Vol. 171, No. 6

Alp, a Suppressor of Lon Protease Mutants in Escherichia coliJANINE E. TREMPY AND SUSAN GOTTESMAN*

Laboratory of Molecular Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892

Received 6 January 1989/Accepted 24 March 1989

Escherichia coli lon mutants lack a major ATP-dependent protease, are sensitive to UV light andmethylmethane sulfonate (MMS), and overproduce capsular polysaccharide. Evidence is presented that an

activity (Alp), cloned on a multicopy plasmid, can suppress the phenotypes of Ion mutants. The sensitivity toUV and MMS is a reflection of the stabilization of the cell division inhibitor SulA, while the capsuleoverproduction arises through the stabilization of a transcriptional activator of capsule biosynthetic genes,RcsA. Multicopy alp (pAlp) suppressed capsule formation in Alon cells, and Alon cells containing the pAlpplasmid were resistant to MMS treatment. The MMS resistance of Alon pAlp+ cells correlates with an increasein the degradation of SulA to that found in lon+ cells. Lon-directed degradation of SulA was energy dependent,as was the increase in degradation of SulA in Alon pAlp+ cells. alp maps close to pheA, at 57 min on the E. colichromosome. Although pAlp can substitute for Lon, cells lacking alp activity did not have the phenotype on aIon mutant. This study demonstrates that at least one activity, when overproduced in the cell, can substitutefor Lon protease.

Protein turnover serves as a regulatory strategy for cellsduring periods of changing conditions or environmentalemergencies. When a crisis is encountered, the cell mayrequire the rapid synthesis or elimination of specific geneproducts to permit survival. The cell can respond quiterapidly by regulating gene transcription through a variety ofinteresting mechanisms, but once a gene product is made itis generally long lived (10, 24). After the cell has adjusted tothe emergency and resumes normal growth, the emergency-activated gene products can be gradually removed by dilu-tion, provided that the accumulation is not harmful to thenormal physiology of the cell. However, if these productsmust be rapidly cleared for normal growth, then some sort ofinactivation of the gene products must occur along with thetermination of their synthesis.The most thoroughly studied proteolytic process in Esch-

erichia coli is that which involves the ATP-dependent pro-tease Lon. Mutations in Ion give rise to a pleiotrophicphenotype (for a review, see reference 10). Two of thesephenotypes, sensitivity to DNA damage and overproductionof capsular polysaccharide, can be explained as a directresult of the unavailability of Lon protein, resulting in thestabilization of two normally unstable proteins (10, 26, 29).When cells experience DNA damage (e.g., through expo-

sure to UV light or methylmethane sulfonate [MMS]), theSOS repair response in E. coli is activated, giving rise to theexpression of specifically repressed genes (19, 31). Oneexpressed gene product is SulA, a cell division inhibitorwhich inhibits septation by interacting with FtsZ, a compo-nent of the cell division apparatus (13-15). In wild-type cells,SulA is very unstable (with a half-life of 1.2 min). Thisnormal instability of SulA permits the cell to resume celldivision once the environmental stress has been alleviated.However, in Ion mutant cells, the SulA half-life is 20 to 30min (21, 26). The stabilization of SulA in Ion mutants resultsin irreversible filamentation and subsequent cell death (14).As is the case with sensitivity to DNA-damaging agents,

overproduction of capsular polysaccharide is dependent on

an unstable protein, RcsA (29). The regulation of the capsu-

* Corresponding author.

lar polysaccharide genes (cps) occurs at the transcriptionallevel (30). Ion+ cells with a lac fusion to one of the cps genes(cpsBJO::lac) are Lac-, whereas Ion mutant cells with thecpsB10: :lac fusion are Lac'. The difference in cps: :lacexpression correlates with differences in the stability of thecps positive transcriptional regulator, RcsA; the half-life ofRcsA is 5 min in lon+ cells, and in Ion mutant cells thehalf-life is increased to 20 min (29). RcsA is apparentlylimiting for cps expression in lon' cells; its stabilization inIon mutant cells allows high-level cps expression. The nor-mal instability of RcsA may permit fine tuning of cps geneexpression in response to changing environmental condi-tions. As seen with the stabilization of SulA protein, thestabilization of RcsA protein in lon mutant cells gives rise toa phenotype (overproduction of capsule) that is a disadvan-tage to the cell for normal growth.SulA and RcsA proteins represent two examples of normal

short-lived proteins whose regulation by Lon-dependentproteolysis may help the cell recover from emergency con-ditions. We have exploited the behavior of these Lon sub-strates to probe the cell for other, similar protease functions.Maurizi et al. (21) have demonstrated residual degradation ofSulA protein in Alon cells. Similarly, RcsA turnover is stillsignificant in cells devoid of lon (29). On the basis of theseobservations, other protease activities that are capable ofdegrading SulA and RcsA must exist in E. coli. Increasingthe activity of such proteases should lead to suppression ofthe Lon- phenotypes. We have screened fragments of the E.coli chromosome for those which, when present on multi-copy plasmids, are capable of suppressing Lon- pheno-types. We have identified an activity capable of suppressingboth sensitivity to DNA-damaging agents and capsule over-production. Therefore, this activity, when present in excessin the cell, can mimic Lon.

