A Picornaviral2A-Like Sequence-Bad Tricistronic Vector Allowing for High-Level Therapeutic Gene Expression Coupled to a Dual-Reporter System
Mark J.Osborn,1Angela Panoskaltsis-Mortari,1Ron T.McElmurry,1Scott K.Bell,1and Dario A.A.Vignali2Martin D.Ryan3Andrew C.Wilber4R.Scott McIvor4
Jakub Tolar1Bruce R.Blazar1,*
惠普电脑u盘启动
1Pediatrics,Hematology–Oncology,Blood and Marrow Transplant Program,and4Gene Therapy Program, Institute of Human Genetics,Department of Genetics,Cell Biology,and Development,University of Minnesota,Minneapolis,MN55455,USA
2Department of Immunology,St.Jude Children’s Rearch Hospital,Memphis,TN38105,USA
3School of Biology,Centre for Biomolecular Sciences,North Haugh,University of St.Andrews,St.Andrews,KY169ST Scotland,UK *To whom correspondence and reprint requests should be addresd at the Department of Pediatrics,University of Minnesota,MMC109 460F Cancer Center Rearch Building,425East River Road,Minneapolis,MN55455,USA.Fax:+16126243913.E-mail:blaza001@umn.edu.
Available online16June2005
The2A-like quences from members of the picornavirus family were utilized to construct a
tricistronic vector bearing the human iduronida(IDUA)gene along with the firefly lucifera and
DsRed2reporter genes.The2A-like quences mediate a cotranslational cleavage event resulting in
the relea of each individual protein product.Efficient cleavage was obrved and all three proteins演播室
不一样的童年were functional in vitro and in vivo,allowing for supratherapeutic IDUA enzyme levels and the
coexpression of lucifera and DsRed2expression,which enabled us to track gene expression.
Key Words:multicistronic,cis-acting hydrola elements,internal ribosome entry site,
iduronida,mucopolysaccharidosis type I,whole body imaging
Therapeutic gene expression can be difficult to detect and often times necessitates coexpression of a reporter gene to allow for indirect detection.While many reporter genes exist there is no one reporter that meets the multiple needs of quential whole-body imaging and unequivocal detection
of gene expression at the cellular level.For instance,the commonly ud green fluorescent protein(GFP)reporter can be uful for cell lection.However,its efficacy is limited in vivo becau the excitation and emission spectra have only a1-to5-mm penetration depth and tissue absorbance can result in high background fluorescence[1,2].In contrast,the Discosoma sp.DsRed2red fluorescent protein is better suited to tissue penetration than GFP[3].Photinus pyralis lucifera is capable of generating light by catalyzing the conversion of d-luciferin to oxyluciferin.The back-ground is very low and the bioluminescent signal can be imaged in real time in vivo.This obviates the need for sacrificing the animal to perform analys and,as such, is a powerful analytical tool.Lucifera is,however, difficult to visualize at the cellular level as it does not posss native fluorescent activity as do GFP and DsRed2 [1].
We sought to determine whether we could couple expression of the lucifera and DsRed2reporter genes to a therapeutic gene—all contained on a single transcript in one vector.Such a multicistronic vector would result in the coordinated expression of all three genes and allow us to(i)exploit the advantages of lucifera by using real-time in vivo whole-body imaging and(ii)detect and quantify gene expression at the cellular level—without antibody staining—by using DsRed2,while achieving expression of the therapeutic iduronida(IDUA)gene. Tracking therapeutic gene expression in this manner would be a particularly uful tool for gene therapy preclinical studies.
Current strategies for creating multicistronic vectors include the u of internal ribosome entry sites(IRES), multiple promoters,and fusion proteins—with or with-out linkages via cellular protea sites.IRES elements can be large,precluding their u in size-restricted vectors (e.g.,viral and transposon-bad vectors).Furthermore, downstream gene expression can be attenuated poten-tially due to the order of the genes on the transcript, which may necessitate extended vector optimization [4,5].Fusion protein production may result in compro-
mid function potentially due to improper protein
folding or trafficking.The u of multiple promoters
results in less intimate linkage between expression of the
reporter genes and that of the therapeutic gene.Impor-
tantly,such a strategy cannot provide a reliable indica-
tion of therapeutic gene expression levels.Using cellular
protea sites as a means to link cistrons together is
complicated by the fact that cleavage will occur only
within cells expressing the protea and the protein must
pass through the cellular compartment in which the
protea is expresd [6,7].
