Government & Nonprofit

Hydrodynamics-Based Transfer of Human Apolipoprotein A-I Gene to Mice: Study of Factors Affecting the Efficiency and Duration of Gene Expression in the Mouse Liver

Hydrodynamics-Based Transfer of Human Apolipoprotein A-I Gene to Mice: Study of Factors Affecting the Efficiency and Duration of Gene Expression in the Mouse Liver
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
   0026-8933/04/3806- © 2004 MAIK “Nauka/Interperiodica” 0921      Molecular Biology, Vol. 38, No. 6, 2004, pp. 921–928. Translated from Molekulyarnaya Biologiya, Vol. 38, No. 6, 2004, pp. 1076–1084.Original Russian Text Copyright © 2004 by Akifiev, Dizhe, Efremov, Mogilenko, Oleinikova, Lapikov, Zhdanova, Kidgotko, Orlov, Perevozchikov.  Currently, methods of gene therapy for various dis-eases are becoming more and more widespread [1].The key moment in this approach is delivery of thegene, whose expression product has a therapeuticeffect, to a patient. The method for delivery of thegene expression vector and the selection of regulatorysites that determine the level and duration of expres-sion of this gene in the body are also important.The delivery procedure for genes contained inrecombinant viruses (adenoviruses, adenoassociatedviruses, retroviruses, etc.) and driven by virus-specificregulatory regions, such as the promoter of humancytomegalovirus (CMV) early genes, is widely usedtoday because of its high efficiency. However, it hassome disadvantages that restrict its use. In particular,this concerns side effects, which, depending on thevirus used, are manifested in toxicity, increasedimmunogenicity, or the ability to activate protoonco-genes and induce the formation of pathogenic variantsof viruses due to recombination with endogenousviruses [2, 3]. In addition, it is believed that, inresponse to the functioning of virus-specific sitesresponsible for regulation of gene activity, the cell candevelop mechanisms for inactivating these sites; thisrestricts the duration of functioning and the level of activity of virus-specific expression vectors [4].Hence, nonviral methods of delivery, based on theuse of cationic DNA carriers, have been developed.They are less effective than viral vectors but havesome advantages, the main one being the possibility  APPLIED MOLECULAR BIOLOGY  Hydrodynamics-Based Transfer of Human Apolipoprotein A-I Gene to Mice: Study of Factors Affecting the Efficiency and Duration of Gene Expression in the Mouse Liver  B. N. Akifiev   1  , E. B. Dizhe   1  , A. M. Efremov   2  , D. A. Mogilenko   1,2  , G. N. Oleinikova   1  , I. A. Lapikov   1,2  , O. Yu. Zhdanova   1  , O. V. Kidgotko   1  , S. V. Orlov   1,2  , and A. P. Perevozchikov   1,2   1   Institute of Experimental Medicine, Russian Academy of Medical Sciences, St. Petersburg, 197376 Russiae-mail:   2  St. Petersburg State University, St. Petersburg, 199034 Russia  Received March 22, 2004  Abstract  —Human apolipoprotein A-I gene (  apoA-I   ) inserted into a plasmid expression vector was transferredin vivo into C57Bl/6 mice using hydrodynamic injections into the tail vein. Two types of plasmid expressionvectors were used: (1) pCMV  capoAI   which contained cDNA of apoA-I   driven by the human cytomegalovirus(CMV) early gene promoter and (2)  pAlg  , which contained a genomic locus of intron-containing apoA-I   drivenby its own extended 5'-regulatory region (APOAI). Hydrodynamic intravenous injections of both expressionvectors led to the appearance of human apoA-I   mRNA in the liver and human ApoA-I protein in the serum of injected mice. The dynamics of human ApoA-I content in the sera of mice injected with pCMV  capoAI   and  pAlg  were different. When pCMV  capoAI   was used, the concentration of human ApoA-I in mouse serum was maxi-mal one day after injection and decreased to zero within the next two weeks. In the case of  pAlg  , the content of human ApoA-I   in serum was maximal (up to 20 µ  g/ml) on days 5–7 after injection and then gradually decreasedfor several months (six months after injection, for example, it decreased to 25% of the maximal value). Exper-iments on “saved”  pAlg  plasmid isolated from the nuclei of hepatocytes 50 