preparation and characterization of human hiv type 1 neutralizing reference sera

AIDS RESEARCH AND HUMAN RETROVIRUSESVolume 11, Number 7, 1995Mary Ann Liebert, Inc.

Preparation and Characterization of Human HIVType 1 Neutralizing Reference Sera

LUBA K. VUJCIC1 and GERALD V. QUINNAN, JR.1,2

ABSTRACT

Reference neutralizing antibody (NA) reagents are needed for laboratories to be able to compare results ofneutralization assays that will be used to monitor HIV-1 vaccine recipients. In an effort to establish such ref-erence reagents two asymptomatic, seropositive patients were identified with medium to high amounts of cross-reactive NA activity against a number of HIV-1 strains. Sera obtained from each individual at three or foursequential phlebotomies were pooled, and the two pools were each distributed in 3000 aliquots into glassampoules and lyophilized, and the ampoules were flame sealed. An HIV-1 antibody-negative reference serum

was prepared in a similar fashion after pooling serum from four individuals. Ampoules were tested for uni-formity of nil, sterility, moisture content, residual oxygen, stability, infectivity, and presence of antibody. Aninternational collaborative study was conducted to determine the potency of the samples in six laboratories,each using their own neutralization assays and reagents. The results indicated reasonable consistency betweenlaboratories and that both sera have sufficient titers against a variety of strains for Use as reference reagents.These reference sera have been included in the World Health Organization (WHO) AIDS Reagent Projectand are available through the three AIDS reagent repositories.

INTRODUCTION

Neutralizing antibody (NA) determinations are an im-portant feature of HIV-1 vaccine clinical trials. Because

NAs play a critical role in protection from viral infections theyhave typically been measured to assess if vaccine recipientsmight be protected sufficiently by immunization. ' Whether NAsare beneficial in HIV-1 infection has not been shown. However,they will be important immunological markers of responses tovaccines.

Presently, there is no single, generally used neutralizing an-

tibody assay and laboratories use assays that may vary in re-

gard to target cells, method of virus preparation, virus strains,virus potency, incubation times, duration of assay, and end-point markers used to determine virus replication.2-8 Three pre-vious international collaborative studies have been conductedto compare neutralization assays using either polyclonal sera or

monoclonal antibodies to HIV-1.4,9'10A World Health Organization (WHO) International

Workshop on Standardization of Neutralizing AntibodyMeasurements to HIV and Related Viruses was held in London

on October 3-5, 1988.n The need for standardized referencereagents was emphasized at that time and as a follow-up to thatmeeting we have prepared the sera described here for use as

reference reagents. Another workshop, entitled "Neutralizationof HIV-1" held on April 19-20, 1993 in Bethesda, Maryland,again stressed the need for standardized reagents to be used inneutralizing antibody assays.7 Monoclonal antibodies could beused for these purposes because they would have the advantageof being in an unlimited supply and lacking the synergistic phe-nomena potentially occurring in polyclonal sera, but they havethe potential disadvantage that they may be too specific.Polyclonal human sera are broadly reactive and may better suitthe need for reference reagents for human vaccine trials.

We selected two HIV-1-infected patients to be bled repeti-tively for sera to be used to prepare the reference reagents. Therepeated serum collections from each patient were pooled andwere treated and tested according to the WHO guidelines to beprepared in aliquots as reference reagents, including testing inan international collaborative study.1216 These samples havebeen designated to be reference reagents for research purposesin the WHO AIDS Reagent Project. They are now available

'Center for Biologies Evaluation and Research, Food and Drug Administration, 1401 Rockville Pike, Rockville, Maryland 20852-1448.2Present address: Department of Preventive Medicine, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road,

Bethesda, Maryland 20814.

783

784 VUJCIC AND QUINNAN

through the three WHO AIDS Reagent repositories (the NIHAIDS Research and Reference Reagent Program [Rockville,MD], the National Institute for Biological Standards andControl [Hertfordshire, England], and the Agence Nationale deRecherches sur le SIDA [Paris, France]). An HIV-1 antibody-negative human serum has been prepared in a similar fashion.

