Linear epitopes in Onchocerca volvulus vaccine candidate proteins and excretory‐secretory proteins

Summary In our previous study, a proteome‐wide screen was conducted to identify linear epitopes in this parasite's proteome, resulting in the discovery of three immunodominant motifs. Here, we investigated whether such antigenic peptides were found in proteins that were already known as vaccine candidates and excretome/secretome proteins for Onchocerca volvulus This approach led to the identification of 71 immunoreactive stretches in 46 proteins. A deep‐dive into the immunoreactivity profiles of eight vaccine candidates that were chosen as most promising candidates for further development (Ov‐CPI‐2, Ov‐ALT‐1, Ov‐RAL‐2, Ov‐ASP‐1, Ov‐103, Ov‐RBP‐1, Ov‐CHI‐1, and Ov‐B20), resulted in the identification of a poly‐glutamine stretch in Ov‐RAL‐2 that has properties for use as a serodiagnostic marker for O. volvulus infection. A peptide ELISA was developed, and the performance of this assay was evaluated. Based on this assessment, it was found that this assay has a sensitivity of 75.0% [95% CI: 64.9%‐83.5%] and a specificity of 98.5% [95% CI: 94.6%‐99.8%]. Furthermore, 8.7% reactivity in Asian parasite‐infected individuals (8 out of 92) was observed. Besides this identification of a linear epitope marker, the information on the presence of linear epitopes in vaccine candidate proteins might be useful in the study of vaccines for river blindness.


| BACKG ROU N D
Of the 20 infectious diseases listed on the World Health Organization where at least 120 million are at risk. 4,5 Current treatment programs are based on mass drug administration (MDA) of the microfilaricidal agents ivermectin (Mectizan, Merck) as no approved macrofilaricide drugs or vaccines are available. However, contraindications in areas co-endemic for loiasis, an inability to break transmission in some foci, and the possible emergence of drug resistance ask for a change in strategy including vaccination in order to control or eliminate onchocerciasis. 6,7 Two basic strategies were used to identify and clone O. volvulus target vaccine antigens: immunoscreening of various O. volvulus cDNA libraries, and rational antigen selection. [8][9][10][11][12][13][14] Also, analysis of Summary In our previous study, a proteome-wide screen was conducted to identify linear epitopes in this parasite's proteome, resulting in the discovery of three immunodominant motifs. Here, we investigated whether such antigenic peptides were found in proteins that were already known as vaccine candidates and excretome/secretome proteins for Onchocerca volvulus This approach led to the identification of 71 immunoreactive stretches in 46 proteins. A deep-dive into the immunoreactivity profiles of eight vaccine candidates that were chosen as most promising candidates for further development (Ov-CPI-2, Ov-ALT-1, Ov-RAL-2, Ov-ASP-1, Ov-103, Ov-RBP-1, Ov-CHI-1, and Ov-B20), resulted in the identification of a poly-glutamine stretch in Ov-RAL-2 that has properties for use as a serodiagnostic marker for O. volvulus infection. A peptide ELISA was developed, and the performance of this assay was evaluated. Based on this assessment, it was found that this assay has a sensitivity of 75.0% [95% CI: 64.9%-83.5%] and a specificity of 98.5% [95% CI: 94.6%-99.8%]. Furthermore, 8.7% reactivity in Asian parasite-infected individuals (8 out of 92) was observed.
Besides this identification of a linear epitope marker, the information on the presence of linear epitopes in vaccine candidate proteins might be useful in the study of vaccines for river blindness. excretory-secretory products (ESPs) in nodule fluid of cows infected with Onchocerca ochengi has led to the identification of 85 proteins with potential pharmacological properties or immunogenic potential. 15 Based on a mouse-Onchocerca model, the top ranking eight Ov protective antigens (Ov-CPI-2, Ov-ALT-1, Ov-RAL-2, Ov-ASP-1, Ov-103, Ov-RBP-1, Ov-CHI-1, and Ov-B20) were chosen for more extensive studies. 16 Using two model systems, O. volvulu in mice and Brugia malayi in gerbils, this list of candidates has been reduced to the final selection of Ov-103 and Ov-RAL-2 for further clinical development. 10,16,17 In our previous work, the entire O. volvulus proteome was screened for the presence of linear epitopes. 18 Based on this work, the serodiagnostic peptides OvMP-1 and OvMP-23 were proposed and evaluated as markers for onchocerciasis. 19 In the study presented here, we use the previously obtained data from the highdensity peptide arrays to investigate whether the vaccine candidates and ESP's described above contain one or more peptide fragments that are recognized by antibodies in chronically infected individuals, whether immunodominant epitopes could be identified and whether some of them might have properties that make them attractive serodiagnostic candidates.

