CA2346220A1 - Engineered bira for in vitro biotinylation - Google Patents

Description

CA Patent File No. 45424.10 EI\fGINEERED BirA FOR L~V VITRO BIOTINYLATION
Inventors: Sau-Ching Wu, Sui-Lam Wong .A major attraction to use Bacillus subtilis as an expression host for heterologous protein production is its capability to secrete e:Ytracellular proteins into the culture medium. To take full ;advantage of this system, an efficient nnethod of recovering the target protein is crucial. For secretory proteins whi<;h cannot be purified by a simple scheme, in vitro biotinylation using biotin ligase (BirA) offers an effective alternative for their purification.
Availability of large ;mounts of quality Bir.A can be critical for in vitro biotinylation. We report here the engineering ;rnd production of an E. coli BirA and its application in the purification of staphylokinase, a ~Ebrin-specific plasminogen activator, from the culture supernatant of B.
subtilis via in vitro biotinylation. BirA was tagged with both a chitin binding domain and a hexahistidine tail to facilitate both its purification and its removal from the biotinylated sample.
We show in this paper how, in a unique way, we solved the problem of protein aggregation in the E. coli BirA
production system to achieve a yield oj'soluble functional BirA hitherto unreported in the literature. Application of this novel BirA to protein purification via in vitro biotinylation in general will also be discussed. Biotinylated staphylokinase produced in the study not only can act as an intermediate for easy purification; it can also serve as an important element in the creation of a blood clot targeting and dissolving. agent.
"Designer affinity purification" of proteins ( 1 ), a strategy which involves fusing the target protein 'Kith an affinity tag to facilitate its puri:6cation, is an attractive approach for efficient purification of target proteins from a crude preparation. Many affinity tags have been developed for this purpose, among which glutathione S-transferase (GST) (2), P-galactosidase (3), maltose binding protein (4), and biotin acceptor domains (5, 6) are more popular. Since some of these proteins tags are relatively large, it is not uncorrtmon for them to contribute to more than 50% of the molecular mass of the protein fusions and create some undesirable side effects. For example, large tags can occasionally cause solubility problem during protein production and adversely affect the conformation and biological activity of the target proteins (7, 8).
Moreover, use of these tags usually requires post-purification tag removal by chemical or enzymatic means which can be a challenging, time-consuming and costly process and which may not be compatible with the target protein ( 1 ). For these reasons, small tags including His-tag (9), strep-tag ( 10) and biotinylation tag (11, l2) are preferred choices. Strep-tags I and II are short peptides (9 amino acids in length) that can bind selectively to streptavidin (Kd -10' M).
Biotinylation tags, identified through screening of a peptide library (11, 12, 13), are peptide tags (13-15 amino acids) that can be biotinylated enzymatically using the E. coli biotin ligase (BirA). Since the affinity of both his- and strep-tags to their affinity matrices is not particularly high, presence of contaminants is a common problem. This shortcoming can be avoided by the use of the biotinylation tag since biotin binds to monomeric avidin ( 14, 15) or nitro-avidin ( 16) with a much higher affinity ('~d 10' to 10' M). A systematic comparison of the use of these three tags to purify a rat neurotensin receptor expressed in E. coli demonstrates that biotinylation tag provides the highest efficiency and purity ( 17;).
Recombinant proteins with biotinylation tags have commonly been biotinylated in vivo using endogenous BirE~ or co-overproduced BirA (6, 17, 18, 19). This approach may not be desirable for some applications as in vivo biotinylation has a number of drawbacks. First, incomplete biotinylation of the target proteins because of limited cellular resources (BirA and ATP) is a common occurrence (18). Limitation caused by BirA deficiency could sometimes be overcome by coexpressing birA (17); hawever, limitation from depletion of intracellular ATP is more difficult to addre~~s. In the biotinylation reaction, BirA uses ATP to introduce the biotin ;moiety to the lysine residue on the biotinylation tag (20). Thus, all events involved in producing the biotinylated protein fusions in vivo (production of BirA, production of the target protein, enzymatic biotinylation) are highly energy-demanding processes. This may explain why in some eases, coexpression of BirA could only partially improve the biotinylation efficiency with still a ilarge amount of the target proteins remaining unbiotinylated ( 17, 18). A
second drawback with in vivo biotinylation is thc~ presence of endogenous biotinylated proteins. For example, E. coli has one (biotin carboxyl carrier protein or BC:CP, a subunit of acetyl-CoA
carboxylase)(5) and B.

subtilis has two such biotinylated entities (BCCP and pyruvate carboxylase)(21). Although these biotinylated proteins are present as a tiny fraction of the total intracellular proteins, the prospect of having these contaminants in an otherwise highly pure sample is definitely undesirable.
Finally, in vivo biotinyiation has a serious limitation: it cannot be used to effectively biotinylate secreted proteins since both BirA and ,ATP are intracellular. Thus, this technique cannot be used to exploit the full advantages of secretary production of heterologous proteins such as ease of protein recovery, reduced cell toxicity and absence of intracellular biotinylated contaminants as mentioned earlier. It may be desirable to avoid these limitations by carrying out biotinylation in vitro.
For in vitro biotinylation, availability of BirA can be a critical factor.
This is particularly true if large scale protein purification via biotinylation is the goal. In this paper, we describe the production and characterization of an ewgineered BirA which contains a C-terminal His-tag and an N-terminal chitin binding domain (CBD). Both tags facilitate the purification of this engineered version of l3irA. In addition, the N-terminal CBD also allows the rapid removal of BirA from the biotinyl~ation mixture after the completion of the reaction.
This engineered BirA
was applied to biotinylate a secretary recombinant protein (staphylokinase) from Bacillus subtilis. The recombin;~nt staphylokinase was tagged with a biotinylation peptide. We show that, by coupling in vitro biotinylation with subsequent affinity purification using monomeric avidin, ~we have been able to rc;cover a large amount of active staphylokinase in high purity.
Construction of pET-BirA-His lPlasmid pET-BirA-His is an expression vector to produce BirA with a C-terminal hexahistidine ~:ag in E. coli using the T7 promoter system. E coli BirA was amplified by PCR
with E. coli ~;enomic DNA as the template and synthetic oligonucleotides ECBIRAF (S' GGGAATTCTAAGGATAACACC(i7'GCCACTG 3') and ECBIRAB (5' GGAAGCTT
'rTGTTTTTCTGCAC'TACGCAGGG 3') as the forward and backward primers, respectively.
The amplified product carried an EcoRI site at the 5' end and a HindIII site at the 3' end. The ~~70-by fragment was digested with EcoRIiHmdIII and inserted in frame to pET-29b (Novagen, l:lSA) to give pET-Birth-His.

