Molec. gen. Genet. 179, 359 368 (1980)

[qlGG

© by Springer-Verlag 1980

Immunological and Genetic Properties of Escherichia coil K12 argE Mutants A.E. Kadikiran* and S. Baumberg Department of Genetics, University of Leeds, Leeds LS2 9JT, England

Summary. C R M + nonsense mutations, as described for E. coli K12 lacZ, in theory permit determination of the direction of transcription of an isolated gene. In the hope of utilising this approach to confirm that the E gene of the E. cob K12 argECBH cluster is transcribed in the opposite direction from the C B H unit, 30 argE mutations were investigated immunologically and genetically. Only three, E l , E25 and E26 (which map close together towards the left-hand end of the gene), were found to be CRM+. Co-suppression of each argE mutation with known strongly polar lacZ amber, ochre and U G A mutations was looked for to distinguish missense from nonsense argE's, Arg + revertants being screened for Lac + and Mel + phenotypes. Of 16 mutations not hitherto characterised as nonsense, frameshift or large deletions, only three, E l , E25 and E26, appeared to be missense. Three of the nonsense mutations were also streptomycin suppressible. It appears, therefore, that among argE mutants so far studied the correlation between CRM + and missense is complete, so that the projected method for determining the direction of transcription of argE could not be applied.

Introduction The argECBH cluster of Escherichia coli K12 has been reported as showing divergent transcription from a common control region at the E - C boundary (Jacoby 1972; Elseviers et al. 1972; Pouwels et al. 1974). When the work to be described here was commenced, the evidence for divergent transcription of argE and argCBH was incomplete. In particular, although polar mutations had shown that the latter three genes are transcribed onto a single polycistronic m R N A molecule in the direction CBH, the direction of tran* Present Address: Biyoloji B61timfi,Bo[~azigiUniversity. Istanbul, Turkey Offprint requests to." Dr. S. Baumberg

scription of the lone E gene could not be established in the same manner (Cunin et al. 1969; Baumberg and Ashcroft 1971). A possible method for determining the direction of transcription of an isolated gene was, however, suggested by the work of Zabin's group (Fowler and Zabin 1968; Berg etal. 1970) on the immunological properties of E. coli K12 laeZ mutants. These workers have shown that many lacZ nonsense mutants are detectably CRM +, i.e. contain material which cross-reacts with anti-/~-galactosidase antibody. Antibody binding capacity does not increase monotonically with chain length. However, in principle the molecular weight of any CRM + lacZ nonsense mutant product could be determined by use of its cross-reacting ability to locate it in e.g. a sucrose density gradient or fractions from a chromatographic column sieving on the basis of molecular size. If a similar situation were to apply for chain termination mutants of some other, isolated, gene, then correlation of molecular weight with map position would indicate the direction of transcription of that gene (it may be noted that this should be the case for eukaryotes as well as prokaryotes). It is true that for most bacterial systems studied, chain termination mutants are invariably C R M - (Yanofsky 1967, and references therein; Jargiello et al 1974). However, it seemed worth attempting this method for the E. coli K12 argE gene for the following two reasons : (i) it is probable (though not certain) that the AcOase 1 enzyme encoded by argE is a monomer of molecular weight approximately 99 kilodaltons (Kadikiran 1977), this large size possibly increasing the chances that some chain termination mutants might be CRM + ; and (ii) preliminary studies (Ashcroft 1970: J.N. Wood, unpublished results) suggested that certain argE frameshift mutants were CRM+. We present here results on the immunological properties of a collection of argE mutants whose missense vs. nonsense character was determined. It is shown that the suggestions that some argE frameshift mu1 Abbreviation: AcOase, acetylornithinase

0026-8925/80/0179/0359/$02.00

360

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE Mutants

Table 1. E. coli strains Strain

Genotype

P4X P4XB2 P4XG P4XB2G P4XSG P4XB2SG P4XGN P4XB2K1 CA7049 CA150 2414

Hfr Hfr Hfr Hfr Hfr Hfr Hfr Hfr Hfr Hfr Hfr

(P4X) metB (P4X) metB argR (P4X) ppc (P4X)ppc argR (P4X) ppc strA (P4X) ppc argR strA (P4X) ppc nal (P4X) argR (H) thi lacZ (U131 ; UAG) (H) thi lacZ (" O°"; UAA) (H) thi trp (UAG) lacZ (UGA)

Source/reference Cunin et al. (1969) Cunin et al. (1969) This laboratory (Met + ppc transductant of P4X) This laboratory (Met + ppc transductant of P4XB2) Strr mutant of P4XG Strr mutant of P4XB2G NaY mutant of P4XG Met + transductant of P4XB2 J.G. Scaife J.G. Scaife M.M. Howe

Strains carrying known argE nonsense mutations E18 E21 HC-D-20-15 E24 P4XB2E34

Hfr (P4X) argE18 F - metB argE21 strA Hfr (P4X) metB argE23 F - metB thi purD argE24 strA Hfr (P4X) argE34 argR

MG183 MG343 MG350

argE183 aroE24 his argR15 strA40 argE343 aroE24 his argR15 strA40 argE350 aroE24 his argR15 strA40

