Precursor Forms of the Subunits of Nitrate Reductase in chlA and chlB Mutants of Escherichia coli K12

Eur. J. Biochem. 105, 297-306 (1980)

Precursor Forms of the Subunits of Nitrate Reductase in chlA and chlB Mutants of Escherichia coli K12

Gerard GIORDANO, Lydia GRILLET, Janine POMMIER, Claude TERRIERE, Bruce A. HADDOCK, and Edgard AZOULAY

Laboratoire de Structure et Fonction des Biomembranes (Equipe de Recherche 143 du Centre National de la Recherche Scientifique), Unite d’Enseignement et de Recherche de Luminy, Marseille, and Department of Biochemistry, University of Dundee

(Received October 2/ December 19, 1979)

The synthesis of nitrate reductase by a parental Escherichia coli K12 strain and its isogenic chlA and chlB mutants has been analyzed by protein double labelling with L- [4,5-3H]leucine and sulphur-35 and by immunoprecipitation using specific antiserum. The chlA and chlB mutants although defective in nitrate reductase activity retain the ability to synthesise the different polypeptides that are normally required for functional enzyme activity. In addition the data shows the following.

1. These polypeptides are present in unequal quantities in the membrane and in the cytoplasm of the cells. The chlB mutant synthesizes three times more nitrate reductase than the chlA mutant.

2. The subunit composition of the membrane-bound nitrate reductase present in the two mutants is different.

3. Membrane preparations from the chZB mutant contain the three subunits c~, P, y in a ratio which is similar to the ’wild type.

4. In the chlA mutant the two subunits p and y are missing and the level of CI subunit is very low. In the same membrane a 48 000-M, subunit (polypeptide p’) precipitable by nitrate reductase antiserum has been found. The chZA and chlB mutants accumulate the three subunits c ~ , /s and y in different proportion and concentrations in the cytoplasm unlike the parental strain.

5 . The cytoplasm from the chlA mutant also contains the P’ polypeptide found in the membrane fraction of this mutant and in addition contain another polypeptide designated c ~ ’ of molecular weight 105 000 which is precipitated by the nitrate reductase antiserum.

The formation of particulate active nitrate reductase can be achieved by mixing the supernatant fractions of the chlA and chlB mutants (complementation) and procedes by two distinct but mutually dependent stages. Following reconstitution of activity the two peptides c ~ ’ and p’ present in the supernatant fraction of the chlA mutant, disappear. Analysis of the immunoprecipitate polypeptides present in both the soluble and particulate nitrate reductase protein after reconstitution suggests that these polypeptides are precursors of the CI and /3 subunits following a process that remains to be elucidated.

Nitrate reductase from Escherichia coli has been solubilized from the cytoplasmic membrane, purified and characterized by several authors [l-41. It is composed of at least three subunits: c~ (molecular weight 150000- 155000), p (60000-65000) and y (20000). According to the method of preparation and solubili- zation used, enzymes with different subunit composition and molecular weight are obtained. This hetero- geneity in the size of purified nitrate reductase is due mainly to the action of proteases during the purification procedure. Indeed MacGregor [5] has made use

Enzymes Nitrite: (acceptor) oxidoreductase (EC 1.7 99.4); NADPH : NAD’ oxidoreductase (EC 1.6.1 . l) ; ATP phosphohydro- lase (EC 3.6.1.3).

of a membrane-bound protease in her purification procedure and suggested that cleavage of the P-subunit is responsible for the release of nitrate reductase from the membrane. In addition, DeMoss [4] has analyzed the effects of limited proteolysis on nitrate reductase, previously solubilized from the membrane by detergent and subsequently purified, and shown that the P-subunit can be cleaved into two fragments of molecular weight 43000 and 17000.

A variety of mutants, designated chlA - G, have been described [5-81 with defects in their ability to synthesize a functional nitrate reductase activity. In particular it has been shown that it is possible to reconstitute both a soluble and a membrane-bound form

298 Precursor Forms of the Subunits of Nitrate Reductase of E. coli

of nitrate reductase by mixing supernatant extracts prepared from chlA and chlB mutants [6,9]. This has been confirmed and extended to show that these mutants accumulate the different polypeptide subunits of nitrate reductase in an inactive form in the cytoplasm during growth anaerobically in the presence of nitrate [lo]. Further analysis of this complementation process has shown that the formation of reconstituted nitrate reductase requires phospholipid [l 11, and that lipids are present in the supernatant fractions of chlA and chlB mutants as small vesicles, as demonstrated by smooth fracture faces in freeze-etching studies [12]. During the reconstitution of membrane- bound nitrate reductase, it is also significant that both ATPase and cytochrome b E:' become incorporated into the vesicles [13,14]. At the morphological level, the reconstitution of membrane-bound nitrate reductase by mixing supernatant fractions from chlA and chlB mutants results in the production of vesicles of asymmetric structure, similar to those of native membranes, as shown by freeze-etching studies of the fracture faces [35].

