Molee. Gen. Genetics 106, 32--47 (1969)

On the Functional Organization of the arg ECBH Cluster of Genes in Escherichia coli K-12 R. Cu~IN*, D. ELSEVI~RS**, G. SA~D, G. FREUNDLICH a n d N. GLA~SDO~FF Laboratory of Microbiology, Brussels University, and Research Institute of the C.E.R.I.A., Brussels 7, Belgium Received October 22, 1969 Summary. Among the four seemingly adjacent loci of the argE C B H cluster of E. coli K-12, the last three are shown to belong to the same unit of coordinated expression; the latter exhibits a clockwise polarity in contrast to all other known E. coli operons, except the cluster governing the synthesis of the pyruvate dehydrogenase complex. The analysis of several deletion and nonsense mutants suggests that argE (the expression of which is not strictly correlated with the functioning of the a r g C B H group) has the same polarity but is not integrated with the three other genes into one operon. Between polar arffCB and B mutants the coefficient of repressibility of enzyme H synthesis varies widely. This feature resembles the reduced repressibility of distal gene activity found in polar mutants in the tryptephan operons of E. coli and S. typhimurium but not in the/ac, gal (E. coli) and his (S. typhimurium) operons. Possible implications of the present results and some relevant data that have appeared in the recent literature are discussed.

Introduction T h e s t u d y of t h e f u n c t i o n a l o r g a n i z a t i o n of t h e genes which control arginine biosynthesis in E s c h e r i c h i a coli K-12 has led to t h e conclusion (Maas, 1961) t h a t genes which are s c a t t e r e d on t h e b a c t e r i a l chromosome m a y nevertheless be s u b m i t t e d to a unique r e g u l a t o r y system. F o r this t y p e of e n t i t y , which e x h i b i t s a degree of c o m p l e x i t y p l a c e d i m m e d i a t e l y a b o v e t h e o r g a n i z a t i o n of genes i n t o operons ( J a c o b et a l , 1960) t h e t e r m " r e g u l o n " has been coined (Maas a n d Clark, 1964). T h e arginine (arg) 1 s t r u c t u r a l genes are m o s t p r o b a b l y s u b j e c t to c o n t r o l of t h e n e g a t i v e t y p e (ibidem ; Maas et al., 1964), t h r o u g h i n t e r a c t i o n of a u n i q u e reprcssor with t h e several o p e r a t o r s i n v o l v e d in t h e expression of each i n d i v i d u a l gene or group of genes. U p to now, t h e search for a possible i n v o l v e m e n t of a r g i n y l t r a n s f e r I~NA (or t h e corresponding a c t i v a t i n g enzyme) in t h e repressive process has given n e g a t i v e results (Boman, B o m a n a n d Maas, 1961; H i r s h f i e l d et al., 1968) a n d t h e chemical i d e n t i t y of t h e repressor as well as of t h e corepressor remains unknown. * Research fellow of the Institute for the Encouragement of Scientific Research in Industry and Agriculture, Belgium. ** Research fellow of the National Fund of Scientific Research, Belgium. 1 Abbreviations used. arg = arginine; glu = glucose; lac -~ lactose; leu ~ leueine; met = methionine; pro ~ proline; thl = thiamine; thr = threonine; try = tryptophane; 8tr ~ streptomycine (r or s: resistant or sensitive); H]r ~ male, donor; F ' = intermediate donor type; - ~ - ~ female, recipient; OTC = ornithine carbamoyl transferase; NG = Nmethyl-N-nitroso-N1-nitroguanidine; DES = diethyl sulfate.

Cluster of arg Genes in Escheriehia eoli thr.

o,gEcB,

33

leu

"%rgF

/

l

(argD) str%araR / arg G-~k,~ /~ arg A " ~ - . . ~ N h ] : r g S

glutamate A.~N_acetyl_ ~ B N_acetyl_ C glutamate glutamy[phosphate N-acetylD glutamate semi-aldehyde

N-acety[- E ~ ornithine ornithine

F 5 J[ ~, citrulline ~

arginino- H succinate ~arginine

Fig. 1. Gene-enzyme relationship in the arg pathway of E. coli K-12; argR = regulatory gene; argS = structure~l gene of the arginine activating enzyme (Hirshfield et al., 1968) In spite of the existence of a unique repressor, the expression of the arg genes is not strictly coordinated; (Maas, 1961 ; Vogel, 1961 ; Glansdorff and Sand, 1965). Our interest has been focused mainly on the functional organization of a group of adjacent arg locil, the study of which had posed some interesting questions regarding coordination of gene expression. Fig. 1 shows the present status of the gene-enzyme relationship in arginine biosynthesis. Nine genes (2 of which determine ornithine transcarbamylase--~ enzyme F -~ OTC; Glansdofff, Sand and Verhoef, 1967) code for the 8 enzymes of the pahtway (the respective roles of the 2 0 T C loci have not yet been elucidated). An interesting feature of the system is the tight cluster formed by argE, C, B and H, in that order (Glansdofff, 1965), in the immediate vicinity of the loci metB, metF, glu and purD (see Fig. 2). Although the 4 loci appear to be adjacent, their expression is not strictly coordinated; Baumberg et al. (1965) reported a two fold difference in the repressibility coefficients u of enzymes E and H; we reached independently the same conclusion, showing, moreover, that the expressions of argC, B and H present a strong degree of coordination, the repressibility coefficient of the corresponding enzymes being about 50, whereas that for the synthesis of enzyme E is about 18 (Glansdorff and Sand, 1965). That argE and H (at least) are independently repressed - - a statement first suggested by the analysis of an apparent translocation of argE (Baumberg et al., 1965) - - might be correct, as we shall see, but was, at that time, based on a methodological artifact due ¢o the pecularities of recombination at the level of the genes transferred 2 (I)erepressed versus repressed levels of enzyme specific activity). 3 l~olec.Gem Genetics10¢i