Cunin et al. (1969) Cunin et al. (1969) Cunin et al. (1969) Cunin et al. (1969) This laboratory (Met + argE34 transductant of P4XB2); Cunin et al. (1969) Jacoby (1972) Jacoby (1972) Jacoby (1972)

Strains carrying known argE frameshift mutations P4XB2E171 P4XB2E176 P4XB2E179 P4XE 181 MG377 MG381 MG387

Hfr (P4X) aidE171 argR Hfr (P4X) argE176 argR Hfr (P4X) argE179 Hfr (P4X) argE179 argE377 aroE24 his argRI5 strA40 argE381 aroE24 his argR15 strA40 argE387 aroE24 his argR15 strA40

Baumberg and Ashcroft (1971) Baumberg and Ashcroft (1971) Baumberg and Ashcroft (1971) Baumberg and Ashcroft (1971) G.A. Jacoby (personal communication) Jacoby (1972) G.A. Jacoby (personal comminication)

Strains carrying deletions known to extend over part or all of argE MN41 MN42 DS11 S1D2 P4XSMN42 P4XB2SMN42

Hfr (P4X) metB AMN41 (ppc-argECBH) Hfr (P4X) metB AMN42 (ppc-argECBH) Hfr (P4X) metB AIO0 (ppc-argE) F - AIO1 (ppc-argE) Hfr (P4X) metB AMN42 strA Hfr (P4X) AMN42 strA

Cunin et al. (1969) Cunin et al. (1969) Elseviers et al. (1972) Elseviers et al. (1972) Stff mutant of MN42 Strr mutant of Met + MN42 transductant of P4XB2

Strains carrying uncharacterised argE point mutations SB166 R19 PA201 R15 2-4E5 RE11 P4XE13 El4 20-BTP-67 10-BTP-67 RE20 E25 MN35 P4XB26169 6196 RE96

Hfr (P4X) metB argE1 Hfr (P4X) argE2 F - argE3 Hfr (P4X) metB argE5 F metB his thi purD ppc argE5 strA F - rnetB his ile argEll argR strA Hfr (P4X) metB argE13 Hfr (P4X) metB argE14 Hfr (P4X) metB argE15 Hfr (P4X) metB argE16 F - metB his ile argE20 argR strA Hfr (P4X) metB argE25 Hfr (P4X) metB argE26 Hfr (P4X) argE69 argR F his ile metB argE96 F - his ile metB argE96 argR

R. Cunin and N. Glansdorff R. Cunin and N. Glansdorff R. Cunin and N. Glansdorff R. Cunin and N. Glansdorff R. Cunin and N. Glansdorff H.J. Vogel This laboratory; Ashcroft (1970) R. Cunin and N. Glansdorff R. Cunin and N. Glansdorff R. Cunin and N. Glansdorff H.J. Vogel R. Cunin and N. Glansdorff R. Cunin and N. Glandsdorff Met + argE69 transductant of P4XB2a Baumberg et al. (1965) H.J. Vogel

a argE69 is the allele number used in this laboratory for the mutation argE3 present in the strain A B l l l 5 and its derivatives see Bachmann (1972)

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE M u t a n t s

361

&[p[x;argE)101

Growth and Harvesting of Cultures, Enzyme and Protein Assays, and Transduction and Conjugation Techniques were as in Bretscher

&lppcorg E]IO0

and Baumberg (1976). 346, ~1 183,343, 2

61 1

21 3 25 23 5 2; 15 16 2Z, 15 34 (,l,)

[

I,,,,,,,,I,I ....

ppc ~ --

<

algE

>

promoleroperotor region

~ - argC - ~ --

Fig. 1. Positions of accurately- or roughly-mapped mutations in argE (Cunin et al. 1969; Jacoby 1972; Elseviers et al. 1972; Cunin 1973). The mutation El4 previously thought to be the rightmost within argE now appears to owe this location from three-point crosses to an unusually strong marker effect (Bretscher and Baumberg 1976; G. Freundlich and N. Glansdorff, personal communication)

tants might be CRM + were incorrect, and that no

argE chain termination mutants examined were detectably CRM +. The method outlined above for determining direction of transcription could not therefore be applied. A preliminary report of this work has been published (Baumberg and Kadikiran 1977). Materials and Methods Bacterial Strains. The strains used are listed in Table 1. All were derivatives of E. coli K12 except for the M G series which are derivatives of E. coli B. The m a p positions of the argE mutations and adjacent markers are shown in Fig. 1. Two sets of derivatives of these were constructed for this study. (1) 32 argE mutations, all those mentioned in Table 1 except MN41 and E176, were transferred to the P4X and P4XB2 backgrounds for C R M estimation. This was done by transduction; P1 was grown on an argE m u t a n t being used to transduce P4XG, P4XB2G, P4XSMN42 or P4XB2SMN42 to Ppc + argE; or by conjugation between an Hfr(P4X) argE donor and P4XSG, P4XB2SG, P4XSMN42 or P4XB2SMN42 in the " F - phenocopy" state, Ppc + argE St~ transconjugants being selected. The strains were checked for argE by growth tests (response to ornithine but not acetylornithine) and enzyme assay, and for argR+/argR - by assay of the enzymes ornithine carbamoyl transferase and argininosuccinase. (2) The 15 uncharacterised argE mutations, six of the argE nonsense mutations (all except E2I and E183), and five of the argE frameshift mutations (all except E171 and E179) were combined with a set of three strongly polar lacZ mutations for cosuppression studies. Nal r derivatives of Hfr(H) strains CA7049, CA150 and 2414, containing respectively lacZ amber, ochre and U G A mutations, were isolated. Also, Rig derivatives of the argE mutations in a P4X or P4XB2 background, obtained as in (1) above, were isolated. The Hfr(H) lacZ Nal r strains were crossed as donors with the P4X/P4XB2 argE Rig strains in the " F phenocopy" state. After allowing mating to proceed for 2 h at 37 ° C, mating mixtures were in most cases diluted 10-fold into nutrient broth and incubated overnight to allow phenotypic expression of the nal marker, Nal r Rig transconjugants then being selected. In a few cases, where the argE parent was already Nal ~ but M e t - , Met + Rig transconjugants were selected immediately after mating. In all cases, argE laeZ Str ~ isolates were taken for the co-suppression studies (as str alleles m a y interfere with suppression). Full details of the strain constructions will be found in Kadikiran (1977).

