298

Biorhimicaet BiophysicaActa, 1076(1991)298-304 ©1991ElsevierScience['ablishersB.V.(BiomedicalDivision)0167-4838/91/$03.50 ADONIS O16748389IO00881

BBAPRO 33829

Glucose dehydrogenase from Bacillus subtilis expressed in Escherichia coli I: purification, characterization and comparison with glucose dehydrogenase from Bacillus mega~erium Wolfgang Hilt l, G e r h a r d Pfleiderer 1 a n d Peter Fortnagel 2 t Institutf ~ OrganischeChemie, Biochemieund lsotopenforschungder Universit~tStuttgart. Stuttgart and 2 lgffilut j ~ allgemeineBolanik. AbteilangMikrobiologie Unirersitfft Hamburg;Hamburg(F i~G.)

(Received9 May1990)

Keywords: Glucosedehydrogenase;pH induceddissociation:( R subtilis) Esckerlcbla calf containing the Bacillus subtilis glucose dehydrogenase gene on a plasmid (prLT) was used to produce

the enzyme in high quantifies. GInc-DH-S was purified from the cell extract by (NH4)zSO4-precipitation, ion-exchange chmnatography and Triazine-dye chromatography to a specific activity of 375 U/rag. The enzyme was apparently homogenous on SDS-PAGE with a subunit molecular mass of 31.5 kDa. Investigation of Ghic-DH-S was perfmmed for comparison with the cotTespmaling Woperties of Gluc-DH-M. The limiting Michaelis constant at pH 8.0 for NAD + is K . = 0.11 mM and for D-glucose K b m 8.7 mM. The dissociation constant for NAD + is K~ -- 17.1 mM. Similar to Gluc-DH-NL Giuc-DH-S is inactivated by dissociation under weak alkaline conditions at pH 9.0. Complete reactivation is attained by readjustment to pH 6.$. Ultraviolet absorption, fluorescence and CD-slax'~ of native Gluc-DH-S, as well as fluorescence- and C D . b a c k b o n e - ~ of the dissociated enzyme were nearly identical to the corresponding spectra of Gluc-DH-M. The aromatic CD-spectmn of dissociated Gluc.DH-S was different, representing a residual elliptieity of tryptophyl moieties in the 290-310 nm region. Density gradient centrifugation proved that this behavionr is due to the formation el inactive dimers in equilibrium with monomers after dissociation. In comparison to Gluc-DH-M, the kinetics of inactivation as well as the time.dependent change .d fluorescence intensity at pH 9.0 of Gluc.DH.S showed a higher velocity and a changed course o! the dissociation process.

Introduction In the last few years we have been investigating the properties of glucose dehydrogenase from B. megaterium (EC 1.1.1.74). This enzyme is an excellent object to sway the dissociation, reassociation, unfolding and refolding processes of an oligomeric protein. Detailed kinetic data from ultraviolet, CD and fluorescence spectroscopy gave us important insights into the structurefunction relationship [1-4]. In addition HiSnes et al. [5] compared the primary structure of pig heart lactic dehydrogenase with glucose dehydrogenase from B. megaterium. They concluded that similar secondary structure

Abbreviations:Gluc-DH-M.glucosedehydrodenasefrom B. raegaterium M1286;Gluc-DH-S,glucosedehydrogenasefrom B. subtilus clonedin E. coll. Correspondence:G. Pfleiderer,lnstitut far OrganischeChemie,Biochemic und lsotopenforschungder UniversitiltStuttgart, P[affenwaldring55, D-7000,Stuttgart80, F.ILG.

features exist between these evolutionary, quite different, enzymes. A possible method to study structure-function relationships, is the comparison of enzymes with identical functions from closely related species if the primary structures are known. We use glucose dehydrogenase from B. megaterium and R sublilis which show 80% identical primary structure [6-8], Glucose dehydrogenase from B. subtilis was first isolated by Ramaley [9]. The gene for this enzyme was cloned and expressed in E. coil [1031], which itself does not contain an NAD(P) dependent glucose dehydrogenase. Using the SPAC-I promoter system of Yansura and Henner [12], Fortnagel and colleagues [13] were able to express the B. subtilis glucose dehydrogenase in E. coli with high efficiency [12,13]. Using this biological enrichment system we describe here a simple purification procedure for the enzyme, certain physical data and the kinetics of its inactivation and reactivation. The results obtained with the B. subtilis glucose dehydrogenase synthesized in E. coil are compared with corre-

