biology, Ohio State University, 484 West 12th Ave., Columbus, OH. 43210. Tel. ... conserved residues (Chapman et al., 1988; Knight et al., 1989;. Schneider et al.
Vol. 266, No. 12, Issue of April 25, pp. 7417-7422, 1991 Printed in U.S.A.
CHEMISTRY THEJOURNALOF BIOLOGICAL
(0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.
Mutations in the Small Subunitof Cyanobacterial Ribulosebisphosphate Carboxylase/Oxygenase That Modulate Interactions with Large Subunits* (Received for publication, September 20, 1990)
Bonggeun Lee#, Randy M. Berkas, and F. Robert TabitaS7l From the tDeoartment of Microbioloev and The Biotechnology Center, Ohio State University, Columbus, Ohio43210 and 94080 §Genencor’InE., South sun Francisco~“Ca~if0rnia
In the cyanobacteriumAnacystis nidulans (Synechococcus PCC6301), ribulose 1,5-bisphosphate carboxylase/oxygenase (Rbu-Pz carboxylase) is composed of eight large subunits and eight small subunits. There are three regions of the small subunit that contain amino acids that are conserved throughout evolution, from bacteria to higher plants. Since the function of the small subunit is not fully understood, site-directed mutagenesis was performed on highly conserved residues in the first and second conserved regions. Ser-16, Pro-19, Leu-21,and Tyr-54 were replaced by Asp-16, His-19, Glu-21, and Ser-54, respectively. Crude extracts containing the recombinant His-19 mutant enzyme indicated that there was little effect on either Rbu-Pz carboxylase activity or interactions between large and small subunits. However, the Asp-16, Glu21, and Ser-54 mutations showed effects on Rbu-P, carboxylase activity and the interaction between large and small subunits. The large and small subunits of the Asp-16, Glu-21, and Ser-54 enzymes were found to dissociate during nondenaturing gel electrophoresis or sucrose density gradient centrifugation. However, the dissociated small subunits remained functional and were capable of reconstituting Rbu-Pz carboxylase activity when added to large subunits. These results indicated that Ser-16, Leu-21, and Tyr-54 might play an important role in interactions between large and small subunits of the A. nidulans enzyme.
subunit is not fully understood. Both subunits contain conserved regions which have common amino acid residues that are foundin awide range of photosynthetic and chemolithoautotrophic organisms. With regard to the small subunit, analysis of the amino acidsequencefroma number of sources indicated that there are three major conserved regions (Nierzwicki-Bauer et al., 1984; Tabita, 1988). Recent x-ray structural studies have focused on the plant L8S8 Rbu-Pp carboxylase, and the structure has been refined enough to analyze detailed domain contacts and suggest a function of various conserved residues (Chapman et al., 1988; Knight et al., 1989; Schneider et al., 1990). Theprimarystructure of Rbu-P2 carboxylase from the cyanobacteriumAnacystis nidulans ( S y nechococcus PCC6301) closely resembles the plant enzyme (Shinozaki and Sugiura, 1983); however, the large and small subunit genes are linked and co-transcribed (Shinozaki and Sugiura, 1985), allowing for the expression of highly active and correctly assembled recombinant Rbu-Pz carboxylase in extracts of Escherichia coli (Christeller et al., 1985; Gatenby et al., 1985; Tabita and Small,1985). In this study, site-directed mutagenesis was performed on highly conserved amino acid residues of the first and second conserved regions of the A . nidulans small subunit; the primary effect of these mutations on catalysis and the interaction between large and small subunitswere analyzed. EXPERIMENTALPROCEDURES
