Gene, 140 (1994) 91-96
0 1994 Elsevier Science B.V. All rights reserved. 0378-l 119/94/$07.00
GENE 07680
A reZkS) suppressor mutant allele of Bacillus subtilis which maps to reZA and responds only to carbon limitation
(Stringent response; (p)ppGpp; amino-triazole; amino-acid starvation; suppressor; transduction and
transformation crosses)
Michal Gropp”, Einat Eizenmana, Gad Glasera, Walied Samarraib and Rivka Rudnerb
“Department of Cellular Biochemistry, The Hebrew University-Hadassah Medical School, Jerusalem, Israel. Tel. (972-2) 758-l 73; and bDepartment oJ Biological Sciences, Hunter College of the City University of New York, New York, NY 10021. USA
Received by R.E. Yasbin: 24 May 1993; Revised/Accepted: 13 September/l4 September 1993; Received at publishers: 28 October 1993
91
SUMMARY
The histidine analog 3-amino-1,2,4-triazole (AT) was used for the selection of spontaneous AT-resistant revertants of
a relA mutant of Bacillus subtilis. One of these revertants, L3, showed a unique phenotype; it did not respond to amino
acid starvation, like the relA mutant, but it did respond to glucose starvation by the accumulation of (p)ppGpp, unlike
its parent. Genetic analysis revealed that this suppressor mutant (relA@‘) allele mapped to the relA locus at 239” on the
B. subtilis chromosome.
INTRODUCTION
Adaptation to starvation conditions is one of the basic
phenomena in bacteria. The response to aa limitation is
known as the stringent response (see review of Cashel
and Rudd, 1987), and is accompanied by increasing
(p)ppGpp levels via a relA gene-dependent mechanism.
The response of E. coli to carbon source shift-down is
also accompanied by increased (p)ppGpp levels via a
spoT gene-dependent mechanism (Xiao et al., 1991). The
relA mechanism involves activation of (p)ppGpp synthe-
tase (PSI) by high ratios of uncharged/charged tRNAs
(Rojiani et al., 1989). Although the spoT gene product is
Correspondence to: Dr. R. Rudner, Department of Biological Science,
Hunter College, 695 Park Avenue, New York, NY 10021. USA. Tel.:
(1-212) 772-5231; Fax: (1-212) 772-5225;
e-mail: [email protected]
Abbreviations: A, absorbance (1 cm); aa. amino acid(s); amG,
a-methylglucoside: AT, 3-amino-1,2,4-triazole; B., Bacillus; bp, base
pair(s); Cm, chloramphenicol; kb, kilobase or 1000 bp; MLS, macro-
lide-lincosamide-streptogramin resistance associated with Tn917; nt,
nucleotide(s); &A, gene encoding (p)ppGpp synthase; relA@‘, allele of
re/A whose phenotype is ATR; R, resistance/resistant: A, deletion.
SSDI 0378-l 119(93)E0676-5
responsible for (p)ppGpp degradation, it appears to func-
tion also as the second (p)ppGpp synthetase (PSII) (Xiao
et al., 1991, Hernandez and Bremer, 1991). In Escherichia
coli, relA mutants are unable to respond to aa limitation
while they continue to respond to carbon source limita-
tion. Like E. coli, the rel system of B. subtilis has a relA
gene responsible for the synthesis of ppGpp and a relC
gene responsible for the ribosomal protein Lll synthesis
which upon mutation confers resistance to thiostrepton
(Swanton and Edlin, 1972; Smith et al., 1978; 1980).
Unlike E. coli, the B. subtilis relA mutant strain lost the
ability to respond to both aa, as well as carbon limitations
(Nishino et al., 1979).
Resistance to the histidine analog AT was described in
both E. coli and Salmonella and found to be associated
with high cellular levels of ppGpp (Hilton et al., 1965;
Rudd et al., 1985). Both relA and relC mutants are sensi-
tive to the analog while mutations in the spoT gene
increase the resistance to AT (Rudd et al., 1985). We
demonstrated that wild-type B. subtilis is resistant to
15530 mM AT. As in E. coli, the relA mutant of the iso-
genie strain of B. subtilis failed to grow at these AT con-
centrations, even after 48 h. In this paper we describe a
92
spontaneous ATR revertant (L3) selected in a relA back-
ground in which the response to aa limitation and the
response to carbon limitation have been dissociated. This
allele was called relAcS) because its phenotype represents
a suppression of the carbon source deprivation response
in the original relA strain.