Preparation of Anti-AcOase Antibody. Adult male New Zealand white rabbits were injected intrascapularly (twice weekly for at least three weeks) with 2.5 ml of an approximately 1 m g / m l protein solution either a crude extract of strain P4XB2KI or a partially (10- to 30- fold) purified AcOase preparation therefrom (Kadikiran 1977 after Vogel and McLellan 1970) - mixed with an equal volume of Freund's adjuvant. The animals were bled from the marginal vein of the ear. The blood was allowed to clot; the serum was clarified by centrifugation and was then dialysed overnight against 18% (wt/vol.) N a 2 SO,. The precipitate (largely IgG) was washed with 18% Na2 SO4, then redissolved in the m i n i m u m possible quantity of 0.1 M potassium phosphate buffer, pH 7.0, containing reduced glutathione at 1 m m (P/G buffer). It was finally dialyzed against the same buffer and stored frozen at - 1 5 ° C. Control sera were also obtained similarly from animals prior to immunisation. Inhibition of AcOase with Anti-AcOase Antibody. In all cases, 0.3 ml of an appropriate dilution in P/G buffer of a crude extract of P4XB2K1 was mixed with 0.3 ml of a solution containing antibody (see below), or with 0.3 ml P/G buffer as a control. The mixture was incubated at 37 ° C for 30 rain, and was then centrifuged at 4,000 r.p.m, for 15 min on a bench centrifuge. The supernatant was assayed for AcOase activity. In preliminary experiments to observe the interaction between AcOase and antibody (Results Section 1), solutions of antibody (including controls from unimmunized animals) in P/G prepared as above were used. In the immunological characterisation of argE mutants (Results Section 2), the following protocol was adopted. 0.3 ml m u t a n t extract adjusted to 4 m g protein/ml was mixed with 0.3 ml antibody solution in P/G prepared as above. The mixture was incubated at 3 7 ° C for 25 rain, and was then centrifuged at 4,000 r.p.m, for 15 rain on a bench ~entrifuge. 0.3 ml supernatant was then added to 0.3 ml diluted P4XB2K1 extract and the aforementioned procedure followed. Results are described involving three batches of antibody: B1M3, prepared against partially purified AcOase; B3P, a pool of antibody preparations against both partially purified AcOase and a crude extract of P4XB2K1; and UI, from unimmunized animals.

Co-suppression Tests. argE/lacZ (nonsense) strains prepared as above were grown up overnight in nutrient broth with shaking, and washed twice with and finally resuspended in minimal salts medium. 0.1-0.5 ml samples were then plated on minimal-salts glucose agar supplemented as appropriate with thiamine, methionine and/or tryptophan but without arginine or a precursor thereof, to select Arg + revertants. In some cases, a drop (0.05 rnl) of Nmethyl-N'-nitro-N-nitrosoguanidine (NG) solution (100 ~g/ml) or of ethyl methanesulphonate (EMS) was placed at the centre of the plate. Plates were incubated for ~ 5 days at 30 ° C, to permit growth of temperature-sensitive revertants. They were then replica plated on to minimal agar similarly supplemented but containing lactose or melibiose in place of glucose, and on to supplemented minimal salts-glucose plates as a control. Glucose and lactose plates were incubated at 30 ° C, melibiose plates at 3 7 ° C and 42 ° C; in all cases incubation was permitted for two days. Arg ÷ revertants in which cosuppression to Lac + or Mel + was occurring gave uniformly solid replicas, while the others gave thin, flat colonies which occasionally contained solid papillae of Lac + or Mel ÷ secondary revertants. Test for Suppressibility of argE Mutations by Streptomycin. The argE/lacZ (nonsense) strains were patched on minimal salts-glucose

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE M u t a n t s

362

16

----T--

I

I

I

m

I

c

I

I

I

I~

'

o 12 L

.=_

.£ 40

-~

E

8 " 30

.gz

2; 6

>,

' ; 20

a -o

~10

2 --

0

10

20 30 Vo(. antibody (~JI)

z.o

50

0

5

10

15

20

25

Vol. extract added (pJl)

