THE JO~RXAL OF kor.ocrca~. CI~EMISTRY Vol. 249, No. 9, Issue of Xhy 10, pp. 2946-2952, 1974

Printed in u.s.n.

Regulation of the L-Arabinose Operon BAD in Vitro*

(Received for publication, October 12, 1973)

SUMMARY

A DNA-directed cell-free system to study the L-arabinose BAD operon has been further developed so that the rate of synthesis of enzymes coded for by the structural genes in the operon in vifro is about 5 % the in vivo rate. L-Arabinose isomerase and L-ribulokinase, the products of the araA and araB genes, are synthesized at the same rate, demon- strating that the operon is coordinately expressed in vitro. The system is completely dependent upon araC protein which can be supplied either by synthesis in vitro from an araC+ DNA template or by the addition of araC protein ex- tracted from whole cells. The properties in the in vitro sys- tem of araC protein from either source suggests that the two products are similar if not identical. Both the repressor and activator forms of the araC protein are demonstrable in vitro using the appropriate templates confirming the es- sential aspects of the model for gene regulation of this operon.

‘I’lic espression of the r,-nrabinosc olx’ron, clraB.+lD,’ in Esche- rich coli 13/r is rcgulatcd by the Ixotcin product of the gcnc aruc (2, 3). It has lxx11 shoed that the nraC protein can exist in two distinct functional forms, activator and rqxcssor. ‘I’ll0

former, in the prcsqlcc of I,-arabinosr , is rcquirrtl for esprcssion of the opcron while the latter prcv(‘llts it (4). The model fol regulation of the IL-arabinosc oprm~ (5, 6) proposes that the transition from one functional state to the other is mediated by tlic inducer, L-arabinosc. In the al~~rc of L-arabinose the quilibriuin between rcprcssor and activator is in the direction of rcprcssor with rcprcssor bouiltl to tllc ara operator (ard).

1x1 the ~1l‘CSCilCC of I,-arabiiiosr thcl quilibrium is shifted toward activator and the replcssor is rcmo~ctl from n~u0. The activator functions at the initiator site (a&) to allow for csprcssioii of the opcro11. AIeasuremcnt of ara mRNA\ ~~~otluccd in viva suggests

that both repressor and activator function at thr lcvcl of traii- scription (7, 8, 9).

Supcrimposctl upon thr specific control exerted 011 the r,-arabi- 110s:~ opwoil by the araC protein is a moor gewral systenl of cntab-

* This investigation was supported by Kational Science Founda- tion (>rnrrt (;B2-1093.

$ llecipient of an 15.M.B.O. fellowship. 1 For the sake of simplicity we shall refer to this operon by its

structural genes HAD to distinguish it from other operons, e.g. WY&“, concerned with the t,ransport of 1,.arabinose and controlled by the uraC gene(l). The araUALl opcron is shown in Fig. 1.

olitc positive control mctliatctl through cyclic AMP2 and the catabolite gcnc activator (CGLi) protein (10). l‘hc opcro11 re- quires the prcsencc of CC.-1 protein aiitl cyclic AAll’ for espres- sion. Some wcrnt cspcrimrnts suggrst that (:GA protein inter- acts at a site which may b(x congruent xvith the cirul site (11) but tlir r&tionship htwmi (‘GA protein and thr activator for111 of the nraC protrill is at pmxwt unclear.

Itcgulation of thr I,-arabillosr operoil ill vitro Ins Imn studied using I)Nh-tlirrcted cell-frcr 1)rotcill-s~lltlicsizillR prrparations. It has been tlcnionstratrd tliat some of the in vivo proprrtics of thr I,-arabiirosc system arc also prrscnt in an ill vitro system which uses mai- DNA and clutlr a,oC 1”otc,ill-colltaillilrg rstracts to stimulate syiithrsis of I,-ril)liloliiliasr, the product of the m-al3 grnr (12). Ilowrxw, in :I similar systrm using ara+ 1)X.\, the synthrsis of I,-ribulokinnsr has bcrn obsnvctl without the adtli- tion of araC protciii (13). The npparrnt conflict may have been clarifictl by the recent tlrmonstration that arnC proteiii can be supplictl by de novo synthrsis tlependentj up011 the arut DNA template (14).

Vc have further dc~rlolxxl the DSh-tlirrctctl cell-fret system to study tlir L-arabiirosc oprmi. ITr have drfmetl optimum conditions for this system so that the rate of synthesis of L-nrabi- nose isomerasc per arafl gcnr in vitro is about Scj+ of the in vivo rate and we can tlctect both IJ-arabinosc isomnasc and I,-ribulo- kinasc activities, thr products of the aru.4 and araU gents, ~~pec-

tivrly. The rate of synthesis of these two riizyincs, cspcsscd in

tcnns of monomers of cilaymr per big of I>NA prr hour is very similar, demonstrating that the operon is coortlinatcly cspressed in vitro. The system is complctclg dependent upon araC protein which can be supplied either by syllthcsis in vitro from an araC+ IIN;\ trmplatc or by the addition of arc& lxotrin estractcd from whole cells. The incrcnscd sensitivity of the ill vitro system has allowed us to cornparr thr arnC 1)rotciii synthrsized in the it2 vitro system with the propertics of araC protein purifirtl from whole cells. The similar behavior in the i71 vitro system of araC protein from eithrr source strongly suggests that the two products are similar if not identical.

