Quantcast

Isolated nucleic acids encoding a secretory signal for expression and secretion of heterologous recombinant proteins

Imported: 24 Feb '17 | Published: 11 May '04

Jeak Ling Ding, Nguan Soon Tan, Bow Ho, Toong Jin Lam

USPTO - Utility Patents

Abstract

A universal secretory signal originally derived from a piscine vitellogenin (Vtg) gene is inserted into various expression vectors for driving the secretion of the recombinant protein into the culture medium. This enhances the detection, quantification and downstream scaled-up purification of a recombinant protein of interest. The secretory signal system is very versatile, being conveniently and widely applicable to an array of heterologous host cells such as bacteria, yeast, insect, piscine, and mammalian cell lines (e.g., COS, CHO, NIH/3T3). The said secretory system is also applicable as a reporter vector for secretion of reporter proteins/enzymes, thus, enabling the detection of the reporter proteins (e.g., CAT, GFP) in the culture medium.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows the construction of pAc5/VtgCrFCES-V5-His. The full-length CrFC21 cDNA encoding Factor C was released by a BglII and NotI double digest. The released fragment was subcloned into the BamHI and NotI sites of VtgEGFP. Consequently, the BglII/BamHI site is destroyed. Then the LPS-binding domain of CrFC21 was released by digestion with Eco47 III and SalI (cutting an internal site). The fragment was subcloned into the EcoRV and XhoI sites of pAc5/V5-HisA to give pAc5/VtgCrFCES-V5-His. Both Eco47 III/EcoRV and SalI/XhoI sites are destroyed.

FIG. 2A shows the amino acid (SEQ ID NO:13) and nucleotide sequences (SEQ ID NO:12) at the junction of Vtgss and CrFCES. The nucleotide junctions were determined by sequencing using the Ac5 forward primer and pcDNA3.1/BGH reverse primer. The CrFCES is cloned in-frame with respect to Vtgss (at the 5′ end) and V5-His (at the 3′ end). The secreted VtgCrFCES protein was purified by affinity chromatography (TALON™, Clontech) under denaturing conditions (6M urea, 250 mM NaCl and 20 mM sodium phosphate buffer, pH7.0). The protein was transferred to a PVDF membrane in transfer buffer (10% methanol, 10 mM CAPS, pH 11.0) and its N-terminal amino acid sequence determined. Using the novel Vtgss as secretory signal, only a single cleavage point was identified (indicated by the arrow). Thus, Vtgss allows the homogenous production of secreted heterologous recombinant protein.

FIG. 2B shows a map of the plasmid pAc5/VtgCrFCES-V5-His.

FIGS. 3A and 3B show the distribution of secreted VtgCrFCES protein.

FIG.

3A—culture medium

Lane M: Benchmark™ prestained marker

Lane 1: control medium (30 g)

Lane 2: VtgCrFCES medium (not purified 30 g)

Lane 3: VtgCrFCES (affinity purified; 1 g)

Lane 4: VtgCrFCES (ISOprime™ purified; 1 g)

FIG.

3B—cell lysate;

Lane M: Bio-Rad prestained marker

Lane 1: control cell lysate (30 g)

Lane 2: VtgCrFCES cell lysate (transient; 30 g)

Lane 3: VtgCrFCES cell lysate (stable; 30 g)

The VtgCrFCES (hereinafter “VtgCrFCES”) protein was effectively secreted into the culture medium, as verified by SuperSignal HisProbe™ Western Blotting Kit (Pierce). The secreted VtgCrFCES was purified to homogeneity by isoelectric focusing (ISOprime™, Hoefer) resulting in a single protein band having a molecular weight identical to that of the protein isolated by affinity column chromatography. The ISOprime™-purified protein was not denatured even when purified in pyrogen-free water. No VtgCrFCES was detected in the cell lysate. The purification process was made easy and more effective by the presence of VtgCrFCES in the culture medium.