Construction of pET-C',BD-BirA-His This vector allows the production of CBD-BirA-His in E. coli. The gene encoding a chitin binding domain (22) was amplified from the pCYBI plasmid carrying the CBD of chitinase Al (New England BioLabs, Canada) using; the forward primer CBDF (5' CCCATATGACGAC.AAATCCTGCsTGTATCC 3') and the backward primer CBDB (5' CCAGATCTTGAAGCTGCCACAAGGCAGGAAC 3'). The 165-by amplified product was then ~3igested by Ndel and ~iglII and inserted into pET-BirA-His. The resulting plasmid, designated pET-CBD-BirA-His, vvas transformed to E. coli BL21(DE3) (Novagen, USA) for expression ;studies.
Construction of pSAK:PFB
'This is a B. subtilis vector for secretoy production of staphylokinase (SAK) containing a C-terminal biotinylation peptide (PFB). This vector used a strong and constitutively expressed promoter (P43) to drive the transcription and a B. subtilis levansucrase signal peptide to direct ~;he secretion. The biotinylation tag wars fused translationally to secretory staphylokinase in the iFollowing manner. The sequence encoding PFB was fused in frame to the 3'-end of the sak gene in pSAK-K1, a pWB980-based vector in B. subtilis (23), by PCR with pSAK-K1 as the template, 5' CAAGCAACAGTATTAACC 3' as the forward primer and 5' CCAAGCTTATCGATGAT'rCCAAA C',CATTTTTTCJTG CAT
CAAGAATATGATG.AAGGGATCCAGAGCCACTAGTAGATCC 3' as the backward primer.
'Che backward primer encodes a 15-amino acid peptide with the amino acid sequence LHHILDAQKMV WNHR ( 1 I ). The amplified 578-by fragment was digested with HindIII and used to replace an equivalent fragment from HindIII digested pSAK-KI. The resulting plasmid, designated pSAKPFB, was transformed to B. subtilis WB800, an eight-protease deficient strain (;24) and the transformamts screened for the right orientation of the insert.
(ell growth E. coli BL21 [pET-CB:D-BirA-His] was grown at 30°C', in Luria broth ( 1 % tryptone, 0.5% yeast extract, 0.2% NaCI) containing 30 ug/ml kanamycin to 150 klett units in a shake flask. IPTG was then added to a final concentration of 0.1 mM and growth continued for 5-10 hours. Cell density was measured using a Klett-Summerson photoelectric colorimeter with a green filter (Klett Mfg.
Co., USA). B. subtilis WB800[pSAKfFB] was cultivated in super-rich medium (25) containing 10 pg/ml of kanamycin at 37°C in a shake flask. Cells were harvested at 5-6 fours after inoculation.
Purification of BirA
Cells of E. coli BL21 [pET-CBD-BirE1-His] were harvested by centrifugation at 10,000 x g for 5 min at 4°C. Cell pellet was resuspended in lysing buffer, disrupted with French press and the crude lysate was separ;~ted into the sohable and insoluble fractions by centrifugation (20,000 x g for 20 min). CBD-Bir~~-His in the soluble fraction was purified by either of two schemes: metal ~:.helation chromatography or chitin affinity chromatography. For metal chelation ~~hromatography, the lysing buffer contained 15 mM imidazole, O.SM NaCI, 0.1%
Triton X, 1 ~mM phenylmethylsulfimyl fluoride (PIvISF), and 20 mM Tris-HCI, pH 8Ø
His.Bind Quick 900 partridges (Novagen, LISA) charged with NiZ+ were used as the affinity matrix.
CBD-BirA-His was eluted stepwise with increasing imidazole concentrations (60 mM, 250 mM, 1 M imidazole) ;according to the manui:acturer's suggestions. For chitin afflinity chromatography, cells were llysed in buffer containing 1 M NaCI, 1 mM EDTA, 0.1 % Triton X, 5 mM (3-mercaptoethanol, 1 mM PMSF and 20 mNf sodium phosphate, pH 7Ø The soluble cellular fraction was loaded to a column packed with chitin beads (New England BioLabs, Canada) equilibrated in lysing buffer.
,After washing with 5-10 column volumes of lysing buffer followed by 20 mM
sodium acetate, pH 5.5, CBD-BirA-His was eluted with 20 mM acetic acid, pH 3Ø
In either scheme, fractions containing pure CBD-BirA-His (confirmed by SDS-PAGE) were pooled, concentrated and buffer-changed to a storage solution containing 50 mM imidazole, :p0 mM NaCI, 5% glycerol and 5 mM ~S-mercaptoethanol, pH 6.8, using Ultrafree-centrifugation tubes (Millipore Corporation, USA). Pure CBD-BirA-His was quantified by its absorbance at 280 run using a molar e:Ktinction coefficient of 68420 M -' cm '' (26).
Purification of SAK-PFB
Culture supernatant of B. subtilis WB800[pSAKPFB) was separated from the cells by centrifugation at 10,000 x g for 10 min at 4°C. SAK-PFB was precipitated with ammonium sulfate to 65% saturation at 4°C, desalted by dialysis, concentrated to appropriate volume and buffer-changed to 10 rnM Tris-HC 1, pH 8.0, using Ultrafree-4 centrifugation tubes (Millipore Corporation, USA). The sample was then biotinylated at 30°C for 4 hours to overnight, using CBD-BirA-His. The reaction mixture contained 50 mM bicine (pH 8.3), 10 mM ATP, 10 mM
magnesium acetate, SG uM biotin, and for every ml of final mix, 500 ug of SAK-PFB and 5 ug of purified E. coli CBD-BirA-His. following the reaction, the sample was mixed with a small amount of chitin bead:. to remove CBD-BirA-His. After a simple centrifugation to remove the chitin beads, the sample was passed over a column containing Sephadex G-25 (Amersham Pharmacia Biotech, Canada) to remove the excess biotin. Biotinylated SAK-PFB
was separated the unbiotinylated prol:eins by passing the sample over a monomeric avidin agarose column {Pierce, USA). Bound biotinylated SAK-PFB was eluted by competition with 2 mM
d-biotin.
Pure biotinylated SAK-PFB, was quantified by its absorbance at 280 nm using a molar extinction .coefficient of 22,900 nM'crri' (26) for calculation.
Determination of the activity of purified CBD-BirA-His 'The activity of purified CBD-BirA-1-its was compared with that of a wild type E. coli BirA
available from a commercial source (Avidity, USA) using an ELISA method (13).
In this assay, maltose binding protein-AviTag fusion (MBP-AviTag, Avidity, USA) was used as the substrate.
.AviTag is a peptide tal; for efficient biotinylation ( 11 ). MBP-AviTag was adsorbed to the wells of a Reacti-bind malefic: anhydride activated polystyrene strip plate (Pierce, USA). Biotinylation was carried out at 30°(' with different amounts of enzymes and different reaction times. The reaction mixture contaiined 50 mM bicine (pH 8.3), 10 mM ATP, 10 mM magnesium acetate, 50 uM biotin and BirA from different sources. Biotin ligated to the AviTag was detected by its interaction with streptavidin-horseradish peroxidase (Pierce, USA) using 1 step slow TMB
ELISA (3,3',5,5'-tetrarnethylbenzidine., Pierce) as the color development reagent. A standard curve of biotinylation reaction was established using known quantities of fully biotinylated MBP-AviTag (Avidity, USA). Readings were taken at end point at 450 nm using a Bio-Tek CERES 900 plate reader (Bio-Tek Instruments, Inc., LISA).
Matrix-assisted laser desorption ionization time-of flight (MALDI-TOF) mass spectrometric analyses Protein samples in 25 mM ammonium acetate and the matrix solution of sinapinic acid were mixed on the MALDI plate and analyzed on a Perseptive Biosystems-(Framingham Mass.) Voyager-DE STR Mass spectrometer equipped with a pulsed nitrogen laser operated at 337 nm in a linear mode. The mass spectrometer was previously calibrated with apomyoglobin (horse skeletal) m/z 16952.56 and its dimer rr~/z 33905.12. These analyses were done in Plant Biotechnology Institute, National Research Council of Canada, Saskatoon, Canada.
Other methods Vent DNA polymerase: (New England BioLabs, Canada) was used for all DNA
amplification reactions. The sequence of all PCR products was confirmed to be free of PCR
errors by nucleotide sequencing based on the dideoxy method using a T7 sequencing kit from Amersham lPharmacia Biotech, Canada. SDS-polyacrylamide gel electrophoresis followed standard procedure based on the Laenimli system. Western blot was done on a nitrocellulose membrane using 4-chloro- 1 -naphthol (Bio-Rad, Canada) as the color development reagent. SAK activity was determined by radial caseinolysis assay on plasminogen-skim milk agarose plate (27).
Results Production of E. coli CBD-BirA-His using the pET expression system.