MATERIALS AND METHODS

Bacteria and phage strains. The relevant bacterial andphage strains used are listed in Table 1. Hfr strains andadditional recipient strains with appropriate markers usedfor mapping were obtained from B. Bachmann (E. coliGenetic Stock Center, Yale University, New Haven,

3348

JOURNAL OF BACTERIOLOGY, June 1989, p. 3348-33530021-9193/89/063348-06$02.00/0Copyright © 1989, American Society for Microbiology

on Septem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

Alp, A SUPPRESSOR OF Lon PROTEASE MUTANTS IN E. COLI 3349

TABLE 1. Bacterial and phage strains

Strain Genotype Reference or source

JT5lOOa Alon-510 Agal-165 SG20250 deriva-tive

SG20780a Alon-510 cpsBIO::lac-Mu-immX 3SG20781a lon' cpsBIO::lac-Mu-immX 3SG20250' Ion' cps' 11C600 F- thi-J thr-I leuB6 lacYl National Institutes

tonA21 supE44X of Health straincollection

JM1o1 A(lac-proAB) supE thi(F' 23traD36 proAB lacJqZAM15)

SG1611 Alon-S10 Agal-165 A(lac-proAB) JM101 derivativesupE thi(F' traD36 proABlacIqZAMl5)

DB1255 recBC sbcBJS hsdR supF8 33XNK1105 cI857 b221 Pam8O nin5 AKan 32

ptac-transposasea MC4100 derivative; contains A1acU169 (4).

Conn.). The procedures for Hfr mating and P1 transductionhave been described elsewhere (25).

Construction of library and selection of Ion-complementingplasmids. Chromosomal DNA was isolated (28) from cellscarrying a Alon-510 mutation (JT5100), partially digestedwith Sau3A (28), and cloned into the BamHI site of pBR322.This multicopy plasmid library was transformed into Aloncells containing the cpsBJO::lac fusion (SG20780). Ampicil-lin-resistant colonies were replica plated onto either Luriabroth plates containing 0.05% methylmethane sulfonate(MMS) (12) or plated directly onto lactose-MacConkey agarplates containing 50 ,ug of ampicillin per ml. Plasmids thatconferred MMS resistance also conferred UV resistance.

Insertional mutagenesis of plasmid DNA. Transposon mu-tagenesis of plasmid-cloned DNA was performed as de-scribed previously (32) by using the phage XNK1105, whichwas provided by N. Kleckner. The mini-transposon carriedon this phage (AKan) confers resistance to kanamycin andcontains the outer ends of Tn1O (32). Wild-type cells (SG20250) containing pAlp were infected with XNK1105, andKanr colonies were selected. Plasmid DNA was isolatedfrom the Kanr cells and retransformed into wild-type cells(C600). Transformants were selected for Kanr and screenedfor Ampr. Plasmids containing the AKan inserts were iso-lated, and restriction analysis was performed (20) to deter-mine the location and orientation of the AKan. alp-210::AKan is a AKan transposon that eliminates alp activity fromthe plasmid, and alp' zJh27::AKan is a AKan transposonthat does not affect alp activity but is located in the chro-mosomal DNA carried by the plasmid.Both AKan insertions in the chromosomal insert were

transferred to the chromosome by linearizing the plasmidwith the restriction endonuclease NheI and transformingDB1255 (33) as previously described (3, 16). The AKan wasmapped by Hfr matings and P1 transductions (25).

In vivo SulA turnover. lon+ (JM101) cells were trans-formed with pBR322, and Ion mutant (JM101 derivative,SG1611) cells were transformed with either pBR322, pAlp,or alp-210: :AKan plasmid. Ampicillin-resistant transfor-mants were transformed with a pBR322-compatible plasmid(conferring chloramphenicol resistance) containing the sulAlocus under the control of an isopropyl-p-D-thiogalactopy-ranoside (IPTG)-inducible lac promoter. Ampicillin-resis-tant, chloramphenicol-resistant cells were grown at 32°C inM56 minimal medium (9) supplemented with 0.4% glycerol,