MATERIALS AND METHODS

Donors

The donors that were selected for the reference sera were

participants in a longitudinal study of progression of HIV in-fection and underwent extensive medical evaluation at each pe-riodic visit, including testing of their immunological status. Thisstudy was conducted at the National Institutes for Health (NIH,Bethesda, MD) with informed consent and it has been reviewedyearly by the Institutional Review Board. A panel of 10 donorswas initially screened in order to select 2 patients with high NAtiters against HIV-lmn- The two asymptomatic, HIV-1-infectedindividuals that were selected were phlebotomized under med-ical supervision for approximately 500 ml of blood on three or

four occasions over the course of 10 months. The units were

collected without anticoagulant and serum was removed aftercentrifugation. Each individual donation was tested for neu-

tralizing antibody titer and the final pooled samples for eachdonor were also tested. The final pooled individual serum col-lections consisted of 350 ml for patient 1 and 375 ml for pa-tient 2. An HIV-1 antibody-negative serum was prepared bypooling sera from four uninfected individuals with hemochro-

matosis who were being maintained in normal iron balance byperiodic phlebotomy. The volume of this serum pool was 720ml. The pooled sera were heat inactivated at 56°C for 30 min.

Distribution of sera into vialsThe sera were diluted 1:4 with half-strength phosphate-

buffered saline (PBS), pH 7.0, made with pyrogen-free doubleglass-distilled water prior to aliquoting. Samples were stirredthroughout the filling process and a Wheaton (Millville, NJ)peristaltic dispenser was used to place 0.5-ml amounts in 2-mlcapacity flat-bottom Wheaton Vacules made of type 1 borosil-icate glass, class A, that were washed and autoclaved prior tothe fill, and then were closed with Wheaton low-extractablegray butyl slotted stoppers. The fills were performed in a lam-inar flow cabinet in a class 100 clean room with a HEPA-fil-tered, positive-pressure air supply. Ampoules were stopperedby hand and placed in a precooled Hull (Hatsboro, PA) freezedryer. Three probes were distributed through the product loadto monitor the product temperature and the lots were subjectedto a 3-day freeze-dry cycle, stoppered under vacuum, and thenremoved from the freeze dryer. The ampoules were stored at4CC until flame-sealed under vacuum, reinforced with a coat ofparaffin, tested for sealing defects with a high-frequency gen-erator (model BD10A; Electro Technic Products, Chicago, IL),and stored upright in boxes at +4°C.

Quality control tests

To ensure uniformity of fill of the liquid samples beforelyophilization at least 3 weight measures per each tray of 942

Table 1. Summary of Neutralization Assay Methods in Collaborative Study

Participant(reference)

Targetcells

Serum-virusincubation

Length ofassay (days)

Replicationmarker

HTV-1strain

A (2)

B(-)bC(3)

Dl(4)

D2(4)

E(5)

F (6)

Molt-3a

C8166cC8166

C8166

C8166

AA-2d

PBMCe

90 min/37 °C

60 min/37 °C60 min/37 °C

90 min/room temperature

90 min/room temperature

120 min/4 °C

60 min/37 °C

Syncytia

p24 AgSyncytiaColorimetric

Syncytia

SyncytiaAcid reduction

p24 Ag

IIIBMNSF2RFZ3IIIBNY5-LAV-1RFMNSF2IIIBRFMNSF2IIIBRFPNL 4-3NY5Z34IIIBJ4499

aHuman lymphoblastoid cell line.bNo reference provided.cHuman umbilical cord blood lymphocyte (T lymphoid).dHuman male spleen cell subclone of AA derived from WIL-2 human splenic EBV+ lymphoblastoid line).Peripheral blood mononuclear cells.

HUMAN HIV-1 NEUTRALIZING REFERENCE SERA 785

ampoules were taken. After lyophilization, the moisture con-tent of the aliquots was analyzed using the gravimetric or losson drying method.16 Samples were tested for sterility in fluidthioglycolate medium at 31°C and soybean casein digestmedium at 22°C for 14 days. Because the samples were sealedunder vacuum, the oxygen content of the ampoules atmospherewas not tested. However, the samples were subjected to thehigh-frequency vacuum testing coil previously described toconfirm that vacuum was present. A short-term stability studywas conducted with five samples each at 56, 35, 27, and 4°C.Samples were reconstituted every 1-4 weeks depending on thetemperature and tested for neutralizing activity.17"19 The serawere also checked for infectivity by adding a vial of each tocultures of phytohemagglutinin (PHA)-stimulated peripheralblood lymphocytes (PBLs) and incubated at 37°C in the pres-ence of a 5% C02-in-air atmosphere for a period of 4 weeks.20Samples were checked daily for indication of giant cell forma-tion and weekly samples were taken for testing on a Du Pont(Wilmington, DE) HIV-p24 core profile enzyme-linked im-munosorbent assay (ELISA) test kit.

ELISAThe reference sera were tested by HIV-1 enzyme im-

munoassay (EIA), HIV-2 EIA, and a combination HIV-1/HIV-2 EIA (all made by Genetic Systems Corporation, Seattle, WA).