Xl.Y~l~:IZIALS ASI) MliTIIOIXj

All of the strains used in this study are E. coli I<12 strains which contain the ara region from E. co/i B/r. It is ncccssary to t,ransfer the ara region from IS/r to I<12 because the Xh8Odara phage Itsed as a source of I)NA is grown in I(12 and such phage I)NA would be destroyed by restriction enzymes in cell-free

2 The abbreviations used are: ryclic AAlP, cyclic adenosine 3’:5’-monophosp~~:~tc; C:(;i\ protein, caatabolite gene act,ivator protein; and IPTC:, isopropyl-1-thio-p-u-galactoI~yranoside.

2946

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

2947

L-ARABINOSE - L-RISULOSE - L-RIBULOSE-S-PHOSPHATE - D-XYLULOSE-S-PHOSPHATE

__ _--__

extracts prepared from B/r strains. SB7219(A7%), SB7228(A719), and SB7223(A7&) were constructed by Pl transduction and are otherwise isogenic derivatives of strain 514 described by Zubay et al. (15). The ara deletions used in this study are shown in Fig. 1.

Phage Strains

Xh80daru+ phage is isolated, as previously described (7), from the doubly lysogenic strain SB75OO(araA7f4) which contains Xh80 and Xh8Odara. The ara region on the Xh8Odara phage is from E. coli B/r. Otherwise isogenic derivatives of the Xh8Odara phage have been constructed which contain various deletions (Fig. 1) in the L-arabinose region. araA766 was transferred to the XhdOdara phage by an F’A766 episome by P. Cleary, University of Minnesota. Xh80daruA766 phage is isolated from SB7600 which is SB7219 lysogenized with Xh80 and Xh80daruA766. Xh%OdaraA7% phage was constructed from a cross of a F’A7SB strain with a WWdaraC- phage by N. Lee, University of California, and is isolated from SB7603 which is SB7228 lysogenized with Xh80 and Xh80daraA7%. Xh80daraZC110~44A766 phage was constructed by Pl transduction and is isolated from SB7604 which is SB7223 lyso- genized with Xh80 and U.8OdaraZ ciioZr4~~766. Xh80dlac phage is isolated from strain RV80 which was obtained from A. Riggs, Cit.y of Hope, Duarte, Calif.

Conditions for in Vitro Protein Synthesis

The procedure used for the preparation of 530 extracts is similar to that described by Zubay et al. (15) except for the following modifications: cultures of SB7223 are grown at 30” in a-liter Kluyver flasks and harvested at an Asoo of 2 with a yield of about 3.5 g of packed cells per liter of medium. The cells are im- mediately washed twice in Buffer I (0.01 M Tris, 0.014 M mag- nesium acetate, 0.06 M KCl, and 0.006 M 2-mercaptoethanol) and resuspended in Buffer II (0.01 M Tris-acetate, 0.014 M magnesium acetate, 0.06 M KCl, and 0.001 M dithiothreitol) at a ratio of 1 g of cells to 2 ml of buffer and lysed by passage through an Aminco French pressure cell at a pressure of 1300 to 2000 psi. The final dialysis is for a period of 4 to 5 hours at 4” against one change of Buffer III (0.01 M Tris-acetate, pH 8.2, 0.01 M magnesium acetate, 0.06 M potassium acetate, and 0.001 M dithiothreitol). The S30 is frozen in liquid nitrogen in 0.5- to l&ml samples and stored in a liquid nitrogen freezer.

The procedures for the preparation of phage DNA have been described (7). A final dialysis against 0.01 M Tris-acetate, pH 8.0, has been added.

Conditions for protein synthesis are similar to those of Zubay et al. (13) and are described in the legend to Fig. 1. The reaction mixture volume is 0.15 ml. The synthesis was terminated by dilution into the appropriate assay mixture which contains chloramphenicol. The activity of the protein-synthesizing sys- tem is initially monitored by measuring the synthesis of p-galac- tosidase as well as L-arabinose isomerase when the system is programmed with XhlOdlac DNA and Xh8Odara DNA, respectively. The amount of @-galactosidase produced is directly proportional

FIG. 1. The L-arabinose gene-en- zyme complex in Escherichia coli B/r. The horizontal line represents the segment of the genome which contains the genes whose products are involved in the catabolism of L-arabinose. Genes araA, araB, and araD code for the enzymes which convert L-arab- inose to n-xylulose 5phosphate. araC is the regulatory gene; ara0 is the operator; and araZ is the site where araC protein interacts. The n,umbers represent mutations.

to the amount of Zac DNA (10 to 50 pg per ml) added and is com- pletely inhibited by the addition of 4 pg per ml of highly purified lac repressor (the generous gift of A. Riggs), chloramphenicol (100 rg per ml) or rifampicin (2 rg per ml). The inhibition ob- served with Zac repressor is reversed by the presence of 6 X 10e4 M

IPTG. If the Xh8Odara and the Xh8OdZac DNA are both present in the same synthesis mixture, one can detect both p-galactosidase and L-arabinose isomerase activities. Thus, we are able to distinguish between specific effects on the expression of the L-

arabinose operon and more general effects on protein synthesis.

