Arch Virol (2003) 148: 2077–2097DOI 10.1007/s00705-003-0182-x

Expression at mRNA level of cytokines and A238Lgene in porcine blood-derived macrophages infected

in vitro with African swine fever virus (ASFV)isolates of different virulence

S. Gil1, M. Spagnuolo-Weaver2, A. Canals3, N. Sepulveda4, J. Oliveira1,A. Aleixo5, G. Allan2, A. Leitao1,6, and C. L. V. Martins1

1Laboratorio Doencas Infecciosas, CIISA, Faculdade de Medicina Veterinaria,Lisboa, Portugal

2Virology Department, Veterinary Sciences Division, Belfast,Northern Ireland, UK

3CISA-INIA, Valdeolmos, Spain4Instituto Gulbenkian de Ciencia, Oeiras Portugal

5Roche Diagnostics, Lisboa, Portugal6CVZ, CIISA, Instituto de Investigacao Cientıfica Tropical, Lisboa, Portugal

Received March 25, 2003; accepted July 3, 2003Published online August 18, 2003 c© Springer-Verlag 2003

Summary. Porcine macrophage cultures were infected with two ASFV isolatesof variable virulence and mRNA levels of several relevant macrophage-derivedcytokines were quantified by real time PCR. At six hours post infection, a clearenhancement of mRNA expression of TNFα, IL6, IL12 and IL15 was observedin macrophages infected with the low virulent ASFV/NH/P68 (NHV) whencompared to those infected with the highly virulent ASFV/L60 (L60). Thesequence of the A238L gene homologue to the cellular IκB was found identicalin both viral isolates and its expression at mRNA level was higher in macro-phages infected with NHV when compared to macrophages infected with L60.Furthermore our results suggest a negative correlation between the mRNA ex-pression of A238L gene and the mRNA expression of the above mentionedcytokines (with the exception of IL10) in L60 infected macrophages in oppo-sition to the positive correlation (with exception of the IL1) suggested in NHVinfection. Overall, our data strongly emphasize that virulence of ASFV isolatesmay depend on their capacity to regulate the expression of macrophage-derivedcytokines relevant for the development of host protective responses by yetunknown mechanisms triggered by the virus at early stages of the cellularinfection.

2078 S. Gil et al.

Introduction

African swine fever virus (ASFV) is an icosahedral, cytoplasmic, double-strandedDNA (170–190 kbp) virus, which is the only member of the family Asfarviridae[12]. This virus infects all members of the Suidae family and different species ofticks [36]. The structure ofASFV genome, enzymes and pattern of gene expressionis very similar to the family Poxviridae [46], but the virion is morphologicallydifferent [3, 31].

Distinct ASFV isolates have been reported to induce a range of immune re-sponses and symptoms from fatal hyper-acute, acute and sub-acute, to chronic orunapparent forms of disease in the host [27].

Studies by different authors have failed to demonstrate classical virus neutral-izing antibodies [8, 9]. Indeed, anti-virus antibodies often coexist with viremia[7, 10]. Nevertheless, pigs surviving infection resist challenge inoculation withthe homologous but not with heterologous virus isolate [23, 45]. Although notexcluding a role for antibodies, these findings strongly postulate a role for cellular-based immune mechanisms in protection, reinforced by the characterisation ofASFV-specific CTL [24, 26] and NK activity in pigs experimentally infectedwith the low virulent ASFV/NH/P68, that survive challenge inoculation withASFV/L60 [20].

Independently of virulence, all ASFV isolates replicate in monocytes andmacrophages in vivo and in vitro [22, 48].

Cells of macrophage lineage play an essential role in both innate and acquiredarms of immune response to pathogens and interactions at different levels betweenASFV and pig macrophages will certainly have an impact on the outcome of thehost response against infection. Previous studies have demonstrated impairment ofchemotatic responses and phagocytic capacity in parallel with decreased releaseof toxic oxygen radicals, in porcine macrophages infected in vitro with ASFVof different virulence [25], and SLA antigen expression on these cells was notsignificantly modified [16].

Macrophage-derived cytokines determine the development of inflammatoryresponses that are the first defence against infection and they influence the de-velopment of acquired immune response towards Th1 or Th2 cell activation.Different authors have previously shown changes in expression at both the mRNAand protein levels of different pro-inflammatory cytokines in porcine macrophagesgrown in vitro under different conditions and infected with differentASFV isolates[11, 37], pointing to the relevance of macrophage-derived cytokines on the patho-genesis of ASFV infection. Interestingly, synthesis of pro-inflammatory cytokineshas been described to be controlled by NFκB, which in turn is blocked by theASFV geneA238L that is structurally and functionally homologous to cellular IκB[37, 38].

In this work we have assessed the impact on the mRNA expression of TNFα,IL1, IL6, IL10, IL12 and IL15 on porcine blood derived macrophages, causedby infection with ASFV isolates with distinct virulence (NHV, L60) and wehave evaluated the expression of the viral A238L gene on those cells. A clearenhancement on the expression of TNFα, IL6, IL12 and IL15 was found in NHV

Cytokines and A238L mRNA expression on ASFV infected macrophages 2079

infected cells in comparison to the effect observed in infection with L60 rein-forcing previous data in that cellular based immune responses are found in NHVinfection in vivo [20, 24, 26]. Similarly, an enhanced expression of A238L genewas found on NHV infected cells. Different correlations suggested between theexpression of this gene and several cytokines in both ASFV infections point to therole of other yet unknown viral factors capable of regulating cytokine expressionat macrophage level. These interactions will be relevant for the development ofprotective responses against infection.

Methods

Virus

Two naturally occurring ASFV isolates were used: the low virulent non-hemadsorbing ASFV/NH/P68 (NHV) and the highly virulent ASFV/L60 (L60) [25]. Virus stocks were grownin blood-derived macrophage cultures as previously described [20], for no more than sixpassages. Supernatants from macrophage cultures infected with stock virus at an approximatemultiplicity of infection of 5 (MOI = 5), were collected when CPE was observed in morethan 80% of cells (around 72 h p.i.), clarified by centrifugation and further centrifuged(11,500 r.p.m, GSA rotor, overnight, Sorvall, 4 ◦C). Pellets were suspended (Tris 20 mM,EDTA 1 mM, NaCl 1 M) and resedimented by centrifugation (28,000 r.p.m., SW 28 rotor,Sorvall, 90 min, 4 ◦C), through a cushion of a 25% sucrose. The resulting pellet was finallyre-suspended in culture medium. Viral suspensions were titrated by observation of cyto-pathic effect (CPE) at end-point dilutions [25]. Suspensions of Pseudorabies virus (PrV)were used as clarified supernatants of Vero cell cultures infected and lysed by the virus,obtained after an undetermined number of passages in these cells (kindly supplied by BeneditaCruz, LNIV, Lisbon, Portugal). Virus titration was performed in Vero cells as above forASFV.

Enriched porcine blood derived macrophage cultures and viral infection

Suspensions of peripheral blood mononuclear cells (PBMC) from ten crossbred LargeWhite×Landrace pigs (six month old) were prepared from heparinised peripheral blood samplescollected in aseptic conditions during bleeding at the abattoir, using slight modificationsof protocols previously described [26]. Briefly, heparinised blood samples were incubated(37 ◦C, 15 min), with 10% (v/v) of a 5% (v/v) Dextran T 500 solution in Hank’s balanced salinesolution (HBSS) and supernatant was collected, centrifuged (400 g, 30 min, room temperature)over Ficoll-Hypaque (d = 1.077, Seromed). The interface PBMC were collected, washedtwice and re-suspended in complete culture medium (RPMI 1640 with 100 IU/ml penicillin,100 µg/ml streptomycin and 20 mM HEPES), supplemented with 10% foetal calf serum(FCS), (Gibco BRL, 102770-106) and seeded in T175 tissue culture flasks (Nunc, 156502)at 5 × 106 cells/ml. Cultures were incubated (37 ◦C, 5% CO2 and >90% relative humidity)during 72 h. After incubation, non-adherent cells were removed by serial washing with pre-warmed phosphate buffered saline (PBS) (8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, 0.24 gKH2PO4 per liter, pH 7.4) and adherent macrophages harvested by treatment with 0.8 mMEDTA in this buffer. Cell viability counting by tripan blue dye exclusion was always >90%.

