HIV-1 gp41 envelope IgA is frequently elicited after transmission but has an initial short response half-life.

0.45 MB·English

by  Yates, N L

ARTICLES
naturepublishinggroup
HIV-1 gp41 envelope IgA is frequently elicited after
transmission but has an initial short response half-life
NLYates1,2,ARStacey3,TLNolen4,NAVandergrift1,2,MAMoody1,5,DCMontefiori1,6,KJWeinhold1,6,7,
WA Blattner8, P Borrow3, R Shattock9, MS Cohen10, BF Haynes1,2,7 and GD Tomaras1,6,7,11
Prevention of HIV-1 transmission at mucosal surfaces will likely require durable pre-existing mucosal anti-HIV-1
antibodies (Abs). Defining the ontogeny, specificities and potentially protective nature of the initial mucosal virus-
specificB-cellresponsewillbecriticalforunderstandinghowtoinduceprotectiveAbresponsesbyvaccination.Genital
fluidsfrompatientswithintheearlieststagesofacuteHIV-1infection(FiebigI–VI)wereexaminedformultipleanti-HIV
specificities.Gp41(butnotgp120)Envimmunoglobulin(Ig)AAbswerefrequentlyelicitedinbothplasmaandmucosal
fluidswithinthefirstweeksoftransmission.However,shortlyafterinduction,theseinitialmucosalgp41EnvIgAAbs
rapidly declined with a t of B2.7 days. B-cell-activating factor belonging to the TNF family (BAFF) was elevated
1/2
immediately preceding the appearance of gp41 Abs, likely contributing to an initial T-independent Ab response.
HIV-1transmissionfrequentlyelicitsmucosalHIV-1envelope-specificIgAresponsestargetedtogp41thathaveashort
half-life.
INTRODUCTION during infection.11,12 Thus, T cell help for maturation of the
Immunoglobulin (Ig)A antibodies (Abs) are a first line of B-cellresponseislikelylimitedduringtheearlystagesofHIV-1
immune protection for many viral infections at mucosal infection. HIV-1 also specifically causes deficiencies in B-cell
surfaces. The mechanisms by which mucosal IgA is induced function,13–15 and viral cofactors such as the viral envelope
and how innate immune recognition of pathogenic micro- glycoprotein gp120 (refs 16 and 17) and Nef18,19 have been
organismshavearoleinIgclassswitchrecombinationarean proposedtomediateB-celldysfunctioninHIV-1infection.A
area of intensive study that ultimately will inform HIV-1 cytokine storm early in AHI20 may also account for some
vaccinedesign1.Thereisaresurgenceofinterestinthepotential aspects of B-cell dysfunction.
roleofpre-existingmucosalIgAAbsinprotectionagainstHIV-1. There is a narrow window of vulnerability after virus
Recent studies in non-human primates indicate that Abs to exposurethatcouldallowAbswithantiviralfunctiontoinhibit
gp41(ref.2)andinparticularIgAAbstogp41(ref.3),ifpre- HIV-1 at mucosal surfaces.21,22 Pre-existing HIV-1-specific
existing, may have important functional roles in protection mucosal Abs present at the time of transmission could block
from infection. In acute HIV-1 infection (AHI), humoral HIV-1acquisition.ThemechanismsbywhichAbscaninhibit
responsesarisetoolatetobeeffectiveinthecontrolofHIV-1 HIV-1 movement across the mucosal barrier include direct
replication,4 and there is significant impairment of the B-cell virus neutralization, viral aggregation, inhibition of transcy-
response(reviewedinrefs5and6).HIV-1transmissionresults tosis, intra-epithelial neutralization, phagocytosis, inhibition
in rapid turnover of CD4þ T lymphocytes7 and specific throughmucus,andFcreceptor-mediatedneutralization(Ab-
depletionofCCR5þ CD4þ Tcellsinthegut.8–10Moreover, dependent cellular cytotoxicity) (reviewed in refs 23 and 24).
HIV-specificCD4þ Tcellsarepreferentiallyinfectedandlost AnalysesoftheinitialmucosalAbresponsestoHIV-1infection
1DukeHumanVaccineInstitute,DukeUniversity,Durham,NorthCarolina,USA.2DepartmentofMedicine,DukeUniversity,Durham,NorthCarolina,USA.3Nuffield
DepartmentofClinicalMedicine,UniversityofOxford,Oxford,UK.4ResearchTriangleInstitute,ResearchTrianglePark,NorthCarolina,USA.5DepartmentofPediatrics,Duke
University,Durham,NorthCarolina,USA.6DepartmentofSurgery,DukeUniversity,Durham,NorthCarolina,USA.7DepartmentofImmunology,DukeUniversity,Durham,
NorthCarolina,USA.8DepartmentofMedicine,InstituteofHumanVirologyEpidemiologyDivision,UniversityofMaryland,Baltimore,Maryland,USA.9Departmentof
Medicine,ImperialCollege,London,UK.10InstituteforGlobalHealthandInfectiousDiseases,UniversityofNorthCarolina,ChapelHill,NorthCarolina,USAand11Department
ofMolecularGeneticsandMicrobiology,DukeUniversity,Durham,NorthCarolina,USA.Correspondence:GDTomaras([email protected])
Received2March2012;accepted8September2012;publishedonline9January2013.doi:10.1038/mi.2012.107
692 VOLUME6NUMBER4|JULY2013|www.nature.com/mi
ARTICLES
will enable a more complete understanding of the effects of TodeterminewhetherotherHIV-1proteins,inadditionto
HIV-1 ontheB-cellarmoftheimmuneresponse.Moreover, the HIV-1 Env, stimulated a humoral response in acute
mucosal Abs that are easily elicited by HIV-1 and target the infection, we assessed Ab responses to Gag, p66 RT, p31
HIV-1 Env may be a potential target for vaccine design if Integrase, Tat, and Nef. The majority of patients developed
thedurabilityoftheresponsecanbeimproved.Antibodiesthat systemicandmucosalIgGandIgAresponsestoGagandp66
areeasytoelicitanddonotrequirethesameextensivelevelof RT,whereasIgGandIgAAbresponsestop31Integrase,Tat,
somatic hypermutation thought to be required for broadly and Nef were less commonly detected by B133 days post
neutralizing Abs24 would, if protective, be ideal as targets for enrollment (Figure 1b and d).