Assay of L-Arabinose Isomerase (EC 6.3.1.4) and L-Ribulokinase (EC d.7.f .f6)

L-Arabinose isomerase is assayed by the method of Cribbs and Englesberg (16). Assay mixture (0.1 ml or 0.15 ml) containing 0.4 M glycylglycine, pH 7.6, 0.1 M MnClz, 1.0 M L-arabinose, and 200 fig per ml of chloramphenicol is mixed with an equal volume of protein synthesis mixture and incubated at 37” for a suitable length of time (5 to 60 min). The assay is stopped by the addi- tion of 0.9 ml of 0.1 M HCl. The amount of L-ribulose produced is determined by the cysteine-carbazole test (17).

L-Ribulokinase is assayed by a radiometric method in which the conversion of L-[i%J]ribulose to L-[i”C]ribulose 5-phosphate is measured (18). Five microliters of chloramphenicol (2.5 mg per ml) and 5 ~1 of DNaee (200 pg per ml) are added to the protein synthesis mixture. After 5-min incubation at 37”, 5 to 20 ~1 of the resulting solution are added to 0.1 ml of L-ribulokinase re- action mixture containing 5.6 X lo4 cpm of L-[i4C]ribulose and incubated for 16 hours at 30”. The assay is linear for this period of time.

Preparation oj Extracts for araC Protein Activity

Step f---Cells of the appropriate strain are grown at 37” to an AWN of 2 in TYEA medium (15 g of tryptone, 10 g of yeast ex- tract, 5 g of NaCl, and 10 g of L-arabinose per liter of glass dis- tilled water) and harvested by centrifugation. All subsequent operations are performed between 0” and 5”. Five grams of cells are suspended in 125 ml of Buffer A (0.01 M Tris-acetate, pH 8.2, at 20”, 0.25 M potassium acetate, 1 mM EDTA, 1 mM dithiothreitol, 2 g of L-arabinose per liter, and 50 mg of phenylmethylsulfonyl fluoride dissolved in 10 ml ethanol per liter) and the resulting sus- pension is centrifuged at 10,000 X g for 15 min. The pellet is sus- pended in 10 ml of Buffer A and the resulting suspension is passed through a French pressure cell at pressures between 4,000 and 8,000 p.s.i. The lysate is centrifuged for 30 min at 30,000 X g. The supernatant is made to 2Oq7c saturation with solid ammonium sulfate, stirred for 30 min, and then centrifuged for 15 min at 15,000 X g. The resulting supernatant is made to 50% saturation by the addition of solid ammonium sulfate, stirred for 30 min, and finally centrifuged for 15 min at 15,000 X g, The precipitate is suspended in 2 ml of Buffer B (Buffer A without the potassium acetate) and dialyzed against this buffer for 8 hours. The dialy- sate is centrifuged for 30 min at 50,000 X g and the resulting super- natant (crude araC protein) can be used directly in the in vitro

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

2948

system, further purified by DNA cellulose chromatography, or stored indefinite1.y in nitrocellulose or polypropylene tubes in a liquid nitrogen freezer.

&en %-DNA cellulose is DreDared according to the method of Alberts el al. (19), except &at the ultraviolet-irradiation step of Litman (20) was included to help the DNA adsorb to the cellulose. The product of “Step 1” is applied to a column (3.7 X 10 cm) containing 50 g of salmon sperm DNA cellulose (8 mg of DNA per g of cellulose) and washed on with 100 ml of Buffer B. The column is further washed with 300 ml of Buffer B containing 0.3 M

potassium acetate before it is eluted with Buffer B containing 0.5 M potassium acetate. The fractions containing araC protein activity, as determined in the i7~ z&o system herein described, are pooled, frozen in liquid nitrogen, and stored in a liquid nitro- gen freezer where araC protein activity is stable for about 2 weeks.

RESULTS

Activator Function of araC Protein in Vitro-The cell-free DNA- dependent protein-synthesizing system is composed of an “S30” (cell-free extract) that cont.ains ribosomes and all of the protein factors required for protein biosynthesis, DNA from defective transducing phages carrying t.he bacterial genes to be studied, and those cofact,ors and substrates necessary for RNA and pro- tein synthesis. Our S30 is prepared from a strain of E. coli that contains deletions of both the L-arabinose and lactose genes and is, therefore, devoid of any of the product,s of t,hese systems. araC protein is obtained from a strain which contains a deletion of the araAB region. The DNA used to program the system con- tains a deletion of the araC gene t.o ensure that any activation of the operon results from the addition of araC protein and not from the synthesis of araC protein in the in vi&o system. Thus, any detectable L-arabinose isomerasc or L-ribulokinase (the products of the arad and araB genes) after the incubation is the result of a net in vitro synthesis stimulated by the addition of araC protein.

Crude araC protein added to the in vitro system stimulates L-arabinose isomerase synthesis at least go-fold after a 60.min incubation (Table I). The amount of L-arabinose isomerase produced is proportional to the amount of XMOdaraC- DNA added in the range of 4 to 60 pg per ml (data not shown). The data presented in Table I also show that zero time for synthesis, the omission of L-arabinose, cyclic AMP, araC protein, or DNA, or the addition of n-fucose or chloramphenicol result in no syn- thesis of i-arabinose isomerase. The above results, all consistent with in viva studies, suggest that the L-arabinose operon is being activated by the same mechanisms in vitro as it is in vivo.