In this study, an IPT(r-induced expression of BirA in a pE'T-29b based vector was used for intracellular producaion of E. coli (:BD-BirA-His in BL21(DE3). CBD-BirA-His (with both CBD- and His-tags) produced migrated as a 40-kDa protein on the SDS gel (Fig.
1). The presence of the His-to~; was found to complicate the production because, at a growth temperature of 30°C, 90% of CBD-BirA-His accunnulated as inclusion bodies (Fig. 1A, lanes 1 and 2). In contrast, over 80% of CBD-BirA (no His tag) produced under the same cultivation condition was in the soluble form (data not shown). 'l'o address the solubility problem, different measures were taken. These include lowering the growth temperature from 30°C
downwards, lowering the salt .concentration in the culture medium, varying the IPTC~ levels, and modifying the cellular .osmotic environment with the use of sorbitol and betaine during cell growth (28). These measures at best yielded marginal improvement with still more than 70% of CBD-BirA-His present as insoluble aggregates. However, supplementing the culture medium with 10-20 uM
biotin not only enhanced the growth rate of the culture (data not shown) but also conspicuously vpromoted solubility of CBD-BirA-His with about 40% of the protein in the soluble fraction (Fig.
1 A, lanes 3 and 4; Fig. 1 B, lanes 1 and 2). Moreover, whereas temperature lowering by itself ~iid not effectively solve the problem of inclusion body formation, a measure combining biotin supplementation and temperature lowering (25°C, post-induction) enhanced BirA solubility significantly. Typically, 70-90% of BirA produced under this condition was in the soluble form ( Fig 1B, lanes 3 and 4), amounting to about 100 mg of soluble CBD-BirA-His per liter of culture.
l~,ngineered E. coli Bir,A could be purified with simple manipulations CBD-BirA-His was equipped with two tags: a 6-amino-acid histidine tag preceded by an 8-amino-acid linker and ;~ 53-amino-acid chitin binding domain followed by an 18-amino-acid linker. These tags allow rapid purification of the protein by either scheme:
metal chelation or chitin affinity chromatography. Fig. '2A shows the purification of CBD-BirA-His using a Niz+
chelation column. CBI)-BirA-His bound to the column effectively with essentially no loss in the slow-through fractions (lane 2). Reasonably pure fractions (over 80% purity) were recovered by elution with imidazole (lanes 3-5). C..'BD-BirA-His in these fractions could be further purified to over 95% purity by repeatedly reloading the purified CBD-BirA-His to the Ni2+
chelation column. The chitin affinity scheme was more efficient. CBD-BirA-His bound to the chitin column with high affinity and great specificity with no CBD-BirA-His detectable in the flow-through and washes (Fig. 2B, lanes 2 and 3). A single-column operation was usually adequate to recover CBD-BirA-His with over 95%~ purity (Fig. 2B, lane 4). Chitin affinity chromatography, however, has a major drawback. About 40-50% of the CBD-BirA-His tended to be retained on the column and could not be recovered) even with extensive washes and elutions at low pH.
Despite this drawback. we have been able to recover 1.5-2 mg of highly pure CBD-BirA-His from 100 ml of shake flask culture using the chitin column, representing an overall recovery yield of 15-20%. The recovery rate with the metal chelation scheme (involving three cycles of Ni' chelation column) ro purify CBD-BirA-His with over 95% purity is similar.
Purified engineered BirA demonstrated high biological activity Activity of CBD-BirA-His was determined by its ability to biotinylate maltose binding protein 'tagged with a short biotinylation peptide designated AviTag (I 1) in an ELISA
study. With vunbiotinylated MBP-A.viTag as the substrate using parameters (amount of enzyme used and reaction time) that ensured a linear rate of enzymatic reaction, the activity of CBD-BirA-His purified from either scheme was found to be 50% more active than that of the natural E. coli BirA from a commercial source (Table I ).
'Table I . Activity of BirA from different sources Source of BirA Specific ActivityRelative Activity ~

