Radiol Oncol 1993; 27: 302-6.
Natural porcine interferon gamma (PoIFN gamma)
Bratko Filipic,1 Sonja Rozman,2 Katarina Carlsson3 and Avrelija Cencic4
1 Institute for Microbiology, Medical Faculty, Ljubljana, 2 BIA Ltd., Ljubljana, Slovenia, 3 Centre of Chemistry, University of Lund, Sweden, 4 College of Agriculture, University of
Maribor, Slovenia
The natural porcine mitogen induced gamma interferon (PoIFN gamma) was studied and compared with the human interferons gamma (HuIFN gamma) and alpha (HuIFN alpha). A comparison was performed by the following criteria : pH 2.0 and 56"C stability, molecular weight, dot-blot reactions and cross reactivity (neutralisation index). The biological activity of porcine ancl human interferons in vitro (antiviral and antiproliferative) was correlated on human nontransformed (HEF) and transformed cells (FL).
Key words: interferon type II-analysis; cells; cultured-drug effects
Introduction
Interferons (IFNs) are defined as proteins/gly-coproteins having an ability to protect the infected cells (causing the antiviral state) or to reduce and/or inhibit the growth of transformed cells in vitro and tumors in vivo.• 2 They are divided into at least two classes (Type I - alpha or beta and Type II - gamma) according to their mode of induction (Type I - with viruses or pI:C, Type II - with T celi mitogens or heterologous celis). 3' 4
Because of clinical use, the main attention was focused on human interferon alpha (HuIFN alpha)5 and much later on the human interferon gamma (HuIFN gamma).6 For the IFN gamma production it was thought that a pure popula-
Correspondence to: Dr. Bratko Filipic, Institute for Microbiology, Medical Faculty, 61105 Ljubljana, Slovenia
UDC: 615.281.7.015.44
tion of T lymphocytes is necessary. Latter on, it was found that partially purified spleenocytes can be used as well. The highest titers (IFN units/ml) were obtained when a mixture of peripheral blood lymphocytes and spleenocytes was used for the IFN gamma production, with specific T celi mitogens as inducers.
A comparison between HulFN alpha and HuIFN gamma shows that they are antigenically completely different: with anti-HuIFN alpha antisera it is impossible to neutralise HuIFN gamma and vice versa.
At the beginning of the clinical use of human IFNs in natural form, the main problem was how to produce enough IFN Concomitantly, the optional solutions arose: to use human-like IFNs. In this respect, many studies were carried out to find the antigenic similarity between human and human-like interferons. Surprisingly, the highest degree of homology was found between murine and human IFNs.7
Natural porcine interferon gamma
303
Similar interferon system as in humans can be found in other animal species, such as rat, horse, cattle, pig, monkey, etc. Porcine and bovine IFNs8-11 became of interest because of their relatively high antigenic similarity with human one. Such a similarity between Hu and PolFN is about 78.5 % (at nucleotide level)12 in the case of IFN alpha. When a recombinant gamma interferons (gamma IFNs) were compared (rPolFN gamma/rHulFN gamma) a homology was estimated in 59 % (at nucleotide level). 13 Correlation of non-recombinant forms of porcine alpha and human alpha IFNs shows even higher levels of homology (86.7 %% at nucleotide level) between them. 14-i5 It seems to be unusual that PoIFN gamma is cross reactive with HulFN alpha and not with PolFN gamma. Experiments presented herein are aimed to compare the natural porcine interferon gamma (PolFN gamma) with the human interferon alpha (HulFN alpha) by the following criteria: pH 2.0 and 56 °C stability, molecular weight, dot-blot analysis and neutralisation on cell culture in vitro.
Material and methods
Blood collection, buffy-coat preparation and interferon induction
Porcine blood was collected aseptically into sterile flasks containing 3.8 % sodium citrate to prevent coagulation. The porcine buffy-coat was separated by introducing 30 % sacha rose. lfi"18 The sedimented buffy-coat was washed twice with saline and resuspended in modified Eagle's medium19 containing 4 % of porcine plasma and antibiotics (Penicillin 5000U/ml, Streptomycin 5 mg/ml, Gentamycin 10 mg/ml, Neomycin l0mg/m1). The induction was performed by LCL (Lens culinaris lectin) at a concentration of 25 [xg/ml. Cell concentration was adjusted to 107-108 cells/ml. After three days of cultivation on a spinner (37°C/140 RPM) the cells were sedimented by centrifugation (2000 RPM/20 minutes) and the supernatant representing the crude IFN was harvested.
