- Email: [email protected]
Quantification of adenovirus particles
Journal of Virological Methods
ELSEVIER
Journal of Virological Methods 50 (1994) 281-292
Quantification
of adenovirus particles
Herbert Liebermann x41e.V./Institut fir
*, Renate Mentel
Medizinische Mikrobiologie der E. -M.-Amdt-Uniuersitiit, D-I 7489 Greifswald, Germany
M. -Luther-Str. 6,
Accepted 8 June 1994
Abstract A physical method for quantification of the adenovirus was tested. It implies a centrifugation of the clarified or chloroform- (or freon-) treated crude virus suspension (0.5 ml) in a gentle sucrose gradient, following the analysis in a sensitive UV flow photometer and calculation of virus mass. The result ( pg/ml) is obtained after about 1 h. The sensitivity of detection at wavelength 254 nm, 278 nm and 226 nm was compared. Virus yield of several serotypes from monolayer cultures of FL-cells was determined in a range of < 0.1 to 7 pg/ml. The ratio of infectious to physically complete virus (about the 770s component), the influence of freezing and thawing, storing at 4°C and the effectiveness of concentration steps were also determined. There was no significant difference between the sucrose density gradient method (sedimentation rate) and the density equilibrium ultracentrifugation in a CsCl gradient. Keywords: Adenovirus; Determination of virus concentration; Sucrose density gradient centrifugation; UVflowing through-photometry
1. Introduction Detailed knowledge of physico-chemical properties of viruses and technical advances facilitated in vitro measurement of concentration of virus particles by a physical method (Fayet et al., 1971; L’le b ermann and Scherff, 1973; Barteling and Meloen, 1974; Doe1 et al., 1984; Smola et al., 1990; Liebermann et al., 1991 and others). This procedure including essentially density gradient centrifugaion followed by analysis in a UV flow calorimeter and calculation of the virus concentration by the specific extinction coefficient is used routinely for the production of footh-and-mouth disease vaccines with
* Corresponding
author.
0166-0934/94/%07.00 0 1994 Elsevier Science B.V. All rights reserved SDf 0166-0934(94)00086-V
282
H. Liebermann,
R. Mentel / Journal of Virological Methods 50 (1994) 281-292
definite antigen content (Liebermann et al., 1988; Pay and Hingley, 1992). The result of virus quantification is delivered usually after about 1 to 3 h in pg/ml units or as virus particles per ml, depending on the kind of virus and chosen conditions of centrifugation and calculation. Current investigations on diagnostic problems as well as persistence of adenoviruses the characterization of the virus samples (stock) is necessary. A number of factors have to be considered, especially the yield of virus from the cell cultures, the efficiency of virus purification, the stability of the virus particle at storage and the number of virus particles for studying the virus cell interactions. This paper reports the establishment of a method for the quantification of adeovirus (Ad) particles comprising serotypes 3, 7, 21, 2 and 5 in the FL-cell system.
2. Materials
and methods
2.1. Cells and virus FL-cells were grown in monolayer cell cultures in minimal essential medium containing 10% heat inactivated fetal calf serum. Adenoviruses were propagated in medium containing 2% heat inactivated fetal calf serum. Titration was carried out in micro titre plates using 200 pi/well (200000 cells/ml) after incubation for 7 days (CPE). 2.2. Pre-treatment
of virus
and density gradient
centrifigation
After three freeze-thaw cycles the virus suspension was treated with freon or chloroform and/or clarified by low speed centrifugation. Gradients were prepared using ultra pure sucrose (ICN Biomedicals, Meckenheim, Germany) in PBS (pH 7.4) or in TN buffer (pH 7.6) (0.02 M Tris-HCl, 0.15 M NaCl) and CsCl (for density gradient centrifugation, Merck, Darmstadt, Germany). 0.5 to 0.7 ml virus samples were overlayed on the 4.5 ml-sucrose gradient (5% to 25%) and about 0.8 to 1.2 ml aliquots on the 10 ml-gradient (5% to 30%) respectively following centrifugation at 20000 rpm for about 35 min at 20°C in swing-out rotors. The time was estimated on the basis of a sedimentation coefficient of the virus particle of about 770s. For CsCl gradient centrifugation, up to 1.5 ml of virus were overlayed on a 1.1 to 1.5 g/ml gradient and spun at 35 000 rpm for about 105 min at 18°C in the 3 X 5 ml rotor. 2.3. Measurement
of W
absorption
and calculation
The gradients were eluted after centrifugation by a Varioperpex pump (PharmaciaLKB). A l-mm tubule was put downward in the tube. The direction of the elution flow must be downward in the measuring cell of the Uvicord. For this reason the tube system at first has to be filled upwards by a solution with a slightly higher density (for example a 5% higher sucrose solution) than that at the bottom of the centrifugation tube. The optical density (Ezsdn,,, , E278nm or Ezlhn,,,) was monitored, using Uvicord SD (Pharmacia-LKB). The area below the virus peak was measured here by hand and used
H. Liebermann,
R. Mentel/Joumal
of Virological Methods 50 (1994) 281-292
283
to calculate the virus concentration as previously described (Liebermann, 1982). Briefly, the virus mass is calculated from the values of extinction Ei of the ‘fractions’ of the virus peak by the simplified formula x(mg)
= (CE, - nE,)V/F
where V= volume of a ‘fraction’; F = factor, resulting from the extinction coefficient, multiplied with the thickness of the measuring cell; IZ= the number of the ‘fractions’. We selected the width of a ‘fraction’ (e.g. 2 mm) on the recorder paper so that about 10 to 20 &-values were obtained. (The volume of such a fraction results from the paper transport/min (e.g. 30 mm/min) and the elution velocity (e.g. 0.68 ml/mitt, here V = 0.0453 ml)). 2.4. ELBA procedure The fractions of the density gradient were analysed using an indirect ELISA with HRP-labelled anti-rabbit mune sera and mock-infected cell culture preparation.
by anti-Ad3 sera from rabbits IgG. Controls included preim-
3. Results
3.1. Optical detection of the 770s unit, the virus particle The principle of the method is shown in Fig. 1. After centrifugation of a clarified crude virus suspension in a gentle sucrose gradient only one peak was present. This peak was not found in the mock-infected culture (Fig. 2). The Ad specificity was tested with anti-Ad3 antibodies by ELISA. The optical peak and the maximum in ELISA were observed in the same fractions (Fig. 3).