| Study samples
All samples used in this study were de-identified before being provided, written informed consent was obtained from all individuals and usage of these samples for research purposes was approved by an ethical committee or Institutional Review Board (IRB). Detailed characteristics of the samples used was published before. 18 Additionally, the second set of plasma samples from O. volvulus infected individuals were collected as part of a field study in Ghana.
This study was undertaken in an Onchocerciasis-endemic community located in Adansi South District along the Pra river basins in the Ashanti Region of Ghana. Physical examinations were performed to identify those subjects having palpable nodules. Skin snips (biopsies) were then taken in order to determine the microfilarial (mf) load in the skin. 20 Most subjects were participating in mass drug administration programs. A total of 97 nodule positive subjects that donated plasma samples were included. Demographic information is provided in Table 1. For the non-Onchocerca endemic control samples, demographic information is also provided in Table 1.
TA B L E 1 Study populations used for determination of diagnostic performance

| Peptide array analysis
Data of high-density peptide arrays were obtained as described previously. 18 For interpretation of the individual results, two cut-offs were used, based on the previous analyses. 18 A first cut-off of 837 Relative Fluorescence Units (RFU) was used as the upper cut-off for healthy control samples to be confirmed as negative. The second cut-off of 4842 RFU was used as the lower cut-off for samples from O. volvulus infected individuals to be confirmed as positive.

| Total IgG peptide ELISA
Biotinylated synthetic peptide OVOC9988;25-33 (QQQQQQQQR) was synthesized by standard procedures and purchased from PEPperPRINT GmbH (Heidelberg, Germany). For determination of peptide specific serum antibody levels, a peptide ELISA was developed as described previously. 18

| Bio-IT analysis
Homologs of Ov-RAL-2 were identified using the BLAST tool in WormBase Parasite (Version WBPS10, WS263) and fasta files of the sequence of these homologs were obtained. Multiple Sequence Alignment was performed using the online Clustal W aligner tool in T-COFFEE, Version_11.00.d625267 (http://tcoffee.crg.cat/apps/ tcoffee/index.html). 21 Identical or similar epitopes in human proteins were searched in the immune epitope database (www.iedb.org). 22

| Linear epitopes in eight selected vaccine candidate proteins
In 2015, an international consortium launched The Onchocerciasis Vaccine for Africa (TOVA) with the goal of evaluating and pursuing vaccine development for onchocerciasis. After a rational design for the antigen discovery and selection process, eight recombinant pro- Ov-RAL-2, Ov-ASP-1, Ov-103, Ov-RBP-1, Ov-CHI-1, and Ov-B20. 16 The corresponding native proteins are encoded by the genes OVOC7453, OVOC12769, OVOC9988, OVOC9575, OVOC4230, OVOC8754, OVOC12569, and OVOC9221/9222, respectively. All peptides used in the high-density peptide array from these proteins were listed, and signals of the onchocerciasis patients and healthy controls were plotted ( Figure 1 and Data S2). Detailed investigation of the immune response against peptides from these eight proteins shows that immunodominant linear epitopes are present in these proteins. A peptide was considered to bear immunodominant epitopes if at least 8 out of 12 cases had a signal above 4842 RFU.
In Ov-RAL-2 there is a poly-glutamine stretch between residues 25 and 33 (QQQQQQQQR) that is recognized by 9 of the 12 infected individuals. Also in Ov-B20 (OVOC9221) one peptide between residues 67 and 80 (WATTENSSSIDSNN) was found to be recognized by 9 of the 12 infected individuals, but an average signal of the onchocerciasis cases was markedly lower than for the peptide from Ov-RAL-2. It must, however, be said that the poly-glutamine stretch in

