INFECTION AND IMMUNITY, JUlY 1976, p. 135-143Copyright © 1976 American Society for Microbiology

Vol. 14, No. 1Printed in U.S.A.

Thermal Inactivation of Rabies and Other Rhabdoviruses:Stabilization by the Chelating Agent

Ethylenediaminetetraacetic Acid at PhysiologicalTemperatures

FRANK MICHALSKI, NANCY F. PARKS, FRANTISEK SOKOL,' AND H. FRED CLARK*

The Wistar Institute ofAnatomy and Biology, Philadelphia, Pennsylvania 19104

Received for publication 21 January 1976

Thermal inactivation of rabies and several other rhabdoviruses was studiedusing virus suspended in several different diluents. Rabies serogroup viruseswere more stable than Kern Canyon or vesicular stomatitis viruses. Limitedstudies of two fish rhabdoviruses requiring low temperatures (<33 C) for repli-cation indicated that they were not markedly more thermolabile than rabiesvirus. Bovine serum protein components in complex cell culture media stabilizedvirus at 56 C, but at temperatures of s37 C, sodium tris(hydroxymethyl)-aminomethane (NT) buffer containing ethylenediaminetetraacetic acid (EDTA)(NTE) was a much more efficient stabilizer of virus infectivity. Chelating agentsEDTA and ethyleneglycol-bis-(,8-aminoethyl ether)tetraacetic acid were equallyefficient in protection of rabies virus infectivity; the effect of each was lost whenexcess Ca2+ was added. Bovine serum in NT or NTE buffers produced a thermo-stabilizing effect at 37 C not provided by the same serum concentration incomplex cell culture media. Bovine serum was more efficient than EDTA instabilizing virus infectivity during repeated cycles of freezing and thawing.

We have previously described the thermoin-activation of several strains of fixed rabies vi-rus and rabies ts mutants, propagated in cellculture and heated at 40.5 C (5, 8); other pub-lished reports on rabies virus thermoinactiva-tion have dealt with virus propagated in mousebrain (21, 30). The analysis of thermoinactiva-tion of vesicular stomatitis virus (VSV) hasbeen restricted to one or a few temperaturesand the influence of different suspending mi-lieu on thermoinactivation has not been evalu-ated (11, 18, 19, 29).Because of the increasing importance of cell

culture-propagated (concentrated or otherwise)rabies virus in vaccine production (33), wewished to determine the thermostability ofsuchvirus at several temperatures, in a variety ofsuspending media. As part ofa continuing com-parative study ofproperties of a variety ofrhab-doviruses (2, 25, 28), we have also characterizedthe thermoinactivation of several other verte-brate rhabdoviruses.The diluents studied include solutions in

standard use in our laboratory-complex cellculture media supplemented with bovine serumproteins, buffered salt solutions with or without

I Deceased 25 May 1974.

chelating agents, and buffered salt solutionssupplemented with serum. Remarkable differ-ences were noted in the thermostabilizing ef-fects of protein in cell culture medium as com-pared with simple buffers. Differences in theeffect of added cations (Mg2+ and Ca2+) on thethermostabilizing efficiency of ethylenedia-minetetraacetic acid (EDTA) and ethylenegly-col - bis(,8 - aminoethylether)tetraacetic acid(EGTA) was also documented.

MATERIALS AND METHODSVirus strains. Rabies virus strains ERA and Pit-

man-Moore (PM) (7), VSV strain Indiana, Lagos batvirus (3, 27) (Wistar clone 3) (9), Mokola virus (20,27) (Wistar clone 1) (9), Kern Canyon virus (KCV;23), spring viremia of carp virus (SVCV; 14), andpike fry rhabdovirus (PFV; 10) were used. All viralstocks were grown in BHK-21 cell monolayers in-fected at a cell multiplicity of approximately 1.0 andfed with Eagle minimal essential medium supple-mented with 0.1% bovine serum albumin and twicethe normal amount of sodium bicarbonate (MEM-0.1 BSA). Mammalian rhabdovirus-infected cellswere incubated at 33 C; SVCV and PFV were grownat 23 C (6). VSV was harvested after 24 h, KCV,SVCV, and PFV after 48 h, and rabies, Mokola, andLagos viruses after 72 h. The viruses were titratedby a plaque technique in agarose-suspended BHK-21/13S cells (26) except for SVCV and PFV, which

135

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

136 MICHALSKI ET AL.

were titrated in BHK-21 cell monolayers (6). Rabies,Lagos, and Mokola virus plaques were read after 6days at 35 C, KCV after 5 days at 35 C, VSV after 1day at 35 C, and SVCV and PFV after 3 days at 23 C.