34

R. Cunin, D. Elseviers, G. Sand, G. Freundlich and N. Glansdorff:

during the first m i n u t e of bacterial c o n j u g a t i o n (G]ansdofff, 1967). We wish to present n e w a r g u m e n t s which s u p p o r t the view t h a t the cluster contains 2 successive entities polarized in the same direction (clockwise on the usual drawings + a n d argCBH. -----> F o r reasons which will appear of t h e bacterial chromosome): argE i n the discussion, we do n o t wish to present t h e m definitely as distinct operons a t the present time. A p r e l i m i n a r y a c c o u n t of this work has been published (Cunin et al., 1968). Materials and Methods Genetic Techniques. Transductions and matings were performed as described previously (Glansdorff, 1965). The genetic nomenclature follows the rules of Demeree et al. (1966). Strains. a) Bacteria: all strains are derivatives of Escherichia coli K-12 (see Table 1). b) Phages: the temperate phage 363 was used to perform generalized transduetions (Jacob, 1955; Glansdofff, 1965). The following set of phage Z strains was used to define the suppressor genotype of various bacterial strains: the clear plaque mutant ~ C 90; the amber mutants 13, 43, T 57, 204, 207, 208, 216, 221. The phages were kindly provided by Professor R. Thomas; they permit the unambiguous characterization of the amber suppressors I, II and III, and of the ochre suppressors B and C (Thomas et al., 1967; Van Montagu et al., 1967). The set of MS2 nonsense mutants described in the last reference has also been used in some instances. Table 1. List o/the strains a

P4XB2 A B 1206 CA2441

© OO 0OO

-~ ~+

-~ -+

~-~-

~-~ UAG

B ~+

~-+

-~ -~ ~-

S R S

P678R2 C600R9

t.1. O©O

--

~-

--

+

-}-

~-

~-

--

-~

--

~-

-~-

-~

-[-

R R

H/r

--

--

Fld

~-

F+

+

n.d. --

F-

--

~-

F-

-t-

-t-

a Arg mutants are described in the text. b The sign ~- means: suppressor present. C) ~ W. K. Maas, © O : E. A. Adelberg, O O O ~ R. Thomas, t.1. ~ this laboratory, n. d. ~ not determined.

Enzymes Assays. For the techniques used in the ultrasonic extraction of cells and the measurements of the specific activities of enzyme B (ATP: c~-N-acetyl-L-glutamate 5-phosphotransferase), C (~-N-acetyl-L-glutamate y-semialdehyde: NADP oxydoreduotase, phosphorylating), D (~-N-acetyl-L-ornithine: 2-oxoglutarate aminotransferase), E (L-ornithine: ~-N-acetyloraithinehydrolase), F (carbamoylphosphate: L-ornithine carbamoyl transferase OTC) and H (L-argininosuccinate arginine-lyase), see Glansdorff and Sand (1965). Some additional comments are necessary in regard to enzyme It. The assay of enzyme H involves the colorimetrie determination of urea formed after splitting by arginase of the arginine produced in the course of the reaction. Urea had been frequently assayed according to the method of Archibald (1945); however, to make reliable estimates of specific activities lower than 0.15 ~zmole/hr/mg protein (repressed level of this enzyme), we used the 10 times more sensitive method of Croekaert and Sehram (1958). The extracts were incubated either after molecular filtration on Sephadex G25 or after a 10 to 30 fold dilution in the phosphate buffer used for the incubation mixture; short time incubation of a filtered extract or relatively long incubation of diluted extracts gave identical results. As blanks, samples without argininosuccinate were used. For the extraction of cells which

Cluster of arg Genes in Escherichia cell

35

have to be used in the assay of several of the six above listed enzymes, Tris buffer 0.05 ~ pit 7.5 and phosphate buffer (K salts) 0.05 ~ ptI 7.5 gave identical results, provided that enzyme tI was preincubated 4 minutes in phosphate buffer after extraction in tris (tIavir et al., 1965). Experimental

Results

We will first review a minimum of relevant data concerning the decoordination of enzyme synthesis in the pathway and comment on the non-triviality of the departure from strict coordination t h a t we observe in the expression of argE and argH; secondly, we shall present the necessary mapping data for argE, C, B or H markers and thirdly, an analysis of the functional organization of the cluster by determining the extent of the polar effects exhibited by various mutations (nonsense, frameshift, deletions). Observations on the coordination of enzyme synthesis in polar a~y mutants will be presented in a fourth section. As will be shown this point is particularly important for the general discussion of the results.

1. Coordination o/Enzyme Synthesis in the Arginine Pathway Table 2 shows values previously given for the specific activities found for enzymes B, C, E and H under various conditions. (Glansdorff and Sand, 1965). In addition enzyme activities are reported for enzyme D. From these results it is possible to evaluate the very large difference which distinguishes OTC from Table 2. Specl/ic activities o] the enzymes speciJied by the loci argE, C, B, H, D, /ound in P d X and P g X B 2 (argR) under various conditions Strain

P4XB2 (argR) P4X (~v.t.) id.

Enzyme activity (y.m/hr/mg protein) Supplemerit to minimal medium

Acetylornithine lyase (locus argE)

OxydoPhospho- Arginino- Transareductase transsuceinate minase (locus ferase lyase (locus argC) (locus (locus argD) argB) argH)

Ornithine transcarbamylase (loci argF, I)

--

84.1

3.8

4.0

9.7

5

1,100

--

28.0

0.7

0.9

1.8

2.7

4.6

0.08

0.1

0.18

0.26

arginine

90 1.5

the other enzymes, as well as the more subtle kind of decoordination between the synthesis of enzymes E and D: although the coefficient of repressibility is the same for both of them, the intermediary values show t h a t the variations resulting from a change in culture conditions are not the same. For the present work our main interest lies, however, in the repressibility coefficients of enzyme E on the one hand and of enzymes C, B and I t on the other hand, given the clustering of the corresponding structural genes. We have a t t e m p t e d to see if this decoordination is not due to an artifact at some stage in the performance of the enzyme assays. We obtained essentially the same type of response as t h a t which Bauerle and Margolin (1966) observed in the study of a similar question concerning the t r y p t o p h a n enzymes in Salmonella: 3*

