Limited Proteolysis and Active-site Studies of the First Multienzyme

T H E JULKN.U. O S BNMKI~AI. CHEMISTKI 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 269, No. 11. Issue of March 18, pp. 8524-8528. 1994 Printed in U.S.A.

Limited Proteolysis and Active-site Studies of the First Multienzyme Component of the Erythromycin-producing Polyketide Synthase*

(Received for publication, October 14, 1993, and in revised form, December 29, 1993)

Jesus F. AparicioSP, Patrick CaffreyS, Andrew F. A. Marsden$, James Stauntonn, and Peter F. Leadlaytll From the Cambridge Centre for Molecular Recognition, and Departments of $Biochemistry and IOrganic Chemistry, University of Cambridge, Cambridge CB2 lQW, United Kingdom

The domain structure of the 6-deoxyerythronolide B synthase 1 component of the erythromycin-producing polyketide synthase from Saccharopolyspora erythraea has been investigated using limited proteolysis and active-site labeling. Trypsin, elastase, endoproteinase Glu-C, and endoproteinase Arg-C were used to cleave the multienzyme, and the sizes of the resulting fragments were assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The location of fragments within the primary structure was established by N-terminal sequence analysis. The cleavage pattern followed domain boundaries previously predicted on the basis of sequence alignments, but many predicted interdomain regions were not cleaved, even under the harshest conditions used. Initial proteolysis generated three large fragments: an N-terminal fragment (about 60 kDa) hous- ing an acyltransferase-acyl carrier protein di-domain; a central fragment (about 90 kDa) containing a ketosynthase-acyltransferase di-domain; and a C-terminal fragment (about 220 kDa) containing the remaining six domains of the multienzyme, including the third acyltransferase. The intact multienzyme behaves as a dimer of molecular mass 660 kDa on gel filtration; and the C-terminal fragment remains dimeric. However, the N-terminal and central fragments appear to be monomeric species. After proteolysis of the multienzyme, the N-terminal di-domain was found to be specifically labeled after incubation with [14C]propionyl-CoA, providing the first evidence for its proposed role as a “loading domain” for the propionate starter unit. In contrast, the other two fragments were specifically acylated by [14C]methylmalonyl-CoA, indicating that both the other two acyltransferases remain enzymatically active after proteolysis.

Erythromycin A is a clinically important macrolide antibiotic produced by Saccharopolyspora erythraea (1, 2). Its biosynthesis involves the initial construction of a polyketide chain from one molecule of propionyl-CoA and 6 molecules of methylmalonyl-CoA to form the 14-membered macrolide ring of the first isolable intermediate, 6-deoxyerythronolide B by a mechanism that resembles the biosynthesis of fatty acids (1,3). Sequencing of the structural genes encoding the chain-building polyketide

Engineering Research Council (United Kingdom). The costs of publica- * This work was supported by project grants from the Science and

tion of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

S: Recipient of a bursary from the European Community.

try, University of Cambridge, Tennis Court Rd., Cambridge CB2 lQW, I/ To whom correspondence should be addressed: Dept. of Biochemis-

UK. Tel.: 44-223-333656; Fax: 44-223-333345.

synthase, 6-deoxyerythronolide B synthase (DEBS),l has revealed that it is composed of three type I (multifunctional) polypeptides (DEBS 1, DEBS 2, and DEBS 3) (4-7) each of which apparently catalyzes two of the six cycles of chain extension required to produce 6-deoxyerythronolide B. Thus, DEBS 1 would catalyze the binding of the propionyl-CoA starter unit and the first two cycles of chain extension. The domain structure for each DEBS multienzyme has been predicted (6, 7) on the basis of their striking similarity with the known organiza- tion of vertebrate fatty acid synthases (8, 9) and of the occur- rence, a t predicted domain boundaries, of unusual alanine-, proline-, and charged residue-rich sequences (7) that have been implicated as inter-domain linker regions in other multi-domain enzymes (10, 11). However, direct evidence for the topology and domain structure of the DEBS multienzymes is required in order to understand how the growing polyketide chain moves between the 30 or so active sites. Limited proteolysis (12, 13) has been used here, in combination with specific active-site radiolabeling, to study the first multienzyme of the erythromycin-producing polyketide synthase (DEBS 1) and to provide the first structural information on any bacterial antibiotic-producing polyketide synthase.

