8 and higher charged nucleotides. The -7 and -8 charged nucleotides were found not to bind to a boric acid gel column, indicating substitution at the 3‘ end. Therefore, it is concluded t 9 t the -7 and*-8 charged nucleotides have the structures of pppGpNp and pppGp(Np),, respectively. When

32

.. a u

E E

0 20 40 60 80 Fractlon number

80

Mlnutes

FIG. 5. The separation and hydrolysis of [y-32P]GTP-termi- nated oligonucleotides. A, the 32P-labeled oligonucleotides were separated in a column of DEAE-Sephadex A-25 (0.6 X 23 cm) con- taining 7 M urea as described under “Methods.” GMP, GDP, GTP, guanosine 5’-tetraphosphate, ppGpp, and pppGpp were used as stand- ards. Radioactive fractions* corresponding to the -7 charge (6ppGpNp) and -8 charge (pppGp(Np)~) nucleotides were pooled, desalted, and lyophilized. B, standard reaction mixtures in 100 p1 containing ether 5ppGpNp (44,000 cpm; O), pppGp(Np)~ (35,000 cpm; X), or pppGpp (38,000 cpm; A) as substrates were incubated with 7.2 units of pppGpp-5’-phosphohydrolase at 30 “C. Five-pl ali- quots were withdrawn every 5 min and mixed with 1 p1 of 20% HCOOH. Five pl of the mixture were then chromatographed on PEI- cellulose developing with 0.5 M KHzPO,, pH 3.5. Per cent hydrolysis corresponds to percentage of input counts hydrolyzed to P,.

2ol l o t

- =L \ € 4 f Q

” b31 X 2

-2 J.

1 :: A

0 2 0 40 60 80 Fraction number

P

ICLL. 2 0 0 10 20 30

Minutes FIG. 6. The separation and hydrolysis of [y-32P]GTP-termi-

nated dinucleotides. A, [y3’P]GTP-terminated dinucleotides were synthesized and were then chromatographed in a column of 7 M urea- DEAE-Sephadex A-25 as described under “Methods.” The same guanine nucleotide mixtures as in Fig. 5A were u:ed as standards. Radioactive fractions corresponding to -5 charge (pppGpN) and -7 charge (;ppGpNp) nucleotides were pooled, desalted, and lyophilized. B, 100 p1 of standard reaction mixtures containing 10 m~ Tris-HC1, PH 7.5, instead of Tris-*OAc, pH 8.1, were incubaied with either PPPGPN (134,000 cpm), PPPGPNP (91,000 cpm), OrpppGpp (43,000 cpm) at 30 “C. The assay mixtures contained 2.4 units of pppGpp-5’- phosphohydrolase and/or 20 units of RNase T2. Five-pl aliquots were withdrawn and assayed for hydrolysis by PEI-cellulose chromatog- raphy as in Fig. 5B. Per cent hydrolysiszorresponds to percentage of radioactive counts released as 32P,. 0, pppGpN + pppGpp-5‘-phos- phohydrolase; A, 5ppGpN + RNase T z ; 0, PPPGPN + PPPGPP-~” phosphohydrolase + RNase TP; a, pppGpNp ,+ pppGpp-5”phospho- hydrolase; A, 6ppGpNp + RNase Tz; B, pppGpNp + pppGpp- 5’-phosphohydrolase + RNase TO; X, CppGpp + pppGpp-5”phospho- hydrolase.

subjected to pppGpp-5’-phosphohydrolase, the dinucleotide gppGpNp was found to be hydrolyzed and the trinucleotide pppGp(Np)~ was much more resistant (Fig. 5B).