days after injection showed that theplasmid was retained for a long time in the form of an episome. A significant content of human ApoA-I in serumand its long-term persistence after injecting mice with  pAlg  may be accounted for by the properties of APOAIand/or the exon–intron structure of the apoA-I   gene. Injecting mice with different variants of APOAI coupledwith the luciferase gene did not lead to long-term expression of luciferase in the liver. It is concluded that thepresence of introns in the apoA-I   gene is required for its efficient and long-term expression after transfer to miceby means of hydrodynamic injections.  Key words  : C57Bl/6 mice, intravenous hydrodynamic injections, plasmid human apolipoprotein A-I geneexpression vectors, gene therapy, significance of the exon–intron gene structure for long-term expression in vivo  UDC 577.122.8   922  MOLECULAR BIOLOGY   Vol. 38   No. 6   2004  AKIFIEV et al  .  of repeated use. This is important for treating chronicdiseases, such as atherosclerosis, diabetes, rheuma-toid arthritis, etc., by gene therapy methods.The recently designed method of hydrodynamicinjections of naked plasmid DNA into laboratory ani-mals (mice) makes it possible to dispense with anycarriers. Furthermore, it ensures the efficient andlong-term expression of genes of interest in the recip-ient [5–8]. The essence of this method is in injectingDNA solution into the blood flow of mice at a highrate (1–3 ml within several seconds) by applyinghydrodynamic force. When injections are made intothe tail vein of mice, almost two-thirds of all DNAenter liver cells (primarily hepatocytes) [5]. This DNAis retained in the cell nuclei as episomes. Although theprocess of naked DNA delivery by this method ispoorly understood, is has been successfully applied inprimates [9]. In the near future this procedure willprobably be adapted for treating humans. In addition,this method of gene transfer is ideal for optimizationof the regulatory regions of genetic constructs,because additional agents (cationic carriers or viralvectors) that affect the expression of the transferredgene are absent.In our previous works, we studied the delivery of the human apolipoprotein A-I (  apoA-I   ) gene, whichexhibits antiatherogenic properties, into the liver of mice and rats in a complex with lysine- or arginine-rich cationic carriers [10, 11].The endogenous apolipoprotein A-I (ApoA-I) issynthesized mainly in the liver and small intestine of adult humans and then secreted into the blood, whereit exists in complexes with high-density lipoproteins(HDLP) and chylomicrons. The human apoA-I   gene,located on chromosome 11 in a cluster with the genes  apoÒ-III   and apo-IV   , is highly homologous to themouse apoA-I   gene [12]. It was shown earlier that theregulatory regions including the specific liverenhancer of apoA-I   (from  –220 to –77  bp relative tothe transcription start point) are sufficient for highlyefficient expression of this gene in hepatocytes [13].The purpose of this work was to study the conditionsof efficient and long-term expression of human apoA-I   delivered to the liver of laboratory mice by the hydrody-namics-based method depending on the nature of thetranscription-regulating regions linked to this gene andthe presence of introns in its coding region.EXPERIMENTALIn this work, we used human hepatocellular carci-noma (hepatoma) cell line HepG2 received from theAmerican collection of cell cultures (ATCC, NIH),which is maintained in the bank of cell cultures at theInstitute of Cytology, Russian Academy of Sciences(St. Petersburg).C57Bl/6 and hybrid C57Bl/CBA mice wereobtained from the Rappolovo nursery of the RussianAcademy of Medical Sciences.  Reagents and media used  in this study were fromSigma (United States), Gibco Life Technologies(United States), Serva (Germany), Pharmacia (Swe-den), and Boehringer Mannheim (Germany). Somereagents of analytical and chemical purity grade weremanufactured in Russia. The enzymes for geneticengineering procedures were from Fermentas (Lithua-nia) and NPO SibEnzim (Novosibirsk).  Genetic constructs and genetic engineering pro-cedures.  