36

R. Cunin, D. Elseviers, G. Sand, G. Freundlich and N. Glansdofff:

the same activities (Table 2) were obtained whether the extracts were prepared in phosphate or Tris buffer, whether they were dialyzed or not (this was verified for enzymes E and H only), whether or not additions of an extract of cells harbouring a deletion of the cluster were made to the incubation mixtures; at most we could detect a 15% decrease in enzyme C activity under the latter conditions. The decoordinations can thus not be attributed to the presence of inhibitors in concentrated extracts. Moreover the specific activities of enzymes E and H do not change when argR mutants (genetically dereprcssed) are shifted from the routinely used minimal medium to broth (medium 853, Glansdorff, 1965) or a synthetic enriched medium (AF medium, Novick and Maas, 1961); the values previously considered as maximal were thus not artificially limited by an insufficient supply of a particular amino acid (for an example of decoordination provoked by such a limitation see Somerville and Yanofsky, 1964). The nontriviality of the decoordination exhibited by enzyme E synthesis is moreover strikingly confirmed by a genetic argument: in a recent communication (Elseviers Table 3. Three point transductions between (gln, arg) recipients and arg donorsa Recipient

Donor

glu+arg+/arg+

%

Order

gln argE-1 g~uargE-2 glu argE-1 glu argE-1 glu argE-26 glu argE-15 glu argE-15 g~u argE-26 glu argE-15 glu argE-16 glu argE-21 glu argE-16 glu argE-23 glu argE-16 glu argE-24 glu argE-24 glu argE-24 glu argE-3 glu argE-21 glu argE-21 glu argE-16 glu argE-16 glu argU-6 glu argB-1 g~uargB-4 glu argB-2 glu argC-9 glu argC-7 glu argC-8

argE-2 argE-1 argE-25 argE-26 argE-1 argE-25 argE-26 argE-15 argE-16 argE-15 argE-16 argE-23 argE-16 argE.24 argE-16 argE-26 argE.3 argE-24 argE.18 argE.3d argE-18 argE-34 argG-1 argB-5 argB-5 argB-5 argU-1 argU-2 argC-2

53/265 192/252 71/94 480/591 29/492 39/151 75/209 147/292 263/299 26/288 37/288 162/335 52/320 159/339 67/352 69/334 231/317 128/298 125/145 62/117 95/108 63/109 64/119 60/107 165/250 30/460 89/120 6/71 28/65

21.8 80.0 89.0 81.5 9.2 25.8 36.0 50.5 88.0 9.0 12.8 53.8 16.8 52.7 20.6 20.6 73.0 43.0 86.2 53.0 88.0 58.0 53.8 56.0 66.0 6.5 74.0 8.4 43.0

2--1 1--25 1--26 15--25 26--15 15--16 16--21 16--23 16--24 26--24 24--3 21--18 21--34 16--18 16--34 6--1 1--5 4--5 5--2 9--1 2--7 8--2

a The markers argE-1, C-1, C-2, C-3, B-l, B-2, B-3, B.4, H-1 to 20 (Fig. 2) have been mapped previously (Glansdofff, 1965, 1966). Additional results or confirmation of 3 point tests data were obtained by deletion mapping (Fig. 2).

Cluster of arg Genes in Escherichia coli 30 rl~

..~rg

/626

8g 52 g5 85 org m~tBF c~JuECBH

E 21" 18" 36 ~: !

18

37

cotransduction index pqrg

arg C 5to20 tl I

II

I

6x

I

I I

vy

polarEC-Idel

1

5 2 polar point mutations

polorCBqlde[ I

I

I

I

non polar SUP-102 del I I I non iI I

potor gtu-arg EC(B) del

I

I po nr

B-25de[

Fig. 2. The argECBH cluster, approximately to scale. The position of the markers has been ascertained by 3 point transductions. (Table 3, Glansdorff, 1965, 1966) and by deletion mapping (this Fig.); *: nonsense (UAG) mutation

et al., 1969) we describe how a deletion destroying the argEC border reduces by a factor of 3 the repressibility coefficient of enzyme It.

2. Extension o/the Topographical Analysis o/the argECBH Cluster In order to appreciate the information revealed by the behaviour of the mutants described in the next sections, an extension of our topographical knowledge of the cluster was necessary. This was performed by pairs of 3 point tests in different coupling phases, as described earlier (Glansdorff, 1965, 1966), or by deletion mapping. Table 3 contains the information pertinent to the present paper; a more extensive analysis, consisting of reciprocal 4 point tests, allowing estimation of negative interference on both sides of the cluster will be published elsewhere (Freundlich and Glansdorff, in preparation). Fig. 2 presents the conclusions of these investigations; utilizing observed recombination frequencies we have drawn the map roughly to scale. With the exception of the mutants treated in the next section, none of those shown (plus half a dozen of yet unmapped ones) exhibit the slightest polar effect; this has been determined by measuring the specific activity of the enzymes controlled by the gene or genes immediately adjacent to the one affected by a given mutation (Glansdorff, 1966; Sand, 1969).

3. Polarity in the arg Cluster To elucidate the functional organization of the argECBH duster, we will successively consider three approaches: 1. Screening for mutants which, on the basis of their phenotype, could be polar. This was possible because the polar mutation argB-2 (Glansdorff and Sand, 1965, and below) dfizdnishes the maximal rate of enzyme H synthesis by about

38

1~. Cunin, D. Elseviers, G. Sand, G. Freundlich and N. Glansdorff:

,

o, /

/ j'

O2

o

,..o 0.1 d

0.05

//

C

/

// I

I

0

I

I

2

3

5

r

6 7 8 time (hours)

