Extracellular accumulation of recombinant proteins ...

27 downloads 0 Views 223KB Size Report
Dec 20, 2005 - no leaking for MalE during its early growth stage6. pYebFH6/15A plasmid-encoded YebF and chromosomal MalE were simultaneously.

© 2006 Nature Publishing Group http://www.nature.com/naturebiotechnology

LETTERS

Extracellular accumulation of recombinant proteins fused to the carrier protein YebF in Escherichia coli Guijin Zhang, Stephen Brokx & Joel H Weiner Bacterial protein secretion is important in the life cycles of most bacteria, in which it contributes to the formation of pili and flagella and makes available extracellular enzymes to digest polymers for nutritional purposes and toxins to kill host cells in infections of humans, animals and plants. It is generally accepted that nonpathogenic laboratory strains of Escherichia coli, particularly K12 strains, do not secrete proteins into the extracellular medium under routine growth conditions1,2. In this study, we report that commonly used laboratory strains secrete YebF, a small (10.8 kDa in the native form), soluble endogenous protein into the medium, challenging the status quo view that laboratory strains do not secrete proteins to the medium. We further show that ‘passenger’ proteins linked to the carboxyl end of YebF are efficiently secreted. The function of YebF is unknown, but its use as a carrier for transgenic proteins provides a tool to circumvent toxicity and other contamination issues associated with protein production in E. coli. The DNA sequence of the yebGFE operon has a typical ribosomebinding site (GGAG, Blattner coordinates 1928424–1928421 in MG1655) upstream from the second ATG (1928414–1928412) in

Figure 1 Subcellular localization and secretion of E. coli YebF. (a) A 200-ml culture of HB101/pYebFH6/MS was harvested 4 h after induction and washed once at 22 1C with 200 ml of 100 mM MOPS (pH 7.0). Cells were subjected to osmotic shock to separate periplasm from spheroplasts. Protein samples from these fractions and immunoblotting were carried out as described in Methods. Lanes were marked as follows: S, spheroplasts; P, periplasm; H, hypertonic solution; M, medium. (b) Ten ml of induced HB101/pYebFH6/MS cells as described above were harvested and then washed once with 35 ml of fresh TB and suspended in 10 ml of fresh TB containing 80 mg chloramphenicol/ml. Cell samples of 1.5 ml each were taken at the time points indicated in the figure. Cells and medium were separated by centrifugation at 22 1C to avoid cold shock. Immunoblotting for His6tag was conducted as described in Methods. Lanes 1–7 show the cells (0–180 min) and lanes 8–14 show the medium (0–180 min) at 30-min time intervals. Lane F is the purified YebF-His6. (c) Protein profiles in the medium from HB101 cells harboring pMS119EH (lanes Ctr) or pYebFH6/MS (lanes YebF). Samples were collected at the indicated time after induction and prepared as in Methods. Proteins were separated on 10% SDS-Tricine gel.

the region between yebG and yebF. Although the first ATG (1928426– 1928424) could initiate in-frame translation of YebF, and this is the way it is predicted in databases, no ribosome-binding site could be found preceding this ATG. We cloned the yebF gene (NCBI (B1847) and Swiss-Prot (P33219)) in the expression vector pMS119EH under lac promoter regulation starting from nucleotide 1928424 without addition of a ribosomebinding site and with a 6-histidine tag at the carboxyl terminus. The His-tagged YebF protein was positively expressed as shown by immunoblotting. The N-terminal sequence of E. coli YebF resembles the universal, bacterial lipoprotein3 motif, which includes a fatty acid-acylation site (Cys-6). This motif is conserved in YebF from Shigella flexneri but the cysteine is missing in other similar proteins: STY2087 from Salmonella typhi, YebF from Salmonella typhimurium and YP01779 from Yersinia pestis. E. coli YebF protein also displays a 21 residues long typical secleader. A similar leader was found in all YebF-related proteins suggesting that E. coli YebF may be secreted to the periplasm. We purified YebF-His6 from the culture medium and carried out N-terminal amino acid sequence analysis that revealed the sequence as Ala-Asn-Asn-Glu-Thr-Ser-Lys-Ser-Val-Thr, which indicates that YebF is cleaved immediately after the 21-amino acid sec-leader and not

a S

P

H

M

b 1

c

2

3

4

Ctr

YebF

5

6

2h

0 Ctr

YebF

7

F

8

9 10 11 12 13 14

4h Ctr

YebF

6h Ctr

YebF

8h Ctr

YebF

F

Membrane Protein Research Group, Department of Biochemistry, 474 Medical Sciences Building, University of Alberta, Edmonton, Alberta, T6G 2H7, Canada. Correspondence should be addressed to J.H.W. ([email protected]). Received 16 August; accepted 28 October; published online 20 December 2005; doi:10.1038/nbt1174