In the present study, we have analyzed the role and fate of the various subunits present in the supernatant fractions of the chlA and chlB mutants during the complementation process so as to define the mechanism of reconstitution of membrane-bound nitrate reductase. To avoid interference from endogenous proteases, we have used routinely extracts prepared in the presence of inhibitors of these enzymes.

MATERIALS AND METHODS

Strains, Growth Conditions and Preparation of Sub-cellular Fractions

The wild-type strain 541 (F-, thr, leu, his, pro, arg, the, ade, gal, lacy, malE, xyl, ara, mtl, str, TIR) and the deriwd chlorate-resistant mutants chlA and chlB have been described [6,16].

The medium used routinely for growth contained per liter: K2HP04 (12 g), KH2P04 (3 g), (NH4)*S04 (4 g), KN03 (10 g), MgC12. 6 H20 (0.2 g), a sulphate- free mineral salts solution (10 ml) as described previously [17], glucose (2 g), vitamin-free casamino acids (1 g), and other supplements required by these strains (all at 10 mg). In addition ammonium molybdate (1 pM) and potassium selenite (1 pM) were present.

For 3H-labelling, the chlA mutant was grown on this medium anaerobically for 12 h, harvested, and re- suspended in the same medium containing ~- [4 ,5-~H]- leucine (150 pM, specific activity 32 Ci/mol). Growth was allowed to proceed for a further 2 h and the cells reharvested. For 35S-labelling experiments, vitamin- free casamino acids and (NH4)2S04 were omitted from the growth medium and the latter replaced with NH4C1 (4 g) and K25S04 (specific activity 25 mCi/

mmol) at a concentration of 40 pM for strain 541 and 200 pM for the chlB mutant.

Cells were harvested, washed, broken by ultra- sonic disruption and membrane particles prepared by differential centrifugation as in [I71 except that all buffers contained benzamidine HCI (5 mM) and phenylmethylsulphonylfluoride (2 mM) to inhibit protease activity.

Triton-X-100-soluble extracts from these membrane particles were prepared by suspending particles (1 - 5 mg protein/ml) in sodium phosphate buffer (50 mM, pH 7.4) containing benzamidine HCl(5 mM) and phenylmethylsulphonylfluoride (2 mM) and Tri- ton X-100 (1 %, w/v) at 0°C for 30 min, and then centrifuging at 230000 x g for 20 min at 4 "C. The supernatants so obtained are referred to as Triton X-100 extracts.

Enzyme Assays

Nitrate reductase activity, NADH-NADP+ transhydrogenase activities, the quenching of acridine dye fluorescence, and the concentrations of protein and cytochromes were measured or performed as described in previous publications from this laboratory [18,19].

Antiserum to E. coli nitrate reductase was raised in rabbits using enzyme purified essentially as in [20].

Double immunodiffusion analysis was performed in agar (1 %, wiv) plates containing sodium phosphate (50 mM, pH 7.4), Triton X-100 (0.5"/,, w/v) [21]. Activity staining of agar plates for nitrate reductase activity was performed as described by Lund and DeMoss [22].

Immunoprecipitation was performed by the addition of 50 pl of antiserum/mg protein to Triton X-100 extracts, incubating for 2 h at room temperature and then for a further 14 h at 4^C. Protein A/Sepharose (17 pg) in Tris buffer (25 mM adjusted to pH 7.5 with HCl) containing ethylene diamine tetraacetate (5 mM), phenylsulphonylfluoride (2 mM) and azide (0.5 g/l) was then added to the reaction mixture, and stirred for 1 h at 4'C. The immunoprecipitates were collected by centrifugation at 10000 x g for 2 min, washed five times with a Tris buffer (100mM) adjusted to pH 6.8 with HCl) containing LiCl (0.5 M) and nonidet P40 (0.5 x, wiv), twice with a second Tris buffer (150 pM adjusted to pH 6.8 with HCl), and finally three times in a sodium phosphate buffer (10 mM, pH 6.8) containing Triton X-100 (1 x, w/v) and NaCl(0.3 M). To confirm that all the antigen had been precipitated, the residual supernatant, following the first centrifugation, was re-incubated with antiserum and protein A/Sepharose. In addition no nitrate reductase activity could be detected in the supernatant following immunoprecipitation of Triton X-100 extracts of particles from strain 541, of reconstituted particles, and of the complementation supernatant.

G. Giordano, L. Grillet, J. Pommier, C . Terriere, B. A. Haddock, and E. Azoulay 299

Reconstitution of Nitrate Reductase Activity by Complementation between Soluble Extracts of the chlA and chlB Mutants

This was achieved using the conditions described by Azoulay et al. [6] with the supernatant fraction of the chlA mutant, the proteins of which were labelled with ~-[4,5-~H]leucine (25.8 x lo6 dis. x min-' x mg protein- ') and the supernatant fraction of the chlB mutant whose proteins were labelled with KP5S04 (8.8 x lo6 dis. x min-' x mg protein-'). After incubation for 2 h, the complementation mixture was centrifuged for 90 min at 220000 x g so as to separate the newly formed particulate fraction from the mixed supernatant fractions.