essential amino acids (50 ,ug/ml) excluding methionine andcysteine, 50 ,ug of ampicillin per ml, and 25 t±g of chloram-phenicol per ml. To induce transcription of SulA, IPTG (finalconcentration, 5 mM) was added to exponentially growingcells (optical density at 600 nm [OD6], 0.6) and incubationwas continued at 32°C for an additional 45 min. Cells werepulse-labeled for 1 min with [35S]methionine (10 ,uCi/ml,Amersham Corp.) and chased with a 103-fold excess ofunlabeled methionine. Samples were removed at varioustimes after the pulse-chase and precipitated with 10% (vol/vol, final concentration) cold trichloroacetic acid. Acetone-washed pellets were suspended in Tris-buffered saline (10mM Tris, pH 7.8, 150 mM NaCl) containing 0.1% sodiumdodecyl sulfate and 1% Nonidet P-40. Suspended pelletswere boiled, and aprotinin (0.1 U; Sigma Chemical Co.),phenylmethylsulfonyl fluoride (1 mM; Sigma), N-ot-tosyl-L-lysine chloromethyl ketone (100 ,uM; Sigma), and N-tosyl-L-phenylalanine chloromethyl ketone (100 ,uM; Sigma)were added to inhibit proteolytic activity. Anti-SulA anti-body, negatively absorbed against total protein from a sulAE. coli strain, and Staphylococcus A protein (IgGsorb;Enzyme Center) were used to immunoprecipitate labeledSulA. Immunoprecipitates were analyzed by electrophoresison sodium dodecyl sulfate-polyacrylamide gels (10 to 20%linear gradient gels). Fluorography with En3Hance (Dupont,NEN Research Products) followed by autoradiography wasused to visualize the labeled protein bands. To determinewhether alp activity required energy, the same procedurewas followed except that at the time unlabeled methioninewas added (chase), 10 mM potassium cyanide (KCN) wasincluded to inhibit ATP generation.

RESULTS

Identification of lon-suppressing activity from a multicopyplasmid library. We have used the two characteristic pheno-types of cells carrying a lon mutation, sensitivity to DNA-damaging agents and high-level expression of genes forcapsular polysaccharide synthesis, to test the hypothesisthat other protease activities in the cell can substitute forLon protease if present in excess. A multicopy plasmidlibrary consisting of chromosomal DNA from Alon cells (JT5100) was transformed into Alon cells containing the cpsBlO::lac fusion (SG20780 Lac'; MMS sensitive). This strainpermitted us to examine the fate of two lon substrates, SulA(through sensitivity to MMS) and RcsA (through P-galactosi-dase levels). Twenty thousand colonies were screened onMacConkey agar-lactose plates containing ampicillin or onLuria broth plates containing MMS. The three classes oflon-suppressing plasmids identified in the screen are listed inTable 2. lon+ cells are MMS resistant and Lac-, whereasAlon cells are MMS sensitive and Lac'. Seven plasmidswere identified (class 1) that suppressed the Lac' phenotypeof Alon cells but did not affect MMS sensitivity. One plasmidwas identified (class 2) that conferred MMS resistance onAlon cells but did not affect the Lac phenotype. Class 1 and2 plasmid activities might represent a protease or otherproteins that are specific to the SOS or capsule pathways.Restriction map analysis of these plasmid inserts (data notshown) suggested that the activities are not known compo-nents of the SOS or capsule pathways. Two plasmids iden-tified in the screen (class 3) suppressed both phenotypes of alon deletion. The two lon-suppressing plasmids of class 3were identified in separate screens. One plasmid was iden-tified by its ability to confer resistance to MMS and thenscreened for the Lac phenotype. The other plasmid was

VOL. 171, 1989

on Septem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

3350 TREMPY AND GOTTESMAN

TABLE 2. Three classes of /on-suppressing mutants'

PhenotypeIsolate

MMSb Lac' No. of isolates'

Wild-type RAlon S +Class1 S - 7Class 2 R + 1Class 3 R - 2

a Chromosomal DNA was isolated from Alon cells, partially digested withSau3A enzyme, and cloned into pBR322. This plasmid library was trans-formed into Alon cells, and plasmids were screened for their ability tosuppress one or both of the described phenotypes.

b R, Resistant; S, sensitive.c -, Lac- (white on lactose-MacConkey agar containing 50 p.g of ampicillin

per ml at 32°C); +, Lac' (red on lactose-MacConkey agar containing 50 jig ofampicillin per ml at 32°C).

d Number of colonies identified out of 20,000 clones.