The in vitro system can be used as an assay for araC protein activity in cell-free est.racts. The results in Fig. 2 show that the amount of L-arabinose isomerase produced is proport.ional to the amount of added araC protein. Although the addit’ion of 1 pg of araC protein purified by affinity chromatography as previously described (21) stimulates the synthesis of 1.8 x 1013 monomers of L-arabinose isomerase per ml of synthesis mixture, the instabil- ity of araC protein purified by this procedure forced us to find an alternative method for purification (see “Materials and Meth- ods”). Partially purified araC protein (1 pg) cout.ained in a 20 to 50% ammonium sulfate cut results in t.he synthesis of 8 x 1O’O monomers of L-arabinose isomerase per ml of synthesis mixture and is very stable when stored in a liquid nitrogen freezer. The crude araC protein is further purified about 300-fold by chro- matography on salmon sperm DNA cellulose. This product (1 pg) stimulates the synthesis of 2.4 x 1013 monomers of L-arabi- nose isomerase per ml of synthesis misture.

Repressor Acfivify of arac Protein in Vi&o-Initiator constitu- tivc (aral”) mutants have been described in strains containing deletions of the araC gene (6). These mutants have a cis-act.ing,

TABLE I Enzymatic activities resulting from cell-free synthesis

The protein synthesis mixture contains per ml: 43 rmoles of Tris-acetate, pH 8.2; 1.3 pmoles of dithiothreitol; 53 pmoles of potassium acetate; 0.22 pmole each of the 20 amino acids; 0.57 amole each of cytidine 5’-triphosphate, guanosine 5’-triphosphate, and uridine 5’-triphosphate, pH 7.0; 2.2 pmoles of adenosine 5’- triphosphate, pH 7.0; 21 pmoles of trisodium phosphoenolpyru- vate; 27 pmoles of ammonium acetate; 107 pg of transfer RNA from E. coli strain K; 0.13 pmole of pyridoxine-HCl; 0.036 pmole of triphosphopyridine nucleotide; 0.033 rmole of flavin adenine dinucleotide; 0.056 @mole of folinic acid; 0.19 rmole of p-amino- benzoic acid; 10 pmoles of magnesium acetate; 7.2 pmoles of cal- cium chloride; 0.2 rmole of diphosphopyridine nucleotide; 15 mg of polyethylene glycol; 20 pmoles of L-arabinose; 0.8 rmole of cyclic AMP; 37 pg of hh80daraD+A+B+Z+O+C-(A766) DNA; and 0.33 ml of S30. In addition to the above amounts the S30 con- tributes per ml of synthesis mixture: 7 to 9 mg of protein; 4.7 pmoles of magnesium acetate; 20 pmoles of potassium acetate; 0.3 rmole of dithiothreitol; and 3.3 pmoles of Tris-acetate, pH 8.2. The S30 is prepared from SB7219 which contains deletions that excise the ZacZZ and the ara A BZOC(A744) regions of the chromo- some. The source of araC protein is 5 pl per synthesis mixture of the product of “Step 1” (see “Materials and Methods”). Pro- tein synthesis is at 37” for GO min and is initiated by the addition of the S30. Enzymatic activities are assayed as described under “Materials and Methods.”

Incubation system 1 L-arabinose isomerase

Complete. Complete at zero time for synthesis. Complete, no DNA. Complete, no L-arabinose. Complete, no araC protein. Complete, no cyclic AMP. . : Complete, + 0.27 M n-fucose Complete, + 85 pg per ml of chloramphcnicol

utzils/ml

9.2 <O.l <O.l

0.1 0.1 0.3 0.1

<O.l

partially constitutivc phenotype. The enzyme levels of the mutants are repressed by the product of the araC gene in the absence of L-arabinose (repressor form) and further induced by the presence of araC gene product in the presence of L-arabinose (activator form). A mutant containing a double mutation in the initiator region (aralclc) has recently been isolated (22). This mutant has a higher constit,utive rate of ara operon espres- sion corresponding to about 20% of the fully induced wild type level but responds to the activator and repressor forms of araC protein the same as the parent aTaIc. The relative enzyme levels of the araICIC strains arc shown in Table II.

I f the in vifro system is programmed with a X1~80daraZcZcC-- (A766) DNA template, L-arabinosc isomcrase is produced in the absence of araC protein and L-arabinose (Table II). The addi- tion of both L-arabinose and araC protein results in a large in- crease in the amount of L-arabinose isomerasc synthesized. However, the addition of araC protein to a reaction mixture con- taining no L-arabinose results in a significant reduction in the synthesis of L-arabinosc isomcrase (Table II). The repression is specific for the ara operon since the synthesis of fl-galactosidase programmed by lac DNA is not affected by the addition of crude araC protein (data not shown). It can also be seen in Table II that the amounts of L-arabinose isomerase produced from the various DNA templates in vifro is proportional to the amount found in the corresponding in vivo situation.