Macrophages were re-suspended at 5 × 106/ml in complete culture medium in sterilepolypropylene centrifuge tubes (Nunc, 366060) and inoculated with working suspensions ofNHV or L60 or PrV at a multiplicity of infection of three (MOI = 3). Tubes kept horizontallywere incubated (37 ◦C, 5% CO2 and >90% relative humidity) with slight mechanical shakingduring 90 min to facilitate virus adsorption. Cells were washed, re-suspended in complete

2080 S. Gil et al.

culture medium and incubation pursued during six hours for mRNA quantification or dur-ing eighteen hours to determine level of infection. For all the experiments, non-inoculatedmacrophages, cultured in complete culture medium were kept as controls.

RNA extraction and quantification

Samples of control and infected macrophages incubated during six hours were washed bycentrifugation in PBS and total RNA from 107 cells was extracted after the addition of1 ml of Trizol (Gibco-BRL, 15596) to individual pellets, following an adapted protocolfrom the manufacturer’s instructions. Briefly, cell disruption was obtained by repetitivepipetting (10 times) and the homogenised samples were kept at room temperature during5 min to allow complete dissociation of nucleoprotein complexes. Chloroform (200 µl) wasadded to the lysates. Tubes capped in “Parafilm” were vigorously handshaked for 15 sec,incubated (15–30 ◦C, 2 min) and finally centrifuged (12,000 r.p.m. Haereus minifuge, 15 min,4 ◦C). Following centrifugation, the upper-phase containing RNA (approximately 600 µl) wastransferred to a new tube, added with isopropanol (500 µl) and glycogen (1µl at 20 mg/ml),incubated (15–30 ◦C, 10 min) and centrifuged (12,000 r.p.m., Haereus minifuge, 4 ◦C). Thesupernatant was carefully removed with a pipette and the precipitated RNA was recovered bycentrifugation with 1 ml of 75% ethanol (7,500 r.p.m., 5 min, 4 ◦C). The RNA was dissolvedin Diethylpirocarbonate (DPC)-treated water and stored at −70 ◦C. Samples of RNA werequantified in spectrophotometer (DU40 Beckman), at 260 and 280 nm (A). Total RNA wascalculated as described in the literature [38].

Reverse transcription of mRNA into cDNA

10 µg of total RNA were mixed with oligo (dT)12−18 primers (Gibco BRL, 18064-014) andincubated (70 ◦C, 10 min). The mixture was chilled on ice and incubated (45 ◦C, 50 min) with 5times concentrated reverse transcriptase buffer (50 mM Tris-HCl pH 8.4, 75 mM KCl, (GibcoBRL, 18064-014); 3 mM MgCl2 (Gibco BRL, 18064-014); 500 µM of each deoxynucleotide;10 mM dithiothreitol (Gibco BRL, 18064-014) and 200 U of Superscript II, RT Moloneymurine leukemia virus reverse-transcriptase (Gibco BRL, 18064-014). The reactions wereincubated (90 ◦C, 5 min) and the samples of the cDNA obtained stored at −20 ◦C until furtheruse.

Oligonucleotide primer pairs

Primer pairs were selected from different exons, as determined from porcine or consensusDNA sequences from other species, to ensure that amplified cDNA could be distinguishedfrom any amplified genomic DNA contaminants. Primers for TNFα, IL1 and IL6 weredesigned on the basis of the sequences obtained from the GenBank Data Bank. Primersfor IL10 and cyclophilin were used as described by Dozois et al. [13]. Primers for IL12 (p40)were obtained as published by Zarlenga et al. [49] and primers for IL15 were used as publishedby Canals et al. [5] (Table 1).

PCR amplification of porcine cytokines and housekeeping gene (cyclophilin)cDNA using a fluorimeter based real-time PCR

Quantitative PCR was performed by using real-time PCR technology in the “Lightcyclerinstrument” (Roche, 2011468). For quantification of mRNA expression of TNFα, IL1, IL6,IL10, IL12, IL15 and cyclophilin, cDNA samples obtained from infected and non-infectedmacrophages, were diluted in DPC treated water (74 µg/60 µl) and used for PCR amplifica-

Cytokines and A238L mRNA expression on ASFV infected macrophages 2081

Table 1. Primers and optimised conditions used for quantification of mRNA expression of TNFα,IL1, IL6, IL10, IL12 and IL15

Cytokine Oligonucleotide sequences (5′–3′) Annealing cDNA PCR(accession temperature (product basenumber) pairs)

Cyclophilin (F)-TAA CCC CAC CGT CTT CTT 55 ◦C 369(F 14571) (R)-TGC CAT CCA ACC ACT CAGTNFα (F)-TGC CTA CTG CAC TTC GAG GTT ATC 65 ◦C 478(M 29079) (R)-TGA GTC GAT CAT CCT TCT CCA GCTIL1 (F)-ACA GAA GTG AAG ATG GCC AAA GTC 55 ◦C 385(X 52731) (R)-TCA TGT TGC TCT GGA AGC TGT ATGIL6 (F)-GTA CAT CCT CGG CAA AAT CTC TGC 55 ◦C 376(M 86722) (R)-GTT CTG TGA ATG CAG CTT ATC CAGIL10 (F)-GCA TCC ACT TCC CAA CCA 62 ◦C 446(L 2001) (R)-CTT CCT CAT CTT CAT CGT CATIL12 (F)-GAT GCT GGC CAG TAC ACC TG 62 ◦C 324(U 08317) (R)-CCT GAT GAA GAA GCT GCTIL15 (F)-CAT ATT TGA GAA GTA CTT CCA TCC AG 62 ◦C 331(U 58142) (R)-GAA GTG TTG ATG AAC ATT TGC AC

tion with the enzymatic system “Lightcycler-FastStart DNA Master Sybr Green I” (Roche,3003230) using an adapted protocol according to manufacturer’s instructions. Briefly, a 3 mMconcentration of MgCl2 reaction mix was prepared: 11.4 µl of water, 1.6 µl of MgCl2,1.5 µl of forward primer and 1.5 µl of reverse primer (5 µM each), 2 µl of master mix(described above) and 2 µl of cDNA. The optimised runs were as follows: 1 cycle of de-naturation (95 ◦C during 600 sec at 20 ◦C/sec ramp rate), 50 cycles of amplification: de-naturation (95 ◦C during 15 sec at 20 ◦C/sec ramp rate), annealing temperature dependingon the cytokine (Table 1), during 5 sec at 20 ◦C/sec ramp rate and extension (72 ◦C during15 sec at 3 ◦C/sec ramp rate) and a final melting program at 65 ◦C during 5 sec at 20 ◦C/secramp rate and heating of the samples until 95 ◦C with 0.1 ◦C/sec ramp rate. The differentquantitative PCRs of the several cytokines tested showed the same PCR efficiency (data notshown).