FIG. 4 shows the lipid A binding properties of ISOprime™-purified VtgCrFCES. Immobilization of lipid A to the HPA sensor chip has been described earlier (Pui et al.,

J. Endotoxin Research 4:391-400 (1997)).

Injection of VtgCrFCES (800 ng/100 l) resulted in a signal of ˜2000 relative response units. This represents a ˜92% saturation of lipid A which clearly indicates that VtgCrFCES binds to lipid A. Subsequently, injection of antibody (INDIA His-HRP Ab) against the poly-His tag of VtgCrFCES resulted in a further increase in the signal. The binding of INDIA His-HRP Ab further confirms that only VtgCrFCES was bound to the immobilized lipid A. The result also indicates that the six additional Vtgss-derived amino acids in the N-terminus of the protein do not interfere with the proper folding of the recombinant protein and its LPS-binding property. The properly folded VtgCrFCES is secreted. Injections of samples at various steps are marked on the sensogram. “Wash” indicates a washing step using pyrogen-free water. The relative response units of the signal are obtained by subtracting the response units recorded just before injection of sample from the response units recorded after injection of sample and a 2 min wash.

FIG. 5A shows the construction of psp-VtgCAT. The CAT gene was isolated by PCR. The start ATG codon of the CAT gene was changed to CTG to ensure translation initiation at Vtgss as indicated in bold in the sequence of the CAT forward primer (SEQ ID NO:1) (5′-GAA GAT CTG CTG GAG AAA AAA ATC ACT GG-3′). The stop codon of the CAT gene is indicated in bold in the CAT reverse primer (SEQ ID NO:2) (5′-GC ATC GGC CGT GCC TTA AAA AAA TTA CGC-3′). The temperature profile used includes a 1

st cycle (94° C./5 min; 50° C./1 min; 72° C./1 min), 29 cycles (94° C./45 sec; 50° C./30 sec; 72° C./30 sec) and final extension at 72° C. for 5 min. The PCR product of ˜675 bp was purified using a Qiaquick™ PCR Purification Kit and subcloned into the HincII site of pBluescript™ II SK. The resultant construct is named pBSCAT. The CAT gene was released by BglII and NotI digestion and subcloned into compatible BamHI and NotI sites of pVtgEGFP. The CAT gene harboring Vtgss (VtgCAT) was subsequently released by using Eco47 III and NotI digest and recloned into the EcoRV and NotI sites of pBluescript™ II SK to form the vector pBSVtgCAT. This recloning step serves two purposes: (1) the sequences between Vtgss and CAT can be determined using a T7 primer; (2) this creates a HindIII site 5′ of VtgCAT and a ClaI site 3′ of VtgCAT for future subcloning into the pSEAP-Promoter vector.

FIG. 5B (SEQ ID NOS:14-17) shows the details of the Vtgss-CAT fusion in the pBSVtgCAT vector.

FIG. 5C shows the plasmid map of the pSEAP-Promoter vector.

FIG. 5D (SEQ ID NOS:14, 16-18) shows the amino acid and nucleotide sequences adjoining Vtgss and CAT in the vector psp-VtgCAT. The sequences between Vtgss and CAT were verified by Taqtrack™ sequencing (Promega) using a T7 primer. Both EcoRV/Eco47 III and BglII/BamHI sites were destroyed.

FIG. 5E shows the detailed plasmid map of psp-VtgCAT. The VtgCAT gene was subcloned downstream of the SV40 early promoter. The unique restriction enzymes in the multiple cloning site (MCS) were as illustrated. This construct can be used to analyze enhancer sequences cloned into any one of the unique sites in the vector.

FIG. 5F shows the plasmid map of ERU-psp-VtgCAT. The estrogen response unit (ERU) of 565 bp from the Xenopus vitellogenin B1 gene was subcloned into the NheI and BglII site of psp-VtgCAT. The NheI/XbaI site was destroyed. The transcription of VtgCAT is under the influence of estrogen in this construct.

FIG. 5G shows restriction digests of various constructs.