In Vifro Synthesis of araC Protein--If the system is pro-

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

grammed with Xh8Odara+ DNA and no cwaC protein is added to the synthesis misture, L-arabinose isomerase and L-ribulokinase activities are present after the incubation (Table III). The amount of L-arabinose isomerasc produced is proportional to the amount of am+ DNA added in the range of 4 to 40 pg per ml (data not shown). The data presented in Table III also show that zero time for synthesis, no DKih, or the addition of chloran- phenicol result in no synthesis of L-arabinose isomerase. The synthesis of L-arabinose isomerase from the aru+ template is ci,cpendent upon the presence of L-arabinose and is inhibited by n-fucose. Furthermore, cyclic XX’ stimulates synthesis from the uruf template. The above results, all consistent with in viva studies, suggest that the L-arabinosc operon is being regulated by the same mechanisms in vi/r0 as it is in vivo. Therefore, aruC

protein must be synthesized in vi/r0 and in the presence of t-arabi-

5 IO 15 20 25

pl of Extract

FIG. 2. Synthesis of L-arabinose isomerase as a function of the amount of added extract-containing araC protein. Synthesis was under standard conditions except that varying amounts of araC protein-containing extracts were added to the synthesis mixture. The strains used as a source of urd protein contain deletions of the araA and aruU genes; thus the extracts contain no L-arab- inose isomerase or L-ribulokinase activities. The concentration of Xh80&ruC-(~%‘6) DNA in each synthesis mixture was 21 pg per ml. O--O, extract prepared from SB7219 (araC+) through “Step 1” containing 5 mg per ml of protein; A- --A, extract prepared from SB7219 (araG’+) after chromatography on salmon sperm DNA cellulose (“Step 2”) containing 200 rg per ml of protein; X-X, extract prepared from SB7219 (araC+) by affinity chromatography as described in Ref. 9 containing 10 pg per ml of protein; O---O, extract prepared from SB7223 (arc&) through “Step 1” containing 7 mg per ml of protein.

2949

nose activates the operon resulting in the synthesis of L-arabinose isomcrase and L-ribulokinase.

Further support for the in v&o synthesis of arc& protein is obtained from the complementation experiments presented in Table IV. Neither XU?OduruC-(A766) nor XhSOduruA-B--

(A?%) is callable by itself of stimulating significant L-arabinose isomerase synthesis, but a mixture of the two DNAs behaves like ara+ DNA. This indicates that the uruC+B-8- trmplate is

TABLE III Syttlhesis of araC protein in vitro

Protein synthesis was as described in the legend to Fig. 1 except that Xh80daraD+A+R+Z+O+C+ DNA replaces the Xh80duraD+A+B+- Z+0Y?(~766) DNA and no araC protein is added. Enzymatic activities are assayed as described under “Materials and Meth- ods.”

Incubation system

Complete.............................. Complete at. zero time for synthesis. Complete, no DNA. Complete, -Xh80clara+ DNA

+Xh80daraC-(A766) DNA.. Complete, - L-arabinose : . Complete, + 0.27~ n-fucose Complete, -CAMP. Complete, + 85 pg/ml chloramphenicol.

I

--

.-Arabinose I-Ribulo- isomerase kinase

ads/ml

4.1 <O.l <O.l

0.15 0.35 0.4 0.3

<O.l

rmils/ml

0.28 <O.Ol

<O.Ol

TABLE IV Template complemenlation studies

Protein synthesis was as described in the legend to Fig. 1 except that. the DNA template is varied as indicated and no araC protein is added. The enzyme assays were incubated for 60 min as de- scribed in “Materials and Methods.”

lZll3f

38

DNA I-arabinose isomerase

I ad-

I WUA-

M/ml 2rrtils per ml

4.1 38 0.15

28 0.1 38 28 3.2

TABLE II Activation and repression in vivo arid ill vitro

Protein synthesis was under standard conditions except that the DNA templates, L-arabinose, and aruC protein are varied as indi- cated. All of the DNAs were present at a final concentration of 37 rg per ml. Five microliters of araC protein-containing solution from “Step 1” of the purification (see “Materials and Methods”) containing40 mg per ml of protein is added just prior to the addition of the S30 where indicated. When araC protein is present in a synthesis mixture containing no L-arabinose, the product of “Step 1” has been dialyzed against Buffer B without L-arabinose for 18 hours.

strain

I-Arabinose isomerase activity in vim

u~uC- (~766) am’ (WT) ~~uZ”“~Z~~~AC-(A~~~) araZC110zC44Ac+

--ara +CiKi

y. induced wild lgpe

0.2 0.2 0.2 100

19 G 78

system I I-Arabinose isomerase synthesized in G&o

Xh8OdaraC-(A766) DNA XhBOdaraC-(A766) DNA + uraC protein Xh80daraZcl’oZC44AC- (A766) DNA Xh80duraZc110Z~44AC-(A766) DNA -I- araC protein

tmifs per ml reacfion m&we

<O.l <O.l 0.3 9.2 2.0 2.2 0.7 7.1

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

2950

synthesis) to a system containing araC- DNA and added araC 100 - x protein obtained from whole cells (in viva araC protein) the two

products can be compared. > 60 - x L 2

/

0 The effects of L-arabinose and n-fucose on “in vivo” and “in v&o” araC protein have been determined. The induction ex-

$60- x periment shown in Fig. 3 shows that the dependence of both

0 5

i

systems on L-arabinose is quite similar. An apparent K, of

I 2 X 10W2 M for the interaction of L-arabinose with both “in vitro” x 40-

2 0 aroC+DNA and “in viva” araC protein may be estimated if we assume that

k x araC- DNA+ araC protein induction of the operon reflects binding of L-arabinose to the