/

9

10

Ill

12 13

Fig. 3. Growth pattern of polar arg mutants in minimal medium, a) argCB-1 with 100 Fg L-arginine per ml; b) id. with 5 ~g L-arginine and 100 Fg L-ornithine per ml; c) argEC-1 with 5 tLg L-arginine and 100 ~g L-ornithine per ml 99 %, which gives rise to very slow growth in the presence of ornithine: the rate of ornithine utilization is indeed limited by the residual capacity of enzyme H synthesis. 2. Examing the suppressibility (by amber suppressors) of various mutations, in order to determine in which of the 4 genes mutations known to be nonsense were possibly not polar. This approach is a check of the first one; it was particularly desirable in the case of argE where it had not been possible to find a mutation leading to slow growth in the presence of ornithine. 3. Analyzing the cluster with the help of deletion mutants. As previous results suggested the existence of a yet undefined genetic punctuation between argE and the group CBH, it was necessary to observe the effects of mutations deleting the boundaries between genes of the cluster. a) Isolation and Analysis o/ Slow Utilizers o/ Ornithine. Prompted by the isolation of m u t a n t argB-2, which already established the existence of a unit of coordinated expression comprising at least argB and H, BH (Glansdorff and Sand, 1965) we looked for similar mutations, in argC and E among a set of arginine auxotrophs isolated after treating H f r P X 4 (sup-) with X rays or with the frameshift inducing agent acridine ICR 191 (see Whitfield et al., 1966, for use and comments). Mutants argC-6, E-18 and E-3d described at the end of this section, are N G induced derivatives of strain CA 2441. The relevant features of the mutants obtained are given in the following paragraphs. Mutant I{C-B-6-21 (argOB-1): the growth response of this organism is shown in Fig. 3, which also holds for I{C-D-16-19, HC-D-20-5 and I-IC-12-67. The

Cluster of arg Genes in Eseherichia coli

39

Table 4. Argininosucci:~sespeci/icactivities/oundin polararg mutants in conditions o/repression

(arginine supplemented minimal medium:100 ~g/mL) and genetic derepression (argR suppressorless derivativesin identical conditions Mutation

argCB-1 aryCB-2 argCB-3 argB-2 argB-5 argC-6

Enzyme tt activity a (~tlYI/hr/mgprotein)

argR-

argR+

ratio

0.59 0.41 0.53 0.02 0.31 1.2

0.01 0.01 0.03 -0.07 --

59 41 17 -4.4 --

a 2 to 4 independent determinations in each case. exhaustion of a limiting amount of arginine is followed by an adaptation period after which growth on ornithine proceeds at a constant rate. The outgrowth on ornithine is paralleled by a gradual increase of enzyme H specific activity to its maximal residual level given in Table 4 (determined in an argR derivative of the strain). Enzyme E is normally produced and regulated (Table 5) ; no enzyme B nor C are detectable. Genetically, the mutation appears as a deletion covering the right end of argC and most of argB (Fig. 2). No known argH mutation is included in the deletion; moreover, argH m a y be fully reactivated b y an extension of the deletion to the left. Mutant ItC-D-16-19 (argCB-2): this organism, independently isolated, is very similar to argCB-1. Mutant tIC-D-20-5 (argCB-3): two features distinguish this m u t a n t from the 2 previous ones: the occurrence of very rare spontaneous revertants, in spite of the multisite nature of the mutation, and a diminution of the repressibility coefficient of enzyme H (this point is discussed at the end of the paper). I n the absence of further information, the m u t a n t m a y be compared to a leu auxotroph (leu-39) described by Margoiin (1963) who considered the possibility of an inversion. Mutant HC-12-67 (argB-5): the mutation has been localized between argB-4 and argB-2 (Table 3, Fig. 2). Synthesis of enzyme E is normal (Table 5). The value obtained for enzyme C in an argR derivative (Table 5) suggests that argB-5 exerts a moderate antipolar effect on the expression of argC. The mutation m a y tentatively be considered as frame-shift (no suppression by amber suppressors, no N G induced reversion). Mutant argB-2: this X - r a y induced m u t a n t gives rare revertants, (0.5 10-1°, rate not increased b y NG, DES or ICl%) the properties of which suggest t h a t argB-2 is a short frame-shift deletion. The revertants are "pseudowfldtypes", they contain a phosphotransferase unidentifiable b y the conventional assay but nevertheless about half as efficient as the wild-type enzyme since the rate of synthesis of enzymes E, C, H and F in unsupplemented minimal medium is almost 2 times higher than in the wild-type.

40

1%.Cunin, D. Elseviers, G. Sand, G. Freundlich and N. Glansdorff:

Table 5. Polar mutants: Specific activities o/the enzymes specified by the loci argE, C, B, H

and Y, I (OTC) at about hal/ their complete derepression level in L-ornithine supplemented minimal medium (lO0~g/mL), and in repression (L-arginine: 100 ~g/mL) l~utant

Condition

Acetylornithinase (locus E)

Oxydoreductase (locus C)

HC-B-6-21

orn arg

41 3.9

<0.04 --

HC-D-16-19

orn arg

66.5 4.4

tIC-D-20-5

orn arg

44.2 2.9

HC-12-67

orn arg

39.5 4.1

0.41 a 0.02

argC-6b

arg

71

--

Phosphotransferase (locus B)

Argininosuccinase (locus H)

Ornithine transcarbamylase (loci, F, I)

<0.02 --

0.21 --

520 4

<0.02 --

<0.02 --

0.21 --

391 2.4

<0.02 --

<0.02 --

0.30 --

405 2.0

<0.05 --

0.17 --

374 1.1

1.2

1.2

700

a 2.1 in an argR derivative.

b argR derivative. The Arg+ phenotype of those strains is cotransdueible with glu at the same frequency as argB-2 itself and m a y thus be attributed to intragenic suppression. As expected from the s t u d y of his frame-shift m u t a n t s (Martin et al., 1966), the polar effect totally disappears in the revertant. Moreover, a derivative of argB-2 in which argH activity is fully restored, harbours a short deletion, covering argB-2 and argB-5, which presumably corrects the flame-shift present in the parental strain. The extremely strong polar effect of argB-2 (Table 4) and its distal position in the gene suggest t h a t the frame-shift effect extends into argH. I n regard to the functional organization of the cluster, the properties of the 5 m u t a n t s lead to the same conclusions: the integration of argB and H, already stated after a preliminary analysis of argB-2 (Glansdorff and Sand, 1965) is confirmed. The antipolar effect of argB-5, the deletion analysis - - (section e) - and the m u t a n t argC-6 show t h a t argC belongs to the same unit of expression. M u t a n t argC-6: the residual level of enzyme I t synthesis is 70% of w h a t is found in the wild-type growing in unsupplemented minimal medium thus one sixth of the maximal activity (Table 5). ArgC-6 is an amber m u t a n t , as shown b y the same tests as the ones we have applied to characterize argE nonsense m u t a n t s (see section 3. b.). The m u t a t i o n lies in the left h a n d portion of the C gene; it shows t h a t argC, B and H belong to the same operon. b) Non polar nonsense arg mutants. ArgE m u t a n t s : none of the mutations mentioned so far, except argC-6, are suppressible b y the amber sup. I I . This does n o t hold for the N G induced m u t a n t s argE-18, 21, 23, 24 and 34. We conclude t h a t t h e y are nonsense (UAG) on the following basis: their Arg p h e n o t y p e becomes leaky against the genetic background of P A 342 (sup II). transductions between 2 - derivatives of the 4 m u t a n t s and a phage 363 lysate of C 6 0 0 R 9 del (carrying sup I I and a deletion of the arg cluster), gives pheno-