PotencyAn international collaborative study was designed with six

laboratories participating from the United States and Europe.The investigators were as follows: our own laboratory at theFood and Drug Administration (FDA) in Bethesda, Maryland;Dr. Gunnel Biberfeld, National Bacteriological Laboratory inStockholm, Sweden; Dr. Robin Weiss, Chester BeattyLaboratories, Institute of Cancer Research, Royal CancerHospital, London, England; Dr. Geoffrey Schild, NationalInstitute for Biological Standards and Control in Hertfordshire,England; Dr. Malcolm Martin, National Institutes of Health,Bethesda, MD; and Dr. Eva Maria Fenyö, Karolinska Institutet,Stockholm, Sweden. These laboratories are designated as lab-oratories A through F, respectively, in the remainder of this ar-

ticle. The sample set included 10 coded human serum samplesalong with a positive and negative control that were freeze-driedand vialed in an identical manner using the same procedure as

for the proposed reference reagents and sent to each participant.Each laboratory submitted a completed questionnaire describ-ing the details of the assay method and this information is sum-marized in Table 1. A variety of target cells, incubation timesand temperatures, days in culture, end-point markers, and lab-oratory and primary virus strains were used in the various tests.Most laboratories used only one method of end-point determi-nation. In one laboratory, two distinct assays were used, eachwith their own end-point marker, and these results were reportedas two separate data sets.

Viruses

Eight strains of HIV-1 were used in the various NA assays:seven laboratory strains (HIV-lnm, HIV-Imn, HIV-1Sf2. HIV-Irf, HIV-1NY5, HIV-1Z34, and HIV-1Z3) and one clinical iso-late (HIV-1,4499)-

RESULTS

The results of NA tests performed in our laboratory on sera

from the two donors are shown in Table 2. The screening (ini-tial) sera were collected as part of a panel of sera from 10 indi-viduals, and were selected on the basis of having the highesttiters among the group against HlV-lum- The later sera from thetwo individuals were tested for NA against a variety of strainsafter each individual bleed and as a final pooled sample. Thepostpooled titers reported in Table 2 are mean titers derived inour laboratory as the result of two to four tests of each.

Quality control tests were done on the reference reagents dur-ing and after they were aliquoted, lyophilized, and sealed.Testing indicated uniformity of fill with variation <2.5% of fillvolume, moisture content of 2.37 and 3.07%, sterility with re-

spect to aerobic and anaerobic bacteria, no residual oxygen, no

detectable HIV-1 infectivity, and levels of antibodies to HIV-1 and HIV-2 in ELISA similar to the serum pools used for fill-ing. The accelerated stability test showed the samples gelled at56CC after 3 weeks but there was no loss in potency in 8 weeksat 35°C or 5 months at 27°C, which would be consistent withlong-term stability at 4°C. They should be handled as poten-tially infectious as any human serum sample.

Some of the results of the international collaborative study forpotency are summarized in Fig. 1. The results showreasonable consistency between laboratories against the HIV-Ilav/ihb strain. The geometric mean titers (GMTs) of the dupli-cate samples tested were similar in all cases and the GMTs forthe 1:2 dilutions of the references {samples 4 and 9) were abouthalf that for the undiluted specimens in each case. Both refer-ences had neutralizing activity in all participating laboratories.

Table 3 shows the results of tests against the virus strainsthat were tested in more than one laboratory and also a primaryisolate, J4499. The results in Fig. 1 are summarized and in-cluded here for comparison. The results indicate that the posi-tive control serum had higher titers against all five viruses thaneither reference, the blinded positive had higher titers againstthe MN strain, and references 1 and 2 were similar with the ex-

ception of J4499, where reference 2 had higher activity thanreference 1 or the blinded positive sample.

In addition to these five viruses, laboratory E also tested NY5and Z34 and laboratory A tested Z3. The results were similar