100

VOLUME 24

NUMBER 1

JANUARY 2006

NATURE BIOTECHNOLOGY

LETTERS 2h F

C

4h M

C

M

6h C

M

YebF

Figure 2 Immunoblotting of YebF and CRP in HB101 (pYebFH6/MS) cells and the medium. Ten ml of cell culture from 50 ml of cells were harvested at 2, 4 and 6 h after induction. Protein samples from cells or media were prepared as in Methods and were loaded onto two parallel gels, which were subsequently subjected to immunoblotting for YebF with anti-His6 antibody or for CRP with anti-CRP antibody. Lane F is the purified YebF-His6. C, cells; M, medium.

Distribution (%)

preceding the Cys-6. The molecular weight of 11,829.2 ± 1.4 Daltons of His6-tagged YebF determined by mass spectrometry also shows no covalently attached fatty acyl chains by comparison to its calculated molecular weight of 11,860.3 Daltons. Therefore, YebF is not a lipoprotein as suggested in NCBI and Swiss-Prot. The presence of a potential sec-leader in YebF prompted us to determine its subcellua 90 b lar localization. E. coli HB101/pYebFH6/MS 80 2h 4h 6h after induction pF_Amy was induced with isopropyl-b-d-thiogalactoA B C A B C A B C 70 pFLS_Amy 60 pyranoside (IPTG) and the cold osmotic pAmy 50 shock method was used to fractionate the YebF-Amy 40 periplasm from the cytoplasm, resulting in 30 spheroplasts. We found that YebF was located 20 in the periplasm and surprisingly also in the 10 0 culture medium. No YebF was detected in the Medium Wash fluid Hypotonic Periplasm Spheroplasts solution spheroplasts after cold osmotic shock Fractions (Fig. 1a). When chloramphenicol was d 1.5 h 3.0 h 6.0 h c added to the cells after induction to stop 2.0 Cells C M C M C M F further synthesis of nascent proteins, we Media found that the level of YebF in the periplasm MalE 1.5 gradually declined to a trace level whereas it progressively accumulated in the medium 1.0 during the subsequent incubation period 0.5 (Fig. 1b). Figure 1c shows the accumulation YebF of YebF protein in the medium during 0.0 Alkaline phosphatase YebF-α-amylase growth. When cells were induced and (nmoles pNP/min/ml) (µmol Glu/min/ml) grown overnight, all the YebF was located in the medium (data not shown). Taken Figure 3 Examination of outer membrane leakage in E. coli cells expressing YebF protein. together, we concluded that YebF is most (a) Localization of chimeric a-amylase in periplasm and spheroplasts. E. coli HB101/pYebF-AmyH6, HB101/pFLS-AmyH6 and HB101/pAmyH6 were used. Data are the average of three replicates with the likely an extracellular, soluble protein. The presence of YebF in the culture med- standard deviation shown. (b) Protein profiles in the media from HB101 cells harboring pAmyH6 (lanes ium could be a result of cell lysis, leakage A), pFLS-AmyH6 (lanes B) and pYebF-AmyH6 (lanes C). Proteins were separated on 10% SDS-PAGE. (c) Distribution of YebF-a-amylase and alkaline phosphatase. GZ39T/pYebF-AmyH6 was grown in 25 ml through the outer membrane, a natural cell of TB. Cells were harvested at 7 h after induction with 0.05 mM IPTG, washed with 100 mM MOPS secretion process or some combination. To (pH7.0) and suspended in 1.5 ml of 30 mM Tris-HCl (pH 8.0). We mixed 200 ml of the cells with test for lysis we compared the subcellular 300 ml of cell lysis buffer (Methods). The cell lysates were used for both a-amylase and alkaline localization (Fig. 2) of YebF and catabolite phosphatase activity assays. To measure a-amylase activity in the medium, 35 ml of the medium was repressor protein (CRP), a cytoplasmic DNA directly used. Alkaline phosphatase in the medium was assayed as in Methods. Activities shown are binding protein. In cells, both CRP and YebF the average of two determinations. (d) Immunoblotting of YebF and MalE in cells and medium from MC4100/pYebFH6/15A. Cells were grown in 10 ml of TB plus 10 mg tetracycline/ml and induced were readily detected. In the medium, YebF with 0.05 mM IPTG and 0.4% maltose. We harvested 1 ml of cell culture at 1.5, 3 and 6 h after was detected at all experimental time points; induction. Protein samples were separated on 8.5% Tricine SDS-gel and subsequently subjected to CRP, however, was not observed in the med- immunoblotting for YebF and MalE with both anti-His6 antibody and anti-MalE antibody. Lane F is the ium at 2 and 4 h after induction. A trace purified YebF-His6. C, cells; M, medium. Enzyme activities