| Peptide ELISA based on poly-glutamine stretch of Ov-RAL-2
Based on the strong and dominant immune response that was observed against the poly-glutamine stretch in Ov-RAL-2, and the fact that the same stretch was identified as immunoreactive peptide in a previous study, we decided to further explore this peptide. 23   (autoimmune related) epitopes did not return any identical hit for this peptide sequence, but one epitope in the human protein Trinucleotide repeat-containing gene 6A was found to have a similar poly-glutamine stretch.
In order to confirm the observation from the high-density peptide array and to further study the specificity of this peptide, we developed a peptide ELISA based on this 9-mer peptide and deter-

| Detection of linear epitopes in extended list of vaccine candidates and ESP's
Since the analysis on the eight selected vaccine candidates demon- are responsible for inducing the immune response or the polyglutamine stretch in OVOC9988 cannot be deduced from these data. OVOC9384 is known as Gln-rich protein and has been proposed by others as serodiagnostic marker for onchocerciasis. 24 Based on our observation here, it appears that linear epitopes in OVOC9384 play an important role in the immune recognition of this protein.
F I G U R E 1 Response against peptides derived from the eight selected vaccine candidate proteins Ov-CPI-2, Ov-ALT-1, Ov-RAL-2, Ov-ASP-1, Ov-103, Ov-RBP-1, Ov-CHI-1 and Ov-B20. For every peptide derived from these Onchocerca volvulus proteins signals in the peptide arrays are plotted as the average of 12 infected individuals (blue circles) and of six healthy controls (red squares). Error bars represent SEM. Numbers in the x-axis indicate the start position of the peptide with reference to the full-length protein F I G U R E 2 Alignment of Ov-RAL-2 (OVOC9988) and its closely related homologs in other helminth species (Brugia malayi, Wuchereria bancrofti, L. loa and Ascaris lumbricoides). The location of the immunodominant peptide has been indicated by the grey box

| D ISCUSS I ON
From the longlist of vaccine candidates for O. volvulus, ultimately 8 proteins have been selected for preclinical evaluation and two of them, Ov-103 and Ov-RAL-2, were picked for further clinical development as vaccine candidate. 12,16 In the study presented here, we investigated in detail the immunoreactivity of peptides derived from these eight proteins. Several peptide stretches in all of these proteins were found that were recognized by a small number of infected individuals (less than half of them). In Ov-103 and Ov-RAL-2, the two candidates in clinical development, it was observed that a stretch is present that is recognized by eight or nine of the 12 infected individuals, respectively. Although we only confirmed the results from high-density peptide array in peptide ELISA for the stretch in Ov-RAL-2, our data indicate that these proteins are obviously already exposed to the immune system upon natural infection and an immune response is being raised against a particular linear epitope in these proteins but clearly this is not able to stop further development of the parasite as the onchocerciasis patients used to perform the high-density peptide arrays had high microfilarial load. 18 Given the large portion of infected individuals that had a strong response to the 9-mer peptide from Ov-RAL-2 with sequence QQQQQQQQR and the fact that this sequence was identified before as immunoreactive peptide, we investigated whether this peptide could be an interesting serological marker for infection with O. volvulus. 23 Of interest, the poly-glutamine stretch in the Nterminal region of Ov-RAL-2, is absent in all its closely related homologs in other helminth species making this peptide very attractive for further diagnostic development. Blast analysis however also showed that this small peptide stretch is found in many proteins, one of them infected individuals raised an immune response against the linear epitopes found in both the vaccine candidate and ESPs. A difference in HLA type of the infected individuals might be responsible for this, 32 but for ESPs it might also be possible that some nodules disseminate more of these proteins in the periphery ('leaky nodules'), making them more accessible for the immune system.
In conclusion, this work demonstrates that several vaccine candidate proteins and excretome/secretome proteins display immunodominant linear epitopes. We also demonstrate that a 9-mer peptide containing a poly-glutamine stretch from the vaccine candidate Ov-RAL-2 might be an interesting serological marker warranting further exploration.

ACK N OWLED G EM ENTS
We