EXPERIMENTAL AND DISCUSSION
(a) Isolation and physiological properties of the relA@)
mutant
Spontaneous ATR mutants in strain IS56 (re/A, trpC2,
lys-3) were selected on lawns of cells [(l-5) x lo’] plated
on a complex minimal medium containing 15 mM AT
(Spizizen, 1958; Rudd et al., 1985). A few colonies isolated
after 48 h at 37”C, were purified and found to be ATR
like the isogenic relA+ strain (IS58). The various isolates
were examined for their ability to respond to aa and
carbon source limitation during growth in a low-
phosphate medium (Nishino et al., 1979) in the presence
of 100 @/ml of 32P followed by 30-min treatments
with either serine hydroxamate (l-2 mg/ml) or
m-methylglucoside (clmG; 1%). Most of the ATR rever-
tants were unable to accumulate ppGpp under both aa
and carbon source limitation similarly to the original
strain IS56. These revertants may represent mutants in
the pathway of the degradation of ppGpp (Rudd et al.,
1985; Belitsky and Shakulov, 1982). One mutant, L3,
differed from the others by its ability to respond to carbon
source limitation while still being unable to respond to
aa starvation (Fig. 1A). The levels of both ppGpp and
pppGpp increased within the first 5 min of exposure to
the inhibitor ctmG with a concomitant decrease in the
levels of GTP (Fig. 1A and B). The L3 revertant desig-
nated as relA@’ behaves like a suppressor mutant and
phenotypically resembles the relA mutant of E. coli. On
the other hand, the original strain IS56 seems to resemble
the ArelA, AspoT mutant of E. coli, in which requirement
for a whole set of aa was found (Xiao et al., 1991).
The aa requirements of B. subtilis strains IS56 and L3
were established in a complex minimal medium (Nishino
et al., 1979). As shown in Table I, the relA mutant
revealed phenotypic requirements for histidine, phenyl-
alanine and valine in addition to its auxotrophic require-
ments (trp, lys), whereas the L3 revertant lost most of
these partial aa requirements, except for phenylalanine.
It should be noted that these aa requirements were found
in an enriched minimal medium containing all other aa,
purines, pyrimidines, etc., and thus reflects a genuine phe-
notypic requirement. We also compared the frequencies
of sporulation and competency in the three B. subtilis
strains (i.e., IS58, IS56 and L3). As shown in Table I, the
A 12a34 1
b 2 3 4
GTP
PPGPP
a?@*@ 0 origin
B 12312 123
IS58 IS56 L3
Fig. 1. Accumulation of (p)ppGpp in strains of B. suhtilis. (A)
Accumulation of ppGpp in the L3 mutant in response to aa starvation
(a) and to carbon source starvation (b). The cell culture was grown in
low-phosphate medium to an A,,, of 0.1. The cells were then labelled
with [32P]phosphoric acid (50-100 pCijm1) for one generation. After
labelling for 1 h, the cells were either starved for amino-acids by the
addition of serine hydroxamate to a concentration of 1 mg/ml (a) or
for carbon source by the addition of rmG to a final concentration of
I % (b). Samples of 100 ~1 were withdrawn at 0, 5, 10 and 15 min (lanes
l-4 respectively), mixed with 13 M formic acid and frozen. After thaw-
ing, samples were centrifuged and 10 ~1 of the supernatants were applied
to polyethyleneimine-cellulose (PEI) plates (Brinkmann Instruments,
Westbury, NY, USA) for separation by thin-layer chromatography of
the phosphorylated guanosine nucleotides in 1.5 M KH2P0,. (B)
Accumulation of (p)ppGpp in all of the three isogenic strains (IS58,
IS56 and L3) during carbon source limitation. Cultures were treated
with clmG ( I “h) as described in panel A. The samples 1, 2 and 3, were
taken after 0, 5 and 15 min in the case of IS58 and L3 and samples 1
and 2 at 0 and 15 min for IS56.
relA mutant is less competent and its ability to sporulate
is slightly impai@d, while L3 exhibits maximal levels of
heat-resistant spores and continues to be competent.