© 2006 Nature Publishing Group http://www.nature.com/naturebiotechnology

CRP

amount of CRP was detected at 6 h, which indicates the beginning of inevitable cell lysis. This rules out the possibility that the YebF found in the medium up to 4 h after induction resulted from cell lysis. To investigate the possible involvement of outer membrane leakage, we constructed two chimeric genes in the vector pMS119EH, under lac-promoter control by fusing the leaderless, mature a-amylase gene from Bacillus subtilis4 either to the 3¢ end of the yebF gene (pYebFAmyH6) or directly to the sec-leader sequence of yebF (pFLS-AmyH6). a-amylase expressed from the leaderless mature a-amylase gene in pMS119EH (pAmyH6) should remain in the cytoplasm. Only aamylase fused to full-length YebF could be found in the medium with 35% of the total activity in the medium. The leaderless a-amylase and a-amylase fused to the leader peptide of YebF were expressed but remained inside cells, and less than 5% of each on average were detected in the medium (Fig. 3a). These data were confirmed by the protein profile in the medium, in which only a-amylase fused to fulllength YebF (lane C) accumulated in the medium (Fig. 3b). Using osmotic shock, the periplasm of cells was separated from the cytoplasm. We recovered 68% of the a-amylase fused to the leader of YebF in the periplasm (Fig. 3a); 85% of leaderless a-amylase remained in the cytoplasm. Twelve percent of the leaderless a-amylase found in the periplasm could be due to cell breakage during the cold osmotic shock. Thus the YebF leader could direct the fusion protein to the periplasm using the sec-dependent translocation system but could not