Table 2. Pre- and Postpooled SerumNeutralization Titers of Reference Sera

Neutralization literagainst HIV-1 strain

Donor Serum IIIB MN SF2 Z-3

ScreeningPrepool 1Prepool 2Prepool 3PoolScreeningPrepool 1Prepool 2Prepool 3Pool

1:1281:321:64 (32)1:321:261:2561:1281:1281:2561:114

>1024>1024

1:1340

>1024>1024

1:2580

1:5121:5121:5121:512

1:10241:10241:5121:512

1:81:32

1:40

1:321:32

1:64

786 VUJCIC AND QUINNAN

1,000a.LU

í= 300

gr-<N

<rrHLU

A £.^ 100

30

10

Ti

Sample Number

Identity

12 1

PositiveControl

10 2

BlindedPositive

5 7 4

Reference 1

6 8 9

Reference 2

3 11

NegativeControl

FIG. 1. Neutralizing antibody titers obtained by participating laboratories against the HIV-1Lav/iiib strain. Participating labo-ratories are as follows: (•) A; (*) B; (A) C; (O) Dl; ( ) D2; (A) E; ( ) F. The samples designated as negative and positivecontrols in the serum panel have been identified here as samples 11 and 12, respectively. Samples 1 and 2 are 1:2 dilutions ofsamples 12 and 10, respectively. Sample 4 is a 1:2 dilution of reference 1, and sample 9 is a 1:2 dilution of reference 2. For thepositive and negative controls, samples 1 and 3 were blinded and samples 11 and 12 were unblinded. Bars indicate geometricmeans.

because the positive control had higher reactivity than the ref-erences against these additional three strains.

DISCUSSION

We prepared and characterized neutralizing and control anti-HTV-1 reference sera for standardization of neutralizing anti-

body assays. An international collaborative study was con-

ducted to assess the levels of neutralizing antibodies in the ref-erence reagents in different laboratories, each using their own

reagents and assay procedure. Six laboratories participated, us-

ing seven different assays to test the panel of sera. Seven dif-ferent strains of virus were used, including six laboratory strainsand one primary isolate. The results of the collaborative studydemonstrated that the two positive reference sera had neutral-

Table 3. Neutralization Assay Titer of Each Reference Sample Against Several HIV-1 Strains Tested

Virus strain*

SampleIIIB/LAV(N = 7)b

MN RF(N = 4)

SF2(N = 5)

J4499(N = /)

Positivecontrol

Blindedpositive

Reference 1

Reference 2

Negativecontrol

385(160-3840)c

81(25-1448)

49(8-160)

137(40-320)4

8640(3840-23170)

4196(1280-11585)

804(240-1626)

1317(240-4096)4

1087(160-7680)

129(10-1280)63(10-480)84(20-480)

4

5549(3840-8192)

648(240-2048)

329(160-512)

410(160-512)

4

113

14

13

80

4

aGMT reported is a grand mean of all participating laboratories using the HIV strain listed. This GMT takes into con-sideration sample set replicates, dilutions, and multiple testing results in each laboratory. J4499 was tested only byLaboratory F.

bNumber of laboratory procedures used to obtain mean.

cRange of individual laboratory means.

HUMAN HIV-1 NEUTRALIZING REFERENCE SERA 787

izing antibody activity against all the strains tested, indicatingthat they should be useful as reference reagents. The virusstrains used included at least five subtype B strains and two

subtype D strains, indicating relatively broad cross-reactivity.An attempt was not made to assign any specific level of po-tency to either of these reagents, because of the wide variationin titer observed against different strains. However, they shouldbe useful for evaluating test-to-test consistency within a labo-ratory and for facilitating comparison of results from differentlaboratories. They have been included as reference reagents forresearch purposes as part of the WHO AIDS Reagent Program.

The WHO guidelines for preparation of reference sera haverequirements for uniformity of fill, moisture content, microbialcontamination, residual oxygen, and stability.13 Although our

test procedures regarding some of these parameters were notthose specified in the guideline, the test results indicate that thesera have properties consistent with those properties specifiedin the guideline. The overall quality control test results for thesesamples have the appropriate characteristics for these referencesera to be suitable for long-term storage and use.

There are approximately 2500 aliquots of each positive ref-erence serum and approximately 5600 aliquots of the negativereference serum available through the three AIDS repositoriesin the United States, France, and England. Owing to the lim-ited supply, these sera should generally be used for standard-ization to establish each laboratory's own in-house referenceserum.

ACKNOWLEDGMENTS

The authors thank the patients who donated the sera used tomake the reference reagents; Harvey Alter, M.D. and SusanLeitman, M.D. for providing the sera used to prepare the ref-erence reagents and other assistance; Dr. Saladin Osmanov ofthe World Health Organization, who provided valuable infor-mation and suggestions for preparation of these samples for theAIDS Reagent Project; Wendy Glass Ryan, M. ChristineAnderson, Julia Gorman, and Mark Heintzelman for assistancewith preparation and testing of the reference reagents; and thecollaborating laboratories that conducted neutralizing antibodytests.