HIV-1 vaccine immunogen design. Thus, understanding the ToevaluatethetimingoftheappearanceofHIV-specificIgA
propertiesoftheinitialIgAresponseinthemucosafollowing duringAHI,wedeterminedthefrequencyofplasma(Table1a)
HIV-1 transmission should enable a better understanding of andmucosal(Table1b)IgAspecificitiesbyFiebigstage.25We
how to best elicit durable humoral immunity for preventing evaluated plasma samples from 30 Center for HIV/AIDS
HIV-1 acquisition and/or virus replication. Vaccine Immunology (CHAVI) 001 subjects (most were
Here we report that IgA Abs to gp41 Env arise initially in enrolledduringstage3orlater)and25plasmadonors(Fiebig
acute infection and are frequently detected in genital fluids. stagesI/IIonwarduntilsamplecollectionstopped).Inplasma,
MucosalIgAAbstogp120EnvandtootherpartsofHIV-1arise gp41 IgA and Gag IgA were present in 26.1% and 8.7% of
laterduringacuteinfection.Moreover,weexaminedpotential subjects during Fiebig stages I/II, respectively. RT- and Nef-
regulators of the initial Ab response and found that B-cell- specific IgA were not detected in samples obtained during
activatingfactorbelongingtotheTNFfamily(B-cell-activating FiebigstagesI/IIandweremorecommonlydetectedinsamples
factor(BAFF))waselevatedprecedingthespikeinanti-HIV-1 fromstagesIV–VI.HIVgp120-andp31-specificIgAwereless
Abs.ThisearlyBAFFelevationcoupledwiththelackofarobust frequentevenatlaterstagesV/VI.Inmucosalsamplesfrom23
and sustained mucosal HIV-1-specific IgA response is con- CHAVI001subjects,weobservedthatgp41,Gag-,andgp120-
sistent with an initial HIV-1-specific Ab response that is T specificIgAwerefirstdetectedduringFiebigstageIVandwere
independent. Thus, the initial mucosal Env Ab response is morefrequentlydetectedduringFiebigstagesV/VI.RT-,Nef-,
limitedinspecificitytogp41duringtheearlystagesofinfection and p31-specific IgA were only detected in samples obtained
and has a short half-life. duringFiebigstagesV/VI.IntheHIV-1seroconversionplasma
donorcohort,wealsoexaminedthetimesinceT (plasmaviral
0
RESULTS RNA¼100 copies/ml)4,20,26 for the appearance of IgA Abs.
Frequency, specificity and timing of anti-HIV-1 mucosal During Fiebig stages I–VI, in those plasma donors with
Abs during AHI detectable HIV-1-specific plasma IgA, the median time to
TodetermineifmucosalAbstotheHIV-1envelopearisewithin appearanceofgp41-specificIgAwas13.5dayspostT (rangeof
0
thefirstfewmonthsofAHIandwhethermucosalAbstoother 9–18 days), whereas the median time for the appearance of
regions of HIV-1 are elicited during AHI, we examined the Gag-specificIgAwas25.5dayspostT (rangeof14–40days)
0
frequencyofdetectionofIgAandIgGresponsestotheHIV-1 (Table1c).Gp120IgAAbresponseswerenotdetectedinthese
envelope proteins, gp41 and gp120 (Figure 1a and c); and to samplesduetotheearlyFiebigstaginginthiscohort,consistent
HIV-1 antigens Gag, Tat, Nef, Integrase (p31), and reverse with the lack of early IgG gp120 responses we previously
transcriptase (RT) (Figure 1b and d) in up to 30 patients reported.4 Taken together, these data show that HIV-specific
(N¼170samples)forplasmaandupto23patients(N¼119 IgAisfrequentlypresentinthegenitaltract,aswellas,present
samples) for mucosal samples (Supplementary Table S1 systemically during AHI. The initial Ab specificities include
online). HIV-1 gp41-specific IgA was detected in the genital predominantlytheimmunodominantepitopeingp41,whereas
secretions of 20/23 (87%) acutely infected individuals and Abs to gp120 and other HIV-1 proteins appear later in some
plasma gp41-specific IgA was present in 29/30 (97%) of AHI individuals.
patients,allofwhomwerepositiveforHIV-1gp41IgGinboth
plasmaandgenitalsecretions.PlasmaHIV-1gp41-specificIgM HIV-1-specific mucosal Ab concentrations
wasfrequentlydetectedin25/28(89%)patients;however,only To examine the concentrations of HIV-1-specific IgA
about half of acute patients had detectable mucosal gp41- responses in both plasma and mucosal samples following
specific IgM (11/21, 52%). Additionally, the vast majority of HIV-1transmission,weevaluatedHIV-1-specificAbresponses
patientshad gp41 IgG andIgA Abs that werespecific for the intheCHAVI001AHIcohortthatwereenrolledduringAHI
immunodominantepitope(Figure1c);whereas,therewereno andfollowedforupto133daysafterenrollment.6,27Mucosal
detectable membrane proximal external region (MPER) Ab concentrations were measured in 64 samples from 12
epitope-specific responses above 1ng/ml of mucosal fluid. patients (seminal plasma from ninemales and cervicovaginal
Anti-gp120 IgA Abs were detected at later stages of acute lavage from three females) to quantify HIV-1-specific Ab
infection in plasma and mucosal samples and they were less concentrations relative to total (non-specific) Ab concentra-
frequentlydetectedthangp41Env-specificAbs.Mucosalgp120 tions. During AHI, HIV-1-specific IgG and IgA were both
IgA was present in 27% of AHI patients, and mucosal gp120 frequently detected in systemic and mucosal compartments,
IgG was present in 40% of AHI patients. and the observedpeak concentrations of anti-gp41 Abs (IgG,
MucosalImmunology|VOLUME6NUMBER4|JULY2013 693
ARTICLES
a b
Plasma Plasma
100 100
IgG
IgA IgG
80 IgM 80 IgA
IgM
60 60
40 40
20 20
e
sitiv 0 gp41 gp120 0 Gag Tat Nef p31 RT
o
p
nt c d
ce Mucosal Mucosal
er 100 100
P
IgG IgG
IgA IgA
80 IgM 80 IgM
60 60
40 40
20 20
nt
0 0
gp41 Immuno- gp120 Gag Tat Nef p31 RT
dominant
gp41 HIV antigen
Figure1 Frequentdetectionofanti-gp41mucosalimmunoglobulin(Ig)A,butinfrequentdetectionofmucosalanti-gp120IgAwithin133days
fromenrollment.FrequencyofHIV-specificantibodiesin(a,b)plasmain14–30CHAVI001subjectsand(c,d)mucosalsitesin13–23
CHAVI001subjectswithin133daysofenrollment.IgGandIgAspecificfortheimmunodominantepitopeofgp41werealsotestedinmucosal
samplesfrom12individuals.nt,nottested.
