Ž .Journal of Immunological Methods 220 1998 115–121

Green fluorescent protein expressed by a recombinant vacciniavirus permits early detection of infected cells by flow cytometry

Javier Domınguez, Marıa del Mar Lorenzo, Rafael Blasco )´ ´Departamento de Sanidad Animal, Centro de InÕestigacion en Sanidad Animal-I.N.I.A., Valdeolmos, Madrid E-28130, Spain´

Received 2 June 1998; revised 30 July 1998; accepted 10 August 1998

Abstract

Ž .We have tested Green Fluorescent Protein GFP expressed by a vaccinia virus recombinant as a marker for viralŽ .infection. Virus recombinants expressing either wild-type GFP, or a Ser to Thr mutated version GFP-S65T were used to65

infect cultured cells, and the appearance of fluorescence was followed during infection by flow cytometry. Although bothversions were detectable in infected cells, GFP-S65T gave up to 26-fold brighter fluorescence than wild-type GFP when

Ž .excited by an argon laser beam 488 nm . In addition, GFP-S65T fluorescence appeared earlier, and infected cells could bedetected above background as soon as 1 h after infection. We have used this construct to infect porcine peripheral bloodlymphocytes, and show its usefulness to study virus tropism when used in combination with cell-type specific markers.Thus, GFP provides a direct, fast and convenient way to monitor infection by flow cytometry. q 1998 Elsevier Science B.V.All rights reserved.

Keywords: Vaccinia virus; Green fluorescent protein; Peripheral blood leukocytes; Virus tropism

1. Introduction

Detection of virus-infected cells usually relies onthe detection of viral proteins, using specific reagentssuch as antibodies. As an alternative, the use ofconvenient readily detectable marker genes has been

Abbreviations: GFP, green fluorescent protein; FB, fluores-cence buffer; BSA, bovine serum albumin; PE, phycoerythrin; Ig,immunoglobulin; MAb, monoclonal antibodies; VV, vacciniavirus; PBMCs, peripheral blood mononuclear cells; EMEM, EagleMinimal Essential Medium; hpi, hours post-infection

) Corresponding author. Tel.: q34-1-620-23-00; Fax: q34-1-620-22-47; E-mail: blasco@inia.es

suggested. The marker gene is incorporated in thevirus genome and its expression used to monitorinfection. Obviously this approach is restricted toviruses, such as vaccinia virus, for which methodsexist that allow insertion, and direct expression, offoreign genes. A number of marker genes have beeninserted in the vaccinia virus genome, and theirexpression and usefulness have been demonstrated indifferent experimental situations. These include se-

Žlectable markers, such as thymidine kinase Mackettet al., 1982; Panicali and Paoletti, 1982; Katz and

.Middle, 1990 , guanine phosphoribosyl transferaseŽ . Žgpt Boyle and Coupar, 1988; Falkner and Moss,

. Ž1988 or neomycin resistance Franke et al., 1985;

0022-1759r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.Ž .PII: S0022-1759 98 00156-2

( )J. Domınguez et al.rJournal of Immunological Methods 220 1998 115–121´116

.Katz and Middle, 1990 and detectable markers,Ž .including b-galactosidase Chakrabarti et al., 1985 ,

Ž .b-glucuronidase Carroll and Moss, 1995 and lu-Ž .ciferase Rodriguez et al., 1988 .

Ž .Aequorea Õictoria green fluorescent protein GFPhas been successfully expressed in a variety of sys-

Žtems Chalfie et al., 1994; Amsterdam et al., 1995;Atkins and Izant, 1995; Cubitt et al., 1995; Chalfie,1995; Hu and Cheng, 1995; Kremer et al., 1995;

.Peters et al., 1995; Stearns, 1995; Yeh et al., 1995 .GFP offers several unique properties that make itideal for certain applications. Because it can bedetected without the need for fixation or processingof the sample, GFP is suited for in vivo applications.GFP fluorescence is stable and is detectable both bylight microscopy and flow cytometry in a wide rangeof laboratory settings.