Metal chelation'40.2 1.48 Chitin affinity242.7 I .58 Commercial3 27.1 1 Activity of BirA was determined by ELISA method (13) using unbiotinylated MBP-AviTag (Avidity, USA) as the substrate. Specific activity of BirA is defined as ng biotinylated MBP-AviTag formed per min per tzg of enzyme at 30°C. 'CBD-BirA-His purified by metal chelation chromatography. ''CBD-BirA-His purified by chitin affinity scheme. 3Wild type E. coli BirA
obtained from a commercial supplier (Avidity, I1SA). Data represent the average of two independent trials.
This shows that the BirA engineered, produced and purified using our purification scheme is of high quality. The presence of His-tag has little effect on the biological activity of the purified enzyme as CBD-BirA and CBD-BirA-flis exhibited similar specific activities on biotinylation of MBP-AviTag (data not shown). 'The readiness of CBD-BirA-His to biotinylate proteins with a biotinylation tag was also demonstrated in a Western blot analysis (Fig. 3).
Two test proteins were used as examples: MBP-AviTag and staphylokinase tagged with another biotinylation tag designated PFB. Probing with streptavidin-horseradish peroxidase showed biotinylation of both proteins with BirA (Fi;~. 3B, lanes 1 arid 2).
Engineered BirA is active in a fairly broad pH range 'The pH activity profile of CBD-BirA-1-Iis was established with an ELISA study similar to the one used for the determination of its biotinylation activity. Different reagents were used to provide buffering capacity for a broad pH range (see legend to Fig. 4). MBP-AviTag was used as 'the substrate. Fig. 4 shows that CBD-I?~irA-His had a pH optimum around 6.5.
It retained a fairly thigh activity at pH 5.5-8.3, but the activity dropped substantially at either ends. This information would be useful for one to tailor an optimal condition for in vitro biotinylation with this enzyme.
'To our knowledge, the pH activity proitile of natural E. call BirA has not been systematically studied before.
Secretory production of staphylokinase-PFB from B. subtilis 'to explore the possibility of purifying a secretory fusion protein carrying a biotinylation tag i~rom a B. subtilis culture supernatant via in vitro biotinylation using the engineered BirA, staphylokinase (SAK), a very promising blood clot dissolving agent (29), was used as a model system. A 15-amino-acid biotinylation tag (PFB) was added to the C-terminal end of SAK
containing an 18-amino-acid C-terminal linker sequence [(GSTSG)3SGS]. Addition of the linker and the biotinylation tag did not affect the secretory production yield of SAK-PFB since SAK
with or without PFB was produced at ;~ comparable level (Fig. 5A, lanes 1 and 2). When analyzed by SDS-PACiE, SAK-PFB showed an apparent molecular mass of 21 kDa.
The calculated molecular mass of SAK-PFB is 18,862 Da. To confirm that the intact form of SAK-PFB was produced from B. subtilis, the molecular mass of SAK-PFB was determined by MALDI-TOF mass spectrometry. The observed molecular mass matched closely with the expected value and was determined to be 18,861.22 Da (data not shown).
Functional SAK-PFB could be purified via in vitro biotinylation using the engineered BirA
After concentrated from the culture supernatant, SAK-PFB was biotinylated in vitro using purified CBD-BirA-His. The rate of biotinylation depends, among other variables, on the amount of enzyme used for the reaction. As SAK-PFB is fairly stable, biotinylation could be carried out using varying amounts of enzyme from several hours to overnight with no apparent adverse effect. Biotinylated SA.K-PFB, with an apparent molecular mass of 21.5 kDa on the SDS gel, emigrated more slowly than its unbiotinylated counterpart (Fig. 5A, lane 3 vs.
lane 2, Fig. 6, lane 2 vs. lane 1). This allows us to easily monitor the extent of biotinylation.
In all biotinylation runs ;attempted so far, over !~5% biotinylation of SAK-PFB could be achieved as demonstrated by the s~bsence of any significant amount of SAK-PFB in the flow-through or washes of the monomeric ;~vidin agarose column (Fig. 6, lanes 3 and 4). The completion of biotinylation was also remonstrated by the M:ALDI-TOF mass spectrometric analysis. The peak with the expected molecular mass corresponding to the unbiotinylated form of SAK-PFB disappeared completely i.n the biotinylated sample while a new peak with the expected molecular mass corresponding to l:he biotinylated form appeared (data not shown). Biotinylated SAK-PFB could be effectively purified using a monorneric avidin agarose column with remarkable specificity (Fig. 6, lanes :>-7). We have been able to recover about 450 ~g of highly pure SAKPFB from a crude sample containing 600 ug of SAK-PFB on a single column, representing an overall yield of 75%. SAK-1?FB purified by this method showed full biological activity as compared with both the unbiotinylated form and the natural, untagged SAK on a plasminogen assay ml of B. subtilis culture.
Discussion 'To capture the full advantages of in vitro biotinylation, a ready source of easily purified, high quality BirA is needed. In this study, we addressed this concern by engineering an E. coli BirA
with a different tag at each end (CBD-l~irA-His). These tags enable easy recovery of the protein by simple column manipulations. Use of the His-tag allows a one-step recovery of large amounts of reasonably pure Bir.A, while use of t:he CBD enables, again, a single-column recovery of a llesser quantity of ultrapure BirA. These two grades of BirA can be found useful in different applications. For example, reasonably pure BirA can be used to biotinylate a crude extract (such as the secreted fraction) as other contaminants can be removed later via the monomeric avidin step. On the other hand, ultrapure BirA. is critical in the biotinylation of pure proteins (such as affinity-purified single chain antibodies). Besides the tag advantage, the production yield and quality of our engineered BirA compare favourably with the literature data. By supplementing t:he medium with biotin and lowering the post-induction temperature to 25°C, the soluble CBD-BirA-His reached a level of 100 nng per liter of culture. This level is double the amount of GST-BirA reported previously (30). Moreover, the specific activity of CBD-BirAHis was found to be more than that of the natural BirA from a commercial source. In one study involving GST-BirA (30), thrombin was applied to cleave off GST from the fusion and the resulting BirA