Materials-Ribulose 1,5-bisphosphate was prepared according to the previously reported method (Horecker et d., 1958). Restriction enzymes were purchased from New England Biolabs (Beverly, MA) Ribulose 1,5-bisphosphate carboxylase/oxygenase (Rbu-Pp or GIBCO/BRL Life TechnologiesInc. SequenaseandT4 DNA carboxylase)’ is a bifunctional enzyme which catalyzes the polymerase were obtained from U. S. Biochemical Cop. and Biofixation and photorespir- Rad, respectively. NaHI4C0:, was purchased from Amersham Corp., first reaction of photosynthetic atory carbon oxidation (Tabita,1988). The predominantform and [3sSS]n-thio-dATPwas obtained fromAmershamCorp. or Du of Rbu-P, carboxylase is ahexadecameric protein (LsSs) Pont-New England Nuclear. BacterialStrains,Phages,andPlasmids-E. coli strain MV1190 composed of eight large (L) subunitsandeightsmall (S) (d(lac-proAB),~hi,supE,~(srl-recA)306::Tn10(tet‘)[F’traD36,proAB, subunits. The small subunits are attached to the top and the lacIqZAM15]); strainCJ236 (dut-l,ung-l,thi-l,relA-l;pCJ105(Cmr)) bottom of the large subunit octamer (Chapman et al., 1988; (Kunkel et al., 1987);helperphageM13K07(Vieira and Messing, Anderson et al., 1989). While the sites for activation and 1987); and plasmid pTZ18R (Mead et al., 1986) were used for this study. E . coli MV1190 and CJ236 were purchased from Bio-Rad. catalysis reside in the large subunit, the role of the small Cell Culture and Preparation of CrudeExtracts-E. coli strains * This work was supported by National Institutes of Health Grant harboring mutated plasmids were grown a t 37 “C overnight in 25 ml GM 24497. The costs of publication of this article were defrayed in of Luria-Bertani medium (Maniatis et al., 1982) containing 100 fig/ ml ampicillin; 1.5 ml of this culture was used to inoculate 250 ml of part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with18 U.S.C. Section Luria-Bertani medium containing 100 pg/ml ampicillin. To this culture, 1 mM isopropyl 6-D-thiogalactopyranosidewas added, when the 1734 solely to indicate this fact. ll To whom correspondence should he addressed: Dept. of Micro- A,,, reached about 0.4. Theculture was incubatedwith vigorous biology, Ohio State University, 484 West 12th Ave., Columbus, OH shaking for 12 h prior to harvesting. Harvested cells were washed in TEM (25 mM Tris-HC1, p H 8.0, 1 mM EDTA, 5 mM P-mercaptoeth43210. Tel.: 614-292-4297; Fax: 614-292-1538. anol), resuspended in TEM buffer containing 1 mM phenylmethylThe abbreviations usedare: Rbu-P2 carboxylase,ribulose1,5bisphosphate carboxylase/oxygenase; L, large subunit; S, small subsulfonyl fluoride, and disrupted by sonication. To prepare thesoluble unit; kb, kilobase(s); SDS, sodium dodecyl sulfate; PAGE, polyacry- protein fraction, the extractswere centrifuged twice a t 12,000 X g for amide gel electrophoresis. 10 min and 100,000 X g for 60 min.
Cop
’
7417
Mutant Rbu-P2Carboxylase Small Subunits
7418
In Vitro Mutagenesis-The Muta-gene phagemid in vitro mutagenesis kit, based upon a previously reported method (Kunkel, 1985; Kunkel et al., 1987), was purchased from Bio-Rad. The mutagenic oligonucleotidesprepared for this study are shown in Fig. 1. The 2.2kb PstI fragment, containing the rbcL and rbcS genes of A. nidulans, was cloned into the phagemid vector pTZ18R in the correct orientation under control of the lac promoter (Lee and Tabita, 1990). The resulting plasmid (pBGL710) was isolated and transformed into the E. coli dut ung strain, CJ236, in order to prepare uracil-containing template DNA. Uracil-containing phagemids were grown by adding helper phage M13K07 to the E. coli CJ236(pBGL710) culture. Phagemids were collected by precipitation with ammonium acetate/polyethylene glycol 8000. Uracil-containing single-stranded template DNA was prepared by extracting the phagemids with phenol and chloroform. Subsequently, phosphorylated mutagenic oligonucleotides were