Plasmids for quantification in RT-PCR

Plasmids for cyclophilin, IL1, IL6 and TNFα, were obtained as previously described [41] andplasmids for IL12 and IL15 were obtained as respectively published by Zarlenga et al. [49]and Canals et al. [5]. Colonies of recombinant E. coli were collected and incubated at 37 ◦Covernight in suspension with vigorously shaking in LB broth medium (10 g of Bacto-trypton,5 g of yeast-extract and 10 g of NaCl) supplemented with ampiciline (0.1 mg/ml). Thesesuspensions were used for amplification with corresponding primers (as described above),using a PCR Perkin-Elmer conventional machine and sequenced to confirm the identityof the amplicon. Dilutions (in number of molecules) of PCR amplicons were quantifiedby spectrophotometry [38] and used as standards in real time PCR for the quantificationof mRNA expression of different cytokines. Quantification of mRNA expression of IL10was assessed using purified amplicons and number of molecules calculated as describedabove.

2082 S. Gil et al.

Optimisation of standard curves and quantification of cytokine cDNA

Quantification of cyclophilin, TNFα, IL1, IL6, IL12 and IL15 was assessed using standardcurves of amplification obtained with 103, 104, 105 and 106 DNA molecules of recombinantplasmids (calculated as described above). For IL10 the standard curves were obtained usingpurified amplicons. Quantification of cDNA of different cytokines was determined as thenumber of molecules using the “fit points method” in the Lightcycler software. The quantifi-cation was obtained with the fluorescence data in the log-phase portion of the curve and acrossing line was set above the level of the noise band to give a value for a crossing point.The crossing point values of the standards (plasmids for TNFα, IL1, IL6, IL12 and IL15and purified amplicon for IL10 at known concentrations of 103, 104, 105 and 106 molecules)were expressed by a standard curve obtained as a fractional cycle number determined at thelog-phase of the product amplification, plotted against log concentration, which was usedto determine the number of molecules of the PCR products. The melting curves obtainedafter amplification were used to verify the specificity of the PCR products/amplicons bycomparison with the melting curves of the correspondent recombinant plasmids.

Control of macrophage infection by ASFV and PrV

Smears of macrophages infected with NHV, L60 and PrV incubated during eighteen hourswere prepared in cytospin slides, fixed in cold acetone, and incubated with FITC-conjugated,anti-ASFV and anti-PrV polyclonal pig sera (kindly supplied by Benedita Cruz, LNIV,Lisbon), following current procedures. Preparations were examined by epifluorescence mi-croscopy and number of infected cells scored in a total of 200 cells/sample.

Sequencing of the ASFV IκB homologue (A238L)

Viral DNA extraction from NHV or L60 was performed using 200µl of virus suspensions at108 CPE50/ml, with the “High Pure Viral Nucleic Acid Kit” (Roche, 1858874) following themanufacturer’s instructions.

Two sets of primers designed in house were used for amplification of viral A238L gene,with sequences complementary to regions in the genome external to the A238L ORF that wasexpected to amplify ASFV genomic DNA, as follows: EIκB (forward): AGA TAA TTA AGAGAT TAC TTT CCA, EIκB (reverse):AAA CAG CTTATC CAAAGAAGT CC (amplicon of818 bp) and IκB (forward): TTA TGT CAT GGA CGC ATT GAA, IκB (reverse): CTC CCCGCT ATA AGG CAT TT (amplicon of 1081 bp). The PCR was performed using 0.06 µg,0.1 µg and 0.3 µg of viral DNA with a proof-reading enzyme: “PWO DNA polymerase”(Roche, 1644947) in a Perkin-Elmer conventional PCR machine using the following cyclingconditions: For the first pair of primers (EIκB): denaturation of the template at 94 ◦C for 2 min,10 cycles of: denaturation at 94 ◦C for 15 sec, annealing at 55 ◦C for 30 sec and elongationat 72 ◦C for 1 min and 30 sec, 20 cycles of: denaturation at 94 ◦C for 15 sec, annealing at55 ◦C for 30 sec and elongation at 72 ◦C for 1 min and 30 sec plus cycle elongation of 5 secfor each cycle and a final elongation 7 min at 72 ◦C. For the second pair of primers (IκB):denaturation of the template at 94 ◦C for 2 min, 10 cycles of: denaturation at 94 ◦C for 15 sec,annealing at 65 ◦C for 30 sec and elongation at 72 ◦C for 1 min and 30 sec, 20 cycles of:denaturation at 94 ◦C for 15 sec, annealing at 65 ◦C for 30 sec and elongation at 72 ◦C for1 min and 30 sec plus cycle elongation of 5 sec for each cycle and a final elongation 7 min at72 ◦C. PCR products were purified using “High Pure PCR Product Purification Kit” (Roche,1732668), following manufacturer’s instructions.

Three samples of each of the strands of A238L of NHV and L60 were sequenced in anautomatic sequencing machine.

Cytokines and A238L mRNA expression on ASFV infected macrophages 2083

Quantification of A238L expression

mRNA coding for A238L was quantified in cDNA samples obtained from the 10 macrophagecultures used for cytokine mRNA studies. This quantification was performed using two setsof primers: 1) EIκB (as described above) and 2) IIκB, a set of primers designed in house andexpected to amplify DNA complementary to mRNA expression of A238L during infection

Table 2. Statistical analysis of results on mRNA expression of IL6 (A), TNFα (B), IL12 (C) andIL15 (D) macrophage-derived cytokines, using the parametrical modified ANOVA test

A

NL (IL6) IL6

Comparison Difference Standard error CI (95%) Ratio CI (95%)

L60 −0.636 0.495 (−2.044, 0.722) 0.529 (0.130, 2.164)C NHV −2.836 0.495 (−4.244, −1.428) 0.059 (0.014, 0.240)

PrV −2.367 0.495 (−3.775, −0.959) 0.094 (0.023, 0.383)

C 0.636 0.495 (−0.772, 2.044) 1.889 (0.462, 7.721)L60 NHV −2.200 0.495 (−3.608, −0.792) 0.111 (0.027, 0.724)

PrV −1.731 0.495 (−3.139, −0.323) 0.177 (0.043, 0.453)

C 2.836 0.495 (1.428, 4.244) 17.047 (4.170, 69.686)NHV L60 2.200 0.495 (0.792, 3.608) 9.025 (2.208, 36.892)

PrV 0.469 0.495 (−0.939, 1.877) 1.598 (0.391, 6.534)

C 2.367 0.495 (0.959, 3.775) 10.665 (2.609, 43.598)PrV L60 1.731 0.495 (0.323, 3.139) 5.646 (1.381, 23.081)

NHV −0.469 0.495 (−1.877, 0.939) 0.626 (0.153, 2.557)

B

NL (TNF) TNFα

Comparison Difference Standard error CI (95%) Ratio CI (95%)

L60 −0.315 0.616 (−2.069, 1.440) 0.730 (0.126, 4.221)C NHV −2.796 0.616 (−4.551, −1.042) 0.061 (0.011, 0.353)

VDA −1.110 0.616 (−2.864, 0.645) 0.330 (0.057, 1.906)

C 0.315 0.616 (−1.440, 2.069) 1.370 (0.237, 7.917)L60 NHV −2.482 0.616 (−4.236, −0.727) 0.084 (0.014, 0.483)

VDA −0.795 0.616 (−2.550, 0.960) 0.452 (0.078, 2.612)

C 2.796 0.616 (1.042, 4.551) 16.379 (2.835, 94.727)NHV L60 2.482 0.616 (0.727, 4.236) 11.965 (2.069, 69.131)

VDA 1.687 0.616 (−0.068, 3.441) 5.403 (0.934, 31.218)