NATURE BIOTECHNOLOGY

VOLUME 24

NUMBER 1

JANUARY 2006

101

LETTERS primarily localized in the cell fraction though some leakage into the medium was observed. In contrast, YebF was almost 350 3.5 entirely released into the medium and 300 3.0 b 100 was nondetectable in the cell fraction. 250 2.5 80 200 These data support the observation that 2.0 60 1.5 150 YebF is secreted rather then leaking across 40 1.0 100 the outer membrane. 20 0.5 50 No difference in membrane sensitivity was 0.0 0 seen for cells expressing YebF, compared to 0 2 4 6 8 Time (hours after addition of 0.1 mM IPTG) the parent or cells harboring only the vector, regarding either the permeability of antibioDilution factor tics including erythromycin, rifampicin and Figure 4 YebF as the carrier protein to produce hIL2 or a-amylase in the medium. (a) Secretion of azithromycin, or the sensitivity to Triton YebF-hIL2 from HB101(pYebF-hIL2H6). We added 10 mM DTT to 25 ml of one culture at the same X-100, cholate, deoxycholate and lysozyme time as addition of 0.05 mM IPTG. Samples were taken 3 h after induction. Localization of YebF-hIL2 in the presence or absence of EDTA (data was shown by immunoblotting using an anti-6-histidine tag antibody. C1, cells + 10 mM DTT; M1, not shown). medium + 10 mM DTT; C2, cells – DTT; M2, medium – DTT. (b) The activities of the secreted YebFWe monitored the expression and localizahIL2 fusion protein in the diluted medium harvested from the E. coli culture without DTT. The E. coli tion of human interleukin-2 (hIL-2) fused to medium was diluted with RHFM at the ratios indicated. Positive, treatment with 600 units of standard hIL-2/ml in RHFM medium; Negative, RHFM without hIL-2; TB, ten times diluted medium from YebF (Fig. 4a). The fusion protein was HB101/pMS119EH. Each bar was averaged from triplicate wells. (c) Secretion of the YebFa-amylase secreted to the medium at 3 h after inducfusion protein. HB101(pYebF-AmyH6) were grown in 25 ml of TB. Samples were taken at 2-h intervals tion. This was prevented by 10 mM dithioafter induction and assayed for a-amylase activity. Growth (——x——) was shown in Kletts. a-amylase threitol (DTT), although comparable activities in the cells (————) and the medium (——J——) were expressed in units per ml of the expression inside cells was seen. Activity original medium. assays for the YebF-hIL-2 fusion protein based on the growth of cytotoxic T-lymphorelease the protein to the medium. Therefore these data together cyte line (CTLL)-2 in vitro showed the secreted protein in the medium with the absence of a-amylase in the medium from cells harboring was active (43,800 units of hIL-2/ml) (Fig. 4b). In the case of YebF-a-amylase fusion protein, by 9 h after induction, plasmid pFLS-AmyH6 indicate that the outer membrane was able to prevent mature a-amylase (the YebF leader peptide was 75% of YebF-a-amylase was secreted from cells (Fig. 4c). The fusion removed by leader peptidase I) from leaking out of cells during protein was stable in the medium and the maximal secretion occurred after the culture reached stationary phase. The specific activity of normal growth. Only the YebF-a-amylase fusion protein was able to cross the outer a-amylase in the initial culture medium varied after induction from membrane from the periplasm (Fig. 3a,b) demonstrating that YebF 150 to 85 mmol glucose/min/mg protein at 4 or 9 h after induction, plays an essential role in targeting the fusion protein for secretion respectively. A B. subtilis culture4 displays a specific activity of across the outer membrane to the medium. Figure 1a shows no 3.4 mmol glucose/min/mg protein. The high specific activity here is apparent difference in the mobility of YebF protein in the periplasm or also reflected in the simple protein profiles of the medium shown in the culture medium indicating that no additional cleavage occurred in Figure 3b. during secretion across the outer membrane. The secretion of chromosomally integrated Erwinia pectate lyase in To demonstrate that YebF in the medium is due to secretion, rather E. coli by a cosmid-based cluster of out genes (B12 kb) encoding the than periplasmic leakage, we compared the simultaneous release of gene secretion apparatus of Erwinia chrysanthemi has been reported7. YebF-a-amylase and alkaline phosphatase, a well-defined periplasmic E. coli cells can secrete Klebsiella oxytoca pullulanase into the medium enzyme, from cells hosting pYebF-AmyH6. As alkaline phosphatase is by coexpressing the K. oxytoca secretion machinery genes cloned on a induced by phosphate starvation, it is poorly expressed in E. coli second plasmid8. The secretion of chitinase into the medium can grown in rich medium containing inorganic phosphate salts. Thus we occur by coexpressing a cluster of E. coli gsp genes, encoding the type isolated a derivative of E. coli TG1, GZ39T, that expresses 10 times II secretion machinery carried on a second plasmid6. These excretion more alkaline phosphatase than TG1 in rich medium. Figure 3c shows studies required the artificial induction of secretion machinery prothat 42.6% of YebF-a-amylase was secreted to the medium compared teins. Herein we discovered YebF to be naturally secreted from E. coli to only 5.6% of alkaline phosphatase. cells without any artificial manipulation to the host cells. The experiWe also transformed E. coli MC4100 with pYebFH6/15A, a pACYC- ments reported rule out the release of YebF secondary to cell lysis or based (P15A replicon) low copy number plasmid5 to reduce the outer membrane leakage. We believe YebF is secreted in a two-step process. First, pre-YebF is expression of YebF. MC4100 was chosen as this strain exhibits the least leakage of maltose binding protein (MalE) of 15 common E. coli transported from the cytoplasm to the periplasm by the sec-dependent strains tested (data not shown) and a derivative strain showed system in the inner membrane and converted to mature YebF by leader no leaking for MalE during its early growth stage6. pYebFH6/15A peptidase I. Next, the leaderless YebF is exported from the periplasm to plasmid-encoded YebF and chromosomal MalE were simultaneously the medium by an unknown system on the outer membrane. The general secretion pathway of Gram-negative bacteria has at expressed by induction of MC4100 with IPTG and maltose. YebF was present in the medium fraction at 1.5 and 3 h after induction, whereas least six different terminal branches depending on the secretion MalE was not present in the medium at 1.5 h and was very weakly pathways in the outer membrane9. Genomic analysis of E. coli has observed at 3.0 h (Fig. 3d). In the cell fraction, however, YebF was identified genes that are homologous to genes encoding secretion poorly detected whereas MalE was readily detected. At 6.0 h, MalE was proteins in other bacteria including gsp and yacC (gspS) homologous C1

M2

c

C2

102

TB at ive eg

N

1/ 10 1/ 10 0 1/ 1, 00 0 1/ 10 , 1/ 000 10 0, 00 0

Po s

iti ve

Surviving fraction (%)

α-Amylase activities (µmolGlu/min/ml)