showed a comparable activity similar to that of the wild type BirA. In another case (19), esST-BirA, used uncleaved, was shown to retain biotin ligase activity but the specific activities of the fused and non-fused versions were not studied.
Several interesting and important observations were made during the development of the engineered BirA. First, supplementation of biotin in the culture medium could help reduce the formation of inclusion bodies. Biotin was commonly included in the culture medium in in vivo biotinylation studies involving the E. c~li system since E. coli has been shown to uptake biotin via an active transport mechanism (31 ). In those studies, biotin served mainly as one of the substrates for BirA in the biotinylation reaction. Our observation in this work suggests that being a substrate, biotin can also possibly enhance the proper folding of BirA in favour of soluble protein formation. Second, presence of small tags at both ends of BirA does not materially affect the biological activity of BirA as a biotin ligase. We designed two small affinity tags for the BirA: a 53-amino acid chitin binding domain and a 6-amino acid His-tag. The engineered BirA, used as such, demonstrated a higher specific activity than that of the natural BirA (from a commercial source). This shows that the engineered BirA retained good biological activity through the purification procedure and, unlike some large tags, can be used uncleaved. Third, although CBD-BirA could be produced as a soluble enzyme in large quantities, addition of a short C-terminal His-tag severely reversed the situation with the problematic formation of inclusion aggregates. This shows that t:he use of small tags does not guarantee that the system will work as expected. Even if the tags do not affect biological activity of the target protein, .complications like prooein insolubility during production can arise and have to be addressed .accordingly.
'Two interesting observations were also made during the purification of the biotinylated proteins.
Occasionally, we detected a biotin-Bir.A complex in Western blot probed with ;~treptavidin-horseradish peroxidase even though the sample had been boiled in the presence of SDS before loading to the SDS-polyacrylamide gel. This complex is likely to be the tight entity i;Kd = 7 x 10-") formed between BirA .and biotinoyl-5'-AMP, an intermediate in the biotinylation reaction carried out by BirA (32). The presence of this complex means that postbiotinylation removal of BirA is necessary not only when pure target protein is involved but also when crude sample is used for biotinylation. The installation of the N-terminal CBD in CBDBirA-His allows rapid removal of BirA by the use of chitin beads. In the purification of SAKPFB, CBD-BirA-His was removed by chitin bead treatment in a simple centrifugation step to avoid the potential problem of contamination. Thus, the tags on CBD-BirA-His facilitate not only purification of CBD-BirA-His but also removal of C'.I?~D-BirA-His from the postbiotinylation reaction mixture.
Another interesting observation is that the biotinylated protein exhibited a small mobility shift on the SDS gel. This has a practical application for the biotinylation of small target proteins as one may be able to monitor the extent of biotinylation, easily by SDS-PAGE. This method worked well for SAK-PFB wil:h a molecular mass of 19 kDa.
In vitro biotinylation offers a general tool to affinity purify secretory proteins not only from E. coli but also from other organiisms such as B. subtilis. This approach is most valuable for the purification of proi:eins (e.g. staphylokinase) which cannot be recovered by other affinity purification methods and which require multiple chromatographic steps for their purification. As demonstrated in this study, addition of the biotinylation tag to staphylokinase affected neither the production yield nor the biological activity of staphylokinase and intact SAK-PFB could be produced as confirmed by mass spectrometric analysis. This system works best when the target protein has a high-level expression, thc~ fusion is stable, and protease activity is absent. The high efficiency biotinylation achieved with our SAK-PFB study may be attributed to the remarkable secretory yield of SAK; in B. subtilis (over 100 mg/I in a shake flask) (3 3), the stability of SAK-PFB, and the use of an eight-protease deficient strain which has been shown to dramatically enhance the yield (24) and stability (unpublished data) of some secretory proteins in B. subtilis.
The high efficiency biotinylation, coupled with the high capacity of monomeric avidin with its exceptional affinity and specificity to biotin, contributes to a remarkable recovery of quantitative .amounts of distinctly pure staphylokinase. This approach can be applied to other secretory proteins from B. subtilis.
Besides protein purification, the homogeneous biotinylated products made possible by the 1'aighly selective, site-specific action of ('.BD-BirA-His on the biotinylation tag offers many other ;applications. They serve as agents in immunoassays, drug delivery, imaging and targeting (34, 3 :>, 36, 37). Biotinylated proteins can also be immobilized in an orientation-specific manner (38) to generate protein or antibody biochips for surface plasmon resonance based biosensor measurements (39, 40), active electronic microchips for biomolecule detection and quantification I 41 ), and high density protein microarrays for high throughput proteomics studies (42).

