Characterization of zebrafish vitellogenin gene family for potential development of receptor mediated gene transfer method 4

Chapter Receptor-mediated gene transfer Chapter Preliminary studies on receptormediated gene transfer using Vtgpolylysine as DNA carrier and identification of receptor-binding domain in fish Vtg 137 Chapter Receptor-mediated gene transfer Abstract The potential of receptor-mediated gene transfer using vitellogenin protein as DNA carrier was explored In these preliminary experiments, tilapia Vtgs were induced, purified and labeled with 125 I After injection, purified Vtgs could be preferably taken up by ovaries By modification with N-succinimidyl 3-(2-pyridyldithio)-propionate (SPDP), three types of Vtg-poly-L-lysine conjugates were constructed and used in complexes preparation However, the efficiency of mediating DNA uptake from the Vtg-poly-L-lysine conjugates by oocytes was not significantly higher than those by other tissues Possible reasons for this were discussed Furthermore, recombinant Vtg fragments covering essentially the full Vtg sequence were produced in E coli and attempts of determining receptor-binding domains were also made by in vivo binding assay, though inconclusive results were observed 138 Chapter Receptor-mediated gene transfer 4.1 Introduction Transgenic fish are not only an important experimental tool for developmental analyses of gene expression and function, but also have enormous potential in aquaculture In 1985, Zhu et al successfully made transgenic gold fish (Carassius auratus) through gene transfer by microinjection Since then, the production of transgenic fish has been reported in many fish species and a variety of gene delivery methods have been developed and successfully employed (Fletcher and Davies, 1991; Maclean and Rahman, 1994; Gong and Hew, 1995; Chen et al., 1996) Common gene delivery methods used in transgenic fish research include microinjection, electroporation, sperm carrier and particle bombardment Each method has its advantages and drawbacks For example, microinjection is a popular gene delivery approach used by many researchers but requires special equipment and skilled personnel Moreover, only a limited number of eggs can be injected at a time for most fish species In contrast, other gene delivery methods such as electroporation and particle bombardment are more efficient in dealing with a large number of eggs but the germ line integration rate of foreign genes is usually very low and special equipment is also required Thus, new gene delivery methods need to be developed that overcome these deficiencies It is well known that receptor-mediated endocytosis (RME) provides a major pathway for trafficking of extracellular molecules or ligands into animal cells Based on the RME process, a novel gene transfer method designated “receptor-mediated gene transfer” was proposed (Wu and Wu, 1987) In their experiment, foreign DNA was transported into hepatocytes by an asialoglycoprotein receptor-mediated pathway after administration of complexes formed between DNA and asialoorosomucoid (ASOR)-poly-L-lysine (pLys) 139 Chapter Receptor-mediated gene transfer conjugates (Wu and Wu, 1987) The main advantage of the receptor-mediated gene transfer is that it can be used to target specific cells in vivo after intravenous injection However, no attempts have been made to apply this gene delivery approach in fish In fish, Vtgs are synthesized in the liver under the control of E2 and incorporated into oocytes via receptor-mediated endocytosis (Selman and Wallace, 1982; Tyler et al., 1987) Thus, Vtg is a candidate protein for a DNA carrier to use in transforming fish oocytes through receptor-mediated gene delivery approach If this gene delivery approach is feasible in fish, foreign genes could be specifically introduced into the oocytes after injection of the Vtg-DNA complexes into blood circulation of female fish and high percentage of transgenic offspring would be expected This gene delivery method is effective, and does not depend on experienced personnel or special equipment In this study, receptor-mediated gene transfer will be tested in red tilapia (Oreochromis mossambica) The reason for shifting the experimental model from zebrafish to tilapia is that it is much easier to inject experimental materials through caudal artery of the tilapia than of the zebrafish Furthermore, the full-length cDNA sequence of tilapia (Oreochromis aureus) vtg1 became available from GenBank during the project and it encodes a Vtg with the homologous subdomains I-V (see Table 2-4 in Chapter 2) Thus, the tilapia vtg1 is a potential orthologue of the zebrafish vtg2 and it also encodes a Vtg that is complete in primary structure 140 Chapter