(b) Mapping studies
We postulated that in B. subtilis relA is responsible for
responses to aa and to carbon source limitations and in
the L3 mutant, the inability to respond to carbon starva-
tion is suppressed. Thus, it became apparent that the
genetic location of relAcS’ was critical to our hypothesis
that the L3 mutant represents either an inter- or intra-
genie suppressor mutation of the relA gene. Early map-
ping experiments with the transducing phage AR9 and
93
TABLE I
The mutant strains of B. subtilis
Relevant
genotype”
Strainb Phenotypic
requirement
of aac
Resistance“
to:
Accumulation Sporulation
of (P)PPGPP” frequency’
in: (%)
Transformation
frequency8
(%)
MM-X AT TS SH nmG NSM
Wild type
(&A +, w/C ‘)
WIA
w/A, w~A’~
IS58
IS56
L3
(trp, 1~s)
(trp, 1~s) phe, his, val
(trp, 1~s)
R S ++ ++ 76 100
S S _ _ 48 9
R S _ + 85 46
CUR101 (trp), phe S S _ +
Tn917-LTV 1 lys, leu, val
in the relA region
mapping
strain
(w/A+, w/C+)
phe
CUR102 (trp)phe R S undetermined
BD79 (leu. phe) R S undetermined
“Tn917-LTV1 is a vector containing transposition-proficient derivative of Tn917 that contains the ColEl-derived cloning vector (Youngman, 1990).
It was transformed into strain CU4159 (trpC2, zig-83::Tn917; Vanderyar and Zahler, 1986), which replaces the MLSR by CmR between leuB and pheA.
bStrains IS58, IS56 obtained from 1. Smith; CU4159 from S. Zahler is the parent strain of CUR101 and 102 constructed in this study, as was L3.
“MM-X. phenotypic aa requirement noted as 20aa-X in minimal medium (Spizizen, 1958) containing an incomplete set of 19 aa with the missing X
indicated; the aa in parentheses represent the auxotrophic requirement of the strain. The medium does not contain AT.
dAT, 3-amino-1,2,4-triazole (15-30 mM); Ts, thiostrepton (l-5 pg/ml).
‘SH, serine hydroxamate (l-2 mg); ccmG= cc-methyl glucoside (1%). The accumulation of (p)ppGpp in the presence of SH or ctmG was carried out
as described in the legend of Fig. 1.
‘Sporulation frequency determined after growth in NSM [Nutrient sporulation medium (Schaeffer et al., 1965)]: % viable cells after heating 1 ml
undiluted sample, at 80°C for 20 min.
8Transformation frequency based on the numbers of MLSR transformants scored per pg DNA.
scoring for the relA marker by means of plate autoradiog-
raphy were reported by Smith et al. (1980). In that report,
relA was placed between aroD and the leu operon at 23.5”
on the map. We repeated the mapping of relA and the
new mutant relAcS) using resistance and sensitivity to AT
as the phenotype of the rel character. We targeted to the
leu-phe region by using an integrative plasmid pERlO
carrying a selectable antibiotic marker (CmR) and a
3.2-kb EcoRI-BamHI fragment homologous to leuA- /euC-led3 (LaFauci et al., 1986). The integrative plasmid
pERlO was transformed into strains IS58, IS56 and L3.
Clones with the CmR character were selected and used
to raise PBS1 donor lysates for transduction crosses. The
donors were designated as strains IS56[pER102] and
L3 [pERlO (Table II). For the crosses various isogenic
recipients were used: (i) BD79 (leuB1, pheAl), (ii) BDR79
(leuB1, pheA1, CmR) constructed by a transduction
cross using IS56[pER102] as donor to transfer the
re/A/ATS gene to that background and (iii) ISR56 (trpC2, /e&l, relA/ATS) constructed by congression with DNA
from strain BD79 to transfer the led gene.