IgA,andIgM)(Table2)weremeasuredduringFiebigstageIII of HIV-1 Env-specific IgA in genital fluids during acute
throughV/VI. Onan average, therewereovertwo logsmore infection were low compared with HIV-1 Env-specific IgG.
HIV-1gp41-specificIgGintheplasmacomparedwithIgAat
thepeakoftheresponseduringAHI.Thiswaspartlyreflective Differences in systemic and mucosal IgA and IgG
ofthedifferenceinthetotalnon-HIV-specificlevelsofIgG/IgA concentrations and specificities
thatispresentinplasma(mean43.8mg/mlofIgGvs.4.9±mg/ml ToexaminewhethertheHIV-1-specificmucosalAbresponse
IgA, an 8.9-fold difference). In the genital secretions, was likely derived locally or transudated from plasma, we
gp41-specific IgG was on an average 11-fold higher than comparedthelevelsofHIV-specificAbsinplasmaandgenital
gp41-specificIgA(Table2).However,totalIgGandIgAlevels tractcompartmentsacrossalltimepointswithinthefirst133
within the mucosal samples that we examined were more days post enrollment. We first compared the antigen-specific
similartooneanother(168±STDEV209.5mg/mloftotalIgG Abresponsesinplasmatothoseinmucosalsamples(Figure2a
vs. 133.2±STDEV 278.2mg/ml of total IgA); consistent with andb).Ofallcomparisons,systemicandmucosalgp41-specific
normalAbranges28andconsistentwithpreviousreportsthat IgG most strongly correlated to one other (Figure 2a), in
demonstratedthat,unlikeintestinalfluidcontaininghighlevels contrasttogp41-specificIgA(Figure2b)thatcorrelatedweakly
ofSIgA,semenandcervicovaginalfluidcontainmoreIgGthan (R¼0.322,P¼0.040),suggestingthatsomeIgAmaybelocally
IgA.29Asexpected,inmucosalsecretionstotalIgMlevelswere producedinthemucosalcompartment.WethenexaminedIgG
lower than total IgG and IgA. In those samples with HIV-1- vs. IgA in plasma and mucosal samples (Figure 2c and d) to
specificIgM,thespecificactivitylevelsarehigherthantheIgG determine whether any correlations existed between IgG and
andIgAmeasurements,inpart,duetothelowerconcentrations IgAlevelsineithercompartment.PlasmaHIV-1gp41EnvIgG
oftotalIgM(mean10.2mg/ml±STDEV11.7mg/ml)thatwere and IgA correlated (R¼0.532, Po0.0001) (Figure 2c).
present. In addition, there were no significant differences in However, mucosal gp41 Env-specific IgG and IgA
total or gp41-specific IgG and IgA in plasma or genital fluid responses correlate weakly (Figure 2d) (R¼0.366 and
collections between males and females from this study R¼0.017). As noted by the circles, Ab concentrations in
(Supplementary Figure S1 online). Taken together, the levels numeroussampleslinedoneaxisortheother(notonanx¼y
694 VOLUME6NUMBER4|JULY2013|www.nature.com/mi
ARTICLES
Table1 Timingofsystemicand mucosalgp41-specific IgAantibodies
gp41 Gag RT Nef gp120 P31
Fiebigstage No./totala % No./total % No./total % No./total % No./total % No./total %
(a)PlasmaIgA
I/II 6/23 26.1 2/23 8.7 NT — NT — NT — NT —
III 7/14 50.0 3/14 21.4 NT — NT — NT — NT —
IV 15/17 88.2 10/17 58.8 3/7 42.9 3/7 42.9 2/8 25.0 0/7 0.0
V/VI 39/40 97.5 35/40 87.5 13/15 86.7 8/14 57.1 11/29 37.9 2/14 14.3
(b)MucosalIgA
I/II 0/1 0 0/1 0 NT — NT — NT — NT —
III 0/2 0 0/2 0 NT — NT — NT — NT —
IV 7/21 33.3 4/8 50.0 0/7 0 0/7 0 1/7 14.3 0/7 0
V/VI 18/23 78.3 17/23 73.9 5/14 35.7 1/14 7.1 4/14 28.6 1/14 7.1
(c)
Antigen No.ofsubjectsb Mediantimetoantibodyresponse(range(days))c
gp41 14 13.5(9–18)
p55 6 25.5(14–40)
Abbreviations:Ig,immunoglobulin;NT,nottested.
(a)FrequencyofHIV-1-specificplasmaIgAresponsesperFiebig25stageinupto30CHAVIand25plasmadonorpatientsisshown.Frequencywascalculatedbythenumberof
plasmasamplespositivepereachFiebigstageoverthetotalnumberplasmasamplesmeasuredinthatstage.(b)FrequencyofHIV-specificmucosalIgAresponseswere
measuredinmucosalsamplesperFiebigstagefromupto23acuteHIV-1infection(AHI)patients.(c)SystemicHIV-1gp41-specificIgAappearsbeforeanti-GagIgAantibodies.
HIV-specificIgAresponsesweremeasuredin25HIVþseroconversionplasmadonorpanelsbystandardenzyme-linkedimmunosorbentassayandalignedtoT (thetimeat
0
whichviralloadisfirstdetectable),aspreviouslydescribed.
aNumberofsubjectshavinganantigen-specificIgApositivesampleinthedesignatedFiebigstagedividedbythetotalnumberofsubjectsinthatdesignatedFiebigstage.
bNumberofplasmadonorsubjectswithpositiveIgAresponse.
cMediantimefromT (thefirstdaytheviralloadreaches100copies/ml).
0
Table2 PeakconcentrationofHIV-1gp41-specificantibodiesinplasmaandmucosalsamplesobservedduringFiebigstageIII
through V/VIofAHI
Plasmagp41-specificAb(lg/ml) Mucosalgp41-specificAb(lg/mg)
Numberpositive(n) Median(range) Mean±s.d. Numberpositive(n) Median(range) Mean±s.d.
IgG
All 23 546.3(5.4–5194.8) 1242.4±1485.8 12 35.6(2.4–444.9) 102.7±133.7
Male 16 1300.1(5.4–5194.8) 1618.8±1632.8 9 32.5(2.4–444.9) 111.2±151.8
Female 7 353.5(128.8–1761.1) 555.4±554.8 3 72.1(10.6–148.1) 76.0±68.9
IgA
All 23 12.5(0.7–41.2) 12.5±9.4 12 4.4(0.4–52.4) 9.2±14.2
Male 16 13.0(0.7–41.2) 14.4±10.3 9 9.0(0.7–52.4) 11.6±15.7
Female 7 4.8(1.2–13.0) 7.0±4.6 3 0.8(0.4–3.2) 1.4±1.5
IgM
All 12 7.9(0.3–26.7) 9.8±8.0 5 26,939(15,110–112,989) 40,719±40,780
Male 10 6.7(0.3–13.7) 6.9±4.2 5 26,939(15,110–112,989) 40,719±40,780
Female 2 24.8(22.9–26.7) 24.8±2.7 0 — —
Abbreviations:Ab,antibody;AHI,acuteHIV-1infection;Ig,immunoglobulin.