299 spondiug data of the B. megaterium glucose dehydrogenase. Materials Bovine serum albumin, ovalbumin, chymotrypsinogen, myoglobin and cytochrome c were commercial products from SERVA Mannheim. NAD + was obtained from Boehringer-Mannheim and DEAE-Sepharose CI 6B from Deutsche Pharmacia, Frankfurt. Blue Fractogel was prepared by coupling Cibacron blue F3G-A to Fractogei TSK from Merck, Darmstadt as described by BiJhme [14]. All other substances were purchased from usual commercial sources in the highest grade available, Partially purified glucose dehydrogenase from B. megaterium M 1286 was kindly suppfied from E. Merck, Darmstadt and further purified by stepwise precipitation with (NH4)2SO4 (specific activty 450 U/mg) [21.

KH2PO4 buffer (pH 6.5). This procedure was used to avoid loss of erc~yme activity by'dialysis against buffers with low NaCI concentrations. Any precipitate formed was re:roved by centdfugation and the sample applied to a column of DEAE-Sepharose Ci 6B (2 x 24 cm), which had been equilibrated with 33 mM KH2PO4 buffer (pH 6.5) containing 0.1 M NaCI. After washing, the column was developed with a 280 ml linear gradient of 0.1 M NaCI to 0.5 M NaCl in the same buffer. Enzyme activity was eluted at approx. 0.25 M NaCI. Triazine-dye chromatography on Blue Fractogel TSK After dialysis against 50 mM pyrophosphate buffer (pH 5.2) and removal of precipitate by centrifugation, the enzyme fraction was loaded on a Blue Fractogel TSK column (3 x 15 cm). The column was washed with the same buffer and 300 ml pH-gradient of 50 mM pyrophosphate buffer (pH 5.2) to 67 mM KH2PO4 buffer (pH 6.5) containing 0.5 M NaCI was applied. Glucose dehydrogenase was eluted at pH 5.4.

Methods Bacterial strains and culture conditions Escherichia coli which conlained the Bacillus subtiiis glucose dehydrogenase gene [10,11], as the only component of the glucose dehydrogenase operon on a plasmid (pRLT) [13] under the control of the SPAC-I promoter [12] was used to produce Gluc-DH-S in high quantities. Cells were grown in a 10 1 fermenter in LB-broth at 37°C, During early exponential 8rowth !PTG (isopropyl-fl-D-thiogalactoside) (1 raM) was added to induce glucose dehydrogenase synthesis. Cells were harvested 4 h later when exponential growth ceased. Extraction procedures and purification Cells (230 g) were washed with 100 mM Tris-HCl buffer (pH 7.4) containing 500 m_MNaCI at 0°C. They were disrupted at 0°C in a french pressure cell at 10 000 PSI. DNA was destroyed by addition of 10/tg/ml DNase It. After 10 rain incubation at room temperature the debris was removed from cell extract by centrifugation. The clear supemataflt was fractionated with solid (NH4)2SO4 at 0°C. The precipitate at the fraction 30-60~ saturation, containing 20 000 U Gluc-DH-$ was used for the further purification of the enzyme. The precipitate was dissolved by dialysis against 67 mM KH2PO4 buffer (pH 6.5) containing 0.5 M NaCl. The enzyme was stable under these conditions and could be stored at 0°C for at least a few months without any loss of enzyme activity. Ion.exchange chromatography on DEAE sepharose CI 6B A third of the protein solution was dialysed against 33 mM KH2PO4 buffer (pH 6.5) containing 0.3 M NaC! and subsequently diluted with a 2-fold vol. of 33 mM