Receptor-mediated gene transfer 4.2 Materials and Methods 4.2.1 E2 induction and blood serum isolation Female red tilapia (Oreochromis mossambica) (body weight 400 – 500g) were purchased from a local fish farm and acclimated in a tank for two weeks They were fed daily with commercial fish food E2 stock solution was prepared as described in Chapter (Section 3.2.1) and was injected intraperitoneally on days 1, 5, and 14 at µg E2/g body weight according to Ding et al (1989) Blood samples were collected through the caudal artery from control and treated fish on days 4, 16 and 20, respectively, using pre-chilled syringes For serum isolation, blood was clotted on ice for 10-15 prior to centrifugation at 6000g for 10 at °C Aprotinin (Sigma) was added to the serum at a final concentration of 40 µg/ml and the fish serum was stored at – 70 °C prior to purification 4.2.2 Vtg purification by anion-exchange chromatography Vtg was purified from tilapia serum by anion-exchange chromatography according to Chan et al (1991) with modifications Briefly, to adjust the pH and ionic strength, buffer A (50 mM Tris-HCl, pH 8.0) was added to the fish serum at a ratio of : (v/v) After filtration, ml of diluted serum sample was injected into an UNO Q-1 column (Bio-Rad) integrated in a HPLC system (Pharmacia) Unbound proteins were washed away with 2.5 ml of buffer A (~ times of bed volume) at a constant flow rate of ml/min For elution, a gradient of – 35% of buffer B (50 mM Tris-HCl, pH 8.0, M NaCl) was applied over 15 ml, followed by holding at 35% of buffer B for ml (see Fig 4-3B) Elutes were collected in ml fraction and stored at – 80 ºC 141 Chapter Receptor-mediated gene transfer 4.2.3 Iodination of Vtg Purified tilapia Vtg was labeled with 125 I using a solid phase oxidizing agent, 1,3,4,6- tetrachloro-3α, 6α-diphenyl glycouril (Iodogen, Sigma) mediated method (Salacinski et al., 1981) Briefly, an Iodogen tube was prepared by dispensing 100 µl of Iodogen solution (0.1 mg/ml in chroloform) onto the bottom of a 12x75 mm glass tube, followed by vacuum evaporation A time course of radioiodination was determined based on the method of Rudick (1998) Briefly, the Iodogen tube was rinsed with ml of 50 mM TrisHCl (pH 8) to remove any loose flakes of Iodogen Then, the following components were added with gentle swirling: 90 µl of 50 mM Tris-HCl (pH 8.0), 10 µl of purified tilapia Vtg (20 µg) and 0.5 µl of Na125I (50 µCi) Immediately, µl of solution was removed and spotted onto a nitrocellulose sheet This process was repeated at intervals for a total of 21 Finally, all nitrocellulose sheets were washed in 50 ml of washing buffer (25 mM Tris-HCl, 192 mM glycine, 12.5 mM NaI, 20% methanol, pH 8.3) and countered by a Gammer counter (1470 Wizard, Wallac) Final labeling reaction was performed according to the optimal duration determined and 20 µg of purified Vtg was radioiodinated at room temperature The 125 I-labeled Vtg was purification on a Sephadex G-25 column (PD-10, Pharmacia Biotech) and stored at °C prior to use 4.2.4 Synthesis of Vtg-pLys conjugates Purified tilapia Vtg was coupled to poly-L-lysine (pLys) through a disulfide bond after modification by a heterobifunctional reagent, N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP, Pierce) (Jung et al., 1981; Wagner et al., 1990) In this experiment, 142 Chapter Receptor-mediated gene transfer Step Vtg Step H2N NH2 poly-L-lysine + SPDP + SPDP O Vtg HN-C-CH2-CH2-S-S- + DTT N O HS-CH2-CH2-C-NH O Vtg Step HN-C-CH2-CH2-S poly-L-lysine O S-CH2-CH2-C-NH poly-L-lysine + N S (pyridin -2-thione) H Fig 4-1 Flow chart depicting the formation of Vtg-poly-L-lysine conjugates Three steps are included in the process, which is described in detail in Materials and Methods (Section 4.2.4) SPDP, N-succinimidyl 3-(2-pyridyldithio)-propionate; DTT, dithiothreitol (adapted from Jung et al., 1981; Wagner et al., 1990) 143 Chapter Receptor-mediated gene transfer two types of pLys with an average chain length of 36 (pLys36, MW 7500 Da, Sigma) and 144 lysine monomers (pLys144, MW 30,100 Da, Sigma) were used Stock solutions of 3.33 nmol/µl for pLys36 and nmol/µl for pLys144 were prepared in sodium phosphate buffer (0.1 M sodium phosphate, pH 7.8, 0.1 M NaCl) The construction process included three steps shown in Fig 4-1 Results are summarized in Table 4-3 Step Modification of Vtg with 3-(2-pyridyldithio)-propionate by SPDP First, a buffer exchange with sodium phosphate was performed on a PD-10 column for the HPLC purified Vtgs and the Vtg solution was concentrated using KwikSpin Micro Ultrafiltration Units (30 kDa MWCO, Pierce) For modification of Vtgs used for type II conjugates (see Table 4-3), µl of SPDP stock solution (2 nmol/µl in 100% ethanol) was gradually