The herpevac trial for women: Sequence analysis of glycoproteins from viruses obtained from infected subjects

Abstract

The Herpevac Trial for Women revealed that three dose HSV-2 gD vaccine was 58% protective against culture-positive HSV-1 genital disease, but it was not protective against HSV-2 infection or disease. To determine whether vaccine-induced immune responses had selected for a particular gD sequence in strains infecting vaccine recipients compared with viruses infecting control subjects, genetic sequencing studies were carried out on viruses isolated from subjects infected with HSV-1 or HSV-2. We identified naturally occurring variants among the gD sequences obtained from 83 infected subjects. Unique or low frequency amino acid substitutions in the ectodomain of gD were found in 6 of 39 HSV-1-infected subjects and in 7 of 44 HSV-2-infected subjects. However, no consistent amino acid change was identified in isolates from gD-2 vaccine recipients compared with infected placebo recipients. gC and gE surround and partially shield gD from neutralizing antibody, and gB also participates closely in the viral entry process. Therefore, these genes were sequenced from a number of isolates to assess whether sequence variation may alter protein conformation and influence the virus strain's capacity to be neutralized by vaccine-induced antibody. gC and gE genes sequenced from HSV-1-infected subjects showed more variability than their HSV-2 counterparts. The gB sequences of HSV-1 oral isolates resembled each other more than they did gB sequences rom genital isolates. Overall, however, comparison of glycoprotein sequences of viral isolates obtained from infected subjects did not reveal any singular selective pressure on the viral cell attachment protein or surrounding glycoproteins due to administration of gD-2 vaccine.

Fig 1. gD amino acid sequences of isolates from the Herpevac Trial.

The U_S6 gene of isolates from subjects infected during the Herpevac Trial was sequenced and then translated. A) Diagram of the gD protein showing boundaries of functional domains; TM, transmembrane. 309t in red indicates the site of truncation in the HSV-2 gD vaccine construct. Substitutions in the gD amino acid sequences of isolates from B) 39 HSV-1-infected subjects compared with reference sequence KOS (HSV-1 laboratory strain), and C) 44 HSV-2-infected subjects compared with G (HSV-2 laboratory strain and derivation of the gD-2 vaccine), and SD90e (HSV-2 primary isolate from South Africa) are shown. Numbers to the right of each sequence representation indicate the number of subjects sharing the amino acid sequence. C, control vaccinated subject; V, gD-2 vaccinated subject.

Fig 2. gC amino acid sequences of isolates from the Herpevac Trial.

The UL44 gene encoding gC was sequenced from selected subjects’ isolates and aligned to reference sequences, then translated. Amino acid substitutions are diagramed for A) gC from HSV-1-infected subjects, B) gC from HSV-2-infected subjects. Numbers to the right of each sequence representation indicate the number of subjects sharing the amino acid sequence. C, control vaccinated subject; V, gD-2 vaccinated subject. The triangle represents a deletion of one amino acid.

Fig 3. gE amino acid sequences of isolates from the Herpevac Trial.

The US8 gene encoding gE was sequenced from selected subjects’ isolates and aligned to reference sequences, then translated. Amino acid substitutions are diagramed for A) gE from HSV-1-infected subjects, and B) gE from HSV-2-infected subjects from the same sets of subjects as described in Fig 2. Numbers to the right of each sequence representation indicate the number of subjects sharing the amino acid sequence. C, control vaccinated subject; V, gD-2 vaccinated subject. The inverted triangle represents an insertion of two amino acids (GE).

Fig 4. gB amino sequences.

The UL27 gene encoding gB was sequenced from selected subjects’ isolates and aligned to reference sequences. Amino acid substitutions in the translated sequence are diagramed for gB from A) HSV-1-infected subjects, and B) HSV-2-infected subjects from the same sets of subjects as described in Fig 2. Numbers to the right of each sequence representation indicate the number of subjects sharing the amino acid sequence. C, control vaccinated subject; V, gD-2 vaccinated subject.

Fig 5. dN/dS ratios for HSV-1 and HSV-2 glycoprotein gene sequences.

Mean ratio of non-synonymous to synonymous evolutionary substitutions (dN/dS) within the U_S27 (gB), U_L44 (gC), U_S6 (gD) and U_S8 (gE) genes is shown for all HSV-1 and -2 strains sampled. The dN/dS ratio for all HSV-2 gD sequences from gD-2 vaccine recipients was not significantly different than control subjects (0.61 versus 0.53, P = 0.554 by unpaired t test).