Introduction

Human papillomavirus (HPV) infection is a prominent concern in both the research and medical fields. Although HPV has been associated with many diseases, cervical cancer has particular significance, being the principal cancer in women in most developing countries. To date, more than 100 types of HPV have been identified. They are divided into two groups. Low-risk types, such as HPV-6, -11, -34, -42 and -44, are associated with the development of benign lesions. High-risk types, such as HPV-16, -18, -31, -33, -45, and -58, can give rise to pre-cancerous lesions and cancer [1]. Two of these HPV types, HPV16 and HPV18, account for 60–70 % of all cervical cancer cases worldwide [1, 2].

HPVs are small DNA viruses, and typically infect hosts via squamous or mucosal epithelia, causing benign and malignant epithelial neoplasia [3]. HPV capsids consist of major (L1) and minor (L2) capsid proteins, synthesized late in the infection cycle, which encapsidate the histone-associated, closed circular and double-stranded DNA mini-chromosome [4]. L1 protein alone or together with the L2 protein can self-assemble into virus-like particles (VLPs) when expressed in a recombinant expression system. As VLPs are structurally and immunologically similar to the native virions, VLPs are able to induce high titres of neutralizing antibodies (IgG and IgA) and effectively protect animals [5–7] and humans [8–12] from papillomavirus infections.

Different methods have been developed to quantify the type-specific neutralizing antibodies for HPVs, such as pseudo-neutralization assays [13] and competitive radioimmunoassays [14]. Each method has limitations concerning time and sensitivity. For this reason, the Luminex (L1)-based competitive immunoassay [15, 16] was developed by Invitrogen (Carlsbad, CA) and used to quantify different type-specific neutralizing antibodies in a single reaction simultaneously. In the assay, different types of VLPs were bound with different kinds of microspheres, then the known, HPV type-specific, biotin-labeled polyclonal antibodies were mixed with VLP-coupled microspheres and the vaccine challenged serum samples for reaction, where the biotin-labeled neutralizing antibodies compete with the serum antibodies to bind to the conformational epitopes on VLPs. After that the microspheres were put into Luminex system for detection of the fluorescence signal.

To develop L1-based immunoassay, it is necessary to obtain large quantities of a pure and active form of the HPV L1 proteins. It has been reported that HPV L1 proteins can be produced in soluble and active form along with glutathione S-transferase (GST) tag in Escherichia coli [17, 18]. In this study, we report that the combination of GST fusion and late log phase culture induction can be used as a means for boosting the solubility and yield of HPV L1 proteins in a bacterial-based expression system. Furthermore, we describe the development of a VLP-based immunoassay based on our previous study [19, 20] and validated the assay to ensure it was sensitive, specific and robust enough to meet the requirement for detection of HPV 16 and 18 antibodies in clinical serum samples.

Materials and Methods

Construction of Recombinant E. coli for HPV L1 Production

The bacterial strains and plasmids used in this study are presented in Table 1. E. coli BL21 and Rosetta (DE3) served as expression hosts for production of HPV L1 proteins, while E. coli DH (5α) was used for plasmid cloning and maintenance. The cells were cultured in Luria–Bertani (LB) medium containing 50 mM 3-(N-morpholino) propanesulphonic acid (MOPS), pH 7.0 at 37 °C with vigorous shaking. When needed, 100 μg/mL ampicillin and/or 20 μg/mL chloramphenicol were added. Growth was monitored at an optical density of 600 nm.

Table 1 Strains, plasmids and primers used in this study
Full size table

Plasmids pBR322-H16 and pBR322-H18 encoding the HPV16 L1 (NCBI gene ID: 1489082) and HPV18 L1 (NCBI gene ID: 1489090) genes, respectively, were kindly provided by Prof. E.M de Villiers (DKFZ, Heidelberg). The L1 genes were amplified by polymerase chain reaction (PCR) and cloned into pGEM-T (Promega, Madison, WI, USA). For construction of the GST-tagged HPV, the primers indicated in Table 1 were used (where the underlined sequence indicates the BglII and SalI restriction site for HPV16 and the EcoRI and SalI restriction site for HPV 18). The amplified PCR fragment was ligated into the pGEM-T vector and transformed into E. coli DH (5α) by electroporation. The resulting plasmids were then digested with the proper restriction enzymes and the each restriction fragment was sub-cloned into the pGEX-4T-3 (GE Healthcare, Buckinghamshire, UK) to form pGEX-HPV16 and pGEX-HPV18 expression vector. The integrity of the HPV L1 genes was confirmed by sequencing.

Cultivation of Recombinant E. coli for HPV L1 Production

For the production of GST-tagged L1 proteins, colonies of Rosetta carrying pGEX-HPV were grown overnight on LB-Amp-Cm plates and were used to inoculate 1 L of prewarmed LB containing 100 μg/mL ampicillin and 20 μg/mL chloramphenicol. The cell cultures were incubated at 37 °C with vigorous shaking. At an optical density of 2.0 at 600 nm, the cultures were rapidly chilled to 20 °C, induced with 0.1 mM isopropyl-beta-d-thiogalactopyranoside (IPTG), and incubated for another 6 h, after which cells were harvested by centrifugation.

Purification of GST-tagged HPV L1

To purify HPV L1 proteins, cells were harvested after overnight induction by centrifugation at 8,000 rpm for 15 min. The cell pellet was washed twice with 1× phosphate buffered saline (PBS; pH 7.3) containing 140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4 and 1.8 mM KH2PO4, and cells were resuspended in the same buffer. Resuspended cells were disrupted using Misonix 3000 sonicator (Misonix, Farmingdale, NY, USA) for 12 cycles, at 600 W for 5 s per cycle. The cell lysate was centrifuged at 15,000 rpm and 4 °C for 30 min to remove the particulate fraction. To purify the GST-tagged HPV L1 proteins, the soluble fraction was subjected to GST affinity column chromatography (732-1010; Bio-Rad, Hercules, CA, USA) in 1× PBS. The GST fusion proteins bound to the column and impurities were removed in washing buffer (1× PBS). Fusion proteins were then eluted with the elution buffer (50 mM Tris–HCl and 10 mM reduced glutathione, pH 8.0) under mild, non-denaturing conditions that preserved both protein antigenicity and function. To obtain the native HPVs, the GST tag was removed by incubation at 4 °C with PreScission Protease Factor Xa (Promega). The free GST tag was then removed from solution by GST affinity column chromatography and the HPV L1 proteins were obtained from the flow-through fractions.

Coupling of Glutathione to Casein

The procedure for coupling of glutathione to casein (GC) has been described [21]. Casein (Sigma-Aldrich, St. Louis, MO, USA) at a concentration of 150 mg in 25 ml PBS was incubated for 15 min at room temperature (RT) with 0.5 ml of 0.4 mM N-ethylmaleimide (NEM; Sigma-Aldrich) to block the single cysteine residue in casein. Thereafter, 50 mg sulfosuccinimidyl 4-[p-maleimidophenyl]butyrate (Sulfo-SMPB; Pierce, Rockford, IL, USA) was added as crosslinker and the reaction proceeded for 30 min at RT. Free Sulfo-SMPB and NEM were separated from casein by size exclusion chromatography on PD10 columns (Pharmacia, Freiburg, Germany). The protein fraction was then supplemented with 10 mM glutathione (Sigma-Aldrich) and the coupling reaction was executed for 1 h at RT. The glutathione casein was separated from unbound glutathione by gel filtration with PD10 columns using PBS as buffer and stored at −20 °C in small aliquots.

Covalent Conjugation of HPV VLPs to Luminex Microspheres

The bead stock was vortexed and 0.1 ml of the homogeneous bead solution was transferred to a 1.5 ml Eppendorf-tube and centrifuged 2 min at 6,000 rpm. The supernatant was discarded and 500 μl activation buffer (0.1 M sodium phosphate, pH 6.2) was added. The mixture was centrifuged for 2 min as before and the supernatant was discarded. Two hundred microliters of activation buffer was added immediately prior to use, followed by 25 μl of a solution containing 50 mg/ml N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide (EDC; Pierce) and 50 mg/ml N-hydroxysuccinimide (NHS) in water-free dimethylsulfoxide (DMSO). The suspension was vortexed it and incubated at RT with mixing at 250 rpm for 20 min in the dark. The activated beads were washed twice with 0.5 mL of coupling buffer (50 mM 2-morpholinoethanesulfonic acid, pH 5.0) and resuspended in 0.25 mL of a 250 μg/ml GC solution in coupling buffer. Coupling was carried out on a shaker at RT for 2 h in the dark. After the coupling procedure, the beads were incubated for 15 min in 0.5 ml of washing buffer 1 (PBS containing 0.05 % Tween 20 and 50 mM Tris, pH 7.4) to block unreacted carboxyl groups with primary amines. The beads were then washed twice with 0.5 ml of washing buffer 2 (PBS containing 0.05 % Tween 20, pH 7.4) and stored in 100 μl of storage buffer (PBS containing 1 mg/ml casein and 0.05 % sodium azide, pH 7.4) at 4 °C in the dark. For coupling, 3 × 103 GC beads and purified HPV L1 proteins were mixed in casein buffer (1 mg/ml casein in PBS, pH 7.4) to a final volume of 1 ml. The mixture was rotated in the dark at 250 rpm at RT for 1 h. The beads were then washed thrice with 0.5 ml of casein buffer and resuspended in 100 μl of the same buffer.

Multiplex Immunoassay

Bead sets carrying different antigens were mixed, and 50 μL aliquots of preincubated diluted serum and mixed beads (3,000 per set) were combined in wells of 96-well plates with filter bottoms (Millipore, Billerica, MA, USA) and incubated on a shaker for 1 h at RT in the dark. The beads were washed thrice in 100 μL of casein buffer on a vacuum manifold (Millipore). Biotinylated secondary antibody [goat anti-human IgA, IgM, IgG (H + L); Thermo, Rockford, IL, USA] diluted 1:1,000 in casein buffer was added and incubated as before. After washing, detection conjugate (streptavidin-R-phycoerythrin) diluted 1:1,000 in casein buffer was incubated with the beads for 30 min. The beads were washed again, and the wells were filled with casein buffer. Reporter fluorescence of the beads was determined with the Luminex analyzer (Invitrogen, Carlsbad, CA, USA) and expressed as median fluorescence intensity (MFI) of at least 50 beads per set per well. To calculate antigen-specific reactivity, the MFI of GST tag was subtracted from the antigen MFI.

Electron Microscopy

The structure of the VLPs was studied by transmission electron microscopy (TEM). A sample was applied to a 200-mesh carbon-coated copper grid, dried, and stained with 2 % phosphotungstic acid. TEM was performed using a model 100CX transmission electron microscope (JEOL, Tokyo, Japan) at 40 kV with a final magnification of ×200,000.

Western blot

Proteins was loaded and separated by 12 % sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The separated proteins were electroblotted to a PROTRAN nitrocellulose membrane (Scheicher & Schuell, Dassel, Germany), and the blots were washed three times with Tris buffered saline-Tween (TBST; 100 mM Tris–HCl, 0.1 % Tween 20). The membranes were blocked with blocking solution (5 % skim milk) for 2 h at RT, further developed with a 1:1,000 dilution of primary rabbit anti-SV40 KT3 polyclonal antibody (American Research Products, Belmont, MA, USA) for 1 h, and then incubated with a 1:5,000 dilution of secondary antibody (goat anti-mouse IgG) for 1 h. Membranes were washed thrice with TBST and developed using enhanced chemiluminescence (ECL) western blotting detection reagents (Amersham Pharmacia Biotech, Piscataway, NJ, USA).

HPV 16 and 18 Antibodies

For anti-HPV16 and anti-HPV18 polyclonal antibodies production, HPV-16 peptide (377–391 L1 amino acid: SETTYKNTNFKEYLR-C) and HPV-18 peptide (498–510 L1 amino acid: C-ENKDPYDKLKFWN) were coupled to carrier protein bovine serum albumin via a Cys residue in the N or C terminus. The coupling was performed with m-maleimidobenzoic acid-N-hydroxysuccinimide as the linking agent. The efficiency of coupling was established after dialysis by amino acid analysis. For rabbit immunizations, each dose contained 1 mg of coupled peptides emulsified with Freund’s complete or incomplete adjuvant for first or second subcutaneously injection, respectively. After two injections, the sera were tested in an enzyme-linked immunosorbent assay (ELISA) for titer of antibody.

Clinical Samples

All tested serum samples were collected at the Catholic Medical University, Seoul, Korea. A total of 22 samples were tested. Serum samples were from women, 10–25 years-of-age who had not been vaccinated against HPV vaccination or enrolled in phase I HPV VLP L1 vaccine clinical trials. Eligibility criteria required that volunteers be in good health and have no history of genital warts or abnormal cervical cytology. This study was conducted in accordance with the Good Clinical Practice guidelines, and the experiment protocol and informed consent documentation were approved by the Independent Human Research Ethics Committee or Institutional Review Board of each study centre. All participants provided written informed consent/assent prior to conduct and performance of any study-related procedures.

The cutoff value to define antibody-positive sera was calculated separately for each antigen as the median fluorescence values of all control sera plus three standard deviations excluding positive outliers. Briefly, control sera with MFI values higher than the calculated cutoff value were omitted and the calculation was repeated with the remaining sera. This procedure was repeated until the absorbance values of all remaining sera were below the last calculated cutoff value, which was used to judge sera as antibody-positive or -negative.

Analyses

Cell concentrations were measured in a 10-mm path length cuvette using an Ultrospec 3000 single-beam spectrophotometer (Pharmacia). One unit of absorbance at 600 nm corresponded to 0.3 g dried cell mass per liter. Band intensities on DNA or SDS-PAGE images were determined using a Gel-Doc instrument (Bio-Rad). Protein concentrations in cell-free extracts were determined by the Bradford assay [22] using bovine serum albumin as a standard on an ELISA microtiter plate reader (Bio-Tek, Winooski, VT, USA).

Results and Discussion

Production and Purification of HPV 16 and 18 L1 Proteins

To date, the expression and purification of HPV L1 proteins have been reported using the baculovirus expression system and Saccharomyces cerevisiae expression system [23, 24]. However, the use of bacterial expression systems offers distinct advantages in relation to vaccine development, which include cost-effective VLP production, high levels of product yield, and relative ease of purification. In addition, a bacterial expression system is much more convenient than any other systems; advantages include cell growth rapidity, simplicity of growth medium, low cost of growth medium and high expression level, even though limits posttranslational modification. Some prophylactic HPV vaccines expressed in yeast and baculovirus expression systems have been commercialized for use in humans. However, the costs of the expression and purification in these systems are exorbitant. In addition, the production efficacy is lower than that of bacteria. To overcome these problems, we presently developed a bacterial expression system suitable for the industrial production of HPV vaccine candidates.

The L1 genes from both HPV16 and 18 were analyzed in parallel to generalize the findings concerning whether a good expression level could be obtained in a normal E. coli expression host strain, such as E. coli BL21. However, it was found that codon adaptation index (CAI) for these L1 genes in E. coli varied from 0.62 to 0.65, which indicated that the number of rare codons was huge and codon bias would occur when normal E. coli was chosen as expression host, resulting in either poor expression or accumulation in insoluble form. Recently, it was reported that pGEX in Rosetta cells provides an alternative for the expression of L1-based antigens from cutaneous species alpha HPVs [25]. Herein, we fused GST tag to the L1 amino N-terminus using the pGEX-4T-3 vector and transformed the recombinant plasmid into Rosetta (DE3) to maximize the production of soluble and active HPV L1 proteins.

In general, the solubility and yield of target proteins can be affected by the host strain, IPTG concentration and duration and temperature. These factors should be systematically varied to optimize expression for the production of proteins. Among these factors, induction temperature may exert great influence on the target protein solubility. When expressed at 37 °C, all the GST-L1 fusion proteins were detected by SDS-PAGE with whole cell lysates but most of them were found in the insoluble pellet for both E. coli and Rosetta. When induction was carried out at 20 °C, protein expression was significantly improved (Fig. 1a). However, L1 protein fractions decreased over time when the induction time was more than 6 h, which might be partially due to protein degradation by native protease in the host strain. To optimize the production condition, various conditions, such as different temperature and IPTG concentration as well as incubation time, were carefully studied (data not shown).

Fig. 1
figure 1

SDS-PAGE and western blot for HPV 16 and 18 L1 proteins: a production of HPV16 L1 at 20 and 25 °C in Rosetta. Lanes 1, 4, 7, and 10: total cell lysate; lanes 2, 5, 8, and 11: soluble fraction of HPV L1 protein; lanes 3, 6, 9, and 12: insoluble fraction of HPV L1 protein. b Comparison of HPV16 L1 production at early and late induction in E. coli and Rosetta. Lanes 1, 5, 8, and 12: total cell lysate; lanes 2, 6, 9, and 13: soluble fraction of HPV16 L1 protein; lanes 3, 7, 10, and 14: insoluble fraction of HPV16 L1 protein; lanes 4 and 11: protein marker. c Cleavage of GST-tagged HPV16 L1 protein by Factor Xa. Lane 1 protein marker, lane 2 GST-tagged HPV16 L1 protein, lane 3 native HPV16 L1 protein, lane 4 GST tag. d Western blot of HPV 16 L1 and HPV 18 L1 proteins. Lane 1 GST-tagged HPV16 L1 protein, lane 2 GST-tagged HPV18 L1 protein, lane 3 native HPV16 L1 protein, lane 4 native HPV18 L1 protein

Full size image

Most existing protocols for expressing recombinant protein recommend inducing protein expression when the cell culture is in mid log phase (i.e., OD600 = 0.3–0.6). This is presumably because the cultures are growing rapidly and protein translation is maximal. In this study, when the cells were induced at low OD600 (0.5) at 37 °C, in which ~90 % HPV16 L1 protein was accumulated as inclusion bodies in E. coli and 70 % insoluble L1 was found in Rosetta, consistent with previous studies on the production of HPV L1 proteins in bacterial expression platforms [25, 26]. The most unexpected finding in this study was the observation that expression of HPV L1 proteins in late log phase cultures yielded higher recovery of soluble target protein than that obtained from mid log phase cultures (Fig. 1b). It was estimated that more than 60 % of HPV16 L1 protein was in the soluble fraction when the cells were induced at high OD600 (2.0) for a short time in Rosetta. Importantly, the aforementioned results demonstrated that the increased yield of soluble HPV L1 protein from late log phase culture induction was not merely due to the greater number of cells. This is most evident in the comparison of the relative recovery of HPV 18 L1 protein (data not shown) as insoluble and soluble fractions from cultures induced at 20 °C at an OD600 of 0.5 or 2.0. Late log phase cells sequestered relatively less of the total expressed protein in inclusion bodies. In late log phase, cells must have undergone a metabolic and growth shift that may have attenuated the response to foreign and potentially toxic proteins.

L1 proteins decreased over time when the induction time exceeded 6 h, which might have been partially due to protein degradation by native protease. Consistent with this speculation, in the presence of 0.1 mM IPTG during induction at 20 °C, when cell density reached 2.0, and further incubation for 6 h, GST fusion protein was most abundantly produced. Consequently, these conditions were used throughout this study. Furthermore, when the soluble fraction of cell lysate was partitioned and chromatographed on a glutathione-sepharose column, the GST fusion proteins were retained on the column, as shown by the detection of an ~80 kDa band upon SDS-PAGE (Fig. 1c). After purification, both GST-tagged HPV16 L1 and GST-tagged HPV18 L1 were successfully cleaved to obtain authentic L1 proteins, which were confirmed by Western blotting using anti-SV40 KT3 antibody (Fig. 1d).

Renaturation of HPV 16 and 18 L1 Proteins

The yield of the fusion L1 proteins expressed in Rosetta cells after IPTG induction was estimated from SDS-PAGE band intensity. HPV L1 proteins were determined as comprising 12–16 % of the total cellular protein. Around 30 % of the L1 proteins accumulated in the cells as inclusion bodies that could be further solubilized by protein renaturation treatment. All the L1 insoluble fractions were isolated and treated with 8.0, 6.0, 4.0, and 2.0 M urea. L1 proteins were effectively solubilized in 8.0 M urea, partially solubilized in 6.0 and 4.0 M urea and remained insoluble in 2.0 M urea (data not shown). To renature the purified L1 proteins, the high concentrations of denaturant (8 M urea) and reducing reagent (10 mM dithiothreitol) were removed by a stepwise dialysis procedure in which four 6 h dialysis steps were performed against buffer A containing 4.0, 2.0, 1.0, and 0 M urea. All steps in the renaturation of L1 proteins were performed at 4 °C. Any aggregated particles formed during the dialysis were removed by centrifugation.

In vitro Assembly of Bacterially Expressed HPV L1 Proteins into VLPs

The purified HPV L1 proteins were pooled up to 1 mg/ml in 0.2 M sodium acetate buffer (pH 5.2) containing 1 M sodium chloride and further incubated at RT for 30 min. Then, VLP development was observed by electron microscopy. The purified L1 proteins were most stable between pH 6.0 and 8.0. Outside this range, the proteins tended to aggregate in an irreversible manner. TEM showed that the VLPs were regular spherical particles with a diameter of 20–45 nm, quite consistent with previous studies [26, 27]. It was confirmed that the VLPs could be assembled in vitro from the purified L1 soluble fraction proteins without any renaturation process (Fig. 2). In addition, the electron microscopy results indicated that the L1 proteins renatured from the inclusion bodies were also able to form VLP structures, but that the solubilization process was required to restore the native folding of the protein and the assembly VLPs.

Fig. 2
figure 2

Electron micrograph of HPV 16 (a) and 18 (b) L1 in vitro self-assembly. The bar length represents 50 nm

Full size image

Optimization for Conjugation between HPVs and Microspheres

In the assay, each type of VLP was coupled with one set of distinct fluorescent Luminex microspheres. The type-specific HPV VLP antibody responses are associated with specific Luminex microspheres that are identified by their distinct red and orange fluorescent dye spectral properties on the Luminex instrument. Known, HPV type-specific and biotin-labeled neutralizing antibodies were mixed with VLP-coupled microspheres and the vaccine challenged serum samples for reaction, where the biotin-labeled neutralizing antibodies competed with the serum antibodies for binding to the conformational epitopes on VLPs. After that, the microspheres were put into the Luminex system to detect the luminescence signal. Different ratios (μg/106 beads) of HPV L1 16 and 18 microspheres were used for coupling. The VLP-coupled microspheres were then reacted with the same amount of related neutralizing antibodies. The coupling ratio at which the VLP conjugated beads could generate the strongest signal was selected as the optimal ratio.

To determine the optimal concentration of VLP, microspheres (3 × 103 per reaction) were coupled with 500 μl of a 0.05, 0.5, 5, or 50 μg/ml solution of HPV 16 or 18 L1. Each VLP-microsphere preparation was then tested with the same reference serum standard preparation using the optimized antibodies concentration of 0.1 μg/ml. The optimized conditions for HPV 16 and 18 are shown in Table 2. The VLP L1 concentrations that achieved acceptable sensitivity while maintaining robust assay performance was determined to be 5 μg/ml for both HPV 16 and 18. The coupling reaction for VLP and beads was saturated when coupling ratio exceeded 40 μg protein/106 beads, while for VLP-18 the optimal coupling ratio was 50 μg protein/106 beads.

Table 2 Optimized conditions in the immunoassay
Full size table

HPV 16 and 18 L1 Antibody Standard Curve

Before analyzing human serum samples, we tested the Luminex-based immunoassay for sensitivity and specificity using antibodies against the HPV 16 and HPV 18 L1 proteins. The specific spectral profiles of the Luminex microspheres and the HPV type specificity of the detection antibody permitted the simultaneous measurement of type-specific antibodies to neutralizing epitopes on HPV16 and 18. As shown in Fig. 3, with HPV L1 proteins, biotin-labeled antibody for HPV 16 could only bind to HPV16 L1 coupled microspheres for signaling, while biotin-labeled antibody for HPV 18 only bound to HPV18 L1 coupled microspheres. These results indicate that the type-specific antibodies were capable of neutralizing the respective HPVs and did not cross-react with the other HPV genotype. The optimized conditions were adopted for standard curve preparation. The coupling ratios for VLPs to microspheres were both 5 μg protein/106 beads. The standard curves in the Luminex-based immunoassay were established successfully with good fit (R 2 > 0.98) for both HPV 16 and 18 within this dynamic range (0–200 ng/mL HPV neutralizing antibody; Fig. 4).

Fig. 3
figure 3

HPV 16 and 18 L1 polyclonal antibody specificity. A single Ab-biotin was incubated with HPV16, and 18 L1 microspheres. Detection of Ab-biotin to any of the two VLP types revealed any cross-reactivity. Detection of fluorescence from the biotin-labeled Abs is shown in median fluorescence intensity (MFI)

Full size image
Fig. 4
figure 4

Standard curve for Luminex-based immunoassay. a Standard curve for HPV16 neutralizing antibody. b Standard curve for HPV18 neutralizing antibody. Detection of fluorescence from the biotin-labeled Abs is shown in median fluorescence intensity (MFI)

Full size image

HPV Clinical Serum Sample Test

The good specificity of the immunoassay allowed us to examine individual HPV clinical samples using HPV type-specific antibody for positive control and normal serum as the negative control to set a serum status cutoff (Table 2). In total, we tested a panel of 22 human sera in duplicate in the assay. The clinical sera were from HPV-negative individuals who had not been vaccinated against the virus, and from HPV-positive individuals enrolled in a trial to receive divalent vaccines for HPV 16 and 18. In HPV vaccine trials, serum antibodies are used to monitor the response to vaccination. Fifteen negative, five medium, and two relatively high titer samples to both HPV genotypes were confirmed by the immunoassay.

Table 3 present data from a representative set of samples that were negative, low positive or positive for antibodies to HPV16 and 18, showing the correlated relationship between the MFI and the calculated serum titer in the assay. The negative serum samples were chosen from non-vaccinated individuals 10–20 years-of-age, who were unlikely to have been exposed to HPV. As expected, the MFI values in the non-vaccinated group were low (<81). Meanwhile, higher MFI values were observed in the vaccinated group, entirely consistent with the expected elevated antibody titers after vaccination. A diversity of antibody titer was observed in the vaccinated, which might be related to the difference in each individual’s immunity. These vaccine-induced immune responses must be accurately measured without interference from the immune response to other HPV types. Using the multiplex Luminex immunoassay, antibody titers to neutralizing epitopes on HPV 16 and 18 virions can be simultaneously measured in the serum of vaccinated or naturally infected subjects.

Table 3 Clinical samples test in the immunoassay
Full size table

The collective data indicates that this immunoassay appears fit for its intended purpose of measuring HPV neutralizing antibodies following natural HPV infection or vaccination. Especially, it may play an important role in post-vaccine surveillance in populations.

Concluding Remarks

In summary, we describe the development of a cost-effective HPV L1 bacterial-based production platform. The combination of GST fusion and late log phase culture induction could be used as a strategy for boosting the solubility and yield of HPV L1 proteins in a bacterial expression platform. Obtained HPV L1 recombinant proteins can be coupled to Luminex beads and used for analysis of HPV-specific antibodies in clinical samples. In addition, the characteristics of the immunoassay were optimized and the absence of cross-reactivity confirmed. This immunoassay was proven to be sensitive, specific and robust enough to meet the needs of testing samples from clinical serum samples. Future work will focus on development of low- and high-risk human HPV reference sera and a formal validation with a larger panel of samples from individuals, as well as HPV vaccine clinical trials using this immunoassay platform.