Partial purification of PolFN gamma
Interferon was isolated as follows: to 100 ml of IFN containing supernatant, 3 g of avtoclaved SiO2 (water glass) were added and incubated overnight at + 4°C. On the next day the suspension was centrifuged at 2500 RPM for 30 minutes to sediment the water glass. The sedimented water glass was resuspended in the 1/8 of the original mixture volume of 50 % ethylene glycol monoethyl ether in 1.4 mM NaCl to elute the interferon. After 2 hours the water glass was sedimented by centrifugation (2500 RPM for 30 minutes). The elution procedure was repeated in 1/16 of the original volume. Both eluates were mixed with 0.1 % of FCS (Fetal calf serum, Flow) and dialysed against distilled water.
Measurement of antiviral and antiproliferative activity
IFNs were tested for antiviral (AV) activity by 50 % cytopathogenic inhibition assay against HSVl (Herpes simplex virus type 1) as a challenge virus.20 The antiproliferative activity (AP) was determined by 50 % growth inhibition assay on HEF (human embryonal fibroblasts) and FL (human amniotic cell line) cells.21 Natural interferons alpha (Institute of Immunology, Zagreb, Croatia; EGIS Budapest, Hungary) at a concentration of 1000 AV (antiviral units/ml) were used as the control standards.
Stability tests
To determine the type of IFNs the samples were exposed to pH 2.0 and heating at 56°C for 30 minutes. Antiviral (AV) and antiprolife-rative (AP) units were determined before and after such treatment.
Protein content
In each sample, the quantity of proteins was determined by the modified Lowry method.21
304
Filipic B. et al.
Serological analyses
To determine the serological similarity and/or differences between porcine and human IFNs (gamma and alpha) the "constant method" according to LaBonnardiere et al.14 was used. In summary, constant (twofold) dilutions of anti-IFN antisera were added to the FL cells in microtiter plates (NUNC). In the next step, three-fold dilutions of IFNs with the virus (HSV l) were added. Simultaneously, IFN titration was performed without antisera. The neutralisation index (NI) was calculated as follows: NI = log3 (antiserum + interferon) - log3 (interferon).
Antigenic properties
Immunoblot tests were performed according to Pretnar et al.22 In summary, the antigens HulFN alpha (Institute for Immunology, Zagreb, Croatia; EGIS, Budapest, Hungary), HuIFN gamma (Sigma, St. Luis, USA), PoIFN gamma (Institute for Microbiology, Medical Faculty, 61000 Ljubljana, Slovenia) in the volume of 10 ¡xl (the concentration of proteins in the samples was 0.9 mg/ml) were spotted on nylon films (Hybond, Amersham) and air dried. The nylon films with the bound antigens were incubated for 60 minutes with the primary antibodies (polyclonal) as follows: Anti-HuAlpha (Boehring, Manheim, FRG; neutralisation titer 10.000U/mg), Anti-HuGamma (Boehringer, Manheim, FRG; neutralisation titer 10.000 U/ mg) and Anti-PoGamma (INRA, Youi-en-Jo-sas, France; neutralisation titer 10.000U/mg), Anti-HuAlphal (Boehring, Manheim, FRG; neutralisation titer 10.000 U/mg), Anti-HuAl-pha 2 (Institute of Virology, Bratislava, Slovakia; neutralisation titer 4000U/mg) and Anti-Acidolabile (Institute of Virology, Bratislava, Slovakia; neutralisation titer 10.000 U/mg). After washing with TTBS (10 mM Tris + HCl, pH 8.0, 150mM NaCl + 0.05 % Tween 80), the films were incubated in anti-rabbit IgG (Boe-hring, Manheim, FRG) conjugated with peroxidase. Following three washes in TTBS, the reactions were developed using DAB (diamino-
benzidine, Sigma, St.Luis, USA) in the concentration of 50 mg/100 ml TBS and 30 fl of hydrogen peroxide.
Results
Stability tests
Table 1 shows the differences between porcine (PoIFNs) and human interferons (HuIFNs, alpha, gamma) as follows: PoIFN gamma is pH 2.0 stable in contrast to HulFN gamma which is not. On the other side, PoIFN alpha is pH 2.0 labile, i.e. different from HulFN alpha which is not. Similar data were found when the resistance to heating to 56°C were compared. It seems that PolFN gamma is more similar to HuIFN alpha than to its luiman counterpart.