Cluster of art Genes in Escherichia coli

41

Table 6. Absence of polar e]]ect in amber argE and argH mutants (argR derivatives) Strain

Reference

argR strain

P4XB2 P678R2

argE-21 argE-23 argE-24 argE-18 argE-34 argH-4 argH-11

--P678R2 P 678 R 2 P 678 R 2 P 678 R 2 P 678 R 2 P4XB2 P678R2

Enzyme specific activity (~M/hr/mg protein) Oxydoreductase (locus C)

Argininosuccinase (locus H)

Phosphotransferase (locus B)

3.7 5.1 5.3 5.2 4.3 -----

9.7 14.5 14.2 13.5 12.0 8.3 8.4 ---

4.0 9.5 -----3.6 11.2

typically Arg+ recombinants which the phage spot test (see Materials and Methods) identifies as sup I I . a m o n g revertants of all 4 mutants, we find sup I I , sup I I I and sup U (ochre suppressor) strains. These organisms are particularly i m p o r t a n t for the analysis of the functional organization of the cluster: t h e y are not polar (Table 6) : the maximal levels of enzymes C and H, measured in argR derivatives are normal. argE-21, 23 and 24 m a p near each other a p p a r e n t l y in the centre of the gene, although t h e y have arisen independently; argE-18 and E-34 are more distal. I f there existed an argECBH operon, polarized from left to right, one would expect some of these mutations (mainly E-21, 23 and 24) to be polar, on the assumption t h a t in argE the gradient of polarity is similar to the one found in the first gene of various operons (Newton et al., 1965; Bauerle and Margolin, 1966; Y a n o f s k y and Ito, 1966; F i n k and Martin, 1967).

argH m u t a n t s : a m o n g 20 argH m u t a n t s of Hfr P 4 X , 2 (argH-4 and H-11) are amber, as shown b y the tests described in the previous subsection. B y examining enzyme B specific activity in argR derivatives of both m u t a n t s and in argR art + strains, it could be established t h a t neither argH-4 nor argH-11 influence the expression of their left side neighbour argB (Table 6). c) Deletion Analysis o/ the arg Cluster. The first strain to be considered is HC-B-7-18 (argEC-i; I C R mutagenesis). I n a preliminary account of p a r t of this work (Sand and Glansdorff, 1967) we considered this organism to carry a deletion covering the left p a r t of argE but not the right one, since rare Arg+ recombinants ( < 0.1%) were obtained in transductions with argE-3 or E-5 donors; because the m u t a t i o n is strongly polar (see below) this favoured the hypothesis of an argECBH operon. I t now appears from reconstruction experiments t h a t the recombinants obtained were due to the occurrence of phages carrying arg+ alleles originating from the lysates used to prepare our stocks; one would tend to minimize the risk created b y this artifact after reading Demerec's first paper on selling (1962) ;

42

R. Cunin, D. Elseviers, G. Sand, G. Freundlich and N. Glansdorff:

we think however t h a t with the inocula routinely used in the preparation of P 1 and 363 transductions, such a risk exists and should not be neglected. Stocks are now routinely recycled or prepared with a lysate of MN41 (glu/argECBH deletion mutant). The features of HC-B-7-18 are: No enzyme E produced; the whole argE gene appears to be deleted: no Arg+ recombinants ( < 0.005%; Met+ as standard) are obtained in transductions with argE-1, 2, 3, 5 and 6 nor (<0.002%) with the argEUB deletion of tIC-B-6-21 sup 102 (Elseviers et al., 1969 ; Fig. 2) which covers so little of ar9E t h a t aeetylornithinase is still partially functional. Although no known argC marker is included in the deleted area, the activity of enzyme C is not detectable ( < 10% of the repressed level) whereas enzyme H under the same conditions constitutes 70 to 80 % of this level, given in Table 2. Moreover, no crossfeeding of argC mutants by HC-B-7-18 could be observed. The deletion thus presumably covers part of argU. The residual level of enzyme t[ synthesis is no longer repressible b y arginine. This is strikingly reflected by the absence of the adaptation lag in the arginineornitMne diauxy (cf. Fig. 3); OTC synthesis, however, is still repressible. These features at first sight suggest that HC-B-7-18 contains a deletion destroying argE, and at least one operator gene; ff an operator existed between argC and argB, the residual level of enzyme H synthesis would not be constitutive in this strain. I t has been verified t h a t the latter does not contain a second polar mutation in argC or argB: not a single Arg and aeetylornithine utilizing recombinant could be found among 1,000 recombinants selected in the presence of aeetylornithine (used by argB and C mutants) from a transduction between HC-B-7-18 (recipient) and argH-1; the latter donor was used instead of the wild type in order to increase the sensivity of the test by a factor of about 6, since argE and H are 85 % cotransducible. The second relevant strain (sup 102: Fig. 2) has already been described (Elseviers et al., 1969). I t is a derivative of the polar deletion m u t a n t argCB-1 (Fig. 2) in which an extension of the deletion has eliminated the argE-C boundary. I n this strain argH is fully reactivated and has a coefficient of repressibility (14) similar to that of argE in wild type strains (18). These features, the lack of polar effect of the 5 amber argE mutations on argU and H, and the properties of the strain HC-B-7-]8 support the view that the cluster consists of 2 successive entities (argE and argCBH), having the same polarity but separated by a punctuation which does not allow the polar effect of argE nonsense mutations to reach argC. The results also suggest t h a t under conditions of repression the molar ratio enzyme E/enzyme H is about 4,0 or a little lower (the repressed levels of enzyme g in sup 102 and in the wild-type are 0.7 and 0.18 respectively). d) Coordination o/Enzyme Synthesis in the Polar art Mutants. Martin et al., (1966) have established that, in Salmonella, the repressibility coefficient of an enzyme produced b y a gene whose activity is dependent on a polar his mutation, is the same as in the wild-type (a two fold difference is observed in some rare exceptions, ibid.) ; a similar statement has been made in the case of the gal operon in Escherichia coli (Michaelis and Starlinger, 1967). This does not hold for the tryptophane (try) operon of Escherichia coli (¥anofsky and Ito, 1966 ; I m m a m o t o et al., 1966) nor for the try operon of Salmonella (Bauerle and Margolin, 1966);

