A BSTRACT

Phosphomannose isomerase (PMI) plays a pivotal role in biosynthesis of GDP-mannose, an important precursor of many polysaccharides. We demonstrate in this study that

Pseudomonas aeruginosa pslB

encodes a protein with GDP-mannose pyrophosphorylase/PMI dual activities. The GDP-man PPase activity is Mg²⁺-dependent, whereas the PMI activity is Co²⁺-dependent and could be inhibited by GDP-mannose in a competitive manner. Furthermore, the PMI activity could be inactivated by 2, 3-butanedione suggesting the presence of a catalytic Arg residue. Site-specific mutations at R373, R472, R479, E410, H411, N433 and E458 increase the KM approximately 8- to 20-fold. The PMI activity of PslB was completely diminished with a R408K or R408A, reflecting the importance of this residue in catalysis. The CD spectra of R408A, R408K and wild type PslB are nearly identical, indicating that there is nearly no alterations of their secondary structures.

Overall, these results provide a basis for understanding the catalytic mechanism of PMI.

1.2 Introduction

Phosphomannose isomerase (PMI) catalyzes the reversible interconversion of fructose-6-phosphate (Fru-6-p) and mannose-6-phosphate (Man-6-p) (1). This reaction links Man-6-p into the mannose metabolism pathway resulting in the generation of GDP-mannose (GDP-Man), an important precursor of many nucleotide sugars such as GDP-rhamnose and

3

GDP-fucose, and for mannosylation of various bacterial structural components such as lipopolysaccharides and glycoproteins (2,3). PMI also plays an essential role in yeasts and therefore can be regarded as an appropriate target to combat both bacterial and mycotic infections (4).

PMI can be classified into three major types based on sequence similarity and domain organization (5,6). Type I PMIs are found in all eukaryotes including yeast Candida albicans, and certain bacteria including Escherichia coli. Type II PMIs are bifunctional enzymes possessing GDP-mannose pyrophosphorylase (GDP-Man PPase)/PMI dual activities and are primarily found in bacteria. Type III PMIs, first identified in Rhizobium meliloti, are evolutionally more distinct from the other two PMI types. The structural basis on how these seemingly unrelated proteins can catalyze an identical reaction is not clear.

Pseudomonas aeruginosa PAO1 contains three genes encoding a product homologous

with the Type II PMI. These genes, designated algA, wbpW and pslB, are located individually within three distinct polysaccharide biosynthesis gene clusters (3). AlgA and WbpW have been shown to participate in the production of alginate and A-band lipopolysaccharide, respectively (3,7). The gene pslB is located in a 15-gene psl operon required for exopolysaccharide synthesis and biofilm formation (8,9). Whether PslB is indeed a bifunctional GDP-Man PPase/PMI has not been verified.

4

The major aim of this study is to address the structural-functional relationship of type II PMIs that is been largely unexplored. We have cloned, overexpressed, and characterized the PMI activity of PslB. This was followed by site-directed mutagenesis of conserved amino acid to identify important residues participating in the PMI activity.

1.3 Materials and Methods

1.3.1 Bacterial strains and growth conditions

E. coli XL-1 Blue was used for plasmid preparation and recombinant DNA manipulation.

E. coli Nova Blue was used for protein expression. All bacteria used in this study were grown

in Luria-Bertain medium containing ampicillin (100 μg/ml) or kanamycin (50 μg/ml), with 150 rpm shaking at 37^oC. Bacteria strains and plasmids used in this study were listed in Table 1.5.1.

1.3.2 Expression and purification of PslB

E. coli Nova Blue harboring the pslB overexpression plasmid was grown in 200 ml

Luria-Bertani broth supplemented with 50

g/ml kanamycin and 100M isopropyl

thio--D-galactopyranoside (IPTG) at 30^oC with vigorous shaking. After 10 h, cells were

collected by centrifugation at 3100 rpm for 15 min and washed with 20 mM Tris-HCl buffer pH 7.5. The pellets were resuspended in 20 mM Tris-HCl buffer and the bacteria were disrupted by sonication. Soluble proteins were separated from non-soluble proteins by

5

centrifugation at 14000 g for 20 min at 4^oC. PslB was purified by nickel-charged affinity chromatography following the standard purification protocol (Novagen). The eluted protein was dialyzed against 20 mM Tris-HCl buffer (pH 7.5) to remove imidazole and the protein concentration was determined by the Bradford method with bovine serum albumin as a standard.