Human apoA-I   cDNA copy (  capoAI   ) and theEcoRI-derived fragment of an extended genomiclocus (11q23) with coordinates 116252251–116239958 (  ),which included the full-length human gene apoA-I   and its 5'-regulatory region, were kindly provided byDr. L. Chan (Baylor College, Texas, United States).The 5'-regulatory region of human apoA-I   gene(APOAI, from −  2497 to +401  ) was the kind gift of Dr.V.I. Shul’-zhenko (Institute of Molecular Biology andGenetics, Kiev). The nucleotide sequence of apoA-I   corresponds to the sequence deposited in GenBank (http://  ).The 5'-regulatory region of early CMV genes rein-forced by the SV40 enhancer (CMV promoter) andlinked to apoA-I   gene cDNA (the resulting plasmid  pCMV  capoAI   ), as well as the genetic constructs con-taining the reporter gene encoding chloramphenicolacetyltransferase (  cat   ) driven by the promoters of CMV (  pCMV  cat   ), mouse ribosomal protein L32(  pL32  cat   ), and apoA-I   (pAPOAI  cat   )  , were describedin our previous work [10].  Genetic engineering manipulations  were per-formed using standard procedures [14]. Plasmidexpression vectors containing the reporter gene of   Photinus piralis  luciferase (  luc  ) driven by the 5'-regu-latory region of the human apoA-I   gene (  pAPOAI  luc  )or the deletion variants of this promoter (  pAPOA-ISma-Sph  luc  and pAPOAISma-  Stu  luc  ) were obtainedas follows: the 5'-regulatory region (  from  −  2497 to+173  ) of the apoA-I   gene and its deletion variants cor-responding only to the core promoter with the hepaticenhancer SmaI   …  SphI   (from −  256 to +173)  and  Sma  …  StuI   (from −  256 to +71)  were linked to cDNAof the reporter gene luc  , which srcinated from thepGL3-basic plasmid (Promega).  Transfection of cells.  Cell cultures were main-tained using standard procedures. Cells were trans-fected with DNA using calcium phosphate coprecipi-tation. To standardize measurements of chloram-phenicol transferase and luciferase activities, cellswere cotransfected with the pCMV   Lac  Z  plasmid con-taining the gene for bacterial β  -galactosidase (   Lac  Z  )driven by the CMV promoter. The activity of β  -galac-   MOLECULAR BIOLOGY   Vol. 38   No. 6   2004  HYDRODYNAMICS-BASED TRANSFER OF HUMAN APOLIPOPROTEIN A-I GENE923  tosidase was determined by standard procedure using  Ó  -nitrophenyl-  β  -D-galactopyranoside as a substrate[15]. The activity of β  -galactosidase in cell lysateswas expressed in units of optical density (  OD   420  ) permilligram total protein per hour. In all experiments,chloramphenicol transferase and luciferase activitieswere normalized with respect to β  -galactosidaseactivity.  Cat  test and determination of luciferase activity.  The test for bacterial chloramphenicol transferaseactivity in lysates of transfected cells (  Cat   test) wasconducted as described in [15]. The luciferase reac-tion was performed using the kits from Promegaaccording to the protocol provided by the manufac-turer. The activity of luciferase in cell lysates wasdetermined by recording luminescence in a scintilla-tion counter (Beckman Instruments) and expressed inrelative light units (RLUs, counts per milligram totalprotein per minute).  Intravenous injections of plasmid DNA.  PlasmidDNA was isolated and purified by equilibrium ultra-centrifugation in a CsCl gradient and additionallypurified by chromatography on Minicolumn-D col-umns (Sigma). Then, plasmid DNA was injected intothe tail vein of mice weighing 12–19 g (10–20 µ  g permouse) either slowly (in 200 µ  l of Ringer’s solution orPBS) or hydrodynamically (in 1–3 ml of Ringer’ssolution within 4–5 s, depending on mouse weight).Hydrodynamic injections were administered to micewhose weight varied no more than by ±  1 g relative tothe mean value (  m   mean ). The exact volume ( V  ) of injected solution was calculated using the equationbased on the data of Liu et al.  [5]: V   (ml) = 0.0057143· m mean  + 0.5143. PCR analysis.  