selected for its ability to reverse the Lac+ phenotype of Aloncells to Lac- and then screened for the MMS phenotype.Restriction analysis revealed that the two lon-suppressingplasmids of class 3 contained an identical 4.5-kilobase DNAinsert. We chose to focus our analysis on the plasmid thatsuppressed both Ion phenotypes (class 3). We reasoned thatthe ability to suppress two independent lon phenotypes wasmost easily explained as an increased protease activity incells carrying the multicopy plasmid. We have named thisplasmid pAlp and the putative gene alp. We confirmed thatthe ability to suppress the Ion phenotypes resided entirely inthe plasmid by isolating pAlp and retransforming it into Aloncells (SG20780). Once again, both MMS sensitivity and highcps::lac expression were suppressed.AKan transposon insertions were isolated in pAlp to

physically map the activity responsible for suppressing theIon phenotypes. The Ampr Kanr plasmids were isolated,transformed into Alon cps::iac cells (SG20780), screened forthe MMS and Lac phenotypes, and analyzed by restrictionanalysis. Two types of AKan insertions in the plasmid wereidentified: (i) insertions that did not affect the MMS or Lacphenotypes and (ii) insertions in pAlp that abolished sup-pression of both Alon phenotypes (Fig. 1). On the basis ofthe phenotype and DNA restriction analysis, we determinedthat alp activity resided in a DNA fragment of 750 bases.Neither of the AKan insertions identified affected just one of

R p PH R

TABLE 3. Effect of multicopy alp on the expressionof the cpsBJO-lac fusion

Host genotype Plasmid Phenotypea -Galactosidase(strain) genotype activity

Ion' (SG20781) Lac- 8Alon (SG20780) Lac' 760Alon pAlp Lac- 9Mlon alp-210::AKan Lac' 604Alon alp' zjh-27::AKan Lac- 18Ion' pAlp Lac- 7Ion' alp-210::AKan Lac- 8Ion' alp' zfh-27::AKan Lac- 8

a Lac phenotype was determined on lactose-MacConkey indicator agar at32°C.

b Cells were grown in glucose minimal medium at 32'C and assayed forP-galactosidase specific activity (expressed as Miller units [25]). Plasmid-containing strains were grown in the presence of ampicillin (50 pLg/ml).

the phenotypes. This suggests that a single gene locus, inmultiple copies, mimics the effects exerted by Lon in thecell. We have named this locus alp (for alternative activity toLon protease).To quantitate the effect of pAlp on cps::iac expression,

isogenic lon+ (SG20781) and Alon (SG20780) cells containinga cps::iac protein fusion were assayed for P-galactosidase inthe presence of the wild-type plasmid and AKan insertionplasmids. In Ion' cells, the cps::iac fusion was expressed atlow levels (Table 3, line 1). In Alon cells, cps::iac expressionwas approximately 100-fold higher than in Ion' cells (Table3, line 2). This difference has been interpreted as a reflectionof the effect of Lon protease on the stability of the regulatorof cps operons, RcsA (29). Alon cells carrying pAlp hadP-galactosidase levels similar to those seen in Ion' cells(Table 3, line 3). When pAlp contained a /Kan that inacti-vates its ability to suppress Ion (alp-210::AKan), cps::iacexpression returned to high levels in Alon cells (Table 3, line4). However, the plasmid suppressed cps::iac expressionwhen the AKan was inserted elsewhere in the plasmid (alp'zjh-27::AKan) (Table 3, line 5). Additionally, lon+ cellscontaining any of the three plasmids expressed cps::iac atthe same levels as those without the plasmid (Table 3, lines6 to 8).

In vivo proteolysis of SulA protein: a biochemical examina-tion of pAlp effects. The decrease of ,-galactosidase levels in

V V

il0.5kb

FIG. 1. Restriction map of cloned insert in pAlp plasmid. Symbols: 210, alp-210::AKan insertion which inactivates alp activity; 27, alp'z,fh-27::AKan insertion which does not inactivate alp. Restriction enzyme sites are abbreviated as follows: R, EcoRI; V, EcoRV; H, HindIll;P, PstI. The following restriction enzymes did not cleave this fragment: BamHI, ClaI, KpnI, NheI, NruI, and PvuII.

J. BACTERIOL.

on Septem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

Alp, A SUPPRESSOR OF Lon PROTEASE MUTANTS IN E. COLI 3351

A

SulA - --

C

SulA - -_ _

B

fm_ _40-_ --_

D

I_ ____- W

E

oft --,Mg . am -.

F

__-psod -.00O

1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8FIG. 2. SulA stability in lon' cells, Alon cells, and A/on cells plus pAlp or alp-210::A&Kan plasmid. IPTG-induced cells carrying a

pBR322-compatible plasmid containing the sulA locus under the control of an IPTG-inducible lac promoter were pulse-labeled for 1 min with[35S]methionine (10 ,uCi/ml) and chased with 103-fold-excess unlabeled methionine. Samples were immunoprecipitated and analyzed byelectrophoresis as described in Materials and Methods. Labeled protein bands were visualized by fluorography. The position to which SulA(18 kilodaltons) migrated in this gel system is indicated on the figure. (A) Ion' (JM101) + pBR322 control vector. (B) Alon (SG1611) + pBR322control vector. (C) A/lon (SG1611) + pAlp. (D) A/on (SG1611) + alp-210::AKan plasmid. (E) Ion' (JM101) + pBR322 control vector. (F) Alon(SG1611) + pAlp. For panels E and F, cells were incubated with potassium cyanide as described in Materials and Methods. Cells were labeledfor 1 min (lane 1) and chased for 1 (lane 2), 2 (lane 3), 3 (lane 4), 5 (lane 5), 7 (lane 6), 12 (lane 7), and 20 min (lane 8).