Polyacrylamide Gels

Electrophoresis in the presence of sodium dodecyl- sulphate was performed with polyacrylamide gels (6.5 x, wiv) as described previously [23]. The distribution of radioactivity in the gels was determined by scintillation counting of 1-mm slices of the gels previously solubilized by treatment with perhydrol(3.34 M) and heat (50 "C) for 14 h. The values, expressed in dis./ min, were calculated from reference curves obtained using Amersham standard samples (3H = 4 x lo6 dis. x min-' x g-', and 35S = 1 x lo6 dis. x min-' x g-') treated in a similar fashion.

RESULTS

Antigenic Characterization of the Nitrate Reductase Activity Reconstituted by Complementation between the Supernatant Extracts of the chlA and chlB Mutants

Incubation of a mixture containing equivalent pro- portions of the supernatant fractions of cell-free extracts of chlA and chlB mutants under the conditions described by Azoulay et al. [6] leads to the reconstitution of nitrate reductase activity, an activity which is normally absent in either of the mutants grown anaerobically in the presence of nitrate (Table l). If this mixture is subsequently centrifuged to yield in turn a newly-formed particulate fraction and a supernatant fraction then it is found that 45 o/, of the reconstituted nitrate reductase activity and 15 % of the protein is associated with the particulate fraction (Table 1). In addition these particles contain b-type cytochromes (0.3 f. 0.1 nmol cytochrome blmg particulate protein). It is also noteworthy that a preparation of the supernatant fractions of the chlA and chlB mutants in the presence of the protease inhibitors benzamidine and phenylmethylsulphonylfluoride results in the reconstitution of a particulate nitrate reductase activity some 35 % higher than that of particles derived from

supernatant fractions prepared in the absence of these compounds (compare the data of Table 1 with that given in [6]).

Using anti-(nitrate reductase) serum raised in rabbits inoculated with nitrate reductase of Escherichia coli K12 prepared according to Clegg [3] and Ouchter- lony double immunodiffusion analysis, it can be shown that Triton X-100 extracts of these reconstituted particles give precipitin lines characteristic of an antigen- antibody reaction (Fig. 1). In addition it will be noted that the supernatant fraction remaining after sedimen- tation of these particles, as well as the supernatant fractions of the chlA and chlB mutants before mixing, all show precipitin lines. Although a Triton X-100 extract of membrane particles derived from the wild- type strain 541 grown anaerobically in the presence of nitrate gives a precipitin line, it is noteworthy that the supernatant fraction derived from this strain does not.

In addition the precipitin lines obtained with (a) membrane prepared from strain 541 and (b) both the supernatant fraction and the newly formed particulate fraction obtained following incubation of supernatants of the chlA and chlB mutants, all possess nitrate reductase activity when assayed directly on the agar plates suggesting that the antigenic sites are distinct from the sites of enzyme activity.

The Nitrate Reductase Polypeptide Composition in the Supernatant and Particulate Fractions of the chlA and chlB Mutants

If membrane particles derived from strain 541 grown anaerobically in the presence of nitrate are solubilized with Triton X-100 and challenged with antiserum to nitrate reductase and protein AISepha- rose, then nitrate reductase activity is precipitated. If proteins are labelled with either L- [4,5-3H]leucine (450 pM, 32 Ci/mol) or KP5S04 (200 pM, 25 Ci/mol) and assuming that proteins are uniformly labelled with these isotopes, it can be calculated that nitrate reductase accounts for 3.5-4.0% of the total membrane-bound proteins in these cells (Table 2). This figure is some 4 - 5-fold lower than that reported previously for another wild-type strain [24].

If membrane preparations from the chZA and chlB mutants grown in the presence of ~-[4,5-~H]lleucine or Kq5S04 are treated in a similar fashion it can be shown that nitrate reductase comprises some 0.25 to 0.4% of the total membrane proteins although, of course, it is without enzymic activity (Table 2). In addition the supernatant fractions from chlA and chlB mutants, unlike the supernatant fraction derived from strain 541, do contain immunoprecipitable nitrate reductase polypeptides, again devoid of enzymic activity. The concentration of these polypeptides is higher in the supernatant fraction from the chlB mutant


Table 1. Distribution of nitrate reductase and protein following complementation of chlA and chlB mutants Reconstitution was performed as described in Materials and Methods in the presence of benzamidine (5 mM) and phenylmethylsulphonylfluoride (2 mM) using the supernatant extracts of the chlA mutant labelled with ~-[4,5-~H]leucine (25.8 x lo6 dis. x min-’ x mg protein-’) and the chlB mutant labelled with K;5S04 (8.8 x 106 dis. x min-’ x mg protein-’). After 2 h incubation the mixed supernatants were separated into the newly formed particulate fraction and the supernatant fraction after complementation by centrifugation. The values in parentheses correspond to the relative proportion of the proteins from the chlA and chlB mutants expressed as a percentage of the total amount of protein in the reconstituted particles assuming that the j5S and 3H tracers uniformly label proteins. The amounts of active or inactive nitrate reductase from the antigen-antibody precipitates are expressed in pg protein and were calculated as described in the legend to Table 2