Virus concentration. For released virus, the me-dium from infected cultures was harvested; for cell-associated virus, the cell sheet was washed threetimes with phosphate-buffered saline (PBS; Dul-becco) (13) and frozen and thawed three times inMEM-0. 1 BSA. Virus preparations were clarified bylow-speed centrifugation (20 min, 500 x g) and thensedimented at 36,000 x g for 1 h. The sediment,suspended overnight at 4 C in PBS, was sonicated at0 C at 1 A for 1 min with a Branson sonifier, againclarified by low-speed centrifugation, and then sedi-mented a second time. The sediment was again sus-pen4ed in PBS or NT buffer [sodium tris(hydroxy-methyl)aminomethane(Tris)] at 4 C and then usedimmediately as the stock virus, concentrated ap-proximately 100-fold.

Diluents. Virus was suspended in PBS (13), pH7.4, MEM-0.1 BSA, pH 7.9, BHK medium (22) con-taining 6% tryptose phosphate broth and 2% fetalcalf serum (BM-Sr2), pH 7.4, or NTE buffer [0.13 MNaCl, 0.05 M Tris and 0.001 M EDTA, pH 7.8].Concentrated virus stocks in PBS or NT buffer werediluted 1:100 in the test diluents prior to inactiva-tion studies.

Heat inactivation. The stock concentrated viruswas diluted and held in an ice bath (no longer than24 h) until the actual time of heating. Replicate 60-mm glass tubes (Wasserman) were filled with 1 ml

ERA-BSA 0.1

0 D

I ______NoE 0

2-z 223- 1560

04 I~~~~~~~o ~~~37~

0

of the diluted virus, tightly stoppered, and exposedto test temperatures for variable periods of time. Forharvest, the tubes were plunged into an ice bath, 1ml of BM-Sr2 was added to stabilize the residualinfectivity, and the sample was frozen at -70 C.

Calculation of half-life. Virus inactivation rateswere compared on the basis of half-lives of infectiv-ity. For a given set of inactivation conditions, thelog1,, of residual infectivities were plotted againsttime (see Fig. 1-3). The time required for 1 log1,decrease in titer was determined from the inactiva-tion curve (the initial phase of two-phase curves wasused); this time period was divided by 3.32 to givethe time required for 1 log, decrease in titer (or one"half-life") (15).

RESULTSEffect of temperature and diluent. A typical

pattem of inactivation of the infectivity of ra-bies virus (strain ERA) suspended in our stand-ard protein-containing virus growth medium(MEM-0.1 BSA) and exposed to various temper-atures is shown in Fig. 1. The virus titer de-creased more than 10')-fold in less than 15 minat 56 C, whereas less than a 102-fold decreasewas noted during 15 days of observation at 4 C.Infectivity inactivation curves were either ir-regularly monophasic or biphasic, with declina-tions in the slope of the inactivation curve oc-curring after varying time intervals.

30 2 3 4 5 6 7 8 9 10 11 12 13 14 15MINS DAYS

FIG. 1. Thermal inactivation of rabies virus (ERA) suspended in MEM-O. % BSA. 4, 9, etc. representimaginary values less than that of the point indicated. Inactivation curves were arbitrarily extended beyondthe last point determined by a real value.

INFECT. IMMUN.

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

THERMAL INACTIVATION OF RABIES VIRUS 137

NT

NTE

a. 2 ....0

0~~~~~~~~~~~~~~

3 0 liL~~~~~~~~~

4 -

U 0~~~~~~~~~~~~~~~~~~~~~~E-S .

a %~~~~~~~~~~~~~~~~~~~~~~~~~6 MEM-BSA o.\ \~~~~~~~~~~~~~ME-OA

0~~~~ ~ ~ ~ ~ ~~~~~~~~

7' '

24 48 72 96 2'4 48 72 96HOuRS

FIG. 2. Thermal inactivation of (a) rabies virus (ERA) and (b) VSV at 37 C. Symbols: @, released virus(RV) in PBS; 0, RV in MEM-Oi1% BSA; U, RV in NTE; 0, RV in BM-Sr2.