Fig. 2. Loss of AcOase activity on adding increasing a m o u n t s of antibody to a fixed a m o u n t of E. call K12 extract. The indicated a m o u n t s of antibody preparation (B1M3) were made up to 0.3 mI in the P / G buffer and these added to 0.3 ml diluted P4XB2K1 extract, as described in Materials and Methods. The activity in the resulting supernatant is given in units/ml

Fig. 3. Increase in AcOase activity on adding increasing a m o u n t s of P4XB2K1 extract to a fixed a m o u n t of antibody. The indicated volume of P4XB2K1 extract was made up to 0.3 ml with P/G buffer, and this was added to 0.3 ml antibody preparation (B1M3) as described in Materials and Methods or to P/G buffer as a control. The activity in the resulting supernatant is given in units/ ml. - o ©-, extract+ antibody; - e - e - , extract + P/G buffer (control)

agar supplemented with ornithine and, as required, thiamine, methionine and/or tryptophan. These plates were replicated on to similar media without ornithine but containing streptomycin sulphate at 0.125, 0.25, 0.5 or 0.75 pg/ml, with appropriate controls. The replica plates were scored after two days incubation at 37 ° C.

to the initial activity or as it approaches zero. A plot of residual AcOase activity against volume of crude P4XB2K1 extract added progressively to a fixed volume of IgG solution is shown in Fig. 3. The vertical distance between the line for the control without IgG and the line for the assay with IgG may be taken as a measure of the anti-AcOase antibody activity of the IgG solution.

Results

l. Interaction of AcOase and Anti-AcOase Antibody On incubation at 37 ° C of crude E. coli extracts or partially purified protein fractions containing AcOase activity with IgG preparations derived from the sera of rabbits inoculated with crude E. coli extracts or partially purified protein fractions of high AcOase activity as described in Materials and Methods, AcOase activity was rapidly lost. At least 90% of the loss in activity took place over the first minute (data not shown). It is assumed that the (AcOase) - (anti-AcOase IgG) complex ended up in the precipitate. A plot of residual AcOase activity against volume of IgG solution added progressively to a fixed volume of crude P4XB2K1 extract is shown in Fig. 2. The relationship differs significantly from the linear decrease that represents the simplest relationship of this type, only where the residual activity is close

2. Immunological Characterisation of argE Mutants The presence in crude bacterial extracts of material cross-reacting with anti-AcOase IgG may in principle be determined as follows. Mutant extracts are treated with excess IgG; then extract containing excess AcOase activity is added, and the residual AcOase activity determined. If this residual activity is greater than in a control where the mutant extract is omitted, then the extract contains CRM. However, preliminary experiments (data not shown) indicated that unless extracts were compared under rigorously controlled condition, unexplained and irreproducible fluctuations in measured final activity were obtained. These appeared to result from the sensitivity of AcOase activity (itself rather labile) and/or its interaction with

363

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE Mutants

Table 2. CRM activity of argE mutant extracts Enzyme activity (units/ml) of final supernatant"

Mutation

argR +

argR-

background

background

12 Deduced CRM status

~B

~6

Deletions AMN41 (ppc - argECBH) AMN41 (ppc argECBH) (control) b AMN42 (ppc - argECBH) AIO0 (ppc - argE) AIOI (ppc argE)

g el ¢:

3.5 10.3 1.5 2.3 2.3

2.0 2.2 2.8

CRMCRM CRM-

1 o

Known nonsense mutations E18 E21 E23 E24 E34 E183 E343 E350

2.8 1.8 3.4 2.4 2.3 2.3 4.2 2.7

2.4 1.8 2.0 1.9 2.4 2.2 2.9 2.8

CRMCRM CRM CRMCRMCRMCRMCRM

2.4 4.7 2.0 2.3 1.9 1.6

2.5 2.8 2.3 2.9 1.3 3.4

CRM C RM CRM CRMCRMCRM

8.5 2.4 2.7 4.4 3.7 2.4 2.2 2.8 2.9 1,8 4,1 2,5 3,2 2,5 2.2

12.0 3.0 2.3 3.9 3.4 2.3 2.2 3.5 3.4 2.3 3.4 8.6 6.9 1.5 2.9

CRM + CRMCRMCRMCRMCRMCRMCRM CRMCRMCRMCRM + CRM + CRMCRM

L

I

[

I 2 3 Protein concentralion (rng/rnl)of rnulant extract

Fig. 4. Increase in residual AcOase activity with protein concentration of argE mutant extract used to treat antibody. The procedure was identical to that given under Table 2, except in that varying protein concentrations of argE mutant extract were employed; the mutations were present in the argR- background (except for the AMN41 control), o - o - , El," - o - o - , E25; - A - A - , E26; -zx-z~-, AMN41

Known frameshift mutations E171 E179 E181 E377 E381 E387

Uncharacterized mutations E1 E2 E3 E5 E6 Ell El3 E14 El5 E16 E20 E25 E26 E69 E96

a 0.3 ml mutant extract diluted to 4 mg protein/m1 with P/G buffer was treated with 0.3 ml antibody preparation (a three-fold dilution of B3P in P / G buffer) as described in Materials and Methods. 0.3 ml supernatant was added to 0.3 ml of a 12-fold diluted P4XB2K1 extract, the procedure in Materials and Methods again followed, and the AcOase activity in the supernatant determined b Here the same protocol was adhered to except that antibody preparation UI, from unimmunized rabbits, was employed