Alon cells carrying pAlp is consistent with an increase inturnover of RcsA. Resistance to MMS,in cells carrying thepAlp plasmid is consistent with an increase in degradation ofthe cell division inhibitor, SulA. The effect of pAlp on thestability of SulA protein was directly tested in lon' and Alonstrains (JM101 and SG,1611, respectively) containing apBR322-compatible plasmid with the sulA locus under thecontrol of an IPTG-inducible lac promoter. The IPTG-induced cells were pulse-labeled with [35S]methionine andchased with an excess of unlabeled methionine, and immu-noprecipitated samples were analyzed by sodium dodecylsulfate-polyacrylamide gel electrophoresis followed by auto-radiography. An example of this experiment is illustrated inFig. 2. SulA is very unstable in Ion' cells (with a half-life ofless than 5 min) (Fig. 2A). The instability of SulA is afunction of Lon (Fig. 2B) in that SulA is stabilized in A/oncells (with a half-life of greater than 20 min). Introducing thepAlp plasmid into A/on cells (Fig. 2C) decreased SulAhalf-life to that which was observed in Ion' cells (less than 5min). A/on cells carrying the alp-210::AKan mutation on theplasmid (Fig. 2D) showed a pattern of SulA decay similar tothat found in the Alon patent, indicating the elimination ofthe multicopy alp effect by the AKan insertion. A similareffect of pAlp on the stability of SulA was seen when StlAwas synthesized from a transducing phage in UV-irradiatedcells (data not shown). In both cases, pAlp increased thedegradation of SulA in Ion mutant cells to that found in Ion'cells.

Proteolytic activity by Lon requires energy in the form ofATP in vitro (5, 6). We wanted to determine whether theincrease in degradation of SulA by pAlp required energy.The same procedure was followed as illustrated in Fig. 2A toD, except that potassium cyanide was added at the time ofthe chase td quench energy-generating systems. Inhibition of

ATP generation by cyanide stabilized SulA in Ion' cells(Fig. 2E). This result was expected, given the in vitro ATPdependence of Lon, and had been previously observed by D.Canceill and 0. Huisman (personal communication). Theactivity from pAlp that decreased the stabilization of SulA inA/on was also reversed by the addition of cyanide (Fig. 2F).This result indicates that, like Lon, the pAlp mechanismrequires energy to affect the stability of SulA.These results, taken together, suggest that we have iden-

tified a gene which, when present in multiple copies, actsdirectly or indirectly to impart a Lon+ phenotype to the cell.The only component that the SOS system and the capsulesystem are known to have in common are unstable sub-strates whose stabilities are affected by Lon. Since multi-copy alp suppresses the /on phenotype for both of thesesystems, alp may code for a protease which, when overpro-duced, is capable of degrading SulA and RcsA. It is alsopossible that the alp product may be a regulatory protein thatincreases synthesis of a protease or a site that titrates anegative repressor from protease genes.Chromosomal alp mutants. The AKan inserts in pAlp were

transferred to the chromosome by linearizing the alp-210::AKan plasmid with a single restriction cut and transformingthe linear plasmid DNA into a nuclease-deficient strain (3,16). PlCMclrlOO was grown on the resulting Kanr Ampstransformants that arose through homologous recombina-tion, and the lysate was used to transduce the AKan into theHfr strains used for mapping by interrupted mating. TheAKan was transferred early in interrupted matings with HfrKL16 (point of origin, 60 min) and PK191 (point of origin, 44min), consistent with alp mapping in the 44- to 60-minregion. P1 transductions of the AKan into appropriate recip-ient strains with markers in this area demonstrated that alpcotransduces with pheA (57 min [1]) at a frequency of 50%.