The effect of the diluent upon the rate ofthermal inactivation of rabies virus (releasedvirus) and released VSV at 37 C is illustratedin Fig. 2a and b. It is clear that both viruses areprotected against thermal inactivation muchmore efficiently by NTE buffer (protein free)than by the other diluents, two of which con-tain protein additives. Cell-associated rabiesvirus (data not shown) was inactivated at aslightly more rapid rate than was released vi-rus.The effect of the diluent upon the rate of

thermal inactivation of these two viruses at56 C (Fig. 3a and b) was markedly different. Atthis temperature each virus was protected moreefficiently by protein-containing diluents thanby either NTE or PBS buffers. After inactiva-tion of the greatest portion of the virus infectiv-ity within 5 to 10 min, a small fraction of ther-mostabile virus sometimes persisted for up to 30min in protein-containing diluents only and inNTE buffer in the case of VSV only. When ra-bies virus that had survived exposure to 56 Cfor 30 min in BM-Sr2 was regrown in BHK-21cells and reheated at 56 C, rapid inactivation ata rate similar to that of the parental virus stockwas obtained, indicating that the thermostable

rabies virus population did not result from ge-netic variance.

Half-life determinations indicating the rela-tive effects oftemperature and diluent upon therates of inactivation of a variety of rhabdovi-ruses are listed in Table 1. The effect of thediluent on the thermostability of the vertebraterhabdoviruses was remarkably similar. Virussuspended in diluents containing bovine serumproteins tended to best resist inactivation toheating at 56 C. At temperatures of 37 C orbelow, NTE was uniformly the most protectivediluent; at these temperatures, protein-con-taining diluents provided a lesser degree or noprotection.

Striking differences in the rate of thermalinactivation of the rabies (PM and ERA), rabiesserogroup (Lagos and Mokola), and fish rhab-doviruses were not noted. VSV and KCV wereinactivated at rates approximately twice thoseof the rabies serogroup viruses. A nonconcen-trated preparation of rabies virus (PM) wasmore stable than concentrated virus regardlessof diluent when heated at 56 C but not at lowertemperatures.

Identification of virus-thermostabilizingcomponent of NTE buffer. NTE buffer, clearly

VOL. 14, 1976

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

138 MICHALSKI ET AL.

superior to other test diluents in its thermosta-bilizing effect on rhabdoviruses tested at physi-ological temperatures, differed from other testdiluents in its content of both Tris and EDTA.However, when rabies virus was heated at 37 Cafter suspension in Tris buffer with or withoutEDTA (Table 2), it became clear that the EDTAwas the thermostabilizing component.

If EDTA stabilizes rabies virus, it should bepossible to titrate the effect in serially de-creased concentrations of EDTA. The results ofa series of experiments, carried out at 37 C,that demonstrate this effect are listed in Table3. EDTA effectively protected rabies virus in-fectivity over a broad range of concentrations,from 0.1 to 0.0001 M, with an optimum effi-ciency observed at a concentration of 0.01 M.The protective effect of EDTA was completelylost only at a dilution to 0.00001 M concentra-tion. An experiment with unconcentrated virus(experiment 3, Table 3) revealed that the pro-tective effect of EDTA was not limited to viruspossibly damaged during the manipulations in-cluded in our concentration procedure.Inasmuch as EDTA appeared to specifically

protect rhabdoviruses, the assumption thatsuch protection might be attributed to its che-

aERA-56'

0-

E

a. 2-

WI ~~0~,-. 3-

<,4- %\@ o B"5r_ "1"a:U

5-

0o )

lating activity for specific cations was tested ina virus-heating experiment in which Ca2+ andMg2+ in various concentrations were added tovirus preparations suspended in NTE (0.001 MEDTA) or NTEGTA (NT + 0.001 M concentra-tion of the chelating agent EGTA) (Fig. 4).Concentrated rabies virus (ERA) in NT bufferwas diluted 100-fold in NT, NTE, NTEGTA, orsodium citrate (SSC, 0.15 M NaCl, 0.015 so-dium citrate, pH 7.4) buffer. To replicate tubesof virus plus NTE or NTEGTA were addedCaCl2, MgCI2, or equal amounts of CaCl2 andMgCl2 to give final concentrations of divalentcation that were equimolar, 10-fold less or 10-fold in excess of the concentration of the chelat-ing agents. The virus tubes were heated at37 C; samples were removed for virus titrationat 24 and 48 h.

