Isolation of Neurocysticercosis-Related Antigens from a Genomic Phage Display Library of Taenia solium

Construction of a phage display library

All molecular biology procedures were carried out using standard protocols or as recommended by manufacturers. Restriction enzymes, DNA isolation/purification kits, T4 DNA polymerase, T4 DNA ligase, calf intestine alkaline phosphatase (CIAP), and M13KO7 helper phage were obtained from Invitrogen (Carlsbad, CA), Qiagen (Valencia, CA), or GibcoBRL (Rockville, MD). The oligonucleotides were synthesized at GibcoBRL. The T. solium genomic DNA of cysticerci dissected from skeletal muscle of infected pigs was isolated using the QIAamp DNA Kit (Qiagen) and used as described. ¹⁹ Briefly, about 10 µg of chromosomal DNA was fragmented randomly by sonication 5 times for 20 s at maximum power with a sonicator Aquasonic 50T (VWR Scientific Products, West Chester, PA). The DNA fragments (100-600 bp) were made blunt ended using T4 DNA polymerase, purified by the Concert Rapid PCR purification System (GibcoBRL), and used in a ligation reaction with 2 µg of SnaBI digested and dephosphorylated pG8SAET vector DNA, using T4 DNA ligase. ²⁰ The ligated DNA was column purified and used to transform Escherichia coli TG1 cells by electroporation using the Gene Pulser II System (Bio-Rad Laboratories, Hercules, CA). Eight electroporations were performed, and the transformed TG1 cells were plated on LB-Amp plates to determine the diversity of the library. The resultant phagemid library was rescued/amplified using M13KO7 helper phage, purified by double PEG/NaCl (20% w/v polyethylene glycol l-8000; 2.5 M NaCl) precipitation and resuspended in Tris-buffered saline (TBS). The phage yield was 10¹² colony-forming units (cfu) per milliliter of culture medium.

Affinity selection of phage clones expressing T. solium–derived antigens by biopanning and ELISA screening

Six-week-old female BALB/cAnN mice, bred in our animal facilities, were immunized with T. crassiceps soluble antigens (TcAg) as described previously. ²¹ Hyperimmune sera from 3 mice were pooled, diluted 1:200 in phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin (BSA), and used in a biopanning assay. Phage biopanning was carried out essentially as described in our previous studies. ²¹ Briefly, the wells of Nunc Maxisorp microplates were coated with goat antimouse IgG (0.5 µg/well; Zymed, South San Francisco, CA), washed with PBS/Tween 0.05%, and blocked with PBS/2% BSA. Then, mice sera described above were added to 4 wells, and after washing with PBS-Tween, 10¹⁰ phage particles per well from the genomic DNA phage library were added. After incubation for 4 h at 4°C, the plate was washed with cold PBS-Tween, and bound phage particles were eluted using glycine-HCl (0.2 M, pH 2.2) and neutralized by adding Tris-HCl (1 M, pH 9.1). The eluted phages were plated on LB-Amp plates, and individual colonies were rescued/amplified using M13KO7 helper phage and used in ELISA screening. Briefly, the bacterial colonies were used to inoculate 96-well cell culture plates (Costar, Corning, NY) containing 150 µL of 2-YT-Amp. Plates were incubated for 4 h at 37°C, and helper phage was added to each well. After 30 min of incubation without shaking and 30 min of incubation with shaking, plates were centrifuged and supernatants were removed. Then, 150 µL of 2-YT-Amp-Kan was added to the cells, and the plates were incubated overnight at 37°C. The supernatants from plates containing the phage were used in ELISA screening with mice anti–T. crassiceps sera as described below. Also, DNA was purified from these clones, and the DNA inserts were amplified by PCR using pG8SAET vector-based primers and analyzed on agarose gel.

Cerebrospinal fluid and plasma samples

Seventeen CSF and 22 plasma samples from patients with NCC confirmed by CT and MRI were obtained at the Instituto Nacional de Neurologia y Neurocirugia, Mexico City. All samples were positive in ELISA with T. solium cyst fluid antigens (TsAg). CSF samples from 18 patients with encephalitis and plasma from 7 healthy individuals were used as negative controls.

Enzyme-linked immunosorbent assay

ELISAs were carried out as previously described. ^5,6,19 Briefly, 10¹⁰ cfu of amplified phage was added to microplates (Nunc), incubated for 1 h at 37°C, and washed with PBS and PBS-Tween 0.1%. After blocking with PBS–2% BSA, CSF and plasma samples diluted at 1:40 and 1:100 in PBS–1% BSA, respectively, were added to the wells and incubated overnight at 4°C. Horseradish peroxidase (HRP)–conjugated antihuman IgG diluted in PBS–1% BSA (1:5000) was added to the wells and incubated 1 h at 37°C. The plates were washed, and ABTS single solution (Zymed) was added. The OD readings at 405 nm were registered using the Opsys MR Microplate Reader (DYNEX Technologies, Inc., Chantilly, VA). The Ig subtypes present in plasma samples from NCC patients were determined using recombinant phage clones and HRP-mouse antihuman IgG1, IgG2a, IgG3, IgG4, and IgE Abs (Zymed). All samples were tested in duplicate, and the reported data are representative of 3 independent experiments. Similarly, the plates were coated with TcAg (1 µg/well), and ELISA was performed using samples from NCC patients as described above. For ELISA screening, the plates were coated with phage particles and incubated with mice hyperimmune sera, and HRP/mouse anti-IgG conjugate was added to the wells.