Lane 1: 100 bp marker

Lane 2: ERU-psp-VtgCAT/HindIII and ClaI

Lane 3: psp-VtgCAT/HindIII and ClaI

Lane 4: pSEAP-Promoter/HindIII and ClaI

Lane 5: cER in pSG1/HindIII

The VtgCAT (˜830 bp) fragment can be released by Hind III and Cla digestion.

The upper band for ERU-psp VtgCAT is slightly larger than psp-VtgCAT because it contains a ˜560 bp ERU.

FIG. 6 shows the secreted VtgCAT expression profile over time in COS-1 cells. The COS-1 cells were cotransfected with ERU-psp-VtgCAT, pSG cER (chicken estrogen receptor expression vector) and pSEAP-Control in a ratio of 6:3:1. The result illustrates several important features. First, the recombinant VtgCAT was effectively secreted and accumulated in the culture medium. Second, the secretion of VtgCAT was not delayed. The uninduced cells (open box) exhibited only a marginal increase in VtgCAT over the period of 24 h. For induced cells (solid box), the increase in the production and secretion of VtgCAT can be detected as early as 2 h after the E

2-induction. A peak 7-fold increase (˜1400 g VtgCAT/U SEAP activity/h) in VtgCAT was observed 12 h post-induction as compared to the corresponding control. By 24 h, the VtgCAT decreased to ˜1000 g VtgCAT/U SEAP activity/h. The secreted VtgCAT was assayed using CAT ELISA (Boehringer Mannheim) SEAP was measured as described in Tan et al.,

Mol. Cell. Endocrinol. 123:146-161 (1996)). The VtgCAT amount was normalized using SEAP activity. 1×10

−8 M or estradiol was used for induction in all experiments.

FIG. 7 shows a Northern blot analysis of E

2-induced VtgCAT expression for ERU-psp-VtgCAT. Total RNA was isolated from the COS-1 cells for E

2-induced and uninduced cells using RNAeasy™ Mini Kit (Qiagen). Ten g of RNA was resolved on a 1.2% formaldehyde/agarose gel and alkaline transferred onto Hybond-N+™ membrane (Amersham). The membrane was probed initially with VtgCAT, stripped and reprobed with a mouse actin gene. The result indicates that the levels of VtgCAT in the culture medium are directly proportional to changes in intracellular concentration of VtgCAT mRNA. Thus, Vtgss can be utilized for reporter gene studies since there is no delay in its production and secretion. An actin cDNA probe was used to normalize the result.

Lane 1: 0 h

Lane 2: 2 h

Lane 3: 4 h

Lane 4: 6 h

Lane 5: 12 h

Lane 6: 24 h

FIG. 8A shows the construction of pVtgEGFP. The Vtgss was isolated by PCR using a vector T3 primer and a OaVtgExon2 reverse primer (SEQ ID NO:3) (5′-CCAAGTTGGACTGGTCCCCCA-3′) using pOaVtg 1 as a template. The PCR conditions used includes a 1

st cycle (94° C./5 min; 52° C./1 min; 72° C./1 min), 29 cycles (94° C./1 min; 52° C./30 sec; 72° C./30 sec) and final extension of 72° C. 5 min. The PCR product was purified using Qiaquick™ Nucleotide Purification Kit (Qiagen) and digested with SacI. This blunt-end (from the PCR) and SacI fragment was subcloned into the SmaI and SacI sites of pEGFP-N1 (Clontech). Consequently, the SmaI site of pEGFP-N1 was destroyed. To reduce the distance between the transcriptional start site and the promoter, the clone was digested with Eco RI and BglII, Klenow end-filled, and religated to give pVtgEGFP.

FIG. 8B (SEQ ID NOS:19-20) shows the details of the Vtg-EGFP fusion in the vector pVtgEGFP.