PeakplasmaandmucosalHIV-1gp41-specificantibodylevelsweredeterminedinCHAVI001patientsduringthefirst133dayspostenrollment.Concentrationisexpressedin
mg/mlequivalentsdeterminedwithastandardcurveusing2F5IgG,IgA,orIgM.Mucosalantibodylevelsareexpressedasspecificactivity(mgofgp41-specificantibody/mgof
totalantibody).
MucosalImmunology|VOLUME6NUMBER4|JULY2013 695
ARTICLES
a b
gp41 IgG gp41 IgA
500 15
R=0.697 R=0.322
cificmg) 400 P<0.0001 cificmg) 10 P=0.040
Mucosal spe(cid:2)activity (g/ 123000000 Mucosal spe(cid:2)Activity (g/ 5
0 0
0 2000 4000 6000 0 10 20 30
Plasma ((cid:2)g/ml) Plasma ((cid:2)g/ml)
c Plasma gp41 d Mucosal gp41
6000 500
R=0.532 R=0.366
P<0.0001 g) 400 P=0.017
(cid:2)IgG (g/ml) 42000000 IgG specific(cid:2)ctivity (g/m 320000
a 100
0 0
0 10 20 30 0 5 10 15
IgA ((cid:2)g/ml) IgA specific activity ((cid:2)g/mg)
Figure2 Differenceinsystemicandmucosalimmunoglobulin(Ig)AandIgGconcentrationsandspecificities.Correlationsofgp41Env-specific
(a)IgGand(b)IgAingenitalsecretionsvs.plasmaduringacuteHIV-1infection(AHI).Correlationsof(c)plasmaand(d)mucosallevelsofHIV-1
Env-specificIgGvs.IgAduringAHI.SpecificactivityinmucosalsamplesiscalculatedasmgofHIV-specificantibody(Ab)permgtotalAb.
Concentrationsofanti-gp41Abcalculatedper2F5monoclonalAbequivalentsandconcentrationsofanti-gp120Absarecalculatedasthebindingunits
inthelinearrangemultipliedbythedilutionfactordividedbytotalAbconcentration.
line). These data indicate that some HIV-1-specific Ab Weappliedanexponentialdecaymodel31todeterminethe
responses present at mucosal sites represent transudate. Ab half-life of gp41-specific IgA in the plasma and mucosal
However, the discordance between mucosal and systemic compartments during AHI among individuals with at least a
HIV-specificAbssuggeststhatsomefractionofthemeasured 2-folddecreaseinAbresponse(Table3).Themodelthatfitbest
HIV-1-specific mucosal response may be locally produced. for the mucosal samples assumes a lower asymptote greater
Thus,measurementofplasmaHIV-1EnvIgAdoesnotentirely thanzero(Abresponsesplateausatanon-zerolevel),whereas
reflect the level or specificity of mucosal HIV-1 Env IgA. the model for the plasma samples assumes that the lower
asymptoteiszero(Abresponseeventuallydeclinesalltheway
Short half-life of the mucosal anti-gp41 Env IgA in AHI tozero).Althoughthehalf-lifeofplasmagp41-specificIgAwas
To address the question of whether the initial mucosal Ab much longer (48.19 days (95% confidence interval
responsetoHIV-1infectionistransientandthereforemayhave (CI)¼34.57–61.81)) than the half-life of mucosal IgA (2.71
been difficult to detect previously30, we examined 12CHAVI days (95% CI¼2.06–3.36)), the fold decline (the delta from
001 patients longitudinally (out to 133 days post enrollment) peak to nadir) of HIV-1-specific IgA was similar in mucosal
(Figure3)todeterminethekineticsoftheHIV-1-specificIgA (6.20-fold (95% CI¼ (cid:2)0.51, 12.92) and plasma (8.65-fold
and IgG responses in both plasma (Figure 3a and c) and (95% CI¼3.38–13.93) samples.
mucosal compartments (Figure 3b and d). To normalize for
changes in total Ab concentrations, specific activity (HIV-1 Confirmation of short half-life of anti-gp41 Env IgA in
gp41-specific Ab/total Ab) was determined for each mucosal plasma in additional AHI cohorts
sample.AlthoughmucosalHIV-1-specificIgAresponseswere To determine whether the overall systemic gp41-specific IgA
detected frequently in AHI, there was an early peak and declineintheCHAVI001patientswascommontootheracute
subsequentdeclineduringthelaterphaseofacuteinfectionin infection cohorts, we examined additional patients for whom
11 out of 12 patients (91.7%). This was in contrast to the wehadlongitudinalsamplesavailableforstudyintheCHAVI
predominantly increasing or steady mucosal gp41 IgG 001cohortand,aswell,comparedAbkineticstothatobserved
response. Likewise, of the 12 patients that we studied with intwoothercohorts:theplasmadonorcohort4andTrinidad
matching longitudinal plasma and mucosal samples, 10 cohort32,33.Insequentialplasmasamplesfrom44patients,we
(83.3%) of these patients had declining gp41-specific IgA in foundthat28patientsdemonstrateddiscordantgp41-specific
the plasma (Figure 3e). IgA and IgG responses (Table 4), in that, gp41-specific IgA
696 VOLUME6NUMBER4|JULY2013|www.nature.com/mi
ARTICLES
a Plasma b Mucosal
10,000 6 100,000
n 703-01-013-1 Plasma 703-01-013-1 Seminal Plasma
o
ntibody concentrati(cid:2)(g/ml)1,010100100 IIIgggGAM 345 Log viral load Specific activity(cid:2)(g/mg)101,,01001000000 IIIgggGAM
A VL
0.1 1
0 20 40 60 80 100 0 20 40 60 80 100 120 140
Days post enrollment Days post enrollment
c 6 d 1,000
703-01-068-1 Plasma 703-01-068-1 CVL
on 1,000 y 100
ody concentrati(cid:2)(g/ml) 11000 IIVggLGA 54 Log viral load Specific activit(cid:2)(g/mg) 01.101 IIggGA
b
nti
A 1 3 0.01
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
Days post enrollment Days post enrollment
e 5
d 4
e
m 3
nsformg) 12
ag/
al-log trgp41 (µ ––210
ur –3
at
N –4
–5
1 51 101 151 201 251 301
Day
Figure3 RapiddeclineinmucosalHIV-specificimmunoglobulin(Ig)AinacuteHIV-1infectionsubjects.HIV-1-specificIgGandIgAantibody
concentrationsareshownfortworepresentativesubjectswith(a,c)pairedplasmaand(b,d)genitalfluid(seminalplasmaandcervicovaginallavage
(CVL).(a)Plasmaand(b)mucosalHIV-specificIgMisshownforonesubject.(e)MucosalHIVgp41-specificIgAkineticsin11patientsalignedtothepeak
response.