IgG, to minute details of procedure. It is likely that fluctuations of this type were responsible for the ear-

lier indications that certain argE frameshift mutants were CRM ÷. The procedure described in Materials and Methods was therefore strictly adhered to. It was also found that differences in genetic background of the argE mutants contributed to these fluctuations. Consequently all argE alleles were put into the same pair of backgrounds, namely P4X and P4XB2 which are argR + and argR- but otherwise isogenic (Cunin etal. 1969); details of strain construction are given in Table 1. Levels of CRM, in cultures grown on exogenous arginine, should be repressed in the former and derepressed in the latter. Results are tabulated in Table 2. The criteria adopted for recognition of a mutant allele as CRM ÷ were (i) residual activity in argR- background greater than 3.5 units/ml (that found with the control containing an extract of MN41, whose deletion spans the entire argECBH cluster), and (ii) residual activity in argR + background less than in argR- background. By these criteria, only three alleles emerge as CRM ÷, namely El, E25 and E26. The determinations were repeated with varying volumes of extracts of P4XB2E1, P4XB2E25 and P4XB2E26 (Fig. 4). The plots confirm that these strains are CRM ÷ and indicate that the relative amounts of CRM per ml extract decrease in that sequence. From the linear portions of these plots (i.e. at low concentration of mutant extracts), the amounts of CRM per mg protein can be roughly quantified in terms of units of AcOase activity protected, as four times the slope of this linear portion. The

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE Mutant~,

364

Table 3, Co-suppression of argE mutations with known lacZ nonsense mutations a

argE mutation

Nature of lacZ mutation

Mutagen None

NG

Total Arg + revertants

Lac +

LacMel +

Total Arg + revertants

EMS Lac +

LacMel +

Total Arg + revertants

Lac +

LacMel +

E1

UAG UAA UGA

431 72 207

0 0 0

0 0 0

2 17 9

0 0 0

0 0 0

9 17 7

0 0 0

0 0 0

E2

UAG UAA UGA

24 16 9

5 3 0

6 11 0

4 1 1

0 1 0

1 0 0

4 9

0

2 0

E3

UAG UAA UGA

None None None

19 35 23

0 0 0

0 0 1

1 1 2

0 0 0

0 1 2

E5

UAG UAA UGA

1 None None

0

0 0

2 0

E6

UAG UAA

2 2

0 0

0 0

0 2

Ell

UAG UAA UGA

None None None

El3

UAG UAA UGA

32 16 50

6 1 0

0 10 0

El4

UAG UAA UGA

8 8 3

0 0 0

3 3 0

52 70 27

0 0 0

0 6 0

El5

UAG UAA UGA

12 12 3

0 0 0

0 0 0

119 97 13

0 0 0

0 0 2

El6

UAG UAA UGA

14 None 17

1

7

0

0

8 12 1t

0 0 0

6 0 0

El8

UAA UGA

13 None

3

0

E20

UAG UAA UGA

4 5 1

0 1 0

0 0 0

E23

UAG UAA

14 None

10

0

E24

UAG UAA UGA

153 13 14

127 2 0

4 0 0

E25

UAG UAA UGA

3 2 None

0 0

0 0

34 41 26

0 0 0

0 0 0

None None None

E26

UAG UAA UGA

32 2 None

0 0

0 0

25 18 7

0 0 0

0 0 0

6 1 2

E34

UAG UAA

None None

48 92

28 5

0 0

E69

UAG UAA UGA

5 15 69

0

0 0

None Nohe 3

0

0

None 2 2

5 12

0 0

0 3

3 6

None None None

5 2 0

0 0 0

None None None

None None None

None None None

0 0 0

0 0 0

365

A.E. Kad ikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE Mutants

Table 3 (continued)

argE

Natu re of

m u tation

lacZ mutation

Mutagen None Total Arg + revertants

E96

NG Lac +

LacMel +

EMS

Total Arg + revertants

Lac +

LacMel +

Total Arg + revertants

Lac +

LacMel +

UAG UAA UGA

None 20 21

5 0

0 0

E176

UAG UGA

2 59

0 0

2 0

None

2

0

0

E181

UAG

2

2

0

None

7

2

3

E343

UAG UAA

16 3

4 0

1 1

E350

UAG UAA

300 137

2 61

0 0

E377

UAG UAA UGA

12 19 4

0 0 0

0 0 0

E38I

UAG UAA UGA

200 150 None

0 0

0 0

E387

UAG UAA UGA

1 None None

0

0

a

N one N one None

The procedure is described in Materials and Methods

amounts of CRM come out as 38, 19 and 8 units/mg for P4XB2EI, P4XB2E25 and P4XB2E26 respectively; these figures may be tentatively compared to the specific AcOase activity of P4XB2K1, characteristically within the range 40 80 units/mg.