1
2
-‘clT?+ I
II
I
Fig. 1. Scheme of the physical determination of virus concentration. (1) Overlay of the virus specimen (e.g. clarified infectious cell culture supematant) on a sucrose gradient; (2) ultracentrifugation in a swing-out rotor; (3) analysis of the contents of the tube in a sensitive UV-flow-photometer, recording of the sedimentation profile. It follows the calculation of the virus concentration (in pg/ml) by hand or by computer.
284
H. Liebermann, R. Mentel / Journal of Virological Methods 50 (1994) 281-292
Fig. 2. Sedimentation profile of a mock-infected culture (a) and of Ad2-infected cell culture (clarified crude virus suspension after three cycles of freezing and thawing) (b). Centrifugation conditions: 0.5 ml virus was overlayed on 5% to 25% sucrose in PBS; 31 min at 20000 rpm and 20°C in a 3 X5 ml rotor of UP65M (Janet&, Leipzig, Germany). Analysis by Uvicord SD (Pharmacia/LKB, Sweden) at a wavelength of 254 nm; cv = 2.34 pg/ml; V: Viruspeak; B: bottom; M: meniscus.
Fig. 3. Sedimentation profile of Ad 3 crude virus suspension (three cycles of freezing and thawing and low speed centrifugation). Conditions of ultracentrifugation: 0.9 ml virus on a 9.5 ml sucrose gradient (5% to 30%); 62 min at 15 000 rpm and 20°C. 0- -: values of ELISA (the virus peak corresponds about 6% from the overall titre); - - -: extinction at 254 nm; Fr.: fraction number.
H. Liebermann, R. Mentel/ Journal of Virological Methods 50 (1994) 281-292
28.5
Fig. 4, UV-absorption spectrum of purified adenovirus in PBS. (The adenovirus was clarified, pelleted, freon treated and two times centrifuged in a CsCl gradient.) 2): Ad2; 7): Ad7. .Ez6,,/Ezsonm is about 1.3.
For the calculation of the virus mass it was neccessary to determine or to estimate the extinction coefficient of the virus. This value at the actual wavelength was estimated from the absorption spectrum of the virus, purified by two cycles of CsCl gradient centrifugations and by the estimation of I&en,,, (cm’/mg) on the basis of the viral dsDNA content of about 13% (Green and Pina, 1963). Fig. 4 displays the absorption spectrum of high purified adenovirus. The extinction coefficients were obtaned from spectra of Ad7, Ad2 and Ad3 samples and resulted in Ez6e = (3.30 & 0.35) cm2/mg, E 254nm= (3.14 f 0.40) cm’/mg, E278nm= (2.79 -t 0.56) cm’/mg, EzzGnm= (8.6 12.2) cm’/mg. 3.2. rn~~e~ce of the wavelength on se~s~tiui~ and li~ea~i~ of ~o~cen~ati5n
ualues at
uirus dilutions
The virus peak at 226 nm is about 3-times higher as at 254 nm in accordance with the absorption spectrum of the virus (Ad2) (Fig. 5). But the baseline at 226 nm is more uncertain in the rule. 83 ng of Ad3 could be detected at this wavelength. Fig. 6 demonstrates the linearity of ~ncen~ation values in the given range. 3.3. Virus yield and ratio of infectious to physical virus particles Table 1 demonstrates the virus yields at different harvests between 0.1 and 7 pg/ml at our conditions of virus multiplications in FL, cells, and they are lower in human lymphocytes. The standard deviation for the quantifi~tion is an_ r = 0.12 pg/ml (= 7.6%) at a mean value of x = 1.58 ,ug/ml; (n = 6). on_ t = 0.20 yg/ml, if different persons perform the determinations. In the first experiments the ratio of infectious to physical virus particles is between 1: 126 and 1: 5100 (Table 2).
286
H. Liebermann,
R. Mentel /Journal
of Virological Methods 50 (1994) 281-292
Fig. 5. Registration of the sedimentation profile of a crude Ad3 suspension by UV-absorption measurement at three different wavelengths. Conditions: 5% to 30% sucrose in PBS (pH 7.2); 62 min at 15 000 rpm and 20°C in a 6 X 11 ml swing-out rotor (TH 641, Sorvall-Du Pant); a: 278 nm; b: 254 nm; c: 226 nm (the virus overlay d: 226 nm (the virus was 0.9 ml each, and the calculated virus concentration cv is about 0.92 pg/ml); overlay was 0.09 ml corresponding about 83 ng). B: bottom; M: meniscus.
3.4. Virus production by one cell
Under the simplified assumption that each cell is infected and the applied method of virus mass determination applies to all viruses, one cell in a stationary FL cell culture produces about 40000 virus particles on average. The volume of these 40000 virus particles (about 8.7 krn3) is about 0.3% of the volume of a FL cell (about 3400 pm3).
&- cglculated virus cont.
real 1
05 Fig. 6. Dependence definite dilutions).
of the determined
virus concentration
rel. virus cont.
r
1:o
from the real concentration
(preconcentrated
Ad5 in
H. Liebennann, R. Mentel/Journal
of Virological Methods 50 (1994) 281-292
287
Table 1 Virus yields from cell cultures No.