1.3.3 Determination of the enzyme activity

The PMI activity was determined in a 1-ml reaction mixture containing 50 mM Tris-HCl

pH 7.0, 2 mM CoCl₂, 1 mM Man-6-p, 1 mM -NADP, 5 U phosphoglucose isomerase and 5 U Glc-6-p dehydrogenase. The reduction of -NADP was measured at 340 nm at 25^oC using

a spectrophotometer. The GDP-Man PPase activity was assayed in a reaction mixture containing GDP-Man, 5 mM sodium pyrophosphate, 2 mM MgCl2, 2 mM 3-phosphoglycerate, 20 U glyceraldehyde-3-phosphate dehydrogenase and 10 U phosphoglycerate kinase. The decrease in NADH absorption at 340 nm was monitored by a spectrophotometer (10). Where indicated, 0.5 mM of Man-1-p, GDP-Man, GTP, UDP-Glc, UDP-Gal, ADP-Glc, TDP-Glc, UDP-GlcN, or sodium pyrophosphate was included in the PMI reaction to determine the inhibition activity of these compounds. For determination of the effects of divalent cation, PslB was dialyzed extensively in 20 mM Tris HCl pH 7.5, 140 mM NaCl and 1 mM EDTA. The enzyme activities were determined in the presence of 2 mM

6

chloride form divalent cation including CaCl2, MnCl2, MgCl2, CoCl2, NiCl2 and ZnCl2 (2,11).

1.3.4 Inactivation of PslB by 2, 3-Butanedione

PslB (3.2 mg/ml) was incubated with different concentrations of 2, 3-butanedione (150 mM borate buffer, pH 9.0) at 22^oC in the dark for 0 to 60 min (12,13), then the excess 2,3-butanedione was removed by passing the reaction mixture through a Sephadex G-25 column (GE heahthcare). The PMI activity was measured as described above.

1.3.5 Amino acid sequence alignment and site-directed mutagenesis

The multiple sequences alignment of several type II PMIs, including Xanthomonas

campestris XanB (14), Helicobacter pyroli HP0043 (2), Gluconacetobacter xylinus AceF

(15), and Vibrio vulnificus YJ016 VV0352, were conducted by using the Vector NTI (Invitrogen) and the result is shown in Fig. 1.6.6. Site-specific mutations were performed using the QuikChange site-directed mutagenesis kit purchased from Stratagene.

Oligonucleotide primer sequences are provided in Table 1.5.2.

1.3.6 Circular dichroism spectrum analysis

The CD spectra of wild type and mutant PslB were recorded by using a CD spectrophotometer (AVIV 62A PS) with 1-mm path length cell, 0.5 nm wavelength step, and an averaging time of 3 × 10^-1 s. The protein samples were adjusted to 5

M before

7

measurement. The CD spectra signals were collected from 190 nm to 260 nm in 20 mM Tris-HCl buffer at 25^oC and averaged over three scans (16).

1.3.7 In vivo protein-protein interaction assay by GFP fragment reassembly

The vectors used in this study were provided by Dr. Lynne Regan of Yale University as a gift (17). Gene algC was cloned into pMRBAD-link-CGPF. Three GMP-PMIs (algA, pslB and wbpW) were cloned into pET11a-link-NGFP. The two plasmids were co-transformed into

E. coli BL21 (DE3) and cultured in LB medium containing 100 μM ampicillin and 35 μM

kanamycin at 37^oC for 12 to 16 h. To perform the GFP reassembly, ten microliters of the overnight cultures were applied to the LB agar plates containing the same antibiotics, 100 μM IPTG and 0.05% arabinose. Then, the plates were statically incubated at 20°C for 2 to 3 days.

The bacteria cells grown on the plates were re-suspended with PBS buffer and examined by fluorescence microscope. All images were captured with an upright microscope (BX-51;

Olympus) connected with a CCD camera. The data were analyzed by using the SPOT Advanced Plus Imaging software (Sterling Heights, MI).

1.4 Result and Discussion

1.4.1 PslB possesses PMI and GDP-Man PPase activity

Based on the domain analysis result of InterProScan, the N-terminal domain of AlgA, PslB and WbpW is GMP domain and the C-terminal domain of those enzymes is the PMI

8

domain. Among the three PMI-GMPs in P. aeruginosa PAO1, the GDP-Man PPase/PMI activities of AlgA have been proved (7), whereas that of PslB has not been reported before.

