All three candidate HIV vaccines included in the current analysis were designed to elicit HIV-1 Env-specific antibody responses (). HVTN 203 was an early phase clinical study using a canarypox prime-protein boost regimen prior to the full-scale RV144 efficacy trial. Volunteers from HVTN203 (Group B) received the canarypox vector expressing a clade B Env, and were boosted with a bivalent clade B/B Env protein formulation from HIV-1 isolates, MN, and GNE8 
, whereas RV144 expressed a clade E Env by canarypox vector, which was then boosted with bivalent clade B/E Env proteins 
. Volunteers in the HVTN 203 trial received a total of four canarypox vector immunizations in addition to two protein boosts adjuvanted with alum that were overlapped with the last two canarypox immunizations. Protein boosts consisted of the same recombinant Env protein vaccine that failed to show protective efficacy in a Phase III clinical trial when used alone 
. HVTN 041 tested the immunogenicity of recombinant Env protein derived from the HIV-1 isolate W61D, adjuvanted in AS02A
, without any prime immunizations 
. The DP6-001 trial used a DNA prime-recombinant protein boost immunization approach delivering a 5-valent Env formulation from HIV-1 isolates of clades A, B, C, and E 
. Human volunteers were first immunized three times with Env-expressing DNA vaccines, followed by two boosts using matched recombinant Env proteins (gp120) in QS-21 adjuvant.
Neutralizing antibody activity has been a key parameter in HIV vaccine research to measure the protective potential of immune sera specific for HIV-1 Env antigens 
. Results of neutralizing antibody activities in three sets of sera included in the current report were previously reported and showed diverse profiles 
. In contrast to sera from the DP6-001 study, which were capable of neutralizing a broad range of T-cell line adapted (TCLA) and primary HIV-1 isolates 
, sera from the HVTN 041 and HVTN 203 studies was only capable of neutralizing autologous and TCLA viral strains 
. Because previous neutralizing activity analyses from each trial were done in different assay systems, making direct comparisons difficult, a new but limited set of neutralization assays were conducted by using pseudotyped viruses expressing three model HIV-1 primary Env antigens with varying degrees of sensitivity to neutralization to confirm the previously reported neutralizing patterns for these three sets of human sera. No extensive NAb analysis was done in the current study, as they have been done in previously published reports 
The vast majority of all sera tested, including 11 out of 12 (92%) from the HVTN 041 study, 10 out of 12 sera (83%) from the HVTN 203 study, and 20 out of 21 (95%) from the DP6-001 study, were capable of neutralizing SF162, a primary isolate highly sensitive to neutralization (). Geometric mean ID50 titers were 1
164 for HVTN 041, 1
62 for the HVTN 203 trial sera, and 1
104 for DP6-001. Sera from the HVTN 041 trial were significantly more potent than those from the HVTN 203 study against the sensitive isolate SF162 (p
0.027), but not significantly different from DP6-001.
Confirmation of neutralizing activities against representative HIV isolates.
Neutralizing activities against SS1196, a primary isolate that is moderately sensitive to neutralization, allowed for some differentiation of the neutralization potential of each trial sera (). Only 4 of the 12 sera (33%) from the HVTN 203 trial were capable of neutralizing SS1196 at a 1
10 dilution but 8 of the 12 sera (67%) from the HVTN 041 trial were capable of neutralizing this virus. In contrast, 18 of the 21 sera (86%) from the DP6-001 trial were capable of neutralizing SS1196. Both the HVTN 041 and DP6-001 trials elicited higher titers than the HVTN 203 trial (p
0.02 and p
The third pseudotyped virus tested in the current analysis expressed Env from the HIV-1 isolate, SC422661.8, a Tier 2 virus representative of those found shortly after the establishment of HIV-1 infection and known to be highly resistant to neutralization 
. A significant drop of neutralizing activities was observed with sera from all three vaccine trials against this virus (). None of the sera from the HVTN 203 trial were capable of reaching 50% neutralization at the lowest dilution tested (1
10). Similarly, neutralizing activity against this isolate was only observed in two sera (17%) from the HVTN 041 trial. However, 10 of the 21 sera (48%) from the DP6-001 trial were capable of neutralizing SC422661.8 at a 1
10 dilution. This occurred despite the fact that, on average, individuals in the DP6-001 had either lower or equivalent titers of Env-specific binding antibodies when compared to other two trial sera ( below). The lack of neutralizing activity from the HVTN 203 and 041 trials against more resistant isolates, and the low titer neutralization seen in the samples from the DP6-001 trial are both consistent with previously reported neutralization profiles 
Geometric mean endpoint binding titers of sera from each of three human vaccine trials against a clade B recombinant gp120 protein (JR-FL).
In order to understand what features of the antibody responses elicited by each of these sera may be responsible for the difference in their neutralization profiles, a wide spectrum of analyses were conducted to understand the quality of different sera. The first was Env-specific binding antibodies. The gp120 protein from the clade B JR-FL strain was chosen as the model antigen to examine binding titers because it is derived from a well-characterized primary isolate and while each trial tested here was formulated with at least one clade B component, JR-FL was not a component in any of the formulations. Antibody levels generated by the HVTN 041 formulation were found to be significantly higher than the titers of binding antibodies generated in either the HVTN 203 or DP6-001 clinical trials (p
0.035 and p
0.0003, respectively) (), suggesting that gp120 adjuvanted with AS02A
is an exceptionally immunogenic formulation.
