Expression of a rK39 homologue from an Iranian Leishmania infantum isolate in Leishmania tarentolae for serodiagnosis of visceral leishmaniasis

Background Kinesin-related gene diversity among strains and species of Leishmania may impact the sensitivity and specificity of serodiagnostic tests for visceral leishmaniasis (VL). Methods In this study, we report on the recombinant expression of this novel Iranian Leishmania infantum (MCAN14/47) homologue of rK39 (Li-rK39), in L. tarentolae. The diagnostic potential of the Li-rK39 antigen was evaluated in an ELISA, using sera from 100 VL patients, 190 healthy endemic controls, 46 non-endemic healthy controls and 47 patients with other infections. Results The results showed a sensitivity of 96% and a specificity of 93.8%. A commercial rK39 immunochromatographic test (ICT) was 90% sensitive and 100% specific on the same cohort. Conclusions Here, we show that the K39 gene from an Iranian L. infantum isolate is heterozygous as compared to the sequence of the Brazilian L. infantum (former L. chagasi), whose antigen is incorporated in most rK39-based immunochromatographic tests. Therefore, Li-rK39 has the potential to be used as an alternative for VL diagnosis in Iran.


Background
Visceral leishmaniasis (VL) or kala-azar is a neglected parasitic disease with an annual worldwide incidence of 500,000 human cases, caused by an obligate intracellular parasite belonging to the Leishmania donovani complex [1][2][3]. Early diagnosis and intervention are critical for symptomatic patients since the disease is lethal if left untreated [4]. The definitive diagnosis of VL is based on the visualization of the parasite in Giemsa-stained smears, obtained from the spleen, liver, bone marrow or lymph node. The main problem with parasitological diagnosis is its variable sensitivity and invasiveness [5,6]. During a VL infection, substantial amounts of specific antibodies are produced that are not protective but can serve as a marker of ongoing or past infection with Leishmania sp. Accordingly, different serological tests such as immunofluorescence assay (IFAT), direct agglutination test (DAT), and enzyme-linked immunosorbent assay (ELISA) have been developed. Problems with IFAT, DAT and some ELISAs that use whole Leishmania parasites or crude extracts are batch to batch variation, the need for equipment and, most importantly, limited sensitivity [7][8][9][10]. Previous meta-analysis reported a sensitivity of 88% for IFAT, 87% for ELISA and 94% for DAT [11]. However, a study conducted in Iran, the region of the present study, reported a sensitivity of 70.5% for DAT [12].
The development of the recombinant antigen rK39 by Burns et al. [13] significantly contributed to the improvement of VL diagnosis. rK39 is a protein containing 39 amino acid repeats derived from a conserved region within a gene coding for a kinesin-related protein of L. infantum (former L. chagasi). Either using ELISA or simple immunochromatographic tests, high levels of antibodies against K39 are detectable in the blood of VL patients but not in those with mucosal or cutaneous leishmaniasis [10,[14][15][16]. Compared to other serological tests and test formats, rapid diagnostic tests (RDT) based on rK39 have the advantages of being fast, simple without the need for equipment and visual reading of the result [8,10]. Nonetheless, rK39 RDTs show varying diagnostic performance in different parts of the world and even within a single region if heterogeneous Leishmania parasite populations are present [17]. Meanwhile, polymorphisms of the kinesin-related gene in various strains of L. donovani in different regions of the world may explain the discrepancies in sensitivity of the rK39 antigen in different serodiagnostic tests [18,19].
Recombinant protein expression systems have been developed in prokaryotic organisms like bacteria and in eukaryotic cells and organisms such as yeast, mammalian, insect and plant cells, and protozoa like L. tarentolae, yet all have their advantages and drawbacks [20]. A major shortcoming of heterologous recombinant expression is codon bias and the processing of post-translational modifications [21]. Leishmania tarentolae, a trypanosomatid protozoan parasite of the Moorish gecko Tarentola mauritanica, was genetically engineered as a eukaryotic expression system (Leishmania expression system, LEXSY) for the production of recombinant proteins [22]. So far, rK39 and rK39-like antigens have been produced in heterologous expression systems [13,23,24]. We hypothesized that using L. tarentolae for expression of rK39 antigens derived from endemic Leishmania species would be advantageous in terms of codon usage and post-translational processing of the recombinant protein. Thus, the present study was undertaken to develop an alternative rK39 (Li-rK39), derived from an Iranian L. infantum strain and expressed in L. tarentolae, for serological diagnosis of VL in endemic regions of Iran.