M1

Kletts

© 2006 Nature Publishing Group http://www.nature.com/naturebiotechnology

a

VOLUME 24

NUMBER 1

JANUARY 2006

NATURE BIOTECHNOLOGY

© 2006 Nature Publishing Group http://www.nature.com/naturebiotechnology

LETTERS to Klebsiella pulS6. Although these genes appear to encode functional proteins10, their transcription is turned off under standard growth conditions1. gsp genes, for example, are silenced by H-NS6. gspD is not involved in the secretion of YebF (G.Z. and J.H.W., unpublished data). The Type I single accessory pathway11 responsible for the secretion of pili, fimbria and adhesin may act as the transporter for YebF. This transporter could reside on the outer membrane and recognize a yet unidentified signal in the YebF protein. There may also be an unknown secretion mechanism for YebF. E. coli is the most widely used bacterium for protein expression. The absence of a secretion system in common laboratory strains has prompted investigators to explore different ways of producing recombinant proteins in the medium2 or display proteins on the cell surface such as using the autotransporter domain of adhesin12. Using YebFhIL2 (hIL-2, 15 kDa and very hydrophobic13), YebF-a-amylase (a-amylase, 48 kDa and hydrophilic) and YebF-alkaline phosphatase (94 kDa), we demonstrated that YebF could carry these fusion proteins in their active states to the medium. These data indicate that YebF may have the ability to carry proteins of varying size and hydrophilicity from E. coli cells into medium. Compared with the expression of a-amylase produced in the medium from Bacillus subtilis X-23 (ref. 4), the YebF fusion protein achieved a higher yield of protein and a higher specific activity in a much shorter incubation time. Finally, the presence of lipopolysaccharide is often a concern in therapeutic proteins overexpressed in E. coli cells14, and hence, for recombinant proteins used as drugs, production of these proteins in the medium could potentially reduce the contamination of the recombinant proteins with lipopolysaccharide in the purified product. METHODS Bacterial strains and growth conditions. The E. coli strains used in this study were HB101 (supE44 hsd20(rB–mB–) recA13, ara-14, proA2, lacY1, gaIK2, rpsL20, xyl-5, mtl-1), BL21(DE3)(hsdS, gal(lcIts857 ind1, Sam7, nin5, lacUV5-T7, gene 1), MG1655(F- l- ilvG- rfb-50, rph-1) and MC4100 (F–, D(argF-lac)U169, araD139, rpsL150, ptsF25, flbB5301, rbsR, deoC, relA1). Cells were grown at 30 1C in Terrific broth (TB). Antibiotics were added at the following concentrations: 100 mg/ml of ampicillin, 80 mg/ml chloramphenicol and 10 mg/ml tetracycline (tet). Expression of genes under lac promoter control was induced with 0.1 mM IPTG or as indicated. Cells growth was monitored with a Klett spectrophotometer equipped with a red filter. Molecular biology techniques and plasmid construction. DNA manipulation, sequencing and bacterial transformation were essentially as described15. Expression plasmids were created based on pMS119EH16 containing the lac promoter or pT7-5 (ref. 17) with T7 promoter. PCR amplifications using the touch-down protocol and DNA purification thereafter were carried out as described18. The yebF gene was amplified from HB101 chromosomal DNA using the following forward and reverse oligomers, respectively, with the indicated restriction sites underlined: yebF5, 5¢-GAGAATTCGGAGAAAAACATGAA AAAAAG-3¢ containing EcoR1; and yebF3, 5¢-ATATCTCGAGACGCCGCTGA TATTC-3¢ (Xho1). To tag the gene with six histidines at its 3¢-end (yebFH6), the amplified DNA, after digestion with EcoR1 and Xho1, was cloned into pCITE2a(+) (Novagen) cut with the same enzymes. The EcoR1-HindIII fragment carrying yebFH6 was then subcloned into pMS119EH to generate pYebFH6/MS, or into pT7-5 to make pYebFH6/T7. The mature truncated form of the a-amylase gene (GenBank accession no. AB015592) of Bacillus subtilis X-23 (ref. 4) was amplified from pAC92 kindly provided by S. Astolfi-Filho, University of Brasilia, Brazil, using the following oligos: amy5, 5¢-TAGAATTCAGGAGAAAAACATGGTCGACTCG GTCAAA AACGGG-3¢ engineered with an EcoR1site, a ribosomal binding site (bold), the in-frame initiation codon ATG and Sal1; amy3, 5¢-ATATCTCGA GATGAGGCGCATT TCC-3¢ (Xho1). The amplified DNA was cut with EcoR1