\. 20 -7 araC protein. The data presented in Fig. 4 show the effect of 0 the anti-inducer, n-fucose, on the expression of the L-arabinose

; operon. Once again a single curve can be drawn for both in vivo 0 4 8 12 16 20 24 26 and in vitro araC protein, corresponding to an apparent K; of

MOLARITY OF L-ARABINOSE DURING PROTEIN SYNTHESIS (X 10-2)

2 X lop2 M if we assume inhibition is the result of the binding of

FIG. 3. Synthesis of L-arabinose isomerase as a function of the n-fucose to the araC protein. The experiments reported above

L-arabinose concentration. Synthesis was under standard con- strongly indicate that the araC protein synthesized in vitro is very

ditions except that the L-arabinose concentration was varied as simi1ar to the in GvO product. indicated. After incubation at 37” for 60 min, the synthesis Kinetics oj Synthesis-The time required to synthesize an was terminated and the enzyme assay initiated by the addition of isomerase reaction mixture. 0, araA+B+C+ DNA template

inducing amount of araC protein in the in vitro system can be

was present at a concentration of 37 pg per ml. Maximum ac- determined by measuring the time required for the first appear-

tivity corresponds to 1.3 X 10” monomers of isomerase per hour ante of L-arabinose isomerase activity in the system programmed

per ml of reaction mix per pg of DNA. X, araA+B%‘- DNA at a with ara+ DNA compared to the time required in the system concentration of 38 pg per ml and 5 pl of crude araC protein were programmed with araG’-(A766) DNA in the presence of added present in each synthesis mixture. Maximum activity car- hracprote;n. responds to 2.3 X 10” monomers of isomerase per hour per ml of

All of the iompdnents of the in e&o system except

reaction mix per rg of DNA. the S30 are mixed together and divided equally among 10 tubes. Synthesis is initiated by the addition.of S30 and stopped at the

100 -A I

0 &DNA

x araC- DNA + aroC protein

I,, I T I I , I, I x

0 4 S 12 I6 20 24 26 MOLARITY OF D-FUCOSE DURING

PROTEIN SYNTHESIS (x10-*)

FIG. 4. Synthesis of L-arabinose isomerase in response to varying n-fucose concentrations. Synthesis was under standard conditions and the D-fucose concentration was varied as indicated. The concentration of L-arabinose was 6.7 X low2 M. After incuba- tion at 37” for 60 min, the synthesis was terminated and the en- zyme assay initiated by the addition of isomerase reaction mix- ture. 0, araA+B+C+ DNA template was present at a concentra- tion of 37 rg per ml. Maximum activity corresponds to 6 X 10”’ monomers of isomerase per hour per ml of reaction mix per pg of DNA. X, araA+B+C- DNA template at a concentration of 38 Mg per ml and 5 ~1 of crude araC protein were present in each synthesis mixture. Maximum activity corresponds to 7 X 10”’ monomers of isomerase per hour per ml of reaction mix per pg of DNA.

able to provide a product, the araC protein, that is required for the synthesis of L-arabinose isomerase from the araC-WA+ template. Thus, just as in vivo, L-arabinose isomerase synthesis is positively controlled by the product of the araC gene.

Comparison of in Vivo and in Vitro araC Protein-By compar- ing the system programmed with ara+ DNA (in vitro araC protein

indicated times by the addition of the L-arabinose isomerase assay mixture which contains chloramphenicol. In the system programmed with ara+ DNA the first appearance of L-arabinose isomerase is at approximately 27 min, if one extrapolates to the time axis, and the synthesis continues at a constant rate until at least 60 min. If the system is programmed with araC-(A766) DNA in the presence of added araC protein, L-arabinose isom- erase activity first appears 14 min after the initiation of syn- thesis. The lag of 13 min reflects the time required to synthesize and assemble sufficient araC. protein to activate the ara operon (Fig. 5).

The kinetics of synthesis of both L-arabinose isomerase and L-ribulokinase from an araf DNA template have also been deter- mined. At various times after the synthesis is initiated, chloram- phenicol and DNase are added to stop the reaction and the syn- thesis mixture is diluted into the appropriate assay mixture. The monomers of enzyme synthesized are calculated from the specific activities of purified L-arabinose isomerase (12 units per pg of protein (23)) and L-ribulokinase (2.6 units per pg of pro- tein3). L-Ribulokinase appears before L-arabinose isomerase and the rate of synthesis of the two enzymes is almost identical, 1.8 x 10” monomers per hour per ml of synthesis mixture per pg of DNA.