All infectivity (104-fold reduction in titer) waslost in less than 24 h in virus suspensions in NTor SSC (data not shown) buffers. In the absenceof added cations, EDTA and EGTA wereequally effective in protecting virus againstthermal inactivation. Furthermore, the virus-protective effect of each chelating agent waslittle affected by added Ca2+ or Mg2+ inamounts up to equimolar concentration.

5 10 15 20 25 30 5 10 15 20 25 30

MINuTES

FIG. 3. Thermal inactivation of (a) rabies virus (ERA) and (b) VSV at 56 C. Symbols: *, released virus(RV) in PBS; 0, RV in MEM-0.1%BSA; 0, RV in NTE; O, RV in BM-Sr2; A, cell-associated virus (CAV) inPBS; A, CAV in MEM-O.1% BSA.

INECT. IMMUN.

I

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

THERMAL INACTIVATION OF RABIES VIRUS 139

TABLE 1. Effect of temperature and diluent on thermal inactivation of rhabdoviruses

Half-lives of infectious titer" at:Virusb Diluent

56 C (min) 37 C (h) 33 C (h) 23 C (days) 4 C (days)

Rabies ERA PBS 0.50 2.4 4.6 0.85 3.1

Rabies PM

PM-NC

Lagos

Mokola

VSV

KCV

SVCV

BM-Sr2NTE

PBSMEM-BSANTEPBSMEM-BSANTE

PBSMEM-BSANTE

PBSMEM-BSANTE

PBSMEM-BSABM-Sr2NTE

PBSMEM-BSANTE

0.750.60

0.300.480.300.720.940.45

0.300.550.37

0.370.720.37

0.220.420.410.30

0.190.680.23

NTMEM-BSANTE

PFV NTMEM-BSANTE

6.214.4

0.483.37.04.32.98.1

5.936.8

0.855.4

4.725.5

<0.724.6

10.1<0.722.76.5

<0.483.57.7

1.84.39.4

1.11.62.6

10.8

1.42.35.1

12.0

0.140.180.460.55

1.12.76.04.6

0.241.58.7

3.63.47.5

4.62.96.6

a A half-life = time for 1 log base2 decrease in titer during the initial phase of the thermal inactivationcurve. Details of technique are given in Materials and Methods.

b Concentrated released virus, except where designated otherwise. NC, Released virus not concentrated.' Abbreviations: PBS, PBS (Dulbecco); MEM-BSA, MEM-0.1% BSA; NT, Na Tris buffer; NTE, Na Tris

buffer with 0.001 M EDTA; BM-Sr2, BHK medium with 2% fetal calf serum.

TABLE 2. Effect ofEDTA component ofNTE bufferon the inactivation of rabies virus (ERA) at 37 C

Infectivity of rabies virus afterheating in diluent

Time ofheating NT NTE

(h)Titer pH Titer

(PFU/ml)a (PFU/ml) pH

0 1.4 x 105 7.8 1.2 x 105 7.872 <5 x 10° 7.8 2.2 x 103 7.8

a PFU, Plaque-forming units.

(Added trace amounts of Ca2+ or Mg2+ appearedto slightly enhance the protective effect ofEDTA.) Addition of a 10-fold excess of Ca2+ (orthreefold excess of Ca2+ combined with a five-fold excess of Mg2+ [data not shown]) com-

pletely abolished the protective effect of EGTAand greatly reduced the protective effect ofEDTA. Addition of excess Mg2+ reduced theprotective effect of EDTA but had no effect onthe efficiency of EGTA.

Protective effect of serum in NT buffer.Standard proteinaceous virus diluents in use inthis laboratory are formulated from complexcell culture media. As a further control on thespecificity of the virus-protective effect notedwith NTE, the rate of inactivation of rabiesvirus was studied in NT and NTE buffers with2% serum added, as well as in the standardprotein-containing diluent MEM (BHK) with 2or 10% serum (Fig. 5).

Surprisingly, 2% serum in NT buffer effec-tively stabilized rabies virus against inactiva-tion at 37 C under conditions where 2 or 10%

VOL. 14, 1976

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

TABLE 3. Effect ofEDTA concentration in NTE buffer on the half-life of rabies virus (ERA)a at 37 C

Expt 1 Expt 2 Expt 3

Diluentb (1:5) Half-life Diluent (1:15) Half-life Diluent (1:10) Half-life

MEM (0.1% BSA) 3.3NT 4.3 NT 2.1 NT 4.3NT(0.1MEDTA) 21.2NT (0.01 M EDTA) 79.5NT (0.001 M EDTA) 17.3 NT (0.001 M EDTA) 11.5 NT (0.001 M EDTA) 15.8NT (0.0001 M EDTA) 13.1 NT (0.0001 M EDTA) 9.6

NT (0.00001 M EDTA) 2.9

aReleased virus. Concentrated virus in NT buffer was used in experiments 1 and 2. Unconcentrated virusin MEM-0.1 BSA was used in experiment 3.bThe pH of all diluents was 7.8.

o448-24 4 NTEGTA

20 NTEC" 0-

30 *~~~~~-mg"IO.3M NT

-1 % & -NNT cT0E

HOURS HOURS

FIG. 4. Effect of added cations upon thermoinactivation of rabies virus suspended in media containingchelating agents (a) EDTA (NTE medium) or (b) EGTA (NTEGTA medium). Chelating agents were presentat 10-3M concentration; Ca2+ or Mg2+ (as chloride salts) was added at concentrations indicated on the figureprior to heating the virus at 37 C.

serum added to BHK cell medium had no ther-mostabilizing effect. Very similar inactivationrates were observed in NTE, NT + 2% serum,and NTE + 2% serum. The thermostabilizingeffects ofEDTA and serum were apparently notadditive in this system.

Effect of EDTA and serum on rabies virusinactivation by freezing and thawing. Obser-vations of thermostabilizing effects of diluentsat different temperatures have suggested thatdifferent factors may account for thermal inac-tivation of viruses at 56 C and at lower, physio-logical temperatures (see below). Thus, an ad-ditional experiment was performed to deter-mine whether NT buffer with EDTA and NTbuffer with serum, equally effective in protect-ing rabies virus against inactivation at 37 C,

could protect virus against loss of infectivitycaused by freezing and thawing (Table 4).The results indicate that serum added to NT

or NTE completely protected virus against lossof infectivity after 10 cycles of freezing andthawing. The titer of virus suspended in NTEalone decreased 5-fold after similar treatmentbut was a full 10-fold less than that in serum-containing diluents because of an apparent in-crease in titer in such diluents. Virus in NTbuffer alone declined 100-fold in titer duringthis treatment.

DISCUSSIONWe have compared, by identical technique,

the effect of heating, at assorted temperatures,upon the viability of rabies and several other

140 MICHALSKI ET AL. INFECT. IMMUN.

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

THERMAL INACTIVATION OF RABIES VIRUS 141

rhabdoviruses propagated under similar condi-tions in BHK-21 cells. To our knowledge, this isthe first report of a systematic study of ther-moinactivation of rhabdoviruses. For eachgiven combination oftemperature and suspend-ing diluent, it was determined that the half-lives of rabies virus (strains ERA and PM) andthe rabies serogroup Lagos bat and Mokola vi-ruses were in general about twice those ob-served with VSV and KCV. Cdhsistent differ-ences in thermolability among the four rabiesserogroup viruses tested, which might haveserved as useful marker characteristics, werenot observed. The fish rhabdoviruses SVCVand PFV, which show much lower optimumtemperature for replication in BHK cells thando mammalian rhabdoviruses (6; H F. Clark,unpublished data), are not obviously more ther-molabile than rabies virus.

Biphasic inactivation curves were commonlyobserved with each virus when treated at 56 Cin diluents other than PBS or NT buffer. It wasdetermined that diluents containing serum pro-tein consistently afforded some protection toviruses heated at 56 C, whereas protein-freediluents did not. At temperatures of 37 C andbelow, NTE provided a remarkable degree ofprotection for each virus, whereas the protein-containing diluents based upon complex cellculture medium were totally inefficacious.These results suggest that different mecha-

nisms of inactivation are operative at 56 C andat lower, physiological temperatures. A similarpattern detectable during heat inactivation ofpicornaviruses has previously been suggestedbased upon observation of: (i) two-componentinactivation curves at high temperatures (12,34); (ii) stabilization of infectivity by treatmentwith 1 M Mg2+, preferentially at high.tempera-tures (12); (iii) inactivation of infectivity of vi-rus and nucleic acid at different rates at high

0o ERA-37

22R-Sr 10

zNT-Sr 2

24

w ~~~~~~~~~~NTE~.5-

06-j

24 48 T2 96 120 144

HOURS

FIG. 5. Inactivation at 37 C ofrabies (ERA) virus

suspended in NT buffer, with or without addedEDTA and/or serum, or in BHK cell medium withadded serum.

TABLE 4. Inactivation ofrabies virus by freezing andthawinga

Freeze and PFU/ml in diluent:thaw cycles NT NTE NT-Sr2 NTE-Sr2

5x 3.1 x l 1.1x104.2 x101.2 x 108lox 2.5 x 10O 5.0 x 106 11.0x 108 5.0x 107

a Rabies virus (ERA) was concentrated by sedimentationand suspended in a minimum quantity of NT buffer. Con-centrated virus was diluted 100-fold in test buffers andrapidly frozen and thawed 10 times, using a dry ice-ethanolbath and a 37 C water bath. Initial titer, 2.6 x 107 plaque-forming units (PFU)/ml. NT-Sr2 and NTE-Sr2 are bufferswith 2% fetal calf serum.

temperature but at similar rates at lower tem-perature (12); and (iv) isolation of virus var-iants whose stability was increased at hightemperature but unchanged at physiologicaltemperatures (35). Dimmock (12) has suggestedthat these data are compatible with virus inac-tivation primarily by protein damage at thehigher temperatures and by disruption of nu-cleic acid at physiological temperatures. Physi-cal fragmentation of the ribonucleic acid ge-nome of rhinoviruses inactivated by heating at34.5 C has recently been demonstrated (17).The mechanism by which EDTA protects

rhabdoviruses against heat inactivation hasnot been determined. The observation is inagreement with findings of others that sub-stances with chelating properties, heparin inthe case of rabies virus (30) and EDTA andbentonite in the case of a variety of viruses (1),protect infectivity. That a chelating agentmight be expected to protect virus infectivitymight also be expected from the reports of Wal-lis and Melnick that 1.0 M concentrations ofCa2+ or Mg2+ enhanced the thermolability ofseveral types of ribonucleic acid or deoxyribo-nucleic acid viruses (31, 32) and from earlierreports that bacteriophage exhibited increasedlability in media containing these ions in con-centrations exceeding 0.15 M (4). Nevertheless,the fact that we observed protection of rhabdo-viruses at EDTA concentrations as low as0.0001 M suggests that we have observed aprotective phenomenon based upon chelation ofmuch lower concentrations of cations. The che-lating effect may act by removing cations (i)that act as necessary cofactors for degradativeenzymes, (ii) that non-enzymatically catalyzedecomposition of virus components, or (iii) thatpoison endogenous virus enzymes necessary forreplication.We do not know which ion is critically impor-

tant; it is difficult to interpret the results of ourexperiment in which excess Mg2+ and Ca2+were added to EDTA- or EGTA-protected virussuspensions. The efficient protective effect ofEGTA, which allegedly complexes only Ca2+

VOL. 14, 1976

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

142 MICHALSKI ET AL.

and Ba2+ (24), suggests that Ca2+ may be thecnritical cation. Two contrary observations are:(i) the fact that excess Mg2+ greatly reduces theprotective effect of EDTA despite the fact thatMg2+ is not capable of displacing Ca2+ fromEDTA (16), and (ii) a slight protective effect ofEDTA and EGTA persists in the presence of 5-to 10-fold excess concentrations of Ca2+, despitethe fact that these chelating agents can com-plex cations only on an equimolar ratio (16).The latter observations suggest that the addedCa2+ and Mg2+ may eliminate the protectiveeffect of the chelating agents by displacingsome other cation, possibly present only intrace amounts.Serum added to NT buffer protects rabies

virus at physiological temperatures under con-ditions where serum added to cell culture me-dium is totally without effect. Serum andEDTA protective effects in NT buffer are notadditive. The explanation of these observationsis uncertain. It is possible that serum itselfexerts a minor chelating effect or that serumand EDTA may act in a similar nonspecificvirus-complexing manner in simple diluentsonly.

Final elucidation of the mechanism of theprotective effect of EDTA and EGTA uponrhabdovirus will require study of the roleplayed by other cations in the thermal inactiva-tion of purified virus and the biochemical andmorphological characterization of virions ex-posed to heat under chelating agent-protectedand unprotected conditions. In the meantime, itis apparent that addition of EDTA or EGTA tothe virus-suspending medium in minimal con-centrations may be expected to very effectivelystabilize the infectivity of rhabdoviruses un-dergoing storage for repaarch purposes or foruse as live virus vaccines. Addition of serum toNTE is indicated for preparations to be frozen.Cell culture medium is a poor milieu for themanipulation of rabies virus in the laboratory.

ACKNOWLEDGMENTS

This work was supported by Public Health Service re-search grants AI 09706, from the National Institute of Al-lergy and Infectious Diseases, and RR 05540, from the Divi-sion of Research Resources, and by funds from the WorldHealth Organization.

LITERATURE CITED

1. Bishai, F. R., and N. A. Labzoffsky. 1974. Stability ofdifferent viruses in a newly developed transport me-dium. Can. J. Microbiol. 20:75-80.

2. Bishop, D. H. L., H. G. Aaslestad, H F. Clark, A.Flamand, J. F. Obijeski, P. Repik, and P. Roy. 1975.Evidence for the sequence homology and genome sizeof rhabdovirus RNAs, p. 259-292. In R. P. Barry andB. W. T. Mahy (ed.), Negative strand viruses. Aca-demic Press Inc., New York.

3. Boulger, L. R., and J. S. Porterfield. 1958. Isolation ofa virus from Nigerian fruit bats. Trans. R. Soc. Trop.Med. Hyg. 52:421-424.

4. Burnet, F. M., and M. McKie. 1930. Balanced saltaction as manifested in bacteriophage phenomena.Aust. J. Exp. Biol. Med. Sci. 7:183-198.

5. Clark, H F., and H. Koprowski. 1971. Isolation of tem-perature-sensitive conditional lethal mutants of"fixed" rabies virus. J. Virol. 7:295-300.

6. Clark, H F., and E. Z. Soriano. 1974. Fish rhabdovirusreplication in non-piscine cell culture: new system forthe study of rhabdovirus-cell interaction in which thevirus and cell have different temperature optima.Infect. Immun. 10:180-188.

7. Clark, H F., and T. J. Wiktor. 1972. Rabies virus, p.177-182. In S. A. Plotkin (ed.), Strains of humanviruses. S. Karger, Basel.

8. Clark, H F., and T. J. Wiktor. 1972. Temperature-sensitivity characteristics distinguishing substrainsof fixed rabies virus: lack of correlation with plaque-size markers or virulence for mice. J. Infect. Dis.125:637-645.

9. Clark, H F., and T. J. Wiktor. 1974. Plasticity ofpheno-typic characters or rabies-related viruses: sponta-neous variation in the plaque morphology, virulence,and temperature-sensitivity characters of seriallypropagated lagos bat and mokola viruses. J. Infect.Dis. 130:608-618.

10. de Kinkelin, P., B. Galimard, and R. Bootsma. 1973.Isolation and identification of the causative agent of"Red disease" of pike (Esox lucius L. 1766). Nature(London) 241:465-467.

11. Deutsch, V., and A. Berkaloff. 1971. Analyse d'un mu-tant thermolabile du virus de la stomatite vesiculaire(VSV). Ann. Inst. Pasteur Paris 121:101-106.

12. Dimmock, N. J. 1967. Differences between the thermalinactivation of picornaviruses at "high" and "low"temperatures. Virology 31:338-353.

13. Dulbecco, R., and M. Vogt. 1954. Plaque formation andisolation of pure lines with poliomyelitis viruses. J.Exp. Med. 99:167-182.

14. FUan, N., Z. Petrinec, ID. Sulimanovic, and L. 0. Zwil-lenberg. 1971. Isolation of the viral causative agentfrom acute form of infectious dropsy of carp. Vet.Arch. 41:125-138.

15. Finney, D. J., T. Hazelwood, and M. J. Smith. 1955.Logarithms to base 2. J. Gen. Microbiol. 12:222-225.

16. Garvan, F. L. 1964. Metal chelates of ethylenediamine-tetraacetic acid and related substances, p. 283-333. InF. P. Dwyer and D. P. Mellon (ed.), Chelating agentsand metal chelates. Academic Press Inc., New York.

17. Gauntt, C. J. 1974. Fragmentation of RNA in virusparticles of rhinovirus type 14. J. Virol. 13:762-764.

18. Holland, J. J., D. Kohne, and M. V. Doyle. 1973. Analy-sis of virus replication in ageing human fibroblastcultures. Nature (London) 245:316-319.

19. Holloway, A. F., P. K. Y. Wong, and D. V. Cormack.1970. Isolation and characterization of temperature-sensitive mutants of vesicular stomatitis virus. Virol-ogy 42:917-926.

20. Kemp, G. E., 0. R. Causey, D. L. Moore, A. Odelola,and A. Fabiyi. 1972. Mokola virus. Further studies onIbAN 27377, a new rabies-related etiologic agent ofzoonosis in Nigeria. Am. J. Trop. Med. Hyg. 21:356-358.

21. Kuwert, E., and W. Liebenow. 1959. Die Stabilitat desTollwutvirus bei verschiedenen H-Ionenkonzentrati-onen, verschiedenen Temperaturen und Formaldehy-deinwirkung. Arch. Exp. Vet. Med. 13:335-344.

22. Macpherson, I., and M. Stoker. 1962. Polyoma transfor-mation of hamster cell clones-an investigation ofgenetic factors affecting cell competence. Virology16:147-151.

23. Murphy, F. A., and B. N. Fields. 1967. Kern Canyon

INFECT. IMMUN.

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

THERMAL INACTIVATION OF RABIES VIRUS

virus: electron microscopic and immunological stud-ies. Virology 33:625-637.

24. Perrin, D. D. 1970. Masking and demasking ofchemicalreactions, theoretical aspects and practical applica-tions, p. 42-44. Wiley-Interscience, New York.

25. Repik, P., A. Flamand, H F. Clark, J. F. Obijeski, P.Roy, and D. H. L. Bishop. 1974. Detection of homolo-gous RNA sequences among six rhabdovirus ge-nomes. J. Virol. 13:250-252.

26. Sedwick, W. D., and T. J. Wiktor. 1967. Reproducibleplaquing system for rabies, lymphocytic choriomen-ingitis, and other ribonucleic acid viruses in BHK-21/13S agarose suspensions. J. Virol. 1:1224-1226.

27. Shope, R. E., F. A. Murphy, A. K. Harrison, 0. R.Causey, G. E. Kemp, D. I. H. Simpson, and D. L.Moore. 1970. Two African viruses serologically andmorphologically related to rabies virus. J. Virol.6:690-692.

28. Sokol, F., H F. Clark, T. J. Wiktor, M. L. McFalls, D.H. L. Bishop, and J. F. Obieski. 1974. Structuralphosphoproteins associated with ten rhabdoviruses.

J. Gen. Virol. 24:433-445.29. Thormar, H. 1967. A comparison of cocal and vesicular

stomatitis virus, serotypes New Jersey and Indiana.Virology 31:323-328.

30. Turner, G. S., and C. Kaplan. 1967. Some properties offixed rabies virus. J. Gen. Virol. 1:537-551.

31. Wallis, C., and J. L. Melnick. 1962. Cationic stabiliza-tion-a new property of enteroviruses. Virology16:504-506.

32. Wallis, C., C-S. Yang, and J. L. Melnick. 1962. Effect ofcations on thermal inactivation of vaccinia, herpessimplex, and adenoviruses. J. Immunol. 89:41.

33. Wiktor, T. J., and H F. Clark. 1975. Growth of rabiesvirus in cell culture, p. 155-179. In G. Baer (ed.),Natural history of rabies, vol. 1. Academic Press Inc.,New York.

34. Woese, C. 1960. Thermal inactivation of animal vi-ruses. Ann. N.Y. Acad. Sci. 83:741-751.

35. Youngner, J. S. 1957. Thermal inctivation studies withdifferent strains of poliovirus. J. Immunol. 78:282-290.

VOL. 14, 1976 143

on April 28, 2021 by guest

http://iai.asm.org/

Dow

nloaded from