Antibodies directed to CD4 inducible (CD4i) epitopes are frequently elicited in HIV-infected individuals 
although their role in controlling viral infection is currently unknown. Prior exposure of pseudovirus to soluble CD4 (sCD4) can expose CD4i epitopes, such as the co-receptor binding site, on the viral envelope 
. Sera from each trial included in the current study were assayed for their ability to outcompete binding to 17b, a mAb that targets the co-receptor binding site. High frequency and titers of 17b-like antibodies were detected in all three vaccine trials (). Seven out of 12 sera (58%) from the HVTN 203 trial, 9 out of 12 (75%) from HVTN 041, and 17 out of 21 (81%) from DP6-001 were able to outcompete binding to 17b. Interestingly, those sera that did compete did so at high titer, indicating an abundance of antibodies with this specificity.
Analysis of antibodies against CD4 inducible (CD4i) epitopes.
Next assay evaluated if the CD4i antibodies found in the sera are functional in a modified neutralization assay. Pseudotyped viruses expressing Env from the JR-FL isolate were treated with sCD4 prior to incubation with serum. While without prior sCD4 treatment, JR-FL was difficult to neutralize by sera from all three trials (), significant neutralizing activities against JR-FL Env pseudotyped viruses upon exposure to sCD4 were found in these sera: 7 out of 12 (58%) from HVTN 203, 10 out of 12 (83%) from HVTN 041, and 20 out of 21 (95%) from DP6-001 with positive neutralizing activities (). Geometric mean neutralizing titers for HVTN 203, HVTN 041, and DP6-001 were 1
44, and 1
49, respectively. This data suggests that under the proper conditions, CD4i antibodies present in vaccinee sera would be capable of neutralizing heterologous isolates of HIV-1.
Because it has been reported that sCD4 treatment leads to increased exposure of the V3 loop 
, we attempted to determine if the neutralizing activity observed after sCD4 treatment was due to recognition of the V3 loop or recognition of the co-receptor binding site by the 17b-like antibodies detected through competition. Vaccinee immune sera were incubated with a synthetic peptide matched to the V3 loop sequence of the JR-FL Env prior to the exposure of sCD4-treated JR-FL. This resulted in a slight drop in the geometric mean NAb titer of HVTN 203 sera to 26, of HVTN 041 sera to 25, and of DP6-001 sera to 34 (). This drop in potency was also accompanied by a drop in the frequency of positive neutralizing sera to 6 out of 12 sera (50%) in the HVTN 041 trial and to 16 out of 21 (76%) in the DP6-001 trial (). This data indicates that both V3 and co-receptor binding site antibodies play a role in neutralizing the sCD4-treated JR-FL virus.
Competitive binding assays were conducted against known broadly neutralizing mAbs. Minimal competition was seen against the glycan-specific 2G12 mAb (). None of the 12 sera from the HVTN 041 trial, 2 of the 12 sera (17%) from HVTN 203, and 5 of 21 (24%) from DP6-001 outcompeted binding to 2G12. In contrast, antibodies with specificities similar to that of the V3-specific mAb, 447-52D, were elicited nearly ubiquitously in all of the vaccinee sera tested (). The geometric mean competitive binding titers against 447-52D were 1
108 for the HVTN 203 sera, 1
409 for HVTN 041, and 1
187 for DP6-001. Statistically significant differences in the titers of V3-directed antibodies were observed in the HVTN 041 sera relative to the HVTN 203 sera (p
0.008) and DP6-001 sera (p
Specificity of vaccine-induced antibody responses as determined through mAb competition.
A unique profile of CD4bs-directed antibodies was observed upon examination of the ability of the immune sera to outcompete binding against mAb b12 (). Only 4 out of 12 sera (33%) from either the HVTN 203 trial or HVTN 041 generated an antibody response capable of outcompeting binding to b12. However, 20 out of 21 sera (95%) from the DP6-001 trial were capable of outcompeting binding to b12 and did so with significantly higher titers, sometimes exceeding a 1
500 dilution (p<0.001 against both HVTN 041 and HVTN 203 sera).
Additional functions of gp120-specific antibodies were analyzed. Immune sera elicited by all three vaccine regimens were capable of mediating ADCC function in an equivalent fashion with 19–21% lysis of the recombinant gp120 protein pulsed CEMNKr target cells (). An additional intrinsic characteristic of antigen-specific antibody is the ability to mediate activation of the complement pathway. Complement activation by gp120-specific antibody was conducted for sera from all three trials; however, they all activated complement in a similar fashion. A representative assay result is shown in . A summary of the antibody profiles for each set of immune sera analyzed in the current study is provided ().
Ability of vaccinee sera to mediate ADCC activity.
The ability of Env-specific antibodies to activate the complement cascade present in complement intact normal human sera was determined using deposition of C4 as a marker for complement activation.
Profiles of antibody responses elicited by three HIV vaccine regimens.