Table3 Half-lifeestimatesforinitialgp41IgAdeclineinAHI Table4 Declineinsystemicgp41-specificIgA inmultiple
(exponentialdecaymodel)forindividualswithatleasta2-fold cohorts
decrease inantibodyresponse
No.ofsubjects
HIV-1gp41IgA Cohort No.Samplespersubjecta IgGþ/IgAþb IgAkc
Half-life(days) Folddecreasea
Plasmadonor 4–29 14 6
Estimate 95%CI Estimate 95%CI Trinidad 8–17 7 6
Plasma(N¼10) 48.19 34.57,61.81 8.65 3.38,13.93 CHAVI001 4–10 23 16
Mucosal(N¼11) 2.71 2.06,3.36 6.20 (cid:2)0.51,12.92 Total 44 28
AHI,acuteHIV-1infection;CI,confidenceinterval;Ig,immunoglobulin. Abbreviation:Ig,immunoglobulin.
Themodelthatfitbestforthemucosalsamplesassumesalowerasymptotegreater aThenumberoflongitudinalsamplespersubjectanalyzedoverthecourseofacute
thanzero(antibodyresponseplateausatanon-zerolevel),whereasthemodelforthe infectionareindicated.
plasmasamplesassumesthatthelowerasymptoteiszero(antibodyresponse bThenumberofsubjectswithatleastfourtimepointsthatweretestedthathad
eventuallydeclinesallthewaytozero). detectableanti-gp41IgGandIgAtomeasuredeclineareindicated.
aAtlowerasymptoteformucosaland150dayspostpeakforplasma. cNumberofsubjectswithincreasing/stableIgGandZ2.5-folddeclineinIgAfrom
peaklevel.
responsesdeclinedwhereasgp41-specificIgGresponseseither Figure4,wheretheHIV-1-specificIgGresponseappearsata
remainedstableorincreasedovertime.ThekineticsofHIV-1- median time of 13 days post T (ref. 4) and continues to rise
0
specificIgGandIgAintheplasmadonorcohortareshownin whereas gp41-specific IgA declined. These data demonstrate
MucosalImmunology|VOLUME6NUMBER4|JULY2013 697
ARTICLES
a b
3.0 1.2
gp41 IgG gp41 IgA
2.5 1.0
2.0 0.8
D 1.5 D 0.6
O O
1.0 0.4
0.5 0.2
0.0 0.0
–80 –60 –40 –20 0 20 40 60 80 –80 –60 –40 –20 0 20 40 60 80
Days post T0 Days post T0
Figure4 Initialdeclineinanti-gp41immunoglobulin(Ig)A,butnotanti-gp41IgGduringacuteHIV-1infection.Differentialkineticsofsystemic
HIVgp41-specificIgAandIgGantibodiesfollowingHIV-1transmissioninplasmadonorsubjectsalignedbyT (plasmaviralRNA¼100copies/ml)
0
inacuteinfection.OD,opticaldensity.
a that the HIV-1 gp41-specific IgA decline that we observedin
Percent change in gp41 IgA levels
theCHAVI001cohortcouldbeconfirmedinadditionalAHI
140
700-01-023-8 gp41
ak leveleak)110200 777700000003----00001111----000000118153----1901 ggggpppp44441111 cSohhoorrttsh.alf-life of IgA Abs in AHI is specific to HIV-1 IgA
ercent of pe(at gp41 p 468000 777000003---000111---000474705---005 gggppp444111 TssuphgeecgidefsiectceldiIngetAhianþtgHpB4IV1c-e-s1lplsed.coiTfeioscIngdoAettetsrhtmiamtinwueleatowebhsloeetrnhvgee-rdlivdtehudirsinHwgIVaAs-H1a-I
P 20 generalized downregulation of all IgA responses, we also
0 examinedmemoryIgAresponsesspecifictotetanustoxoidand
influenza antigens from vaccination (Figure 5a–c). Tetanus
0 20 40 60 80 100
toxoidandinfluenzaIgAAbresponsescouldnotbefittothe
Days post enrollment
b Percent change in Tetanus toxoid IgA levels exponential decay model and did not decline during AHI
duringthetimeofgp41IgAAbdecline.Moreover,theaverage
140
700-01-023-8 TT
700-01-008-1 TT levels of total plasma IgA and IgG within this cohort did not
el 120 700-01-001-9 TT
ak leveak)100 777000030---000111---000114537---010 TTTTTT dinectlhineedIguArinrgesapcuontesienfiesctsipoenc.iTfihcesfoerdaHtaIVim-1plyanthdatitshendoetcldinuee
ercent of pe(at gp41 p 846000 770003--0011--007405--05 TTTT trlioaktehaleyrgd,etunheeertosahlEodnrotvwhsatnilmrfe-lguiuflealtaoitofionHnoIVfoa-f1sa-hslolpremtc-ileifmviceodIgrByA-cIingeAllArHerseIpsiposonmnsesoetrsoe;
P HIV-1.