Virus concentration
Remarks
( pLg/mt) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Ad3 Ad 3 Ad3 Ad3 Ad 21 Ad 21 Ad 7 Ad7 Ad 7 Ad7 Ad2 Ad5 Ad 2 Ad2 Ad2
2.42 4.33 3.79 20.1 6.11 0.12 5.78 0.85 6.12 4.32 6.54 2.85 < 0.1
from FL stat. culture yield 11.8.92 165 d later at 4-7°C cl. sed. 13.5% of the virus 4th pass 5th pass 4th pass 5th pass yield 5.11.92 81 d later at 4-7°C
from human lymphocytes from human lymphocytes from human lymphocytes
3.5. Storage of virus The virus particle in the crude virus suspension stored at 4°C was stable for months. The virus concentration was only decreased by 20% after 165 days (Table 1). In contrast, the 770s component in partially or highly purified virus specimens already disappeared after 9 weeks. The influence of freeze-thaw cycles was also tested for crude virus suspensions (Fig. 7). Frequent freezing and thawing can diminish the virus content markedly, particularly if the virus is clarified ppreviously. 3.6. Effect of virus enrichment
and solvent treatment
The yield of virus mass after ultracentrifugation (65 min at 18000 rpm at 6°C in a 3 X 30 ml swing-out rotor) was determined for an Ad3 virus sample to 26%, for Ad7 to 33% and for Ad21 to 33% and 70%. (The last result was obtained after resuspension of the pellet in PBS with 0.5 M instead of 0.15 M NaCl.) In comparison, the yield of virus
Table 2 Number of infectious/number Virus
Ad 3 from FL-cells Ad 3 from FL-cells Ad 2 from FL-cells
of physical titre log CID,, /50
particles Virus cont. ~1
(a)
kg/ml (b)
6.5 5.0 5.5
2.35 3.01 6.54
Particle 50 /.Ll
Relation (a)/(c)
(c) 3.99 x 10s 5.12 X 10’ 11.1 x108
1: 126 1:5100 1:3500
288
H. Liebermann, R. Mentel/
Journal
of Virological Methods 50 (1994) 281-292
Fig. 7. Influence of freezing and thawing on the virus mass. Ad2 and Ad5 virus samples were unclarified crude virus suspensions; ordinate: virus (about 770s unit) concentration in pg/ml; abscissa: the number of cycles of freezing and thawing.
mass was about 69 + 12% for Ad5 after precipitation by PEG 6000 and resuspension in l/10 vol. TEN buffer (pH 7.6) (0.025 M Tris-HCl, 0.15 M NaCl, 1 mM EDTA). The virus concentration was not changed after treatment of crude virus samples with chloroform or freon (trichloro-trifluoro-ethane) for partial purification (for example, 1.13 pg/ml without and 1.07 pg/ml after chloroform treatment of Ad21, and 0.95 pg/ml without and 1.08 pg/ml after treatment of Ad5 respectively). 3.7. Comparative centrifugation
quantification
of the adenovirus
by sucrose and CsCl density gradient
The separation of the components in a virus specimen on the basis of their density can be used for the control of the accuracy of virus quantification by the used sucrose gradient centrifugation. The values of virus mass obtained were similar in the first results (Fig. 8; Table 3). (The slightly higher values after CsCl gradient centrifugation are not significant.)
4. Discussion Up to now the concentration of physical virus particles was determined by UV spectrum measurement after several purification steps of virus samples from more than 10 to 100 ml in studies of the Ad-host cell interactions (Defer et al., 1990). Here the quantification of the about 770s unit of Ad in a crude virus suspension (from cell cultures) is simple and fast by sucrose density gradient centrifugation and UV-photometry. The result is obtained in pg/ml or as a number of virus particles per ml afer about 1 h, using for example 0.5 ml of clarified or chloroform-treated virus (two- or threefold determination). It is possible to detect about 80 ng of virus at 226 nm. The gradient must
H. Liebermann, R. Mentel/Journal
of Virological Methods 50 (1994) 281-292
289
-006
Fig. 8. Absorption profile after CsCl gradient centrifugation. Conditions: gradient 1.1 to 1.5 g/ml; overlay 0.8 ml; 34000 ‘pm at 20°C for 1.75 h in a 3X5 ml swing-out rotor; V: virus peak; calculated concentration: (1.74 * 0.34) kg/ml.
Ad7 virus
be optically pure and the pump must have low pulsation, such as ‘Perimax 12’ (Knauer, Berlin, Germany) or ‘Minipuls 3’ (Abimed/Gilson, Langenfeld, Germany). The reasons for possibly higher values after CsCl gradient centrifugation could be an additional light scattering effect of the virus band in the CsCl by partial precipitation and/or the measurement of free and aggregated virus together. Aggregates of viruses and free virus in the same peak can be detected in all density equilibrium ultracentrifugations. This was done for the quantification of foot-and-mouth disease virus by partially comparing with the sedimentation rate method in sucrose gradient (Liebermann, 1978; Strobbe, 1983). Lei (1978) found with the latter procedure about 80% of the CsCl-method. Disadvantages of the CsCl gradient centrifugation are the greater time required, the higher costs and a possible partial precipitation of the virus. The separation of the virus from the ‘contaminants’ is usually better than by sucrose gradient centrifugation but it can be sometimes not so good as by the latter (Liebermann et al., 1992). An important quality criteria for high performance measurements, besides precision, is accuracy. The latter only can be guaranteed by use of an appropriate standard (Bickel and Mayer,
Table 3 Comparative centrifugation
quantification
of Ad3
and Ad7 virus by sucrose
(S DGC)
and by CsCl density
(CSCI-DGC)
No. Virus
S-DGC
CsCl-DGC
1
2
3 ( fig/ml)
4 ( pg/ml)
Relation 3:4 (%)
1 2 3 4
Ad3 Ad3 Ad7 Ad7
3.06 2.37 1.18 1.27
4.08 2.97 1.42 1.74
75 80 83 73
gradient
290
H. Liebermann,
R.