NATURE BIOTECHNOLOGY

VOLUME 24

NUMBER 1

JANUARY 2006

and Xho1 and used to replace the yebF fragment (EcoR1-Xho1) in pYebFH6/MS to create pAmyH6. pYebF-AmyH6 containing the yebF-amy fusion gene was created by replacing the EcoR1-Sal1 fragment in pAmyH6 with the EcoR1-Xho1 fragment carrying the yebF gene from pYebFH6/MS. To fuse the sec-leader sequence of the yebF gene to the mature a-amylase gene, the leader sequence was first amplified by PCR using pYebFH6/MS as the template with Fls (5¢-ATATCTCGAGAGCGAAAACTGATGC-3¢) containing a Xho1 restriction site and yebF5 (above) as the oligomers. After digestion with EcoR1 and Xho1, the amplified DNA fragment was inserted into pAmyH6 to replace the EcoR1- Sal1 sequence, which generated pFLS-AmyH6. The sequence of this hybrid gene was verified by sequencing. pYebFH6/15A was constructed by replacing the sequence between Dra 1 and EcoR V containing part of the b-lactamase sequence and the ColE1 origin in pYebFH6/MS with the sequence of Pvu II–Sty 1 (blunted with Klenow) containing replication origin of p15A and the entire tetracycline cassette in pACYC184 (ref. 5). To construct pYebF-hIL2H6, the hIL-2 gene was amplified by PCR from pBM806, a plasmid containing hIL-2 gene provided by Biomira, using the following oligonucleotides: hil5 with Sall, 5¢-GATCATGTCGACGCTCC GACCTCCAGC-3¢ and hil3 with Xhol, 5¢-ATATCTCGAGGGTCAGGGTGGA GAT-3¢. The Sall-Xhol fragment of the PCR product was inserted into the Xhol site in pYebFH6/MS to generate pYebF-hIL2H6. To make pYebF-phoAH6/T7 plasmid, the phoA gene was PCR-amplified with the following oligos: phoA5 with Xho1, 5¢-GCATATCTCGAGCGGACACCA GAAATGCC-3¢; and phoA3 with Sal1, 5¢-CGATAGTCGACTTTCAGCCCCA GAGC-3¢. The Xhol-Sal1 fragment of the PCR product was inserted into Xhol site in pYebFH6/T7 to generate pYebF-phoAH6/T7. Protein expression and treatment of samples. A single colony from LuriaBertani plates or a small aliquot of cold cell suspension stored in 40% glycerol at –20 1C was inoculated into 1.5 ml of TB and grown overnight at 30 1C. After harvesting by centrifugation, the cells were washed once in 1.5 ml of fresh TB, and were suspended in 1.5 ml of fresh TB and inoculated at 1% by volume into fresh TB for expression studies. Protein expression was induced by addition of 0.1 mM or 0.05 mM IPTG at 0.6 OD600. Cells were harvested at the indicated time points in the experiments. After harvesting by centrifugation at 22 1C to avoid cold shock, cell pellets were washed once with a volume equal to the original culture volume of 100 mM MOPS (pH 7.0). The resulting medium supernatant was filtered using a 0.22 mm Millipore filter to remove unpelleted cells. Subcellular fractionation. Periplasm was separated from cytoplasm by the osmotic shock method19. a-amylase activity assay. This assay is based on the 3,5-dinitrosalicylic acid (DSNA) method as described4, using starch as the substrate for the enzyme. The reaction was stopped by adding 200 ml of DNSA in 0.4 N NaOH to 200 ml of the reaction mixture after 10 min incubation at 58 1C with shaking once per minute. The mixture was boiled for 5 min thereafter, diluted with 1 ml of water, and spun for 1 min to pellet starch. Finally, readings were taken at 540 nm. The enzyme activity was expressed in moles of glucose (Glu) released per minute, which is equivalent to the amount of reducing sugar released by the enzyme from starch in the reaction. A pair of every sample to be assayed was set up. One is for the reaction, the other for background reading, to which 200 ml of DNSA was added before the reaction starts. To measure the activity in the medium or in the osmotic shock fluid, we added 35 ml of the liquid directly to the reaction mixture. To assay the activity in cells or in the spheroplasts resulting from the osmotic shock, we mixed the cell suspension or the spheroplast suspension in 10 mM Tris-HCl (pH 7.0) with equal volume of lysis buffer (pH 8.0) consisting of 70 mg/ml lysozyme, 6 mM EDTA, 2% Triton X-100 to break cells or spheroplasts. After lysis was complete on ice, 1 ml of 210 mM CaCl2 was added to every 70 ml of the lysate to overcome the EDTA and allow for 0.1 mM Ca2+ in the reaction mixture as required for a-amylase activity. Alkaline phosphatase activity assay. 1 ml of the reaction consists of 100 mM Tris-HCl (pH 8.0), 10 mM MgCl2, 2 mM ZnCl2 and 0.04% pNPP. The reaction

103

© 2006 Nature Publishing Group http://www.nature.com/naturebiotechnology

LETTERS was run in the dark for 60 min at 37 1C and stopped by addition of 40 ml of 5 M NaOH followed by 1 min of centrifugation at 21,910g before taking readings at 410 nm. To measure the background readings, samples were added to the mixtures after the reactions were stopped by NaOH. Enzyme activities were calculated using e410 ¼ 17,800 M–1 cm–1 for p-nitrophenolate anion (pNP), the yellow product released from pNPP. To measure alkaline phosphatase activity in the medium, we first cleared the medium to reduce phosphate and colored substances and concentrated it as follow: 8 ml of the medium was filtered three times through an Amicon ultracentrifugal filter (MWCO 30KD) with 50 mM MOPS (pH 7.0) and finally concentrated to 0.6 ml 200 ml of this preparation was used for the assay.