FIG. 8C shows the detailed plasmid map of pVtgEGFP. The VtgEGFP was subcloned downstream of the CMV promoter. Several unique sites flank both the Vtgss and EGFP to allow more flexibility in cloning. The sequence junction of Vtgss EGFP was determined using Taqtrack™ sequencing and an EGFP reverse primer (SEQ ID NO:4) (5′-CCCTCGCCGGACACGCTGA-3′). The transcriptional start site was determined by Teo et al. (“A Novel Piscine Vitellogenin Gene: Structural and Functional Analyses of Estrogen-Inducible Promoter”,

J. Mol. Cell. Endocrin. (1998), in press).

FIG. 8D shows western blot analysis of VtgEGFP expression in COS-1 cells. Transient transfection of COS-1 cells with pVtgEGFP was performed using Lipofectamine™ (Gibco). Forty-eight hours post transfection, both the medium and cell lysate were sampled. The result indicates that majority of VtgEGFP was secreted into the culture medium. A weaker immunoreactive band was also identified in the cell lysate using GFP antibodies. This is most likely due to the translation initiation at the start ATG codon of native EGFP. This shows that Vtgss can direct secretion of another reporter gene, EGFP. However, the optimal result for Vtgss-directed secretion can be achieved if the start codon of the EGFP gene is replaced or removed so that translation begins at the start codon of the Vtgss.

Lane M: Bio-Rad Kaleidoscope™ marker

Lane 1: control culture medium (24 h, 30 g)

Lane 2: VtgEGFP culture medium (24 h, 30 g)

Lane 3: VtgEGFP culture medium (48 h, 30 g)

FIG. 9 shows the secreted VtgCAT expression profile over time in NIH/3T3 (normal mouse fibroblast) and CHO-B (Chinese hamster ovary) cells. NIH/3T3 and CHO-B cells were cotransfected with ERU-psp-VtgCAT, pSG cER (chicken estrogen receptor expression vector) and pSEAP-Control in a ratio of 6:3:1 using Lipofectamine™ (Gibco). The result shows that recombinant VtgCAT was effectively secreted and accumulated in the culture medium. Thus, secretion of VtgCAT was not limited to COS-1 cells; estrogen-induced expression of ERU-psp-VtgCAT can also be observed in other mammalian cells. The uninduced cells (open box) exhibited only a marginal increase in VtgCAT over the period of 24 h. For induced cells (solid box), the increase in the production and secretion of VtgCAT can be detected in the earliest indicated time point of 6 h after the E

2-induction. A peak showing 4-fold (˜400 g VtgCAT/U SEAP activity/h) and 5.5-fold (˜550 g VtgCAT/U SEAP activity/h) increase in VtgCAT were observed for NIH/3T3 and CHO-B cells, respectively. In contrast to COS-1, the lower level of induction is due to the lack of the large T-antigen which would otherwise amplify the signal. Importantly, the results indicate that the secretion of VtgCAT is not limited to a particular cell type. The secreted VtgCAT was assayed using CAT ELISA (Boehringer Mannheim). SEAP was measured as described by Tan et al.

Mol. Cell. Endocrinol. 123:146-161 (1996)). The VtgCAT amount was normalized using SEAP activity.

FIG. 10 shows the secreted VtgCAT expression profile over time in EPC (carp epithelial cells), a piscine cell line. (Open box, uninduced; solid box, estrogen-induced cultures.) A detectable amount of VtgCAT can be observed at 6 h. The EPC cells were cultured at a low temperature of 25° C. A 3-fold increase in VtgCAT can be detected by 24 h after estrogen-induction. The efficient production and secretion of VtgCAT was again demonstrated in this piscine cell line. The low metabolic rate of EPC would conceivably produce less VtgCAT, but this does not affect the secretion efficiency conferred by the SS of the invention. The secreted VtgCAT was assayed using CAT ELISA (Boehringer Mannheim). SEAP was measured as described in Tan et al.,

Mol. Cell. Endocrinol., 123:146-161 (1996)). The VtgCAT amount was normalized using SEAP activity.