References:
The following references are incorporated herein as if reproduced in their entirety.
1. Sharma, S. K. (1997) Designer affinity purifications of recombinant proteins in "Affinity Separations: a practical approach" (Matejtschuk, P., Ed.), pp. 197-218, IRL
Press, Oxford.
2. Smith, D. B., and Johnson, K. .S. (1988) Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transferase. Gene 67, 31-40.

3. Hirel, P. H., Le:veque, F., Mellot, P., Dardel, F., Panvert, M., Mechulain, Y., and Fayat, G. (1988) Genf;tic engineering of methionyl-tRNA synthetase: in vitro regeneration of an active synthetase by proteolytic; cleavage of a methionyl-tRNA synthetase-beta-galactosidase chimeric protein. Biochimie 70, 773-782.

4. Sachdev, D., and Chirgwin, J. A (2000) Fusions to maltose-binding protein:
control of folding and solubility in protein purification. Methods EnzymoL 326:312-321.

5. Cronan, J. E., Jr. (1990) Biotination of proteins in vivo: a post-translational modification to label, purify and study proteins. J.BioI~ Chem. 265, 10327-10333.

6. Cronan, J. E., Jr., and Reed, K. E. (2000) Biotinylation of proteins in vivo: a useful posttranslational modification for protein analysis. Methods EnzymoL 326, 440-458.

7. Yu, L., Deng, ><:., and Yu, C. A. (1995) Cloning, gene sequencing, and expression of the small molecular mass ubiquinone-binding protein of mitochondrial ubiquinol-cytochrome c reductase. JBi.oLChem. 270,25634-25638.

8. Corchero, J. L.., Viaplana, E., H'~enito, A., and Villaverde, A. (1996) The position of the heterologous domain can influence the solubility and proteolysis of beta-galactosidase fusion proteins in E. coli. J Biotechnol 48, 191-200.

9. Bornhorst, I A., and Falke, J. J. (2000) Purification of proteins using polyhistidine affinity tags. Methods Enzymol 326, 245-254.

10. Skerra, A., and Schmidt, T. U. (2000) Use of the Strep-Tag and streptavidin for detection and purification of recombinant proteins. Methods Enzymol 326, 271-304.

11. Schatz, P. I (1~~93) Use of peptide libraries to map the substrate specificity of a peptide-modifying enzyme: a 13 residue consensus peptide specifies biotinylation in Escherichia coli. Biotechnology (N. Y.) 1 l, 1 13 8-1143.

12. Beckett, D., Kovaleva, E., and Schatz, P. 1 (1999) A minimal peptide substrate in biotin holoenzyme sy:nthetase-catalyzed biotinylation. Protein Sci. 8, 921-929.

13. Cull, M. G., and Schatz, P. J. (2000) Biotinylation of proteins in vivo and in vitro using small peptide tags. Methods Enzymol. 326, 430-440.

14. Henrikson, K. P., Allen, S. H., and Maloy, W. L. (1979) An avidin monomer affinity column for the purification ok~b~iotin-containing enzymes. Anal.Biochem. 94, 366-370.

15. Kohanski, R. ~~., and Lane, M. D. (1990) Monovalent avidin affinity columns. Methods Enzymol. 184,194-200.

16. Morag, E., Bayer, E. A., and V~lilchek, M. (1996) Immobilized nitro-avidin andnitro-streptavidin as :reusable affinity matrices for application in avidin-biotin technology.
Anal.Biochem. 243, 257-263.

17. Tucker, J., and Grisshammer, R. ( 1996) Purification of a rat neurotensin receptor expressed in Escherichia coli. Biochem.J 317, 891-899.

18. Chapman-Smith, A., Turner, L). L., Cronan, J. E., Jr., Morris, T. W., and Wallace, J. C.
( 1994) Expression, biotinylation and purification of a biotin-domain peptide from the biotin carboxyl carrier protein of Escherichia coli acetyl-CoA carboxylase.
Biochem.J
302, 881-887.

19. Saviranta, P., Haavisto, T., Ra~ppu, P., Karp, M., and I,ovgren, T. (1998) In vitro enzymatic biotinylation of recombinant Fab fragments through a peptide acceptor tail.
Bioconjug. Chem. 9,725-735.