Recent strategies have involved the u of the foot-
and-mouth dia virus 2A or 2A-like cis -acting hydrola
elements (CHYSEL)to create multicistronic vectors capa-
ble of generating multiple proteins from the same tran-
script,including tho capable of coexpressing functional
enzymes [8–11].Many virus encode multiple proteins
that are cleaved into individual protein products at 2A or
2A-like quences [12–14].The 2A-like quence contains
a canonical Asp-Val/Ile-Glu-X-Asn-Pro-Gly (2A)A Pro (2B)
motif,which results in cleavage between the 2A glycine
and the 2B proline [15,16].The cleavage mechanism is
thought to occur as the result of a ribosomal b skipping Q
孝顺父母的成语mechanism whereby ribosome activity is modified by the
2A-like quences,preventing peptide bond formation
between the 2A Gly and the 2B Pro,causing relea of the
upstream protein while allowing continued translation of
the downstream gene [13,17,18].We sought to determine
whether we could utilize the 2A-like CHYSEL quences of
定位手机号
porcine teschovirus (P2A)and Thoa asigna virus (T2A)to create a tricistronic plasmid vector encoding a single,long,ORF consisting of the therapeutic human IDUA gene,the firefly lucifera gene,and the DsRed2gene to direct high-level therapeutic gene expression while utiliz-ing the lucifera and DsRed2genes as direct markers of vector expression (Fig.1).IDUA is a lysosomal enzyme involved in glycosamino-glycan (GAG)degradation and is deficient in mucopoly-saccharidosis type I (MPS I),causing intracellular GAG accumulation,resulting in organomegaly and neuro-logical and skeletal defects [19].Currently,bone marrow transplant (BMT),which is a standard therapy for MPS I,is not uniformly effective in preventing the quelae of MPS I and can lead to morbidity and mortality as complications of the procedure [20].Therapeutic vectors capable of replacing IDUA enzyme levels are therefore potentially advantageous.Our vector design strategy may be particularly uful in that it can express functional therapeutic enzyme while at the same time allowing for direct detection utilizing the lucifera and DsRed2reporter genes and may be uniquely uful for preclinical and gene therapy studies.We created a CMV promoter-regulated IDUA –P2A–lucifera–T2A–DsRed2construct termed pLB (Fig.1A)that was engineered for cotranslational cleavage,releas-ing each individual protein (Fig.1B).We generated mutated constructs in which the Pro-Gly (2A)A Pro (2B)(PGP)motif of P2A was changed to Ala-Val-Pro
小飞象英文(AVP),
FIG.1.Tricistronic vector constructs.(A)Schematic of
足球英语怎么读the cleavable pLB construct containing the IDUA
downstream of the CMV promoter in the pCDNA3.1D
TOPO vector (Invitrogen).Subquently the P2A–
lucifera–T2A–DsRed2fragment,generated by
recombinant PCR as described by Szymczak [8]and
Higuchi [26],was cloned in-frame,downstream of the
IDUA gene,resulting in the plasmid pLB.The IDUA
cDNA was a gift from E.Neufeld [27],lucifera was
derived from the pGL-3vector (Promega),and DsRed2
was subcloned from the pIRES-DsRed2vector (Clon-
tech).Arrow indicates the cleavage sight between the
2A glycine and the 2B proline.(B)Individual cleavage
products with associated 2A-like quences at the C
(IDUA and lucifera)and/or N (lucifera and DsRed2)
terminus of the protein products.(C)The P2A
quence was mutated from PGP to AVP or (D)
completely deleted,which should decrea the cleav-
age efficiency between IDUA and lucifera [13].The
T2A quence was maintained to monitor transfection
外派劳务efficiency by DsRed2expression.
termed p2A-M (Fig.1C),or removed altogether,termed
D PGP (Fig.1D).Mutation/deletion of the residues
should decrea cleavage efficiency [13].To determine
functionality of the multigene construct we transfected
NIH 3T3cells with each plasmid and assd gene
expression at 48h.We assd the activity of creted
IDUA in the cellular supernatant,as lysosomal enzymes
are capable of being creted and taken up by cells distant
from the site of production via the ubiquitous manno 6-phosphate receptor,using 4-methylumbelliferyl a -l -iduronide as the substrate as described [21,22].Cells transfected with pLB (cleavable form)exhibited substantial IDUA activity compared to untransfected 3T3control cells and total bone marrow (BM)(Fig.2A).The mutated,p2A-M and D PGP,plasmids conferred higher IDUA activity than that in the untransfected cells and BM,but this level was markedly reduced compared to the cleavable form of IDUA (Fig.2
A).