FIG. 11 shows the construction of pYEX-VtgCAT. The pYEX-VtgCAT vector basically used the backbone from pYEX-S1 (Clontech) except that the original

K. lactis killer toxin signal sequence was replaced by Vtgss. Due to the limited number of cloning sites in the original pYEX-S1, an intermediate vector pBSPGK was constructed. pBSPGK consists of the PGK promoter, released via HindIII and Eco RI digestion from pYEX-S1, and subcloned into the corresponding sites of pBluescript™ II SK. This construct is a very useful intermediate as it provides more unique sites for cloning genes of interest. The VtgCAT fragment was released from pVtgCAT by Eco47 III and EagI digestion and subcloned into the SmaI and EagI sites of pBSPGK, respectively. The entire fragment harboring the PGK promoter and VtgCAT was released by HindIII and SacI digestion and subcloned into the similar site of pYEX-S1. Consequently, expression of pYEX-VtgCAT was driven by the strong constitutive PGK promoter and transcription terminated by the PGK terminator.

FIG. 12 shows the expression profile of VtgCAT in two different yeast transformants over 72 hours. The construct pYEX-VtgCAT was transformed into

S. cerevisiae strain DY150 using the single-step transformation method described by Chen et al.,

Curr. Genet. 21:83-84 (1992)). The transformants were selected on synthetic minimal medium agar containing all the required supplements except uracil. 100 ml YEPD medium (pH 5.0) contained in a 500 ml baffled flask was inoculated with a single colony of pYEX-VtgCAT and grown for 16 h at 30° C. with vigorous shaking. Subsequently, 2×50 ml of the yeast were collected independently by centrifugation for 10 min at 800×g. One 50 ml pellet was resuspended in 200 ml YEPD medium and grown for 72 h at 30° C. with vigorous shaking in a 1 L baffled flask. The other 50 ml pellet was resuspended in 200 ml minimal medium (MM) and grown as above. At 24, 48, and 72 h, two ml aliquots of the culture were removed. The yeast and medium were separated by centrifugation. The cell lysate was obtained by lysing the yeast pellet (resuspended in phosphate buffered saline) with glass beads, while the culture medium was collected and frozen without any pre-treatment. The pH of the culture was also monitored using appropriate universal pH indicator paper and adjusted to pH 5.0 using 1 M potassium phosphate buffer (pH 8.0). The amount of VtgCAT secreted into the culture medium and in the cell lysate were assayed using CAT ELISA (Boehringer Mannheim). Total soluble protein was used to normalize the data.

FIG. 13 shows the construction of pBADVtgblactKana. The -lactamase gene was isolated by PCR using blactfor (SEQ ID NO:5) (5′-CCGGGATCCAGAAACGCTGGTGAAAGTAA-3′) and blactrev (SEQ ID NO:6) (5′-GCGGCCGTTACCAATGCTTAATCAGTGAG-3′) using pBluescript™ II SK as a template. The PCR primers were designed to exclude the native -lactamase secretory signal. The PCR conditions used include a first cycle (94° C./45 sec; 50° C./30 sec; 72° C./1 min), 29 cycles (94° C./45 sec; 50° C./1 min; 72° C./30 sec) and final extension (72° C./5 min). The approximately 790 basepair PCR product was purified using a Qiaquick™ PCR Purification Kit (Qiagen) and digested with BamHI and EagI. This BamHI-EagI PCR fragment was subcloned into the BamHI and NotI sites of pVtgEGFP. To maintain optimal distance between the promoter and the ATG start codon, a second round of PCR was performed using BspSSFor (SEQ ID NO:7) (5′-GGGTCATGAGGGTGCTTGTACTAGCTCTT-3′) and blactrev primers and using pVtgblact as template. The second PCR follows a thermal regime of a first cycle (94° C./5 min; 60° C./1 min; 72° C./1 min), 29 cycles (94° C./45 sec; 60° C./30 sec; 72° C./30 sec) and a final extension of 72° C./5 min. The resulting PCR fragment of about 865 basepairs was purified as above and cloned into Bluescript™ II SK. This Vtg--lactamase fragment was released by BspHI and EcoRI digestion and subsequently subcloned into the NcoI and EcoRI sites of pBAD/myc-His B. The vector's -lactamase gene was released by BspHI digestion, and the vector ends were blunt-ended by Klenow enzyme. Then a kanamycin resistance gene obtained as a 2461 basepair DraI fragment from pGFP-N3 was inserted into the blunt-ended vector. The resulting plasmid, named pBADVtgblactKana was transformed into