Since the rate of pppGpNp hydrolysis was only Y7 that of pppGpp or pppGp hydrolysis and, furthermore, the hydrolysis rate rapidly declined with an increase in chain length of the nucleotide, it was interesting to determine whether the nu- cleotide pppGpN containing no 3’ substitution at the N nu- cleotide might be a better substrate for pppGpp-5”phospho- hydrolase. pppGpN was prepared by incubating E. coli RNA polymerase and calf thymus DNA with CTP, UTP, and [ y - ”PIGTP, but in the absence of ATP. The products of polym- erase reaction were chromatographed directly on a 7 M urea- DEAE-cellulose column (Fig. 6A). Two products having -5

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from

pppGpp-5‘-phosphohydrolase 1683

and -7 charges were found and both products were converted to CppGp when further digested by RNase Tz. The -7 charged nucleotides wre found not to bind to$he boric acid gel column. They are, therefore, similar to the pppGpNp fraction of Fig. 5A. The -5 charged nucleotides were found to bind to the boric acid gel column, indicating that they have unsubstituted 2’,3’-OH*groups. The -5 charged nucleotides should, there- fore, be pppGpN.

Fig. 6B shows that $ppGpNp was hydr2lyzed by pppGpp- 5’-phosphohydrolase, but surprisingly, pppGpN was not. 6ppGpN became sensitive to* pppGpp-5’-phosphohydrolase only after it was converted to pppGp by RNase Tz treatment (Fig. 6B).

DISCUSSION

Our results confirmed the presence of a specific pppGpp-5’- phosphohydrolase that is the gene product of gpp (14). The enzyme does not hydrolyze GTP, ATP, or pppApp nor is it inhibited by these nucleotides. These results suggest that, unlike the elongation factors, pppGpp-5’-phosphohydrolase is probably not a nonspecific GTPase or other similar type of enzyme. The enzyme has an apparent K, of 0.11 mM for pppGpp and requires both Mg2+ and NH4+ for maximal activ- ity. It hydrolyzes the 3’-monophosphate analog pppGp equally well with an apparent Km of 0.13 mM. This indicates that pppGpp-5”phosphohydrolase is not specific toward the num- ber of phosphate groups at the 3”OH position and, indeed, the enzyme is active toward pppGpNp, where the bulky pNp group replaced the 3’-pyrophosphate moiety. Furthermore, the ability of pppGpp-5’-phosphohydrolase to hydrolyze pppGp and pppGpNp but not pppG and pppGpN indicates that a free, unsubstituted phosphate group at the 3’-OH position is essential for activity. It is interesting to note that the two enzymes in the ppGpp metabolic cycle which act on the 3’-OH position, the ppGpp-3’-pyrophosphohydrolase and ATP:GTP (GDP)-3’-pyrophosphoryltransferase (ppGpp-syn- thetase), have an analogous nonspecificity except that they are nonspecific toward the 5’-OH substitutions. For example, the ribosome-dependent ppGpp-synthetase will pyrophos- phorylate GMP, guanosine 5’-tetraphosphate, although not as well as when GTP or GDP are substrates (17). The ribosome- independent ppGpp-synthetase of Bacillus brevis can also pyrophosphorylate GDP-glucose.’ ppGpp-3’-Pyrophospho- hydrolase hydrolyzes pppGpp, ppGpp, and pGpp equally well (2).

The ability of pppGpp-5’-phosphohydrolase to readily hy- drolyze both pppGp and pppGpNp raises the possibility that these nucleotides may be present under in vivo conditions. Both pppGp and pppGpNp can be considered as Y-terminal fragments of GTP starting mRNAs. Therefore, these frag- ments may be released by RNase action during the rapid turnover of mRNA in bacterial cells. In the absence of the stringent response, PppGpp-5”phosphohydrolase could then

* J. Sy, unpublished observation.

be involved in the normal Catabolism of the 5”terminal frag- ments.

The catabolism of RNA may also be the source of the recently identified MS I11 nucleotide ppGp (23), which was shown to be a possible RNA synthesis regulator (24). The nucleotide ppGp may be released by RNase from GTP- terminated RNA fist as pppGp which is later hydro- lyzed by pppGpp-5’-phosphohydrolase to ppGp: pppGp(Np), - ( N P ) ~ + PPPGP PPGP.