VOL. 171, 1989

on Septem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

3352 TREMPY AND GOTTESMAN

Comparison of the pAlp restriction map (Fig. 1) with that ofthe published restriction map of the E. coli chromosome (18)permitted us to accurately locate alp at kilobase pair 2765.Saito et al. have reported that grpE, a gene encoding a heatshock protein, cotransduces with pheA at a frequency of61% (27). grpE is carried on clones E8F2 and 22D7 ofKohara et al. (18); these phage overlap in a region at least 10kilobase pairs from alp (S. Gottesman and C. Georgopoulos,unpublished observations).The ability to transfer the alp::AKan insertion from the

plasmid to the chromosome suggested that alp is not essen-tial for normal cell growth. To confirm that secondarymutations or rearrangements were not occurring, we carriedout P1 transductions from the original alp::AKan isolatesinto pheA: :TnlO recipients, selected either kanamycin resis-tance or prototrophy for phenylalanine and screened for thenonselected marker. Control transductions were carried outusing the closely linked alp' zjh-27::AKan insertion, whichwas also transferred to the chromosome by linear transfor-mation. Both alp-210::AKan and alp' zlh-27::A&Kan showedessentially identical linkage to pheA regardless of the origi-nal selection. Therefore, if secondary mutations are presentin cells carrying alp-210::AKan, they must be tightly linkedto the kanamycin resistance marker. Duplications in the alpregion would be expected to be unstable and might showunusual transduction behavior; we have not encounteredproblems associated with these characteristics. alp-210::AKan was also introduced into Alon-S10 hosts by P1 trans-duction, once again by selecting PheA+ and screening forthe nonselected kanamycin resistance marker. The pheA-AKan linkage was identical in lon' and Alon strains. Appar-ently, the alp activity, as defined by the plasmid insertionmutations, is not essential for cell growth in the presence orabsence of Lon. Finally, lon' cells containing the alp-210::AKan mutation did not show any lon phenotypes (sen-sitivity to MMS or capsule overproduction). Thus, the pAlpsuppression of lon mutations was not due to any direct andnecessary role of alp in degrading Lon substrates.

DISCUSSION

Proteolysis, combined with controls on gene expression,can serve as an important mechanism for the fine tuning ofprotein availability and plays a necessary role in the rapidresponse to emergency conditions and recovery from theseemergencies (10). In E. coli, Lon has been identified as amajor, ATP-dependent protease. Ion mutants directly orindirectly affect the stability of at least two proteins indifferent global response networks, SulA and RcsA. In theabsence of Lon, the stability of SulA interferes drasticallywith the ability of the cell to recover from even mildDNA-damaging treatments. Capsule synthesis in the ab-sence of Lon is carried out at a rate too high for rapid cellgrowth. The residual turnover of SulA, RcsA, and manyabnormal proteins in cells devoid of Ion, as well as thecontinued rapid turnover of unstable lambda proteins unaf-fected by lon mutations (10, 21), suggests that proteasesother than Lon must be present in E. coli. Identification ofthe natural cellular substrates of these proteases will be animportant step in understanding the full possibilities forregulation by proteolysis. In addition, some foreign proteinscloned into E. coli are unstable; while lon mutations havesome stabilizing effect on this degradation, residual degra-dation is frequently significant (for a review, see reference24). We reasoned that we could identify other activitieswhich may overlap with Lon in substrate specificity by

overproducing these activities in the cell. We have identifieda gene present on a multicopy plasmid that can suppress thelon phenotypes of MMS sensitivity and overproduction ofcapsule. This activity is a newly identified locus of E. coliwhich we have named alp.When alp is present on a multicopy plasmid, it functions

similarly to Lon, as demonstrated in this study, by suppres-sion of two Lon phenotypes. We have demonstrated directlythat suppression of MMS sensitivity is correlated with anincrease in SulA degradation in Alon cells carrying alp on amulticopy plasmid and that energy is required for pAlpactivity. We noted other similarities between cells carryingalp on a multicopy plasmid and those carrying lon onplasmids. Goff and Goldberg reported that lon cloned onhigh-copy-number plasmids acquires insertion elements thatinactivate the Ion gene (8). Similarly, the pAlp plasmidrapidly acquires mutations which significantly reduce itsability to suppress lon mutations, although these mutationsdo not appear to be due to insertion elements (J. Trempy,unpublished observations). Presumably, increased levels ofLon or Alp lead to degradation of essential cell components.Additionally, we have observed that cells carrying the lon'or alp' plasmid are temperature sensitive and are unable togrow on minimal media even at lower temperatures (unpub-lished observations), suggesting once again that overproduc-tion of either Ion or alp is similarly unfavorable for normalcell growth.A function coded for by the Alp plasmid and inactivated

by insertions at one site within the plasmid can increasedegradation of the SulA protein. The easiest way to explainthis effect is that a limiting component of a protease,normally with only limited ability to degrade SulA or RcsA,is coded for by the alp gene. If so, it would seem that thisprotease is not normally essential for cell growth, sincetransfer of the alp-210: :AKan mutation into the chromosomewas easily accomplished. Similarly, if a positive activator fora protease or proteases were present on the plasmid, muta-tions that inactivate it might be defective for the proteasewhen present on the chromosome but might still give signif-icant levels of activity. However, we have not yet identifiedan alp product, although numerous attempts have been madeby using maxicell analysis and in vitro transcription-transla-tion systems. It is possible that a site rather than an openreading frame for a protein is on the plasmid and is inacti-vated by the alp-210: :AKan insertion. If a site for a negativeregulator were on the plasmid such that a protease repressorwere being titrated, it might not be surprising to find thatinactivation of the single site on the chromosome had nophenotype. Since five AKan insertions which inactivate alpappear to be at the same point in the plasmid, as determinedby restriction mapping (possibly reflecting a transposon hotspot), we cannot rule out a site. The absence of lon pheno-types in lon' cells carrying the alp-210::AKan insertion inthe chromosome demonstrated that alp is not necessary forIon synthesis or activity and that lon does not act indirectlyby activating or increasing synthesis of alp.What might alp do for the cell? No known proteases or