Specimen collection
A total of 383 sera from 100 VL patients, 190 endemic healthy controls, 46 non-endemic healthy controls and 47 non-VL patients were included. Five ml of venous blood were collected in plain tubes to prepare serum. VL patients consisted of 48 male and 52 female children of 1 month to 16 years-old, all referred from VL-endemic regions and admitted to Nemazee Hospital, Shiraz University of Medical Sciences, Shiraz, Iran. All patients had ≥ 14 days' fever, 88 had hepatosplenomegaly and 92 had anaemia (Hb < 11g/l). They were all positive in IFAT (titer > 64). After treatment with either antimonial or amphotericin B therapy, they were all cured with decreased spleen size. Endemic healthy controls were 190 children (55% male and 45% female) with age range of 1-16 years from endemic areas for VL in southern Iran, with no clinical symptoms and no history of VL. Non-endemic healthy controls included 46 individuals (50% male and 50% female) with an age range of 15-35 years, originating from non-endemic regions. Non-VL patients (n = 47) suffered from other infections: toxoplasmosis (n = 10); malaria (n = 10); cutaneous leishmaniasis (n = 11); fascioliasis (n = 7); and hydatidosis (n = 9). All sera from the endemic healthy controls, non-endemic healthy controls and non-VL patients were negative in IFAT (titer < 64).

rK39 rapid diagnostic test
The test was conducted according to the manufacturer's instructions (Kalazar Detect ™ Rapid Test for VL, InBios, Seattle, USA).

Leishmania infantum strain, DNA isolation and PCR amplification
A L. infantum strain (MCAN/IR/14/M14) was isolated from a domestic dog in Meshkin-Shahr area from northwestern Iran in 2015 [25]. Promastigotes were cultured in 10 ml of HOMEM medium (GE Healthcare) supplemented with 10% heat-inactivated fetal calf serum and incubated at 26 °C. Pellets of 10 9 L. infantum promastigotes were washed and suspended in PBS and genomic DNA was extracted using QIAamp Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions. DNA quantity was measured using Nan-oDrop spectrophotometer (NanoDrop ND-1000 UV-Vis spectrophotometer, NanoDrop Technologies, Wilmington, DE, USA).

Construction of the recombinant expression vector
An analogous fragment of the K39 gene, as described by Burns et al. [13], was amplified from the Iranian L. infantum strain, using a forward primer selected to bind to the upstream non-repetitive part of Gen-Bank sequence L07879, and a reverse primer able to target within most of the repetitive 117 bp regions (Table 1) [13]. As required for In-Fusion cloning, each primer was modified at the 5'-end by addition of a 15-bp sequence including a restriction site complementary to the place of integration in the expression vector, pLEXSY-hyg2 (Jena Bioscience, Jena, Germany). Using XbaI-rK39 F and KpnI-rK39 R primers, integration in pLEXSY-hyg2 provided a L. mexicana signal peptide at the N-terminal and added a hexa histidinetag at the C-terminal, thus allowing IMAC purification. For cytoplasmic expression in pLEXSY-hyg2, the M14/47 sequence was re-cloned from pLEXSY-hyg2-M14/47 using NcoI-rK39 F and KpnI-rK39 R primers. This reconfiguration to pLEXSY-hyg2-CYTO-M14/47 cleaved off the existing signal peptide at the N-terminal, and provided a NcoI start codon. PCR reactions were performed in a final volume of 20 µl including 4 µl of 5× Phusion buffer (NEB), 200 µM of dNTPs (Eurogentec, Seraing, Belgium), 0.5 µM of each primer, 0.5 U of Phusion DNA polymerase (NEB), 4 µl of Q-solution 5× (Qiagen) and 10 ng genomic DNA. The final PCR program was run as follows: 1 cycle of 98 °C for 10 s; 35 cycles of 98 °C for 10 s, 67 °C for 30 s, 72 °C for 1 min; 1 cycle of 72 °C for 5 min. PCR products were separated on 1.5% agarose gels by electrophoresis and visualized, using ethidium bromide staining.
PCR products were purified, using the NucleoSpin PCR Clean-Up Kit (Macherey-Nagel, GmbH & Co. KG, Germany), according to the manufacturer's instructions. Next, the purified amplicons were integrated into a XbaI or NcoI and KpnI double digested pLEXSY-hyg2 vector, using In-Fusion recombinase. Finally, the completed vector was transformed into Stellar ™ Competent Cells (Takara Bio Company, CA, USA) by heat shock and plated on Luria Bertani (LB) agar supplemented with 100 µg/ml carbenicillin at 37 °C overnight. At least 48 isolated Escherichia coli colonies were screened using colony PCR with primers that surround the site of integration in the pLEXSY vector (P1442 and A264). Using gel electrophoresis, all colonies that contained inserts larger than 300 bp were selected for further subculture in 5 ml LB medium overnight. Plasmids were harvested using the QIAprep Spin Miniprep (Qiagen) and sent for bi-directional sequencing (VIB Genetic Sequencing Facility, Antwerp, Belgium) using the P1442 and A246 primers. Next, translated protein sequences of the K39 gene were analysed by MUSCLE alignment with Gen-Bank: L07879 reference sequence in CLC Sequence Viewer (Qiagen).