C 1.110 0.616 (−0.645, 2.864) 3.034 (0.525, 17.532)VDA L60 0.795 0.616 (−0.960, 2.550) 2.214 (0.383, 12.807)

NHV −1.687 0.616 (−3.441, 0.068) 0.185 (0.032, 1.070)

(continued)

2084 S. Gil et al.

Table 2 (continued)

C

NL (IL12) IL12

Comparison Difference Standard error CI (95%) Ratio CI (95%)

L60 −1.034 0.630 (−2.827, 0.758) 0.356 (0.059, 2.134)C NHV −4.200 0.630 (−5.993, −2.408) 0.015 (0.002, 0.090)

PrV −1.674 0.630 (−3.466, 0.118) 0.187 (0.031, 1.125)

C 1.034 0.630 (−0.758, 2.827) 2.812 (0.469, 16.895)L60 NHV −3.166 0.630 (−4.958, −1.374) 0.042 (0.007, 0.253)

PrV −0.639 0.630 (−2.432, 1.153) 0.528 (0.088, 3.168)

C 4.200 0.630 (2.408, 5.993) 66.686 (11.112, 400.62)NHV L60 3.166 0.630 (1.374, 4.958) 23.712 (3.951, 142.31)

PrV 2.526 0.630 (0.734, 4.319) 12.503 (2.083, 75.113)

C 1.674 0.630 (−0.118, 3.466) 5.333 (0.889, 32.008)PrV L60 0.639 0.630 (−1.153, 2.432) 1.895 (0.316, 11.382)

NHV −2.526 0.630 (−4.319, −0.734) 0.080 (0.013, 0.480)

D

NL (IL15) IL15

Comparison Difference Standard error CI (95%) Ratio CI (95%)

L60 −0.017 0.736 (−1.527, 1.492) 0.983 (0.217, 4.446)C NHV −3.122 0.736 (−4.632, −1.613) 0.044 (0.010, 0.199)

PrV −1.380 0.736 (−2.889, 0.130) 0.252 (0.056, 1.139)

C 0.017 0.736 (−1.492, 1.527) 1.017 (0.225, 4.604)L60 NHV −3.105 0.736 (−4.615, −1.596) 0.045 (0.010, 0.203)

PrV −1.363 0.736 (−2.872, 0.147) 0.256 (0.057, 1.158)

C 3.122 0.736 (1.613, 4.632) 22.692 (5.018, 102.72)NHV L60 3.105 0.736 (1.596, 4.615) 22.309 (4.933, 100.99)

PrV 1.743 0.736 (0.233, 3.252) 5.714 (1.262, 25.842)

C 1.380 0.736 (−0.130, 2.889) 3.975 (0.878, 17.975)PrV L60 1.363 0.736 (−0.147, 2.872) 3.908 (0.863, 17.672)

NHV −1.743 0.736 (−3.252, −0.233) 0.175 (0.039, 0.792)

NL – Natural logarithm

(amplicon of 292 bp) as follows: (forward): CTG GAA GAG GTG TTG GGC TA (reverse):ACC GGT GGA AAA CAT CCA G.

The optimised runs for amplification of contaminant viral A238L (double stranded withEIκB primers) from viral infection were: 1 cycle of denaturation (95 ◦C during 600 sec at20 ◦C/sec ramp rate), 50 cycles of: denaturation (95 ◦C during 15 sec at 20 ◦C/sec ramp rate),annealing (55 ◦C during 5 sec at 20 ◦C/sec ramp rate) and extension (72 ◦C during 1 m 30 secat 3 ◦C/sec ramp rate), melting at 65 ◦C during 5 sec at 20 ◦C/sec ramp rate and heating

Cytokines and A238L mRNA expression on ASFV infected macrophages 2085

of the samples until 95 ◦C with 0.1 ◦C/sec ramp rate. For amplification of A238L derived fromviral replication (single stranded with IIκB primers): 1 cycle of denaturation (95 ◦C during600 sec at 20 ◦C/sec ramp rate), 50 cycles of: denaturation (95 ◦C during 15 sec at 20 ◦C/secramp rate), annealing (62 ◦C during 5 sec at 20 ◦C/sec ramp rate) and extension (72 ◦C during15 sec at 3 ◦C/sec ramp rate), melting at 65 ◦C during 5 sec with 20 ◦C/sec ramp rate andheating until 95 ◦C with 0.1 ◦C/sec ramp rate.

Calculation and statistical analysis

Quantification of mRNA expression of different cytokines was defined as the ratio betweenthe number of molecules of each cytokine and the number of molecules of cyclophilinas housekeeping gene (N molecules of mRNA of each cytokine/N molecules of mRNAof cyclophilin) for each cDNA sample. Their statistical analysis was based on repeated-measures models as published by Neter et al. [33]. To attain a good validation of our fittedmodels, we applied the natural logarithm to the data (NL). Moreover, that transformationleads us to a comparison of the four populations in terms of mean ratios (e.g., ratio betweenmean expression of IL1 in PrV and in L60), facilitating the interpretation of the results(Table 2). For the comparison of the expression of a given cytokine, we computed thesimultaneous confidence intervals, adjusted by the Bonferroni method, for all the ratiosbetween the mean expressions in a pair of populations. For the quantification of the expressionof A238L in macrophages experimentally infected with NHV or L60, the results were definedas the number of molecules of A238L amplified with IIκB primers (Table 3) subtractedby the number molecules of A238L amplified with EIκB primers; these results were thennormalised by cyclophilin and also logaritmised (LN). Results obtained (NHV versus L60)

Table 3. A238L expression in macrophages inoculated with NHV and L60

cDNA samplesa (IIκB-EIκB)/cyclophilin (IIκB-EIκB)/cyclophilinNHV infected L60 infectedmacrophagesb macrophagesc

1 2.91 × 10−01 0.97 × 10−01

2 1.49 × 10−01 0.46 × 10−01

3 1.17 × 10−01 1.90 × 10−01

4 0.25 × 10−01 0.03 × 10−01

5 0.28 × 10−01 0.02 × 10−01

6 0.05 × 10−01 0.004 × 10−01

7 0.04 × 10−01 0.001 × 10−01

8 0.40 × 10−01 0.90 × 10−01

9 7.64 × 10−01 0.37 × 10−01

10 1.80 × 10−01 13.9 × 10−01

acDNA from macrophage cultures of ten different pigs infected in vitro with NHV andL60

bNumber of molecules of A238L amplified with IIκB primers (IIκB) subtracted by thenumber molecules of A238L amplified with EIκB primers (EIκB); these results were thennormalised by cyclophilin: (cDNA A238L-viral DNA NHV A238L)/cyclophilin

cNumber of molecules of A238L amplified with IIκB primers (IIκB) subtracted by thenumber molecules of A238L amplified with EIκB primers (EIκB); these results were thennormalised by cyclophilin: (cDNA A238L-viral DNA L60 A238L)/cyclophilin

2086 S. Gil et al.

Table 4. Correlation between the mRNA expression of TNFα, IL1, IL6, IL10, IL12 and IL15 withthe A238L expression in macrophage cultures inoculated with NHV

Virus Coefficient NL (TNFα) NL (IL1) NL (IL6) NL (IL10) NL (IL12) NL (IL15)

NHV Pearson 0.421 −0.047 0.240 0.652 0.157 0.334L60 Pearson −0.388 −0.586 −0.561 0.615 −0.329 −0.192

NL Natural logarithm

were compared using the unilateral paired T test (H0: NHV = L60 versus H1: NHV > L60).Correlation between mRNA expression quantification of each cytokine and expression ofA238L in macrophages experimentally infected with NHV or L60 was assessed by thePearson correlation coefficient test. The level of significance α used in all tests was 5%(Table 4).