Enzyme assay As previously described for Gluc-DH-M [1], enzyme activity was determined at 25°C in 35 mM Tris-HCI buffer (pH 8.0/3 mM NAD+/140 mM 81ucose) by measuring the rate of NADH formation at 366 rim. Protein concentration Protein concentration was determined by the method of Bradford [15] or for pudfied samples by absorption measurements at 280 nm using the absorption coefficient reported in this paper. SDS-PA GE Electrophoresis in 0.1% sodium dodecyl sulfate was performed as described by Laenunli [16] on separation gels containing 15% acrylamide. Protein was fixed and stained with a 0.25% solution of Commassie brilliant blue R-250 in methanol/acetic acid/water (5:1:5; v / v / v ) . Destaining was achieved by soaking in 10~ acetic acid. Inactivation Inactivation of Gluc-DH-S was measured at 0°C after 20-fold dilution of enzyme samples containing 1 to 4 mg/ml protein in 67 mM phosphate buffer (pH 6.5) containing 0.5 M NaCl into 50 mM glycine-Tris buffer (pH 9.0)(adjusted at 25°C) Reactivation After incubation of GIuc-DH-S for I0 rain at pH 9.0 and 25°C, reactivation was attained within 70 rain by 20-fold dilution in 67 mM phosphate buffer (pH 6.5) containing 0.5 M NaCI. The final protein concentrations reached were 0~05-0,2 mg/ml.

300

Ultraviolet absorpiion UV-specta were measure~l at 25°C with a Jasco Uvidce 610 spcctro-photometer in 1 cm pathlength cuvettes. Fluorescence Measurements of fluorescence were performed with a Spex fluorolog with DM 1B Spectroscopy laboratory coordinator in the S1/R1 mode using a 1 × 1 cm cuvette and a 2.25 nm excitation and 4.5 nm emissior~ bandwith. Fluorescence spectra (excitation at 290 nm) were recorded at 25°C at a protein concentration of 30 /~g/ml. Time dependence of fluorecence intensity a-: 340 nm after dilution of a Ghc-DH-S sample in 50 mM glycine-Tris buffer pH 9.0 (excitation 290 rim) was measured at 0°C at a protein concentration of 60

~g/mt. Circular dichroism A Jasco spectropolar;meter J 500 A equipped with a data processor DP 500 N was employed for circular dichroism measurements at 25°C. The pathlength of the cuvettes was 10 era for measurements in the aromatic region (250 to 330 nm) at a protein concentration of 30 /tg/ml. CD backbone spectra (260 to 200 nm) were recorded at a protein concentration of 1 mg/ml using a 0.01 cm cuvette. Density gradient centrifugation 5-15~ linear sucrose gradients (4 ml per tube) were prepared in 50 mM glycine-Tris buffer (pH 9.0). After application of 100 pl samples containing 100 ~g protein per tube, gradients were centrifugated using an SW 60 Ti rotor in a Beckmann L 5-50 ultracentrifuge at 45 000 rpm and 5 °C for 21 h. The gradients were fractionated (3 drops per fraction) by removing the contents of the tubes from the bottom by a capillary tube. Allocation of the protein bands was determined by measuring the concentration by the method of Bradford. The enzymatic activity was additionaly determined for samples containing Ghc-DH-M and Ghc-DH-S after reactivation by 10-fold dilution of the fractions in 67 mM phosphate buffer (pH 6.5)/0.5 M NaCI.