added to 430 µl of Vtg solution (0.86 nmol) The reaction mixture was vigorously mixed and kept at °C for hr Modified Vtgs were purified on PD-10 column with sodium phosphate buffer After an aliquot of modified Vtg was reduced by dithiothreitol (DTT), the amount of dithiopyridine linkers per Vtg molecule was determined based on a molar absorbance coefficient of 8.08 x 103 M-1• cm-1 at 343 nm for the released product pyridin2-thione (Carlsson et al., 1978) For modification of Vtgs used for conjugates of types I and III, the molar ratios between Vtg and SPDP were adjusted accordingly (see Table 4-3) and the similar process was followed SPDP modified Vtgs were stored at °C before use Step Modification of pLys with 3-mercaptopropionate by SPDP and DTT treatment First, for modification of pLys36 by SPDP at a molar ratio of pLys36 to SPDP of 1:1, 15 µl of SPDP (20 nmol/µl) was mixed with 90 µl of pLys36 (3.33 nmol/µl) and 83 µl of sodium 144 Chapter Receptor-mediated gene transfer phosphate buffer The solution was vigorously mixed and kept at room temperature for 2.5 hr, followed by gel filtration on PD-10 column with sodium acetate buffer (20 mM sodium acetate, pH 5.2, 0.1 M NaCl) For detection of the modified pLys, absorption at 211 nm was measured for each elute For modification of pLys144 by SPDP at a molar ratio of pLys144 to SPDP of 1:2, 30 µl of SPDP (20 nmol/µl) was mixed with 150 µl of pLys144 (2 nmol/µl) and the products were purified accordingly afterwards Two standard curves were prepared for quantification of pLys: 1) Y = 0.1035X-0.0061 for pLys36 and 2) Y = 0.3633X + 0.0139 for pLys144 (X: concentration of pLys in pmol/µl, Y: absorption at 211 nm) The amount of dithiopyridine linkers in modified pLys was determined as described in step Second, SPDP modified pLys was reduced by DTT to form 3mercaptopropionate modified pLys Briefly, 500 µl of SPDP modified pLys36 (8.21 nmol) or pLys144 (3.01 nmol) was mixed with 12.5 µl of M DTT and the solution was kept under N2 for hr at room temperature After gel filtration on PD-10 column with sodium acetate buffer, the 3-mercaptopropionate modified pLys was stored at – 20 °C until use Step Synthesis of Vtg-pLys conjugates Three types of Vtg-pLys conjugates were synthesized, Vtg-pLys36 (type I), Vtg-pLys144 (H) (type II) and Vtg-pLys144 (L) (type III) (see Table 4-3) Briefly, for making type I conjugates, ml (1.95 nmol) of 3-(2pyridylditho)-propionate modified Vtg (with 4.9 linkers per molecule) was mixed with 100 µl (1.93 nmol) of 3-mercapto-propionate modified pLys36 and the reaction was kept under N2 at °C for ~20 hr For making type II conjugates, the reaction was performed by mixing 351 µl (0.8 nmol) of modified Vtg (with 10.1 linkers per molecule) with 175 µl (0.832 nmol) of 3-mercaptopropionate modified pLys144 Similarly, for making type III 145 Chapter Receptor-mediated gene transfer conjugates, 467 µl (0.214 nmol) of modified Vtg (with 1.1 linker per molecule) was mixed with 50 µl (0.238 nmol) of 3-mercapto-propionate modified pLys144 The Vtg-pLys conjugates were separated from uncoupled 3-mercaptopropionate modified pLys by gel filtration on a Bio-Gel P-100 column (with exclusion limit of 100 kDa) and eluted using 20 mM HEPES, pH 7.4, 0.15 M NaCl The coupling degree was estimated based on the increased OD343 value as described above 4.2.5 Formation of complexes between Vtg-pLys conjugates and DNA Preparation of complexes between Vtg-pLys conjugates and DNA and gel retardation assay were carried out according to Wagner et al (1990, 1991) In order to form the complexes, Vtg-pLys and DNA were directly mixed at a certain ratio in a buffer containing 20 mM HEPES (pH 7.4) and 0.15 M NaCl, followed by incubation for hr at room temperature The optimal ratio for neutralization between Vtg-pLys conjugates and DNA was determined by gel retardation assay Briefly, µg of DNA (100-bp long, cut from a plasmid by restriction enzymes) was labeled by [α-32P]dCTP using Nick Translation Reagent Kit (BRL) according to the manufacturer’s protocol and purified by a NICK Column (Pharmacia Biotech) A series of complexes were prepared between 0.45 µl of 32 P-labeled DNA (~ 0.02 pmol) and increasing amount of Vtg-pLys conjugates After that, the complex mixture was loaded into a 1% agarose gel and resolved by gel electrophoresis in 1x TAE buffer at 30 V for hr Finally, the agarose gel was dried and autoradiography was performed at – 70 ˚C with Kodak's BioMax MS film 146 Chapter Receptor-mediated gene transfer 24 hr, DNA~Vtg-pLys144 24 hr, DNA~Vtg-pLys144 24 hr, DNA 24 hr, DNA 24 hr, DNA 48 hr, DNA~Vtg-pLys144 48 hr, DNA~Vtg-pLys144 48 hr, DNA 48 hr, DNA 60 55 % of total radioactivity radioactivity % of recovered of whole fish 50 45 40 35 