DISCUSSION

The in vitro protein-synthesizing system can produce as many as 50 monomers of L-arabinose isomerase per araA gene per hour of incubation when araC protein purified on salmon sperm DNA cellulose is added. This estimate is based on a subunit molecular weight for L-arabinose isomerase of 60,000, a specific activity for the purified enzyme of 12 units per pg of protein, and a molecular weight of 30 x lo6 for the phage DNA which contains one copy of the araA gene. At least 1 unit of L-arabinose isomerase per pg of DNA can be produced during a l-hour incubation (Fig. 2)

3 N. Lee, personal communication.

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

2951

0 5 IO 20 30 40 50 60 TIME (men) AFTER INITIATION

OF SYNTHESIS

FIG. 5. The rate of synthesis of t,-arabinose isomerase and L-ribulokinase. Synthesis was under standard condition. The synthesis was stopped at the indicated times by the addition of 5 ~1 of chloramphenicol (2.5 mg per ml) and 5 ~1 of DNase (200 pg per ml). Samples (10 and 20 ~1) were diluted into the kinase reaction mixture and 100 ~1 were added to 100 ~1 of isomerase reaction mixture. The final concentration of M80dara+ DNA (--) was 37 rg per ml. Crude araC protein (5 ~1) was present in the experiments with Xh80daraC-(A766) DNA (- - -) which was present at a final concentration of 38 rg per ml. L-arabinose isomerase (0) and L-ribulokinase (X) are reported as monomers synthesized per hour per ml of synthesis mixture per rg of DNA.

which corresponds to 1 x lOI* monomers of L-arabinose isom- erase from 2 X lOlo molecules of XhSOdaraA+C- DNA.

The differential rate of synt,hesis of L-arabinose isomerase in a wild type strain (C:PlOOO) growing with a generation time of 1 hour is 40 units per mg of protein. This corresponds to 3.3 pg or 3.3 x 1013‘monomers of L-arabinose isomerase per mg of pro- tein. Sittcc there is about 1 mg of protein in 10”’ cells, t,here are 3.3 X lo3 monomers of L-arabinose isomerase per cell. A cell in the espottetttial phase of growvth cont’aitts 3 to 4 chromosomes each of which contains one araA gene. Thus, 1000 monomers of L-arabinose isomerase are produced per araA gene in vivo. Therefore, the number of monomers of L-arabinose isomerase produced per araA gene in vitro is approximately 50/O (50/1000) of the nutnber produced in vivo if me assume that specific activi- ties of t.he in vitro and in vivo enzymes are the same. The rate of synthesis in vitro may be an underestimate since we are assum- ing that all of the DNA molecules added arc act,ive as templates.

The DNA-directed proteill-synthesizing system used in this paper appears to be more efficient than those used iu the past to study the regula,tion of the L-arabinosc operott. The most im- port.ant differences that resulted in a tnorc efficient. system arc in the preparation of the S30. Processing the cells immediately rather t’han freezing them and the use of very low pressures when lysing the cells in the French press are of primary importance. The S30 is stable for at least 6 mont.hs when stored in liquid nitrogen. The inclusion of polyethylene glycol in the synthesis mixture products a 2- to 3-fold increase in the rate of enzyme synthesis. A similar effect has also been described in in vitro studies on the tryptophan opcrou (24). The assay for L-arabi- nose isomerase is much more convenient and rapid than the assay for L-ribulokinasc. Since the two enzymes arc coordinately produced, the induct.ion of the operon is most easily monitored by assaying for L-arabinose isomerase act,ivity.

quantitative assay for araC protein when it is programmed with araP DNA. To routinely use the in vitro systetn as an assay it is necessary to standardize each DNA preparation and each S30 as stnall variations may occur. It is also necessary to deter- mine the effect of a given sample on geueral protein synthesis by measuring /3-galactosidase synthesis frotn a lac DNA template in the same synthesis mixture. Although the assay is tedious it has two advantages over the DNA bittding assay previously described for the uraC protein (21) : one can detect araC protein activity in crude estracts and the biological activity of the protein is being measured.

The csperimcnts reported in this paper demonstrate that many of the in vivo properties of the L-arabinose operon are also present in our in vifro system. TVe have shown that the in vitro system is absolutely dependent upon ard protein which can be supplied either by de novo synthesis from an araG’+ DNA template or by the addition of arc& protein purified from whole cells. arc& protein can activate the operon from the trans position in vitro as demonstrated bg the complemetttation esperiments presented iu Table III. In the absence of L-arabinose the araC protein has been shown to act as a repressor. The operon is expressed coordinately in the in vilro system and the appearance of the promoter prositnal gene product, L-ribulokinase, precedes the first appearance of the tnore distal gene product, L-arabinose isomerasc.

We find, as have others (12, 13, 14), that activation in vitro requires L-arabinosc and cyclic AMP and that t)-fucose inhibits activation in the presence of L-arabinose. Kc have compared the araC protein made in vifro with uraC protein purified from whole cells and find the two products to be very similar--the apparent I<, for L-arabittose and n-fucose is 2 X 1OP’ M for both products. It has been shown that i?z vivo the L-arabinose opcron is half-induced at, an intracellular concentration of 6 x 10e3 M L-arabinosc (25). The interaction of L-arabinosc and n-fucose with elcctrophoretically pure araC protein has been studied by fluorescence techniques4 An apparent K, of 3 x 10e3 XI for L-arabinose and 6 X 10h3 nr for I)-fucose was found. Thus, a relative high concentration of the inducer, L-arabinose, or the anti-inducer, n-fucose, is required before binding to the araC protein occurs and the results obtained both in vivo and in vitro arc very sitnilar.