Results

Purification The purification procedure employed (NH4)2SO4precipitation, ion-exchange chromatography on DEAESepharose CI 6B and Triazin dye chromatography on Blue Fractogel TSK, resulting in a specific activity of 375 U/mg. The results of a typical purification are summarized in Table I. Purity and molecular weight The purified enzyme was apparently homogenous on SDS-PAGE with a subunit molecular mass of 31 500 Da. This conesponds to the value of 28 196 Da calculated from the amino acid sequence [6]. FPLC gel chromatography of GIuc-DH-S at pH 6.5 on Superose 12 preparative grade showed one homogenous peak with the same position of ehtion as Ghc-DH-M, corresponding to a molecular mass of 120 000 Da and revealing that the native enzyme is tetrameric. pH optimum The pH-activity profile of Ghc-DH-S (data not shown) is very similar to the corresponding curve of Ghc-DH-M [1]. Both enzymes show an optimum at pH 8.5. Inactivation of Ghc-DH-S in the test medium was observed at values above pH 9.0. Kinetic constants Initial rate measurements were made in Tris-HCl buffer (pH 8.0) at 25°C. The results are shown in Fig. 1. The forward reaction in the absence of products is described by the initial velocity equation of Cleland

lrrl: v=

vo-IAI-[BI K~- K~ + K~. [A] + ZC..[B] + [A]- [B]

No substrate inhibition was observed up to 200 mM D-glucose and 4 mM NAD +. The kinetic coefficients determined from the slopes (Fig. IB) and intercepts (Fig. 1C) of the secondary plots are the limiting Michaelis constant for NAD+: K, -- 0.11 mM, for gin-

TABLE! Purificationof B. subtilisglucosedehydrogenaseclonedin E. coil Fraction

Cellextract (NH4)2SO4 precipitate DEAESepharo~CI 6B BlueFractoge|TSK

Total activity (U) 7000 6750 5684 4140

Total protein (rag) 450 187 26 11

Specific activity (U/rag) 15 36 217 375

Yield (%)

Purification factor

100 96 81 59

1.0 2.4 14.4 25.0

301

lfiO.[OSf]

A

2-

1ZE6 ~.

._.

280

300

320

D

3~0

360

3B~

~0

~20

WAv'ELEN~TH {nml

Fig. 2. Fluorescenceemission spectra of native and dissociatedGIucDH-S. (a) native enzyme in 67 raM phosphate buffer (pH 6.5) containing 0.5 M NaCI, (b) dissociatedenzymein .50 mM glycin-Tris buffer (pH 9.0) (protein concentration: 30/zg/ral 25°C, cuveues 1 x I cm, excitationat 290 ran).

5

01

10

YINAOI[llmk'l]

cose: K b = 8.7 mM and the dissociation constant for NAD+: Kia = 17.1 raM.

Ultraviolet absorption Gluc-DH-S in 67 rnM phosphate buffer (pH 6.5) containing 0.5 M NaCI showed a typical ultraviolet absorption spectrum, with a maximum at 278 nm and a minimum at 252 nm (ration 2.4/1). The spectrum was nearly identical to that of Gluc-DH-M [1]. The molar absorption coefficient at 280 nm is 131 800 1 / M . cm assuming a molecular mass of I ". 700 Da. This value corresponds to an ab~rpti~m coefficient of .~(l cm-I m g / m l ) = 1.17. The protein concentration was determined by the method of Bradford [15] using GlucDH-M as standard protein assay.

o

Fluorescence

!

0

0.5

0,1

[l/mHl 1/l~UCOSf/

C

-,..,

"6 O,5

0,I

./° I

0

0.7

...

I

0,5

[t/mM] 1/ [6Z~.OSEI

The fluorescence emission spectra of Gluc-DH-$ for the native letrameric and the inactivated enzyme at pH 9.0 are shown in Fig. 2. The curves are nearly identical with the corresponding spectra of Gluc-DH-M. As observed for Gluc-DH-M, Gluc-DH-$ produces a red shift of the emission maximum by 13 nm and a 2.2-fold increose of fluorescence intensity after dissociation at pH 9.0. The 7 nm shift to red of both Gluc-DH-S spectra in comparison with the spectra of Gluc-DH-M puplished by Maurer [1] are due to the device. The fluorescence spectra of native and dissociated Glue-

Fig, !. PrimaryploLVariationof th~ reciprocalinitial rate in Tris-HCI buffer (pH 8.0) at 25°C with the reciprocalof the NAD+ concentration, for constant a-glucoseconcentrations.(B) Secondaryplot, Variao don of the slopes of the primary plot with the raciprocal of the o-glucose conccnvration.(C) Secondaryplot. Variation of the intercepts of the primary plot with the r~iprocal of the D-glucoSecon. ccntration.