Wild-type GFP has an excitation maximum of395 nm, and a minor excitation peak at 470 nm. Theexcitation and emission spectra of GFP can be modi-fied by mutation in the sequence of the protein, and amutation Ser to Thr in position 65 has been shown tomodify the excitation maximum to 489 nm, and tomodify the emission maximum from 508 nm to 511

Žnm Heim et al., 1994; Cubitt et al., 1995; Ehrig et.al., 1995; Heim et al., 1995 . Because of the change

in the excitation spectrum, these GFP variants areespecially suited for flow cytometry using an argon

Ž .laser line 488 nm .In this report we describe the use of GFP ex-

pressed by a vaccinia virus recombinant to serve asan infection tag. In addition, we show the utility ofthat recombinant to study virus tropism in a complexcell population such as porcine peripheral blood

Ž .mononuclear cells PBMCs .

2. Materials and methods

2.1. Viruses and cells

The construction and characterization of recombi-nant vaccinia viruses expressing GFP has been previ-

Ž .ously described Lorenzo and Blasco, 1998 . Vac-cinia viruses were routinely grown and plaqued on

ŽBS-C-1 monolayers American Type Culture Collec-. Žtion , following standard protocols Earl and Moss,.1991 .

2.2. Kinetics of expression in CV-1 cells

CV-1 cells in 12-well plates were infected withvaccinia viruses at a multiplicity of infection of 5PFU per cell. After 1 h adsorption, virus inoculumŽ .500 ml was removed, and 1 ml EMEM containing2% foetal calf serum was added. At different timespost-infection, cells were washed twice with PBS,detached from the plastic by incubation with PBScontaining 0.5 mM EDTA, and analyzed by flowcytometry.

2.3. Animals and cells

Large White pigs with an average weight of 30 kgwere used as blood donors. Peripheral blood

Ž .mononuclear cells PBMC were isolated on Percolldiscontinuous gradients after blood sedimentation in

Ždextran as previously described Gonzalez et al.,.1990 .

2.4. Monoclonal antibodies

ŽMonoclonal antibodies to porcine SWC3 74-22-. Ž .15 and CD3 8E6 were kindly provided by Drs. J.

Ž .Lunney ARS, USDA, Beltsville, USA and M.D.Ž .Pescovitz Indiana University, USA , respectively.

MAb 5C9 to pig IgM was obtained from ATCC.Ž . Ž .MAbs to porcine CD45RA 3C3r9 , LFA-1 BL2F1

Ž .and SLA-II DR 2E9r13 have been produced in ourŽlaboratory Bullido et al., 1997a,b; Alvarez et al.,

.unpublished results .

2.5. Flow cytofluorometry

For surface labelling with monoclonal antibodies,5=105 cells were incubated on ice with 50 ml ofhybridoma supernatants for 30 min. An irrelevant,isotype-matched, monoclonal antibody was used asnegative control. After two washes in PBS contain-

Žing 0.1% BSA and 0.1% sodium azide fluorescence.buffer, FB , cells were incubated for 30 min at 48C

Ž .with 50 ml of phycoerythrin PE -conjugated goatŽ .anti-mouse Ig Caltag, USA diluted 1r50 in FB.

Cells were washed three times in FB and fixed in0.3% paraformaldehyde prior to be analyzed in a

( )J. Domınguez et al.rJournal of Immunological Methods 220 1998 115–121´ 117

Fig. 1. Flow cytometric detection of GFP and GFP-S65T in CV-1 cells. CV-1 cells were infected with vaccinia viruses and analyzed by flowŽ . Ž .cytometry at different times post-infection. Fluorescence was recorded in the FL1 channel 515–545 nm range . A Histograms display the

Ž . Ž .cell number Y-axis vs. the relative fluorescence intensity X-axis . The time post-infection, in hours, is indicated above the panels, and theŽ . Ž .viruses used are indicated on the left. B The mean fluorescence intensity at each time point is represented for uninfected cells l or cells

Ž . Ž . Ž .infected with vaccinia WR B , VV-GFP ' or VV-GFP-S65T v .