Table l. Stability tests for PoIFNs and HuIFNs.
IFN type	pH2.0	,56° C	M.W.1 (Daltons)
PolFN gamma	Stable	Stable	21000
PoIFN alpha	Labile	Labile	20000
HuIFN gamma	Labile	Labile	19000
HulFN alpha	Stable	Stable	19500
1 Molecular weight was determined by PAG-SDS electrop-horesis
Dot-blot analysis
Using dot-blot tests, the antigenic properties are described (Table 2) in terms of positive or negative reactions. PoIFN gamma shows a positive reaction with antisera against PoIFN gamma, HulFN alpha and HuIFN alpha 2. HuIFN gamma reacts only with the antiserum against HuIFN gamma, and not with PolFN gamma.
Table 2. Dot-blot analysis of Po and Hu IFNs.
IFN typc		Rcaction with Anti1			
	Po	Hu	Hu	Hu	Hu
	gamma	gamma	alpha	alpha l	alpha2
PoIFN gamma	+	_	+	-	+
PoIFN alpha	-	-	+	-	+
HuIFN gamma	-	+	-	-	+
HuIFN alpha	+	-	+	+	+
1 Dot-blot reaction: "+" = gives positive reaction "-" = gives negative reaction
Natural porcine interferon gamma
30.5
In the case of alpha IFNs, PoIFN reacts with the antiserum against HuIFN alpha and HulFN alpha 2. HulFN alpha gave a positive reaction with the antisera against PoIFN gamma, HulFN alpha, HuIFN alpha l and HuIFN alpha 2.
Serological tests
By serological tests in vitro (on FL cells) (Table 3) the level of the neutralisation of IFN's antiviral activity with anti-IFN antisera was quantified by the Neutralisation index (NI). When PoIFN gamma was tested, the following Nls
Table 3. Neutralisation indexes for Po and Hu IFNs.
IFN type	Neutralisation with Anti				i
	Po	Hu	Hu	Hu	Hu
	gamma	gamma	alpha	alpha 1	alpha 2
PoIFN gamma	-115	o	-1.94	o	-0.29
PoIFN alpha	o	o	-2.43	o	---0.53
HuIFN gamma	o	-0.98	o	o	o
HuIFN alpha	-0.92	o	-2.85	-2.00	-1.60
1 NI = log3 (antiserum + IFN) - log3 (IFN)
were found: With anti-PoIFN gamma - 1.15, with anti-HuIFN alpha - 1.94 and with anti-HuIFN alpha 2, -0.29. PoIFN alpha shows much higher NI for anti-HuIFN alpha (NI = -2.43) as well as for anti-Hu alpha 2 (---0.53). HuIFN alpha shows the following values for Nls: Anti-PoIFN gamma + 0.92, anti-HuIFN alpha -2.85, anti-HuIFN alpha 1, -2.00, and anti-HuIFN alpha2, -l.60
Discussion
PoIFN alpha and HuIFN alpha were found to be antigenically similar, though the differences between them were found when some of the physico-chemical characteristics were tested (temperature stability, acid-resistance).9"11 Contrarily, natural PoIFN gamma is different from its natural counterpart in humans. Its pH and temperature stability are more similar to that of HuIFN alpha. PoIFN gamma also shows cross reactivity in vitro with HuIFN alpha (complete) and its natural subtype HuIFN alpha 2 When these data were correlated with those
obtained by comparison of recombinant IFNs (human, porcine) in the case of gamma interferons, homology was established in 59 % (at nucleotide level). It seems possible that in the case of recombinant forms only selected molecules, in contrast to the natural ones (partly purified or purified) when complete molecules composed from different numbers of natural subtypes, were compared.
In this respect it should be mentioned that the comparison by NI (Neutralisation index) in vitro gives a picture of biological activity.
Future experiments with pure porcine interferons (alpha and gamma) are expected to disclose the real level of similarity/differences with human interferons (alpha, gamma), and thus enable clinical use of porcine IFNs in veterinary and human medicine.
Acknowledgment
This work was supported by the grant from the Slovenian Ministry for Research and Technology (URP Molecular Biology, Cl-509/381-93; Research field: Biochemistry with Molecular Biology, PI-5064-0381-93).
References
1.	Gresser 1, Tovey MG. Antitumour effccts of interferon. Biochem Biophys Acta 1978; 516: 231-47.
2.	Hubbell RH, Graft AJ, Leibowitz .IP, Gillespie DH. Synergistic antiproliferative effect of recombinant gamma-interferons. J Biol Resp Mod 1987; 6: 141-53.