EXPERIMENTAL PROCEDURES

Materials

~-1-To~ylarnido-2-phenylethyl chloromethyl ketone-treated bovine trypsin, l-chloro-3-tosylamido-7-amino-2-heptanone-treated chymotrypsin, thioglycollic acid (mercaptoacetic acid), and 3-cyclohexylamino- 1-propanesulfonic acid buffer were purchased from Sigma; high range M , standards for SDS-PAGE and phenyl-Superase, Superdex 75, Su- perose 6, and Superose 12 columns were from Pharmacia LKJ3 Biotech- nology Inc.; Elastase, endoproteinase Arg-C, and endoproteinase Glu-C were from Boehringer Mannheim; ProBlott polyvinylidene difluoride membranes and sequencing reagents were from Applied Biosystems Inc., Foster City, CA. All other reagents were from commercial sources and of analytical grade. DEBS 1 protein was purified from S. erythraea as previously described (14). [l-’4C]Methylmalonyl-CoA was prepared from [l-’4C]sodium propionate (DuPont-NEN, 51 Ci mol-’) via a [1-14C]propionyl-CoA intermediate, using acetyl-coA synthetase and transcarboxylase.‘

Limited Proteolysis of DEBS 1 mpsin-Purified DEBS 1 was incubated with L-1-tosylamido-2-phe-

nylethyl chloromethyl ketone-treated trypsin at an enzymehbstrate ratio of 1/350 (w/w) in 1 mM EDTA, 2 mM dithiothreitol, 50 mM Tris-HC1 buffer, pH 7.5, 20% glycerol (buffer A). Reactions were carried out at 30 “C for various times and terminated by heating at 100 “C for 3 min in electrophoresis sample buffer unless otherwise indicated. The reac-

thase 1; ACP, acyl carrier protein; AT, acyltransferase; KR, p-ketore- The abbreviations used are: DEBS, 1,6-deoxyerythronolide B syn-

ductase; KS, 0-ketoacyl ACP synthase; PAGE, polyacrylamide gel electrophoresis.

J. F. ADaricio, P. Caffrey, A. F. A. Marsden, J. Staunton, and P. F. Leadlay, manuscript in preparation.

8524

Proteolysis of Erythromycin-producing Polyketide Synthase 8525

tion products were separated by SDS-PAGE and either examined after staining with Coomassie Blue R-250 or transferred to ProBlott membranes for N-terminal sequencing.

Elastase-DEBS 1 was treated at an elastase/substrate ratio of 1/200 (w/w) in buffer A. Reactions were performed, terminated and the products analyzed as described above.

Endoproteinase Glu-C-An enzyme/substrate ratio of 1/40 in bufferA was used. Reactions were performed a t 37 "C, and the mixtures treated as described above.

Endoproteinase Arg-C-We used a proteinase/substrate ratio of 1/30 (w/w) in buffer A. Reactions were performed a t 37 "C and treated as described above. In some cases the conditions of proteolysis were modi- fied as indicated.

SDS-PAGE and Determination of N-terminal Sequence

SDS-PAGE was performed in 5, 7, or 10% polyacrylamide gels (15). After electrophoresis, gels were stained either with Coomassie Brilliant Blue R-250 or with silver (16). Alternatively gels were transferred to ProBlott polyvinylidene difluoride membranes according to Matsudaira (17) and subjected to Edman degradation on an Applied Biosystems 477A pulsed-liquid protein sequencer. Molecular weights of protein fragments were evaluated by using calibration curves from measurements with commercially available standard proteins. Protein concentrations were determined by the methods of Bradford (18) and Lowry et al. (19).

Acylation of DEBS 1

Labeling was carried out at 0 "C, after limited proteolysis, using (2RS) [l-'4Clmethylmalonyl-CoA or Il-14Clpropionyl-CoA as described elsewhere (20).

Fragment Separation under Non-denaturing Conditions-The products of proteolysis of DEBS 1 were separated from each other by micro- fast protein liquid chromatography using a Smart system (Pharmacia). Proteolysis of DEBS 1 was performed as described above, stopped by rapid freezing on dry icdacetone, and the reaction mixture was stored a t -80 "C.

Hydrophobic Interaction Chromatography-The proteolysis mixture was applied to a phenyl-Superose (Pharmacia) column a t 1 M (NH,), SO, in buffer A. The column was then washed with 250 pl of the same buffer and the products eluted with a linear gradient of decreasing salt a t a flow rate of 50 pl/min at 4 "C.