Mentel/ Journal of Virological Methods 50 (1994) 281-292
1993). An internal standard could be a crude virus suspension with a predetermined virus mass, rapidly frozen and stored at - 80°C. The method described for virus quantification is faster and is likely to be more accurate than the adenovirus particle quantification by electron microscopy. Green and Pina (1963) reported that one A,,, unit of the virus is regarded as 0.28 mg of virus protein or 1.1 X 1012 particles. That means E,,, = 3.57 cm2/mg. This is in a good agreement with our estimation. The results of virus concentration by ultracentrifugation illustrate that the quality of resuspension may be essential. A cushion of RbCl solution (Green and Pina, 1963) of CsCl, KBr or sucrose should be used. Very different values can be obtained for the ratio of infectious to physical adenovirus particles in dependence of the titration method, the cell strain, the titration moment and the step of purification etc. Green et al. (1967) obtained especially for Ad3 a virion/PFU ratio of 11: 1 for a clonal line of KB cells. Reported ratios range from 11: 1 to 2300 : 1 for the different adenovirus types. They discussed also possible differences in the stability of adenoviruses which varies considerably from type to type. Disintegration of virions by freezing and thawing (of purified virus) has been demonstrated (Prage and Pettersson, 1971). Our first estimated value of the number of adenovirus particles per FL-cell seems to be high. However, Myerowitz et al. (1975) reported that it may be as many as 10000 particles per cell. Green and Pina (1963) obtained yields of 5 to 15 mg of purified virus from 6 to 9 X lo8 KB cells, i.e., about 43 500 virus particles were produced per cell. The losses during purification are not considered.
Acknowledgements We wish to thank Prof. Dr. W. Seidel grateful to Miss Dr. U. Wegner for Liebermann for careful assistance and lymphocytes. Thanks are also given to investigation.
for critical reading of the manuscript and are help in the propagation of virus, Mrs. M. Miss Bauer for virus samples from human Prof. Dr. L. DGhner for the support of this
References Barteling, S. and Meloen, R.H. (1974) A simple method for the quantification of 140s particles of foot-and-mouth disease virus (FMDV)). Arch. ges. Virusforsch 45, 362-364. Bickel, M. and Mayer, K. (1993) 1st Genauigkeit genug? - Wozu brauchen wir Standardreferenzmaterialien? GIT Fachz. Lab. 34-35. Defer, Ch., Belin, M.T., Caillet-Boudin, M.L. and Boulanger, P. (1990) Human adenovims-host cell interactions: comparative study with members of subgroups B and C. J. Virol. 64, 3661-3673. Doel, T.R., Osterhaus, A.D.M.E., van Wezel, A.L., van Steenis, G. (1984) The evaluation of a physical method for the quantification of inactivated poliovirus particles and its relationship to D-antigenicity and potency testing in rats. J. Biol. Standard. 12, 93-99. Fayet, M.T., Farglaud, D., Louibot, P., Stellmann, C. and Roumianzeff, M. (1971) Mesure physico-chimique des particules 140s du virus de la fievreaphteuse. Ann. Inst. Pasteur. 121, 107-l 18.
H. Liebermann, R. Mentel/Journal
of Virological Methods 50 (1994) 281-292
291
Green, M. and Pina, M. (1963) Biochemical studies on adenovirus multiplication. IV. Isolation, purification and chemical analysis of adenovirus. Virology 20, 199-207. Green, M., Pina, M. and Kimes, R.C. (1967) Biochemical studies on adenovirus multiplication. XII. Plaquing efficiencies of purified human adenoviruses. Virology 31, 562-565. Lei, J.C. (1978) Comparative Investigation of virus content in samples of FMDV from BHK21-suspension cells by sedimentation rate and density equilibrium ultracentrifirgation. FAO ECC FMD, Uccle, Belgium. Liebermann, HT. (1978) Die Ultrazentrifugation in der Vimsdiagnostik. Arch. exp. Vet. med. 32, 427-434. Liebermann, HT. (1982) Reinigung und Konzentrierung animaler Viren. Fischer, Jena, p. 19 and pp. 87-94. Liebermann, HT. and Scherff, J.(1973) Verwendung des Universalspektralphotometers als Durchfhmphotometer zum empfindlichen Registrieren von Virusproteinen. Arch. exp. Vet. med. 27, 855-859. Liebermann, HT., Thalmann, G., Nockler, A. and Felfe, P. (1988) Korrelation zwischen 146SAntigendosis und Antikiirpertiter bei gegen MKS vakzinierten Rindern und Schweinen. Arch. exp. Vet. med. 42, 528-536. Liebermann, HT., Bergmann, H., Lange, E. and Schirrmeier, H. (1991) Hamorrhagische Septiklmie der Kaninchen (Rabbit Heamorrhagic Disease, RHD) - Untersuchungen zum quantitativen Antigennachweis. J. Vet. Med. B 38, 621-629. Myerowitz, R., Stadler, H. and Oxman, M. (1975) Fatal disseminated adenovirus infection in a renal transplant recipient. Am. J. Med. 59, 591-598. Pay, T.W.F. and Hingley, P.J. (1992) Foot and mouth disease vaccine potency tests in cattle: the interrelationship of antigen dose, serum neutralizing antibody response and protection from challenge. Vaccine 10, 699-706. Prage, L. and Pettersson, U. (1971) Structural proteins of adenoviruses. VII. Purification and properties of arginine-rich core protein from adenovirus type 2 and type 3. Virology 45, 364-373. Smola, T., Specht, G. and Liebermann, HT. (1990) Bestimmung der Viruskonzentration unter Verwendung eines Kleincomputers. Arch. exp. Vet. med. 44, 311-317. Strobbe, R.(1983) Use of short analytical ultracentrifugation runs for the isopycnic determination of FMDV concentration and density in autoformed CsCl gradients. Ann. Rech. Vet. 14, 233-237.