Therefore, we first examined whether PslB is indeed a GDP-Man PPase/PMI bifunctional enzyme. Recombinant PslB was purified from E. coli by nickel affinity chromatography and its purity was confirmed on a SDS-polyacrylamide gel (Fig. 1.6.1). Under the standard assay conditions, both GDP-Man PPase and PMI activities could clearly be detected for PslB. The

K

M value of Man-6-p for the PMI activity was 1.18 mM, and the KM of GDP-Man for GDP-Man PPase was 0.11 mM, both values are in good agreement with that reported for AlgA in P. aeruginosa 8822 (7). Bacteria with a large genome commonly harbor duplicated genes encoding functionally identical enzymes. For example, there are at least two UDP-Glc dehydrogenases in P. aeruginosa PAO1. The enzyme kinetic parameters and expression profiles are somewhat different for these UDP-Glc dehydrogenases that ensure the bacterium can adapt to broader environments (18). A similar mode of regulation may also occur for PMI.

It has been demonstrated previously that a P. aeruginosa strain deficient in WbpW could not produce normal amounts of A-band lipopolysaccharide despite the presence of functional

algA and pslB (3), suggesting that WbpW has distinct expression profiles or kinetic

parameters from PslB and AlgA.

1.4.2 Divalent cation requirement for PslB

9

We measured the PMI and GDP-Man PPase activities of PslB at a fixed concentration (2 mM) of different divalent ions to evaluate cation dependence of the protein. The GDP-Man PPase of PslB could utilize different divalent metal ion with the highest activity in the presence of Mg²⁺, followed by Co²⁺, then Mn²⁺ (Table 1.5.3). The PMI activity of PslB showed high specificity to Co²⁺, yielding about 3-fold higher activity than that observed with Mn²⁺ (Table 1.5.3). Other divalent cations, including Mg²⁺, Zn²⁺, Ca²⁺, and Ni²⁺, could not serve effectively as a cofactor for the PMI activity of PslB, consistent with several previous reports (2,7,19). The optimal Co²⁺ concentration for the PMI was 0.2 mM, which yielded an activity approximately 1.5 fold higher than that of 2.0 mM (Fig. 1.6.2). Mg²⁺ at 2.0 mM showed little interference on the PMI activity in the presence of either 0.2 mM or 2.0 mM Co²⁺

(Fig. 1.6.3), indicating that Co

²⁺ is capable of functioning as a cofactor for the PMI in bacterial cells which commonly contain millimolar Mg²⁺ (20). Interestingly, both PMI and GDP-Man PPase activities in PslB could not be activated in ZnCl2 despite the fact that the protein contains a zinc-binding sequence Q⁴⁰³XH⁴⁰⁵ (X represents any amino acid), a feature that is also present in type I PMIs (6).

1.4.3 The PMI activity of PslB is inhibited by GDP-Man

Several molecules participating in GDP-Man biosynthesis and six nucleotide sugars commonly found in bacteria were tested for their effects on the PMI activity of PslB. At 0.5

10

mM, only GDP-Man significantly inhibited the PMI activity, which decreased to approximately 47% than that observed for the control (Table 1.5.4). Further investigation on the GDP-Man inhibition type demonstrated a competitive inhibition (Fig. 1.6.4) indicating that the GDP-Man biosynthesis pathway is well controlled by the end product, GDP-Man, as also is found for Helicobacter pylori PMI (2). A slight deviation of the 250

M GDP-Man

lines from the other three lower concentration lines was observed suggesting that the inhibition is attenuated in higher concentrations of the inhibitor (GDP-man). The results suggest that the mannose group in GDP-Man competes with Man-6-p at the PMI active site to regulate the mannose utilization in P. aeruginosa PAO1.

1.4.4 An Arg residue participates in the P. aeruginosa PMI activity

To identify if PslB is a useful target to combat P. aeruginosa infections, it is essential to locate the active site participating in the PMI catalytic activity. It has been shown previously that R304 is an active site residue in a type I PMI of C. albicans (21). Although type I and type II PMI share low amino acid sequence homology, their catalytic mechanisms may be similar. To determine whether an Arg is involved in the PMI activity of PslB, this study utilized 2, 3-butanedione to modify Arg in the protein and the PMI activity was determined (22). Figure 1.6.5 shows that the PMI activity decreased with increasing 2, 3-butanedione concentrations and the modification time, revealing that at least one Arg residue in PslB is

11

involved in PMI catalysis.