30 25 20 15 10 24 48 ovary 24 48 liver 24 48 gut 24 gill 48 24 48 heart 24 48 spleen 24 48 kidney 24 48 others remains Fig 4-6 Relative radioactivities in seven tissues and the other parts of fish examined at 24 and 48 hr after injection with complexes (formed between 32P-labeled 100-bp DNA and Vtg-pLys144 (H) conjugates) and 32P-labeled 100-bp DNA, respectively (~ 200,000 cpm/fish) Relative radioactivity is presented as mean ± SD % For information about the number of fish examined and their mean gonad indexes, see legend of Table 4-4 169 Chapter Receptor-mediated gene transfer aprotinin had been added to the serum, it was diluted in the subsequent steps Furthermore, in some modification reactions by SPDP, various degrees of precipitation for Vtgs were observed Thus, degradation or denaturation of Vtgs was a concern in this experiment 2) Destruction of the receptor-binding function by SPDP modification It is known that the hydroxysuccinimide ester in SPDP reacts with primary amino groups to give amide bonds and the 2-pyridyl disulphide group in SPDP reacts with aliphatic thiols to form aliphatic disulphides (Carlsson et al., 1978) Based on the amino acid sequence of O aureus Vtg1, among its 1,788 amino acid residues, 118 are lysines and 104 arginines, which are two kinds of targets for SPDP modification It has been demonstrated that for Vtg of an insect Locusta migratoria, progressive modification of its lysyl and arginyl residues caused reduction (when modifying 5-15% of the lysine or arginine residues) or loss (when modifying 34-50% of the lysine or arginine residues) of the capacity for the derived Vtg to bind with its receptor (Roehrkasten and Ferenz, 1992) The authors further suggested that these residues (lysine or arginine) are involved in expression of a Vtg receptor recognition site (Roehrkasten and Ferenz, 1992) A similar phenomenon was also observed in chicken Vtg and in human apoB and apoE in which their abilities to interact with receptors were abolished after reductive methylation of the lysine and arginine residues of these proteins took place (Steyrer et al., 1990 and references within) In the type II Vtg-pLys144(H) conjugates, the Vtg moieties were modified by SPDP with a high modification extent, resulting in 10.1 dithiopyridine linkers per Vtg molecule which means about 4.5% of the total number of lysine and arginine residues in the tilapia Vtg were modified Thus, similar adverse effects caused by SPDP modification may result in reduction or loss of effective interaction of Vtgs with their receptors 170 Chapter Receptor-mediated gene transfer 3) Improper charge ratio of DNA to poly-L-lysine Zauner et al (1998) pointed out that in order to allow receptor-mediated uptake in mammalian cells, the DNA~poly-L-lysine complexes have to be negatively charged; otherwise, absorptive endocytosis will take place with or without the presence of the ligand in the complex During the preparation of complexes between DNA and Vtg-pLys conjugates in this experiment, excessive amounts of Vtg-pLys conjugates were used based on the consideration that neutral particles intend to aggregate in solution (Wagner et al., 1991) However, this strategy may have had some unexpected drawbacks, such as potential non-specific uptake of the complexes due to the net positive charges of the complexes Successful gene transfer using vitellogenin as a ligand through receptor-mediated approach has not been previously reported The present initiative of developing this gene delivery approach in tilapia will provide some ideas about how to improve the construction of conjugates and complexes for the next phase of the project Obviously, attention must be paid to minimizing protein degradation and denaturation, minimizing the extent of modification of Vtg by SPDP and rendering slightly net negative charges in the complexes Alternatively, the conjugation process may be circumvented by expression of recombinant Vtg fragments fused with a polylysine tail which serves as a DNA binding moiety (Zeng et al., 2004) Thus, the risk of over-modification of Vtgs by SPDP potentially can be completely eliminated 4.3.6 Expression of recombinant tilapia Vtg fragments in E coli An alternative to using native purified Vtg in conjugate preparation, efforts were also made to express recombinant tilapia Vtg fragments in E coli There are two major 171 Chapter Receptor-mediated gene transfer advantages of using recombinant Vtgs: a) the availability of Vtg proteins in large quantity without sacrificing fish and b) the potential to reduce the size of the carrier protein can be reduced, facilitating experimental manipulation of the protein Thus, the tilapia Vtg proteins were expressed as seven recombinant protein