20
0 Specific elevations in BAFF before initial HIV-1 Env Ab
0 20 40 60 80 100
ToinvestigatepotentialmechanismsofHIV-1antigen-specific
Days post enrollment
c Percent change in Fluzone IgA levels IgAdecline,wemeasuredtheplasmaandmucosallevelsoftwo
cytokines,BAFFandAPRIL(aproliferationinducingligand),
140
700-01-023-8 flu
700-01-008-1 flu known to regulate B-cell survival and activation. We first
ak leveleak)110200 777700000030----00001111----000001141537----9010 fffflllluuuu mfoueansdurthedatlBonAgFiFtuldeivnealsltsyapmicpallelsyfrreommaisneerdonsteagbalteivuenddoern1ornsga/mndl
ep 700-01-070-0 flu
ercent of p(at gp41 6800 703-01-045-5 flu o6h0vepe–ra1t2tii0tmisdeaB(yNsv)i.¼ruWs7e(dHtohBneoVnr)se,(xN1a0m¼–i1n12e0dt)impanleadspmaocaiunsttaesmehapecplheas,tifftrrioosmmCuavcpiruuttoes
P 40 (HCV)(N¼10)infectionandfoundthatlevelsofBAFFwere
alsostableduringtheacutephaseoftheseinfections,remaining
20
0 20 40 60 80 100 atlevelssimilartothosefoundinuninfectedpatientswhereas
Days post enrollment viremia increased (Figure 6a). We then examined levels of
Figure5 Immunoglobulin(Ig)AdeclineisspecificforHIV-1infection. plasmaBAFFinAHI(N¼20)andfoundthattherewasapeak
DuringacuteinfectionHIV-1gp41-specificantibodies(a)declineover in the levels of BAFF, above 1ng/ml, soon after HIV-1
time,incontrasttoIgAantibodiesspecificfor(b)tetanustoxoidand(c)
transmissionthatprecededthepeakintheHIV-1-specificAb
influenza(Fluzone).Allantibodyresponsesareplottedasapercentageof
levelsmeasuredatthepeakofthegp41-specificIgAresponse. response(Figure6b).ElevationsinBAFFwerealsoobservedin
698 VOLUME6NUMBER4|JULY2013|www.nature.com/mi
ARTICLES
a Acute hepatitis B plasma Acute hepatitis C plasma
10,000 7 10,000 8
BAFF BAFF
VL 6 VL 7
FF (pg/ml) 1,000 45 g viral load FF (pg/ml) 1,000 56 g viral load
A o A 4 o
B L B L
3
3
100 2 100 2
–20 –10 0 10 20 30 40 –20 –10 0 10 20 30 40
Days post T200 Days post T600
b Acute HIV-1 plasma Acute HIV-1 plasma
1.8 100,000 0.6 10,000
O.D.) 1.6 ggpp4411 IIggGA O.D.) 0.5
HIV gp41 antibody ELISA ( 00000111........02468024 BAAPFRFIL 1110,000,00000 BAFF/APRIL pg/ml HIV gp41 antibody ELISA ( 00000.....01234 ggBAppAP44FR11FI LII gg GA 11,00000 BAFF/APRIL pg/ml
–30 –20 –10 0 10 20 30 –20 –15 –10 –5 0 5 10 15 20
Days post T Days post T
0 0
c Acute HIV-1 cervicovaginal lavage Acute HIV-1 seminal plasma
1,000,000
100 10,000
µ1-specific Ab (g/ml) 0.010.1011 ggBAppAP44FR11FI LIIggGA 110,0000 BAFF/APRIL pg/ml µ41-specific Ab (g/ml) 00.10.1011 ggBAppAP44FR11FI LII gg AG 11100,0,000,000000 BAFF/APRIL pg/ml
p4 0.001 gp
g
0.0001 10 0.001 100
0 20 40 60 80 100 0 20 40 60 80 100 120
Days post enrollment Days post enrollment
d Plasma e Plasma
4,000 25,000
R = 0.32 R= 0.085
P = 0.0003 P = 0.358
20,000
3,000
ml) ml)
g/ g/ 15,000
F (p 2,000 L (p
F RI 10,000
BA 1,000 AP
5,000
0 0
101 102 103 104 105 106 107 101 102 103 104 105 106 107
VL RNA copies/ml
VL RNA copies/ml
Figure6 B-cell-activatingfactor(BAFF)iselevatedbeforeantibodyproductioninacuteHIV-1infection,butisnotelevatedduringacutehepatitisBvirus(HBV)and
hepatitisCvirus(HCV)infections.(a)PlasmapanelsfromsubjectswithacuteHBVinfection(N¼10donors,10–12timepointseach,upto50days))andacuteHCV
infection(N¼10donors,7–15timepointseach,upto50days)weremeasuredforlevelsofBAFF.Datafromonerepresentativesubjectinfectedwitheachvirusare
shown.Sampletimecoursesarealignedrelativetothetimewhenvirusfirstreacheddetectablelevelsinplasma(200DNAcopies/mlforHBV(T ),600RNAcopies/ml
200
forHCV(T )).(b)AnelevationinplasmaBAFFlevelsprecedestheriseinanti-HIVimmunoglobulin(Ig)GandIgAantibodylevelsduringacuteHIV-1infection.(N¼21
600
donors,8–22timepointseachupto40dayspostT weretestedforBAFF;antibodylevelswerealsodetectedin9ofthese;datafromtworepresentative
0
subjectsisshown).(c)ElevationinBAFFlevelsduringthetimeofHIV-specificantibodyelevationsinmucosalsamples.Longitudinalcervicovaginallavage
samples(N¼5patients)andseminalplasma(N¼4patients)weretested(datafromtworepresentativeindividualsisshown).Correlationbetweenplasma(d)BAFF
and(e)APRIL(aproliferationinducingligand)levelswithviralload(VL).Datafrompatient(outlier)arecircledandwhenthesepointsareremoved,R¼0.27,P¼0.0032.
OD,opticaldensity.