1.4.5 Arg408 substitutions abolished the PMI activity of PslB

Based on sequence alignment with other type II PMI, R373, R408, R472 and R479 were chosen and changed to Ala by site-directed mutagenesis. In addition, this study adopted the molecular structure information of Pyrobaculum aerophilum bifunctional enzyme phosphoglucose/phosphomannose isomerase (PaPGI/PMI) (23), and selected E410, H411, N433 and E458 for site-directed mutagenesis. The kinetic constants of the wild type and mutant PslB are shown in Table 1.5.5. Substitution of Arg at position 408 to either a Lys or an Ala abolished nearly all of the PMI activity. Other mutant proteins still retained PMI activities and the KM values were 8- to 20- fold larger compared to those values observed for wild type. The CD spectra (Fig. 1.6.7) of R408A, R408K and wild type PslB are nearly

identical, all showing a minimum point at 208 nm and 222 nm, a typical spectrum of high

-helix content protein. The data indicate that there is no major alteration of the secondary

structures of the mutant proteins.

Overall, the results indicate that R408, which lies within the well-conserved motif shared by most, if not all, type II PMI is an important residue for PMI catalysis. The residue is likely to participate in the interconversion of sugar moieties by providing the binding of the sugar phosphate group or forming the hydrogen bond of sugar hydroxyl group, stabilizing the

12

binding between substrates and enzymes.

1.4.6 The R408 mutations did not affect GDP-Man PPase activity of PslB

The fusion of the GDP-Man PPase and PMI domains in the type II PMI implies these two activities may interact with each other. The GDP-Man PPase activity of the PslB deficient in PMI activity was determined and no difference was observed. The KM values of wild type, R408A and R408K were 0.068 mM, 0.072 mM and 0.078 mM, respectively. The similar KM

values of wild type and mutant PslB provide strong evidence that PMI and GDP-Man PPase are actually two independent domains.

1.4.7 The GMP domain and the PMI domain influence the activities of each other

Based on the domain analysis result of InterProScan, the N-terminal domain (9-297 a.a.) of PslB is GMP domain and the C-terminal domain (325-475 a.a.) of PslB is the PMI domain.

Although previous literatures also reported that Type II PMIs own two independent domains (2,6,14,24,25), no article has reported if the PMI and GDP-Man PPase activities would be influenced when Type II PMIs lacked the PMI domain or the GMP domain. To further understand if the separate domains influence the activities of each other, the GMP domain and the PMI domain were cloned into the protein expression vector individually. According to the amino acid sequence alignment result reported by Dr. Leitao J. H. and colleagues (25), the pslB DNA fragment of 1-951 (PslB^1-317) and 1-1086 (PslB^1-362) were cloned into pET30a

13

to constructed a C-terminal His tag fusion protein. To increase the protein solubility, the pslB DNA fragment of 930-1464 (PslB^310-488) was cloned into pGEX-5X-1 to constructed an N-terminal GST tag fusion protein. As shown in Fig. 1.6.8, PslB and PslB^1-362 display the GDP-man PPase activities, but PslB³¹⁷ donot. The KM values (GDP-man) of PslB and PslB^1-362 were 0.089 mM and 0.078 mM, respectively. The similar KM values of PslB and PslB^1-362 indicated that the GDP-man PPase activity would not be influenced when lacking the C-terminal domain from 363-488. Interestingly, GDP-man PPase activity was lost when lacking the fragment residues from 317-362 predicted to be belonged to the PMI domain (25).

Taking those results together, the complete GDP-man PPase activity still required the existence of partial PMI domain. In the case of PMI domain, the time dependent profile (Fig.

1.6.9) displayed that the PMI activity of PslB^310-488 were relatively poor than that of PslB.

The KM

(Man-6-p) values of PslB and PslB

^310-488 were 0.92 mM and 2.89 mM, respectively.

The KM value of PslB^310-488 decreased 3 folds, indicating that the N-terminal domain is also important for the complete PMI activity. PslB^1-362 had no PMI activity, revealing that the C-terminal domain was important for PMI activity. May T. B et al. made the similar conclusion using chymotrypsin digestion to produce a 1-kDa C-terminal deletion of AlgA.

The chymotrypsin digestion of AlgA retained 81% GMP activity but lost 92% PMI activity (24). In summary, PslB was indeed divided into two independent domains, but they influence

14

the activities of each other. This result may explain why the two domains fused together.

