Biologic& (1995) 23, 75-81

In Vitro Inactivation of Human lmmunodeficiency Virus

by Ascorbic Acid

Bhupat D. Rawalt, Francesco Bartolini$. and Girish N. Vyast*

tDeparfment of Laboratory Medicine, University of California, San Francisco CA 94 143-0134, U.S.A., and
$Centro Trasfusionale E Di lmmunologia Dei Trapianti, via Francesco Sforza 35, Milano, ha/y

Abstract. In vitro inactivation of cell-free human immunodeficiency virus (CFHIV) was investigated by
mixing replication-competent virions with aliquots of a culture medium (RPMI) containing increasing
amounts (625500 pug/ml) of ascorbic acid (AA) at pH7. Similarly, mixtures of CFHIV and 500 pg/ml
AA in whole blood (WB) and leukocyte depleted blood (LDB) were made; control mixtures containing
either CFHIV or AA alone in each experiment were included. After holding the mixtures for 3 h at 4°C
the tubes containing WB and LDB mixtures were centrifuged to remove the blood cells. The respec-
tive supernatants, including the control aliquots, were layered over 05 x lo6 MT2 cells in quadrupli-
cate wells in microtitre plates. After 1 h of incubation at 37°C in an atmosphere of 5.0% carbon dioxide
to permit contact of viable virions, the fluid in each well was replaced with RPM1 containing 20% fetal
bovine serum (FBS). The incubation was then continued at 37°C for 5 days. On the basis of (1)
absence of syncytia formation, (2) 100% viability of MT2 cells as compared with the cell controls, (3)
absence of p24 antigen in the culture supernates, and (4) absence of HIV DNA in MT2 cells, we con-
clude that 500 pg/ml AA, in (a) RPMI, (b) WB, or (c) LDB, inactivated CFHIV in vitro. Furthermore, we
determined that addition of 500 pg/ml AA to platelet concentrates did not adversely affect the platelet
function tests during 5 days of storage at room temperature. These data warrant further work to
evaluate the mechanism of CFHIV inactivation by treatment of blood products with AA..

Introduction functional changes,‘j we studied the effect of the

The risk of acquiring transfusion-transmitted
human immunodeficiency virus (HIV) infection is
estimated to range between 1 in 38 000 and 1 in
300 000 donor blood units.’ The rare infected unit
could contain cell-associated (CAHIV) as well as cell-
free HIV (CFHIV). In our earlier studies we have
demonstrated that the removal of leukocytes by
filtration of blood through leukocyte-removal
also removed CAHIV from experimentally infected
blood;2 these filters, however, do not retain cell-free
HIV (CFHIV).3 Auxillary procedures for the inacti-
vation of CFHIV in blood remain to be developed.
The continuous presence of ascorbic acid (AA) (150
&ml) in the medium has been reported to inhibit
reverse transcriptase (RT) activity in uitro, with
consequent suppression of CAHIV replication
chronically infected cells. 4-5We report here the effect
of higher concentrations of AA with regard to the in
vitro inactivation of CFHIV in experimentally inocu-
lated culture medium (RPMI), whole blood WB), or
leukodepleted blood (LDB). Since red cells stored in
preservative solution containing AA do not undergo
CFHIV inactivating concentration of AA on the sta-
bility of platelets (PL) over a 5-day storage period.
Materials and methods
Blood products
filters Units of WI3 drawn from anti-HIV negative
donors were purchased from the Irwin Memorial
Blood Centers, San Francisco. LDB was obtained by
filtering WB through a new generation leukocyte
removal filter (prototype filter, Pall Biomed Corp.,
NY) that yielded LDB with less than 4.0 leuko-
cytes/ml.3
in Chemicals
AA and poly-L-lysine were purchased from Sigma
Chemical Co., St. Louis, MO. AA stock solution (10.0
mg/ml) was freshly prepared in RPM1 and sterilized
by membrane filtration (0.2~~ Millipore, Bedford,
MA); the pH of this solution was 7.0. Further dilu-
tions were prepared aseptically in RPMI, WB, or

1045-1056/95/010075+07 $08.00/O @ 1995 The International Association of Biological Standardization

76 B. D. Rawal et al.

LDB for use in the virus inactivation assay RPM1 the assay. CFHIV in WB and LDB was tested only in
1640, Hanks Balanced salt solution (HBSS), fetal mixtures containing a final concentration of 500
bovine serum, and gentamicin were purchased from ,ug/ml AA because lower doses were determined to be
the UCSF cell culture facility. ineffective in preliminary experiments with CFHIV-
RPM1 mixtures. Generally a lag of at least 3 h occurs
Cell lines and virus stock between the collection and the processing of donor
blood at blood centers, hence we empirically selected
We used RPM1 1640 with 20% fetal bovine serum
it as the contact period for virus inactivation.
and 40 units of penicillin/ml as our basic culture
Control mixtures included in each experiment were
medium and Ca” and Mg+--free HBSS for the
(a) AA 500 pg/ml in virus-free RPMI, WB or LDB, (b)
washing steps. CFHIV stock (HTLV,i,a strain) and
untreated RPMI, WB and LDB, and (c) CFHIV
MT2 cells were kindly provided by Dr Michael
(0.25 X lo5 sfu/ml) (as a positive control). Each mix-
Busch, Irwin Memorial Blood Centers, San
ture (1.0 ml) was kept for 3 h at 4°C. Following incu-
Francisco.
bation the tubes containing WB and LDB were
centrifuged at room temperature at 1OOOg for 15
Titration of stock virus
minutes. The respective supernatant (1.0 ml) was
Ten-fold serial dilutions of the virus stock were inoculated in 0.25 ml aliquots into four microtitre
prepared in pre-warmed RPM1 at 37°C and 0.1 ml plate wells (Nunc, 24 well plates) with each well con-
aliquots were inoculated in quadruplicate wells in taining 0.5 X lo6 MT2 cells; these plates were previ-
24-well microtitre plates (Nunc); each well cont.ained ously treated with poly-L-lysine (3.0 pg/well) to
05 X lo6 MT2 cells that were stuck on to the bottom ensure cell adhesion. After incubating for 1 h at
of the well with poly-L-lysine. After incubating for 37°C we removed the liquid from each well, rinsed
1 h at 37°C we washed the cells three times with the cells twice with HBSS and then added 1.0 ml
HBSS, added 1.0 ml RPM1 and continued incubation RPM1 with 20% FBS for continued cell propagation
up to the 5th day. Each well was scored for syncytia at 37°C in an atmosphere of 5% carbon dioxide for 5
formation at 48 h after inoculation. The contents of days. On day 2 we scored the syncytia formation in
quadruplicate wells respectively inoculated with each well and on day 5 we pooled the contents of each
each virus dilution, were pooled and centrifuged at quadruplicate set of wells and centrifuged at 800 g
800 g for 5 min on day 5. The microtitre well inocu- for 5 min to separate the cells. The supernatants
lated with the highest dilution of the virus stock that were tested for p24 antigen by ELISA (Du Pant) and
showed syncytia formation was designated to con- the cells were examined for p24 antigen, MT2 cell
tain 1 syncytia forming unit (sfu) per 0.1 ml. Further viability and for HIV DNA by polymerase chain reac-
confirmation of virus replication in each culture was tion (PCR) as described earlier.3 The AA-CFHIV
obtained by testing (a) culture supernates by ELISA mixture that yielded cultures showing (i) absence of
(Du Pant) and the cells by immunocytochemical p24 antigen both in the culture supernatant and in
staining for p24 antigen7 and (b) MT2 cell viability the cells, (ii) 100% host cell viability as compared
in each culture by neutral red uptake measured with that of the virus-free cell control, and (iii) the
spectrophotometrically. The optical density readings absence of HIV DNA in the host cells, was
corresponding to the number of viable cells in each determined to have contained the optimal CFHIV
preparation can be read from the reference curve. inactivating dose of AA under our experimental
conditions.
Virus inactivation assay
To study the inactivation of CHFIV in a static in Effect of the virus-inactivating dose of AA on platelet
vitro environment simulating the ‘blood bag’, we viability
determined the optimal CFHIV-inactivating dose of We monitored in vitro platelet viability in the
AA using a virus inoculum containing 0.5 X lo5 syn- platelet concentrates (PC) stored in the presence or
cytia forming units (sfu). This inoculum was mixed absence of AA to evaluate if the presence of CFHIV-
with solutions of increasing amounts of AA to obtain inactivating concentration of AA affected platelet
final mixtures containing 0.25 X lo5 CFHIV and AA quality during a 5-day storage. The WB (450 2 45
concentrations (62.5, 125, 250, and 500 pug/ml) in ml) from healthy blood donors was collected in
RPMI. In preliminary experiments we had deter- quadruple PVC bags containing 63 ml of citrate-
mined that AA concentrations exceeding 700 pg/ml phosphate-dextrose as an anticoagulant (Fenwal,
were progressively cytotoxic to the host cells used in Deerfield, IL). Identical PC were prepared according
 Inactivation of HIV by ascorbic acid 77

to methods described earlier.g In each of six experi- we found that inactivation of CFHIV by AA was dose
ments, two Al30 identical blood units were cen- dependent. The culture supernates from mixtures
trifuged for 7 min, and platelet-rich plasma (PRP) containing 250 pg/ml or less AA and CFHIV showed
units were collected in satellite bags and pooled in a syncytia formation, lowered host cell viability, HIV
1000 ml PVC transfer bag. Pools were split in two DNA integration with host cell DNA, and p24 anti-
PL-732 polyolefin 300 ml containers (Fenwal) which gen in cells as well as the supernatants. Thus, con-
were then centrifuged at 3000 g for 18 min to pre- centrations of 250 pglml of AA did only partially
pare identical PC. The volume of platelet concen- inactivate the CFHIV inoculum under our experi-
trates was 565 It 1 ml (n = 6). One hour following mental conditions. In contrast, similar cultures from
PC preparation, 500 ,ug/ml AA was added through a mixtures containing the same inoculum of CFHIV
sampling coupler to one of the identical PC (V-PC), and 500 pug/ml AA did not show evidence of virus
while the other served as control (C-PC). PCs were replication as assessed by the above criteria. The
stored at 22 ? 2°C in a tumbler (Helmer Labs, St. observed virus replication in the cultures from the
Paul, MN) for 5 days. The following measurements mixtures containing less than 250 pglml AA shows
were performed on the day after PC preparation (day that the virions not only attached to MT2 cells dur-
1) and after 5 days of storage: platelet count and ing the 1 h incubation at 37”C, but also that the rins-
mean platelet volume (MPV) by a Sysmex TOA NE- ing procedure employed to remove AA did not
8000 automatic counter (Sysmex, Kobe, Japan);‘O pH interfere with the virus adsoprtion to the cells and
at 22’C;” p0, and pC0, by an IL1302 automated gas integration with host cell DNA. Under the same con-
analyser (Instrumentation Laboratory, Milan, Italy); ditions, the cultures from the mixtures of 500 ,uglml
glucose concentration by an Olympus AU510 auto- AA and CFHIV in the RPM1 medium, or WB or LDB
matic analyser (Olympus, Tokyo, Japan); platelet reacted for 3 h were uniformly negative for any evi-
morphology, expressed as percent of platelets show- dence of virus replication. The data, presented in
ing discoid shape (% discs); osmotic reversal; ATP Table 1, show that the CFHIV inoculum, either in
release in response to thrombin (5 U/m1);12~‘3 extent RPMI, WB or LDB, was consistently inactivated by
of platelet aggregation-expressed as percent trans- 500 pg/ml AA in a contact time of 3 h at 4°C. Further,
mission maximum (T,,), in response to collagen and 500 pg/ml AA kept in contact with MT2 cells for 1 h
ADP -in combination with 2 ,uM epinephrine-by a at 37°C did not adversely effect cell viability during
lumiaggregometer (Chronolog Co., Havertown, subsequent culture, i.e, viability of MT2 cells inocu-
PA).14 The amount of P-selectin (GMP-140) present lated with the supernatants from mixtures contain-
in PC plasma was also evaluated by a commercial kit ing AA remained the same as that of the untreated
(Takara Shuzo Co. Ltd., Kyoto, Japan) to determine and uninfected cell controls. Thus, 500 ,ug/ml AA was
the extent of PC activation. Results were expressed not cytotoxic to the MT2 cells.
as a mean -I 1 S.D. and compared with the paired
Wilcoxin test.15 A P value of 0.05 or less was Effect of virus inactivating dose of AA (500 pg/ml) on
considered significant. platelet stability
Platelet and leukocyte counts were 61 2 8 X 10’
and 38 2 21 x 106, respectively, per unit of PC.
Results Figure 1 shows changes in platelet count, discoid
morphology, and mean platelet volume during
Virus stock
storage, together with the evaluation of plasma
An inoculum of 0.1 ml of the virus stock was seri- P-selectin levels as a marker of platelet alpha
ally diluted in 10 X steps. The highest dilution ( 10m5) granule secretion. 12*13Identical increase in P-selectin
that induced syncytia formation in the MT2 cells in PC stored in the presence of AA and controls
was designated to contain 1 sfu/Oal ml; thus, the confirmed that AA did not affect normal platelet
virus stock was determined to contain lo5 sfu/O.l ml. activation known to occur during storage.i2v16
Platelet metabolism, as judged by oxygen consump-
Inactivation of CFHIV by ascorbic acid in RPM/, Wl3 tion, CO2 production, pH, and the rate of decrease in
and LDB glucose concentration was similar in V-PC and con-
By mixing 0.5 ml of the virus stock with O-5 ml trols (Fig. 2). Measurements reflecting platelet
AA (concentrations ranging between 62.5 pug- 500 function and aggregation response in the presence
pg/ml), we tested each concentration of AA against of different stimuli are presented in Fig. 3. A
0.25 x 105 sfu of CFHIV. Under our assay conditions significant reduction of osmotic response was
78 B. D. Rawal et al.

Table 1. In vitro inactivation of human immunodeficiency virus by ascorbic acid

Ascorbic acid (Fug/ml) in

RPM1 WB LDB
--
HIV markers 0 62.5 125 250 500 500 500

Syncytia formation + + + + - - -
HIV DNA + + + + - - -
p24 antigen (pg/ml)
in culture supernatant 413 375 300 13 0 0 0
p24 antigen in + + + + - - -
MT2 cells
MT2 cell culture 20.7 38.4 51.63 53.26 96.5 94.26 100
viability’ 23.0 26.0 T3.8 24.4 53.9 26.5 26.5

’ % Viability of MT2 cells on day 5

a b
Platelet Count Morphology
% (% of Day 1) % (% discs)
100

80

60

40

20

1 2 3 4 -5
1 2 3 4 5
Day Day

C d
P-Selectin
Mean Platelet Volume nglmL
fL, (GMP-140)
ml

f +....A
6

“I 1 2 3 4 5
I
4 5
3 1 2
Day Day
Figure 1. Platelet count (a), morphology (b,, and volume (c) during storage of PC containing 500 pug/ml ascorbic acid (V-
PC, -o-, ra=6) and controls (C-PC, ..x.., n = 6). Results are expressed as a mean + 1 S.D. Note that the platelet
morphology is identical in V-PC and C-PC after 5 days of storage; about 40% of platelets showed discoid morphplogy.
The figure (c,d) shows a slight platelet swelling and a marginal increase in the concentration of P-Selectin (GMP-140), a
marker of platelet alpha granule secretion in stored plasma.
 Inactivation of HIV by ascorbic acid 79

a b

::” 1 . +___,,
mg/dL
1 + .“F . . . . . . .:‘::: .,.j

6.0
200
6.7
100
6.6 ,
#
1 2 3 4 5 0
Day 1 2 3 4 5
Day

C
d
pco2 mm Hg PO2

0' 5f
1 2 3 4 5 1 2 3 4 5
Day Day

Figure 2. Evaluation of pH (a) and measurements reflecting platelet metabolism (b-d) during storage of PC containing
500 pg/ml ascorbic acid (V-PC, -o-, n = 6) and controls (C-PC, ..x.., n = 6). Results are expressed as a mean 2 1 SD. A
similar pattern of oxygen consumption, CO, production, and glucose consumption was observed in V-PC and in C-PC; a
pH value above 7.0 was maintained throughout storage:

observed in V-PC compared to the controls on day 5 ture was further incubated for 1 h at 37°C with MT2
(P = 0.02). cells to capture the vii-ions that survived the action
of AA. At the end of this incubation the cells were
Discussion rinsed and fed with 1.0 ml RPMI. Thus, by removing
residual AA we were able to incubate the MT2 cells
Previous studies have reported the suppression of in in cultures devoid of significant amounts of AA. This
vitro replication of HIV in H9 cells by as little as 150 feature distinguishes our virus inactivation protocol
pg/ml of AA if it was continuously present in the cul- from the protocols previously applied for evaluating
ture medium during the incubation period.4B5 To anti-viral effect of kA or other chemicals; the latter
measure virus inactivation we reacted CFHIV assays are performed in the continuous presence of
(0.25 x lo5 sfu) with four incremental doses of AA the test compound during the culturing period.4*5
viz 62.5, 125, 250 and 500 pg/ml for 3 h at 4°C in The CFHIV inactivating dose of AA, as determined
RPM1 and 500 pg/ml in WB and LDB. Following this herein using the standard MT2 cell infectivity assay
interaction the supernatant of each virus-AA mix- for detecting CFHIV,17 is 500 wrnl and is certainly
80 6. D. Rawal et a/.

a b
Osmotic
ATPRdeasm
inResponse
to
x ““^‘f mdn (5 U/mL)
-1
‘001

0’
1 2 3 4 5
Day

AggregationADP (10 la)

d e f
Figure 3. Measurements reflecting identical platelet function (a-0, including aggregation response (c-D in the
presence of different stimuli, during storage of PC containing 500 pg/ml ascorbic acid (V-PC, -o-, IZ = 6) and
controls (C-PC, ..x.., n = 6). Results are expressed as a mean 2 1 SD. Response to hypotonic shock was lower in
V-PC when compared to C-PC on day 5 of storage.

higher than the 150 pg/ml replication inhibitory Based on our in vitro data, further in viuo studies in
dose; this is to be expected as the inactivating dose chimpanzees are warranted.
represents the virucidal concentration of AA. Although our data illustrate the virucidal effect of
The lack of syncytia formation, HIV DNA, and p24 500 pg/ml of AA on CFHIV, the precise mechanisms
antigen in the culture supernates and the cells, as of virus inactivation by AA remain to be defined.
shown in Table 1, experimentally establishes the Interestingly, AA chelates Mg” in the cell envelope
virucidal effect of 500 pg/ml A4 on CFHIV RPMI, of lipid-rich bacteria-Pseudomonas aeruginosa.‘g It
WB, and LDB. We also found that the addition of 500 is likely that AA also chelates Mg- + in CFHIV to dis-
pg/ml of AA to donor blood used for making the PCs able its Mg-+-dependent RT activity that is essential
had no significant effect on platelet stability or func- for the replication of retroviruses.20 AA also causes
tion during the 5-day storage period. This obser- the inhibition of avian RNA tumor viruses,21 and the
vation complements the previous report of no inactivation of herpesviruses and paramyxoviruses
functional ill-effects of ascorbate-2-phosphate on the in uitro.22 It will, therefore, be interesting to deter-
red cell preservation, storage and post-transfusion mine if AA acts on these viruses through the chela-
survival.6 Because the virucidal close of AA added tion of Mg +’ ions and that such in vitro treatment is
to the units of blood diminishes by only 10% over a broadly applicable to inactivation of other blood-
24-h period,18 the virus inactivation accomplished in borne virus infections.
3 h should remain uninfluenced by the subsequent
oxidative breakdown of AA. It is also noteworthy Acknowledgements
that our CHFIV inactivation dose of 500 pg/ml AA is The study was supported by NIH grant ROl-HL-
2000 times smaller than the 1 000 000 pug approved 41365. We with to thank Leopold0 Azucena and
for intravenous administration in human beings.‘” Chris Dang for their technical assistance.
 Inactivation of HIV by ascorbic acid 81

References 11. Holmes S, Vaidya K, Murphy S. Platelet storage at

208C: Effect of type of agitation on morphology, via-
1. Cumming PD, Wallace EL, Schorr JB, Dodd RY. bility and function in vitro. Blood 1978; 52: 435.
Exposure of patients to human immunodeficiency 12. Bertolini F, Rebulla P, Porretti L, Murphy S. Platelet
virus through the transfusion of blood components quality after 15-day storage of platelet concentrates
that test antibody negative. N Engl J Med 1989; 321: prepared from buffy coats and stored in a glucose-free
941-946. crystalloid medium. Transfusion 1992; 32: 9.
2. Rawal BD, Schwadron R, Busch MP, Endow R, Vyas 13. Holme S, Heaton WA, Courtright M. Improved in vivo
GN. Evaluation of leukocyte removal filters modelled and in vitro viability of platelet concentrates stored
by use of HIV-infected cells and DNA amplification. for seven days in a platelet additive solution. Br J
Blood 1990; 76: 2159-2161. Haematol 1987; 66: 233.
3. Rawal BD, Vyas GN, Yen TSB, Endow R, Heitman J, 14. Di Minno G, Silver MJ, Murphy S. Stored human
Busch MP. Leukocyte filtration removes HIV-infected platelets retain full aggregation potential in response
lymphocytes and infectious particulate cell debris but to pair of aggregating agents. Blood 1982; 59: 563.
not free virus. VOX Sanguinis 1991; 60: 124-128. 15. Glantz SA. Primer of biostatistics. New York,
4. Harakeh S, Jariwala RJ, Pauling L. Suppression of McGraw-Hill, 1981.
human immunodeficiency virus replication by ascor- 16. Finjheer R, Modderman PW, Veldman H, Ouwehand
bate in chronically and acutely infected cells. Proc WH, Nieuwehuis HK, Roos D, de Korte D. Detection of
Nat1 Acad Sci USA 1990; 87: 7245-7249. platelet activation with monoclonal antibodies and
5. Harakeh S, Jariwala RJ. Comparative study of the flow cytometry changes during platelet storage.
anti-HIV activities of ascorbate and thiol containing Transfusion 1990; 30: 20.
reducing agents in chronically HIV-infected cells. Am 17. Harada S, Koyangi Y, Yamamoto N. Infection of
J Clin Nutr 1991; 54: 1231S-1235s. HTLV-III/LAV in HTLV-I carrying cells MT-2 and MT-
6. Moore JL, Marks DH, Carmen RA, Ledford ME, 4 and application in a plaque assay. Science 1989; 229:
Nelson EJ, Fishman RJ, Leng BS. Ascorbate-2-phos- 217.
phate in red cell preservation. Clinical trials and 18. Reynolds JEF. Ascorbic acid. In J.E.F. Reynolds (Ed)
active components. Transfusion 1988; 28: 221-225. Martindale. The Extra Pharmacopoeia. The Pharma-
7. Busch MP, Rajagopalan MS, Gantz DM, Shuyuarn F, ceutical Press, London, 1993; pp 1057.
Steiner KS, Vyas GN. In situ hybridization and 19. Rawal BD. Bactericidal action of ascorbic acid on
immunocytochemistry for improved assessment of Pseudomonas aeruginosa - alteration of cell surface
human immunodeficiency virus cultures. Am J Clin as possible mechanism. Chemotherapy 1978; 24:
Path 1987; 88: 673-680. 166-172.
8. Montefiori DC, Robinson WE, Jr, Sshuffman SS, 20. Varmus H, Swanstrom R. Replication of retroviruses.
Mitchell WM. Evaluation of antiviral drugs and neu- In Weiss R. (Ed) RNA Tumor Viruses; Cold Spring
tralizing antibodies to human immunodeficiency Harbor Laboratory, 2nd ed. 1984, p 385.
virus by a rapid and sensitive microtiter infection 21. Bissell MJ, Hatie C, Farson DA, Schwarz RI, Soo W.
assay. J Clin Microbial 1988; 26: 231-235. Ascorbic acid inhibits replication and infectivity of
9. Slichter SJ, Harker LA. Preparation and storage of avian RNA tumor virus. Proc Nat1 Acad Sci USA
platelet concentrates. I. Factors influencing the har- 1988; 77: 2711-2715.
vest of viable platelets form whole blood. Br J 22. White LA, Freeman CY, Forrester BD, Chappell WA.
Haematol 1976; 34: 395. In vitro effect of ascorbic acid on infectivity of her-
10. Bertolini F, Rebulla P, Porretti L, Sirchia G. pesviruses and paramyxoviruses. J Clin Microbial
Comparison of platelet activation and membrane gly- 1986; 24: 527-531.
coprotein Ib and IIb-IIIa expression after filtration
through three different leukocyte removal filters. VOX Received for publication 28 June 1994;
Sang 1990; 59: 201. accepted 19 August 1994.