Fraction Radioactively Nitrate reductase activity Antigen-antibody protein complex

jH 35s Specific activity Total 3H j5S activity

mg (%I pmol x h-’ x mg protein-’ pmolih pg protein

Supernatant fraction of chlA mutant 7.0 0 0 0 10.6 0 Supernatant fraction of chlB mutant 0 7.0 0 0 0 37.0 Supernatant fraction

after complementation and centrifugation 5.46 6.12 13 150.5 5.76 + 28.41

Newly formed particulate fraction 1.29 0.83 60 127.8 5.43 + 15.66 (60.8) (39.2)

Fig. 1. Antigenic detection for the presmce of nitrate reductuse in various preparations, Precipitation was made by double diffusion in agar medium according to the technique of Ouchterlony [22]. The centre well contained antinitrate reductase serum and the other wells contained: (a) membrane particles from strain 541 ; (b) newly formed particulate fraction following complementation of supernatants from the chlA and chlB mutants and centrifugation; (c) supernatant fraction from wild-type strain 541 ; (d) supernatant fraction from the chlB mutant; (e) the resultant supernatant fraction following complementation of the supernatants from the chlA and chlB mutants and centrifugation; and (f) the supernatant fraction of the chlA mutant. All extracts were first incubated with Triton X-100 (1 w/v)

(0.6 to 0.7 % of the total soluble proteins) than in the equivalent fraction from the chlA mutant (0.15 to 0.20% of the total soluble proteins). From the data of Table 2 it can be calculated that the total quantity of defective nitrate reductase synthesized by the chlB mutant (membrane fraction plus supernatant) is about 3.5-times greater than that produced by the chlA mutant. However, the level of defective nitrate reduc-

tase present in membranes derived from the chlA and the chlB mutant is similar and only 10 ”/, of that found in membranes from the wild-type strain 542. The data also indicate that in the chlA and chlB mutants the greater part of the polypeptides of nitrate reductase accumulate in the cell cytoplasm rather than the cytoplasmic membrane confirming previous observations [7,9,11,24], indeed the cytoplasm of the wild-type strain contains no nitrate reductase (Table 2).

The various precipitates obtained following incubation of antiserum to nitrate reductase and protein A/ Sepharose with Triton-X-1 00-solubilized membranes from strain 541 and its derived chlA and chlB mutants were submitted to electrophoresis in 6.5 ”/, (w/v) polyacrylamide gels in the presence of sodium dodecyl sulphate to determine the relative number and size of the nitrate reductase polypeptides in the different preparations. The data obtained are summarized in Fig. 2 and Table 3 and indicate the following. In accor- dance with previous reports [2] , the subunit composition of the membrane-bound nitrate reductase of the wild-type strain 541 (Fig.2A) is composed of three subunits, CI, f i and y that are present in the ratio 0.9 : 1 : 2 (Table 3) assuming a uniform distribution of the radioactive 35S label between the different subunits. These same three subunits are present in the membrane of the chlB mutant (Fig.2B) in an approximately similar ratio (Table 3). In contrast the polypeptide composition obtained with membranes from the chZA mutant is different in that the a-polypeptide is much reduced in amount, the y and b polypeptides appear to be absent, and the latter is replaced by a P’ polypeptide of molecular weight about 48 000 : the obser-

G. Giordano, L. Grillet, J. Pommier, C. Terriere, B. A. Haddock, and E. Azoulay 301

Table 2. The subcellular distribution of nitrate reductase in cells,from strain 541 and the derived chlA and chlB mutants The particulate extract (membrane) and supernatant (cytoplasm) fractions were prepared from cells grown anaerobically with nitrate in the presence of either ~-[4,5-~H]leucine (150 pM, 32 Ci/mol) or KySO4 (200 pM, 25 Ci/mol). Antiserum to nitrate reductase (50 pl) was added to the membranes previously solubilized with Triton X-100 (1 %, w/v), either 1 mg membrane protein for strain 541 or 5 mg membrane protein for the chlA and chlB mutants. These mixtures were incubated for 2 h at 20"C, then overnight at 4°C. Protein A/Sepharose (17 pg) was added to the mixture and the incubation continued with stirring for a further 1 h to precipitate soluble antigen-antibody complex. The amounts of active and/or inactive nitrate reductase polypeptides were determined from the quantity of radioactivity in the antibody-antigen precipitates and are expressed as percentage, calculated from the ratio of the radioactive proteins in the antigen-antibody complex to the total protein of the initial extracts, or as mg protein, calculated for 1 g dry weight of cells corresponding to 150 mg membrane particles or 600 mg of soluble proteins

Fraction x Radioactivity Nitrate reductase activity Radioactive proteins incorporated precipitated by antiserum into the proteins (lo6 dpm/mg protein)

to nitrate reductase

3H 35s

dis.jmin-' xmg protein-' pmolx h-' xmg protein-' % mg

Membranes from strain 541 16.80 300 Cytoplasm from strain 541 17.66 0 Membrane from chlB mutant 9.30 0 Cytoplasm from chlB mutant 8.81 0 Membrane from chlA mutant 26.62 0 Cytoplasm from chlA mutant 25.80 0