suspected proteases of E. coli for which genes have beenidentified correspond to alp (10, 17, 24). If we assume thatalp is a component of a new protease, the absence ofdramatic phenotypes may suggest that its natural substratesare not detected under ordinary growth conditions. SulA, forinstance, is a lethal Lon substrate, but the Lon effect onSulA is not visible until induction of synthesis by DNAdamage. If alp substrates are the members of other emer-gency-response systems, the detection of a phenotype may

J. BACTERIOL.

on Septem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

Alp, A SUPPRESSOR OF Lon PROTEASE MUTANTS IN E. COLI 3353

require the appropriate induction protocol. A better under-standing of the nature of the alp locus and the conditionsunder which it normally functions would help us to identifyputative substrates for Alp.The use of an overproduction phenotype provided by the

presence of a locus on a multicopy plasmid is simple and hasbeen successfully used in other systems to either suppressmutant functions or inhibit wild-type activities (2, 7, 22). Inour case, we did not have information about the naturalsubstrates of the putative protease, which might have al-lowed development of a selection for inactivation of theprotease. In addition, we were anxious not to eliminate fromour screening procedure genes essential for E. coli growth.Selection for increased activity of a given function andscreening for crossover activities of the plasmid-bome func-tion, as was described here, serve as useful alternatives toidentification of genes by inactivation.

ACKNOWLEDGMENTS

We thank Sankar Adhya and Nancy Trun for critical readings ofthe manuscript, Mary Lee Lanigan for assistance in preparation ofthe manuscript, Michael Maurizi for the gift of anti-SulA antibody,and Olivier Huisman and Danielle Canceill both for the sulA plasmidand for the protocol for SulA turnover experiments.

J.E.T. was supported by a fellowship award from the AmericanCancer Society.

LITERATURE CITED1. Bachmann, B. J. 1983. Linkage map of Escherichia coli K-12,

edition 7. Microbiol. Rev. 47:180-230.2. Berg, C. M., M.-D. Wang, N. B. Vartak, and L. Liu. 1988.

Acquisition of new metabolic capabilities: multicopy suppres-sion by cloned transaminase genes in Escherichia coli K-12.Gene 65:195-202.

3. Brill, J. C., C. Quinlan-Waishe, and S. Gottesman. 1988. Fine-structure mapping and identification of two regulators of capsulesynthesis in Escherichia coli K-12. J. Bacteriol. 170:2599-2611.

4. Casadaban, M. J., and S. N. Cohen. 1979. Lactose genes fusedto exogenous promoters in one step using a Mu-lac bacterio-phage: in vivo probe for transcriptional control sequences.Proc. Natl. Acad. Sci. USA 76:4530-4533.

5. Charette, M. F., G. W. Henderson, and A. Markovitz. 1981.ATP hydrolysis-dependent protease activity of the lon (capR)protein of Escherichia coli K-12. Proc. Natl. Acad. Sci. USA78:4728-4732.

6. Chung, C. H., and A. L. Goldberg. 1981. The product of the lon(capR) gene in Escherichia coli is the ATP-dependent protease,protease La. Proc. Natl. Acad. Sci. USA 78:4931-4935.

7. Gaur, N. K., E. Dubnau, and I. Smith. 1986. Characterization ofa cloned Bacillus subtilis gene that inhibits sporulation inmultiple copies. J. Bacteriol. 168:860-869.

8. Goff, S. A., and A. L. Goldberg. 1987. An increased content ofprotease La, the lon gene product, increases protein degrada-tion and blocks growth in Escherichia coli. J. Biol. Chem.262:4508-4515.

9. Gottesman, M., and M. Yarmolinsky. 1968. Integration-negativemutants of bacteriophage lambda. J. Mol. Biol. 31:487-505.

10. Gottesman, S. 1987. Regulation by proteolysis, p. 1313-1317. InF. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M.Schaechter, and H. E. Umbarger (ed.), Escherichia coli andSalmonella typhimurium: cellular and molecular biology. Amer-ican Society for Microbiology, Washington, D.C.