Cultivation and transfection of L. tarentolae
Leishmania tarentolae strain P10 was cultivated in brain heart infusion (BHI) medium supplemented with hemin and penicillin-streptomycin. pLEXSY-hyg2-CYTO-M14/47, was linearized by SwaI digestion and concentrated to 330 ng/µl by QIAquick PCR purification kit (Qiagen). Transfections were performed with 10 µg of linearized plasmid via electroporation using the high voltage protocol on freshly passaged promastigotes, according to the Jena Bioscience manual [22]. Transgenic LEXSY strains were generated by polyclonal selection in BHI suspension culture, using 100 µg/ml hygromycin, after they were maintained under hygromycin pressure.

Purification and characterization of recombinant protein
Production of recombinant protein Li-rK39 was carried out in 500 ml BHI medium, supplemented with 1.5% yeast extract [26], 2 g/l glucose [27], 100 µg/ml hygromycin and 0.2% hemin. After 72 h, when the OD 600 reached 4, the cells were harvested by centrifugation for 10 min at 7500×g at 4 °C. The pellet was re-suspended in 50 ml of binding buffer (20 mM phosphate buffer pH 7.4, 500 mM NaCl, 20 mM imidazole), supplemented with 3 tablets of a protease inhibitor cocktail (Cat No. 05892791001, Roche Diagnostics, Mannheim, Germany), and lysed by five freeze-thaw cycles using liquid nitrogen, followed by 30 s sonication. After centrifugation for 30 min at 15,000×g at 4 °C, the supernatant was applied to an equilibrated 5 ml Ni-NTA agarose column. Bound proteins were eluted using 6 ml elution buffer (20 mM phosphate buffer pH 7.4, 500 mM NaCl, and 500 mM imidazole). Protein fractions were separated on 15% pol-
Sensitivity and specificity at different cut-off points were calculated for the Li-rK39-ELISA and used to construct a receiver-operator characteristic (ROC) curve and to calculate the area under the curve (AUC) with its 95% CI.