MucosalImmunology|VOLUME6NUMBER4|JULY2013 699
ARTICLES
mucosal fluid near the time of the initial HIV-1-specific Ab vaccines; thus, measuring the Ab half-life of HIV-1-specific
response (N¼5 cervicovaginal lavageand N¼4 longitudinal responses systemically and at mucosal sites is important for
seminalplasmasamples)(Figure6c).LevelsofAPRILinthese understanding the mechanisms to induce long-lived effective
samepatientsalsoshowedanincreaseatsimilartimestothatof Abresponses.Here,wefoundthatinacuteinfection,gp41Env-
BAFF, although the levels of APRIL were generally higher specific IgA responses had a very short half-life both
(peakingat2–12ng/ml).ThetransientupregulationofAPRIL systemically and mucosally: B48 days in the blood and 2.7
duringthisperiodwasspecifictoAHI,astherewerenotsimilar daysinthemucosalsecretionsduringtheacutephaseofHIV-1
elevations in APRIL in acute HBV and HCV plasma donor infection.Thehalf-lifeoftotalIgAinthebloodisB4–9days,39
panels at similar time points. To determine whether higher whereasthe predominant IgGsubclasses(IgG1, IgG2) have a
transientlevelsofBAFFcouldbeaconsequenceofhigherT-cell longerhalf-life(21days)40duetotheabilityofIgGAbstobind
activationandIFNgproductionduetoviralreplicationduring to FcRn, preventing lysosomal degradation.41
AHI, we examined whether plasma BAFF and APRIL cor- WepreviouslydemonstratedthatinanHIV-1vaccinetrial
relatedwithviralload(Figure6dande).BAFFlevels,butnot and anti-retroviral-treated chronic HIV-1 infection, the half-
APRIL,wereweaklycorrelatedwithviralload.HIV-1gp41Env lifeofIgGEnvgp120Absisshortcomparedwithtetanusand
IgA Ab levels did not correlate with viral load (R¼ (cid:2)0.140, influenzaAbs,isdependentonspecificantigendrive,andhas
P¼0.174).ThesedatasuggestthatthelevelsofBAFF,APRIL thehallmarkofbeingmaintainedbyshort-livedmemoryBcells
andHIV-1-specificIgAdonotdirectlyreflectincreasedvirus rather than long-lived plasma cells.42 This finding that gp41-
replication and immune activation in these patients. specificmucosalIgAresponseshavearemarkablyshorterhalf-
life (2.7 days) than gp41-specific IgG is extraordinary and
DISCUSSION suggestsisotype-specificrapidcatabolismofmucosalIgA.Also,
InmucosalsamplescollectedwithinweeksofHIV-1transmis- thesedatasuggestthatlong-livedplasmacellssecretingHIV-1-
sion, we show that mucosal IgA HIV-1-specific Abs were specific IgA are not generated immediately following HIV-1
frequentlydetected(87%)inacuteinfection,albeitatrelatively transmission.However,itisimportanttonotethattheHIV-1-
low levels compared with anti-Env IgG. Similar to the initial specificIgAresponsedoesnotdeclinetonegative.Inchronic
plasma IgG response to gp41,4 and the initial plasma B-cell HIV-1infection,Env-specificIgAresponsesarepresentinboth
responsetogp41(ref.34),wedemonstrateherethattheinitial plasma and at mucosal sites such as breastmilk;43 thus, likely
mucosalAbresponseisalsodirectedtogp41Env.Thereason ongoing virus replication can stimulate HIV-1-specific IgA
forthefocusoftheinitialAbresponsetogp41,ratherthanto responses.Furtherstudiesareneededtoquantifyandcompare
gp120, is not fully understood. However, we have previously thelevelsofHIV-1-specificsecretory,monomeric,anddimeric
reported34thattheinitialgp41IgGAbresponsecanbederived IgAinmucosalsecretionsthroughoutinfectionandinresponse
frompre-existingcross-reactivememoryBcellsthatarefurther to HIV-1 vaccination.
drivenbyHIV-1infection.Previousreportsindicatedthatthere T-cell-independentclassswitchingcanoccurbyactivationof
was a paucity of HIV-1-specific IgA responses in HIV-1 BcellsbyBAFFandAPRILsecretionfromdendriticcells.44,45
infection;30 however, the HIV-1 antigens used and the later Wefoundthatthecytokines BAFF andAPRILare elicited in
timing during acuteinfection ofthesamples likelyaccounted plasmaandmucosalsecretionsjustbeforetheinitialriseinthe
for the low detection of Ab responses in prior studies. HIV-1 Env-specific IgG and IgA responses during AHI. In
Moreover, previous studies have reported the detection of acute and recent HIV-1 infection, naive B cells were
high frequencies of HIV-1-specific IgA responses in seminal significantly reduced relative to the total B-cell population
plasma from chronically infected patients.35,36 In agreement in both the blood and terminal ileum.15 Scholz et al.46
with this, we found that in the majority of patients, gp41- demonstratedthatinhibitionofBAFFcanresultindecreased
specificIgAresponsespeakedandthendeclinedinconcentra- numbersofnaiveBcellsandprimaryimmuneresponseswith
tion in both the mucosal and systemic compartments during littleeffectonmemoryBcellsandlong-livedplasmacells.One
AHI, whereas gp120 IgA appeared later. The ratio of HIV-1- possibilityfortheshorthalf-lifeoftheanti-EnvIgAresponsein
specificIgA:IgGwasrelativelylowingenitalfluids;although acuteinfectionisthatalthoughinnateimmunestimulationcan
there is a higher natural ratio of IgA: IgG in mucosal fluids releasefactorsthatpromoteB-cellclassswitching,thisoccurs
comparedwithplasma.Themechanismfortherelativelylow throughaT-cell-independentpathway.47AsubsequentT-cell-
HIV-1-specific IgA, compared with HIV-1-specific IgG, in dependent pathway for HIV-1-specific Ab induction may be
mucosalfluidsduringHIV-1infectionisnotwellunderstood. hamperedduetothemassiveCD4T-celldepletionthatoccurs
The ratio could be reflective of contributions from plasma early in infection.10,48,49 The transient peak in cytokines that
transudate or a result of mechanisms that cause defects in stimulateB-cellresponses,andtheshorthalf-lifeofHIV-1Env
mucosal class switching, such asHIV-1 Nef-mediatedinhibi- IgA coupled with massive CD4 depletion and damage to
tion of class switching to IgA37 or potentially destruction of mucosal generative B-cell environments15 in AHI are con-
T-regulatory cells important in the development of IgAþ sistentwithaT-independentpathwayfortheinitialHIV-1Ab
mucosal B cells.38 response.
Along-livedprotectiveAbresponseishighlydesiredforany We recently reported on the concentrations and kinetics
vaccine and has been particularly problematic for HIV-1 of the IgG3 subclass of the anti-HIV response during recent
700 VOLUME6NUMBER4|JULY2013|www.nature.com/mi
ARTICLES
HIV-1 infection.31 Anti-gp41 IgG3 appears first among the METHODS
different antigen-specific IgG3 responses, and the anti-gp41, Participants and specimen collection. HIV-1-infected individuals
anti-Gag and anti-RT IgG Abs decline with defined kinetics fromthreedifferentacuteinfectioncohortswereexamined:25HIV-
1þ plasma donors (Clade B),20 8 AHI patients from the Trinidad
withinthefirstyearofinfection,whichcouldbeusefulaspartof
cohort(CladeB),66and30AHIpatientsfromtheCHAVI001acute
an algorithm for measuring HIV-1 incidence.31 The half-life
infectioncohort(CladesBandCfromUnitedStates,SouthAfrica,and
estimatesandconcentrationsofanti-gp41IgAAbssystemically Malawi).4,62,67HIV-1,HBV,andHCVacuteinfectionplasmadonor
andingenitalfluidsduringacuteinfection,asreportedhere,are samplesanduninfectedcontrolplasmadonorsampleswerepurchased
additionalmeasurementsthathavepotentialaspartofanHIV- fromZeptometrix(Franklin,MA).20Pairedbloodandseminalplasma
orectocervicovaginallavagefluidswereobtainedfromtheCHAVI001
1incidencealgorithm.Thereisprecedencefortheutilizationof
acute infection cohort.4,62,67 Seminal plasma samples were obtained
pathogen-specificIgAintestsforrecentinfection,suchasinthe
from ejaculate collected directly into a refrigerated sterile cup con-
case of dengue,50 which utilizes the short half-life of IgA for taining 2.5ml of viral transport medium (Roswell Park Memorial
determining recent infection. Additional characterization of Institutemedium,1,000Upenicillin,1mg/mlstreptomycin,200U/ml
HIV-1-specificIgAresponsesatlaterstagesinHIV-1infection nystatin).Afterliquefaction,thesamplewastransferredtoacentrifuge
tubecontaining2.5mlof2(cid:3) proteaseinhibitorbufferandthetotal
isneededtodeterminewhethertheIgAprofilecandistinguish
volumeanddilutionfactorwerecalculated.Semenwascentrifugedat
recent from chronic HIV-1 infection. 800gfor10min(18–261C),andsupernatantwasaliquotedandstored
The protective properties (e.g., inhibition of ((cid:2)801C). Ectocervicovaginal lavage fluids (cervicovaginal lavage)
transportthroughmucus,Ab-dependentcellularcytotoxicity, wereobtainedthroughrepeatedrinsingofthecervixandectocervix
neutralization,phagocytosis)ofmucosalIgAAbswithdifferent with10mltotalofsalineorbufferedsaline.Fluidwasthentransferred
to a sterile tube with 100(cid:3) protease inhibitor and centrifuged at
HIV-1 specificities have not been fully delineated.