ELSEVIER
Journal of Virological Methods 50 (1994) 281-292
Quantification
of adenovirus particles
Herbert Liebermann x41e.V./Institut fir
*, Renate Mentel
Medizinische Mikrobiologie der E. -M.-Amdt-Uniuersitiit, D-I 7489 Greifswald, Germany
M. -Luther-Str. 6,
Accepted 8 June 1994
Abstract A physical method for quantification of the adenovirus was tested. It implies a centrifugation of the clarified or chloroform- (or freon-) treated crude virus suspension (0.5 ml) in a gentle sucrose gradient, following the analysis in a sensitive UV flow photometer and calculation of virus mass. The result ( pg/ml) is obtained after about 1 h. The sensitivity of detection at wavelength 254 nm, 278 nm and 226 nm was compared. Virus yield of several serotypes from monolayer cultures of FL-cells was determined in a range of < 0.1 to 7 pg/ml. The ratio of infectious to physically complete virus (about the 770s component), the influence of freezing and thawing, storing at 4°C and the effectiveness of concentration steps were also determined. There was no significant difference between the sucrose density gradient method (sedimentation rate) and the density equilibrium ultracentrifugation in a CsCl gradient. Keywords: Adenovirus; Determination of virus concentration; Sucrose density gradient centrifugation; UVflowing through-photometry
1. Introduction Detailed knowledge of physico-chemical properties of viruses and technical advances facilitated in vitro measurement of concentration of virus particles by a physical method (Fayet et al., 1971; L’le b ermann and Scherff, 1973; Barteling and Meloen, 1974; Doe1 et al., 1984; Smola et al., 1990; Liebermann et al., 1991 and others). This procedure including essentially density gradient centrifugaion followed by analysis in a UV flow calorimeter and calculation of the virus concentration by the specific extinction coefficient is used routinely for the production of footh-and-mouth disease vaccines with
* Corresponding
author.
0166-0934/94/%07.00 0 1994 Elsevier Science B.V. All rights reserved SDf 0166-0934(94)00086-V
282
H. Liebermann,
R. Mentel / Journal of Virological Methods 50 (1994) 281-292
definite antigen content (Liebermann et al., 1988; Pay and Hingley, 1992). The result of virus quantification is delivered usually after about 1 to 3 h in pg/ml units or as virus particles per ml, depending on the kind of virus and chosen conditions of centrifugation and calculation. Current investigations on diagnostic problems as well as persistence of adenoviruses the characterization of the virus samples (stock) is necessary. A number of factors have to be considered, especially the yield of virus from the cell cultures, the efficiency of virus purification, the stability of the virus particle at storage and the number of virus particles for studying the virus cell interactions. This paper reports the establishment of a method for the quantification of adeovirus (Ad) particles comprising serotypes 3, 7, 21, 2 and 5 in the FL-cell system.
2. Materials
and methods
2.1. Cells and virus FL-cells were grown in monolayer cell cultures in minimal essential medium containing 10% heat inactivated fetal calf serum. Adenoviruses were propagated in medium containing 2% heat inactivated fetal calf serum. Titration was carried out in micro titre plates using 200 pi/well (200000 cells/ml) after incubation for 7 days (CPE). 2.2. Pre-treatment
of virus
and density gradient
centrifigation
After three freeze-thaw cycles the virus suspension was treated with freon or chloroform and/or clarified by low speed centrifugation. Gradients were prepared using ultra pure sucrose (ICN Biomedicals, Meckenheim, Germany) in PBS (pH 7.4) or in TN buffer (pH 7.6) (0.02 M Tris-HCl, 0.15 M NaCl) and CsCl (for density gradient centrifugation, Merck, Darmstadt, Germany). 0.5 to 0.7 ml virus samples were overlayed on the 4.5 ml-sucrose gradient (5% to 25%) and about 0.8 to 1.2 ml aliquots on the 10 ml-gradient (5% to 30%) respectively following centrifugation at 20000 rpm for about 35 min at 20°C in swing-out rotors. The time was estimated on the basis of a sedimentation coefficient of the virus particle of about 770s. For CsCl gradient centrifugation, up to 1.5 ml of virus were overlayed on a 1.1 to 1.5 g/ml gradient and spun at 35 000 rpm for about 105 min at 18°C in the 3 X 5 ml rotor. 2.3. Measurement
of W
absorption
and calculation
The gradients were eluted after centrifugation by a Varioperpex pump (PharmaciaLKB). A l-mm tubule was put downward in the tube. The direction of the elution flow must be downward in the measuring cell of the Uvicord. For this reason the tube system at first has to be filled upwards by a solution with a slightly higher density (for example a 5% higher sucrose solution) than that at the bottom of the centrifugation tube. The optical density (Ezsdn,,, , E278nm or Ezlhn,,,) was monitored, using Uvicord SD (Pharmacia-LKB). The area below the virus peak was measured here by hand and used
H. Liebermann,
R. Mentel/Joumal
of Virological Methods 50 (1994) 281-292
283
to calculate the virus concentration as previously described (Liebermann, 1982). Briefly, the virus mass is calculated from the values of extinction Ei of the ‘fractions’ of the virus peak by the simplified formula x(mg)
= (CE, - nE,)V/F
where V= volume of a ‘fraction’; F = factor, resulting from the extinction coefficient, multiplied with the thickness of the measuring cell; IZ= the number of the ‘fractions’. We selected the width of a ‘fraction’ (e.g. 2 mm) on the recorder paper so that about 10 to 20 &-values were obtained. (The volume of such a fraction results from the paper transport/min (e.g. 30 mm/min) and the elution velocity (e.g. 0.68 ml/mitt, here V = 0.0453 ml)). 2.4. ELBA procedure The fractions of the density gradient were analysed using an indirect ELISA with HRP-labelled anti-rabbit mune sera and mock-infected cell culture preparation.
by anti-Ad3 sera from rabbits IgG. Controls included preim-
3. Results
3.1. Optical detection of the 770s unit, the virus particle The principle of the method is shown in Fig. 1. After centrifugation of a clarified crude virus suspension in a gentle sucrose gradient only one peak was present. This peak was not found in the mock-infected culture (Fig. 2). The Ad specificity was tested with anti-Ad3 antibodies by ELISA. The optical peak and the maximum in ELISA were observed in the same fractions (Fig. 3).