Table II summarizes the transductional crosses per-
formed with the parental (BD79) and the constructed
(BDR79) recipients. As expected, there was tight linkage
between the integrated CmR and the led marker (co-
transduction values of 62-86%) and less tight to the pheA marker with cotransduction values of 25-41%. The relA gene as followed by the ATRjs phenotype showed the
weakest linkage to CmR with cotransduction values of
17% (24/138, Table II). This relationship places the relA gene downstream from the leuB-pheA region. As shown
in Table II, the L3 donor gave rise to similar numbers of
AT’ recombinants as those produced by the IS56. The
appearance of a class of AT’ clones clearly indicates that
L3 is a double mutant with the relA@’ contributing to
the ATR phenotype and the original relA retaining the
AT’ phenotype.
In the cross with the second BDR79 recipient the se-
lected markers were either Leu+ or Phe+. Higher co-
transduction values (92 and 97%) were obtained with the
ATR marker of L3. The Phe+ATR recombinant class was
the highest (Table II). We note that difference between
the lysates with and without the 8.8-kb plasmid (pER102)
which influence the absolute cotransduction values exist.
For example, the cotransduction values for leuA-pheA are
higher without the plasmid (53-62%, data not shown)
compared to 18&43% in the presence of the integrated
plasmid (Table II). Biases associated with CmR have been
94
TABLE II
Transduction crosses involving the w/A and ~l.4’~ alelles
Donor
(relevant phenotype)
Recipientb Selected
marker’
Recombinant classesd Probable order”
ATR /el_lB pheA No.
L3[pER102]
(Leu+, Phe’. CmR, w/A’~‘/AT~)
BD79 CmR 1
1
1
0
0
0
L3[pER102]
(Leu+, Phe’, CmR, rr/A@‘/ATR)
BDR79 Leu +
Phe+ I 1
0
0
IS56[pER102]
(Leu+, Phe’, CmR, relA/ATs)
BD79 CmR 0
0
0
1
1
1
1 48
0 68
0 46
I 11
0 24
1 0 leuB(CmR)-pheA-w/A
I 22
0 56
0 II
1 6
0 11
1 4
1 15
0 49
1 4
0 0
1 13
1 71
1 0
1 2
“The phenotypes of the donors are: prototrophy, i.e., Leu+. Phe’ (on minimal plates without the appropriate aa), CmR resistance to 10 pg Cm/ml
(Sigma, St. Louis, MO, USA) in LB plates, ATR” resistance or sensitivity to 15-30 mM AT (Sigma) in complex minimal medium (Rudd et al., 1985).
“Recipient strains are BD79 (leuB, pheA1, CmS, relA+/ATR) and the constructed BDR79 (leuB1, pheA1, CmR, relA/ATS).
‘Transduction crosses as described by LaFauci et al. (1986). The mixture was centrifuged and resuspended mixture in 1 ml dilution salts, 0.05- and
O.l-ml aliquots were plated onto selective plates, i.e., LB plus Cm, minimal medium plus the aa phenylalanine or leucine.
dDonor and recipient phenotypes are indicated by symbols 1 and 0. No.. number of transductants tested for cotransfer of unselected markers.
“The deduced gene order from the rare class of recombinants.
observed in similar crosses, and they depend on the rela-
tive size of the heterologous portion in the integrative
plasmid (LaFauci et al., 1986). The transduction crosses
clearly show a closer linkage between the pheA gene and
the ATR phenotype as compared to the leuB gene. The
data presented in Table II are consistent with our notion
that L3 represents a double mutant, since a small number
of AT’ recombinants (8%) were detected when L3 was
the donor. Finally, using the third constructed recipient
ISRR.56 (trpC2, leuB1, relA/ATS) we found that the same
L3 lysates reported in Table II yielded recombinants that
were CmR/ATR and Leu’ /ATR with average values of 4
and 2%, respectively (data not shown).
The suggested gene order: leuA-pheA-relA/relAcS’ was
further confirmed by transformational crosses using the
same parental strain BD79 as a recipient and donor DNA
from the two relevant strains IS56-relA/ATS and
L3-relA(S)/ATR. In the transductional crosses the plating
is done at no dilution, and therefore spontaneous ATR
clones yield high backgrounds. In transformational
crosses, upon dilution of the mixture, primary selection
for the drug resistance is possible and the number of ATR
colonies produced represents the true recombinant class.