fragments in E coli in this study Briefly, eight pairs of primers were designed (Table 4-1) and eight tilapia vtg cDNA fragments were amplifed by RT-PCR (Fig 4-8A) Five cDNA fragments (LVIa, LVIb, LVIc, PV and LVII) cover almost the entire vtg coding region with overlaps between adjacent fragments of LVIc, PV and LVII The remaining three (LVIb1, LVIb2 and LVIb3) cover the LVIb region and the gap between LVIa and LVIb (Fig 4-7) To produce recombinant GST-Vtgs, cDNA fragments LVIa, LVIb, LVIc, PV amd LVII were separately inserted into an expression vector pGEX-2TK Four GST-Vtg fusion proteins were expressed successfully in E coli, namely GST-LVIa (~50 kDa), GST-LVIb (~78 kDa), GST-LVIc (~60 kDa) and GST-LVII (~67 kDa) (Fig 4-8B) However, bacteria transformed with pGST-PV grew very slowly and no obvious recombinant GST-PV proteins were induced even after supplement with or 10 mM of L-serine in the culture medium (Fig 4-8B) After affinity purification, soluble native GST-Vtg fusion proteins were obtained except for GST-PV which appeared as multiple truncated fragments after SDS-PAGE analysis (Fig 4-8C) The proportions of soluble to insoluble proteins were low for GST-LVIb, GST-LVIc and GST-LVII except for GST-LVIa (data not shown) In addition, even in the presence of proteolytic enzyme inhibitors such as PMSF and aprotinin, protein degradation was still severe especially for those with high molecular weight such as the GST-LVIb (Fig 4-8C) To overcome the problems of low yield of 172 Chapter Receptor-mediated gene transfer SP LVI 1F MRVLVLALAVALAVGDQSNLAPGFASVKTYMYKYEAVLMGGLPEEGLARAGVKIRGKVLISATSANDYILKLVDPQLLEY SGIWPKDPFHPATKLTTALATQLSTPIKFEYTNGVVGRLAAPPGVSTTVLNIYRGIINLLQLNVKKTQNVYEMQESGAHG 1R VCKTNYVIREDARAERIHLTKTKDLNHCQEKIMKAIGLEHVEKCHDCEARGKSLKGTASYNYIMKPAPSGSLIMEAVARE 2aF Site A 2F VIEFSPFNILNGAAQMESKQILTFLDIENTPVDHARYTYVHRGSLQYEHGSEILQTPIHLLRVTHAEAQIVSTLNHLVAS 80 160 240 320 NVAKVHEDAPLKFVELIQVMRVARFETIESLWAQFKSRPDHRYWLLNAVPHIRTHAALKFLIEKLLANELSETEAAMALL 2bF 2aR ECLHSVTADQKTIELVRSLAENHRVKRNAVLNEIVMLGWGTVISRFCKAQPSCSSDLVTPVHRQVAEAVETGDIDQLTVT 400 LKCLDNAGHPASIKTIMKFLPGFGSAAARVPLKVQVDAVLALRRIAKREPKMVQEIAAQLLMEKHLHAELRMVAAMVLFE * Site B * * 2cF 2bR TKLPVGLAASISTALIKEKNLQVVSFVYSYMKAMAKTTSPDHVSVAAACNVALRFLNPKLGRLNFRYSRAFHVDTYNNAW 560 480 640 MMGAAASAVLINDAATVLPRMIMAKARTYMAGAYVDAFEVGVRTEGIQEALLKRRHENSENADRITKIKQAMRALSEWRA 2R 2cR 3F NPSSQALASMYVKVFGQEIAFANIDKSKVDQLIQFASGPLRNVFRDAVNSVLSGYATHFAKPMLLGELRLILPTTVGLPM 720 EISLITSAVTAASVDVQATVSPPLPVNYRVSQLLESDIQLRATVAPSLAMQTYAFMGVNTALIQAAVMTKAKVYTAVPAQ 880 IKARIDIVKGNLKVEFLSLQGINTIASAHAETVAIARNVEDLPAARSTPLISSETASQLSKASLNSKISRMASSVTGGMS 960 ASSEIIPADLPSKIGRKMKLPKTYRKKIRASSRMLGFKAYAEIESHNAAYIRDCPLYALIGKHAASVRIAPASGPVIEKI LVI PV 4F 3R EVEIQVGDKAAENMIKAIDMSEEEEALEDKNVLLKIKKILAPGLKNTTSSSSSSSSSSSSSSSSNKSSSSSSRSSSSQSS Site C * SSRSHRSRSRKSQSSSSQSSRSPSSSSSSSSSSSSRSSSRSSSRSSSRSSSRSSSRSRTKMADIVAPIITTSTRVSSSSS 1040 1200 PV RSASNSSSSSASYLLSSSKRRSRSRSSSSSSSSSSSSSSSSSSSSSSSKNSKRSKSSNSKSSSSRSSRRSAQSKQQLLAL LVII 1280 KFRKNHVHRHAISTQRGSSHSSARSFDSIYNKAKYLANTLTPAMSIAIRAVRVDHKVQGYQLAAYLDKQTNRLQLIFARV 5F 4R AEKDNWRICADIVQLSSHKLMAKTAWGAECKQYSTMIVAETGLLGHEPAARLKLTWDKLPGSIKHYAKRALKSIVPIAQE 1360 1440 YGVNYAKAKNPRNQIKLTVAVATETSMNIVLNTPKAIVYKRGVCLPVALPIGNTAAELQATRDNWADKMSYLVTKANAVE 1520 CSLINNTLTTFNNRKARDELPHSCYQVLAQDCTPELKFMVLLKKDQIQDQNQINVKISDIDVDMYRKNNAIAVMVNGVEI 1600 PNSNLPYLHPSGNIHIRQSNEGITLNAPSHGLQEVFLGFNELRVKVADWMKGKTCGACGTASGNVGDEYRTPSEQVTKDA 1680 ISYAHSWVLSSNTCRDPSECSIKQESVKLEKRVIFEGVESKCYSVEPVLQCLPGCIPVRTTTVNVGFHCLPSDTTVDRSG 5R LSSFFEKSIDLRDTAEAHLACRCTPQCA* 1760 800 1120 1788 Fig 4-7 Deduced amino acid sequence of tilapia (Oreochromis aureus) Vtg1 (GenBank accession No AF017250) The locations of eight pairs of primers used in amplification of eight vtg cDNA fragments from tilapia (Oreochromis mossambica) liver total RNA by RT-PCR were indicated Double-headed arrows indicate the division of four regions, signal peptide (SP), lipovitellin I (LVI), phosvitin (PV) and lipovitellin II (LVII) A reported N-terminal receptor-binding region (158-242 amino acid residues), including eight critical residues (site A), is in italic letters (Li et al., 2003) Potential receptorbinding sites (sites B and C) are underlined The conserved lysine residues in the above three sites are highlighted by bold letters Proline residues flanking the potential receptorbinding sites are marked by asterisks 173 Chapter Receptor-mediated gene transfer A bp LVIa M LVIb LVIc PV LVII LVIb1 LVIb2 LVIb3 (690 bp) (1467 bp) (963 bp) (1139 bp) (1304 bp) (625 bp) (585 bp) (547 bp) 2072 1500 * * 600 * * * * * * 100 B kDa M V I a c b -P VI VI VI VI -L ol T-L T-L T-L GST r ST S S p nt S pG pG pG Co pG mM 10 mM 220 67 60 36 C kDa M 220 Ia V -L ST G G -L ST Ib V Ic II PV V V T-L -L S G ST ST