Cluster of art Genes in Escherichia eoli

43

Table 4 shows t h a t some of the polar art mutants exhibit a similar decoordination of enzyme synthesis; whereas in IIC-D-16-19 and HC-B-6-21 the assays do not detect a significant variation in the repressibility eoeffieienb of enzyme H, this coefficient decreases to 17 in tIC-D-20-5 and 4.5 in HC-12-67. This point not only requires an explanation but, as stated in the discussion, takes on a certain importance when we come to discuss the functional organization of the art cluster in the perspective of arguments presented by others (Vogel et al., 1967). Discussion

The properties of the polar argC or B mutations, of the non polar amber argE or H mutations, combined with the deletion analysis of the argECBH cluster, favour the view t h a t the latter is composed of two adjacent operons, identically polarized. Other interpretations are however not excluded, and two of them will be considered at tile end of the discussion. The polarity of the art cluster is clockwise, as in the case of the pyruvage dehydrogenase operon (Henning and Herz, !964), whereas the polarity of the other known Escherichia coli operons is counterclockwise (references in Taylor and Trotter, 1967); this adds to the growing evidence t h a t the informational DNA strand is not the same throughout the entire bacterial chromosome (Margolin, 1965; Sanderson, 1965). Vogel et al. (1967) have described an art auxotroph exhibiting a strong reduction of both the rate and the repressibility of enzyme H synthesis while preserving normal enzyme E ;production and concluded, therefore, that argE and H were independently repressed; however, since m a n y mutations now have been shown to produce similar effects within the boundaries of one operon (see argCB-3, argB-5 in Table 4, and the try-mutants refered to below) this Mnd of evidence has become insufficient to support the latter conclusion. On the other hand, the enzymological properties of the m u t a n t are in full agreement with the idea t h a t argC, B and H belong to the same operon. Recent unpublished evidence, acquired in collaboration with S. Baumberg and E. Ashcroft, suggest t h a t the strain harbours a polar argC mutation. I n addition, the same authors (personal communication) have isolated three IClg acridine-induced argE mutants t h a t show no detectable polar effect on argH; their results therefore are in accord with ours in suggesting t h a t argE and argH are in separate operons. The pattern of repressibility observed in a strain carrying a mutation leading to the synthesis of a thermosensitive product of the a~VR gene points in the same direction (Jacoby and Gorini, 1969). Reduced repressibility coefficients for enzymes produced under the influence of polar mutations have also been observed in the try operon of E. coli (Yanofsky and Ito, 1966) and S. typhimurium (Bauerle and Margolin, 1966). I n the latter ease (Bauerle and Margolin, 1967) and, m a y be, the former too (2¢Iorse and Yanofsky, 1968) a plausible explanation seems at hand since an internal initiator site (p2) exists between the second and the third gene of the operon; this punctuation could assure a basal level of expression of distal genes if a proximal polar mutation prevented translation initiated at the first site (p l, near the operator) to reach p2. I n polar mutants, this basal level would thus result in a lowered repressibility for the synthesis of enzymes coded by distal genes. I n the

44

1~. Cunin, D. Elseviers, G. Sand, G. Freundlich and N. Glansdorff:

case of the argCBH subeluster, we have no direct evidence for such an internal site between argB and H but the possibility remains t h a t the coordination of argH with argB and C, however strong (Glansdorff and Sand, 1965) is not perfect; very low levels of repression are indeed more difficult to estimate accurately for argB and C than for argE and H; refinements in methodology will be needed to answer this question. A last point to discuss is the cause or selective pressure which could explain how the argECBH cluster is maintained, since it contains two groups of genes, the expression of which at first sight seems to be independently controlled b y the same repressor; moreover, artificial dispersal of the cluster by duphcation has been described (Gl~nsdorff, 1966; Glansdorff and Sand, 1968; Elseviers et al., 1968--1969). The model advanced by Vogel and Bacon (1966; see also Vogel et al., 1967) who propose a kind of equivalence between clustering and a postulated aggregation of individually repressible genes in the compartimentalization of metabolic pathways (see last reference p. 228) is an interesting but not yet well established hypothesis. Moreover, we feel t h a t the evidence on which the latter model is based - - the proportion of double art mutants in a random sample - could be explained b y other hypotheses than the one presented; indeed, in a score of 340 U.V. induced auxotrophs, 3 % of which arc Arg mutants, we found a proportion of multiple auxotrophs (for unrelated requirements) high enough (7.5%) to make the 0.7% double Arg mutants found b y the authors (ibid.) among 600 Arg auxotrophs, appear not particularly remarkable. At any rate even a really unexpectedly high frequency of double mutants affected in the genetic control of a particular pathway could be due to other factors than a t e m p o r a r y aggregation of functionally related loci, for example to a possible selective advantage of these organisms over single mutants; this has already been observed in m a n y cases (Gillespie et al., 1968). The clustering of the four art loci might be attributed to a specific, yet obscure feature (new kind of punctuation or selective pressure at some level in their expression) or simply consist in the apposition of two functionally related operons. I t is possible to visualize models integrating the expression of the four loci into one unit of expression, yet not into an operon of the classical type. The two most obvious ones m a y be formalized as follows; One common operator at the left of urgE, but a very efficient initiator site at the argE-argC junction, allowing maximal expression of argC, B and H in nonsense argE mutants, thanks to attachment of ribosomes at the corresponding messenger site. This model might be analogous, at least formally, to the situation encountered in bacterial I~NA viruses. One operator for urgE, one for argCBH, but only one attachment site of RNA polymerase on the left side of argE. Experiments are in progress in order to test these and other possibilities. Useful information is expected to be obtained by the study of argE or argEC deletions as well as other argE nonsense mutations; it is indeed not yet excluded t h a t urge contains a portion within which nonsense mutations are polar. Operator mutations affecting the genes of the cluster would of course settle the question, but have not yet been found, although mutations which specifically