1.4.8 PslB cannot form protein complex with AlgC

Many of the bifunctional enzymes have been discovered that they catalyze the continuous reactions in the biosynthesis pathway. For example, tryptophan synthase which is commonly found in bacteria, fungi and plants is a bifunctional enzyme that catalyzes the final two steps in the biosynthesis of tryptophan (26). Some bifunctional enzyme, such as PMI-GMPs, catalyzes the noncontinuous reactions. PMI-GMPs participates the first and third steps of the GDP-mannose biosynthesis pathway (Fru-6-p PslB (PMI) Man-6-p AlgC

(PMM) Man-1-p + GTP PslB (GMP) GDP-man + ppi). It is reasonable to speculate that PMI-GMPs are associated with PMM (AlgC) by protein-protein interaction to form the channeling of Fru-6-p to GDM-man (7,27). If PMI-GMP from a protein complex with PMM, it could increase the reaction efficiency by avoiding the loss of any reaction intermediates. In order to clarify this speculation, three PMI-GMPs (algA, pslB and wbpW) and algC in P.

aeruginosa PAO1 were cloned into pET11a-link-NGFP and pMRBAD-link-CGFP

respectively for performing the GFP fragment reassembly assay. The bacteria cells harboring the two plasmids were examined by a fluorescence microscope. Only the bacteria harboring pET11a-Z-NGFP and pMRBAD-Z-CGPF displayed the green fluorescence emission (28).

None of the bacterial cells containing the PMI-GMP and PMM plasmid pairs emitted

15

fluorescence signals. Our result does not prove if the PMI-GMPs form protein complex with PMM or not.

16

1.5 Tables

Table 1.5.1 Bacterial strains and plasmids used in this study

pET11a-Z-NGFP Ap^r, plasmid vector that expresses a fusion of an antiparallel leucine zipper peptide to NGFP (29) pET11a-link-NGFP Ap^r, plasmid vector designed for fusion of a target protein to the N-terminal fragment of GFP (1-157) (29) pMRBAD-link-CGFP Km^r, plasmid vector designed for fusion of a target protein to the C-terminal fragment of GFP (158-238) (29) pMRBAD-Z-CGFP Km^r, plasmid vector that expresses a fusion of an antiparallel leucine zipper peptide to CGFP (29) pETPslB Km^r, a fragment containing entire pslB (PA2232) coding region cloned into pET30a (NdeI&EcoRI) this study

pHL17 Km^r, pslB with the mutation E410A cloned into pET30a this study

17

pHL30 Km^r, pslB with the mutation E408A cloned into pET30a this study

pHL31 Km^r, pslB with the mutation E408K cloned into pET30a this study

pHL24 Km^r, a fragment containing entire pslB (PA2232) coding region cloned into pET30a (NdeI&HindIII) this study pHL29 Km^r, a fragment containing residues 280-488 of PslB coding region cloned into pET30a this study pHL75 Km^r, a fragment containing residues 1-362 of PslB coding region cloned into pET30a this study pHL76 Km^r, a fragment containing residues 1-317 of PslB coding region cloned into pET30a this study pHL77 Km^r, a fragment containing residues 310-488 of PslB coding region cloned into pET30a this study pHL78 Km^r, a fragment containing residues 1-362 of PslB coding region cloned into pET28a this study pHL79 Km^r, a fragment containing residues 1-317 of PslB coding region cloned into pET28a this study pHL80 Km^r, a fragment containing residues 310-488 of PslB coding region cloned into pET28a this study pHL81 Ap^r, a fragment containing residues 310-488 of PslB coding region cloned into pGEX-5X-1 this study pHL87 Km^r, a fragment containing entire algA (PA3551) coding region cloned into pET30a this study pHL88 Km^r, a fragment containing entire wbpW (PA5452) coding region cloned into pET30a this study pHL89 Km^r, a fragment containing entire algC (PA5322) coding region cloned into pET30a this study pHL90 Ap^r, a fragment containing entire algC (PA5322) coding region cloned into pGEX-5X-1 this study pHL91 Km^r, a fragment containing entire algC (PA5322) coding region cloned into pMRBAD-link-CGFP this study pHL92 Ap^r, a fragment containing entire algA (PA3551) coding region cloned into pET11a-link-NGPF this study pHL93 Ap^r, a fragment containing entire pslB (PA2232) coding region cloned into pET11a-link-NGPF this study

pHL94 Ap^r, a fragment containing entire wbpW (PA5452) coding region cloned into pET11a-link-NGPF this study