Size Exclusion Chromatography-Proteolysis products were applied onto either a Superdex 75, Superose 12, or Superose 6 (Pharmacia) column previously equilibrated with buffer A containing 150 mM NaCl, and chromatography was carried out at a flow rate of 40 pYmin a t 4 "C. Superose 12 and a Superose 6 column were calibrated for molecular mass determination using the following proteins: thyroglobulin, ferri- tin, catalase, aldolase, and bovine serum albumin. The Superdex 75 column was calibrated for molecular mass determination using as standards bovine serum albumin, ovalbumin, carbonic anhydrase, myoglo- bin, and cytochrome c.

RESULTS

Limited Proteolysis of DERSl DEBS 1 was digested with either trypsin, elastase, endopro-

teinase Glu-C, or endoproteinase Arg-C at various molar ratios at either 30 or 37 "C as described under "Experimental Proce- dures'' for various lengths of time, and the resulting fragmentation patterns are shown in Fig. 1. Fragment patterns were unaltered when SDS-PAGE was carried out in the presence of dithiothreitol or 2-mercaptoethanol.

IFypsin-The fragmentation pattern generated by tryptic cleavage is shown in Figs. 1B and 2. The initial cleavage introduced by trypsin resulted in the release of a 57-kDa fragment (TO), and a 219-kDa fragment (T5) as the first stable products (see Fig. 1B). The size on SDS-PAGE and the N-terminal sequence of TO showed that it comprises the di-domain AT1-ACP1 thought to be responsible for loading the propionyl starter unit. Similarly, the N-terminal sequence of T5 indicated that this fragment comprises the domains KRl-ACP2-KS2-AT3-KR2- ACP3. As the hydrolysis continued, an N-terminal piece of the loading domain was removed leading to the generation of a 56-kDa loading domain fragment (Tl). Several other fragments appeared concurrent with the disappearance of both the re-

2-ADLSKL To 1443-STEVXEV T5 - I

12-TAQPGRIVRP TI 1443-STEVDEVXAL I

I I X-548-ARXNEAAPGE T I 0

T 9

1925-VGALAXLP T6 I 4

1443-STEVXEV TZ

1443-STEVDEV T 3 -

*- 1500-VRELXVD T 7 I I

1500-VRELEVD T8 I

2-ADLSKLSDSR E l I

2 0 2 6 - T T A P V X E P I A E2 I I

2-AOLSKLS F 5

2-ADLSKL . E 4 1 4 4 0 - A R R S T E 3 E 4 . 2-ADLSKLS 63 I 4 3 I -RVXLEPK , I - G2

I 4 3 " R V X L E P K P . G 4 I

5 4 6 - T R A R T N G5

546-TRARTNE 66 I , 20 13-LGGATGAEO GB I

2-ADLSKLS E I

A 5

' 1443-STEVOE A I

2;ADLSKLSDS A6 - 96 I -RAGVSSFGI

- A 7

1443-STEVD A9

961-RAGVSS ~ 1 2

*""""+

FIG. 1. DEBS 1 domain organizat ion as revealed by l imited proteolysis.A, DEBS 1 map. Each activity is represented by a rectan- gle whose length is proportional to the proposed length of the "domain"

posed domains indicate their distance from to the N terminus of the (6). Linker regions are also shown in proportion. Numbers in the pro-

protein. The ruler indicates the residue number within the primary structure of the protein. B, DEBS 1 fragmentation with trypsin. The representation is proportional. The N-terminal sequence of each fragment is included; C , the same with elastase; D, the same with endoproteinase Glu-C; E, the same with endoproteinase Arg-C. Asterisks denote fragments stable under the harshest conditions used (see "Experimen-

Dashed lines indicate minor species. tal Procedures"). Thickened Eines indicate the first fragments generated.

c""-,

maining DEBS 1 protein and the transiently produced 300-kDa species generated by the release of TO, namely T10 (94 kDa comprising AT2-KS1) and T6 (168 kDa comprising ACP2-KS2-

As the time of incubation was increased further fragments were generated. T2 (51 kDa) and T3 (49 kDa) have the same N terminus as T5 and contain the KR1 domain. T9 (193.5 kDa) possesses the same N terminus, but was generated by removal of the C-terminal ACP domain from T5. Other fragments generated were T7 (161 kDa) and T8 (156 kDa). They have the same N terminus a t Val'500 and contain the fragment KR1- ACP2-KS2-AT3. Apart from the KR1 domain, no single domains were observed, suggesting that they are rapidly de- graded. Some fragments such as T1, T3, and the tetra-domains T7 and T8 remain resistant to further hydrolysis even using 10-fold more trypsin and 2 h of incubation.