Amplification and sequencing of the L. infantum K39 gene fragment
PCR amplification of the L. infantum K39 fragment using the XbaI-K39F and KpnI-K39R primers resulted in a multi-banded amplification product, which could not be reduced to single bands by varying the annealing temperature ( Fig. 1) or by using a different cloning polymerase. Therefore, PCR products were co-purified and used for cloning in the pLEXSY-hyg2 expression vector. Individual recombinant clones screened by colony PCR revealed in each the integration of a single fragment; however, between the different clones these fragments had various sizes, usually differing in 117 bp from each other, corresponding with the size of one 39 aa repeat. In total, six plasmids yielded rK39 sequences from MCAN/IR/14/ M14 with a length greater than 300 bp, for which the largest, MCAN14/47 and MCAN14/22 had a size of 550 bp. Sequences of these two genes were deposited in the GenBank database with accession numbers MN585730 (MCAN14/47) and MN585731 (MCAN14/22).
When translated to protein, the sequences revealed that the Iranian L. infantum K39 appears to have at least 2 alleles. The first allele, from clone MCAN 14/22, had 100% identity at 180 amino acids positions of the corresponding translated L. chagasi kinesin-like gene (GenBank: L07879). In contrast, the second allele, from clone MCAN14/47 had no substitutions in the N-terminal non-repetitive part, 5 amino acids substitutions in the first repeat, 4 substitutions in the second repeat, none in the third repeat and 2 substitutions in a partial fourth repeat (Fig. 2), thus, resulting in only 94% protein similarity to the corresponding fragment from GenBank (L07879). The MCAN14/47 from pLEXSY-hyg2 was recloned using NcoI-K39 F and KpnI-K39 R primers into pLEXSY-hyg2-CYTO-MCAN14/47. Here, the artificial start codon provided by NcoI substituted the first amino acid, E, and the second, L, from MCAN14/47 to M and V, respectively in order to allow cytoplasmic expression of the 188 amino acid long Li-rK39 construct.

Expression and purification
The pLEXSY-hyg2-MCAN14/47 plasmid was transfected to L. tarentolae and polyclonal selection using hygromycin yielded a recombinant population. This recombinant population expressed a cytosolic recombinant protein, Li-rK39, which could be purified on Ni-NTA resin. The yield of this purified protein was 250 µg/ml in 500 ml culture. The presence of the recombinant Li-rK39 protein and its purity was confirmed by SDS-PAGE and Coomassie blue staining (Fig. 3a) and by Western blot, using an anti-his-tag antibody (Fig. 3b). The apparent size of the protein was 25 kDa although the expected size was 21 kDa (including the 6 histidine residues), probably due to posttranslational modification like glycosylation or due to the relatively high content in basic amino acids (34%).
Two of the non-VL patients, including malaria and cutaneous leishmaniasis showed a positive reaction in the Li-rK39 ELISA.

Concordance between different tests
By including all the samples including endemic, nonendemic and non-VL patients, the concordance between the Li-rK39 ELISA and the rK39-ICT was good with k = 0.80 and P < 0.0001. Concordance between the Li-rK39 ELISA and IFAT and between rK39-ICT and IFAT was excellent with k = 0.86 and 0.92, respectively, and P < 0.0001.