18

Table 1.5.2 Primers used in this study

Primer name Primer sequence (5'3') Descriptions

Or488-F1 CATATGAACGCCGTCGCCCCGCTG clone pslBinto pET30a

Or488-R1 AAGCTTGGCTTTCTTCTCGTCGCTGG clone pslB and PslB^280-488 into pET30a

R373A-F2 CACCGCACGGTCAGCGCGCCCTGG pslB- pET30a-R373A mutant

R373A-R2 CCAGGGCGCGCTGACCGTGCGGTG pslB- pET30a-R373A mutant

R408AF1 TGCACCACCATGCCAGCGAGCACTGGAT pslB- pET30a-R408A mutant

R408AR1 ATCCAGTGCTCGCTGGCATGGTGGTGCA pslB- pET30a-R408A mutant

R408KF1 ATGCACCACCATAAGAGCGAGCACTGGAT pslB- pET30a-R408K mutant

R408KR1 ATCCAGTGCTCGCTCTTATGGTGGTGCAT pslB- pET30a-R408K mutant

IsoE410A-F1 CATCGCAGCGCGCACTGGATCGTGGTC pslB- pET30a-E410A mutant

IsoE410A-R1 GACCACGATCCAGTGCGCGCTGCGATG pslB- pET30a-E410A mutant

IsoH411A-F1 CATCGCAGCGAGGCCTGGATCGTGGTC pslB- pET30a-H411A mutant

IsoH411A-R1 GACCACGATCCAGGCCTCGCTGCGATG pslB- pET30a-H411A mutant

IsoN433A-F1 CTCCTCAACACCGCGGAATCCACCTTCATCC pslB- pET30a-N433A mutant

IsoN433A-R1 GGATGAAGGTGGATTCCGCGGTGTTGAGGAG pslB- pET30a-N433A mutant

IsoE458A-F1 GGTGATGATCGCGGTACAGAGCGGCGAG pslB- pET30a-E458A mutant

IsoE458A-R1 CTCGCCGCTCTGTACCGCGATCATCACC pslB- pET30a-E458A mutant

R472A-F1 GGACGACATCGTCGCGTTCAACGACATC pslB- pET30a-R472A mutant

R472A-R1 GATGTCGTTGAACGCGACGATGTCGTCC pslB- pET30a-R472A mutant

R479A-F1 CGACATCTACGGTGCCGCCCCCGCCAGC pslB- pET30a-R479A mutant

R479A-R1 GCTGGCGGGGGCGGCACCGTAGATGTCG pslB- pET30a-R479A mutant

PMI-f3 CATATGAGCGACATCGGCTCCTGGCA clone PslB^280-488 into pET30a

GMP-F TTACATATGAACGCCGTCGCCCCGCTG clone PslB^1-317 and PslB^1-362 into pET vector

19

GMP317NR ATTCTCGAGTCACGAATCGATGTAGCAGTTGC clone PslB^1-317 into pET-28a

GMP317CR AATCTCGAGCGAATCGATGTAGCAGTTGCTG clone PslB^1-317 into pET-30a

GMP362NR AATCTCGAGTCAGTGGCCGCGACGCTTG clone PslB^1-362 into pET-28a

GMP362CR AATCTCGAGGTGGCCGCGACGCTTGAG clone PslB^1-362 into pET-30a

PMI-F ATTCATATGGTCAGCAACTGCTACATCGATTCG clone PslB^310-488 into pET vector and pGEX-5X-1

PMI-NR AATGCGGCCGCTCAGGCTTTCTTCTCGTC clone PslB^310-488 into pET-28a

PMI-CR AATGCGGCCGCGGCTTTCTTCTCGTC clone PslB^310-488 into pET-30a

PMI GST F GCGGATCCACGTCAGCAACTGCTACATCG clone PslB^310-488 into pGEX-5X-1

3551 F GCCATATGATCCCAGTAATCCTTTC clone algA into pET30a

3551 R AATCTCGAGGCGGCTGCCGGCGACC clone algA into pET30a

5452 F AATCATATGCTGATTCCCGTGGTG clone wbpW into pET30a

5452 F AATCATATGCTGATTCCCGTGGTG clone wbpW into pET30a

在文檔中 綠膿桿菌PAO1中兩個雙功能酵素PslB和PA3346功能特性之研究與克雷白氏肺炎桿菌CG43 galU突變株在半乳糖壓力下其轉錄體之分析 (頁 21-0)