( )J. Domınguez et al.rJournal of Immunological Methods 220 1998 115–121´118

Ž .FacsScan Becton-Dickinson, San Jose, CA, USA .The excitation wavelength was 488 nm, and fluores-cence was recorded in the 515–545 nm range forGFP, and in the 565–608 nm for phycoerythrin.Routinely, 5000 cells were analyzed for each sample.

3. Results and discussion

3.1. Detection of infected cells by flow cytometry

We have recently isolated vaccinia virus recombi-nants expressing GFP or a mutant version containing

Ž .a Ser to Thr mutation in position 65 GFP-S65TŽ .Lorenzo and Blasco, 1998 . We tested the course ofappearance of fluorescence in cells infected withboth recombinant viruses. CV-1 cells were syn-chronously infected, and analyzed by flow cytometry

Ž .at different times post-infection Fig. 1 . Cells in-Ž .fected with wild-type WR vaccinia virus did not

display significant fluorescence, whereas cells in-fected with VV-GFP or VV-GFP-S65T producedfluorescence above background. Wild-type GFP flu-orescence could be detected at about 4 h post-infec-tion, and increased thereafter. As expected, fluores-cence of GFP-S65T was stronger than that of the

ŽFig. 2. Infection of porcine PBMCs with VV-GFP-S65T. Infection of porcine PBMCs was followed by flow cytometry at the times in.hours indicated in the right upper corner in each panel. Below, the mean fluorescence intensity of positive cells is plotted vs. the time

post-infection.

( )J. Domınguez et al.rJournal of Immunological Methods 220 1998 115–121´ 119

wild-type protein. In addition, GFP-S65T fluores-cence appeared earlier in infection, in good agree-ment with previous reports showing that formationof the chromophore is accelerated by the S65T muta-

Ž .tion Heim et al., 1995 . The fluorescent signal inVV-GFP-S65T infected cells was detectable as soon

as 1 h after infection, and continued to increaseduring infection. Because of the infection, after

Ž .longer incubation periods )8 h the cells wereprone to rupture, and in our hands the results ob-

Ž .tained were difficult to interpret not shown . There-fore, we chose GFP-S65T and 5–6 h post-infection

Fig. 3. Analysis of VV-GFP-S65T-infected porcine PBMCs with cell-specific monoclonal antibodies. Infected PBMCs were incubated for 5h, and then fixed and incubated with the monoclonal antibodies indicated in each panel. GFP fluorescence was recorded in the FL1 channelŽ .515–545 nm , and the antibody markers, which were revealed with a PE-labelled second antibody, were recorded in the FL2 channelŽ .565–608 nm . Numbers inside the panels indicate the percentage of cells within the respective regions.

( )J. Domınguez et al.rJournal of Immunological Methods 220 1998 115–121´120

as optimal to detect vaccinia virus-infected cells byflow cytometry.

3.2. Infection of porcine PBMCs

The emergence of fluorescence was followed afterfreshly isolated porcine PBMCs were infected withVV-GFP-S65T. GFP fluorescence was detectable assoon as 1 h post-infection in a proportion of cellsŽ .Fig. 2 . Although fluorescence increased over aperiod of 2–3 h, a significant percentage of cellsremained fluorescence-negative, suggesting the pres-ence of a population of cells refractory to infection.In contrast to the results obtained with CV-1 cells,fluorescence did not increase significantly beyond 3h of infection.

In order to determine the susceptibility of differ-ent PBMC subsets to infection, freshly isolatedPBMC were infected, and at 5 h post-infection cellswere analyzed by flow cytometry with a panel ofmonoclonal antibodies specific for particular subpop-

Ž .ulations Fig. 3 . Most GFP-positive cells wereq Ž .CD45RA a marker for B cells and naive T cells

Ž .and MHC class II-positive SLA-II-DR . With theamount of virus used, about half of the B cellsŽ q. Ž q.sIgM and monocytes SWC3 expressed GFP.