E. coli. LM194 competent cells.

FIG. 14A shows the amino acid (SEQ ID NO:22) and nucleotide (SEQ ID NO:21) sequences at the junction of Vtgss and -lactamase in pBADVtgblactKana.

FIG. 14B shows the plasmid map of pBADVtgblactKana. The Vtg--lactamase gene was cloned downstream of the araBAD promoter (pBAD). The translation initiation site of Vtgss was constructed so that it is 9 basepairs from the optimized ribosome-binding site. The sequence junction of Vtg--lactamase was determined using Taqtrack™ sequencing. Both the BspHI site of Vtg--lactamase and NcoI site of the pBAD/myc-His B vector were destroyed.

FIG. 15 shows a plating assay of Vtg--lactamase induction. In the absence of inducer or at very low levels of inducer (arabinose), no colonies are observed. Increasing amounts of inducer provides for survival of increasing numbers of colonies. Thus, when a selectable marker gene is used as a reporter gene, the plating assay provides an easy semi-quantitative assay for reporter gene activity.

FIG. 16 shows the growth profile of the bacteria after induction with arabinose. The growth profile was monitored at O.D.

600nm. It is clear from the graph that a pBADVtgblactKana clone grown in RM medium (▪) with 0.2% glucose (i.e., no induction) exhibited a growth profile similar to the control LM194 host bacteria (). Addition of arabinose, at various concentrations, affected the growth of the pBADVtgblactKana clone. Best growth is seen only when 0.0002% arabinose was used (▴). Higher concentrations of arabinose (≧0.02%) resulted in a decrease in cell density. This is probably due to either (1) compromised growth of the cells due to overexpression of Vtg--lactamase or (2) toxicity due to a high level of Vtg--lactamase activity. The decrease in cell density has a pronounced effect on secretion of Btg--lactamase. The induction procedure was as described by the manufacturer (InVitrogen).

FIG. 17 shows the expression profile of Vtg--lactamase in the periplasmic space of bacteria. From the graph, no Vtg--lactamase was produced when a pBADVtgblactKana clone was cultured in RM medium with 0.2% glucose (). This is expected since no inducer is added. Addition of arabinose resulted in expression and accumulation of Vtg--lactamase in the periplasmic space. The early expression profile is dose-dependent, with the highest expression observed when 0.2% arabinose was used (♦). However, by 4 h, this dose-dependent profile is not observed. This could be directly due to the growth retardation as seen in FIG.

15. Importantly, the recombinant Vtg--lactamase is functional and exhibited properties similar to native -lactamase when measured using the standard calorimetric assay for -lactamase (Cohenford et al.,

Anal. Biochem., 168, 252-258 (1988)). The periplasmic space fraction was isolated as described by Laforet and Kendall (

J. Biol. Chem., 266, 1326-1334 (1991)). The values obtained from the various time points for pBADVtgblactKana were subtracted from those of the corresponding LM194 controls.