20. Chapman-Smith, A., and Cronan, J. E., Jr. ( 1999) The enzymatic biotinylation of proteins: a post-translational modification of exceptional specificity. Trends Biochem.Sci.
24, 359363.

21. Marini, P., Li, S. J., Gardiol, D., Cronan, J. E., Jr., and De Mendoza, D.
(1995) The genes encoding the biotin carboxyl carrier protein and biotin carboxylase subunits of Bacillus subtilis acetyl coenzynne a carboxylase, the first enzyme of fatty acid synthesis.
J.Bacteriol. 177, 7003-7006.

22. Watanabe, T., Ito, Y., Yamada, I'., Hashimoto, M., Sekine, S., and Tanaka, H. (1994) The roles of the; C-terminal domain and type III domains of chitinase A1 from Bacillus circulans WL- l2 in chitin degradation. J.Bacteriol. 176, 4465-4472.

23. Wu, S.-C., and Wong, SA. ( 1999) Development of improved pUB 110-based vectors for expression and secretion studies in Bacillus subtilis. J.Biotechnol. 72, 185-195.

24. Wit, S.-C., Yeung, J. C., Szarka, S. J., and Wong, S.-L. (2000) Functional production and characterization of a fibrin specific single-chain antibody fragment from Bacillus subtilis. 11th International conference of antibody engineering, San Diego, California.

25. Halting, S. M... Sanchez-Anzaldo, F. J., Fukuda, R., Doi, R. H., and Meares, C. F. ( 1977) Zinc is associated with the beta subunit of DNA dependent RNA polymerase of Bacillus subtilis. Biochemistry 16, 2880-2884.

26. Gill, S. C., and Von Hippel, P. H. (1989) Calculation of protein extinction coefficients from amino acid sequence data. Anal.Biochem. 182, 319-326.

27. Wu, X-C., Ye, R., Duan, Y., and Wong, S.-L. (1998) Engineering of plasmin-resistant forms of streptokinase and thei.° production in Bacillus subtilis:
streptokinase with longer functional half life. Appl.Environ.Microbiol. 64, 824-829.

28. Blackwell, J. P:., and Horgan, R. ( 199 1 ) A novel strategy for production of a highly expressed recombinant protein in an active form. FEBS Lett. 295, 10-12.

29. Collen, D. (19!8) Staphylokinase: a potent, uniquely fibrin-selective thrombolytic agent.
Nat.Med. 4,279-284.

30. O'dallaghan, C. A., Byford, M. F., Wyer, J. R., Willcox, B. E., Jakobsen, B. K., Mcmichael, A. J., and Bell, J. 1. (1999) BirA enzyme: production and application in the study of membrane receptor-ligand interactions by site-specific biotinylation.
Anal.Biochem. 266, 9-15.

31. Piffeteau, A., and Gaudry, M. (1985) Biotin uptake: influx, efflux and countertransport in Escherichia c;oli K 12. Biochem.Biophy.Acta 816, 77-82.

32. Xu, Y., and Beckett, D. (1994;1 Kinetics of biotinyl-5'-adenylate synthesis catalyzed by the Escherichia coli repressor of biotin biosynthesis and the stability of the enzyme-product complex. Biochemistry 33, 7354-7360.

33. Ye, R., Kim, J. H., Kim, B. G., Szarka, S., Sihota, S., and Wong, S.-L.
(1999) High-level secretory production of intact, lbiologically active staphylokinase from Bacillus subtilis.
Biotechnol.Bioeng. 62, 87-96.

34. Wilchek, M., and Bayer, E. A. ( 1988) The avidin-biotin complex in bioanalytical applications. Anal.Biochem. 171, 1-32.

35. Yao, Z., Zhang, M., Kobayashi, H., Sakahara, H., Nakada, H., Yamashina, L, and Konishi, J. (19!5) Improved targeting of radiolabeled streptavidin in tumors pretargeted with biotinylatc~d monoclonal antibodies through an avidin chase Improved targeting of radiolabeled streptavidin in tumors pretargeted with biotinylated monoclonal antibodies through an avidin chase. J.Nucl..Med. 36, 837-84 1.

36. Ohno, K., Levin, B., and Meruelo, D. (1996) Cell-specific, multidrug delivery system using streptavidin-protein A fusion protein. Biochem.Md.Med. 58, 227-23 3.

37. Smith, I S., Ke:ller, J. R., Lohrey, N. C., McCauslin, C. S., Ortiz, M., Cowan, K., and Spence, S. E. (:1999) Redirected infection of directly biotinylated recombinant adenovirus vectors through cell surface receptors and antigens. Proc.Natl.Acad.Sci.U.S.A.
96, 8855-8860.

38. Turkova, J. (1999) Oriented immobilization of biologically active proteins as a tool for revealing proteiin interactions and function. J Chromatogr.B Biomed.Sci.AppL
722, 11-3 1.

39. Myszka, D. G. (1997) Kinetic analysis of macrornolecular interactions using surface plasmon reson~~nce biosensors. ('urr. Opin.Biotechnol. 8, 50-57.

40. Schultz, J., Lin, Y., Sanderson, J., Zuo, Y., Stone, D., Mallett, R., Wilbert, S., and Axworthy, D. (2000) A tetravalent single-chain antibody-streptavidin fusion protein for pretargeted lymphoma therapy.. Cancer Res. 60, 6663-6669.

41. Ewalt, K. L., Haigis, R. W., Rooney, R., Ackley, D., and Krihak, M. (2001) Detection of Biological Toxins on an Active Electronic Microchip. Anal.Biochem. 289, 162-172.

