Skip to main content
main-content

01.12.2011 | Research | Ausgabe 1/2011 Open Access

Journal of Translational Medicine 1/2011

Defining epitope coverage requirements for T cell-based HIV vaccines: Theoretical considerations and practical applications

Zeitschrift:
Journal of Translational Medicine > Ausgabe 1/2011
Autoren:
Jeffrey R Currier, Merlin L Robb, Nelson L Michael, Mary A Marovich
Wichtige Hinweise

Electronic supplementary material

The online version of this article (doi:10.​1186/​1479-5876-9-212) contains supplementary material, which is available to authorized users.

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

JRC: Conceived and designed the study, developed the model and wrote the manuscript. MAM: Wrote the manuscript and provided important intellectual input into the overall concept of the study. MLR and NLM: Provided support for the study and participated in interpretation of the data and its implication for study vaccine design and assessment. All authors read and approved the final manuscript.

Abstract

Background

HIV vaccine development must address the genetic diversity and plasticity of the virus that permits the presentation of diverse genetic forms to the immune system and subsequent escape from immune pressure. Assessment of potential HIV strain coverage by candidate T cell-based vaccines (whether natural sequence or computationally optimized products) is now a critical component in interpreting candidate vaccine suitability.

Methods

We have utilized an N-mer identity algorithm to represent T cell epitopes and explore potential coverage of the global HIV pandemic using natural sequences derived from candidate HIV vaccines. Breadth (the number of T cell epitopes generated) and depth (the variant coverage within a T cell epitope) analyses have been incorporated into the model to explore vaccine coverage requirements in terms of the number of discrete T cell epitopes generated.

Results

We show that when multiple epitope generation by a vaccine product is considered a far more nuanced appraisal of the potential HIV strain coverage of the vaccine product emerges. By considering epitope breadth and depth several important observations were made: (1) epitope breadth requirements to reach particular levels of vaccine coverage, even for natural sequence-based vaccine products is not necessarily an intractable problem for the immune system; (2) increasing the valency (number of T cell epitope variants present) of vaccine products dramatically decreases the epitope requirements to reach particular coverage levels for any epidemic; (3) considering multiple-hit models (more than one exact epitope match with an incoming HIV strain) places a significantly higher requirement upon epitope breadth in order to reach a given level of coverage, to the point where low valency natural sequence based products would not practically be able to generate sufficient epitopes.