''O 1 6.0 3.5 -4 0 0 0.25 - 0.40 0.60-0.70 0.28 - 0.40 0.58 0.1 5 -0.20 1.02

Table 3. The subunit composition of nitrate reductase in various fractions and preparations Nitrate reductase proteins from membranes of strains 541, the chlA mutant and the chlB mutant, from supernatant extracts of the chlA and chlB mutants, and from the membrane fraction and supernatant fraction obtained after mixing the supernatant extracts derived from the chlA and chlB mutants were precipitated and analyzed as indicated in Table 2 (see also the legends to Fig. 2, 3 and 4). The values given in parentheses indicate the molar ratio of the subunits in each preparation and were calculated from their respective molecular weights and the amount of protein present in each subunit assuming equivalent labelling of each subunit with either isotope

Fraction Labelled proteins contained in the subunits of nitrate reductase

G! G!' P P' Y

'H 35s 3" 35s 3H 35s 3H 3 5 s 3H 35s

Membrane 541 - - 15.44 - 0

Membrane chlB -

Membrane chlA 0.30 -

Supernatant chlA 2.42 -

Supernatant chlB -

Supernatant after complementation 1.42 + 11.06 0 0

Newly formed particles 1.46 + 2.42 0 0

(0.9)

(0.7) 0.75 - 0

0 1.02 -

12.10 - 0 (0.6)

-

(0.7)

(0.7)

(0.3)

- 1.61 - 0 (1) 0.45 - 0

(1)

-

0 - 0.93 -

1.42 - 1.03 -

(1) 8.94 - 0

0.93 + 6.71 0 0

1.31 + 4.53 0 0

-

(1)

(1)

(1)

- 4.71

- 0.33

0 1.86 -

(4) - 4.11

(1.9)

(2.3) -

(1 3 ) 0.68 + 1.89

(1) 1.41 + 2.42

(2)

vation is at variance with data obtained previously by MacGregor [24] with a different chlA mutant.

If the supernatant extracts obtained with the chZA and chlB mutants are analyzed in a similar way then the following conclusions are apparent. Firstly the supernatant extract of the chZB mutant contains the same three polypeptides of nitrate reductase (Fig. 3 A) in a molar ratio of 0.6 : 6.0 : 1.5 (Table 3). On the other

hand the supernatant fraction of the chlA mutant contains five immunoprecipitable polypeptides (Fig. 4A). Three of these correspond to the a, p, y subunits of nitrate reductase and two others designated a' (molecular weight 105 000) and p' (molecular weight 48000), the latter component being similar in size to that identified in the membrane fraction of this mutant (Fig. 2C).


The Distribution of Polypeptides from Supernatant Extracts from the chlA and chlB Mutants Following Complementation and Reconstitution of Nitrate Reductase Activity

When supernatant extracts of the chlA and chlB mutants grown anaerobically with nitrate in the presence of ~-[4,5-~H]leucine and K25S04 are mixed under conditions when proteolysis is essentially prevented with protease inhibitors, it is possible to reconstitute and isolate both membrane-bound and soluble forms of nitrate reductase activity that are labelled with 3H and 35S (Tables 1 and 3).

The relative distribution of 3H-labelled polypeptides (derived from the chlA supernatant extract) and 35S-labelled polypeptides (derived from the chlB mutant) in these newly formed particles and in the supernatant fraction were analyzed. The data indicate that for the newly formed particles 60.8 % of the total membrane protein is derived from the supernatant extract of the chlA mutant (Table 1) whereas the nitrate reductase protein content of these particles is composed largely (in the ratio 3 to 1) of polypeptides from the chlB supernatant extract rather than from the chlA mutant (Table 3). A similar analysis for the soluble form of nitrate reductase, following complementation, indicates again that it is composed chiefly of polypeptides from the chlB mutant rather than from the chlA mutant in a ratio of 5 to 1.

The individual polypeptides present in these newly formed membrane-bound (Fig.3C and 4C) and soluble (Fig. 3 B and 4B) types of nitrate reductase activity were characterized using antiserum and the tech- niques described above. Both nitrate reductase com- plexes contain the characteristic three subunits a, b, and y in the ratio 0.3: 1 : 2 and 0.7: 1 : 1 respectively (Table 3).