annealed with this template, and thecomplementary DNA strand was synthesized with T4 DNA polymerase. The resulting reaction mixture was transformed intoE. coli MV1190.Mutants were screened by DNA sequencing of the target region. The dideoxynucleotide chain termination method was used for DNA sequencing (Sanger et al., 1977). The pBGL710 plasmid which carried the desired mutation was isolated, and a portion of this plasmid, other than the 0.7-kb EcoRIIPstI fragment containing the entire rbcS sequence and a small fragment at the 3' end of rbcL, was exchanged with the 1.5kb BamHIIEcoRI fragment of plasmid pBGL520 and the BanHI/ PstI fragment of vector plasmid pTZ18R (Fig. 2). This construction ensured that there was no possibility for secondary mutations in the large region of the rbcL gene that was exchanged and obviated the need to sequence the entire large subunit gene. The EcoRI/PstI fragment which underwent mutagenesis was then completely sequenced to exclude the possibility of a secondary mutation at other than the target site. A single amino acid change was confirmed for each mutagenesis; Residue numbers refer tothe Anacystis small subunit sequence (Shinozaki and Sugiura, 1983). Quantitation of Rbu-Pp Carboxylase in E. coli Crude ExtractsUsing the purified holoenzyme as a standard, the amount of Rbu-Pp carboxylase protein in E. coli crude extracts was measured by rocket immunoelectrophoresis (Jouanneau and Tabita,1986). Antibodies to the recombinant holoenzyme were raised in rabbits. About 20%of the soluble protein of E. coli crude extracts was found to be Rbu-PZ carboxylase. Specific Activityof Rbu-Pz Carboxylase-Carboxylase activity was Dl6 S'TTC GAG ACT TTC GAC TAC CTG CCT CCC3'
ASP E19
5'TC TCG TAC CTG
3'
C X CCC CTC AGC GAT
EIS P2 1
5'TAC
CTG
CCT
CCC GAA AGC GAT CGC CAA3 '
GLD 5'TTC GAG ACT TTCTCG TAC CTG CCT CCC CTC AGC 3' PHE GLU THR PHE SPR TYR LEU PRO PRO LEO SER
c
I Plac
ACG
ATG
TGG
AAG
CTC
CCC CTG TTT
GAC
TGC
AAG3 '
TYR TRP THR MET TRP LYS LEU PRO LEU PHE ASP CYS LYS ACG ATG TGG AAG CTC CCC CTG TTT GAC TGC AAG 3 ' SPR TTC TAC T K ACG ATG TGG AAG CTC CCC CTG TTT GAC TGC AAG 3' PEE
S 5 4 5'CCG GAA TTC T U TGG
F55 S'CCG GAA
Sequenced ~ragrnent
pTZ18R
FIG. 2. Subcloning of mutated pBGL710. The 1.5-kb EcoRI fragment was removed from mutagenized plasmid pBGL710 to form plasmid pBGL711. The 0.7-kb EcoRI/PstI fragment from pBGL711 was ligated to the 1.5-kb BarnHI/EcoRI fragment from pBGL520 (Lee and Tabita,1990) and theBamHI/PstI fragment of pTZ18R ( B , BamHI; E, EcoRI; H, HindIII; P , PstI). In the final construct, only the 0.7-kb EcoRI/PstI fragment had undergone mutagenesis. Finally, the 0.7-kb EcoRI/PstI fragment was sequenced. measured by l4COZincorporation into acid-stable 3-phosphoglyceric acid as previously described (Whitman and Tabita, 1976). Specific activity was expressed as micromoles of l4CO2fixed per min per mg of immunologically quantitated enzyme protein in crude extracts. Reconstitution assays were performed with recombinant large and small subunit preparations using plasmids pBGL520 and pBGL535, which encoded the large and small subunit genes, respectively (Lee and Tabita, 1990). When E. coli crude extracts containing recombinant large or small subunits were added to the Rbu-Pp carboxylase assay mixture, incubation proceeded at 30 "C for 5 min prior to activating the enzyme. Other Analyses-The protein concentrationof E. coli crude extracts was determined by either the Lowry method (Lowry et al., 1951) or Bradford procedure (Bradford, 1976). Sucrose density gradient centrifugation in 0.2-0.8 M sucrose step gradients was performed as previously described (Tabita and Small, 1985). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed according to a method previously described (Laemmli, 1970).