3.	Bonnem EM, Spiegel RJ. Interferon-alpha: Current status and future promise. J Biol Resp Mod 1984; 3: 580--94.
4.	Toth M, Endresz V, Toth S, Beladi l. Human interferon alpha and beta have more potent priming activities than interferon-gamma. .T Gen Viral 1985; 66: 893-6.
5.	Borden EC. Interferons: In pursuit of the promise. In: Mirand EA,Hutchinson WB,Mihrich E. Eds. 13th International Cancer Congress, Part E., Cancer Management. New York: Alan R. Liss 1983: 287-296.
6.	Osborne LC, Georgiades JA, Johnson HM. Large scale production and partial purification of mouse immune interferon. Infect Immunol 1979; 23: 76-80.
306
Filipič B. et al.
7.	Solovicv VD, Pokidyshcva LN, Volodnina TN. On thc antigenic similarity of some alpha interferons. Vapr Virusal 1982; 27: 526-8.
8.	Richmond .FY. An Interferon-like inhibitor of foot and-mouth disease virus induced by phytohaem-maglutinine in swine leukocyte cultures. Arch Ges Virus Fars;h 1969; 27: 282-9.
9.	Soloviev VD, Ogarkov VI, Marchenko VI, Mona-styreva LN, Parfenov VV, Bukharova II, Preo-brazhenskeja NK, Zubanova NA Swine blood leukocytes, a new source of production of interferon active in human cells. Vapr Virusal 1980; 25: 716-20.
10.	Soloviev VD, Pokydisheva LN, Volodnina TN.On antigenic similarity of some alpha-interferons. Vapr Virusal 1982; 27: 526-8.
11.	Piasecki E. Properties of natural porcine interferons. J IFN Res 1988; 8: 61-73.
12.	Lefevre F, LaBonnardiere C. Molecular cloning and sequencing of a gene encoding biologically active porcine-interferon. / IFN Res 1986; 6: 34960.
13.	Charley B, McCullough K, Martinod C. Antiviral and antigenic properties of recombinant porcine interferon gamma. Vet Immunal Immunalapathal 1987; 7: 357-68.
14.	LaBonnardiere C, Laude H, Berg K. Biological and antigenic relathionship between virus-induced porcine and human interferons. Ann Inst Pasteur! Viral 1986; 137 E: 171-80.
15.	Lavrukhina LA, Gutman NR, Manakhanova LS. Antiviral effect of swine leukocyte interferon in experimental mice. Vapr Virusal 1981; 20: 414- 8.
16.	Filipič B, Golob A, Toth S, Mecs I, Bcladi I, Likar M Interactions between human and porcine interferons. Acta Viral 1991; 35: 19-26.
17.	Filipič B, Golob A, Golcc-Wondra M, Vitez Lj, Keše D, Likar M. Preparing and partially purification of swine immune interferon. In: Filipič B. cd. Yugoslav Colloquium on IFN. Ljubljana: Slovenian Microbiological Society 1986: 137-9.
18.	Filipič B, Golob A, Sinadinovska R, Babič D, Struna A, Pretnar G. Preparing and purification of porcine immune interferon. In: Likar M. ed. Posvetovanje ob 45. obletnici Instituta za Mikrobiologijo, Medicinske Fakultete v Ljubljani. Ljubljana: Institute for Microbiology, Medical Faculty: 1990: 71-80.
19.	Filipič B, Carlsson K, Zupan S, Kneževič M, Raspor P. Preparation and partial purification of porcine interferon gamma (PoIFN gamma). Acta Pharm 1993; 43: 71-4.
20.	Forti LR, Schuffman SS, Davies AH, Mitchell MW. Objective antiviral assay of the interferons by computer assisted data collection and analysis. In: Pestka S. ed. Methad in Enzimalagy, vol. 119, part C. London, New York: Academic Press 1986: 533-40.
21.	Filipič B, Carlsson K, Hartman-Pretnar K, Pod-gornik A, Koselj P. A novel protein adetermination micromethod. Acta Pharm 1992; 42: 355-60.
22.	Pretnar G, Filipič B, Golob A, Škodič A, Toth S, Mecs I, Suhar A. Electroinduction of Interferon-like proteins. Biaelectrachem Biaenerg 1991; 25: 183-92.