RNase pppGpp-5”phosphohydrolase

An interesting parallel in substrate specificity is shown between elongation factor G (13,20) and pppGpp-5’-phospho- hydrolase. Both enzymes hydrolyze pppGpp, pppGp, and pppGpNp very well. The two enzymes differ significantly in that elongation factor G catalyzes the hydrolysis of GTP and requires ribosomes for activity, while pppGpp-5’-phosphohy- drolase does not have the ribosome requirement, nor does it hydrolyze GTP.

REFERENCES I. Silverman, R. H., and Atherly, A. G. (1979) Microbiol. Rev. 43,

27-41 2. Sy, J. (1980) in Molecular Biology and Biophysics (Chapeville,

F., and Haenni, A. L., eds) Vol. 32, pp. 197-204, Springer-Verlag, New York

3. Gallant, J. (1979) Annu. Rev. Genet. 13, 393-415 4. Richter, D. (1980) in Ribosomes (Chambliss, G., Craven, G. R.,

Davies, J., Davies, K., Kahen, L., and Nomura, M., eds) pp. 743-765, University Park Press, Baltimore

5. Sy, J. (1979) in Regulation of Macromolecular Synthesis by Lou) Molecular Weight Mediators (Koch, G., and Richter, D., eds) pp. 95-106, Academic Press, New York

6. Fehr, S., and Richter, D. (1981) J. Bacteriol. 145,68-73 7. Kari, C., Torok, I, and Travers, A. (1977) Mol. Gen. Genet. 150,

8. Sy, J. (1977) Proc. Natl. Acad. Sei. U. S. A. 74, 5529-5533 9. Heinemeyer, E. A,, and Richter, D. (1978) Proc. Natl. Acad. Sci.

249-255

U. 5’. A . 75,4180-4183 10. Sy, J. (1980) J. Biol. Chem. 255, 10056-10059 11. Fiil, N. P., Willumsen, B. M., Friesen, J. D., and von Meyenburg,

K. (1977) Mol. Gen. Genet. 150,87-I01 12. Weyer, W. J., de Boer, H. A,, de Boer, J. G., and Gruber, M.

(1976) Biochim. Biophys. Acta 442, 123-127 13. Hamel, E., and Cashel, M. (1973) Proc. Natl. Acad. Sei. U. S. A.

70,3250-3254 14. Somerville, C. R., and Ahmed, A. (1979) Mol. Gen. Genet. 141,

277-284 15. Zubay, G., Chambers, D. A., and Cheong, L. C. (1970) in The

Lactose Operon (Beckwith, J. R., and Zipser, D., eds) pp. 375- 391, Cold Spring Harbor Laboratory, Cold Spring Harbor NY

16. Sy, J. (1974) Proc. Natl. Acad. Sci. U. S. A . 71, 3470-3473 17. Sy, J. (1975) Biochemistry 14,970-973 18. Bradford, M. (1976) Anal. Biochem. 72, 248-254 19. Burgess, R. R. (1969) J. Biol. Chem. 244, 6160-6167 20. Maitra, U., and Hurwitz, J. (1973) J. Biol. Chem. 248, 3893-3903 21. Cashel, M., and Kalbacher, B. (1970) J. Biol. Chem. 245, 2309-

22. Sy, J., and Lipmann, F. (1973) Proc. Natl. Acad. Sci. U. S. A . 70,

23. Pao, C. C., and Gallant, J. (1979) J. Biol. Chem. 254, 688-692 24. Pao, C. C., Dennis, P. P., and Gallant, J. A. (1980) J. Biol. Chem.

2318

306-309

255, 1830-1833

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from

A Hara and J Sysubstrate specificity.

Guanosine 5'-triphosphate, 3'-diphosphate 5'-phosphohydrolase. Purification and

1983, 258:1678-1683.J. Biol. Chem. 

  http://www.jbc.org/content/258/3/1678Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/258/3/1678.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on July 10, 2020http://w

ww

.jbc.org/D

ownloaded from