Discussion
Given the fact that VL has a fatal outcome if not treated, its early and accurate diagnosis is necessary. Immunochromatographic and ELISA tests, based on a recombinant antigen containing 39 amino acid repeats of a kinesin-related protein of a Brazilian L. infantum strain, are commercially available. Previous studies showed a considerable variation in the sensitivity and specificity of such tests, depending on test format and the geographical distribution of VL with reported sensitivities between 67.6-100% and specificities between 59-100% for rK39-ICTs. The rK39-ELISA has been shown to have sensitivity and specificity ranges between 88.6-100% and 84.5-100%, respectively [17]. A prospective study conducted in Iran on 17 VL patients and 137 patients with other diseases reported a sensitivity of 82.4% and specificity of 100% for the InBios rK39-ICT [28]. Such results imply that due to the suboptimal diagnostic accuracy of rK39-based tests, other diagnostic tests such as IFAT or quantitative PCR need to be performed, which are time-consuming, require equipment and well-trained personnel. Bhattacharyya et al. [19] illustrated an extensive diversity of K39 among various strains of L. donovani which could affect the accuracy of rK39-ICTs. Indeed, in the present study, we confirm that there is also diversity in the K39 gene within an Iranian L. infantum isolate. Cloned 550-bp fragments of the K39 gene from the Iranian isolate MCAN/IR/14/M14 showed that this locus is heterozygous, carrying a sequence identical to the one described for Brazilian L. infantum, and also a variant that differed by 6% with the amino acid sequence of the Brazilian K39 sequence. Surprisingly, these amino acid substitutions were mostly found at the same positions to those noted in the L. donovani study by Bhattacharyya et al. [19], yet, not all substitutions were same as the ones in L. donovani. The observed difference in protein sequence within this limited range (3.5 repeats) might already affect the performance of ICTs. Therefore, we chose this variant, called Li-rK39, for expression in L. tarentolae.
In the present study, the ELISA with the purified recombinant Li-rK39 fragment was more sensitive (96%). One of the four false negative serum samples in the ELISA was also negative in ICT. The reason that some IFAT-positive VL patients remained negative in the Li-rK39 ELISA is not clear. Elfadi et al. [18] pointed out that it is hard to select a diagnostic test based on an antigen of a single strain of Leishmania in places where a population of heterogeneous parasites exists. This is illustrated by a study conducted in India, where a sensitivity of 100% was reported for an ELISA using an rK39 derived from an Indian isolate of L. donovani KE16 [24]. It may therefore be interesting to investigate the heterogeneity within the kinesin-related K39 gene of L. infantum isolated from a human patient in a given area, and to express the dominant sequence as a recombinant antigen for diagnostic purposes. Furthermore, it should be kept in mind that some cases of VL in Iran are caused by L. tropica [29,30]. A study previously conducted in Iran on 12 parasitologically confirmed VL patients reported a sensitivity of 91% for ELISA with a recombinant antigen consisting of one 39 amino acid repeat of the K39 of an Iranian L. infantum strain, expressed in E. coli [31]. Compared to our study, the low sensitivity obtained in their study may be due to the use of only one repeat unit of K39, derived from a different L. infantum isolate and expressed in a prokaryotic host.
Another study showed that a higher copy number of K39 repeats in a recombinant antigen derived from L. infantum increased the affinity of antibodies to the antigen [32]. In the present study, the sensitivity of Li-rK39-ELISA is greater than the one conducted in Brazil and equal to the one in India which used a commercial rK39 of L. infantum in their ELISA system [24,33]. Also, the sensitivity of the Li-rK39-ELISA was greater than that of rK39-ICT (InBios) (90%). This might be due to the higher analytical sensitivity of the ELISA compared to the ICT format. However, in contrast, we found three VL patients positive in the rK39-ICT that were negative in the Li-rK39-ELISA. This discordance may be due to: (i) small modifications of the original sequence by adding an N-terminal methionine for cytoplasmic expression; (ii) the different solid phases in ELISA (polystyrene) and ICT (nitrocellulose) that may influence the antigen conformation; (iii) the more stringent antibody-antigen binding conditions in the ELISA; and (iv) the fact that we expressed and evaluated only one of the two alleles of the kinesin related protein gene, which is different from the one used to produce the Brazilian L. infantum rK39 in the rK39-ICT. It follows that combining in one test recombinant proteins derived from both alleles may further increase the sensitivity of the ELISA.
The Li-rK39-ELISA revealed a specificity of 100% with the 46 non-endemic controls but 14 of the 190 endemic controls and two of the non-VL patients including cutaneous leishmaniasis and malaria were positive, although with OD below 0.5, resulting in a specificity of 93.8% which is not unexpected since in VL endemic regions a considerable number of infected people remain asymptomatic but develop a detectable antibody response [34].
In the present study, we used the L. tarentolae expression system (pLEXSY) to produce Li-rK39 and have evaluated its diagnostic performance. So far, the rK39 antigen has been produced in prokaryotic expression systems. Post-translational modification can affect an antigen in many aspects like stability, solubility, resistance against proteases etc. but doesn't occur in prokaryotic expression systems. Also, codon bias is an important issue related to expression systems [21,35]. The main advantage of prokaryotic expression systems is their low cost, easy handling and high yield [21,35]. Leishmania tarentolae is a relatively easy eukaryotic expression system but compared to bacteria, it has a lower growth rate (doubling time of 6 h in agitated culture) and lower yield. On the other hand, it can be considered as a suitable expression system for rK39 of L. infantum, so the recombinantly produced protein may be more similar to its native counterpart. All these factors make this expression system suitable for the expression of proteins of kinetoplastid organisms as Rooney et al. [36] have shown in their study.