As summarized in Fig. 1, the primary cleavage sites are located at the C-terminal end of the loading di-domain and at

AT3-KR2-ACP3).

8526 Proteolysis of Erythromycin-producing Polyketide Synthase

Flc:. 2. SDS-PAGE of DEBS 1 tryptic fragments and labeling studies. A, SDS-PAGE of DEBS 1 fragments after limited proteolysis with trypsin. DEBS 1 was digested with trypsin at an substrate/ enzyme ratio of 350/1 as indicated under “Experimental Procedures.” Digests were stopped at the times indicated below by boiling in electrophoresis sample buffer. The digestion products, denatured and reduced in this way, were separated on 7% linear polyacrylamide gels, and stained with Coomassie Blue R-250. Lane 1 , un- cut DEBS 1; lanes 2-6,2,5, 15,30, and 60 min. of incubation. B, a 15-min digest under the conditions stated above was stopped by sudden freezing with dry ice/ acetone and incubated with either 11-’“Clpropionyl-CoAor 11-’“Clmethylma- lonyl-CoA a t 0 “C for 5 min. Electrophore- sis was performed on a 7 0 polyacrylamide gel in the presence of SDS, the gel was then soaked in Amplify solution tAmer- sham Corp.) for 30 min, dried, and ex- posed to P-Max x-ray film at -70 “C for 1 week. Lane I, digest incubated with I 1-’.’Clmethylmalonyl-CoA; lane 2, digest incubated with Il-’”C Ipropionyl-CoA.

A

Mr 1 2 I O - 3 x

2 0 5 -

116 -

9 7 . 4 -

B 3 4 5 6 1 2

+T IO”)

66 -

4 5 -

the N-terminal end of fragment T5 ( S ~ I - ’ ~ ~ ~ ) ) , dividing the molecule into three independent parts, namely the loading di-domain (AT1-ACPl), the di-domain comprising KS1 and AT2 (fragment TlO), and a hexa-domain (fragment T5) that includes the rest of the protein. The same fragmentation pattern was observed in the absence of dithiothreitol or when the concen- tration of glycerol in the reaction mixture was reduced to lo%, although the reaction was then much faster (data not shown).

Elastase-The fragmentation pattern generated by elastase cleavage of DEBS 1 is shown in Fig. 1C. Interestingly, initial cleavage between the ACP2 domain and the C-terminal KS domain (KS2) at the Ala2025-Thr2w2fi bond yielded two large fragments, E l (212 kDa) and E2 (158 kDa).

Digestion with 10-fold more elastase generated a compli- cated proteolytic pattern (Fig. 1C) as a result of the more com- plete degradation of the N-terminal fragment, with the C-terminal half remaining stable under these conditions.

Endoproteinase Glu-C-The fragmentation pattern resembles the tryptic fragmentation but with three new fragments. So, G8 that includes KS2-AT3-KR2-ACP3, G5 tha t comprises KS1-AT2-KR1-ACP2 (see Fig. 1D) and G11, a 47-kDa fragment that constitutes the AT1 domain.

Endoproteinase Arg-C-The fragmentation pattern with endoproteinase Arg-C was different to that obtained with trypsin (Fig. E). The initial cleavage introduced by this protease takes place at Arg’442-Ser’44‘3, cutting the molecule into two large fragments: A5, a 151-kDa fragment containing AT1-ACP1-KS1- AT2, and Al, a 219 kDa fragment whose N terminus is at

when harsher conditions were used (ratio of proteinase/DEBS 1 (1:20 w/w)) were other fragments obtained (see Fig. E). In

Ser1443 , and contains KRl-ACP2-KS2-AT3-KR2-ACP3. Only

contrast to the results of the tryptic fragmentation, no fragment was found equivalent to T10, but this is probably due to the bigger “loading domain” generated by this proteinase in comparison with the others, which would make the KS1-AT2 fragment smaller and perhaps less structured. Limited proteolysis was also attempted using chymotrypsin, but no stable fragments were observed.

