Table 1.
Serum samples used in this study.
Details of all serum samples used in the western blots, peptide arrays and RDT prototypes.
Fig 1.
In silico filter applied to select desired protein features.
Selected protein properties (yellow and green nodes) were searched in silico on all protein hits identified by MS from selected IgG1 sero-reactive western blot bands (MS output—top node), with the ultimate goal of shortlisting protein candidates most likely to be antigenic (Shortlisted proteins—bottom node).
Table 2.
Information on the B-cell epitope prediction algorithms employed.
A total of 80 peptides with high antigenic scores, independently predicted from four different B-cell prediction algorithms, were selected for synthesis. AA: amino acid; AUC: area under the curve.
Fig 2.
Gel and western blot strips immunoassayed with sera from Indian VL and EHC patients.
The western blot strips incubated with individual active VL sera revealed the L. donovani protein bands reacting with IgG1 from Indian patients (VL WB). Three antigenic bands, B1, B2 and B3 were excised from gels to have their constituent proteins identified by MS (Gel). The strips incubated with EHC sera did not develop any band (EHC WB). Molecular mass markers are given in kDa.
Fig 3.
Selected proteins for in silico epitope mapping.
65 proteins (‘Final 65 v2 IDs’) satisfied the criteria in either branch shown in Fig 1 and were shortlisted to have their epitopes mapped with multiple in silico B-cell prediction algorithms. In bold are the features added after each step of the filter.
Fig 4.
Peptides reactivity with a pool of Sudanese VL and NEHC sera diluted at 1:100 showed to be unspecific.
[VL]: Section of an array hybridised with pooled Sudanese VL serum at 1:100 dilution. rk39 and a whole L.donovani lysate (CLA) were spotted on the diagonal as positive controls as well as for orientation purposes. [NEHC]: the same array section hybridised with a pool of NEHC sera at same dilution. The red circles indicate the spotting position of a peptide specific for T. cruzi [42], employed here as negative control. Both images indicate the array section where the peptide epitopes predicted from the lbtope algorithm were spotted.
Fig 5.
Hybridisation of the peptide arrays with a pool of serum samples diluted 1:200 revealed the most sensitive peptide.
[VL]: Section of a peptide array hybridised with pooled Sudanese VL serum at 1:200 dilution. The most reactive peptide (pep) was spotted in duplicates. rk39 and CLA spotting positions are shown. [NEHC]: the same array hybridised with a pool of NEHC sera at same dilution. Both images are from the array section where the peptide epitopes predicted from the EpiQuest algorithm were spotted.
Table 3.
Information on the most antigenic peptide identified visually from the pilot screening with desalted peptides.
Column ‘v2 ID’ and ‘v2 gene product’ refer to the improved LdBPK reference genome.
Fig 6.
Peptide specificity for VL IgG1 was highly concentration dependent.
Peptide EpQ11 was spotted at three different concentrations in eight replicates each onto dry NC membranes, previously soaked with NLA. The highest discrimination between VL and NEHC IgG1 (Ratio VL/NEHC) was obtained when the peptide was spotted at 1.15 mg/ml.
Fig 7.
Peptide EpQ11 is specifically detected by Sudanese VL IgG1 in cassette as well as in dipstick format.
‘T’ indicates the location of the NLA at which, in a positive test, there is a red coloured line due to the presence of peptide-IgG1 complex. Successful migration was ensured by the development of the control line at ‘C’.
Table 4.
EpQ11 performance in RDTs with Sudanese VL and NEHC patient sera.
Summary of the RDT prototypes tested with individual VL sera from Sudan or NEHC serum samples from Europe.
Fig 8.
The L. donovani protein containing the EpQ11 sequence shows low similarity to proteins from organisms causing potentially cross-reacting diseases.
The L. donovani protein harbouring the EpQ11 peptide (LdBPK_360019900.1—v2 ID) was aligned against the proteome of organisms causing potentially cross-reacting diseases with VL using MAFFT v7.222 [61]. Column 1: GenBank accession number; column 2: protein length (position of first amino acid | position of last amino acid, 784 for LdBPK_360019900.1), column 3: alignment around the region where the EpQ11 peptide is located. Numbers on top (470-560) indicate the alignment position. Screenshot from belvu, SeqTools—4.44.1 [62] showing the most similar proteins to LdBPK_360019900.1.