ACKNOWLEDGMENTS J.H.W. is a Canada Research Chair in Membrane Biochemistry. This work was supported by the Canadian Institutes of Health Research and the Alberta Heritage Foundation for Medical Research. We thank the following people for their valuable contribution to this study. Len Wiebe and Aihua Zhou provided technical assistance with the hIL-2 bioassay. Chris Bleackley, Irene Shostak and Jonathan Hooton provided the materials and methods for the hIL-2 bioassay. Lorne Burke and Paul Semchuk provided technical support for HPLC and mass spectrometry. We thank Biomira for kindly allowing us to use plasmid pBM806.

Purification of YebFH6 and protein chemistry. Total protein in the culture medium of E. coli HB101/pYebFH6/MS grown overnight (15 h) after induction with IPTG was precipitated by ammonium sulfate at 70% saturation at 4 1C in the presence of 0.2 mM PMSF. The pellet was dissolved in 20 mM Tris-HCl (pH 7.5) containing 0.2 mM PMSF, 10 mM imidazole and 500 mM NaCl. Purification of YebFH6 was achieved using a nickel affinity column. This partially purified protein was further purified using reversed-phase high-performance liquid chromatography (HPLC) on a 300 A˚ C8 column (1 mm  15 cm) eluted with a gradient of acetonitrile from 0% to 2% in the presence of 0.05% TFA at the flow rate of 100 ml/min. A single major peak was observed at the 25th min. This peak fraction was collected and used for both matrix-assisted laser desorption ionization/time-of-flight (MALDI-TOF) linear mode mass spectrometry (Voyager 6064, Applied Biosystems) and N-terminal amino acid sequence determination using Routine 3.0 on a Hewlett Packard G1000A Protein Sequencer.

Published online at http://www.nature.com/naturebiotechnology/ Reprints and permissions information is available online at http://npg.nature.com/ reprintsandpermissions/

Protein electrophoresis and western blotting. The protein samples from either the culture medium or the cells were prepared by precipitation with 5% trichloroacetic acid (TCA). After centrifugation in an Eppendorf tube at 21,910g at 22 1C for 5 min, the pellet was washed once by suspending it in 1 ml of water and centrifuging for 5 min to remove residual TCA. To completely extract TCA from the pellets, 1.5 ml of –20 1C cold acetone was added to each tube. The tube was then vortexed and kept at –20 1C for 20 min before centrifugation as above. The pellet was retained and subjected to one additional acetone wash. The final pellet was dissolved in 50 mM Tris-HCl, pH 8.0 before adding an equal volume of electrophoresis loading buffer. Proteins were separated on 8.5% SDS-Tricine gel20 and then transferred to a nitrocellulose membrane for immunoblotting. To detect the histidine tag, monoclonal mouse antibody against histidine tag (Qiagen) was used. CRP was detected by using a rabbit polyclonal antiserum against CRP, kindly provided by Hiroji Aiba, Nagoya University, Japan. Immunoblots were developed by enhanced chemiluminescence. IL-2 bioassay. The biological assay of YebF-hIL2 was based on a published method21 using the IL-2–dependent T-lymphocyte cell line CTLL-2 with modifications. Before the assay, the YebF-hIL2 sample was prepared by passing the E. coli culture medium harvested 3 h after induction with 0.1 mM IPTG from either HB101/pYebF-hIL2H6 or pMS119EH (control) through a 0.22-mm Millipore filter. The cell culture medium (RHFM) for CTLL-2 uses RPMI 1640 supplemented with 10% (vol/vol) fetal calf serum, 0.1 mM b-mercaptoethanol, 25 mM HEPES pH 7.5, 2 mM L-glutamine. We placed 50 ml of fresh CTLL-2 cells (106 cells/ml) in RHFM in a well of a 96-well plate and added 50 ml of standard hIL-2 (kindly provided by C. Bleackley, University of Alberta) or YebF-hIL2 samples diluted with RHFM. The cells were incubated at 37 1C with 5% CO2 for 20 h. Cell proliferation was measured using the MTT (3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay. We prepared 5 mg of MTT/ml in PBS and sterilized it. We added 10 ml of MTT to each well followed by 4 h incubation at 37 1C. We added 150 ml of isopropanol containing 0.04 N HCl to each well and mixed thoroughly to dissolve the resulting crystals. The absorbance was read at 560 nm of wavelength. Each value is the average of three assays.