The data presented above on the antigens precipitable with antiserum to nitrate reductase present in the supernatant fractions of the chlA and chZB mutants before and after complementation show that the a' and p' polypeptides of molecular weight 105000 and 48 000 respectively are only present in the supernatant fraction of the chlA mutant and certainly are not present in the newly formed functional enzyme. One explanation for this is that the a' and p' polypeptides are converted into a and b polypeptides following association with other polypeptides which will of course be radioactively labelled but which do not have antigenic sites recognized by the antiserum to nitrate reductase. This suggestion obtains some support from the observation that, following complementation, the amount of immunoprecipitable nitrate reductase (inactive and active) in the newly formed particles and associated supernatant fraction has increased by about 15 o/, as compared with the amount of immunoprecipi-

Fig. 2. Polypeptide proji'les 017 sodium dodecyl sulphate/polyacrylamide gels of antibody precipitahle protein f i o m the Triton-X-100- soluhilized cytoplasmic membranes of strain 541 and its derived chlA and chlB mutants. Cells were grown anaerobically in the presence of nitrate with either K:5S04 (for strain 541, and the chl5 mutant) or ~-[4,5-~H]leucine (for the chlA mutant). The membrane fraction was prepared as indicated in Materials and Methods. Aliquots of 1 mg protein for strain 541 (16.8 x 10" dis. x min-' x mg protein-') or 5 mg protein for the chlA and chl5 mutants (26.62 x 10" dis. x min ' x mg protein-' or 9.30 x 10" dis. x min-' x mg protein-' respectively) were solubilized with Triton X-100, and the resultant incubated antiserum to nitrate reductase as indicated in the text. After treatment with protein A/Sepharose the mixture was incubated for 1 h at 4 "C to precipitate the soluble antigen-antibody complex. The immunoprecipitates were collected by centrifugation, treated as described in Materials and Methods, and submitted to electrophoresis on polyacrylamide gels in the presence of sodium dodecyl- sulphate. The profiles obtained are for membranes from the wild- type strain 541 (A), the chl5 mutant (B) and the chlA mutant (C) with the migration front indicated by the arrow

table nitrate reductase (inactive) present in the supernatant extracts of the chlA and rhlB mutants before complementation (Table 1). The data given in Table 3


0

I C

15.0 -

25 -

0 0 K) 20 30 40 50 60

I I I I I 10 20 30 40 50 60

- 1

0 Slice number Slice number

Fig. 3. Polypeptide profiles on sodium dodecyl sulphnte~polyuc~ryI- amide gels of35S-labelled antibody-precipitable proteinsjrom difj'erent preparations. Cells were grown and treated as indicated in the legend to Fig. 2 and in Materials and Methods to obtain the following traces: (A) supernatant extract of the chlB mutant; (B) the supernatant fraction obtained after complementation between the supernatant extracts of the chlA and chlB mutants followed by centrifugation; and (C) the newly formed particulate fraction obtained after complementation between the supernatant extracts of the chlA and chlB mutants followed by centrifugation. Complementa- tion was achieved in the presence of benzamidine (5 mM) and phenylmethylsulphonylfluoride (2 mM) with the supernatant extract of the chlA mutant labelled with ~-[4,5-~H]leucine (25.80 x lo6 dis. x min-' x mg protein-') and the chlBmutant labelled with K?'SO4 (8.81 x lo6 dis. x min-' x mg protein-'). Immunoprecipitation was achieved under the same conditions as described in the legend to Fig.2 with 5 mg protein for the supernatant extracts and 1 mg protein for the reconstituted particles. Note that B and C are for double-labelled gels and the corresponding 31-I-labelled polypeptide profiles are given in Fig.4B,C respectively

show in addition that this increase is linked for the most part with an increase in the amount of the p-subunit rather than the %-subunit.

25

20

15

10

5

0 I

c .- . .- 2 lo I 0

Tg

5

0

15

10

5

~

A

I I I I I

A B

Fuwtional Characterization of Membrane-Bound Nitrate Reductuse Activity Reconstituled after Complementation between the Supernatant Extracts of the chlA and chlB Mutunts

The nitrate reductase protein reconstituted after complementation and incorporated into the newly formed particles, comprises only 0.93 % of the total particulate protein (Table 1). It has been shown previ-


Table 4. Quenching of 9-amino-6-chloro-2-methoxyacridine fluorescence by membrane partirles from strain 541 grown anaerobically in the presence of nitrate and by particles obtained after mixing the supernatant extracts of the chlA and chlB mutants Respiration-driven quenching of dye (0.25 mM) fluorescence was measured in the presence of oxygen or nitrate (50 mM) with NADH (5 mM), D-lactate (2.5 mM) or formate ( 5 mM) as substrate. ATP-dependent quenching was measured aerobically in the absence of electron donors following ATP (2.5 mM) addition. Membrane particles were present at a final concentration of 0.4 mg protein/ml. The values obtained for the newly formed particles are expressed as a percentage of the values obtained with particles prepared from strain 541

Fraction Quenching of 9-amino-6-chloro-2-methoxyacridine fluorescence

Respiration dependent ATP-dependent

Oxygen Nitrate

D-lactate NADH formate D-hCtate NADH formate

Particles from strain 541 100 100 100 100 100 100 100 Newly formed particles 53 48 8 10 15 5 43

ously that these particles incorporate other membrane- bound proteins during their formation such as cytochromes [14], ATPase activity [13] as well as phos- pholipids. In addition, these particles have an asymmetric structure as revealed by freeze etching [15]. It was, therefore, of interest to determine if these newly- formed particles were capable of those energy-linked functions that are characteristic of the bacterial cytoplasmic membrane.