RESULTS
a
S'TAC TGG
c
FIG. 1. Plasmid construction and mutagenic oligonucleotides. The 2.2-kb PstI DNA fragment containing the large and small subunit genes ofA. nidulans Rbu-Pz carboxylase was cloned into pTZ18R (Lee and Tabita, 1990). Mutagenic oligonucleotides of Asp16, His-19, Glu-21, Ser-54, and Phe-55 are shown above and below the nucleotide and amino acid sequence of the first and second conserved regions of the A. nidulans small subunit. The underlined nucleotides indicate the mismatches. bp, base pairs.
Site-directed Mutagenesis-Five different mutant plasmid constructs, which carried an intact large subunit gene and a mutated small subunit gene,were generated as described under "Experimental Procedures" (Fig. 1).Eachmutation gave rise to a single amino acid substitution in the small subunit. Effect of Mutations on Carboxylase Actiuity-The large subunit andmutated or wild type small subunit gene, both under control of the lac promoter, were expressed in E. coEi, and the carboxylase activity was measured in crude extracts (Table I). In this experiment, the amount of enzyme protein was measured by rocket immunoelectrophoresis to provide a direct comparison with the extracts from the different constructs. Each of the mutant enzymes exhibited different levels of carboxylase activity, with mutant HI9 exhibiting nearly the same specific activity as thewild type enzyme. However, more dramatic effects on enzyme activity were observed with the Asp-16, Glu-21, Ser-54, and Phe-55 mutant enzymes. These mutant enzymes showed 74, 30, 70, and 71%, respectively, of the specific activity of the wild type enzyme. Yet,it should be noted that all mutant enzymes showed levels of activity that greatly exceeded the activity of large subunits alone, which ranged from 0.15 to 0.5% of the activity of the holoenzyme
Rbu-P2 Mutant
Carboxylase Small Subunits A
TABLEI Carboxylase activity of mutant enzymes in E. coli crude extracts One unit represents 1 pmol of CO, fixed per min; the amount of Rbu-PI carboxylasefrom each crude extract was measuredby rocket immunoelectrophoresis (Jouanneau and Tabita,1986). Strain
7419
1
2
3
4
5
6
7
8
Specific activity unitsf m g enzyme
MV1190(pBGL710)
Asp-16 (Ser + Asp)
His-19 (Pro + His) Glu-21 (Leu + Glu) (Tyr Ser-54 + Ser) Phe-55 (Trp 4Phe)
1.29 0.96 1.24 0.39 0.91 0.92
(Andrews, 1988, Lee and Tabita, 1990), suggesting that the mutated, small subunits continued to bind to large subunit octamers in E. coli crude extracts. Influence of Mutations onthe Interaction of Large and Small Subunits-Nondenaturing gel electrophoresis was performed with E. coli crude extracts containing mutant enzymes. Extracts from each strain showed a distinct band at about the position of the Lass wild-type Rbu-P2 carboxylase; however the Rbu-PZ carboxylase band from extracts of strains Asp-16, Glu-21, and Ser-54 appeared at a slightly higher position than the Rbu-P2 carboxylase band found in crude extracts of the wild type, His-19, or Phe-55 mutants (Fig. 3A). When SDSPAGE was performed with an excised Rbu-P, carboxylase band from the nondenaturing gel, only the Rbu-P, carboxylase band from the His-19 and Phe-55 E. coli crude extracts was found to contain small subunits (Fig. 3B). The putative RbuP, carboxylase bands from mutants Asp-16, Glu-21,and Ser54 contained only large subunits, presumably in an La structure. These results indicated that the Asp-16, Glu-21, and Ser-54 small subunits were dissociated from large subunits during nondenaturinggel electrophoresis. By extension, these results suggest that the large and small subunits of Asp-16, Glu-21, and Ser-54 are associated less tightly than those of the wild type enzyme, mutant His-19, or mutant Phe-55. To further investigate the association of the large and small subunits, sucrose density gradient centrifugation was performed with the various crude