11. Gottesman, S., P. Trisler, and A. S. Torres-Cabassa. 1985.Regulation of capsular polysaccharide synthesis in Escherichiacoli K-12: characterization of three regulatory genes. J. Bacte-riol. 163:1111-1119.

12. Gottesman, S., and D. Zipser. 1978. Deg phenotype of Esche-richia coli lon mutants. J. Bacteriol. 113:844 851.

13. Huisman, O., and R. D'Ari. 1981. An inducible DNA replica-

tion-cell division coupling mechanism in Escherichia coli. Na-ture (London) 290:797-799.

14. Huisman, O., R. D'Ari, and S. Gottesman. 1984. Cell divisioncontrol in Escherichia coli: specific induction of the SOSfunction SfiA protein is sufficient to block septation. Proc. Natl.Acad. Sci. USA 81:4490-4494.

15. Jones, C., and I. B. Holland. 1985. Role of the SulA (FtsZ)protein in division inhibition during the SOS response in Esch-erichia coli: FtsZ stabilizes the inhibitor SulA in maxicells.Proc. Natl. Acad. Sci. USA 82:6045-6049.

16. Kanaya, S., and R. J. Crouch. 1984. The rnh gene is essential forgrowth of Escherichia coli. Proc. Natl. Acad. Sci. USA 81:3447-3451.

17. Katayama, Y., S. Gottesman, J. Pumphrey, S. Rudikoff, W. P.Clark, and M. R. Maurizi. 1988. The two-component, ATPdependent Clp protease of Escherichia coli: purification, clon-ing, and mutational analysis of the ATP-binding component. J.Biol. Chem. 263:15226-15236.

18. Kohara, Y., K. Akiyama, and K. Isono. 1987. The physical mapof the whole E. coli chromosome: application of a new strategyfor rapid analysis and sorting of a large genomic library. Cell50:495-508.

19. Little, J. W., and D. W. Mount. 1982. The SOS regulatorysystem of Escherichia coli. Cell 29:11-22.

20. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

21. Maurizi, M. R., P. Trisler, and S. Gottesman. 1985. Insertionalmutagenesis of the lon gene in Escherichia coli: lon is dispens-able. J. Bacteriol. 164:1124-1135.

22. Meeks-Wagner, D., J. S. Wood, B. Garvik, and L. H. Hartwell.1986. Isolation of two genes that affect mitotic chromosometransmission in S. cerevisiae. Cell 44:53-63.

23. Messing, J., R. Crea, and P. H. Seeburg. 1981. A system forshotgun DNA sequencing. Nucleic Acids Res. 9:309-321.

24. Miller, C. G. 1987. Protein degradation and proteolytic modifi-cation, p. 680-691. In F. C. Neidhardt, J. L. Ingraham, K. B.Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.),Escherichia coli and Salmonella typhimurium: cellular andmolecular biology. American Society for Microbiology, Wash-ington, D.C.

25. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

26. Mizusawa, S., and S. Gottesman. 1983. Protein degradation inEscherichia coli: the lon gene controls the stability of SulAprotein. Proc. Natl. Acad. Sci. USA 80:358-362.

27. Saito, H., Y. Nakamura, and H. Uchida. 1978. A transducinglambda phage carrying grpE, a bacterial gene necessary forlambda DNA replication, and two ribosomal protein genes, rpsP(S16) and rpsL (L19). Mol. Gen. Genet. 165:247-256.

28. Silhavy, T. J., M. L. Berman, and L. W. Enquist. 1984.Experiments with gene fusions. Cold Spring Harbor Labora-tory, Cold Spring Harbor, N.Y.

29. Torres-Cabassa, A. S., and S. Gottesman. 1987. Capsule synthe-sis in Escherichia coli K-12 is regulated by proteolysis. J.Bacteriol. 169:981-989.

30. Trisler, P., and S. Gottesman. 1984. lon transcriptional regula-tion of genes necessary for capsular polysaccharide synthesis inEscherichia coli K-12. J. Bacteriol. 160:184-191.

31. Walker, G. C. 1987. The SOS response of Escherichia coli, p.1346-1357. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B.Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Esche-richia coli and Salmonella typhimurium: cellular and molecularbiology. American Society for Microbiology, Washington, D.C.

32. Way, J. C., M. S. Davis, D. Morisato, D. E. Roberts, and N.Kleckner. 1984. New TnlO derivatives for transposon mutagen-esis and for construction of lacZ operon fusions by transposi-tion. Gene 32:369-379.

33. Wyman, A. R., L. B. Wolfe, and D. Botstein. 1985. Propagationof some human DNA sequences in bacteriophage X vectorsrequires mutant Escherichia coli hosts. Proc. Natl. Acad. Sci.USA 82:2880-2884.

VOL. 171, 1989

on Septem

ber 9, 2020 by guesthttp://jb.asm

.org/D

ownloaded from