The quenching of acridine dye fluorescence has been used widely to study the functional proton trans- locating ability of membrane vesicles that have an inside-out orientation with respect to the original cell [25]. The quenching of 9-amino-6-chloro-2-methoxy acridine fluorescence by membrane particles derived from the wild-type strain 541 grown anaerobically in the presence of nitrate was compared with that by these newly formed membranes obtained after complementation (Table 4). Data obtained for particles from strain 541 were similar to those obtained previously [26], and it is significant that D-lactate and NADH-dependent respiration linked either to oxygen or nitrate as well as ATP hydrolysis could give rise to acridine dye fluorescence quenching in these newly formed particles. The extent of quenching obtained with these particles was less than that obtained for particles prepared from strain 541, and formate was a poor electron donor, however, the data indicate that at least a proportion of these newly formed particles are vesicular (inside out with respect to the original cell) and capable of establishing and maintaining a proton gradient.

In addition, energy-linked respiration-dependent (10.3 nmol NADPH formed x min-' x mg protein-') and ATP-dependent (8.7 nmol NADPH formed x min- ' x mg protein-') transhydrogenase activities could be demonstrated in these newly formed membrane particles, values that are about 50 % those obtained with membrane particles derived from strain 541, confirming that at least a proportion of these

newly formed vesicles are capable of energy-linked functions.

DISCUSSION

During the formation of a functional membrane- bound nitrate reductase following the mixing of the supernatant extracts of the chlA and chlB mutants, two steps that may be distinct or mutually dependent, can be identified. Firstly the formation of a functional enzyme and secondly the insertion of this enzyme into the membrane. It has been known for some time that mutations in both the chlA and chlB genes are pleio- tropic in that they result in the loss of formate dehydrogenase activity as well as nitrate reductdse activity. It has been suggested that the product of the chlA gene is some molybdenum cofactor, that it is inserted into both apo-formate dehydrogenase and apo-nitrate reductase by the product of the chlB gene, an association factor [9].

In agreement with others [27] we have shown here that chlA and chlB mutants lack nitrate reductase activity yet retain, at least in part, the ability to synthesize the various polypeptides that are normally required for the functional activity of this enzyme. However, some of the data is at variance with that reported previously with respect to the amount and size of subunits present in the membrane and cytoplasm of the chlA and chlB mutants. In particular the two mutants differ in the following respects: (a) the chlA mutant synthesizes three-times-less nitrate reductase protein than the chlB mutant; (b) the membrane- bound form of inactive nitrate reductase in the chZB mutant resembles in composition if not total amount that of the wild type whereas in the chlA mutant the p and y polypeptides are essentially missing, the amount of u-subunit diminished and a new polypeptide, designated p', is observed; and (c) the soluble form of inactive nitrate reductase present in the cytoplasm of the chlB mutant again shows the characteristic a, p, y


subunit composition of functional nitrate reductase whereas the corresponding fraction from the chlA mutant shows two additional immunologically iden- tifiable polypeptides designated a‘ and p’ as well as the a, p, y polypeptide pattern. Following reconstitution of nitrate reductase activity and the characterization of both soluble and membrane-bound forms of the enzyme by mixing the supernatant extracts of the two mutants, it was shown that these two additional polypeptides disappear. It is, therefore, tempting to specu- late that these two polypeptides are precursors of functional a and B polypeptides that accumulate in the chlA mutant in the absence of the chlA gene product. It appears unlikely that a’ and p’ polypeptides are proteolytic degradation products since protease inhibitors were present throughout the preparation proce- dures, they are associated uniquely with the chlA mutant, and they disappear following complementation. In addition it is unlikely that they are contami- nants precipitated by some impurity present in the antiserum to nitrate reductase since again they are only found associated with fractions from the chlA mutant. Direct evidence that the a‘ and p’ polypeptides are indeed precursors in the formation of functional nitrate reductase is not yet available. It is significant, however, that following complementation the amount of immunoprecipitable material present in the newly formed particles and the derived supernatant fraction is greater than that found in the original supernatant extracts of the chlA and chlB mutants.

Little information exists to define how nitrate reductase integrates asymmetrically into the cytoplasmic membrane to catalyze a transmembranous electron flow. Certainly the polypeptides are arranged across the membrane such that the y subunit is located on the periplasmic surface and the a subunit is located on the cytoplasmic surface [28]. It has been suggested that the y subunit or cytochrome h E:’ is inserted into the membrane first and the other subunits follow, based upon the observation that in a hemA mutant, which lacks the ability to synthesize functional cytochromes, the a and p subunits of nitrate reductase accumulate in the cell cytoplasm [27]. However, closer analysis of the data suggests that there is still as much a and p subunits present in the membrane of the hemA mutant as in the wild type and that the hemA mutation results in an overproduction of a and p subunits rather than their altered sub-cellular distribution (R. A. Clegg and B. A. Haddock, unpublished observations). An alternative proposal is that the functional enzyme is a dimer and that aggregation is dependent upon a short fragment of the p subunit that can be removed by trypsin : proteolytic cleavage, such cleavage, results in the inability of the enzyme to dimerize although enzymic activity is retained [4]. The physiological significance of this observation remains to be determined, however.