104

COMPETING INTERESTS STATEMENT The authors declare that they have no competing financial interests.

1. Pugsley, A.P. & Francetic, O. Protein secretion in Escherichia coli K-12: dead or alive? Cell Mol. Life Sci. 54, 347–352 (1998). 2. Shokri, A., Sanden, A.M. & Larsson, G. Cell and process design for targeting of recombinant protein into the culture medium of Escherichia coli. Appl. Microbiol. Biotechnol. 60, 654–664 (2003). 3. Brokx, S.J. et al. Genome-wide analysis of lipoprotein expression in Escherichia coli MG1655. J. Bacteriol. 186, 3254–3258 (2004). 4. Ohdan, K. et al. Characteristics of two forms of alpha-amylases and structural implication. Appl. Environ. Microbiol. 65, 4652–4658 (1999). 5. Chang, A.C.Y. & Cohen, S.N. Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J. Bacteriol. 134, 1141–1156 (1978). 6. Francetic, O., Belin, D., Badaut, C. & Pugsley, A.P. Expression of the endogenous type II secretion pathway in Escherichia coli leads to chitinase secretion. EMBO J. 19, 6697–6703 (2000). 7. He, S.Y., Lindeberg, M., Chatterjee, A.K. & Collmer, A. Cloned Erwinia chrysanthemi out genes enable Escherichia coli to selectively secrete a diverse family of heterologous proteins to its milieu. Proc. Natl. Acad. Sci. USA 88, 1079–1083 (1991). 8. Poquet, I., Faucher, D. & Pugsley, A.P. Stable periplasmic secretion intermediate in the general secretory pathway of Escherichia coli. EMBO J. 12, 271–278 (1993). 9. Stathopoulos, C. et al. Secretion of virulence determinants by the general secretory pathway in gram-negative pathogens: an evolving story. Microbes Infect. 2, 1061–1072 (2000). 10. Francetic, O. & Pugsley, A.P. The cryptic general secretory pathway (gsp) operon of Escherichia coli K-12 encodes functional proteins. J. Bacteriol. 178, 3544–3549 (1996). 11. Thanassi, D.G. & Hultgren, S.J. Multiple pathways allow protein secretion across the bacterial outer membrane. Curr. Opin. Cell Biol. 12, 420–430 (2000). 12. Lattemann, C.T., Maurer, J., Gerland, E. & Meyer, T.F. Autodisplay: functional display of active b-lactamase on the surface of Escherichia coli by the AIDA-I autotransporter. J. Bacteriol. 182, 3726–3733 (2000). 13. Robb, R.J. Interleukin 2: the molecule and its function. Immunol. Today 5, 203–209 (1984). 14. Masui, Y. et al. Microheterogeneity of recombinant products: human interleukin 1a and 1b. in Current Communications in Molecular Biology—Therapeutic Peptides and Proteins. (eds. Marshak, D. and Liu, D.) 167–172 (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1989). 15. Sambrook, J. & Russell, D.W. Molecular cloning, a laboratory manual, edn. 3 (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2001). 16. Strack, B., Lessel, M., Calendar, R. & Lanka, E. A common sequence motif, -E-G-Y-A-TA-, identified within the primase domains of plasmid-encoded I- and P-type DNA primases and the a protein of the Escherichia coli satellite phage P4. J. Biol. Chem. 267, 13062–13072 (1992). 17. Tabor, S., Huber, H.E. & Richardson, C.C. Escherichia coli thioredoxin confers processivity on the DNA polymerase activity of the gene 5 protein of bacteriophage T7. J. Biol. Chem. 262, 16212–16223 (1987). 18. Zhang, G. & Weiner, J.H. CTAB-mediated purification of PCR products. Biotechniques 29, 982–986 (2000). 19. Neu, H.C. & Heppel, L.A. The release of enzymes from Escherichia coli by osmotic shock and during the formation of spheroplasts. J. Biol. Chem. 240, 3685–3692 (1965). 20. Schagger, H. & von Jagow, G. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379 (1987). 21. Lei, H. et al. Induction of potent antitumor response by vaccination with tumor lysatepulsed macrophages engineered to secrete macrophage colony-stimulating factor and interferon-g. Gene Therapy 7, 707–713 (2000).

VOLUME 24

NUMBER 1

JANUARY 2006

NATURE BIOTECHNOLOGY

Suggest Documents