3. Characterization of argE Mutations as missense or Nonsense It was next necessary to determine the missense/nonsense nature of the as yet uncharacterized argE mutations, since if at least two of the detectably CRM-alleles El, E25 and E26 proved to be nonsense, the method of determining the direction of transcription of argE outlined in the Introduction could be applied. In attempting to distinguish between missense and nonsense mutants, it was felt unsatisfactory merely to introduce known suppressors and look to see whether any of these restored functional AcOase activity in the argE mutant under test. It is quite possible that a given suppressor may cause the insertion of an amino acid that does not permit the enzyme to function, and in principle one can only surmount this problem by using sets of suppressors (one set for ochre mutations, since they will also suppress

ambers, and another for the U G A triplet) which introduce all the possible amino acids in each case. The simpler method used here was to look for cosuppression of each argE mutation with known nonsense mutations, and to overcome the above difficulty in regard to the latter by looking not only for restoration of function but also for loss of polar effect the latter not of course requiring anything other than suppression of chain termination. For these reasons, double mutations were constructed carrying each argE mutation in turn together with the strongly polar lacZ mutations contained in the suppressor-free Hfr H strains CA7049 (U131, amber), CA150 (" o 2'' ochre), and 2414 (an uncharacterised UGA). The argE mutations had already been obtained in the suppressor-free P4X or P4XB2 backgrounds to test for CRM, as described in the previous section. The resulting strains were in all cases Lac ÷, in most cases NaP Str s, and in some cases M e t - . A spontaneous R i f derivative was selected from each argE mutant, and Nag derivatives were obtained from the three Hfr H lacZ (nonsense) strains. The crosses Hfr H lacZx argE Rig were then performed, and recombinants selected as Nag Rig or (for those argE's available only in an already Nal r background) as

366

A.E. Kadikiran and S. Baumberg: Immunologicaland Genetic Properties of E. coli K12 argEMutants

Met + RiP, as described in Materials and Methods. Recombinants carrying argE, L a c - and Stl~ as unselected markers were taken for the reversion studies described below - Str ~ because it has been shown (Gorini 1971) that suppression may be severely affected in a Str r background. Some combinations proved refractory to isolation and will therefore be found missing from Table 3. As described in Materials and Methods, Arg + revertants of these strains were then selected in 25 days' growth on salts-glucose agar plates supplemented with thiamine, methionine and tryptophan as required but without arginine or ornithine, at 30 ° C (to include possible temperature sensitives) and in some cases with the addition of the mutagens N G or EMS to the centre of the plate. These plates were then (if they showed revertant colonies) replica plated directly on to similarly supplemented minimal plates containing as carbon source lactose, melibiose or glucose (as a control). Glucose and lactose replica plates were incubated at 30 ° C, melibiose plates at 37 ° C. At and above this temperature, melibiose uptake depends on the galactoside permease product of lacY; this protein will be absent if the strongly polar lacZ's are unsuppressed, but will be present if co-suppression has occurred, even if the /~-galactosidase produced is non-functional (as shown by absence of growth on minimal lactose medium). With the considerable time allowed for growth of the Arg + revertant colonies, these had grown to considerable size. It was therefore easy to distinguish their replicas on minimal lactose or melibiose where co-suppression had occurred, these being uniformly solid, from the infrequent cases where a mutation of Lac ÷ or Mel ÷ (whether same-site or suppressor) had occurred as a secondary event during growth of the Arg + revertant; the replica here showed generally thin growth, typical of the L a c - or Mel- phenotype, with a small solid papilla at some point. The occurrence of co-suppression is shown in Table 3, the characterised argE nonsense and frameshift mutations being included as controls. It may be noted that the technique used here does not distinguish between argE amber or ochre mutations. As far as previously characterised argE mutations are concerned, the amber mutations E18, E24, E34, E343 and E350 are clearly confirmed as such by frequent co-suppression. The frameshift mutations E176 and E181 unexpectedly also showed co-suppression, although E377, E381 and E387 did not. This phenomenon may relate to, for instance, the suppressor reported by Atkins and Ryce (1974) that suppresses both U G A and certain frameshift mutations. Of the hitherto uncharacterized argE mutations, E2, E5, E6, E13, El4, El6, E20, E23, E69 and E96 showed co-

suppression with the lacZ amber and/or ochre themselves. E15 showed co-suppression only with the lacZ U G A mutation and is therefore probably UGA. E3 showed co-suppression with both the ochre and U G A lacZ mutations. This behaviour is far from unique; many instances of co-suppression of U G A with amber and/or ochre mutations are given for example by Jargiello et al. (1974). E11 could not be made to revert and may be a small deletion. Finally, El, E25 and E26 gave no co-suppression with the lacZ nonsense mutations out of many revertants tested, and hence are likely to be missense mutations. We see, therefore, that the three CRM ÷ mutations appear to be missense while the C R M - mutations appear to be nonsense. No mutations are revealed that are nonsense but C R M ÷, as would be necessary to determine the direction of transciption of argE.