Acylation of DEBS 1 DEBS 1 is rapidly acylated by the substrates methylmalonyl-

CoA and propionyl-CoA (20). To determine which domains were being acylated, various proteolytic digests of DEBS 1 were labeled as indicated under “Experimental Procedures.” When a tryptic digest was treated with [l-14Clpropionyl-CoA, the ra- dioactivity was found specifically associated with the TO or T1 fragments (Fig. 2B 1, indicating that at least one of the compo- nents of the loading domain was acylated. With the endoproteinase Glu-C digest, in which an individual AT1 is obtained (fragment Gll), analysis showed that the [l-’4C]propionyl-CoA labeling was found specifically within AT1 and in none of the fragments lacking this AT domain.

In contrast, when the same digests were treated with [l-’4C]methylmalonyl-CoA, labeling was observed in T5, T6, T7, T8, and T10 (Fig. 2B). This indicates that any fragment containing an AT domain other than AT1 could be labeled. The labeling observed in T10 is significant, because this di-domain contains only KS1 and AT2. This labeling was always weaker than that obtained in other fragments, perhaps indicating that the ACP domains might play a role in the stability of this acyl intermediate.

Labeling of other proteinase digests was fully consistent with

Proteolysis of Erythromycin-producing Polyketide Synthase 8527

the labeling ofATl by [l-'4Clpropionyl-CoA and ofAT2 andAT3 by [ l-'4C]methylmalonyl-CoA. No labeling was observed when [2-'4C]malonyl-CoA was used as a substrate (not shown).

Separation of the Proteolytic Fragments Tryptic digestion of DEBS 1 resulted in the initial generation

of three main fragments, T1, T10, and T5 (see above), which were separated using a phenyl-Superose column (see "Experi- mental Procedures"). When a Superdex 75 gel filtration column was used, the relative elution volume of fragments T1 and T10 indicated that they behave as monomers since their predicted sizes, 56 and 94 kDa, respectively (also determined by SDS- PAGE), were very similar to the observed ones (62 2 5 and 95 2 4 kDa). The fact that T1 and T10 were also separated from each other, and not associated with T5 or T6, indicates that under the conditions used, they behave as independent fragments. T2 and T3 (KR1 fragments) were also monomeric, with a native molecular weight of approximately 53,000.

Similar results were obtained upon gel filtration of a n endoproteinase Glu-C digest of DEBS 1. Both G2 and G3 were also independent domains, and G3 behaved as a monomer with a molecular mass of 65 kDa. Analogously, G11 and G10 (AT1 and KR1 fragments respectively) were also independent and monomeric, with apparent molecular masses of 48 and 54 kDa, respectively.

On a Superose 12 filtration column, fragments TO, T1, T10, T2, T3, G3, G11, and G10 all behaved again as independent monomers. However, fragments T5 (219 kDa) and T6 (168.5 kDa) both showed a native molecular weight of approximately 460,000. Although this is not within the optimal separation range, it indicates that they behave as dimers. Fragments T7 (161.5 kDa) and T8 (156 kDa) both eluted from this column as a globular protein with a molecular mass of 240 kDa would do, while fragments G5 (154 kDa) or G8 (159 kDa) showed a native molecular weight of approximately 280,000. Whether those fragments were independent or not, could not be decided because of the similarity of their sizes.

Gel filtration on Superose 6 allowed the fractionation of the largest fragments. For example, fragments G2 (220.5 kDa) and G4 (194 kDa) both showed an apparent molecular mass of 440 kDa while T6 (168.5 kDa) showed a molecular mass of approximately 400 kDa and G5 and G8 a mass of 320 kDa. Comparison of these values with the predicted sizes clearly indicated that these fragments are either dimeric or highly asymmetric monomers. Importantly, fragments T7 and T8 showed a molecular weight of 360,000 using this column, which could indicate that they are dimers as well. Fragments E l (212 kDa) and E2 (159 kDa) also eluted independently and behaved as dimers with molecular masses of 460 and 300 kDa, respectively. From the position of elution of DEBS 1 from this column it appeared to have a molecular mass of 660 kDa. All determinations were performed at least four times and with different enzyme prepa- rations.