From the results presented here it can be concluded that as a result of mutation in the chlA and chlB genes the polypeptides of nitrate reductase accumulate in the cell cytoplasm. Reconstitution of a membrane- bound functional nitrate reductase can be achieved following incubation of a mixture of the supernatant fractions of the two mutants and the polypeptides of this newly-formed enzyme come largely from the chlB mutant. This suggests that the product of the chlB gene is required for membrane insertion as well as enzyme activation; conversely it can be argued that the product of the chlA gene is required for a’B’ to aB conversion. Such suggestions are clearly speculative but the experimental system clearly provides an oppor- tunity for resolving the question of the method of assembly of this membrane-bound multienzyme complex.

The authors wish to thank Dr Marcelle Zacek for interesting discussions and Dr David H. Boxer for a gift of nitrate reductase antiserum.

REFERENCES

1

L

3 4 5 6

7 8

9

10

11

12

13

14

1s

16

17 18

19

20

21 22

MacGregor, C. H., Schnaitman, C. A,, Normansell, D. E. &

Enoch, H. G. & Lester, R. L. (1975) J . Bid. Chem. 250,

Clegg, R. A. (1976) Biochem. J . 153, 533-541. DeMoss, J . A. (1977) J. Biol. Chem. 252, 1696-1701. MacGregor, C . H. (1975) J . Bacteriol. 121, 1101 -1110. Azoulay, E., Puig, J. & Couchoud-Beaumont, P. (1969) Bio-

Guest, J. R. (1969) Mol. Gen. Genet. 105, 285-297. Glaser, J. H. & De Moss, J. A. (1972) Mol. Gen. Genet. 116,

Riviere, C., Giordano, G., Pommier, J. & Azoulay, E. (1975)

MacGregor, C. H. Sr Schnaitman, C. A. (1973) J. Bacteriol.

Azoulay, E., Pommier, J. & Riviere, C. (1975) Biochim. Bio-

Mutaftschiev, S., Olive, J., Bertrand, J. C. & Azoulay, E. (1977)

Giordano, G., Riviere, C. Sr Azoulay, E. (1975) Biochim. Bio-

Azoulay, E., Riviere, C., Giordano, G. & Pommier, J . (1977)

Mutaftschiev, S., Olive, J. Sr Azoulay, E. (1976) C.R. Hehd.

Giordano, G., Rosset, R. & Azoulay, E. (1977) FEMS Micro-

Poole, R. K. & Haddock, B. A. (1974) Biochem. J . 144,77- 85. Kemp, M. B., Haddock, B. A. & Garland, P. B. (1975) Bio-

Giordano, G., Grillet, L., Rosset, R., Dou, H., Azoulay, E.

Enoch, H. G. & Lester, R. L. (1976) J . Biol. Chem. 250,6693-

Ouchterlony, R. (1958) Prog. Allergy, 5, 1-78. Lund, K. Sr DeMoss, J. A. (1976) J . Biol. Chem. 251, 2207-

Hodgens, M. G. (1974) J . Biol. Chem. 249, 5321 - 5327.

6693 - 6705.

chim. Biophys. Acta, 237, 579- 590.

1 - 10.

Biochim. Biophys. Actu, 389, 219-235.

114, 1164-1176.

phys. Acta, 389, 236-250.

J . Biol. Celluluire, I , 17-22.

phys. Actu, 389, 302-318.

FEBS Lett. 79, 321 - 326.

Seances Acud. Sci. 283,825 - 828.

hiol. Lett. 2, 21 -26.

chem. J . 148, 329 - 334.

& Haddock, B. A. (1978) Biochem. J . 176, 553-561.

6705.

2216.

306 G. Giordano et al. : Precursor Forms of the Subunits of Nitrate Reductase of E. coti

23. Cohen, P. (1973) Eur. J . Biochem. 34, 1 - 14.

25. Azzone, G. F. s( Massari, S. (1973) Biochim. Biophys. Acta,

26. Haddock, B. A. & Kendall-Tobias, M. W. (1975) Biochem. J .

27. MacGregor, C. H. (1976) J . Bacteriol. 126, 122-131. 28. Boxer, D. H. & Clegg, R. A. (1975) FEBS Lett. 60, 54-57.

24. MacGregor, C. H. (1975) J. Bacteriol. 121, 1117-1121. 1.52, 655 - 659.

301, 195-226.

G. Giordano, L. Grillet, J. Pommier, C. Terriere, and E. Azoulay*, Laboratoire de Structure et Fonction des Biomembranes (Equipe de Recherche 143 du Centre National de la Recherche Scientifique), Unit6 d’Enseignement et de Recherche de Luminy, 70 Route Leon-Lachamp, F-13288 Marseille-Cedex-2, France

B. A. Haddock, Department of Biochemistry, University of Dundee, Medical Sciences Institute, Dundee, Great Britain, DD1 4HN

* To whom correspondence should be sent.

Documents

Precursor Forms of the Subunits of Nitrate Reductase in chlA and chlB Mutants of Escherichia coli K12