ŽIn contrast, a significantly lower percentage 13–. Ž q.18% of T cells CD3 expressed GFP. In addition,

GFP fluorescence intensity was about two-fold higherin infected B cells or monocytes than in infected Tcells. Therefore, the data reveal a preference byvaccinia virus for porcine B cells and monocytesover T cells. This preference is apparent both fromthe number of infected cells and from the amount ofGFP expressed in different cell populations.

Our results are in good agreement with previousŽ .reports Alonso et al., 1991 in which vaccinia virus

tropism was investigated using several human leuko-cyte cell lines. Remarkably, infection of B cell linesresulted in more viral gene expression and virusproduction than infection of T cell lines. Thoseresults are consistent with the infection pattern de-scribed here. Virus production in the different PBMCsubpopulations remains to be determined, and wouldrequire preparative isolation of cell subsets, or theuse of cell lines. However, the fact that GFP fluores-cence in infected PBMCs does not increase beyond3–4 h could be indicative of an abortive infection.

Vaccinia virus tropism is particularly importantconsidering its potential use as a vector both forimmunization and for in vitro immunological studies.In addition, because of the availability of techniquesto incorporate and express foreign genes in vacciniavirus, this virus offered a convenient model to testthe potential of GFP in tropism studies. This reportillustrates the usefulness of GFP as an infectionmarker, especially when flow cytometry is the tech-nique of choice.

Acknowledgements

We thank Angel Ezquerra for the critical readingof the manuscript. This work was supported bygrants PB95-0237 and BIO97-0402-CO2-01 fromDireccion General de Investigacion Cientıfica y´ ´ ´

Ž .Tecnica DGICYT and Contract CT98-0225 from´the European Commission.

References

Alonso, J.M., Rodriguez, J., Vinuela, E., Kroemer, G., Martinez,C., 1991. Highly efficient expression of proteins encoded byrecombinant vaccinia virus in lymphocytes. Scand. J. Im-munol. 34, 619–626.

Amsterdam, A., Lin, S., Hopkins, N., 1995. The Aequorea Õicto-ria green fluorescent protein can be used as a reporter in livezebrafish embryos. Dev. Biol. 171, 123–129.

Atkins, D., Izant, J.G., 1995. Expression and analysis of the greenfluorescent protein gene in the fission yeast Schizosaccha-romyces pombe. Curr. Genet. 28, 585–588.

Boyle, D.B., Coupar, B.E.H., 1988. A dominant selectable markerfor the construction of recombinant poxviruses. Gene 65,123–128.

Bullido, R., Domenech, N., Alvarez, B., Alonso, F., Babin, M.,Ezquerra, A., Ortuno, E., Dominguez, J., 1997a. Characteriza-tion of five monoclonal antibodies specific for swine class IImajor histocompatibility antigens and crossreactivity studieswith leukocytes of domestic animals. Dev. Comp. Immunol.21, 311–322.

Bullido, R., Gomez del Moral, M., Domenech, N., Alonso, F.,Ezquerra, A., Dominguez, J., 1997b. Monoclonal antibodies toa high molecular weight isoform of porcine CD45: biochemi-cal and tissue distribution analyses. Vet. Immunol. Im-munopathol. 56, 151–162.

Ž .Carroll, M.W., Moss, B., 1995. E. coli beta-glucuronidase GUSas a marker for recombinant vaccinia viruses. Biotechniques19, 352–356.

Chakrabarti, S., Brechling, K., Moss, B., 1985. Vaccinia virusexpression vector: coexpression of beta-galactosidase provides

( )J. Domınguez et al.rJournal of Immunological Methods 220 1998 115–121´ 121

visual screening of recombinant virus plaques. Mol. Cell. Biol.5, 3403–3409.

Chalfie, M., 1995. Green fluorescent protein. Photochem. Photo-biol. 62, 651–656.

Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W., Prasher, D.C.,1994. Green fluorescent protein as a marker for gene expres-sion. Science 263, 802–805.