DISCUSSION

The recent purification to near-homogeneity of the three con- stituent multienzymes of the erythromycin-producing polyketide synthase from S. erythraea (14) has provided the first opportunity to explore the structure and assembly of these unusual multifunctional polypeptides. In general, such multifunctional proteins are composed of independently folded, compact domains connected by linker regions (21,221 which are readily cleavable by limited treatment with proteolytic enzymes (23). From X-ray diffraction studies, it has been demonstrated (24) that segments of the polypeptide chain of high flexibility, as determined by mean atomic displacements, are correlated with those regions that are vulnerable to limited proteolysis. Evi-

dently, many interdomain loops are located at the surface of the protein and adopt conformations that facilitate the interaction of proteases with the polypeptide substrate.

Proteolytic studies on DEBS 1 were carried out using four different proteinases, chosen for their relatively broad primary specificity, so that accessibility to potential target sites, rather than the specificity of the protease itself, is the factor likely to determine the sites of cleavage. N-terminal sequencing of the fragments generated allowed them to be placed within the con- text of the known primary structure of DEBS 1 (6), allowing the rapid identification of the regions particularly susceptible or resistant to proteolysis. The initial cleavages observed indicated that the DEBS 1 molecule (370 kDa) could be divided into three large parts. For example, trypsin cleaves the protein ini- tially into three fragments of approximately 57, 94 and 219 kDa, with cutting sites at Ala54H and The smallest (N-terminal) fragment comprises the di-domain AT1-ACP1, which is proposed to function as a loading domain for the propionate starter unit. The central fragment comprises KS1 and AT2; and the large C-terminal fragment contains the rest of the domains originally present in DEBS 1. This overall fragmentation pattern was confirmed by the results of digestion with endoproteinase Glu-C. As with trypsin, two main sites of cleavage were observed. Cleavage occurred a t Thr546, two positions away from the N terminus of T10, and at Arg'431, 12 residues away from Ser'44". Both linker regions were also proteolyzed when the protein was treated with elastase, but harsher conditions were required to release the N-terminal loading di- domain. With endoproteinase Arg-C, only the second linker region was cut, unless harsh conditions were used, when a second cleavage was observed within the first ketosynthase domain KS1, releasing a larger loading domain than with the other enzymes. Remarkably, the C-terminal part of the molecule remained very stable and highly resistant to hydrolysis by any of the proteinases used (no cleavage was observed between KS2 andAT3, and only under harsh conditions were cuts observed within some other parts of this fragment).

Sequence comparisons have been used to predict the boundaries of the individual domains present in DEBS 1 (6, 7), and the results presented here are broadly consistent with those predictions: only a few minor cuts were found inside the predicted extent of some domains (data not shown). Moreover, one of the two main linker resons, the C-terminal part of the loading domain, is flanked by two regions rich in alanine, proline, aspartic and glutamic acids (ie., 532AEALAAGTE and 552EAAPGEPVA), which are thought to provide flexibility (7, 25). In the same way, the main point of cleavage with elastase (Ala2025-Thr'00"6) is located just at the C-terminal end of an- other putative linker region (""'"AEQAAPA) (7).

The three main fragments obtained after digestion with trypsin (see Fig. 1B) were readily separated by gel filtration under non-denaturing conditions, indicating that the fragments are independently folded and that they do not interact strongly with each other. The N-terminal and central di-domains behaved on gel filtration as monomeric species, while the large C-terminal fragment was apparently dimeric. The individual KR1 fragments generated by trypsin or by endoproteinase Glu-C cleavage are also monomers. DEBS 1 itself behaves on gel filtration as a homodimer of molecular mass 660 kDa, so these results taken together suggest that the dimer is held together only by strong interactions within the C-terminal half of the molecule. This is the first evidence of any kind for the structure in solution of a bacterial polyketide synthase.

After digestion with trypsin, the N-terminal acyltransferase- acyl carrier protein di-domain of DEBS 1 was specifically acylated by [l-14Clpropionyl-CoA, providing the first evidence that this portion of DEBS 1 functions as a "loading domain" for the

8528 Proteolysis of Erythromycin-producing Polyketide Synthase

propionate starter unit, as previously proposed (6), and con- firming that the acyltransferase remains active after limited proteolysis. We have recently shown (20, 26) that all of the purified DEBS multienzymes are transiently acylated by the substrate methylmalonyl-CoA, presumably at the active sites of each of the six acyltransferases proposed to load methylmalonyl-CoA extender units onto the enzyme. The finding that, after trypsin digestion, both the central ketosynthase-acyltransferase di-domain and the large C-terminal fragment (which contains the third acyltransferase activity of DEBS 1) were specifically labeled by [l-14Clmethylmalonyl-CoA, similarly provided evidence that both of these acyltransferases re- tain activity after limited proteolysis, and supports the as- sumption that the acyltransferases function in the loading of methylmalonyl-CoA extender units (4, 6, 7, 20, 26).