extracts andcarboxylase activity was measured in each sucrose density gradient fraction (Fig. 4). The E. coli crude extract containing the His-19 and Phe-55 mutant enzymes showed carboxylase activity at the same position as theE. coli crude extract containing the wild type (LESS) enzyme (Fig. 4, A , C, and E ) . This indicated that the His-19 and Phe-55 mutant enzymes remained intact during sucrose density gradient centrifugation, in agreement with the results from the electrophoresis experiment (Fig. 3). However, extracts from mutants Asp-16, Glu-21, and Ser-54 did not show carboxylase activity in anyfraction from the sucrose gradient (Fig. 4, B, D, and F). This indicated that either the undissociated Asp-16,Glu-21, and Ser-54 mutant enzymes were inherently unstable and lost all enzyme activity or that the large and small subunits of these mutant enzymes dissociated during sucrose density gradient centrifugation. Recently, we showed that active Rbu-P2 carboxylase may be reconstituted in vitro using separately expressed recombinant large and small subunits; this reconstitutionsystem was further found to beuseful for localizing large and small subunits after sucrose density gradient centrifugation of E. coli crude extracts containing Rbu-P, carboxylase subunits (Lee and Tabita, 1990). The in vitro reconstitution assay was thus employed to localize the large and small subunits of sucrose gradient fractionated extracts from mutants Asp-16, Glu-21, and Ser-54 simply by adding recombinant wild-type
B 1 2 3 4 5 6 7
ss + FIG.3. Nondenaturing and SDS-PAGE of E. coli crude extracts containing mutant enzymes. A, PAGE was performed in 5%acrylamide gels with 200 pg of each E. coli crude extract using the Laemmli (1970) gel system in the absence of SDS. Lane I, purified recombinant holoenzyme; lane 2, E. coli MV1190(pBGL710);lane 3, E. coli MV1190(Asp-16);lane 4, E. coli MV1190(His-19); lane 5, E. coli MV1190(Glu-21);lane 6,E. coli MV1190(Ser-54);lane 7,E. coli MV1190(Phe-55);lane 8,E. coli MVllSO(pTZ18R). The left arrow indicates the Rbu-Pz carboxylase band. B, after nondenaturing gel electrophoresis, the gel was briefly stained with 0.1% Coomassie Blue
in 45% methanol and 9% glacial acetic acid for 30 min, destained with 5% methanol and 10% acetic acid for 20 min, and then rinsed with water. Finally Rbu-P, carboxylase bands were removed, equilibrated with SDS buffer (0.15 M Tris-HC1, pH 6.8, 2% SDS, 1% 0mercaptoethanol, 10% glycerol) for 1.5 h, and transferred to a 14% acrylamide slabgel. SDS-PAGE wasthen performed. Lane I , purified holoenzyme; lanes 2-7 indicate Rbu-Pz carboxylase bands fromlanes 2-7, respectively, of A. Large subunits ( L S )and small subunits( S S ) are indicatedby the arrows.
large and small subunits to each fraction. The results indicated that the large (La) and small subunits eluted at the expected position on the gradient (Fig. 5). In addition, differential levels of reconstituted Rbu-P2carboxylase activity were observed with different mutant small subunits with relative activities similar to thatobserved in crude extracts (Table I). These results strongly suggest that the large and small subunits of the Asp-16,Glu-21, and Ser-54 mutant enzymes dissociated during sucrose density gradient centrifugation. Thus, the Asp-16, Glu-21, and Ser-54 mutations seemed to have a detrimental effect on large and small subunit interactions, manifested by a reduced ability of the subunits to bind and hence remain associated after nondenaturing gel electrophoresis or sucrose density gradient centrifugation. DISCUSSION
Although the smallsubunitis required forfull Rbu-P, carboxylase activity (Andrews and Ballment, 1983),the func-
7420
Mutant Rbu-Pz Carboxylase Small Subunits A
80000
B 4
00000
4
3
60000
3
40000
2 I
1
20000
1
0
0
60000
E
40000
2
u 20000
0
IO
0
20
30
W
4
h
0
0
0
40
10
Fraclion no.