The nuclei of hepatocytes were iso-lated by centrifugation of cell homogenate 2000 ·g,20 min, 4°C ) in a stepwise sucrose gradient (0.5–2.2 M).The pellet was resuspended in hypotonic buffer con-taining 20 mM KCl and 10 mM Tris–HCl (pH 7.4).Low-molecular-weight nuclear DNA (lmwnDNA)was isolated by the method of Hirt as described [16]and used for PCR analysis of sequences specific forthe human apoA-I   gene. Primers for PCR and retro-PCR were selected using the Primer program devel-oped by V. Prutkovskii and O. Sokur (Research Insti-tute for Influenza, RAMS, St. Petersburg). As a result,we selected the direct external primer 5'-CCTGGGATC-GAGTGAAGGAC-3', the reverse external primer5'-CGTGCTCAGATGCTCGGTGG-3' (annealingtemperature 62°C ), the direct nested primer 5'-CGC-CTTGGGAAAACAGCTAA-3', and the reversenested primer 5'-GCTCCATCTCCTCCTGCCAC-3'(annealing temperature 57°C ). Retro-PCR analysis.  Total RNA from mousehepatocytes was isolated by the guanidine isothiocy-anate procedure and treated with RNase-free DNase I(Roche Applied Science).The reverse transcriptionreaction was run using reverse transcriptase (Invitro-gen) as recommended by the manufacturer; apoA-I  cDNA was identified by nested PCR analysis usingthe primers listed above. Determination of human ApoA-I and anti-ApoA-I antibodies in mouse blood serum.  HumanApoA-I and antibodies specific for this protein weredetermined in mouse blood serum by enzyme-linkedimmunosorbent assay (ELISA) using polyclonal goatantibodies to human ApoA-I and peroxidase-conju-gated rabbit IgG against goat immunoglobulins.RESULTS AND DISCUSSION Activity of the 5'-Regulatory Region (Promoter) of Human  apoA-I   Gene in HepG2 Cells Viral 5'-regulatory regions are frequently used toachieve the maximal levels of expression of the genesof interest in eukaryotic systems. One of the strongest5'-regulatory regions is the human CMV early genepromoter reinforced by a SV40 enhancer [17, 18].Figure 1a shows the results of comparing the strengthof the CMV promoter and mammalian promoters—the extended promoter of the human  ApoA-I   gene(APOAI, from  –2497 to +173 ) and mouse ribosomalprotein L32 gene (L32)—in the Cat   test on humanhepatoma HepG2 cells. Our results confirmed the datareported by other authors, that the virus-specific CMVpromoter is significantly stronger than the hepatospe-cific gene promoters [17, 18]. Furthermore, the CMVpromoter proved to be stronger than many other suffi-ciently strong mammalian promoters, including thenonspecific promoter of the gene encoding mouseribosomal protein L32 [19]).In this study, we also compared the hepatospecificactivity of the extended APOAI promoter (from − 2497to +173) and its two deletion variants containing themost proximal region of the hepatic enhancer [13, 20]and differing from one another in the length of thenoncoding region located downstream of the tran-scription srcin (from  –2497 to +173  and from − 256 to+71 ). The results of determining the activity of APOAI variants in the luciferase test on HepG2 cellsare shown in Fig. 1b. The activities of pAPOAISma- Sph luc  and pAPOAISma- Stu luc  (deletion variants of the APOAI promoter) were considerably higher thanthe activity of pAPOAI luc  (the variant with theextended APOAI promoter), which was consistentwith the idea concerning the presence of repressorregions in the extended distal region of APOAI [21]. Weused these deletion variants of APOAI to test the activityof the luc  gene, which was linked to them, in mouse hepa-tocytes.  924 MOLECULAR BIOLOGY   Vol. 38   No. 6   2004 AKIFIEV et al . Expression of Human  apoA-I   Gene in the Liver after Delivering It into Mouse Cells by Means of Intravenous Hydrodynamic Injection Plasmids pCMV capoAI   (with the apoA-I   cDNAdriven by the CMV promoter) and p  Alg  (with theintron-containing extended human apoA-I   gene drivenby its own 5'-regulatory region APOAI) were injectedinto the tail vein of male mice of different ages bymeans of hydrodynamic injections (see Experimen-tal). It is known that, as a result of hydrodynamicinjections of naked plasmid DNA into the tail vein of mice, two-thirds of this DNA enters the liver (goingmainly into hepatocytes), with the remaining one-third being distributed in the heart, lungs, and someother tissues [5]. We obtained similar results whenstudying the distribution of intravenously injectedDNA ( pCMV ÒapoAI   and p  Alg ) in mouse tissues byPCR (data not shown).After