DNA sequencing

The DNA sequences of the inserts of phage clones were determined using automated the ABI Prism 310 Genetic Analyzer (Applied Biosystems, Foster City, CA), miniprep-purified double-stranded DNA, and a pair of pG8SAET vector-based primers 5′(ggaaaaacatttattcaattcgtaaactag) and 3′(ggtttatcagcttgctttcgaggtgaatttc). The DNA and deduced amino acid sequences were analyzed with the ExPASy Molecular Biology server and BLAST database. The DNA and amino acid sequences of isolated clones were submitted to the GenBank database.

Statistical analysis

The correlation between variables was verified by using the chi-square test for contingency tables.

Generation of T. solium genomic DNA phage display library and biopanning with hyperimmune mouse sera

To generate the genomic DNA phage library, T. solium genomic DNA was randomly fragmented by sonication (100- to 600-bp DNA fragments) and cloned in pG8SAET phagemid vector as 5′ gene fusions with M13 gVIII. A genomic expression library with the complexity of about 1.25 × 10⁷ independent clones was generated, and the rescue/amplification of the library resulted in a phage library of 10¹² clones.

To isolate phage clones specifically recognizing Abs in the hyperimmune sera, a 2-step selection strategy consisting of biopanning and ELISA screening was applied. Then, 4 × 10¹⁰ phage particles from the library were biopanned against the hyperimmune sera from mice immunized with TcAg. Forty individual clones were randomly picked from the library after the first round of selection and used in an ELISA screening with the mouse sera. This resulted in isolation of 11 positive clones. PCR and NcoI/BamHI enzyme digestion analysis showed that all clones were carrying 100- to 400-bp DNA inserts (data not shown). Three of 11 clones were selected based on their highest reactivity in ELISA with 5 CSF samples from NCC patients and further characterized. The DNA and the deduced amino acid (aa) sequence analysis of these clones revealed that 2 clones expressing a 54 aa peptide (clone EK26, GenBank accession number AY699707) were identical, and the clone designated EK5 (GenBank accession number AY699706) carried a 63 aa peptide. Interestingly, for the correct expression of identified open reading frames, a –1 shift at 3′ of the DNA insert in EK5 and a +1 shift at both ends of the insert in EK26 are required. This is not unusual since the same phenomenon was regularly observed, possibly as a result of ribosomal slippage, in studies where the expression system and vector used in our study were analyzed. ^19,22 Sequence similarity search in the public sequence databases revealed a homology (36% of identical and 68% of positive aa in a 41 aa long stretch) between EK5 and a hypothetical protein of intestinal parasite Giardia lamblia (G. lamblia ATCC 50803; hypothetical protein GL50803_112258, GenBank), whereas EK26 did not show any significant homology to reported sequences.

Immunoreactivity of phage clones

For the evaluation of the antigenic properties of isolated phage clones, their immunoreactivities were tested in an ELISA using 17 CSF and 22 plasma samples from patients with NCC confirmed by CT and MRI. As shown in Figure 1 , 9 of 17 (52.9%) CSF samples from NCC patients reacted with the phage clones EK5 and EK26, whereas only 5 and 4 of 18 CSF samples (27.8% and 22.2%) from patients with encephalitis were recognized by EK5 and EK26, respectively. EK5 and EK26 recognized 36.4% and 18.2% of plasma samples from NCC patients, respectively, whereas both phage clones did not react with plasma samples from 7 healthy individuals ( Fig. 1 ). The examination of ELISA data showed a statistically significant correlation (p < 0.05) between the reactivity against clone EK26 and the presence of arachnoiditis ( Fig. 2A ), as well as between the reactivity against clone EK5 and the presence of hydrocephalus ( Fig. 3A ), whereas the same samples were equally recognized by TsAg ( Fig. 2C and Fig. 3C ). Also, phage EK26 did not recognize preferentially antibodies present in plasma from NCC patients with hydrocephalus, and phage EK5 did not preferentially recognize Abs present in plasma from NCC patients with arachnoiditis ( Fig. 2B and Fig. 3B ).

OPEN IN VIEWER

Fig. 1.

Enzyme-linked immunosorbent assay binding of phage clones EK5 and EK26 to antibodies present in cerebrospinal fluid (CSF) samples from 17 patients with confirmed neurocysticercosis and 18 patients with encephalitis and plasma samples from 22 patients with neurocysticercosis and 7 control healthy individuals. Each sample was also tested against M13 control phage, and a final OD value is a difference between phage-bearing peptide and control M13. Data are expressed as OD/cutoff values, so values above 1 (dashed line) are positive cases and values below 1 negative cases. The cutoff values were determined as the mean + 2 SD of negative controls. NCC, neurocysticercosis.

OPEN IN VIEWER

Fig. 2.

Enzyme-linked immunosorbent assay binding of phage clones EK26 and EK5 and TcAg to antibodies present in plasma samples from patients with and without arachnoiditis (A-C). A correlation between binding of phage clone EK26 to antibodies in plasma samples and the presence of arachnoiditis was observed (p = 0.0062). Importantly, none of the 12 patients without arachnoiditis bound EK26. No correlation was observed between binding of phage EK5 to antibodies in plasma samples and the presence of arachnoiditis (p = 0.093).

OPEN IN VIEWER

Fig. 3.

Enzyme-linked immunosorbent assay binding of phage clones EK5 and EK26 and TcAg to antibodies in plasma samples from patients with and without hydrocephalus (A-C). A correlation between binding of phage clone EK5 to antibodies in plasma samples and the presence of hydrocephalus was observed (p = 0.0023). All 5 patients with hydrocephalus were positive, whereas only 3 of 14 patients without hydrocephalus bound EK5. No correlation was observed between binding of phage EK26 to antibodies in plasma samples and the presence of hydrocephalus (p = 0.063).

The IgG subtyping showed that the predominant subtype present in plasma samples from NCC patients was IgG1 (data not shown).