FIG. 4. Sucrose density gradient centrifugation of E. coli crude extracts containing mutant enzymes. 25 mg of each E. coli crude extract was loaded onto a 0.2-0.8 M sucrose step gradient. Fractionsof 1ml werecollected and samples were assayed for Rbu-P, carboxylase activity. A , E. coli MV1190(pBGL710); B, E. coli MV1190(Asp-16); C, E. coli MV1190(His-19); D, E. coli MV1190(Gl~-21);E, E. coli MV1190(Phe-55); F, E. coli MV1190(Ser-54). Rbu-P:! activity carboxylase was expressed as counts/min of I4CO2fixed in a5-min assay (+); l3, absorbance at 280 nm.
"
3
40000
2
20000
0 20
30
4
3
60000
8 4
10
40000
2
0
1
20000
0
0
0 0
40
10
20
30
40
Fraclion no.
E
F
80000
4
80000
4
60000
3
6nooo
3
40000
2
0
8 4
20000
0
0
W
1
Fraction no.
z
40
D
80000
4
60000
0
30
Fraction no.
C
80000
20
10
30
20 Fraclion no.
tion of the small subunit is not fully understood. Recombinant large subunit octamers contain weak but measurable carboxylase activity (Andrews, 1988; Leeand Tabita,1990), strongly suggesting that thesmall subunit does not directly participate in the catalytic mechanism. The small subunits from various organisms do share a number of conserved residues, found in three majorregions in the amino acid sequence. Thus, to further probe the role of the small subunit, we have begun a program of site-directed mutagenesis and have focused on these conserved residues. In particular, Ser-16, His-19, Leu21, and Tyr-54of the A . nidulans small subunit arecompletely conserved in small subunits of other organisms ranging from bacteria to plants, with the exception of Chromatium uinosum, which hasa conservatively substituted methionine at the position corresponding to Leu-16 of the A. nidulans small subunit (Viale et al., 1989); red algae (Valentin and Zetsche, 1989, 1990) and Cyanophora paradoxa (Starnes et al., 1985) have phenylalanine and valine, respectively, at the position corresponding to Tyr-54 of the A . nidulans small subunit. According to the x-ray structure of the homologous spinach LA&Rbu-Py carboxylase, several conserved residues appear to participate in subunit interactions (Knight et al., 1989; Schneider et al., 1990). Indeed, by analogy to thespinach L8S8 Rbu-P2 carboxylase, Ser-16, Pro-19, and Tyr-54 of the A .
40
40000
2
0
1
20000
0
0
I
1
0 0
10
20
30
40
Fraction no.
nidulans small subunit are amino acid residues that form the major interaction locus with large subunits (Knight et al., 1989).In particular, Pro-19 and Leu-21 of the spinach enzyme showed side chain interactions with large subunits within 4.0 A (Schneider et al., 1990). Since small subunits are required for full Rbu-P2 carboxylase activity, a perfect match between the large and small subunits appears to be a prerequisite for full Rbu-Pz carboxylase activity. The His-19 mutation, which involved replacement of Pro-19 with a histidine residue, did not affect RbuP2 carboxylase activity or perturb any discernible interaction between the subunits. Although the change from proline to histidine was initially expected to produce dramatic effects, the His-19 mutation apparently provided nearly perfect subunit interactions, such that theHis-19 mutant enzyme showed about the same activity as the wild type enzyme. Perhaps Pro-19 might not actually play an important role in interactions between the large and small subunits; alternatively, a more drastic mutation might influence this interaction. The Phe-55 mutantenzyme showed 71%of the activity of the wild type enzyme in E. coli crude extracts and was stable throughoutnondenaturing gel electrophoresis and sucrose density gradient centrifugation. The structural stability of the Phe55 enzyme was thus consistent with previously reported re-
Rbu-P2 Mutant
Carboxylase Small Subunits
7421 B
A 20000
100000 80000
60000
g
B
10000
40000
FIG. 5. Localization of the large
and small subunits after sucrose gradient fractionation of E. coli crude extracts containing the Asp16, Glu-21, or Ser-54 mutant enzymes. 25mg of E. coli crude extract was loaded onto a 0.2-0.8 M sucrose step gradient, and fractions of 1 ml were collected. To localize the large subunits, samples of each 1-ml fraction were assayed in the presence of 20 pl of E. coli MVllgO(pBGL535) crude extract (34.9 mg/ml) containing recombinant small subunits. Rbu-Pn carboxylase activity was expressed as counts/min of “Con fixed in a 10-min assay (El). A , E. coli MV1190(Asp-16); C, E. coli MV1190(Glu-21);E, E. coli MV1190(Ser-54). To localize the small subunits, samples of each 1-ml fraction were assayed in the E. coli MV1190presence of 20plof (pBGL520) crude extract (30.4 mg/ml) containing recombinant large subunits. Rbu-Pzcarboxylase activity was expressed as counts/min of 14C02fixed in a 10-min assay (El). B , E. coli MV1190(Asp-16); D,E. coli MV1190(Glu-21); F, E. coli MV1190(Ser-54).