4. Suppressibility of argE Mutations by Streptomycin It has been suggested (Gorini and Beckwith 1966) that streptomycin-suppressible mutations are usually nonsense; and although it was later stated (Gorini 1970) that this correlation does not hold in that some streptomycin- suppressible mutations are missense, the major piece of evidence cited (Whitfield et al. 1966) does not rule out that the mutations described therein as "missense" are UGA, or indeed amber or ochre though not suppressed by the suppressors tested. It therefore seemed worth while to test the set of argE mutations for suppressibility by streptomycin. The method employed is described in Materials and Methods, all the argE/lacZ (nonsense) combinations available being studied. The only combinations that showed any sign of streptomycin suppressibility were: E13/lacZ (UAG), (UAA), (UGA): slight suppression at 0.125 ~tg/ml streptomycin sulphate, increasing to almost completely Arg + phenotype at 0.75 gg/ml. E69/lacZ (UAA), (UGA): very weak suppression at 0.5 gg/ml and weak suppression at 0.75 gg/ml streptomycin sulphate. E350/lacZ (UAG), (UAA): weak suppression at 0.75 gg/ml streptomycin sulphate. Since E13, E69 and E350 were shown above to be amber or ochre mutations, these results are in accord with the suggestion of Gorini and Beckwith (1966) that streptomycin-suppressible mutations are nonsense. Discussion

Extracts of 32 argE mutants of E. coli K12, three carrying deletions, seven known nonsense mutations,

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE Mutants six frameshift mutations, and 16 mutations of unk n o w n type, were examined under carefully controlled conditions for the presence of C R M , detected by its reduction of the k n o w n AcOase-inactivating power of an I g G standard. N o n e of the m u t a n t s carrying deletions, k n o w n nonsense mutations, or frameshifts showed detectable C R M . O f the 16 u n k n o w n mutations, only three, El, E25 and E26, were C R M + A r o u g h estimate indicated that the a m o u n t s of AcOase-like protein in extracts corresponded to approxinately 50-100%, 2 5 - 5 0 % and 10-20% of wildtype respectively. Two points m a y be raised here relating to sensitivity of the m e t h o d s employed. First, m o r e sensitive m e t h o d s such as double diffusion on Ouchterlony plates and immunoelectrophoresis might be able to detect lower levels of C R M if such were present in any of the other m u t a n t extracts. To use these to best advantage, a n t i b o d y to purified AcOase would be necessary, and in spite of repeated attempts a complete purification of AcOase remained elusive. Purifications of up to 150-fold were attained by methods based on those of Vogel and McLellan (1970); the partially purified enzyme was used to raise a n t i b o d y in rabbits, but this gave ambiguous results (Kadikiran 1977). Second, the possibility exists that C R M present in m u t a n t extracts could have been degraded, as is k n o w n to h a p p e n in E. coli and Salmonella typhimurium to nonsense fragments (Goldschmidt 1970; Platt et al. 1970) and some but n o t all missense polypeptides (Zipser and Bhavsar 1976; Bergquist and Trum a n 1978). A n attempt could be made to obviate this by use o f deg- m u t a n t s which have a reduced rate of degradation of nonsense fragments (Bukhari and Zipser 1973; Miller and Zipser 1977). The u n k n o w n argE mutations were then characterized as to missense/nonsense type by looking for their co-suppression with k n o w n lacZ nonsense mutations. 10 showed co-suppression with the lacZ amber or ochre mutations, and so are p r o b a b l y amber or ochre themselves; one was co-suppressed with lacZ ( U G A ) , and so is presumed to be U G A ; one was co-suppressed with the lacZ ochre and U G A , which is ambiguous but n o t unprecedented (Jargiello et al. 1974); one never reverted and m a y be a small deletion; and three, El, E25 and E26 gave m a n y revertants but no co-suppression and so are p r o b a b l y missense mutations. It is these three that were also C R M ÷. It is interesting that these mutations m a p close together near the left-hand end of the argE gene (Fig. 1; Elseviers etal. 1969; Cunin 1973). Such a clustering of missense mutations has been reported in other systems, e.g. the E. coli K12 lacZ gene (Langridge and Campbell 1969) and the deoC gene of Salmonella typhimurium (Jargiello et al. 1974). It is

367

possible that such clusters define the active site, t h o u g h other interpretations are not excluded. All the mutations were tested for streptomycin suppressibility. Only three were streptomycin-suppressible; these, E13, E69 and E350, are all nonsense mutations, in accord with the suggestions of Gorini and Beckwith (1966). The absence of C R M ÷ nonsense mutations a m o n g the set of argE mutations studied prevended determination of the direction of transcription of argE by their use. There is n o w overwhelming physical and genetic evidence that the argECBH cluster is divergently transcribed f r o m a c o m m o n control region at the E-C b o u n d a r y (Jacoby 1972; Elseviers et al. 1972; Pouwels et al. 1974; Panchal et al. 1974; Cunin et al. 1975; Bretscher and B a u m b e r g 1976; Boyen etal. 1978). H o w e v e r the technique p r o p o s e d could prove useful with an organisn~, whether p r o k a r y o t i c or eukaryotic, less amenable to physicochemical extensions of genetics than E. coli. Acknowledgements.We are grateful to Drs. R. Cunin and N. Glansdorff for unpublished information, helpful comments and strains, J. Hewitt for assistance and advice in raising antibody, and Drs. M.M. Howe, G.A. Jacoby, J.G. Scaife and H.J. Vogel for strains. A.E.K. is indepted to the Turkish Government for a grant.