The isolation of the araIclc mutants has allowed us to develop an assay for the repressor form of arc& protein using the in vitro system. The coustitutive synthesis of L-arabinose isomerase which is observed in viva in araIclcC- strains is also observed in vifro when the system is programmed with uraZcIcC- DNA. The rate of synthesis of L-arabinose isomcrase both in vivo and in vitro is decreased if araC proteiu is present in the absence of L-arabinosc.

rlcknowledgmenfs-We would like to thank Ruth Ehrittg and Geoffrey Zubay for advice during the early part of this study aud Marvin Cassman for suggesting the use of phenylmethylsulfonyl fluoride.

REFERENCES

1. ~NGLESBISRG, 15. (1971) in Metabolic Pathways (VOGEL, H., ed) 3rd Ed. Vol. 5, p. 257, Academic Press, New York

2. RNGLESIIERG, E., Itttt, J., POWICR, J., AND LEE, N. (1965) J. Bucteriol. 90, 946

3. IILR, J., AND ENGLESIHCI~G, I<. (1970) Gertelics 66, 27 4. ENGLESUEI~G, E., SQ~~IRES, C., AND MERONK, F., Jn. (1969)

Proe. Mat. Acud. Sci. U. S. A. 62, 1100

The data shown in Fig. 2 demonstrate that the system is a 4 G. Wilcox, manuscript in preparation.

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

2952

5.

6.

7.

8.

9. 10.

11.

12.

13.

14. 15.

SIIEPPARD, I$. E., AND J~~NGI,ICSI~EI~G, E. (1967) J. Mol. Ijiol. Cold Spring Harbor Laboratory of Quantitative Biology, 26, 443 New York

J’:NGLESBERG, I5., SHRPP.\III), I)., SQUIRES, C., .\ND MISIIONK, 16. F.. JI<. (1969) J. Mol. Biol. 43. 281 17.

CIUISHS, R. M., AND I*:NGI,F:SI~ERG, 15. (1964) Gene/k 49, 95 I)IX:III:, A., AND BOIWNFIII.:UND, 15. (1051) J. Biol. Chem. 192,

583 W~&ox, 6., ~~NCXR, J., AND ~~IWFI~XNAN, 1,. (1971) J. Lilac- teriol. 108, 1

CLEARY, P., AND J’:NGLIISI~IGI~G, 12. (1974) J. llacleriol., in press

Ponm, J., ,\ND IRR, J. (1973) J. Biol. Chem. 248, 7806 S~IWARTZ, I>., AND BECIWITI~, J. (1970) in The I,aclose Operon

(BECKWITH, J., AND ZIPSICIL, I)., eds) p. 417, Cold Spring Harbor Laboratory of Quantitative Biology, New York

Bass, It., HEFFERNAN, L., SW.U~NI~;R, K., AND ENGLI<;SBIGILG, E. (1974) J. Bacterial., in press

GREEN~LATT, J., AND SCHLEIF, I{. (1971) n’ature Xew Viol. 233, 166

ZURAY, G., GIICLOW, L., :\ND I~;NGLEsBERG, 1’. (1971) i\‘alure New Biol. 233, 164

YANG, H., AND ZUIL\Y, G. (1973) 1Vol. Get/. Gelret. 122, 131 ZUBAY, G., CIIAMBERS, I)., AND CWONG, I,. (1970) in The

Laclose Opera/L (BECKWITH, J., AND ZIIWX, I>., eds) p. 375,

18. LIZJC, N., AND I~NGI,ES~I~:RG, E. (1963) Proc. Nat. Acud. Sci. U. S. A. 60, 696

19. ALI~KRTS. B.. AMODIO. F.. JENKINS. M.. GUTM.IN. E.. .\ND

20. 21.

22.

23.

24.

25.

FI~MU~, F.’ (1968) Cbld k’pring Harbor ‘Symp. Qkm1. ’ Biol. 33, 289

LITMAN, It. M. (1968) J. l?ioZ. Chem. 243, 6222 WIM:OX, (;., CI,MU:TSON, K., K.\NTI, Il., .\NI) J*:NGLIGSUERG, 15.

(1971) hoc. l\;al. Acud. Sci. I/. S. A. 68, 2145 COLOMJ~:, J. (1974) Ph.D. thesis, University of California,

Santa Barbara P.\TIUCIC, J. W., .\ND LEE, N. (1968) J. Biol. Chem. 243,

4312 Poun~sr,s, P., ,\ND V.\N ROTTEI~D.\M, J. (1972) Proc. Nat. Acad.

Sci. Ii. S. A. 69, 1786 I)OYLIC, RI. E., BI~O~VN, C., HOGG, R. W., .\ND HEI,LING, R. B.

(1972) 1. Bacterial. 110, 56

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from

Gary Wilcox, Pierre Meuris, Richard Bass and Ellis EnglesbergBAD in VitroRegulation of the l-Arabinose Operon

1974, 249:2946-2952.J. Biol. Chem. 

  http://www.jbc.org/content/249/9/2946Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

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

  http://www.jbc.org/content/249/9/2946.full.html#ref-list-1

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

by guest on Novem

ber 21, 2018http://w

ww

.jbc.org/D

ownloaded from