The transformation data similarly showed a closer link-
age of the ATsiR phenotype to pheA than to leul? (for
IS56 DNA, it was 34 and 12%, respectively; data not
shown). Among the Leu+ or Phe+ transformants replica-
plated on AT plates, the Phe+/ATS class was larger than
Leu+/AT” (14% compared to 6%, respectively; data not
shown). As noted above, when L3, which is phenotypi-
tally ATR, served as the donor, some of the transformants
were AT’, thus indicating that the original relA/ATS mu-
tation is now suppressed by a secondary mutation, i.e.,
relA@‘. All the genetic evidence allowed us to place the
relA gene, as well as relA (‘) downstream from the pheA
gene.
The present studies with a B. subtilis strain CU4159
(trpC2, zig-83::Tn917) which contains a transposon in
the leuB-pheA region (Vanderyar and Zahler, 1986) were
relevant to the characterizations of the relA/relA’S’ locus.
The strain exhibits resistance to MLS (1 yg crythromycin
and 25 pg lincomycin per ml) and to AT like a normal
95
r&i+ strain. Its DNA (and other Tn9f7-LTV1 deriva-
tives constructed by Youngman, 1990) generated trans-
formants that became AT’ and also exhibited aa
requirements like IS56 and L3 (CUR101 and CUR102,
see Table I). The Tn917 transposon as followed by the
MLS phenotype cotransferred with Leuf and Phe’ to
the extent of 46 and 24%, respectively (data not shown).
Among the MLSR transformants 45% became ATS and
a greater number among them showed a tighter linkage
to pheA (Phe+/ATR 52% versus Leu+/ATR 96%; data
not shown). One may conclude that upon the insertion
of Tn9f7 at the relA locus various phenotypes are pro-
duced similar to relA,/ATS and relA’S’fATR which map at
the same location, i.e., near pheA.
(c) Concluding remarks
The mapping data reveals the following gene order: ifv-
Ecu operon( 247”) . . . pheA( 240”) . . . relA/reiAcS’ (239”-235”)
. . . aroD(226”). This order is consistent with the original
report (Smith et al., 1980) although we placed the relA
gene closer to pheA. The relA locus appears to be subdi-
vided into two domains. The first, is involved in the re-
sponse to aa limitation, and the second is the putative
reZAfS’ allele involved in the response to carbon source
limitation. Thus, the original relA mutant isolated in B.
subtilis (Swanton and Edlin, 1972) can be regarded as a
double mutant with defects in both activities. The L3
revertant (r~~A’~re~A(‘)) represents an intragenic suppres-
sion event that yielded partial restoration of the wild-
type phenotype specifically, the ATR property; the partial
loss of aa requirement; and the ability to respond to
carbon limitation. Cloning of DNA from strains CUR101
and CUR102 (Table I), with Tn917-LTV1 located pre-
sumably in either of the functional domains, should con-
clusively verify our conclusions regarding the relA/relA@)
locus in B. suhtilis.
In E. coli, the spoT gene plays a central role in the
relA-independent response to carbon starvation, growth
rate regulation and the stringent response (Cashel and
Rudd, 1987). To date, a clear demonstration of a similar
gene in the Gram+ B. suhtilis has not been made, al-
though a ppGppase activity associated with a ribosomal
fraction has been reported in B. subtifis and other Bacillus
strains (Richter, 1979). A claim of a functioning spoT
gene product has been made based on studies with chelat-
ing agents that abolish the activity of a Mn2+-dependent
ppGppase (Belitsky and Shakulov, 1982). The other
spontaneous revertants isolated in this study that were
ATR and were slow growers had phenotypic properties
resembling spoTmutants. It was noted that these putative
spoTmutants showed a range of aa requirements between
IS56 and L3 (see Table I). A systematic study of these
ATR mutants with respect to their ability to hydrolyze
ppGpp versus the range of aa requirements they exhibit,
as well as their response to carbon starvation should be
undertaken. Mapping of these putative spoT mutants
would reveal whether the reld locus in B. subtilis is com-
posed of additional functional domains involved in the
metabolic pathway of the (p)ppGpp cycle.