Cluster o~ arg Genes in Escherichia coli

45

release one or two arg s t r u c t u r a l genes from t h e pleiotropie repression o p e r a t i n g in t h e w i l d - t y p e are a l r e a d y k n o w n : argH a n d argBH d u p l i c a t i o n s (Glansdorff a n d Sand, 1968; Elseviers et al., 1968, 1969), a n ornithine t r a n s e a r b a m y l a s e m u t a n t ( J a c o b y a n d Gorini, 1969) a n d t h e deletion argEC-1 described in t h e present paper. I t would be p a r t i c u l a r l y interesting to c o m p a r e our results w i t h those o b t a i n e d in similar a n a l y s e s of Proteus a n d Salmonella. M a p p i n g studies h a v e shown t h a t t h e l a t t e r o r g a n i s m h a r b o u r s a cluster of arg loci v e r y similar to t h e one of Escherichia coli (Armstrong, 1968); Proteus mirabilis contains a group of five closely l i n k e d genes: argECBGH (Prozesky, 1968). I t is still difficult to see t o w h a t e x t e n t t h e arg cluster could be c o m p a r e d to t h e group of ilva genes a n a l y z e d b y R a m a k r i s h n a n a n d A d e l b e r g (1965); a l t h o u g h in b o t h cases u n c o o r d i n a t e d genes h a v e been f o u n d to be t i g h t l y clustered, i t is l i k e l y t h a t t h e ilva cluster will p r o v e to be m o r e complex t h a n t h e arg one: a t least two k i n d s of repressor molecules influence different sections of it (ibid.).

Acknowledgements. We wish to express our gratitude to Professor W. K. ~[aas for his criticisms and help during the preparation of the manuscript, to Professor J.-M. Wiame, R. Lavall4, Dr. S. Baumberg and E. Ashcroft for their interest and many helpful discussions. We acknowledge with deep appreciation the amber mutants and the bacterial strains kindly provided by Professors R. Thomas and M. van Montagu, as well as a generous gift of ICR acridines sent by Dr. H. Creech. References Archibald, R. M. : Colorimetric determination of urea. J. biol. Chem. 157, 507--518 (1945). Armstrong, P. B. : Orientation and order of the met-arg region in the Salmonella typhimurium linkage map. Genetics 56, 463--466 (1967). Bauerle, R. H., Margolin, P. : The functional organization of the tryptophan gene cluster in Salmonella typhimurium. Proc. nat. Acad, Sci. (Wash.) 56, 111--118 (1966). - - - - Evidence for two sites for initiation of gone expression in the tryptophan operon of Salmonella typhimurium. J. molee. Biol. 26, 423--436 (1967). Baumberg, S., Bacon, D.F., Vogel, H. J.: Individually repressible enzymes specified by clustered genes of argininc biosynthesis. Proe. nat. Acad. Sci. (Wash.) b3, 1029--1032 (1965). Boman, H. G., Boman, I. A., Maas, W . K . : Studies on the incorporation of arginine into aeceptor RNA of Escherichia coll. In: Biological structure and function (Goodwin and Lindberg, eds.), p. 297--308. New York: Acad. Press 1961. Crockaert, R., Schram:, E. : Dosage des N-carbamoyld6riv6s d'acides amin6s par la diac6tylmonoxime. Bull. Soc. Chim. biol. (Paris) 45, 1093--1106 (1958). Cunin, R., Elseviers, I)., Sand, G., Freundlich, It., Glansdorff, N. : Organisation fonctionelle d'un groupe de loci arginine contigus chcz Escherichia coll. Arch. int. Phys. Biochem. 76, 927--928 (1968). Demerec, M.: Sellers attributed to unequal crossovers in Salmonella. Proc. nat. Aead. Sei. (Wash.) 48, 1696--X704 (1962). - - Adelberg, E. A., Cl~wk, A. J., Hartman, P. E. : A proposal for a uniform nomenclature in bacterial genetics. Genetics 54, 61--76 (1966). Elseviers, D., Cunin, R., Glansdorff, N. : Reactivation of arginine genes under the influence of polar mutations. FEBS Letters 3, 18--20 (1969). - - Sand, G., Glansdorff, N. : Reactivation of genes under influence of polar mutations. Arch. int. Phys. Biochem. 76, 929--930 (1968). Fink, G. R., Martin, R. G. : Translation and polarity in the histidine operon. I I Polarity in the histidine operon.. J. molec. Biol. 97--107 (1967). Gfllespie, D., Demerec, M., Itikawa, H. : Appearance of double mutants in aged cultures of Salmonella typhimu~'ium cysteine-requiring strains. Genetics 59, 433--442 (1968). -

-

46

1%. Cunin, D. Elseviers, G. Sand, G. Frenndlich and N. Glansdorff:

Glansdorff, N. : Topography of cotransducible arginine mutations in Escherichia coli K-12. Genetics 51, 167--179 (1965). - - Le eontr51e g~nStique des biosynth~ses de l'arginine et du carbamoylphosphate chez Escheriehia coll. Thesis, University of Brussels (1966). - - Pseudoinversions in the chromosome of Escherichia coli K-12. Genetics 55, 49--61 (1967). - - Sand, G. : Coordination of enzyme synthesis in the arginine pathway of Escherichia coli K-12. Biochim. biophys. Acta (Amst.) 108, 308--311 (1965). - - - - Duplication of a gene belonging to an arginine operon of Escherichia coli K-12. Genetics 60, 257--268 (1968). --Verhoef, C.: The dual genetic control of ornithine transcarbamylase synthesis in Escherichia coli K-12. Mutation. Res. 4, 743--751 (1967). Havir, E. A., Tamir, H., Ratner, S., Warner, R. C.: Biosynthesis of urea. XI. Preparation and properties of crystalline argininosuccinase. J. biol. Chem. 240, 3079--3088 (1965). Henning, U., Herz, C. : Ein Strukturgen-Komplex ffir den pyruvat-Dehydrogenase Komplex yon Escherichia coli K-12. Z. Vererbungsl. 95, 260--275 (1964). Hirshfield, I. N., Deken, R. de, Horn, P. C., Hopwood, D. A., Maas, W. K. : Studies on the mechanism of repression of arginine biosynthesis in Escherichia coll. III. Repression of enzymes of arginine biosynthesis in Arginyl-tRNA Synthetase mutants. J. molec. Biol. 35, 83--93 (1968). Immamoto, F., Ito, J., Yanofsky, Ch. : Polarity in the tryptophan operon of Eseherichia coli. Cold Spr. Harb. Symp. quant. Biol. 31, 235--249 (1966). Jacob, F. : Transduction of lysogeny in Escherichia coll. Virology 1, 207--220 (1955). - - Perrin, D., Sanchez, C., Monod, J.: L'op6ron, groupe de g~nes £ expression eoordonn~e par un op6rateur. C.R. Aead. Sei. (Paris) 250, 1727--1729 (1960). Jacoby, G.A., Gorini, L.: A unitary account of the repression mechanism of arginine biosynthesis in Escherichia coll. J. molec. Biol. ~9, 73--87 (1969). Maas, W. K.: Studies on repression of arginine biosynthesis in Eseherichia coll. Cold Spr. Harb. Syrup. quant. Biol. 26, 183--191 (1961). - - Clark, A. J. : Studies on the mechanism of repression of arginine biosynthesis in Escherichia coll. II. Dominance of repressibility in diploids. J. molee. Biol. 8, 365--370 (1964). --Maas, R., Wiame, J.-M., Glansdorff, N.: Studies on the mechanism of repression of arginine biosynthesis in Escherichia coll. I. Dominance of repressibility in zygotes. J. molec. Biol. 8, 359--364 (1964). Margolin, P. : Genetic fine structure of the leucine operon in Salmonella. Genetics 48, 441--457 (1963). - - Bipolarity information transfer from the Salmonella typhimurium chromosome. Science 147, 1456--1458 (1965). ~¢[artin, R. G., Silbert, D. F., Smith, D. W., Whitfield, H. J. : Polarity in the histidine operon. J. molec. Biol. 21, 357--369 (1966). Michaelis, G., Starlinger, P. : Sequential appearance of the galact~se enzymes in Escherichia coll. Molec. Gen. Genetics 100, 210--215 (1967). Montagu, M. van, Leurs, C., Brachet, 1~., Thomas, 1~. : A set of amber mutants of bacteriophages ,~ and MS2 suitable for the identification of suppressors. Mutation. Res. 4, 698--700 (1967). Morse, D. E., ¥anofsky, C. : The internal low-efficiency promoter of the tryptophan operon of Escherichia coll. J. molec. Biol. 38, 4 4 7 ~ 5 1 (1968). Newton, :N. A., Beckwith, J. 1~., Zipser, D., Brenner, S.: :Nonsense mutants and polarity in the lae operon of Escherichia coli. J. molec. Biol. 14, 290--296 (1965). Novick, R. P., Maas, W. K. : Control by endogenously synthesized arginine of the formation of ornithine transcarbamylase in Escherichia coll. J. Bact. 81, 236--240 (1961). Prozesky, O. W. : Transductional analysis of arginine less mutants in Proteus mirabilis. J. gen. Microbiol. 54, 127--143 (1968). Ramakrishnan, T., Adelberg, E. A. : Regulatory mechanisms in the biosynthesis of isoleucine and valine. II. identification of two operator genes. J. Bact. 89, 654--660 (1965). Sand, G. : Le contrSle g~n~tique de la biosynth~se de l'arginine ehez Escherichia coli K-12. Thesis, University of Brussels, 1969.

Cluster of a~'g Genes in Escherichia coli

47

Sand, G., Glansdorff, N. : L'op6ron arginine d'Escherichia coli. Arch. int. Phys. Biochem. 75, 568--569 (1967). Sanderson, K. E.: Information transfer in Salmonella typhimurium. Proc. nat. Aead. Sci. (Wash.) 58, 1335--1340. Somerville, F. L., Yanofsky, Ch. : On the translation of the A gene of the tryptophan messenger RlqA. J. molec. Biol. 8, 616--619 (1964). Taylor, A. L., Trotter, C. D. : ~evised linkage map of Escherichia eoli. Bact. Rev. 31, 332--353 (1967). Thomas, R., Leurs, C., Dambly, C., Parmentier, D., Lambert, L., Brache~, P., Lefebvre, N., Mousset, S., Porcheret, J., Szpirier, J., Wauters, D. : Isolation and characterization of new sns (amber) mutants of bacteriophage. Mutation. Res. 4, 735--741 (1967). Vogel, H. J.: Aspects of repression in the regulation of enzyme synthesis: pathwaywide control and enzyme specific response. Cold Spr. Harb. Syrup. quant. Biol. 26, 163--172 (1961). - - Bacon, D. F. : Gene aggregation: evidence for a coming together of functionally related, not closely linked genes. Proe. natl. Acad. Sci. (Wash.) 55, 1456--1459 (1966). - - Baumberg, S., Bacon, D.F., Jones, E.E., Unger, L., Vogel, R. H. : Gene-~ibosomeEnzymes organization in the arginine system of Escherichia coli, in Organizational Biosynthesis (Vogel, Lampen, Bryson, eds.), p. 223--234. New York: Acad. Press 1967. Whitfield, H. J., Jr., Martin, ]~. G., Ames, B. N. : Classification of aminotransferase (C gene) mutants in the histidine operon. J. molec. Biol. 21, 335--355 (1966). Yanofsky, Ch., Ito, J.: Nonsense codons and polarity in the tryptophan operon. J. molee. Biol. 2L 313--334 (1966).

Communicated by W. ~iaas Dr. R. Cunin Laboratory of Microbiology Brussels University l, Avenue E. Gryzon, Brnssels/Belgien