20000 0
0 0
10
20
30
0
40
10
g
i
80000
60000
B
loooo
0
30
40
D
C 20000
20 Fraclion no.
Fraclion no.
i
40000
20000
4,
o{ 0
0
20
30
F
E
20000
looooo
Fraclion no.
sults where the purified Phe-55 enzyme exhibited 44% V,,, of the wild-type enzyme (Voordouw et al., 1987). The x-ray structure of spinach LESERbu-Pa carboxylase indicated that each small subunit forms extensive interactions with two large subunits and also interacts with the amino terminus domain of a third large subunit (Knight et al., 1989). In our experiments, it was apparent that the Asp-16, Glu-21, and Ser-54 mutations affected interactions between large and small subunits, yet the Asp-16, Glu-21, and Ser-54 small subunits did show some ability to bind to large subunit octamers in E. coli crude extracts,since each mutant enzyme showed carboxylase activity at levels which necessitated the presence of small subunits in the enzyme structure. Since one small subunit seems to bind to more than one large subunit, it is possible that small subunits which cause mismatch in one interaction area may still bind to large subunit octamers in E. coli crude extracts. The decreased amounts of carboxylase activity observed in the Asp-16, Glu-21, and Ser-54 mutant enzymes may simply be due to the differential degree of mismatch between large and small subunits, unless small subunits are actually involved in catalysis. In this connection, it was recently suggested from comparisons of the crystal structures of the L2 andLESEenzymes from Rhodospirillum rubrum and spinach, that small subunits produce small conformational changes which might influence function (Schneider et al., 1990). In the Asp-16, Glu-21, and Ser-54 mutant enzymes,
10
Fraclion no.
Fraclion no
7
Fraction no.
the interactions between the large and small subunits were not strong enough to retain the holoenzyme structure upon nondenaturing gel electrophoresis or sucrose density gradient centrifugation. Thus,further purification of these mutant enzymes was impossible. However, dissociated mutant small subunits, isolated from the sucrose gradient, were able to reconstitute significant Rbu-P2carboxylase activity with large subunits, indicating that the dissociated mutant small subunits still retainedsome ability to confer Rbu-Ppcarboxylase activity to large subunit octamers. In any case, the results of this study and recent x-ray crystallographic investigations indicate that Ser-16, Leu-21, and Tyr-54 of the small subunit may participate in interactions between large and small subunits in A. nidulans. Previous results of site-directed mutagenesis on cyanobacterial small subunit residues have been reported (Voordouw et al., 1987; McFadden and Small, 1988; Fitchen et al., 1990). However, until the present study, there has been no demonstration that mutated small subunits may reassociate and reconstitute Rbu-P2carboxylase activity with wild type large subunits in vitro. Recently, we purified recombinant large subunit octamers capable of reconstituting with separately expressed recombinant small subunits to form highly active Rbu-P2 carboxylase (Lee and Tabita, 1990). I n vitro reconstitution of large subunit octamers with mutant small subunit proteins may facilitate future subunit interaction studies and