302

-7O

,'.

,,

i

Q.

//

2

;

/,;

8

6

'

~

~?

g

" , ,.

16

~0 0

18

FRACTIONIVUMBER

*6

260

250

260

230

220

270

200

WAVELENGTH (nm) Fig. 3. Far altraviotet circular dichroisra spectra of native Gluc-DH-S

Fig. 5. Density gradient centrifugafion of dissociated Gluc-DH-S and Gluc-DH-M. Sucrose density gradients 5-15% in 50 mM glycine,Tris buffer {pH 9.0) 21 h centrifngated at 45000 rpm, 5°C. O ~ O GIuc-DH-S; o o, Gluc-DH-M; x - - - x, Gluc*DH-s; × ~ x. Gluc-DH-M enzyme activity after reactivation of the fractions in 67 mM phosphate buffer pH 6..5/0.5 M NaCI. Reference proteins: bovine serum albumin (BSA) and myglobin (MYO).

in 67 mM phosphate buffer (pH 6.5)/0,5 M NaCI (a) and the dissociated enzyme in 50 raM glyeine-Tris buffer (pH 9.0) (30 rain incubation) (b). (protein concentration: I rag/ral pathlength of the cuvenes 0~0) cm, 2.5° C).

+8 "~

.,,j

1.6

k

,ooi

1,

i

1.2

\

\

:1 ~"

250

260 270

280 290 300 310

320

I

\,

0

1

2

3

330

WgVELEN6"/'H (r~n/ Fig. 4. Circular dichroism speczra (aromatic region) of native and dissociated Gluc-DH-S. (a) native enzyme in 67 ram phosphate buffer (pH 6.5)/0.5 M NaCI, (b) dis,~viated enzyme in 50 mM giycine.Tris buffer (pH 9.0) (protein concentration: 30/tg/ral, pathlength of the cuvctte~: l0 cm, 25 o C).

~

5

6

7

7i T/NE [rain) Fig. 6. Inactivation of Gluc-DH-S in glycine-Tris buffer pH 9.0 at 0°C. (protein concentration: 50 t~g/ral). Insert: First order plo! of the inactivation of Gluc-DH-S in glycine-Tris buffer pH 9.0 at 0 ° C (protein concentration 50 #g/ml).

303 DH-M represent a corresponding 7 nm shift by measuring on SPEX-Fluorolog (data not shown).

7oo'

t.

~

\

~,

1.6

~" "'.

c~

\

~

' /

%

\

~

~21-

\

~ o,eI-

"" Circular dichroism (backbone region) The circular dichroism spectra of the peptide backbone of Gluc-DH-S at pH 6.5 (native) and pH 9.0 (dissociated) are presented in Fig. 3a and b. The curves correspond to the backbone spectra of Gluc-DH-M [1]. Circular dichroism (aromatic region) The CD-spectrum in the 250-320 nm region of native Gluc-DH-S, shown in Fig. 4a, is identical with the corresponding spectrum of Gluc-DH-M [11. After inactivation at pH 9.0 the CD-spectrum of Gluc-DH-S is diminished for the whole region (Fig. 4t)). But in contrast to Gluc-DH-M, which demonstrates a total lack of CD-signals in the region from 290-310 nm [l], GlucDH-S produces a residual of 30% positive circular dichroism related to the native spectrum in this region after inactivation at pH 9.0. Even after 24 h incubation in pH 9.0 buffer or inactivation under more alkaline conditions in diethanolamine-glycine buffer (pH 9.4) a residual of CD-signals in the 290-310 nm region was observed. Density gradient centrifugation Molecular weight of Gluc-DH-S and Gluc-DH-M after dissociation at pH 9.0 was determined by sucrose density gradient centrifugation. Ghc-DH-M demonstrated a typical peak of monomers with a molecular mass of 30 kDa (Fig. 5). In contrast Gluc-DH-S produced a broader peak under identical conditions, with a maximum correspondi~-g to 48 kDa, suggesting incomplete dissociation and formation of a direct-monomer equilibrium of this enzyme at pH 9.0. Inactivation The inactivation of Gluc-DH-S (Fig. 6) by shifting the pH from 6.5 to 9.0 proceeded at a faster rate and with a different kinetic behaviour compared to GlucDH-M [2]. Whereas Gluc-DH-M showed two distinct kinetic phases (K1 = 2.88.10 -3 s - t / K 2 = 1.09.10 -3 s-t at 0 ° C), Ghc-DH-S is inactivated in a monophasie first-order reaction (Fig. 7+ insert), 3.3-fold faster than the rapid phase of Gluc-DH-M inactivation, with a rate constant of K = 9 . 5 . 1 0 -3 s -t at 0°C. Reactivation As observed for Gluc-DH-M [1], the inactivation of Gluc-DH-S proved to be completely reversible, by shifting the pH from 9.0 to 6.5 (data not shown). Gluc-DH-S was reactivated to 100% activity for protein concentrations higher than 0.1 mg/ml in the reassociation samples. For lower protein concentrations the yield of recovered activity was decreased.