The idea of a compact, relatively proteinase-resistant C-terminal half of the DEBS 1 molecule, involved in inter-subunit interactions, receives additional support from the finding that tryptic fragments T6 and T5, and endoproteinase Glu-C fragments G5 and G8, have native molecular masses appropriate for dimeric species. The results do not as yet allow discrimina- tion between the two possibilities of parallel and antiparallel arrangement of the polypeptides within the DEBS 1 dimer. Further work is required to establish the detailed topology of the enzyme, and thus to gain insight into how the growing polyketide chain is efficiently transferred to each active site in turn.

REFERENCES 1. Corcoran, J. W. (ed) (1981) Antibiotics, Vol. 4, pp. 132-174, Springer-Verlag,

New York

2

4 3

5

6 7

8.

9. 10.

11. 12. 13.

14.

15. 16. 17. 18. 19.

20.

22. 21.

23. 24.

25.

26.

Seno, E. T., and Hutchinson, C. R. (1986) in The Bacteria (Queener, S. W., and

Hopwood, D. A., and Sherman, D. H. (1990) Annu. Rev. Genet. 24, 37-66 Cortes, J., Haydock, S. F., Roberts, G., Bevitt, D. J., and Leadlay, P. F. (1990)

Donadio, S., Staver, M. J., McAlpine, J. B., Swanson, S. J., and Katz, L. I19911

Donadio, S., and Katz, L. (1992) Gene 111, 5 1 4 0 Bevitt, D. J., Cortes, J., Haydock, S. F., and Leadlay, P. F. (1992) Eur J.

Witkowski,A., Rangan, V. S., Randhawa, Z. I., Amy, C. M., and Smith, S. (1991)

Wakil, S. J. (1989) Biochemistry 28, 45234530 Bleile, D. M., Munk, P., Oliver, R. M., and Reed, L. J. (1979) Proc. Natl. Acad.

Perham, R. N. (1991) Biochemistry 30, 8501-8512 Rossmann, M. G., and Argos, P. (1981) Annu. Reu. Biochem. 50, 497-532 Rubenstein, D. S., Enghild, J. J., and Pizzo, S. V. (1991) J. Biol. Chem. 266,

Caffrey, P., Bevitt, D. J., Staunton, J., and Leadlay, P. F. (1992) FEBS Lett. 304,

Morrisey, J. M. (1981)Anal. Biochem. 117, 307-310 Laemmli, U. K. (1970) Nature 227,680485

Matsudaira, P. (1987) J. Biol. Chem. 262, 10035-10038 Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 Lowry, 0. H., Rosebrough, N. J., Far?., A. L. and Randall, R. J. (1951) J. Biol.

Marsden,A. F. A,, Caffrey, P.,Aparicio, J. F., Loughran, M. S., Staunton, J., and

Edelman, G. M. (1973) Science 180, 830-840 Coggins, J. R., and Hardie, D. G. (19861 in Multidomain Proteins: Structure

and Evolution (Hardie, D. G., and Coggins, J. R., eds) pp. 1-12, Elsevier Science Publishing Co., New York

Day, L. E. eds) Vol. 9, pp. 231-279, Academic Press, New York

Nature 348, 176-178

Science 252, 675479

Biochem. 204, 3 9 4 9

Eur J. Biochem. 198, 571-579

Sci. U. S. A. 76, 43854389

11252-11261

225-228

Chem. 193,265-275

Leadlay, P. F. (1994) Science, in press

Fontana, A,, Fassina, G., Vita, C., Dalzoppo, D., Zamai, M., and Zambonin, M. Porter, R. R. (1959) Biochem. J . 73, 119-126

Stephens, P. E., Darlison, M. G., Lewis, H. M., and Guest, J. R. (1983) Eur J.

Roberts, G . A,, Staunton, J., and Leadlay, P. F. (1993) Eur: J. Biochem. 214,

(1986) Biochemistry 25, 1847-1851

Biochem. 133,481489

305-311

Documents

Limited Proteolysis and Active-site Studies of the First Multienzyme