G G ⊲ 67 60 36 * * * Fig 4-8 Amplification of eight tilapia vtg cDNA fragments and expression of four of them as GST fusions in E coli A: Eight vtg cDNA fragments (marked by asterisks) were amplified by RT-PCR The names and lengths of cDNA fragments are listed above B: SDS-PAGE analysis of whole lysates from bacteria transformed with pGEX-2TK (control) and recombinant pGEX-2TK vectors harboring vtg cDNA fragments L-serine of or 10 mM was supplemented to the culture medium for synthesis of GST-PV Bands corresponding to GST-Vtgs are marked by arrowheads The GST band in the control is not shown C: SDS-PAGE analysis of purified GST-Vtgs (marked by arrowheads) eluted from glutathione Sepharose 4B beads mM (1) or 0.5 mM (2) of L-serine was added to the medium for synthesis of GST-PV Asterisks and an empty arrowhead indicate degraded proteins and the size of predicted GST-PV, respectively 174 Chapter Receptor-mediated gene transfer soluble proteins and protein degradation, the 6xHis expression vector pRSET-A (Invitrogen) was used, and a ~ 21 kDa reduction in molecular weight was expected for each expressed 6xHis-tagged recombinant Vtg fragment when compared with GST-Vtg fusions To facilitate the labeling of 6xHis tagged recombinant Vtgs by 32 P, the original pRSET-A vector was modified with an insertion of a protein kinase recognition motif coding sequence 5’-CGTCGTGCATCTGTT-3’ (Fig 4-2C) Respective vtg cDNA fragment was then cloned into this modified vector and seven 6xHis-tagged Vtg fragments were expressed in E coli (Fig 4-9A,B) Native soluble 6xHis-tagged Vtg fragments were further purified by affinity purification using Talon Metal Affinity Resins (Fig 4-9A,B) The estimated molecular weight was 29.1 kDa for 6xHis-LVIa, 57.3 kDa for 6xHis-LVIb, 39 kDa for 6xHis-LVIc, 45.4 kDa for 6xHis-LVII, 27.7 kDa for 6xHis-LVIb1, 25.1 kDa for 6xHis-LVIb2 and 24.3 kDa for 6xHis-LVIb3 Protein degradation was still present (Fig 4-9A, right panel) Purified 6xHis-tagged Vtg fragments were further confirmed by Western blot analysis using Anti-HisG antibody (Fig 4-9C) Western blot analysis also indicated that most of the degraded proteins may also bear 6xHis-tags (Fig 4-9C) In summary, seven 6xHis-tagged recombinant Vtg fragments were synthesized in E coli and all of them belong to the lipovitellin part (lipovitellin I or II) of the tilapia Vtg protein 4.3.7 No preference was observed for ovary incorporation of 32 P-labeled recombinant Vtg fragments Before the recombinant Vtg fragments can be used for conjugate preparation, the location of the receptor-binding domain should be determined Thus, four of the seven 6xHis-Vtg fragments (6xHis-LVIa, 6xHis-LVIb, 6xHis-LVIc and 6xHis-LVII), covering almost the 175 Chapter Receptor-mediated gene transfer A Bacteria total lysate kDa M nt Co l ro p ’-L ET RS 250 150 a VI SE pR -L T’ Purified 6xHis-Vtg fragments Ic II V V Ia Ic Ib II -L -L V V V T’ T’ LV -L -L -L is is is isSE SE H H H H pR pR 6x 6x 6x 6x b VI 100 75 50 * 37 * 25 * * * B Bacteria total lysate kDa M l tr o on C Purified 6xHis-Vtg fragments Ib Ib Ib VIb Ib Ib Ib Ib LV LV ’-L ’-LV LV T ET LV s-LV LV T’ T’ isi isisSE RSE RS SE H H H H pR pR p p 6x 6x 6x 6x 250 150 100 75 50 37 25 10 176 C kDa 6x H isLV Ia 6x H isLV 6x Ib H isLV Ic 6x H isLV 6x II H is LV 6x Ib H isLV 6x Ib H isLV Ib Chapter Receptor-mediated gene transfer 250 150 100 75 * 50 * * 37 * 25 * * D is xH kDa 250 150 100 75 V -L b Ic II V VI LV L s-L isisH H Hi ST 6x 6x 6x G Ia 50 37 * 25 * * * * * * Fig 4-9 Expression, immunodetection and radioisotope labeling of 6xHis-Vtg fragments A,B: SDS-PAGE analysis of seven recombinant 6xHis-tagged Vtg fragments expressed in E coli Total lysate from bacteria transformed with pRSET’ was used as a control Names of expression constructs and recombinant Vtg fragments are listed above the lanes Bands corresponding to 6xHis-Vtg fragments are marked by arrowheads Asterisks (in A) mark the degraded proteins C: Seven recombinant 6xHis-Vtg fragments (marked by arrowheads) were detected by anti-HisG antibody (1:5000 dilution) in Western blot analysis Degraded 6xHis-tagged proteins are marked by asterisks D: SDS-PAGE analysis of four [γ-32P]ATP labeled 6xHis-Vtg fragments and GST (marked by arrowheads) (~ 95000 cpm/lane) Degraded proteins are marked by asterisks 177 Chapter Receptor-mediated gene transfer entire Vtg molecule except for the PV domain were labeled with [γ-32P]ATP and injected into fish blood vessel to determine whether one would be preferentially taken up by the ovary As shown in Fig 4-9D, all four 6xHis-Vtg fragments can be successfully labeled with 32P in vitro, indicating all fragments bear the inserted protein kinase recognition site As a control, the GST protein, which is encoded by the pGEX-2TK vector bearing an authentic sequence coding for a protein kinase site, was also labeled by [γ-32P]ATP (Fig 4-9D) Despite certain degree of protein degradation observed after the labeling reaction, full-length recombinant proteins were predominantly present, except for 6xHis-LVIb, at least 50% of which appeared to be degraded after the labeling reaction (Fig 4-9D) To localize the receptor-binding domain in tilapia Vtg, the above four 32 P-labeled Vtg fragments were injected separately into caudal artery fish blood circulation and relative radioactivities in the ovaries and other tissues were compared, assuming the radioactivity recovery rates were similar for all injection groups when examined at 48 hr after injection As shown in Table 4-5 and Fig 4-10, at 48 hr after injection, high relative radioactivities were observed in ovaries and “other” parts of fish in all injected groups except for the one receiving 6xHis-LVIb In the ovaries, the relative radioactivities varied (25.2-42.3%) after injection with 32 P-labeled 6xHis-LVIa, 6xHis-LVIc, 6xHis-LVII or GST (Table 4-5) Athough the value was low (25.2%) in the ovary after injection with 6xHis-LVII, there was no significant difference in ovarian relative radioactivities after injection with each 32 P-labeled recombinant Vtg fragment or GST by Student T-test After injection with 32P- labeled 6xHis-LVIb, extremely low relative radioactivity (5.5%) was observed in the ovary, which may be caused by the severe degradation of this recombinant protein after 178 Chapter Receptor-mediated gene transfer Table 4-5 Relative radioactivities in seven tissues and the other parts of fish examined at 48 hr after injection with 32P-labeled recombinant proteins of 6xHis-LVIa, LVIb, LVIc and LVII, and 32P-labeled GST, respectively Group Tissue Ovary Liver Gut Gill Heart Spleen Kidney Others 6xHis-LVIa† (n = 4) 6xHis-LVIb (n = 4) 6xHis-LVIc (n = 4) 6xHis-LVII (n = 4) 33.9 ± 8.0‡ 15.1 ± 4.0 8.5 ± 2.9 9.5 ± 3.2 2.9 ± 0.7 1.2 ± 0.2 2.1 ± 1.0 28.4 ± 1.5 5.5 ± 4.5 8.4 ± 1.2 12.0 ± 3.6 12.0 ± 2.0 4.3 ± 2.0 0.9 ± 0.4 1.3 ± 0.4 55.7 ± 5.7 38.5 ± 7.9 13.6 ± 5.9 8.9 ± 4.8 9.1 ± 2.6 3.1 ± 1.2 1.4 ± 0.5 1.4 ± 0.5 23.9 ± 3.8 25.2 ± 9.4 13.2 ± 3.6 9.0 ± 2.3 11.0 ± 3.0 2.4 ± 0.9 0.9 ± 0.4 1.2 ± 0.4 37.2 ± 3.4 GST (n = 3) 42.3 ± 16.5 12.6 ± 3.0 9.3 ± 4.8 5.9 ± 1.1 0.5 ± 0.4 0.4 ± 0.2 1.5 ± 0.1 27.5 ± 7.1 † The mean gonad indexes of fish in each group (from left to right in the Table) were 2.95 ± 1.31%, 1.71 ± 0.70%, 3.54 ± 0.78%, 2.28 ± 1.34% and 1.98 ± 0.87% (mean ± SD %), respectively ‡ Mean ± SD (%) 179 Chapter Receptor-mediated gene transfer 6xHis-LVIa 70 6xHis-LVIb 6XHis-LVIc 60 6xHis-LVII % of total radioactivity of whole fish % of recovered radioactivity GST 50 40 30 20 10 ovary liver gut gill heart spleen kidney others Fig 4-10 Relative radioactivities in seven tissues and the other parts of fish examined at 48 hr after injection with 32P-labeled recombinant 6xHis-Vtg fragments (LVIa, LVIb, LVIc and LVII) and 32P-labeled GST, respectively (250,000 cpm/fish) Relative radioactivity is presented as mean ± SD % For information about the number of fish examined in each group and their mean gonad indexes, see legend of Table 4-5 180 Chapter Receptor-mediated gene transfer labeling reaction (Fig 4-9D) Low values of relative radioactivity (8.4-15.1%) were present in the liver, gut and gill, and very low values (0.4-4.3%) were observed in the heart, spleen and kidney after injection with recombinant Vtg proteins Thus, it appeared that there was no preference in incorporation of the recombinant Vtg fragment (LVIa, LVIc or LVII) by the ovary For recombinant Vtg fragment LVIb, no conclusion could be drawn from this experiment Furthermore, it seemed that none of these recombinant Vtg fragments were taken up by oocytes specifically, since the relative radioactivities in ovary were similar between GST and recombinant Vtg injected groups It has been reported that two regions in lipovitellin I of chicken VtgII, LKRILK (residues 493-498) and KLKRIL (residues 1079-1084), are involved in interaction with its cellular receptors (Stifani et al., 1990; Steyrer et al., 1990) After sequence alignment between chicken VtgII and tilapia (O aureus) Vtg1, two homologous regions were found in the tilapia Vtg1, i.e IKTIMK (residues 493-498, site B) and KIKKIL (residues 1075-1080, site C) (Fig 4-7) Obviously, these two sites are quite conserved among the chicken and several other teleost fish Vtgs, especially for the positively charged lysine and arginine residues (multiple sequence alignment not shown) Furthermore, proline residues were found within the flanking regions of sites B and C (Fig 4-7), and the proline residues are usually present near protein interaction sites (Kini and Evans, 1995; Kini and Evans, 1996) Thus, it is possible that sites B and C in tilapia Vtg1 are two potential receptorbinding sites A recent study revealed that there is another receptor-binding region located in the lipovitellin I domain from residue 158 to 242, and residues 178-185 (site A) were further suggested to be critical for receptor interaction (Li et al., 2003; Fig 4-7) In this experiment, the three 6xHis-tagged Vtg fragments, LVIa, LVIb and LVIc, harboring the 181 Chapter Receptor-mediated gene transfer receptor-binding site A and two potential receptor-binding sites B and C, respectively (Fig 4-7) However, based on this preliminary experiment, none of these fragments seemed to be internalized by the ovary specifically, although the incorporation of 6xHisLVII by the ovary was relatively low (Table 4-5) The reasons for possible loss of receptor-binding function of these Vtg fragments may be the following: 1) Lack of proper posttranslational modification of the recombinant Vtg fragments expressed in E coli This is the major drawback of expression of recombinant proteins in E coli system It is known that vitellogenins undergo substantial modifications such as glycosylation and phosphorylation after their synthesis in the rough endoplasmic reticulum in most species examined (Byrne et al, 1989) Thus, without proper posttranslational modification, these recombinant Vtgs produced in E coli may not be able to interact effectively with their ovaries receptors in vivo Alternatively, realization of receptor-binding may require the interactions of distal amino acid residues with the residues within the receptor-binding domain, which may not be available in truncated recombinant Vtg fragments 2) Insufficient zinc or calcium ions in the recombinant Vtg fragments expressed in E coli It has been demonstrated that zinc and cadmium are present in lipovitellin I of Xenopus laevis Vtg (Sunderman Jr et al., 1995), while calcium is exclusively detected in phosvitin (Montorzi et al., 1995) Montorzi et al (1995) suggested that zinc or calcium or both may be required for receptor binding of Xenopus laevis Vtg The recombinant Vtg fragments expressed in prokaryotic cells may not carry these ions after synthesis Thus, it is possible that the receptor-binding capabilities may be lost for these fragments 182 Chapter Receptor-mediated gene transfer 3) Possible lack of lipids in the recombinant Vtg fragments expressed in E coli It was noted earlier that the association with lipid was necessary for apoE to present its receptorbinding activity (Innerarity et al., 1979) Ligand blotting experiments also showed that the lipid reconstituted chicken lipovitellin can strongly bind to its oocyte receptor but such interaction was abolished without lipid (Steyrer et al., 1990) Since fish vitellogenins share a common structure with other vertebrate vitellogenins, the requirement of bound lipids for effective interaction with Vtg receptors may also apply to fish Vtgs However, recombinant Vtg fragments expressed in prokaryotic cells may not contain sufficient lipids which may adversely affect their receptor-binding activities Considering the complexity of the in vivo system, future attempts to elucidate the presence of potential multiple receptor-binding sites in fish Vtg could be made using in vitro approaches such as the ligand blotting approach in which solubilized egg membrane proteins or recombinant Vtg receptor proteins are immobilized and the interaction between Vtg receptors and various Vtg fragments assayed in vitro In addition, recombinant Vtg proteins expressed in E coli may be coupled with iron or incubated with lipid prior to interaction with these receptors Alternatively, an eucaryotic expression system may be used To overcome the degradation problem associated with the largest Vtg fragment LVIb, three small recombinant proteins (LVIb1, LVIb2 and LVIb3) encompassing the entire LVIb were expressed in E coli and purified as 6xHis-tagged Vtg fragments (Fig 49B,C) which will facilitate the localization of possible receptor-binding sites in this region 183 ... Receptor- mediated gene transfer 24 hr, DNA~Vtg-pLys 144 24 hr, DNA~Vtg-pLys 144 24 hr, DNA 24 hr, DNA 24 hr, DNA 48 hr, DNA~Vtg-pLys 144 48 hr, DNA~Vtg-pLys 144 48 hr, DNA 48 hr, DNA 60 55 % of total... radioactivity % of recovered of whole fish 50 45 40 35 30 25 20 15 10 24 48 ovary 24 48 liver 24 48 gut 24 gill 48 24 48 heart 24 48 spleen 24 48 kidney 24 48 others remains Fig 4- 6 Relative radioactivities... 211 nm was measured for each elute For modification of pLys 144 by SPDP at a molar ratio of pLys 144 to SPDP of 1:2, 30 µl of SPDP (20 nmol/µl) was mixed with 150 µl of pLys 144 (2 nmol/µl) and the