Cubitt, A.B., Heim, R., Adams, S.R., Boyd, A.E., Gross, L.A.,Tsien, R.Y., 1995. Understanding, improving and using greenfluorescent proteins. Trends Biochem. Sci. 20, 448–455.

Earl, P., Moss, B., 1991. Expression of proteins in mammaliancells using vaccinia viral vectors. In: Ausubel, F.M., Brent, R.,Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A.,

Ž .Struhl, K. Eds. , Current Protocols in Molecular Biology.Wiley-Interscience, New York.

Ehrig, T., DJ, O.K., Prendergast, F.G., 1995. Green-fluorescentprotein mutants with altered fluorescence excitation spectra.FEBS Lett. 367, 163–166.

Falkner, F.G., Moss, B., 1988. Escherichia coli gpt gene providesdominant selection for vaccinia virus open reading frameexpression vectors. J. Virol. 62, 1849–1854.

Franke, C.A., Rice, C.M., Strauss, J.H., Hruby, D.E., 1985.Neomycin resistance as a dominant selectable marker forselection and isolation of vaccinia virus recombinants. Mol.Cell. Biol. 5, 1918–1924.

Gonzalez, S., Mendoza, C., Sanchez-Vizcaıno, J.M., Alonso, F.,´1990. Inhibitory effect of African Swine Fever virus onlectin-dependent swine lymphocyte proliferation. Vet. Im-munol. Immunopathol. 26, 71–80.

Heim, R., Prasher, D.C., Tsien, R.Y., 1994. Wavelength mutationsand posttranslational autoxidation of green fluorescent protein.Proc. Natl. Acad. Sci. USA 91, 12501–12504.

Heim, R., Cubitt, A.B., Tsien, R.Y., 1995. Improved green fluo-rescence. Nature 373, 663–664.

Hu, W., Cheng, C.L., 1995. Expression of Aequorea green fluo-rescent protein in plant cells. FEBS Lett. 369, 331–334.

Katz, J.B., Middle, L.A., 1990. Evaluation of thymidine kinaseand neomycin phosphotransferase II positive selection systemsfor recovery of genetically atypical and recombinant DNAvaccine viruses. Biologicals 18, 301–304.

Kremer, L., Baulard, A., Estaquier, J., Poulain Godefroy, O.,Locht, C., 1995. Green fluorescent protein as a new expressionmarker in mycobacteria. Mol. Microbiol. 17, 913–922.

Lorenzo, M.M., Blasco, R., 1998. PCR-based method for theintroduction of mutations in genes cloned and expressed invaccinia virus. Biotechniques 24, 308–313.

Mackett, M., Smith, G.L., Moss, B., 1982. Vaccinia virus: aselectable eukaryotic cloning and expression vector. Proc.Natl. Acad. Sci. USA 79, 7415–7419.

Panicali, D., Paoletti, E., 1982. Construction of poxviruses ascloning vectors: insertion of the thymidine kinase gene fromherpes simplex virus into the DNA of infectious vacciniavirus. Proc. Natl. Acad. Sci. USA 79, 4927–4931.

Peters, K.G., Rao, P.S., Bell, B.S., Kindman, L.A., 1995. Greenfluorescent fusion proteins: powerful tools for monitoringprotein expression in live zebrafish embryos. Dev. Biol. 171,252–257.

Rodriguez, J.F., Rodriguez, D., Rodriguez, J.R., McGowan, E.B.,Esteban, M., 1988. Expression of the firefly luciferase gene invaccinia virus: a highly sensitive gene marker to follow virusdissemination in tissues of infected animals. Proc. Natl. Acad.Sci. USA 85, 1667–1671.

Stearns, T., 1995. Green fluorescent protein. The green revolution.Curr. Biol. 5, 262–264.

Yeh, E., Gustafson, K., Boulianne, G.L., 1995. Green fluorescentprotein as a vital marker and reporter of gene expression inDrosophila. Proc. Natl. Acad. Sci. USA 92, 7036–7040.