FIG. 18 shows the expression profile of Vtg--lactamase in culture medium. From the graph, no Vtg--lactamase can be detected when a pBADVtgblactKana clone was grown in 0.2% glucose (). Again, this is expected since no inducer is present. Surprisingly, Vtg--lactamase was detected in the culture medium, although not at as high a level as in the periplasmic space. But Vtg--lactamase in the culture medium was measured in the dilute medium. Even more surprising is that the Vtg--lactamase expression profile is reversed as compared with the Vtg--lactamase accumulation in the periplasm. The highest concentration of Vtg--lactamase was detected in the medium when the lowest concentration of 0.0002% arabinose was used in the culturing (▪). Obviously, the ratio of induction and growth profile of bacteria are important factors. Rapid and high accumulation of. Vtg--lactamase in the periplasmic space does not translate into efficient secretion into the medium. Instead, low-level induction is preferred. The medium was clarified of bacteria by centrifugation at 4° C. for 10 min at 12,000 ×g. The filtered medium was kept in ice until all the1 samples were collected for assay.

FIG. 19 shows the cloning strategy of pVtg-Gal. The -galactosidase coding sequences was PCR amplified using ForBamGal (SEQ ID NO:8) (5′-UCATGGATCCCGTGATTTCGTTGCCGGTCT-3′) and RevEagGal (SEQ ID NO:9) (5′-GCCGGGCAGACATGGCCTGC-3′) using pgal-promoter (Clontech) as template. The PCR product was digest with BanHI and EagI and ligated into similar sites in pVtgblact. Using pBluescript II SK as an intermediate vector, a ˜1.3 kb fragment harboring Vtgss and the 5′ coding region of -galactosidase was subcloned into pgal-promoter at the HindIII and EcoRV sites. This vector expressing secreted -galactosidase was designated pVtg-Gal.

FIG. 20 shows western blot analysis of Vtg-Gal using mouse anti--galactosidase. Fifty ug of medium was loaded and electrophoresed on a 10% SDS-PAGE gel. Lane 1: Day 5 medium; Lane 2: Day 4 medium; Lane 3: Day 3 medium; Lane 4: Day 2 medium; Lane 5: control medium; Lane 6: Blank; Lane 7: 20 ug of cell lysate from day 5 culture. The western blot was developed using Goat anti-mouse -HRP and Chemiluminescence Substrate.

FIG. 21 shows the rate of secretion of VtgCAT and Vtg-Gal in comparison with SEAP. COS-1 cells were transfected with 1 g of Vtg-fusion construct: control vectors in a ratio of 8:2 by liposome-mediated transfection with 5-6 l of Lipofectamine® reagent (Gibco-BRL) and incubated at 37° C. for 5 h as described by the manufacturer. After 36 h, the medium from cells of each time point was removed and replaced with 1 ml of fresh medium at an interval of 15 min over a period of 2 h. After the last time point, which should represent 0 min, an additional 1 h incubation was employed for all transfections to avoid low reading variations. At the end of incubation, the medium was retrieved and transferred to a microcentrifuge tube. Centrifugation was carried out as 12,000 ×g for 30 sec to pellet any detached cells present in the culture medium. 100 l of supernatant was used for ELISA. SEAP was assayed using Great Escape™ SEAP Fluorescence Detection Kit (Clontech) as described by the manufacturer. Both SSCAT and SS-Gal were assayed using CAT ELISA and -Gal ELISA (Boehringer Mannheim), respectively. For better comparison, the CAT ELISA and -Gal ELISA were adapted for fluorescence using DIG Fluorescence Detection ELISA (Boehringer Mannheim). The procedures were as described by the manufacturer. SEAP was determined using a fluorimeter (Perkin Elmer) at Ex

360nm and EM

449nm. VtgCAT and Vtg-Gal were detected at Ex

440nm and EM

550nm.

Claims

1. An isolated nucleic acid comprising a nucleotide sequence encoding a secretory signal sequence comprising the amino acid sequence SEQ ID NO:10, or variants of said amino acid sequence that comprise conservative replacements thereof that retain the biological activities of directing secretion of a fusion protein from a cell and cleavage of the secretory signal sequence from the fusion protein, wherein the variations in said variants

2. The isolated nucleic acid of claim 1, wherein said secretory signal sequence comprises the amino acid sequence SEQ ID NO:10, or variants of said amino acid sequence wherein the arginine at the second position is replaced by lysine and/or the glycine at the fifteenth position is replaced by alanine or valine and/or the aspartic acid at the sixteenth position is replaced by glutamic acid.