g 2,OL

i! ~.

~ 70

,.e~.'..

0 1

-.4~..

",.

"--..,

\

2 3 # 5 6

~

".

7 8 917 ff f2 TIHE(rain)

re~E (mm.I Fig. 7. Time dependenceof the change of fluorescence intensity at 340 nra after incubation of Gtuc-DH-S in glycine-Trisbuffer pH 9.0 at

0°C. Insert: Firstorder plot of the changeof fluorescenceintensityat 340 nm. (protein concentration: 60 ~ug/ml.cuvettes ! xl ¢m. excitation: 290 nm).

Kinetic of fluorescence change The time dependence of changes in fluorescence intensity at 340 nm, which is due to the exposure of typtophan residues to the solvent [1,18-20] after incubation in pH 9.0 buffer at 0 ° C was measured. At the beginning of dissociation Gluc-DH-S showed a lag phase, but no jump in fluorescence intensity as observed for Gluc-DH-M [2] (Fig. 7). The first-order plot (insert Fig. 7) shows two distinct kinetic phases. The fast phase, with a rate constant of K] = 8.3.10 -3 s -t corresponds well to the rate of inactivation, whereas the slower process with a rate constant of K 2 = 3.2.10 -3 s- mis not related to the inactivation of this enzyme.

Discussion in comparison with Gluc-DH-M, Gluc-DH-S represents similar biochemical properties, as well as nearly identical UV-, CD- and fluorescence spectra. The conservation of amino acid residues, which in our opinion are essential for the activity of Gluc-DH-M, like Gly-20 [5], His-148, Lys-201, Tyr-254 [7] and Trp154 accounts for the similarity of biochemical properties. A further aspect may be the very high homology (96%) over the 107 earhoxyterminal amino acid residues, where the substrate binding site is expected. The loss of Trp-150 in Gluc-DH-S reveals that this amino acid residue is not important for enzymatic activity. Because of the nearly identical fluorescence and aromatic CD-spectra, we suggest that the contribution of Trp-150 to fluorescence and CD-signals must be, if at all, very slight. In contrast to the similarity of biochemical and spectroscopic properties Gluc-DH-S showed a quite different behaviour of the dissociation process at pH 9.0.