References Ashcroft E (1970) Studies on the arginine cluster in Escherichia coli K-12. Ph D thesis, University of Leeds Atkins JF, Ryce S (1974) UGA and non-triplet suppressor reading of the genetic code. Nature (London) 249 : 527 530 Bachman BJ (1972) Pedigree of some mutant strains of Escherichia coli K12. Bacteriol Rev 26:525 557 Baumberg S, Ashcroft E (1971) Absence of polar effect of frameshift mutations in the E gene of the Escherichia coli argECBH cluster. J Gen Microbiol 69:365-373 Baumberg S, Kadikiran AE (1977) Analysis of Escherichia coli argE mutants. Heredity 39 : 187 Baumberg S, Bacon DF, Vogel HJ (1965) Individually repressible enzymes specified by clustered genes of arginine synthesis. Proc Natl Acad Sci USA 53:1029-1032 Berg AP, Fowler AV, Zabin I (1970) /3-galactosidase: isolation of and antibodies to incomplete chains. J Bacteriol 101:438 443 Bergquist PL, Truman P (1978) Degradation of missense mutant /Lgalactosidase proteins in Escherichia coli K-12. Mol Gen Genet 164:105-108 Boyen A, Charlier D, Crabeel M, Cunin R, Palchaudhuri S, Glansdorff N (1978) Studies on the control region of the bipolar argECBH operon of Escherichia coli. I. Effect of regulatory mutations and IS2 insertions. Mol Gen Genet 161:185-196 Bretscher AP, Baumberg S (1976) Divergent transcription of the argECBH cluster of Escherichia coli. K12. Mutations which alter the control of enzyme synthesis. J Mol Biol 102:205-220 Bukhari AI, Zipser D (1973) Mutants of Escherichia coli with a defect in the degradation of nonsense fragements. Nature (London) New Biol 243:238 241

368

A.E. Kadikiran and S. Baumberg: Immunological and Genetic Properties of E. coli K12 argE Mutants

Cnnin R (1973) Control of gene expression and chromosomal duplications in Escherichia coli K12 : the argECBH gene cluster. PhD thesis, Vrije Universiteit Brussel Cunin R, Boyen A, Pouwels P, Glansdorff N, Crabeel M (1975) Parameters of gene expression in the bipolar argECBH operon of E.coli K12. The question of translational control. Mol Gen Genet 140:51~60 Cunin R, Elseviers D, Sand G, Freundlich G, Glansdorff N (1969) On the functional organization of the argECBH cluster of genes in Escherichia coli K12. Mol Gen Genet 106:3247 Elseviers D, Cunin R, Glansdorff N, Baumberg S, Ashcroft E (1972) Control region within the argECBH gene cluster of Escherichia coli K I2. Mol Gen Genet 117:349 366 Fowler AV, Zabin I (1968)/~-galactosidase: immunological studies of nonsense, missense and deletion mutants. J Mol Biol 33:35 47 Goldschmidt R (1970) In vitro degradation of nonsense fragments in E. coli. Nature (London) 228 : 1151-1154 Go rini L (1970) Informational suppression. Ann Rev Genet 4:107134 Gorini L, Beckwith JR (I966) Suppression. Ann Rev Microbiol 20:401-422 Jacoby GA (1972) Control of the argECBH cluster in Escherichia colL Mol Gen Genet 117:337-348 Jargiello P, Stern MD, Hoffee P (1974) 2-Deoxyribose 5-phosphate aldolase: genetic analyses of structure. J Mol Biol 88:671 691 Kadikiran AE (1977) Escherichia coli acetylornithinase and the direction of transcription of its structural gene. Ph D thesis, University of Leeds Langridge J, Campbell JH (1969) Classification and intragenic posi-

tion of mutations in the fl-galactosidase gene of Escherichia coli. Mol Gen Genet 103:339-347 Miller CG, Zipser D (1977) Degradation of Escherichia coli ~galactosidase fragments in protease-deficient mutants of Salmonella typhimurium. J Bacteriol 130: 347-353 Panchal CJ, Bagchee SN, Gnha A (1974) Divergent orientation of transcription from the arginine gene ECBH cluster of Escherichia coli. J Bacteriol 117:67~680 Platt T, Miller JH, Weber K (1970) In vivo degradation of mutant lac repressor. Nature (London) 228:1154-1156 Pouwels P, Cunin R, Glansdorff N (1974) Divergent transcription in the argECBH cluster of genes in Escherichia coli. K12. J Mol Biol 83:42t 424 Vogel HJ, McLellan WL (1970) Acetylornithinase (Escherichia coli). In : Tabor H and Tabor CW (eds) Methods in enzymology, vol XVII A. Metabolism of animo acids and amines. Academic Press, New York and London, pp 265-269 Whitfield HJ Jr, Martin RG, Ames BN (1966) Classification of aminotransferase (C gene) mutants in the histidine operon. J Mol Biol 21:335-355 Yanofsky C (1967) Structural relationships between gene and protein. Annu Rev Genet 1:117 t38 Zipser D, Bhavsar P (1976) Missense mutations in the lacZ gene that result in degradation of fl-galactosidase structural protein. J Bacteriol 127:1538-1542

Communicated by G. O'Donovan Received April 4, 1980