ACKNOWLEDGEMENTS
We thank Dr. M. Cashel for critically reading the
manuscript and to B. Studamire, a student in R.R.‘s labo-
ratory, for technical and editorial assistance. This work
was partially supported by grants from NIH grant
GM25286 to G.G. and Minority Research Centers in
Minority Institutions NIH grant RR03037, and by
CUNY Research awards 668151 and 662159 to R.R.
REFERENCES
Betitsky, B.R. and Sakuiov, RX: Functioning of spt>T gene product in
B. suhtilis cells. FEBS Lett. 138 (1982) 226-228.
Cashel, M. and Rudd, K.: The stringent response. In: Neidhart, F.D.,
Ingraham, J.L., Low, K.B., Magasanik, B., Schaechter, M. and
Umbarger, H.E. (Ed%), Eschrrichia coli and Salmon& typhimurium:
Cellular and Molecular Biology. American Society for
Microbiology, Washington, DC, 1987, pp. 1410- 1438.
Hernandez, V.J. and Bremer, H.: E. coli ppGpp synthetase 11 activity
requires spoT. J. Biol. Chem. 266 (1991) 5991-5999.
Hilton, J.L., Kearney, P.C. and Ames, B.N.: Mode of action of the
herbicide 3-amino 1,2,4 triazole: inhibition of an enzyme of histidine
biosynthesis. Arch. Biochem. Biophys. 112 (1965) 544-547.
LaFauci, J., Widom, R.L., Eisner, R.L., Jarvis, E.D. and Rudner, R.:
Mapping of rRNA genes with integrable plasmids in B. s~bt~~~~. J. Bacterial. 165 (1986) 204-214.
Nishino, T., Gallant, J., Shalit, P., Palmer, L. and Wehr, T.: Regulatory
nucleotides involved in the Rel function of B. suhtilis. J. Bacterial.
140 (1979) 671-679.
Rojiani, M.V.. Jakubowsky, H. and Goldman, E.: Effect of variation of
charged and uncharged tRNA TRP levels on ppGpp synthesis in E.
coti. J. Bacterial. 171 (1989) 6493.-6502.
Rudd, K.E., Bechner, B.R., Cashel, M. and Roth, J.R.: Mutations in the
spoT gene of S. typhimurium: effects on his operon expression.
J. Bacterial. 163 (1985) 534-542.
Schaeffer, P., Millet, J. and Aubert, J.: Catabolic repression of bacterial
sporulation. Proc. Natl. Acad. Sci. USA 54 (1965) 704-711.
Smith. t.. Paress, P. Cabane, K. and Dubnau. E.: Genetics and physiol-
ogy of the rel system of B. suhtilis. Mol. Gen. Genet. 178 (1980)
271-279.
Smith, I., Paress, P. and Pestka, S.: Thiostrepton-resistant mutants ex-
hibit relaxed synthesis of RNA. Proc. Natl. Acad. Sci. USA 75 (1978)
5993-5997.
Spizizen, J.: Transformation of a biochemically deficient strain of
Bacilltrs suhtilis by deoxyribonucleate. Proc. Natl. Acad. Sci. USA
41 (1958)4787-4791.
Swanten, M. and Edlin, G.: Isolation and characterization of an RNA
relaxed mutant of B. suhtilis. Biochem. Biophys. Res. Commun. 146
(1992) 583-58X.
96
Vandeyar, M.A. and Zahler, S.A.: Chromosomal insertions of Tn917 in
B. subtilis. J. Bacterial. 167 (1986) 530-534.
Xiao, H., Kalman, M., Ikehara, K., Zemel, S., Glaser, G. and Cashel,
M.: Residual guanosine 3’5’-bispyrophosphate synthetic activity of
re/A null mutants can be eliminated by spoT null mutations. J. Biol.
Chem. 266 (1991) 5980-5990.
Youngman, P.: Use of transposons and integrational vectors for muta-
genesis and construction of gene fusions in Bnrillus species. In:
Harwood, C.R. and Cutting, SM. (Eds.), Molecular Biological
Methods for Bacillus. Wiley. New York, 1990, pp. 220-257.