Results

Expression of mRNA coding for TNFα, IL1, IL6, IL10,IL12, IL15 and cyclophilin in porcine blood derived

macrophages infected with L60, NHV and PrV

Quantification by real-time PCR of mRNA expression of pro-inflammatory cy-tokines TNFα, IL1, IL6, and IL10, IL12 and IL15 was determined in non-stimulated, three days-old blood derived macrophage cultures from ten pigs. Eachculture was divided into four samples for RT-PCR analysis: non-infected controls,infected in vitro for six hours (after an adsorption period of ninety minutes) withNHV, L60 and PrV. Approximately 70–80% of cells were infected in all cases asidentified by direct immunofluorescence (data not shown).

Preliminary analysis of results shown as the average ratio ± standard error(AVG ± SE), between the number of molecules of each cytokine and the numberof molecules of the housekeeping gene cyclophilin, (Fig. 1 and Fig. 2) suggestedan enhancement of mRNA expression of the majority of cytokines tested inmacrophages infected with the NHV isolate. This was confirmed by the resultsof Table 2, which show that there are evidences favouring a general enhance-ment of the cytokines production in the macrophages infected with less virulentvirus (NHV) compared with the macrophages infected with the highly virulentvirus (L60). The results of Table 2A reveal that control cells seem to induceidentical mRNA production of IL6 as the L60 infected macrophages and lessthan PrV (11 times less) and NHV (17 times less) infected macrophages; L60infected macrophages seem to induce less mRNA expression than the other in-fected macrophages (9 times less than the NHV infected macrophages and 6 timesless than the PrV infected macrophages); NHV and PrV infected macrophagesseem to produce identical mRNA expression of IL6. In summary, the resultson IL6 mRNA expression suggest two clusters: one composed with the con-trol and L60 infected macrophages and another composed with the NHV and

Cytokines and A238L mRNA expression on ASFV infected macrophages 2087

Fig. 1. Quantification of mRNA expression of pro-inflammatory cytokines TNFα (A), IL1(B) and IL6 (C) in macrophage cultures from ten different pigs. Each macrophage culture wasdivided in four samples: control (non-infected) macrophages , macrophages infected withNHV , macrophages infected with L60 and macrophages infected with PrV . Resultsare shown as average ratio ± standard error (AVG ± SE), between the number of molecules

of each cytokine and the number of molecules of the house keeping gene cyclophilin

PrV infected macrophages (cluster with the highest mRNA expression of thiscytokine).

TNFα mRNA expression (Table 2B) was identical in L60 and PrV infectedmacrophages when compared to control macrophages, but NHV infected macro-phages seem to induce higher mRNA expression of this cytokine in comparison tocontrol macrophages (16 times more). L60 infected macrophages seem to inducethe same mRNA expression of TNFα as PrV infected macrophages and less thanin NHV infected macrophages (12 times less), however NHV and PrV infectedmacrophages seem to induce the same mRNA expression of TNFα. Although theexpression of TNFα was higher in NHV infection when compared to L60, theabove mentioned inferences, did not allow statistical differentiation of the viralinfected macrophages in terms of clusters.

2088 S. Gil et al.

Fig. 2. Quantification of mRNA expression of IL12 (A), IL15 (B) and IL10 (C) in macrophagecultures from ten different pigs. Each macrophage culture was divided in four samples:control (non-infected) macrophages , macrophages infected with NHV , macrophagesinfected with L60 and macrophages infected with PrV . Results are shown as averageratio ± standard error (AVG ± SE), between the number of molecules of each cytokine and the

number of molecules of the house keeping gene cyclophilin

Statistical analysis of mRNA expression of IL12 (Table 2C) shows that controlmacrophages seem to induce the same mRNA expression as L60 and PrV infectedmacrophages and lower than NHV infected macrophages (67 times less); L60infected macrophages seem to induce identical mRNA expression of IL12 thanPrV infected macrophages and less than NHV infected macrophages (24 timesless); NHV infected macrophages seem to induce higher mRNA expression ofIL12 than PrV infected macrophages (13 times more). Resuming, the results onmRNA expression of IL12 suggest two clusters: one composed with the control,L60 and PrV infected macrophages and another composed with the NHV infectedmacrophages (cluster with the highest mRNA expression of IL12).

Analysis of mRNA expression of IL15 (Table 2D) shows that control macro-phages seem to induce the same mRNA expression as L60 and PrV infected

Cytokines and A238L mRNA expression on ASFV infected macrophages 2089

macrophages and less than NHV infected macrophages (23 times less); L60 in-fected macrophages seem to induced the same mRNA expression of IL15 than PrVinfected macrophages and less than NHV infected macrophages (22 times less);NHV infected macrophages seem to induced higher mRNA expression of IL15than PrV infected macrophages (5 times more). In summary, the mRNA expressionof IL15 suggests two clusters: one composed with the control macrophages, L60and PrV infected macrophages and another composed with the NHV infectedmacrophages (cluster with the highest mRNA expression of IL15).

Statistical analysis of mRNA expression of IL1 and IL10 (data not shown)didn’t reveal a clear difference among the different viral infections (NHV, L60 andPrV). The IL1 mRNA expression seems identical in NHV, L60 and PrV infectedmacrophages when compared to control macrophages; L60 infected macrophagesseem to induce a lower mRNA expression of IL1 than the other virus-infectedmacrophages (16 times less than NHV and 12 times less than PrV); NHV andPrV infected macrophages seem to induce the same mRNA expression of IL1; inthis case, the identical levels of viral infected and control macrophages did notallow to discriminate the four populations. IL10 mRNA expression is statisticallyequivalent in L60 and PrV infected macrophages and lower than NHV infectedmacrophages (5 times less); L60 infected macrophages seem to induce less mRNAexpression of IL10 than the other infected macrophages (9 times less than NHVinfected macrophages and 4 times less than PrV infected macrophages), but asNHV and PrV infected macrophages seem to induce the same mRNA expressionof IL10 we can’t clearly discriminate the 4 populations.

According to 5% level of significance, similar levels of mRNA expressionof cyclophilin, used as housekeeping gene to normalize results obtained in cy-tokine expression, were accepted in non-infected control macrophages and in

Fig. 3. Expression of cyclophilin in number of molecules of mRNA in control (C) non-infected and in NHV, L60 and PrV infected macrophage cultures from ten different pigs.

Results are shown as average ± standard error (AVG ± SE), for p = 0.06 (ANOVA)

2090 S. Gil et al.

macrophages infected with NHV, L60 and PrV (Fig. 3) in which the p valuedetermined by ANOVA was 0.0643.

Sequencing of A238L

A238L sequences of NHV and L60 were compared to those previously publishedfor L60 [32] using the multi align software (Corpet, 1988) and they showed 100%identity in both strands (confirmed at least once).

A238L quantification

A238L gene expression in macrophages inoculated with NHV and L60 wasquantified by real-time PCR in the same macrophage cultures used for quan-tification of mRNA expression of different cytokines (see above). Expression ofA238L gene in these cells was calculated by the number of molecules of A238Lamplified with IIκB primers (cDNA) minus the number of molecules of A238Lamplified with EIκB primers (double stranded viral DNA contaminant), theseresults were normalised by reference to cyclophilin levels. Reactions with EIκBprimers showed only marginal amplification usually 103 times less than with IIκBprimers (data not shown).

The hypothesis of higher mRNA expression of A238L in macrophages in-fected with NHV than in macrophages infected with L60 was accepted using theunilateral paired T test (1.22 times more and p = 0.02).