3. The isolated nucleic acid of claim 2, wherein the amino acid sequence is the amino acid sequence of SEQ ID NO:10.

4. The isolated nucleic acid of claim 1, wherein the nucleotide sequence encoding the secretory signal sequence is SEQ ID NO:11.

5. The isolated nucleic acid of claim 1, wherein the cell from which secretion is directed is a eukaryotic cell.

6. The isolated nucleic acid of claim 1, wherein the cell from which secretion is directed is a prokaryotic cell.

7. The isolated nucleic acid of claim 3, wherein the secretory signal sequence is cleaved between the glycine and aspartic acid residues in the valine-glycine-aspartic acid-glutamine portion thereof.

8. An isolated nucleic acid comprising a nucleotide sequence encoding a fusion protein comprising a secretory signal sequence and a desired heterologous protein, wherein said secretory signal sequence, comprises the amino acid sequence SEQ ID NO:10, or variants of said amino acid sequence that comprise conservative replacements thereof that retain the biological activities of directing secretion of a fusion protein from a cell and cleavage of the secretory signal sequence from the fusion protein, wherein the variations in said variants

9. The isolated nucleic acid of claim 8, wherein said secretory signal sequence comprises the amino acid sequence SEQ ID NO:10, or variants of said amino acid sequence wherein the arginine at the second position is replaced by lysine and/or the glycine at the fifteenth position is replaced by alanine or valine and/or the aspartic acid at the sixteenth position is replaced by glutamic acid.

10. The isolated nucleic acid of claim 9, wherein said secretory signal sequence comprises the amino acid sequence of SEQ ID NO:10.

11. The isolated nucleic acid of claim 10, wherein the nucleotide sequence encoding the secretory signal sequence is SEQ ID NO:11.

12. The isolated nucleic acid of claim 8 wherein said desired heterologous protein is a reporter protein.

13. The isolated nucleic acid of claim 12, wherein the reporter protein is selected from the group consisting of chloramphenicol aminotransferase, green fluorescent protein or another aequorin, -amylase, -lactamase, luciferase, glucuronidase, alkaline phosphatase, and -galactosidase.

14. The isolated nucleic acid of claim 8 wherein said desired protein is a lipopolysaccharide-binding protein.

15. The isolated nucleic acid of claim 14, wherein the lipopolysaccharide-binding protein is Factor C.

16. A recombinant vector comprising the isolated nucleic acid of any one of claims

8-

11.

17. A host cell transformed with the recombinant vector of claim 16.

18. The recombinant host cell of claim 17, wherein said cell is selected from the group consisting of a bacterial cell, a COS cell, a Chinese hamster ovary (CHO) cell, a NIH/3T3 cell, a Schneider 2 cell, a

S. cerevisiae cell, and an endothelial progenitor cell (EPC).

19. A method for producing a desired protein comprising

20. A fusion protein comprising

21. The fusion protein of claim 20, wherein said secretory signal sequence polypeptide comprises the amino acid sequence SEQ ID NO:10, or variants of said amino acid sequence wherein the arginine at the second position is replaced by lysine and/or the glycine at the fifteenth position is replaced by alanine or valine and/or the aspartic acid at the sixteenth position is replaced by glutamic acid.

22. The fusion protein of claim 21, wherein said secretory signal sequence comprises the amino acid sequence of SEQ ID NO:10.

23. The fusion protein of claim 22, wherein the nucleotide sequence encoding the secretory signal sequence is SEQ ID NO:11.

24. The fusion protein of claim 20, wherein the heterologous polypeptide is a lipopolysaccharide binding protein.

25. The fusion protein of claim 20, wherein the heterologous polypeptide is a protein selected from the group consisting of chloramphenicol aminotransferase, green fluorescent protein or another aequorin, -amylase, -lactamase, luciferase, glucuronidase, alkaline phosphatase, and -galactosidase.