304 It could be demonstrated by density gradient centrifugation that the dissociation of Gluc-DH-S is incomplete under identical conditions. Whereas Gluc-DH-M shows a dissociation to monomers, without emergence of dimers in the dissociation pathway, GlucoDH-S forms dimers in equilibrium with monomers. We suggest, that the only difference in spectroscopic behaviour, a residual of 30~ aromatic CD signal of dissociated Gluc-DH-S is due to the presence of the partial dimeric form. The formation of dimers in the case of Gluc-DH-S is confmned by successful hybridization experiments, from which we expect more detailed insight into the subunit interaction (work in progress). As a further aspect of the different dissociation behaviour, the alkaline inactivation of GlucoDH-M shows two distinct kinetic phases, whereas Gluc-DH-S is inactivated in a monophasic f'n'st-order reaction, proceeding 3.3-fold faster than the rapid phase of Gluc-DH-M inactivation. Recording the change of fluorescence intensity at pH 9.0, due to the exposure of tryptophyl moieties to the solvent, Gluc-DH-M showes a jump of tryptophan signals, representing a rapid rearrangement within the active tetrameric enzyme. No jump phase and with it no corresponding rapid rearrangement could be observed in the case of Gluc-DH-S. This very different behaviour of the dissociation process must be caused by a variation of the intersubunit contacts and may be due to the loss of homology in the aminoterminal site of both enzymes. Comparison of the properties with a third glucose-dehydrogenase with known amino acid sequence from B. megaterium cloned in E. coli [21] as well as X-ray investigation of the structure of Gluc-DH-M [22] may open further discrimation of structure-function relationships of this enzyme.

A

~

t

This work was gratefully supported by research grants from the Deutsche Forschungsgemeinschaft. References I Maurer, F.- and Pfl~derer, G, (1985) Biochim. Biophys.Acta 827, 381-388.

2 Maurcr. F.. and Pfleider~. (3. 0987) Z, Naturforschung42, 907915. 3 Paaly, H.E. and Pfleide~r, G. (1975) Hoppe SeylersZ. Physiol. Chem 356. 1613-1629. 4 Pauly, I-LE.and Pfleiderer,G. (1977) Biochemistry16, 4599-4604. $ HSnes, J. Jany, K.-D. Pfieiderer, G. and Wagner A.F.V. (1987) FEBS Lett 212, 193-198. 6 LampeL KA~ Uratani. B~ Char,dry. G.R.. Ramaley. R.F. and Rudikoff. $. (1986)J. Bacteriology166. 238-243. 7 Porlnagel, P. Lampel,ICA. Neitzke, K.-D. and Freese, E. (1986) J. theor. Biol. 120, 489-497. 8 Jany, K.-D. Ulmer, W. Fri~chle, M. and Pfleiderer, G., (1984) FEBS Lett 165, 6-10. 9 Ramaley, R.F. and Vasantha, N. (1983) J. Biol. Chem. 258, 12558-12565. 10 Vasantha, N. Uratanl, 13. Rama[ey. R_E and Freese. E. (1983) Prec. Natl. Acad. Sci. USA 80, 785-789. 11 Uratani, B. Lampel, K.A., Lipsky. R.H. and Freese, E. (1984) in Molecular Biology of Microbial Diffczenfiation(Hoch, J.A. and Setlow, 13. eds. pp. 71-76. Am. Soc. Microbial, Washington D.C 12 Yansura, D.C. and Henner, DJ, (1984) Proc, Natl. Acad. Sci, USA 81, 439-443. 13 Lehmann, R. Ph.D. Thesis Hamburg(1988). 14 BOhme., HJ., K o p p ~ e r , (3. Schulz, J. and Hofmann, E (19"/2)J. Chromat. 69, 209-214. 15 Bradford, M.M. (1976) Analytical Biochemistry72. 248-254. 16 LaemmlLU.K. (1970) Nature 227. 680-685. 17 Cleland. W.W. (1963) Biechinc Biophys.Acta 67, 104-t37. 18 Donovan, J.W. Laskowski, MJr. and Schemga, H.A. (1961) J. Am. Chem. Soc. 83, 2686-2694. 19 Brand. L. and Withold, B. (196"/)Methods EnzymoL1L 776-856. 20 [rac~ (3. Balestrieri,C. Parlato, G. Servillo, L. and Colonna, G. (1981) Biochemistry20, 792-799. 21 Heilmann, HJ. M~igerC HJ. and Gassen, H.G. (1988) Eur. J. Biochem. 1"/4,485-490. 22 Pal, G.P. Jany, K.-D. and Saenger, W. (1987) Eur. J. Biochem. 167, 123-124.