600–800g for 10min (18–261C) to remove cells. The supernatant,
However, some studies of persistently highly HIV-1 exposed includinganymucus,wasaliquotedandstoredat (cid:2)801C.Allwork
but uninfected individuals demonstrated that HIV-1-specific performed as part of this study was reviewed and approved by the
mucosalAbscouldbedetected,andinsomecases,neutralizing institutional review boards of each participating center, Duke
IgAresponseswerefoundinbothseraandmucosalfluidsinthe University Medical Center, and the Division of AIDS, NIH.
absence of detectable HIV-1-specific IgA responses.51–54 In
SpecimenpreparationforIgGremoval.FordetectionofIgAandIgM
addition to having antiviral properties, antigen-specific IgA
Abs, IgG was removed using protein G columns, as previously
Abs may block some IgG functions.55,56 Recent results described.4,57
from the RV144 trial57 raise the hypothesis that plasma
BindingAbassay.CustomizedmultiplexHIV-1-bindingassays(Bio-
anti-gp120 IgA Abs with specific epitope specificities
Plex instrument (Bio-Rad, Hercules, CA)) were performed as pre-
may have potentially blocked anti-gp120 IgG Abs with viouslydescribed4todetermineIgG,IgA,andIgMresponsesspecific
functional activity. However, it is unknown whether mucosal for recombinant HIV-1 p55 or p24 Gag (BD Biosciences, San Jose,
IgAhadaroleinobservedprotectioninRV144,asnomucosal CA), recombinant HIV-1 gp41 MN (Immunodiagnostics, Woburn,
samples were collected. In a recent report of a heterologous MA),apreviouslydescribedartificialmulti-cladegroupMconsensus
gp120 Env protein (Con6 gp120) (ref. 68), HIV-1 p66RT (Protein
prime boost, rectal challenge study in non-human primates,
Sciences,Meriden,CT),HIV-1recombinantNef(Genway,SanDiego,
rectal anti-Env IgG was among a number of variables that
CA), recombinant HIV-1 Tat (Advanced Bioscience, Kensington,
correlatedwithreducedacquisitionrisk.58Moreover,IgAAbs MD),andrecombinantHIV-1p31Integrase(Genway).4Todetermine
with specificity for gp41 Env have been reported to have concentrationofAbsbindingtothe2F5epitope,2F5monoclonalAbs
potentially protective functional properties;59–61 thus the (IgG,IgA,IgM)(PolymuneScientific,Vienna,Austria)weretitrated
ongp41-coupledbeadstogenerateastandardcurve.Inaddition,the
focused specificity of IgA Abs to gp41 during acute infection
followingpeptides(PrimmBiotech,Cambridge,MA)wereused:(gp41
suggests that these easy to elicit Abs at the mucosal surface
immunodominant region, RVLAVERYLRDQQLLGIWGCSGKLI
couldpotentiallyhavesomeprotectiveeffectifpresentbefore CTTAVPWNASWSNKSLNKI), SP62 (gp41 MPER, QQEKNE-
HIV-1transmissionandifadurableHIV-1Env-specificB-cell QELLELDKWASLWN). Immunodominant and MPER tetramers
responsecanbeelicitedbyHIV-1vaccination.Wepreviously wereutilizedforthedetectionofepitope-specificresponses.69HIV-1-
binding Ab measurements and measurement ofIgA Absto tetanus
reported that the initial gp41-specific IgG Abs can capture
toxoid and influenza (Fluzone 2007, Sanofi Pasteur) by standard
infectiousvirions;62however,additionalworkisneededtofully
enzyme-linked immunosorbent assay (ELISA) were performed as
determine the spectrum of inhibitory capabilities of both previously described.42 Total IgG and IgA Ab measurements for
systemic and mucosal IgA and IgG Abs induced in acute calculatingspecificactivitywereperformedusingBio-PlexProhuman
infection,includinganypotentialeffectsofmucosalHIV-1Abs isotyping 7-plex panel (Bio-Rad) according to the manufacturer’s
instructions.ThepositivitycriteriaperHIV-1antigenperAbisotype
on transmission.
was determined by screening Z30 seronegative patients. A stan-
Finally, our data highlight the need to determine the dardized HIV-1positive (HIVþ) control was titrated on each assay
mechanisms that lead to the induction of long-lived (trackedwithaLevy-Jenningsplotwithacceptanceoftiteronlywithin
effective mucosal antibody responses. Whereas vaccination 3 s.d’s of the mean). Standard curves generated using titrated
for influenza63 and polio64 or infection with rotavirus65 can monoclonal Abs and concentrations were used to obtain a four-
parameter logistic equation to determine concentrations of Ab in
elicit IgA responses that can be protective, as yet it
patientsamples(SigmaPlot,SystatSoftware,Chicago,IL).Thecoef-
remains uncertain whether either infection or vaccination
ficientofvariationpersamplewasr15%.Twonegativeseraandtwo
with HIV-1 Env can generate protective mucosal or systemic HIV-1þcontrolserawereincludedineachassaytoensurespecificity
IgA responses. andformaintainingconsistencyandreproducibilitybetweenassays.
MucosalImmunology|VOLUME6NUMBER4|JULY2013 701