42. Cahill, D. J. (2;001 ) Protein and antibody arrays and their medical applications.
J.Immunol.Methods 250, 81-9 1.

Figure legends Fig. I
Effects of (A) biotin and (B) growth temperature on the distribution of CBD-BirA-His in the intracellular fractions of E coli. (A) Cultures were grown at 30°C
throughout. Lanes 1 and 2: no biotin added to culture medium. Lanes 3 and 4: biotin added at 12 uM to culture medium. (B) Growth medium contained 12 uM biotin for all samples. Lanes I and 2: cultures grown at 30°C
throughout. Lanes 3-6: cultures grown at 30°C and shifted to 25°C post IPTG induction. Lanes 1-4: E. coli BL21(DE3;)[pET-CBD-Bir.A-His]; Lanes 5 and 6: negative control, E. coli BL21 (DE3)[pET29b]. Samples were analyzed on a 10% SDS-polyacrylamide gel and stained by Coomassie blue. M: molecular weight marker; S: soluble fraction; I: insoluble fraction. Arrow indicates CBD-BirA-His.
Fig. 2 :Purification of E. coli CBD-BirA-His using (A) Ni2+ column and (B) chitin affinity column. (A) '.Lane 1: crude lysate. Lane 2: column flow-through. Lanes 3: eluates (60 mM
imidazole). Lane 4:
eluate (250 mM imida;~ole). Lane 5: elvate (yM imidazole). (B) Lane 1: crude lysate. Lane 2:
~~olumn flow-through. :Lane 3: pooled washes (2-column volumes). Lane 4:
eluate. Samples were analyzed on a 10% SDS-polyacrylamide gel. M: molecular weight marker. Arrow indicates CBD-BirA-His.
l~ ig. 3 1?rotein biotinylation using CBD-BirA-flis purified on chitin affinity column, (A) Coomassie blue-stained gel. (B) Vfestern blot probed with streptavidin-horseradish peroxidase using ~l-chloro- 1 -naphthol (Bio-Rad, Canada) as the color development reagent.
Samples were analyzed on a 12% SDS gel. M: molecular weight marker. Lane I : MBP-AviTag (Avidity) as the substrate; Lane 2: SAK: with 15-mer biotinylation peptide tag as the substrate. Biotinylation reaction was carried ou.t at 30°C for ? l~crurs using 1 ~g of the substrate, 50 ng of CBD-BirA-His and other components :in the reaction mixture as described in Materials and Methods.

Fig. 4 pH profile of engineered E coli CBD-BirA-His. 100 ng of unbiotinylated MBP-AviTag (Avidity, USA) was coated on the wellls of Reacti-bind malefic: anhydride activated polystyrene strip plate (Pierce, US~~) to act as the substrate. Reaction mixture contained 10 mM ATP, 10 mM
magnesium acetate, 50~ uM biotin and 10 ng CBD-BirA-His purified by chitin affinity chromatography. Biotinylation reaction was carried out at 30°C for 20 min. Bound biotin was detected by streptavidin-horseradish pf~roxidase (Pierce) with 1-step slow TMB-ELISA (Pierce) as the color development reagent. The following buffers were used at 50 MM to provide buffering capacity for ;~ pH range 2.5 - I 1: glycine (2.5), NaAc (4.5), MES
(5.5), BIS-TRIS (6.5), TRIS-HCl (7.5), bicinE; (8.3, 9), CAPS ( 10, 11 ). Data represent the average of three independent trials.
Fig. 5 ~Staphylokinase activity as determined by the radial caseinolysis assay. (A) Coomassie blue-;~tained SDS gel showing SAK produced by B. subtilis. Amounts of samples loaded on the lanes 'were normalized to cell density. M: molecular weight marker. Lane 1: natural, untagged SAK
iproduced by WB800[pSAKP] (33). Lane 2: unbiotinylated SAK-PFB produced by WB800[pSAKPFB]. Lane 3: purified biotinylated SAK-PFB produced by WB800[pSAKPFB].
Lane 4: negative control WB800[pWB980]. (B) SAK activity was estimated using the top ;rgarose plasminogen-skim milk plate method. The amounts of SAK in the individual wells were identical to those in the corresponding lanes shown in (A). Picture was taken at 10 hours after incubation at 37°C. Nu.inbers 1-4 correspond to the numbering in (A).
l~ fig. 6 1?urification of SAK-PFB from the culture supernatant of B. subtilis WB800[pSAKPFB] by in vitro biotinylation and monomeric avidin agarose chromatography. Samples were analyzed on a 12% SDS polyacrylameide gel and stained by Coomassie blue. M: molecular weight marker. Lane 1: ammonium sulfate precipitate before biotinylation. Lane 2: ammonium sulfate precipitate after biotinylation. Lane 3: column flow-through. Lane 4: 1-column volume wash.
Lanes 5 and 6:
eluate. Lane 7: concentrated pure SAK-PFB.

Claims (7)

1. A polypeptide comprising a biotin ligase having a C-terminal His-tag and an N-terminal chitin binding domain.

2. An isolated, purified or recombinant DNA which encodes for a a biotin ligase having a C-terminal His-tag and an N-terminal chitin binding domain.

3. An isolated, purified or recombinant DNA which is substantially homologous to the DNA of claim 2.

4. An expression vector comprising the DNA molecule of claim 2 or 3.

5. A host cell comprising the expression vector of claim 4.

6. A method of producing and recovering a protein comprising the steps of:
(a) culturing cells which express and secrete the protein tagged with a biotinylation peptide;
(b) biotinylating the protein with a biotin ligase as claimed in claim 1;
(c) separating the biotinylated protein by passing over an avidin column;
(d) eluting the the biotinylated protein.

7. The method of claim 6 wherein the protein is staphylokinase produed by B.
subtilis.