7422 Lowry, 0.H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193,265-275 Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) M o l e c ~ ~ r Cloning: A Laboratory Munual, p. 440, Cold Spring Harbor LaboAc~no~ledgmen~-We thank the organic chemistry group at Geratory, Cold Spring Harbor, NY nentech for synthetic oligonucleotides, McFadden, B. A., and Small, C. L. (1988) Photosynth. Res. 18, 245260 REFERENCES Mead, D. A,, Szczesna-Skorupa, E., and Kemper, B. (1986) Protein Eng. 1,67-74 Anderson, I., Knight, S., Schneider, G., Lindqvist, Y., Lundqvist, T., Nierzwicki-Bauer, S. A., Curtis, S. E., and Haselkorn, R. (1984) Proc. Branden, C., and Lorimer, G. H. (1989) Nature 337,229-234 Nutl. Acad. Sei. U. S*A. 81,5961-5965 Andrews, T. J , (1988) J. Biol. Chern. 263,12213-12219 Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Andrews, T. J., and Ballment, B. (1983) J. Biol. Chem. 258, 7514Sci. U. S. A. 74,5463-5467 7518 Schneider, G., Knight, S., Anderson, I., Branden, C., Lindqvist, Y., Bradford, M. M. (1976)Awl. Bioc~em.72,248-254 and Lundqvist, T. (1990) E M 3 0 J. 9,2045-2050 Chapman, M. S., Suh, S. W., Curmi, P. M. G., Caacio, D., Smith, W. Shinozaki, K., and Sugiura, M. (1983) Nucleic Acids Res. 11, 6957W., and Eisenberg, D. S. (1988) Science 241,71-74 6964 Christeller, J. T., Terzaghi, B. E., Hill, D. F., and Laing, W. A. (1985) Shinozaki, K., and Sugiura, M. (1985) Mot. Gen. Genet. 200,27-32 Plant MOLBiOL 5,257-263 Starnes, S. M., Lambert, D. H., Maxwell, E. S., Stevens, S. E., Jr., Fitchen, J. H., Knight, S., Anderson, I., Branden, C., and McIntosh, Porter, R. D., and Shiveiy, J. M. (1985) FEMS ~ i c ~ Lett. ~ ~ 28, l . L. (1990) Proc. Natl. Acad. Sei. U. S. A. 87,5768-5772 165-169 Gatenhy, A. A., Van der Vies, S. M., and Bradley, D. (1985) Nature Tabita, F. R. (1988) Microhiol. Reo. 52, 155-189 314,617-620 Tabita. F. R.. and Small. C. L. (19851 . , Proc. Natl. Acad. Sci. U. S. A. Horecker, B.L., Hurwitz, J., and Weissbach, A. (1958)Biochem. 82,6100-6103 Prep. 6,83-90 Valentin. K.. and Zetsche. K. (1989) Curr. Genet. 1 6 , 203-209 Jouanneau, Y., and Tabita, F. R. (1986) J, Bucteriol. 165,620-624 Valentin; K.; and Zetsche; K. (1990) Mol. Gen. Genet: 222,425-430 Knight, S., Anderson, I., and Branden, C.-I. (1989) Science 244, Viale, A.M., Kobayashi, H., and Akazawa, T. (1989) J. BacterioL 702-705 171,2391-2400 Kunkel, 1'. A. (1985) Proc Natl. Acad. Sei. U. 8. A. 82,488-492 Vieira, J., and Messing, J. (1987) Methods Enzymol. 153, 3-11 Kunkel, T. A,, Roberts, J. D., and Zakour, R. A. (1987) M e t h ~ s Voordouw, G., De Vries, P. A,, Van den Berg, W.A.M., and De Clerck, E. P. J. (1987) Eur. J. Biochem. 163,591-598 Enzymol. 154,367-382 Whitman, W., and Tabita,F. R. (1976) Biochem.Biophys. Res. Laemmli, U,K. (1970) Nature 227,680-685 C o ~ ~ u71,1034-1039 n. Lee, B., and Tabita,F. R. (1990) n i o c h e ~ ~29,9352-9357 ~ry
may also provide information on how small subunits influence Rbu-P, carboxylase activity.