FIG.2.In vitro analysis of IDUA,lucifera,and DsRed2gene expression in NIH 3T3cells.Endotoxin-free plasmid (Qiagen)was ud to transfect 1063T3cells using the Effectene transfection reagent (Qiagen)and monitoring for gene expression 48h posttransfection.(A)In vitro IDUA activity.Cellular supernatant was removed and assayed in triplicate for IDUA activity using 4-methylumbelliferyl a -L -iduronide as the substrate as described [22].Controls include an equal number of untransfected 3T3s and 106BM cells.p2A-M reprents cells transfected with the mutated (PGP Y AVP)plasmid,and D PGP reprents cells transfected with the P2A PGP-deleted mutant plasmid.IDUA activity was determined using a standard curve of known amounts of 4-methylumbelliferone (Sigma).Values are prented as IDUA/nmol/A g of protein/4h and standard deviation is shown.(B)In vitro lucifera activity.Cells transfected with the cleavable (pLB),uncleavable p2A-M,or uncleavable D PGP plasmid were harvested and lyd in 1ÂReporter Lysis Buffer (Promega)and eded in triplicate in flat-bottom plates followed by addition of Lucifera Assay Buffer (Promega)and photon emission was measured for 5s using a Chameleon 425–100multilabel counter (Hidex).Cell lysate protein content was assd using the Proteostain Protein Quantification Kit (Dojindo Molecular Technologies)and lucifera activity is expresd as relative light units (RLU)/mg of protein F SD.(C)In vitro DsRed2expression.3T3cells 48h posttransfection (or untransfected cells as a control)were trypsinized and analyzed on a FACSCalibur (Becton–Dickinson)
for DsRed2activity.(D)Western blot analysis of cleavage activity.Antibody specificity for lucifera was determined using recombinant lucifera (lane 1).Protein concentration was determined from cell lysates using the Proteostain Protein Quantification Kit (Dojindo Molecular Technologies).Equal amounts of protein (50A g)from pLB-(lane 2),p2A-M-(lane 3),or D PGP-(lane 4)transfected or untransfected (not shown)cells were parated on a NuPAGE 10%Bis–Tris gel (Invitrogen),transferred to a PVDF membrane (Invitrogen),and probed with a biotinylated polyclonal anti–lucifera antibody (Abcam),followed by addition of streptavidin-conjugated HRP (Abcam),and then developed using the ECL Western Blotting Analysis System (Amersham Biosciences).Cleaved and uncleaved lucifera is indicated by labeled arrows.
We assd lucifera activity in vitro by adding d-luciferin(Promega)to cell lysates and measuring photon emission of the oxyluciferin metabolite.Cells transfected with pLB exhibited high levels of lucifera activity,while untransfected controls exhibited none(Fig.2B).Both mutant plasmids exhibited lucifera activity;however, the number of photons emitted was less than from the cleavable pLB plasmid(Fig.2B).
We confirmed DsRed2expression in cells transfected with the pLB tricistronic construct by fluorescence microscopy as well as by flow cytometry(data not shown and Fig.2C).We ud Western
blot analysis to determine the cleavage efficiency for each construct using an antibody to the lucifera protein,encoded by the gene located in the middle of the construct(Fig.1). We obrved efficient cleavage in pLB cell lysates(Fig. 2D,lane2)resulting in a lucifera species with a slightly higher molecular weight than recombinant lucifera(Fig.2D,lanes1and2).This is consistent with this system,as2A-like quences are incorporated into the gene products,resulting in an increa in molecular weight.Both of the mutated plasmids(p2A-M and D PGP)resulted in production of a high-molec-ular-weight product only,indicating that cleavage into each individual gene product did not occur(Fig.2D, lanes3(p2A-M)and4(D PGP)).This analysis confirmed (i)that efficient cleavage occurred(Fig.2D)and(ii)that disrupting the PGP site,either by mutation or by deletion,prevents cleavage(Fig.2D,lanes2,3,and4). The low level of IDUA and lucifera noted in cells transfected with either mutated construct may be the result of activity of the IDUA–lucifera fusion product (Figs.2A,2B,and2D).
After establishing functionality of each of the three genes in vitro we sought to validate the system in vivo as well as perform comparison studies between the tricis-tronic plasmid and the corresponding monocistronic form of each construct.We created plasmids bearing monocistronic IDUA,lucifera,or DsRed2,such that the vector backbone and regulatory ,promoter and polyadenylation signal)were identical to tho of the tricistronic vector.
Fluorescence microscopy revealed DsRed2-positive cells in vitro mediated by both the tricistronic(Fig.2C) and the monocistronic form(data not shown).To compare tricistronic versus monocistronic IDUA and lucifera we injected equimolar amounts of each plasmid parately into the tail vein of nonobe diabetic–vere combined immunodeficiency(NOD-SCID)mice as described[23].We obtained rum samples48h after injection and assayed for IDUA activity.Animals receiving the pLB plasmid exhibited high-level IDUA activity compared to endogenous IDUA prent in the rum;however,the highest levels noted were from the animals receiving the monocistronic IDUA(Fig.3A).