1
2
-‘clT?+ I
II
I
Fig. 1. Scheme of the physical determination of virus concentration. (1) Overlay of the virus specimen (e.g. clarified infectious cell culture supematant) on a sucrose gradient; (2) ultracentrifugation in a swing-out rotor; (3) analysis of the contents of the tube in a sensitive UV-flow-photometer, recording of the sedimentation profile. It follows the calculation of the virus concentration (in pg/ml) by hand or by computer.
284
H. Liebermann, R. Mentel / Journal of Virological Methods 50 (1994) 281-292
Fig. 2. Sedimentation profile of a mock-infected culture (a) and of Ad2-infected cell culture (clarified crude virus suspension after three cycles of freezing and thawing) (b). Centrifugation conditions: 0.5 ml virus was overlayed on 5% to 25% sucrose in PBS; 31 min at 20000 rpm and 20°C in a 3 X5 ml rotor of UP65M (Janet&, Leipzig, Germany). Analysis by Uvicord SD (Pharmacia/LKB, Sweden) at a wavelength of 254 nm; cv = 2.34 pg/ml; V: Viruspeak; B: bottom; M: meniscus.
Fig. 3. Sedimentation profile of Ad 3 crude virus suspension (three cycles of freezing and thawing and low speed centrifugation). Conditions of ultracentrifugation: 0.9 ml virus on a 9.5 ml sucrose gradient (5% to 30%); 62 min at 15 000 rpm and 20°C. 0- -: values of ELISA (the virus peak corresponds about 6% from the overall titre); - - -: extinction at 254 nm; Fr.: fraction number.
H. Liebermann, R. Mentel/ Journal of Virological Methods 50 (1994) 281-292
28.5
Fig. 4, UV-absorption spectrum of purified adenovirus in PBS. (The adenovirus was clarified, pelleted, freon treated and two times centrifuged in a CsCl gradient.) 2): Ad2; 7): Ad7. .Ez6,,/Ezsonm is about 1.3.
For the calculation of the virus mass it was neccessary to determine or to estimate the extinction coefficient of the virus. This value at the actual wavelength was estimated from the absorption spectrum of the virus, purified by two cycles of CsCl gradient centrifugations and by the estimation of I&en,,, (cm’/mg) on the basis of the viral dsDNA content of about 13% (Green and Pina, 1963). Fig. 4 displays the absorption spectrum of high purified adenovirus. The extinction coefficients were obtaned from spectra of Ad7, Ad2 and Ad3 samples and resulted in Ez6e = (3.30 & 0.35) cm2/mg, E 254nm= (3.14 f 0.40) cm’/mg, E278nm= (2.79 -t 0.56) cm’/mg, EzzGnm= (8.6 12.2) cm’/mg. 3.2. rn~~e~ce of the wavelength on se~s~tiui~ and li~ea~i~ of ~o~cen~ati5n
ualues at
uirus dilutions
The virus peak at 226 nm is about 3-times higher as at 254 nm in accordance with the absorption spectrum of the virus (Ad2) (Fig. 5). But the baseline at 226 nm is more uncertain in the rule. 83 ng of Ad3 could be detected at this wavelength. Fig. 6 demonstrates the linearity of ~ncen~ation values in the given range. 3.3. Virus yield and ratio of infectious to physical virus particles Table 1 demonstrates the virus yields at different harvests between 0.1 and 7 pg/ml at our conditions of virus multiplications in FL, cells, and they are lower in human lymphocytes. The standard deviation for the quantifi~tion is an_ r = 0.12 pg/ml (= 7.6%) at a mean value of x = 1.58 ,ug/ml; (n = 6). on_ t = 0.20 yg/ml, if different persons perform the determinations. In the first experiments the ratio of infectious to physical virus particles is between 1: 126 and 1: 5100 (Table 2).
286
H. Liebermann,
R. Mentel /Journal
of Virological Methods 50 (1994) 281-292
Fig. 5. Registration of the sedimentation profile of a crude Ad3 suspension by UV-absorption measurement at three different wavelengths. Conditions: 5% to 30% sucrose in PBS (pH 7.2); 62 min at 15 000 rpm and 20°C in a 6 X 11 ml swing-out rotor (TH 641, Sorvall-Du Pant); a: 278 nm; b: 254 nm; c: 226 nm (the virus overlay d: 226 nm (the virus was 0.9 ml each, and the calculated virus concentration cv is about 0.92 pg/ml); overlay was 0.09 ml corresponding about 83 ng). B: bottom; M: meniscus.
3.4. Virus production by one cell
Under the simplified assumption that each cell is infected and the applied method of virus mass determination applies to all viruses, one cell in a stationary FL cell culture produces about 40000 virus particles on average. The volume of these 40000 virus particles (about 8.7 krn3) is about 0.3% of the volume of a FL cell (about 3400 pm3).
&- cglculated virus cont.
real 1
05 Fig. 6. Dependence definite dilutions).
of the determined
virus concentration
rel. virus cont.
r
1:o
from the real concentration
(preconcentrated
Ad5 in
H. Liebennann, R. Mentel/Journal
of Virological Methods 50 (1994) 281-292
287
Table 1 Virus yields from cell cultures No.