Conclusions

When HIV vaccine sequences are compared against datasets of potential incoming viruses important metrics such as the minimum epitope count required to reach a desired level of coverage can be easily calculated. We propose that such analyses can be applied early in the planning stages and during the execution phase of a vaccine trial to explore theoretical and empirical suitability of a vaccine product to a particular epidemic setting.
Zusatzmaterial
Additional file 1: Subtype and isolate sequence of origin of the natural sequence-based HIV vaccine products assessed in the study. (PDF 21 KB)
12967_2011_977_MOESM1_ESM.PDF
Additional file 2: Comparison of Epicover algorithm with a "hit/no-hit" model of N-mer epitope identity matching. A stringent "hit/no-hit" model of 10-mer coverage was applied to a dataset of 2285 Gag sequences using a tetra-valent ACDE natural sequence vaccine formulation as the query set. As shown below a near-Gaussian distribution of the percent coverage of all isolates in the dataset was obtained. Near identical mean coverage results were obtained using the Epicover algorithm (as implemented in the LANL HIV Sequence Database) and a larger dataset of 3585 Gag sequences. While a broad distribution of "hits per isolate" was obtained (range = 13%-75%), the near identical mean coverage results for the Epicover algorithm and the "hit/no-hit" model resulted in no difference between the two methods for the calculation of epitope coverage requirements and the multiple epitope hit requirement model. (PDF 124 KB)
12967_2011_977_MOESM2_ESM.PDF
Additional file 3: Global group M Gag coverage analysis. Potential HIV isolate coverage provided by mono-valent (panels A, B and C), di-valent (panels D, E and F) and multi-valent (Panels G, H and I) formulations of the four natural sequence based products is shown. Theoretical coverage (90% in the examples shown here) is profoundly dependent upon the number of epitopes generated and the number of exact epitope (10-mer) matches that are required for infected cell recognition. For example if a 1-Hit model is considered, then the tetra-valent product would reach 90% coverage by generating four epitopes per subject on average (Panel G, orange diamonds). In contrast, if a 3-Hit model is considered the mono-valent subtype A product does not reach 90% coverage even if 20 epitopes are generated on average per subject (Panel C, blue circles). (PDF 144 KB)
12967_2011_977_MOESM3_ESM.PDF
Additional file 4: Global group M Env coverage analysis. Potential HIV isolate coverage provided by mono-valent (Panels A, B and C), di-valent (panels D, E and F) and multi-valent (panels G, H and I) formulations of the four natural sequence based products is shown. Theoretical coverage (90% in the examples shown here) is again dependent upon the number of epitopes generated but there is a much greater epitope requirement than for Gag. For example if a 1-Hit model is considered, then the tetra-valent product would reach 90% coverage by generating 7 epitopes per subject on average (Panel G, orange diamonds). If a 3-Hit model is considered the mono-valent subtype products face an intractable problem with extreme epitope requirements for 90% global coverage (Panel C). Even the multi-valent products have stringent epitope requirements (17-20) for reaching 90% coverage (Panel I). (PDF 142 KB)
12967_2011_977_MOESM4_ESM.PDF
Additional file 5: Gag epitope mapping for post- in vitro stimulated cells from a vaccine trial responder subject. PBMC from a vaccinee in the RV158 trial (received 3x rMVA-CMDR) were subjected two rounds of in vitro stimulation, first with rMVA-CMDR for 14 days, then with autologous, irradiated BLCL pulsed with peptide pools matching the Gag (CM240) insert in the vaccine. Each stimulation cycle was 14 days, with rIL-7 added during the first week and rIL-2 added during the second week. Effector cells were tested for responses to a matrix of peptides (11 × 11 peptide pools) matching the Gag insert sequence. Epitopes were counted and identified from the de-convoluted peptide matrix (Panels A and B) in an IFN-γ Elispot assay. Two epitopes were identified within the Gag-specific effector cells and are shown in Panel C. (PDF 193 KB)
12967_2011_977_MOESM5_ESM.PDF
Additional file 6: Env epitope mapping for post- in vitro stimulated cells from a vaccine trial responder subject. PBMC from a vaccinee in the RV158 trial (received 3x rMVA-CMDR) were subjected two rounds of in vitro stimulation, first with rMVA-CMDR for 14 days, then with autologous, irradiated BLCL pulsed with peptide pools matching the Env (CM235) insert in the vaccine. Each stimulation cycle was 14 days, with rIL-7 added during the first week and rIL-2 added during the second week. Effector cells were tested for responses to a matrix of peptides (14 × 13 peptide pools) matching the Env insert sequence. Epitopes were counted and identified from the de-convoluted peptide matrix (Panels A and B) in an IFN-g Elispot assay. Two epitopes were identified within the Env-specific effector cells and are shown in Panel C. (PDF 196 KB)
12967_2011_977_MOESM6_ESM.PDF
Authors’ original file for figure 1
12967_2011_977_MOESM7_ESM.pdf
Authors’ original file for figure 2
12967_2011_977_MOESM8_ESM.pdf
Authors’ original file for figure 3
12967_2011_977_MOESM9_ESM.pdf
Authors’ original file for figure 4
12967_2011_977_MOESM10_ESM.pdf
Authors’ original file for figure 5
12967_2011_977_MOESM11_ESM.pdf
Literatur
Über diesen Artikel

Weitere Artikel der Ausgabe 1/2011

Journal of Translational Medicine 1/2011 Zur Ausgabe