At48h we performed whole-body lucifera imaging using the Xenogen IVIS system.While the , monocistronic IDUA injected)animal showed no luci-fera expression(Fig.3B,animal2),the animals receiving tricistronic plasmid(Fig.3B,animal1)or monocistronic lucifera(Fig.3B,animal3)exhibited high levels of lucifera expression in the liver,the anticipated site of highest gene expression,after hydro-dynamic injection of plasmid[23].We quantified mono-versus tricistronic lucifera activity by lecting a region of interest(red circles in Fig.3B)from the sternum to the midabdomen of each mou using the Xenogen Living-Image software(version 2.50),which allows for the determination of the number of photons/s/cm2within that
region.Exposure times were varied to allow for a time-cour analysis and the data are reprented in graphical form(Fig.3C).In vivo imaging(Fig.3B)and graphical data(Fig.3C)reveal that lucifera gene expression,whether mediated by the tricistronic or the monocistronic vectors,is comparable.
At48h we anesthetized the animals and made a midline incision to expo the liver for DsRed2imaging as described[24].DsRed2activity can be en throughout the liver in the pLB-injected animals,while uninjected and control animals lacked detectable DsRed2expression (Fig.3D and data not shown).We discted a portion of the liver and analyzed it for DsRed2activity in individual cells by fluorescence microscopy,and DsRed2-positive hepatocytes are shown in Fig.3E.
The ability to generate multicistronic vectors holds great potential for the field of gene therapy as a whole. In this report we have demonstrated that a tricistronic vector carrying a therapeutic gene along with two reporter genes is functional both in vitro and in vivo. Our coexpression strategy resulted in a unique vector capable of high-level therapeutic gene expression and allowed us to track gene expression noninvasively using lucifera and at the cellular level using DsRed2.The multitude of options that exist for creating multicis-tronic vectors carry with them associated potential pitfalls,and2A-bad vectors are not exempt in this respect.The cotranslational cleavage mechanism results in the addition of a2A-like quence to the upstream gene and addition of the2B proline to th
e N-terminus of the translation product downstream of2A,which may potentially interfere with protein folding or activ-ity.In the ca of our vector this resulted in a C-terminal extension of IDUA by the20amino acids of the porcine teschovirus2A-like quence(Fig.1B). Incorporation of the quences may inhibit IDUA activity,as IDUA activity was higher in the monocis-tronic ,lacking the2A-like quences)com-pared to the tricistronic form(Fig.3A).However,we still achieved IDUA activity using the tricistronic vector, which surpasd the endogenous levels produced in vitro by3T3cells(Fig.2A)and in vivo by NOD-SCID mice
FIG.3.Tricistronic versus monocistronic vector gene expression.NOD-SCID mice were infud by hydrodynamic tail vein injection with equimolar amounts (50A g)of pLB,monocistronic IDUA(35A g),or monocistronic lucifera(37A g)as described[23]and analyzed at48h postinjection.(A)In vivo IDUA activity. Serum was harvested and IDUA activity was assd as before[22].Animals receiving monocistronic lucifera rved as controls.Duplicate experiments with three animals each were ud and SEM is shown.(B)Whole-body in vivo lucifera imaging.Animals were anesthetized and administered150A g/g D-luciferin (Xenogen)by intraperitoneal injection and imaged for lucifera activity using the Xenogen IVIS imaging system.The scale bar on the right reprents photon emission as photons/s/cm2/sr.Reprentative animals of tricistronic plasmid-injected(animal1),control(IDUA injected,animal2),and monocistronic lucifera-injected(animal3)recipients are shown.(C)Comparison of lucifera activity between mono-and tricistronic-injected animals.A region of interest(ROI;red circle in B)was placed around the area of lucifera expression and the number of photons/s/cm2was determined within the ROI using LivingImage software version2.50(Xenogen).Graphical analysis ud the mean of photons counted in the ROI of four animals for each group over exposure times of0.1–90s with SEM shown.(D)Whole-organ DsRed2expression.Animals were anesthetized and the liver was expod following midline incision and imaged for DsRed2 expression using a Retiga EXi Fast1394camera(Q Imaging)mounte
d to a Leica MZFLII stereomicroscope as described[24].No fluorescence was obrved in uninjected or control plasmid-receiving animals(data not shown due to the fact that the background is nil).(E)Cellular DsRed2expression.A region of liver, determined by whole-organ imaging to be expressing DsRed2,was discted and homogenized so that a cytospin could be performed.DAPI nuclear staining (Molecular Probes)was performed followed by microscopy analysis using a BX51fluorescence microscope(Olympus).Cells expressing DsRed2are on the top,
the DAPI image is in the middle,and a merged image showing DsRed2-positive hepatocytes with nuclei is shown at the bottom.