Virus concentration
Remarks
( pLg/mt) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Ad3 Ad 3 Ad3 Ad3 Ad 21 Ad 21 Ad 7 Ad7 Ad 7 Ad7 Ad2 Ad5 Ad 2 Ad2 Ad2
2.42 4.33 3.79 20.1 6.11 0.12 5.78 0.85 6.12 4.32 6.54 2.85 < 0.1
from FL stat. culture yield 11.8.92 165 d later at 4-7°C cl. sed. 13.5% of the virus 4th pass 5th pass 4th pass 5th pass yield 5.11.92 81 d later at 4-7°C
from human lymphocytes from human lymphocytes from human lymphocytes
3.5. Storage of virus The virus particle in the crude virus suspension stored at 4°C was stable for months. The virus concentration was only decreased by 20% after 165 days (Table 1). In contrast, the 770s component in partially or highly purified virus specimens already disappeared after 9 weeks. The influence of freeze-thaw cycles was also tested for crude virus suspensions (Fig. 7). Frequent freezing and thawing can diminish the virus content markedly, particularly if the virus is clarified ppreviously. 3.6. Effect of virus enrichment
and solvent treatment
The yield of virus mass after ultracentrifugation (65 min at 18000 rpm at 6°C in a 3 X 30 ml swing-out rotor) was determined for an Ad3 virus sample to 26%, for Ad7 to 33% and for Ad21 to 33% and 70%. (The last result was obtained after resuspension of the pellet in PBS with 0.5 M instead of 0.15 M NaCl.) In comparison, the yield of virus
Table 2 Number of infectious/number Virus
Ad 3 from FL-cells Ad 3 from FL-cells Ad 2 from FL-cells
of physical titre log CID,, /50
particles Virus cont. ~1
(a)
kg/ml (b)
6.5 5.0 5.5
2.35 3.01 6.54
Particle 50 /.Ll
Relation (a)/(c)
(c) 3.99 x 10s 5.12 X 10’ 11.1 x108
1: 126 1:5100 1:3500
288
H. Liebermann, R. Mentel/
Journal
of Virological Methods 50 (1994) 281-292
Fig. 7. Influence of freezing and thawing on the virus mass. Ad2 and Ad5 virus samples were unclarified crude virus suspensions; ordinate: virus (about 770s unit) concentration in pg/ml; abscissa: the number of cycles of freezing and thawing.
mass was about 69 + 12% for Ad5 after precipitation by PEG 6000 and resuspension in l/10 vol. TEN buffer (pH 7.6) (0.025 M Tris-HCl, 0.15 M NaCl, 1 mM EDTA). The virus concentration was not changed after treatment of crude virus samples with chloroform or freon (trichloro-trifluoro-ethane) for partial purification (for example, 1.13 pg/ml without and 1.07 pg/ml after chloroform treatment of Ad21, and 0.95 pg/ml without and 1.08 pg/ml after treatment of Ad5 respectively). 3.7. Comparative centrifugation
quantification
of the adenovirus
by sucrose and CsCl density gradient
The separation of the components in a virus specimen on the basis of their density can be used for the control of the accuracy of virus quantification by the used sucrose gradient centrifugation. The values of virus mass obtained were similar in the first results (Fig. 8; Table 3). (The slightly higher values after CsCl gradient centrifugation are not significant.)
4. Discussion Up to now the concentration of physical virus particles was determined by UV spectrum measurement after several purification steps of virus samples from more than 10 to 100 ml in studies of the Ad-host cell interactions (Defer et al., 1990). Here the quantification of the about 770s unit of Ad in a crude virus suspension (from cell cultures) is simple and fast by sucrose density gradient centrifugation and UV-photometry. The result is obtained in pg/ml or as a number of virus particles per ml afer about 1 h, using for example 0.5 ml of clarified or chloroform-treated virus (two- or threefold determination). It is possible to detect about 80 ng of virus at 226 nm. The gradient must
H. Liebermann, R. Mentel/Journal
of Virological Methods 50 (1994) 281-292
289
-006
Fig. 8. Absorption profile after CsCl gradient centrifugation. Conditions: gradient 1.1 to 1.5 g/ml; overlay 0.8 ml; 34000 ‘pm at 20°C for 1.75 h in a 3X5 ml swing-out rotor; V: virus peak; calculated concentration: (1.74 * 0.34) kg/ml.
Ad7 virus
be optically pure and the pump must have low pulsation, such as ‘Perimax 12’ (Knauer, Berlin, Germany) or ‘Minipuls 3’ (Abimed/Gilson, Langenfeld, Germany). The reasons for possibly higher values after CsCl gradient centrifugation could be an additional light scattering effect of the virus band in the CsCl by partial precipitation and/or the measurement of free and aggregated virus together. Aggregates of viruses and free virus in the same peak can be detected in all density equilibrium ultracentrifugations. This was done for the quantification of foot-and-mouth disease virus by partially comparing with the sedimentation rate method in sucrose gradient (Liebermann, 1978; Strobbe, 1983). Lei (1978) found with the latter procedure about 80% of the CsCl-method. Disadvantages of the CsCl gradient centrifugation are the greater time required, the higher costs and a possible partial precipitation of the virus. The separation of the virus from the ‘contaminants’ is usually better than by sucrose gradient centrifugation but it can be sometimes not so good as by the latter (Liebermann et al., 1992). An important quality criteria for high performance measurements, besides precision, is accuracy. The latter only can be guaranteed by use of an appropriate standard (Bickel and Mayer,
Table 3 Comparative centrifugation
quantification
of Ad3
and Ad7 virus by sucrose
(S DGC)
and by CsCl density
(CSCI-DGC)
No. Virus
S-DGC
CsCl-DGC
1
2
3 ( fig/ml)
4 ( pg/ml)
Relation 3:4 (%)
1 2 3 4
Ad3 Ad3 Ad7 Ad7
3.06 2.37 1.18 1.27
4.08 2.97 1.42 1.74
75 80 83 73
gradient
290
H. Liebermann,
R.