conventional intravenous injections of theabove plasmid DNAs, the protein product of thehuman apoA-I   gene appeared neither in the mousehepatocytes nor in the blood. By contrast, when thehydrodynamics-based procedure was used, humanApoA-I synthesized de novo  in the liver and thensecreted into circulating blood was detected in thesera of mice injected with these plasmids. Figure 2shows the results of ELISA for human ApoA-I inmouse blood serum after hydrodynamic injections of  pCMV ÒapoAI   and p  Alg  into male C57Bl/6 mice. Becauseof individual differences in the efficiency of gene transfer,we performed long-term monitoring of changes inApoA-I content in the serum of the same mice, periodi-cally taking blood samples from the tail vein.When mice were injected with pCMV capoAI  ,human ApoA-I was detected in blood serum as earlyas one day after injection, whereas injections with theexpression vector of the chromosomal gene ( p  Alg )resulted in the appearance of human ApoA-I only onthe third day after injection (Fig. 2). Moreover, thetype of injected plasmid had a considerable effect onthe dynamics of the human protein in mouse blood.When pCMV ÒapoAI   was injected, the content of human ApoA-I reached a peak as early as on the firstday after injection and gradually decreased to zero bythe end of the second week. When mice were injectedwith p  Alg , the content of human ApoA-I increasedgradually, reaching a peak on days 5–7, and thenbegan to decrease slowly. The highest serum level of ApoA-I detected after intravenous injections of p  Alg was comparable with (and even exceeded) thatobserved after injection of pCMV capoAI  . Repeatedinjections of pCMV ÒapoAI   at two- to three-day inter-vals maintained the level of human ApoA-I in mouseblood for some time but did not increase it. Con-versely, repeated injections of p  Alg  at small doses(10–20 µ g per mouse) caused an increase in ApoA-Ilevel in mouse blood (data not shown). The serumlevel of human ApoA-I markedly varied depending onmouse body weight, age, and sex: it was higher inyoung mice and, after p  Alg  injections, could reach 15–20 µ g/ml in some cases. The higher viability of hybridC57Bl/CBA mice allowed the expression of the for-eign gene to be monitored for a longer time. When 1pAPOAI luc 0pAPOAlSma-pAPOAlSma-Sph luc Stu luc 2345675pCMV cat  0pL32 cat  pAPOAl cat  1015203025    C   h   l  o  r  a  m  p   h  e  n   i  c  o   l  a  c  e   t  y   l  a   t   i  o  n ,   %    R   L   U ,       ×     1   0   –   6 (‡)(b) Fig. 1.  Comparison of the activities of virus-specific andcell promoters in HepG2 cells. (a) Cat test with genetic con-structs containing the reporter gene cat   driven by CMV pro-moters ( pCMV cat  ), the gene encoding mouse ribosomalprotein L32 ( pL32 cat  ), and human apoA-I   gene( pAPOA1 cat  ). Ordinate shows the activity of chloram-phenicol acetyltransferase expressed as the percentage of radioactive chloramphenicol acetylation. (b) The luciferasetest with genetic constructs containing the reporter geneencoding luciferase driven by the full-length promoter of the apoA-I   gene (from  –2497 to +173 ) ( pAPOAI luc ) and itsdeletion variants pAPOAISma- Sph luc  (from –256 to +173) and pAPOAISma- Stu luc  (from –256 to +71 ). Ordinateshows luciferase activity in lysates of transfected cellsexpressed in relative light units (RLUs) (see Experimental).  MOLECULAR BIOLOGY   Vol. 38   No. 6   2004 HYDRODYNAMICS-BASED TRANSFER OF HUMAN APOLIPOPROTEIN A-I GENE925 C57Bl/CBA mice received single injections of p  Alg ,human ApoA-I was detected in their blood for sixmonths after injection and even later, and its contentby that time was at least 25% of the peak valueobserved in the first week after injection.The delayed appearance of human ApoA-I inmouse blood upon intravenous injections of p  Alg could result from delayed transcription of the corre-sponding gene (the chromosomal gene apoÄ-I  ),which has an exon–intron structure.To test thishypothesis, we isolated RNA from the liver of p  Alg -injected mice one and three days after injections andtested it for the presence of apoA-I  -specific sequencesusing nested RT-PCR (Fig. 3). On the first day after p  