REFERENCES

1. Quinnan GV: Immunization against viral diseases. In: AntiviralAgents and Viral Diseases of Man (Galasso G, Whitley R, andMerigan TC, eds.), 3rd Ed. Raven Press, New York, 1990, pp.727-769.

2. Vujcic L, Shepp D, Klutch M, Wells M, Hendry M, Wittek A,Krilov L, and Quinnan G: Use of a sensitive neutralization assayto measure the prevalence of antibodies to the human immunode-ficiency virus. J Infect Dis 1988;157:1047-1050.

3. Weiss RA, Clapham PR, Cheingsong-Popov R, Dalgleish AG,Carne CA, Weiler IVD, and Tedder RS: Neutralization of humanT-lymphotropic virus type III by sera of AIDS and AIDS-risk pa-tients. Nature (London) 1985;316:69-72.

4. D'Souza MP, Durda P, Hanson CV, Milman G, and collaboratinginvestigators: Evaluation of monoclonal antibodies to HIV-1 byneutralization and serological assays: An international collabora-tion. AIDS 1991;5:1061-1070.

5. Langlois A, Matthews T, Weinhold K, Chaffee S, Hershfield M,and Bolognesi D: Detection of HIV-1 neutralizing antibodies by a

simple, rapid, colorimetric assay. AIDS Res Hum Retroviruses1988;4:63-69.

6. Albert J, Abrahamsson B, Nagy K, Aurelius E, Gaines H, NystromG, and Fenyö EM: Rapid development of isolate-specific neutral-izing antibodies after primary HIV-1 infection and consequentemergence of virus variants which resist neutralization by autolo-gous sera. AIDS 1990;4:107-112.

7. Golding H, D'Souza MP, Bradac J, Mathieson B, and Fast P:Meeting report: Neutralization of HIV-1. AIDS Res HumRetroviruses 1994;10:633-643.

8. Hanson CV: Measuring vaccine-induced HIV neutralization:Report of a workshop. AIDS Res Hum Retroviruses 1994; 10:645-648.

9. Vujcic L, Katzenstein D, Martin M, Quinnan G, and collaboratinglaboratories: International collaborative study to compare assaysfor antibodies that neutralize human immunodeficiency virus.AIDS Res Hum Retroviruses 1990;6:847-853.

10. D'Souza MP, Geyer SJ, Hanson CV, Hendry RM, Milman G, andcollaborating investigators: Evaluation of monoclonal antibodies toHIV-1 by neutralizing and binding assays; an international collab-oration. AIDS 1994;8:169-181.

11. World Health Organization Global Programme on AIDS: Reportof a WHO workshop on the measurement and significance of neu-

tralizing antibody to HIV and SIV—London, 3-5 October 1988.AIDS 1990;4:269-275.

12. WHO Expert Committee on Biological Standardization: Guidelinesfor the preparation and establishment of reference materials andreference reagents for biological substances. Twenty-ninth report.World Health Organization Technical Report Series, No. 626,1978.

13. WHO Expert Committee on Biological Standardization: Proposedrevised guidelines for the preparation, characterization and estab-lishment of international and other standards and reference reagentsfor biological substances. Geneva, October 24-31, 1989.

14. Campbell P: International biological standards and reference prepa-rations. I. Preparation and presentation of materials to serve as stan-dards and reference preparations. J Biol Standard 1974;2:249-258.

15. Campbell P: International biological standards and reference prepa-rations. II. Procedures used for the production of biological stan-dards and reference preparations. J Biol Standard 1974;2:259-267.

16. Specifications for the quality control of pharmaceutical prepara-tions. In: International Pharmacopoeia, Vol. 1, 3rd Ed. WorldHealth Organization, Geneva, 1979, p. 134.

17. Kirkwood TBL: Predicting the stability of biological standards andproducts. Biometrics 1977;33:736-742.

18. Kirkwood TBL and Tydeman MS: Design and analysis of accel-erated degradation tests for the stability of biological standards II.A flexible computer program for data analysis. J Biol Standard1984;12:207-214.

19. Kirkwood TBL: Design and analysis of accelerated degradationtests for the stability of biological standards. III. Principles of de-sign. J Biol Standard 1984;12:215-224.

20. Sawyer LA, Carrow E, Snoy PJ, Hewlett IK, Quinnan GV, andAnand R: Variable infections with and serological responses of rab-bits to five different HIV-1 strains, including one neural tissue iso-late. Microbios 1992;71:243-255.

Address reprint requests to:Luba K. Vujcic

Center for Biologies Evaluation and ResearchFood and Drug Administration

1401 Rockville PikeRockville, Maryland

20852-1448