Mentel/ Journal of Virological Methods 50 (1994) 281-292
1993). An internal standard could be a crude virus suspension with a predetermined virus mass, rapidly frozen and stored at - 80°C. The method described for virus quantification is faster and is likely to be more accurate than the adenovirus particle quantification by electron microscopy. Green and Pina (1963) reported that one A,,, unit of the virus is regarded as 0.28 mg of virus protein or 1.1 X 1012 particles. That means E,,, = 3.57 cm2/mg. This is in a good agreement with our estimation. The results of virus concentration by ultracentrifugation illustrate that the quality of resuspension may be essential. A cushion of RbCl solution (Green and Pina, 1963) of CsCl, KBr or sucrose should be used. Very different values can be obtained for the ratio of infectious to physical adenovirus particles in dependence of the titration method, the cell strain, the titration moment and the step of purification etc. Green et al. (1967) obtained especially for Ad3 a virion/PFU ratio of 11: 1 for a clonal line of KB cells. Reported ratios range from 11: 1 to 2300 : 1 for the different adenovirus types. They discussed also possible differences in the stability of adenoviruses which varies considerably from type to type. Disintegration of virions by freezing and thawing (of purified virus) has been demonstrated (Prage and Pettersson, 1971). Our first estimated value of the number of adenovirus particles per FL-cell seems to be high. However, Myerowitz et al. (1975) reported that it may be as many as 10000 particles per cell. Green and Pina (1963) obtained yields of 5 to 15 mg of purified virus from 6 to 9 X lo8 KB cells, i.e., about 43 500 virus particles were produced per cell. The losses during purification are not considered.
Acknowledgements We wish to thank Prof. Dr. W. Seidel grateful to Miss Dr. U. Wegner for Liebermann for careful assistance and lymphocytes. Thanks are also given to investigation.
for critical reading of the manuscript and are help in the propagation of virus, Mrs. M. Miss Bauer for virus samples from human Prof. Dr. L. DGhner for the support of this
References Barteling, S. and Meloen, R.H. (1974) A simple method for the quantification of 140s particles of foot-and-mouth disease virus (FMDV)). Arch. ges. Virusforsch 45, 362-364. Bickel, M. and Mayer, K. (1993) 1st Genauigkeit genug? - Wozu brauchen wir Standardreferenzmaterialien? GIT Fachz. Lab. 34-35. Defer, Ch., Belin, M.T., Caillet-Boudin, M.L. and Boulanger, P. (1990) Human adenovims-host cell interactions: comparative study with members of subgroups B and C. J. Virol. 64, 3661-3673. Doel, T.R., Osterhaus, A.D.M.E., van Wezel, A.L., van Steenis, G. (1984) The evaluation of a physical method for the quantification of inactivated poliovirus particles and its relationship to D-antigenicity and potency testing in rats. J. Biol. Standard. 12, 93-99. Fayet, M.T., Farglaud, D., Louibot, P., Stellmann, C. and Roumianzeff, M. (1971) Mesure physico-chimique des particules 140s du virus de la fievreaphteuse. Ann. Inst. Pasteur. 121, 107-l 18.
H. Liebermann, R. Mentel/Journal
of Virological Methods 50 (1994) 281-292
291
Green, M. and Pina, M. (1963) Biochemical studies on adenovirus multiplication. IV. Isolation, purification and chemical analysis of adenovirus. Virology 20, 199-207. Green, M., Pina, M. and Kimes, R.C. (1967) Biochemical studies on adenovirus multiplication. XII. Plaquing efficiencies of purified human adenoviruses. Virology 31, 562-565. Lei, J.C. (1978) Comparative Investigation of virus content in samples of FMDV from BHK21-suspension cells by sedimentation rate and density equilibrium ultracentrifirgation. FAO ECC FMD, Uccle, Belgium. Liebermann, HT. (1978) Die Ultrazentrifugation in der Vimsdiagnostik. Arch. exp. Vet. med. 32, 427-434. Liebermann, HT. (1982) Reinigung und Konzentrierung animaler Viren. Fischer, Jena, p. 19 and pp. 87-94. Liebermann, HT. and Scherff, J.(1973) Verwendung des Universalspektralphotometers als Durchfhmphotometer zum empfindlichen Registrieren von Virusproteinen. Arch. exp. Vet. med. 27, 855-859. Liebermann, HT., Thalmann, G., Nockler, A. and Felfe, P. (1988) Korrelation zwischen 146SAntigendosis und Antikiirpertiter bei gegen MKS vakzinierten Rindern und Schweinen. Arch. exp. Vet. med. 42, 528-536. Liebermann, HT., Bergmann, H., Lange, E. and Schirrmeier, H. (1991) Hamorrhagische Septiklmie der Kaninchen (Rabbit Heamorrhagic Disease, RHD) - Untersuchungen zum quantitativen Antigennachweis. J. Vet. Med. B 38, 621-629. Myerowitz, R., Stadler, H. and Oxman, M. (1975) Fatal disseminated adenovirus infection in a renal transplant recipient. Am. J. Med. 59, 591-598. Pay, T.W.F. and Hingley, P.J. (1992) Foot and mouth disease vaccine potency tests in cattle: the interrelationship of antigen dose, serum neutralizing antibody response and protection from challenge. Vaccine 10, 699-706. Prage, L. and Pettersson, U. (1971) Structural proteins of adenoviruses. VII. Purification and properties of arginine-rich core protein from adenovirus type 2 and type 3. Virology 45, 364-373. Smola, T., Specht, G. and Liebermann, HT. (1990) Bestimmung der Viruskonzentration unter Verwendung eines Kleincomputers. Arch. exp. Vet. med. 44, 311-317. Strobbe, R.(1983) Use of short analytical ultracentrifugation runs for the isopycnic determination of FMDV concentration and density in autoformed CsCl gradients. Ann. Rech. Vet. 14, 233-237.