Alg  injection, human apoA-I   transcripts were scantand could be detected only by nested RT-PCR withtwo pairs of primers (Fig. 3, lanes  3  and  4 ). Thereafter,the amount of specific RNA increased; on the thirdday after injection, apoA-I   mRNA was detected byboth nested RT-PCR and usual RT-PCR with externalprimers (Fig. 3, lane  5 ). However, when mice wereinjected with pCMV capoAI  , apoA-I   mRNA wasdetected by PCR with external (primary) primers asearly as on the first day after injection (Fig. 3, lane 1 ).Thus, the results of these experiments showed thattranscription of the chromosomal gene apoA-I   in theliver of injected mice began later than transcription of  apoA-I   cDNA driven by the CMV promoter.The delay in the appearance in mouse blood of human ApoA-I synthesized in the liver after transferof the human apoA-I   gene, which has an exon–intronstructure, is an srcinal fact, and further investigationis required for its interpretation.Comparison of our results and the data obtained byDe Geest et al.  [4, 22, 23], who transferred similargenetic constructs, inserted into recombinant adenovi-ruses, into the liver of C57Bl/6 mice, allowed us toconclude that the nature of differences in the durationof expression of the chromosomal gene (which has anexon–intron structure and intrinsic regulatory ele-ments) and its cDNA copy driven by the CMV pro-moter was similar. When mice were injected with apoA-I   driven by the CMV promoter, expression wasshort-term and decreased to minimal values by the endof the second week, whereas the transfer of the chro-mosomal gene resulted in longer expression. Theabove authors attributed the decrease in the expressionof apoA-I   cDNA driven by the CMV promoter to thesuppression of this promoter by as yet unknownmechanisms of antiviral cell activity [4].Unlike these researchers, we observed an increase notonly in the duration, but also in the level of expression of the chromosomal gene apoA-I   driven by its own promoter,which was comparable with the initial level of apoA-I  cDNA expression driven by the CMV promoter (Fig. 2).When analyzing this difference, it should be notedthat, in higher eukaryotes, the presence of introns inthe coding region (and, consequently, their splicing)may stimulate the expression of a number of suchgenes in different ways (specifically, at the levels of transcription, processing, mRNA transport into thecytoplasm, accumulation of mRNA in the cytoplasm,and translation of this mRNA) [24]. Assessment of the Efficiency of Luciferase Gene Expression Driven by the Hepatic Enhancer  apoA-I in Mouse Hepatocytes In accordance with the hypothesis of De Geest et al.  [4, 22, 23] concerning the suppression of viral pro-moter in hepatocytes, it was of interest to study therole of the 5'-regulatory region of human apoA-I  (APOAI) in maintaining the expression of the markergene luc  linked to this region in nondividing hepato- 522 bp214 bp 123456  Fig. 3.  Analysis of human apoA-I   gene expression in themouse liver by ( 1 ) RT-PCR with external primers and RNAisolated from the liver of mice injected with pCMV capoAI  24 h after injection, ( 2 ) nested RT-PCR with external andnested primers and RNA isolated from the liver of miceinjected with pCMV capoAI   24 h after injection, ( 3 ) RT-PCR with external primers and RNA isolated from the liverof mice injected with p  Alg  24 h after injection, ( 4 ) nestedRT-PCR with external and nested primers and RNA isolatedfrom the liver of mice injected with p  Alg  24 h after injec-tion, (  5 ) RT-PCR with external primers and RNA isolatedfrom the liver of mice injected with p  Alg  72 h after injec-tion, and ( 6  ) nested RT-PCR with external and nested prim-ers and RNA isolated from the liver of mice injected with p  Alg  72 h after injection. 102357914211234 Time after injection, daysHuman ApoA-I in bloodserum, µ g/ml Fig. 2.  Dynamics of human apoA-I   content in the bloodserum of male C57Bl/6 mice after hydrodynamic tail-veininjections of human apoA-I   plasmid expression vectors.Light and dark columns show data on mice injected with pCMV capoAI   and p  Alg , respectively.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks