Introduction

Porcine parvovirus (PPV) causes reproductive failure in pregnant sows. The reproductive failure is characterized by embryonic and foetal death, mummification, stillbirths, and delayed return to oestrus [1]. Although acute infection of postnatal, non-pregnant pigs is usually subclinical, PPV has also been linked to the occurrence of skin lesions in piglets [2], interstitial nephritis in slaughter aged pigs, and non-suppurative myocarditis in lactating piglets. More recently, PPV has gained importance as an agent with ability to increase the effects of porcine circovirus type 2 infection in the clinical course of postweaning multisystemic wasting syndrome (PMWS) [3, 4], a significant economical disease worldwide [5]. Because of its association with the abovementioned clinical and pathological conditions, PPV is recognized as an economically important cause of reproductive failure, and vaccines containing inactivated PPV are marketed worldwide.

PPV is a small, non-enveloped, single-stranded, negative-sense DNA virus. Capsids of PPV are assembled from three viral proteins (VP1, VP2, and VP3). VP2 is the major capsid protein and consists of an eight-stranded antiparallel β-barrel motif with four large insertions between β-strands; the insertions (called loops) contain many B-cell epitopes [6]. Epitope-mapping experiments show that all of the epitopes generating neutralizing antibody are within VP2 [6, 7, 8].

Porcine parvovirus was initially described from sows with reproductive problems in Germany [9], but the virus has been subsequently found and studied in many parts of the world, including China, where it causes substantial economic loss. Although there is only one serotype of the virus, PPV can be classified into five clinical pathotypes (biotypes). The nonpathogenic NADL-2 strain, which is currently used as an attenuated vaccine, causes only limited viremia and does not cross the placental barrier in experimental infections [10]. NADL-8 and other virulent strains isolated from mummified and dead fetuses do cause a viremia and can cross the placenta and fatally infect fetuses [11]. A third group of virulent PPV strains, including the Kresse and IAF-A54 stains, has been associated with dermatitis [2] and, in contrast to the other virulent strains, kills immunocompetent fetuses [12]. The fourth group of PPV are the enteric strains, such as IAF-A83. Finally, a kind of PPV, which as yet lacks type strains, has been associated with PMWS and porcine respiratory disease complex (PRDC).

Porcine parvovirus cell or tissue-tropism determinants, host-range determinants, and determinants that confer hemagglutination properties have all been shown to be located in the capsid proteins [13]. The high degree of identity (97%) between the Kresse and NADL-2 PPV strains indicates that any difference in tropism must map to minor genomic differences. The NADL-2 and Kresse strains differ by five amino acids, all of which are located within the VP2-coding region [13].

Because VP2 is the main structural protein of PPV and constitutes most of the viral capsid, VP2 produced in vitro could self-assemble into virus-like particles [14]. These virus-like particles could then be used as a vaccine or as a diagnostic agent to detect the antibody produced by PPV infection or vaccination [15]. Similar set-up in the production and purification of VP2 VLPs using IMAC has been applied successfully for other parvovirus such as for B19 [16] and for infectious bursal disease virus [17]. In order to facilitate the use of VP2 for diagnosis and vaccination, the goal of this study was to find one better procedure to express VP2 in vitro and to purify this fusion protein.

Materials and methods

Materials

Strain 20-06 of PPV was isolated from a dead fetus delivered from a sow that was characterized by reproductive failure. The PPV-positive pig sera and Escherichia coli Rosetta were prepared and stored by our laboratory. Restriction endonucleases, polymerase, and DNA and protein weight markers were purchased from TaKaRa Biotechnology (Dalian) Co., Ltd. (Dalian, China). The plasmid pET-32a (+) was obtained from Novagen (Darmstadt, Germany). HRP-labeled anti-pig serum was purchased from Sigma (St. Louis, Missouri, USA). Ni-NTA His Bind resin was obtained from Invitrogen (Carlsbad, California, USA). Prestained protein ladder was purchased from Fermentas International Inc. (Burlington, Canada).

Plasmid construction

Genomic DNA was extracted from the cell-cultured strain 20-06 of PPV using the classical phenol–chloroform extraction method and was used as a template to amplify the VP2 fragment by PCR. The sense strand primer (5′-TGAGGATCCATGAGTGAAAATGTGGAAC-3′) includes a BamHI restriction site (underlined), and the antisense strand primer (5′-CGCGTCGACTTCTAGTATAATTTTCTTG-3′) includes a SalI restriction site (underlined). The template was denatured at 95°C for 5 min, followed by 30 PCR amplification cycles (30 s at 94°C, 30 s at 30°C, and 72°C for 2 min) and a final extension at 72°C for 10 min.

The cloning strategy for constructing the recombinant plasmid is shown in Fig. 1. The PCR product and plasmid pET-32a(+) were both digested with BamHI and SalI, and then ligated with T4 DNA ligase to yield the construct. The construct was transformed into E. coli, and transformed bacteria were identified using both restriction enzyme digestion and PCR. Further confirmation of transformation was performed by sequencing.

Fig. 1
figure 1

The cloning strategy for constructing the recombinant plasmid

Full size image

Expression

Expression of the VP2 gene was carried out according to conventional protocol. In brief, E. coli Rosetta transformed with pET-VP2 was cultured at 37°C in culture medium supplemented with ampicillin (100 μg/ml). The transformed bacteria were induced by adding IPTG at a final concentration of 1 mM for 5 h at pH 7.0. After the cells were harvested by centrifugation at 4,000×g at 4°C for 20 min, the pellet was suspended in 10 ml buffer (20 mM Tris–HCl, pH 7.4, plus 200 mM NaCl) and then lysed by sonication in an ice water bath. The suspension was then centrifuged at 9,000×g for 30 min at 4°C, and the pellet was kept on ice. The pellet was suspended in 10 ml buffer, 15 μl aliquots were centrifuged, and the pellet was resuspended in an equal volume of 2× SDS loading buffer. The samples (plus a set of protein molecular weight standards in SDS-PAGE sample buffer) were subjected to SDS-PAGE and visualized using Coomassie Brilliant Blue, and then for analysis of the quantities of the fusion protein VP2.

Western blots

Western blots were performed according to standard procedure. Protein samples (suspension of the recombinant pET-VP2, pellet of the recombinant pET-32, protein of transformant with recombinant plasmid before induction) were separated by SDS-PAGE with 12% gel before electrophoretic transfer to a nitrocellulose membrane. Western transfer was carried out in cold transferring buffer (0.025 M Tris and 0.19 M glycine, 20% methanol). The nitrocellulose membrane was then blocked overnight at 4°C with 10% skimmed milk in TBST (Tris-buffered saline with 0.1% Tween 20, pH 8.0). The membrane was washed three times in 10 ml of TBST for 15 min each and then incubated with PPV-positive pig sera at 4°C for 60 min. The membrane was then washed and incubated for 60 min with horseradish peroxidase-conjugated rabbit anti-pig antibody. After further washing, immunoreactive proteins were visualized using DAB.

Purification

After induction by IPTG, the bacterial pellets were resuspended in guanidinium lysis buffer and shaken for 5–10 min at room temperature to ensure thorough cell lysis. The cell lysate was then sonicated on ice with 5-s pulses at high intensity and centrifuged at 3,000×g for 15 min to pellet the cellular debris. The supernatant (10 ml) was mixed with 2 ml Ni-NTA resin in the column at 4°C and mixed slowly for 2 min. The resin was settled by gravity, and the supernatant was carefully removed by aspiration. The column was sequentially washed with 4 ml each of three wash buffers (in order, pH 7.8, 6.0, and 5.3). Each wash was repeated one more time. The fusion protein was eluted by 5-ml denaturing elution buffer (pH 4.0). The expression and purification protocol for His-tagged VP2 is shown in Fig. 2. Samples collected at different elution times, along with a set of protein molecular weight standards in SDS-PAGE sample buffer, were subjected to SDS-PAGE; the proteins were visualized using Coomassie Brilliant Blue.

Fig. 2
figure 2

Diagram of expression and purification protocol for pET-VP2

Full size image

Results and discussion

The High fidelity PrimeSTAR HS DNA ploymerase and fewer than 30 amplification cycles were used to ensure accurate amplification of the VP2 gene. In addition, the cell-cultured virus was harvested 48 h after inoculation, when virus content in cells was high and when most viruses were present as mature virions. The final PCR result showed that the target band of VP2 gene was specific and bright (Fig. 3).

Fig. 3
figure 3

Amplification of the whole sequence of VP2 gene. M DL Marker 2000, 1 the PCR product of VP2 gene, 2 negative control

Full size image

In order to obtain the fusion protein, VP2 was expressed using different expression host bacteria, including E. coli BL21 and Rosetta, and different plasmid vectors, including pET-30a (+), pET-32a (+), and pGEX-6p-1. After selection, the capsid protein of PPV was fused to a polyhistidine tag, and the position of the affinity tag is in N-terminus, and only the fusion protein inserted into pET-32a (+) was expressed well in E. coli Rosetta. Similar to other organisms, E. coli uses only some of the 64 genetic codons. Those codons that organisms use most frequently are called optimal codons, while those that organisms rarely use are called rare or low-usage codons. For E. coli, low-usage codons include AGA, AGG, CGG, and CGA coding Arg; AUA coding Ile; GGA coding Gly; and CCC coding Pro. E. coli Rosetta is derived from E. coli BL21 carrying plasmid pRARE2, which could supply the tRNA corresponding to the rare codon in the E. coli. This tRNA can elevate the expression of foreign genes, especially eukaryotic genes, in the prokaryotic expression system. The sequencing results showed that many rare codons of E. coli occur in the VP2 gene of PPV strain 20-06. These rare codons include 15 AGAs, 18 GGAs, 20 AUAs, and 24 CUAs [18]. Some of the rare codons are consecutive, and this increased the difficulty in expressing VP2 (Fig. 4).

Fig. 4
figure 4

Sequencing result of the pET-VP2

Full size image

We tested a series of expression conditions that differed in induction time, IPTG concentration, and induction temperature. As expected, transformants grew fastest at 37°C. The optimum cell density for pET-VP2 induction was reached at OD600 of 0.5–0.6, and a time course and IPTG concentration study established that optimal protein expression occurred 5 h after induction with 1 mM IPTG. The molecular weight of the target protein was about 82 kDa, which was coincident with the theoretical value (the molecular weight of the VP2 is 64 kDa, while the His-tag is 17 kDa, so the His-tag-VP2 may be 81 kDa). The target protein appeared both in the pellet of the induced pET-VP2 after sonication and in the supernatant of induced pET-VP2 (Fig. 5). The recombinant bacterium produced high quantities of the fusion protein VP2, about 8% in total. Similar set-up in the expression and purification of VP2 VLPs has been applied successfully for others parvovirus such as for B19 and CPV [19]. To the authors’ knowledge, this is the first report of expression of fusion PPV VP2 in E. coli.

Fig. 5
figure 5

SDS-PAGE analysis of recombinant proteins. M low molecular protein marker, 1 the whole bacterium of pET-32a(+) after induction, 2 the control of pET-VP2 before induction, 3 the pellet of induced pET-VP2 after sonication, 4 the supernatant of induced pET-VP2 after sonication

Full size image

In order to confirm the identity of His-tagged VP2, the purified fusion protein was exposed to Western blot assay using PPV-positive pig sera. The polycolonal antibodies recognized His-tagged VP2, and the band has the appropriate molecular weight. Immunoblot of these membranes using anti-PPV antibodies showed that the fusion protein had epitopes derived from PPV (Fig. 6).

Fig. 6
figure 6

Western blot analysis of recombinant proteins. M prestained protein ladder, 1 the whole bacterium of pET-32a(+) after induction, 2 the control of pET-VP2 before induction, 3 the pellet of induced pET-VP2 after sonication, 4 the supernatant of induced pET-VP2 after sonication. The first antibody is PPV-positive pig sera, the second antibody is horseradish peroxidase-conjugated rabbit anti-pig antibody

Full size image

In order to purify the fusion protein (His-tagged VP2), we used Ni-NTA agarose. The results showed that the target protein could be conjugated to the resin. A single armed band was detected by SDS-PAGE (Fig. 7), Staining and destaining were done as described [20].

Fig. 7
figure 7

SDS-PAGE analysis of the purified recombinant VP2 protein. Staining and destaining were done as described in the text

Full size image

In order to confirm the identity of VP2 cut His-tag tail, the purified cut protein was exposed to Western blot assay using PPV-positive pig sera. The polyclonal antibodies recognized cut VP2, and the band has the 64 kDa as molecular weight. The assay therefore provides evidence that the protein could be used as an efficient immunological reagent (Fig. 8).

Fig. 8
figure 8

Western blot analysis of Expressed VP2. M prestained protein ladder, 1 the control of pET-VP2 before induction, 2 the purified pET-VP2 cut the His-Tag. The first antibody is PPV-positive pig sera, the second antibody is horseradish peroxidase-conjugated rabbit anti-pig antibody

Full size image

At present, vaccines against PPV are produced by chemically inactivating isolated virus particles grown in primary cell cultures of porcine origin. The method is both labour intensive and costly, with the additional hazard of requiring the handling of large quantities of infectious virus [21]. Economic and safety considerations, as well as practical limitations associated with low yields of PPV particles from in vitro cultures, led us to the investigate recombinant sub-unit vaccines for PPV. The VP2 protein of PPV was shown to self-assemble into virus-like particles when expressed in insect cells by baculovirus infection [22]. In addition, the virus-like particles of PPV were found to be highly immunogenic, and breeding sows were protected against reproductive failure in PPV challenge experiments [23]. Nonetheless, baculovirus-based systems for the production of recombinant proteins are still technically demanding, requiring sterile bioreactors that may be prohibitively costly for the production of vaccines for farm animals. Given that PPV causes serious economic losses for swine producers, development of safe, effective, and inexpensive methods for producing vaccines and diagnosing the disease is warranted.

In conclusion, we have established a procedure to produce the VP2 protein of PPV using plasmid pET-32a (+). After expression was optimized, a His-tagged VP2 was obtained. The fusion protein could be a useful antigen for detecting PPV.