Analysis of correlation between mRNA expressionof TNFα, IL1, IL6, IL10, IL12 and IL15

with the expression of A238L

Attempting to correlate mRNA expression of the different cytokines and expres-sion of A238L gene in the ten macrophages cultures infected with NHV and L60,the observed values, shown in Fig. 1, Fig. 2 and Table 3, were statistically analyzedby the Pearson correlation coefficient test. As shown on Table 4, observed corre-lation between the mRNA expression A238L and mRNA expression of cytokinesin L60 infected macrophages was in general negative (with the exception IL10,which was positive). The observed correlation between the mRNA expressionA238L and mRNA expression of cytokines in NHV infected macrophages wasalmost null for IL1 and positive for the other cytokines. Despite of their positivityor negativity, those observed values aren’t statistically different from zero (absenceof correlation). This fact can be due to the inaccuracy of the null large sampledistribution for the Pearson correlation coefficient.

Discussion

Regulation of pro-inflammatory cytokines at both mRNA and product levelshas been demonstrated in porcine macrophages infected with virulent ASFV

Cytokines and A238L mRNA expression on ASFV infected macrophages 2091

isolates [11, 37], pointing to the role of the A238L gene in those events [37].Our studies extend these findings specifically by examining a broader range ofrelevant cytokines, by comparing both low and highly virulent ASFV isolatesand finally attempting to relate how these phenomena may correlate with theexpression of the ASFV A238L gene, homologue to cellular IκB. Experimentswere deliberately performed in cultures of non-stimulated porcine blood de-rived macrophages, infected in vitro with the highly virulent ASFV/L60 andthe less virulent ASFV/NH/P68 at six hours post infection. At this time point,other authors have shown that early viral proteins are already synthesised andviral DNA replication is already concluded in different in vitro infection models[14, 44]. Studies on mRNA expression of several cytokines in porcine PBMCstimulated with ConA was characterized from two till seventy two hours andthe majority of cytokines was already expressed at six hours of stimulation [13].Furthermore as reported by Powell et al. [37], inhibition of pro-inflammatorycytokine gene expression attributed to A238L expression is observed as soon astwo to four hours in porcine macrophages infected with ASFV isolates of distinctvirulence.

Our data clearly suggest a different effect of the infection by both isolates inthat statistically significant increase of mRNA expression of TNFα, IL6, IL12 andIL15 was identified in macrophages infected with the low virulent NHV, whencompared with control non-infected macrophages and with macrophages infectedwith the highly virulent L60 isolate.

It is of interest to consider how these changes may relate to the pathogenesis ofinfection and to the development of the host immune responses.Amongst its manyactivities, IL6 promotes the final maturation of B cells into plasma cells, but alsopromotes IL2 production and T cell differentiation [18, 19]. The increased mRNAexpression of IL6 in NHV infected macrophages (17 times more than in controlcells and 9 times more than in L60 infected cells) may correlate with the overactivation of the B cell compartment and the associated hypergammaglobulinemiaobserved in chronic infections by others [20, 34, 35].

TNFα mRNA expression in NHV infected macrophages was 16 times in-creased in relation to control cells and 12 times increased when compared toL60 infected macrophages. These results clearly suggest a higher expression ofthis cytokine in NHV infection. Previous studies by del Moral et al. [11] showincreased mRNA expression of TNFα in macrophages infected with the virulentE-75 ASFV isolate. In contrast, LPS – stimulated macrophages, infected with thehighly virulent isolates OUR T/88-1 and Malawi LIL/20/1, show decreased mRNAexpression of pro-inflammatory cytokines (TNFα, IFNα, IL8) and increased ex-pression of anti-inflammatory TGFβ [37].

TNFα is a mediator of several inflammatory and immune functions, it reg-ulates the growth of many cell types, activates macrophages and it can triggerapoptosis [47]. This cytokine acts in a complex way through two receptors intarget cells (TNF-R1 and TNF-R2), involved in the activation of NFκB pathwayand in apoptosis [1]. It is possible that upper regulation of this cytokine by NHVinfection may induce beneficial inflammatory responses in the host followed by

2092 S. Gil et al.

the activation of cellular protective mechanisms potentially responsible for animalsurvival, in contrast to the fatal effect of L60 infection [20, 26].

mRNA IL12 expression was found to be increased 67 times more in NHVinfected macrophages when compared to L60 infected macrophages and controlcells. IL12 induces the production of IFNγ, favours the differentiation of Th1 cells,it enhances NK cell proliferation and cytotoxicity, immunoglobulin secretion andinduces LAK cell proliferation in association with TNFα and IL15 [43].

IL15 plays a very important role on the proliferation of both cytotoxic andhelper T cells, activation of NK cells [6], recruitment of neutrophils and lym-phocytes to sites of inflammation and it is responsible for the appearance oflymphokine-activated killer (LAK) cells [4]. In our experiments, as described forIL12, significant increased levels of IL15 mRNA in NHV infected macrophageswere found when compared to all groups tested (23 times more than in L60 infectedcells). Altogether results on IL12 and IL15 mRNA expression may in part explainthe development of previously reported relevant cellular immune responses inpigs inoculated with the NHV isolate i.e. ASFV-specific CTL activity [26] andNK activity [20]. Moreover they may also contribute to a better understanding ofthe mechanisms underlying LAK cell activity previously reported by Scholl et al.[40].

IL1 is a major co-stimulator of Th2 cells, which stimulates an acute phaseresponse and antibody production [17, 21]. Our results didn’t show significantdifferences on mRNA expression of this cytokine in macrophages infected withNHV and L60 when compared to control cells, which may suggest that TH2responses are not blocked in both infections.

Relevant activities of IL10 include inhibition of activation of Th1 and NK cells,suppression secretion of IL1, IL6, TNFα in activated macrophages and down-regulation of MHC class II expression [30]. Results obtained in our experimentsdid not show a significant difference on the mRNA expression of this cytokineamong the four populations of macrophages tested, and thus a possible role ofIL10 in ASFV pathogenesis is not clear.

As a comparison to ASFV infection, macrophages were infected with atten-uated PrV. In this case higher expression of IL1 and IL6 mRNAs was observedwhen compared to control cells suggesting an activation of a Th2-type cytokineresponse. These results contrast with the observation of a Th1-type response inmice inoculated with virulent PrV [2, 15] and may be related to the use of anattenuated strain in our experiments. Importantly, the cytokine mRNA profilesof PrV infected macrophages are in between those observed in NHV and L60infections (similar to TNFα, IL1, IL6, IL10 in NHV infection and similar to IL12and IL15 in L60 infection) and thus strengthen the validity of the results obtainedin macrophage infected with the ASFV isolates, which in turn is reinforced bythe fact that similar levels of cyclophilin mRNA (housekeeping gene) expressionwere identified in all groups of macrophages tested.

Synthesis of pro-inflammatory cytokines is controlled by NFκB pathway.NFκB is a dimmer of two heterologous proteins (p65 and p50) held in an inactivecomplex with an endogenous inhibitor, IκB, in the cytoplasm.After cell activation,

Cytokines and A238L mRNA expression on ASFV infected macrophages 2093

IκB is phosphorylated and subsequently degraded by cytoplasmic proteosomes.The released NFκB translocates to the nucleus and activates the synthesis of pro-inflammatory cytokines [37]. The A238L gene of ASFV has 21% of identity and40% of homology at the amino acid level to porcine IκBα [36] and can replaceand bind to the p65 sub-unit of NFκB [38, 42] blocking the NFκB translocationto the nucleus which leads to a reduction of pro-inflammatory cytokines mRNAlevels at 4 h after infection [37]. As the A238L sequence lacks the phosphorylationsites it may indicate that this IκB homologue fails to respond to stimulation of thecells and retains his ability to bind to NFκB [37]. Furthermore A238L inhibits theactivity of the calcium/calmodulin-regenerated phosphatase calcineurin (CaN)[28] that regulates a number of different pathways including the NFAT familyof transcription factors. These facts reinforce that A238L protein is a potentimmunosuppressor that may contribute for the viral evasion to the host immuneresponse [29].

As reported by others for different ASFV isolates [32, 38], we found 100%homology in both strands of the gene A238L in NHV and L60. The expression ofthis gene was quantified in macrophages infected with both isolates and we havefurther attempted to correlate it’s expression with the expression of macrophage-derived cytokines at mRNA level in both ASFV infections. For this purpose,DNA samples from macrophage cultures used to quantify mRNA expression ofcytokines as described above were used to quantify mRNA expression of A238Lderived from replication of NHV and L60. Significantly higher mRNA expressionof A238L in macrophages infected with NHV was observed using the statisticalunilateral T test when compared to the mRNA expression of this gene in L60infected macrophages (1.22 times more, p = 0.02). As expected, quantificationof mRNA expression of A238L in non-infected control macrophages was alwaysnegative (data not shown). Higher levels of A238L expression in macrophagesinfected with NHV may be due to differences in kinectics of infection, althoughin our experience similar patterns of replication in terms of viral yelds are observedon porcine macrophages infected with both NHV and L60 isolates.

Although not statistically validated due to the small dimension of the pop-ulations tested, results on the correlation between A238L and cytokine geneexpression suggest a negative correlation between the mRNA expression ofA238Land pro-inflammatory TNFα, IL1 and IL6 in L60 infected macrophages. Thismay relate to the inactivation of NFκB by the A238L gene in accordance withstudies by others using highly virulent ASFV [37]. Similar results observed inrelation to IL12 and IL15 may suggest a role of the A238L on the regulation ofthese cytokines. In contrast, results on NHV infected macrophages suggest thatexpression of A238L gene is positively correlated with all cytokines tested withthe exception of IL1 in which it was found null. The positive correlation suggestedbetween the mRNA expression of pro-inflammatory cytokines and A238L ismost interesting. The statistically significant higher expression of A238L genein NHV versus the L60 infected macrophages and the differences suggested inthe correlation of the gene’s expression in both infections, with the expressionof several cytokines at mRNA level, may indicate that regulation of cytokine

2094 S. Gil et al.

expression trough NFκB may be inhibited later in infection in NHV infectedcells when compared to L60 or alternatively, it is possible that inactivation ofNFκB is blocked by the over synthesis of A238L through an yet unknown viralderived regulatory mechanism(s). The suggested positive correlation between theexpression of A238L and IL10 mRNA in both ASFV infections points to a role ofthe viral gene in the regulation of this cytokine. Development of studies similarto the ones reported on this paper, using A238L-deleted ASFV should help toclarify the role of this gene on the regulation of macrophage-derived cytokineexpression.

Overall our data clearly suggest significant increased expression at mRNAlevel of porcine macrophage-derived cytokines relevant for the development ofinflammatory and cellular immune responses, early post infection in vitro withthe low virulent NHV in comparison to the effect observed in infection with theL60 isolate. Differences in quantification of A238L gene in both viral infectionsand different correlations suggested between its expression and the expression ofseveral cytokines open new insights for the characterisation of virulence factors inASFV isolates able to regulate fundamental macrophage functions at early stagesof infection.

Acknowledgements

This work was supported by the projects FAIR5-CT97-3441, QLK2-CT-2001-02216, by thePhD grant BD/PRAXIS/11528/97 and by CIISA. The authors are gratefull to Dr. MichaelParkhouse from Instituto Gulbenkian de Ciencia for his suggestions in reviewing this paper, toDr. Jorge Carneiro from the same institution for his help in facilitating the statistical analysisof our data, to Dr. Luis Sardinha from CASO for providing the conditions for blood samplingand to Miss Maria de Jesus Silva for the excellent technical assistance.

References1. Ashkenazi A, Dixit VM (1998) Death receptors: signalling and modulation. Science 281:

1305–13082. Bianchi AT, Moonen-Leusen HW, van Milligen FJ, Savelkoul HF, Zwart RJ, Kimman

TG (1998) A mouse model to study immunity against pseudorabies virus infection:significance of CD4+ and CD8+ cells in protective immunity. Vaccine 16: 1550–1558

3. Breese SS, Pan IC (1978) Electron microscopic observation of ASFV development invero cells. J Gen Virol 40: 499–502

4. Burton JD, Bamford RN, Peters C, Grant AJ, Kurys G, Ygoldman CK, Brennan J,Roessler E, Waldmann TA (1994) A lymphokine, provisionally designated interleukinT and produced by a human adult T-cell leukemia line, stimulates T-cell proliferationand the induction of lymphokine-activated killer cells. Proc Natl Acad Sci USA 91:4935–4939

5. Canals A, Grimm DR, Gasbarre LC, Lunney JK, Zarlenga DS (1997) Molecular cloningof cDNA encoding porcine interleukin-15. GENE 195: 337–339

6. Carson WE, Fehniger TA, Haldar S, Eckhert K, Lindemann MJ, Lai C, Croce CM,Baumann H, Caligiuri M (1997) A potential role for interleukin-15 in the regulation ofhuman natural killer cell survival. J Clin Invest 99: 937–943

Cytokines and A238L mRNA expression on ASFV infected macrophages 2095

7. Coggins L (1974) African swine fever pathogenesis. Prog Med Virol 18: 48–638. De Boer CJ (1967) Studies to determine neutralizing antibody in sera from animals

recovered fromAfrican swine fever and laboratory animals inoculated withAfrican swinefever with adjuvants. Arch Ges Virusforsch 20: 164–170

9. De Boer CJ, Hess WR, Dairdiri AH (1969) Studies to determine the presence ofneutralizing antibody in sera and kidneys from swine recovered from African swinefever. Arch Ges Virusforsch 27: 44–54

10. De Boer CJ, Pan IC, Hess WR (1972) Immunology of African swine fever. J Am Vet MedAssoc 160: 528–532

11. Del Moral G, Ortuno E, Fernandez-Zapatero P, Alonso F, Alonso C, Ezquerra A,Domınguez J (1999) African swine fever virus infection induces Tumor Necrosis Factoralpha production: implications in pathogenesis. J Virol 73: 2173–2180

12. Dixon LK, Costa JV, Escribano JM, Rock DL, Vinuela E, Wilkinson PJ (2000) FamiliyAsfarviridae. In: van Regenmortel MHV, Fauquet CM, Bishop DHL, Carstens EB,Estes MK, Lemon SM, McGeoch DJ, Maniloff J, Mayo MA, Pringle CR, Wickner RB(eds), Virus taxonomy: classification and nomenclature of viruses. Seventh Report ofthe International Committee on Taxonomy of Viruses, pp 159–165. Academic Press,San Diego

13. Dozois MD, Oswald E, Gautier N, Serthelon J, Fairbrother J, Oswald IP (1997) Areverse transcription-polymerase chain reaction method to analyse porcine cytokine geneexpression. Vet Immunol Immunopathol 58: 287–300

14. Esteves A, Marques MI, Costa JV (1986) Two-dimensional analysis of African swinefever virus proteins and proteins induced in infected cells. Virology 152: 192–206

15. Fischer T, Buttner M, Rziha HJ (2000) T helper 1-type cytokine transcription in peripheralblood mononuclear cells of pseudorabies virus (Suid herpesvirus 1)-primed swineindicates efficient immunization. Immunology 101: 378–387

16. Gonzalez-Juarrero M, Mebus CA, Pan IC, Revilla Y, Alonso JM, Lunney JK (1992)Swine leukocyte antigen (SLA) and macrophage marker expression on bothAfrican swinefever virus (ASFV) infected and not infected primary porcine macrophage cultures. VetImmunol Immunopathol 32: 243–259

17. Hultuer L, Kolsch S, Stassen M, Kaspers U, Kreuer J, Mailhammer R, Moller J, BroszeitH, Schmidt E (2000) In activated mast cells, IL1 up-regulates the production of severalTh2-related cytokines including IL9. J Immunol 164: 5556–5563

18. Kishimoto T, Hibi M, Murakami M, Narazaki M, Saito M, Taga T (1992) The molecularbiology of interleukin 6 and its receptor. Ciba Found Symp 167: 5–16

19. Kuhweide R, Van Damme J, Ceuppens JL (1990) Tumor necrosis factor-alpha andinterleukin 6 synergistically induce T cell growth. Eur J Immunol 20: 1019–1025

20. Leitao A, Cartaxeiro C, Coelho R, Cruz B, Parkhouse RME, Portugal FC, VigarioJD, Martins CLV (2001) The non-hemadsorbing African swine fever virus isolateASFV/NH/P68 provides a model for defining the protective anti-virus immune response.J Gen Virol 82: 513–523

21. Litchman AH, Chin J, Schmidt JA, Abbas AK (1988) Role of IL1 in the activation of Tlymphocytes. Proc Natl Acad Sci USA 85: 9699–9703

22. Malmquist WA, Hay D (1960) Hemadsorption and Cytophatic effect produced byAfricanswine fever virus in swine bone marrow and buffy coat cultures.Am JVet Res 21: 104–108

23. Malmquist WA (1963) Serologic and immunologic studies withAfrican swine fever virus.Am J Vet Res 24: 450–459

24. Martins C, Mebus C, Scholl T, Lawman MJP, Lunney J (1988a) Virus-specific CTLin SLA-inbred swine recovered from experimental African swine fever virus (ASFV)infection. Ann NY Acad Sci 532: 462–464

2096 S. Gil et al.

25. Martins C, Scholl T, Mebus CA, Fish H, Lawman MJP (1988b) Modulation of porcineperipheral blood derived macrophages functions by in vitro infection with ASFV isolatesof different virulence. Viral Immunol 1: 177–190

26. Martins C, Lawman M, Scholl T, Mebus C, Lunney JK (1993) African swine fever virusspecific porcine cytotoxic T cell activity. Arch Virol 129: 211–225

27. Martins CLV, Leitao A (1994) Porcine immune responses to African swine fever virus(ASFV) infection. Vet Immunol Immunopath 43: 99–106

28. Miskin JE, Abrams CC, Goatley LC, Dixon LK (1998) A viral mechanism for inhibitionof the cellular phosphatase calcineurin. Science 281: 562–565

29. Miskin JE, Abrams CC, Dixon LK (2000) African swine fever virus protein A238Linteracts with the cellular phosphatase calcineurin via a binding domain similar to thatof NFAT. J Virol 74: 9412–9420

30. Moore KW, O’Garra A, del WaalMalefyt R, Vieira P, Mosmann TR (1993) Interleukin10. Annu Rev Immunol 11: 165–190

31. Moura Nunes JF, Vigario JD, Terrinha AM (1975) Ultraestructural study of Africanswine fever virus replication in cultures of swine bone marrow cells. Arch Virol 49:59–66

32. Neilan JG, Lu Z, Kutish GF, Zsack L, Lewis TL, Rock DL (1997) A conserved Africanswine fever virus IkB homologue, 5EL, is nonessential for growth in vitro and virulencein domestic swine. Virology 235: 377–385

33. Neter J, Kutner M, Nachtsheim C, Wasserman W (1996) In: Applied statistical linearmodels, 4th ed. Irwin

34. Pan IC, DeBoer CJ, Heuschele WP (1970) Hypergammaglobulinemia in swine infectedwith African swine fever virus. Proc Soc Exp Biol Med 134: 367–371

35. Pan IC (1987) Spontaneously susceptible cells and cell culture methodologies forAfricanswine fever virus. In: Becker Y (ed), African swine fever. Martinus Nijhoff Publishing,Boston, 9–126

36. Plowright W, Thomson GR, Neser JA (1994) African swine fever. In: Coetzer JAW,Thomson GR, Tustin RC (ed), Infectious diseases of livestock. Oxford University Press,Oxford

37. Powell PP, Dixon LK, Parkhouse RM (1996) An IκB homologue encoded by Africanswine fever virus provides a novel mechanism for downregulation of proinflamatorycytokine responses in host macrophages. J Virol 70: 8527–8533

38. RevillaY, Callejo M, Rodriguez JM, Culebras E, Nogal ML, Salas ML, Vinuela E, FresnoM (1998) Inhibition of nuclear factor kB activation by a virus-encoded IkB-like protein.J Biol Chem 273: 5405–5411

39. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning, a laboratory manual.Second Edition, Cold Spring Harbor Laboratory Press, New York

40. Scholl T, Lunney JK, Mebus CA, Duffy E, Martins CLV (1989) Virus-specific cellularblastogenesis and interleukin-2 production in swine after recovery from African swinefever. Am J Vet Res 50: 1781–1786

41. Spagnuolo-Weaver M, Fuerst R, Campbell ST, Meehan BM, McNeilly F, Adair B, AllanG (1999)A fluorimeter-base RT-PCR method for the detection and quantitation of porcinecytokines. J Immunol Methods 230: 19–27

42. Tait SWG, Reid EB, Greaves DR, Wileman TE, Powell PP (2000) Mechanismof inactivation of NF-kB by a viral homologue of IkBα (signal-induced releaseof IkBα results in binding of the viral homologue to NF-kB). J Biol Chem 275:34656–34664

43. Trinchieri G (2003) Interleukin-12 and the regulation of innate resistance and adaptativeimmunity. Nat Rev Immunol 3: 133–146

Cytokines and A238L mRNA expression on ASFV infected macrophages 2097

44. Urzainqui A, Tabares E, Carrasco L (1987) Proteins synthesised in African swine fevervirus-infected cells analysed by two-dimensional gel electrophoresis. Virology 160:286–291

45. Vigario JD, Vasco SG, Portugal FC, Festas MB (1983) Experimental studies with Africanswine fever virus carriers. In: Wilkinson P (ed), African swine fever. CEC, EUR8466EN,63–66

46. Vinuela E (1987) Molecular biology of African swine fever virus. In: Becker Y (ed),African swine fever. Martinus Nijhoff Publishing, Boston, 31–49

47. Wallach D, Varfolomeev EE, Malinin NL, GoltsevYV, Kovalenko AV, Boldin MP (1999)Tumor necrosis factor receptor and Fas signalling. Annu Rev Immunol 17: 331–367

48. Wardley RC, Hamilton F, Wilkinson PJ (1979) The replication of virulent and attenuatedstrains of African swine fever virus in porcine macrophages. ArchVirol 61: 217–225

49. Zarlenga DS, Canals A, Gasbarre LC (1995) Method for constructing internal standardsfor use in competitive PCR. BIOTECHNIQUES 19: 324–326

Author’s address: Carlos L. V. Martins, Laboratorio de Doencas Infecciosas, CIISA,Faculdade de Medicina Veterinaria, UTL, Rua Professor Cid dos Santos, 1300-477 Lisboa,Portugal; e-mail: cmartins@fmv.utl.pt