The general aim of any heterologous expression system is to obtain a purified protein in its native conformation. Z is a small (approx. 11 kDa) viral protein normally expressed in the late phases of the arenavirus infection in mammalian cells. The recombinant overexpression of native Z protein requires an adequate eukaryotic expression system. However, there are some applications for which it is not necessary to obtain the protein in the native conformation, for example for antibodies production. Also, it has been reported that post-translational modifications are few in many proteins; therefore in these cases the bacterial expression systems are really advantageous because they have higher expression levels and lower costs when compared to eukaryotic expression systems [43
]. Thus, the system will be selected depending on the application of the recombinant product.
Using the sequence logo () the conservation of amino acids for the arenaviral Z proteins was described. Z proteins could be divided into three characteristic regions named Amino, Core and Carboxyl regions. Each region comprises previously described Z protein domains. The Myristoylation domain into the amino region, the RING finger domain into the Core, and the Late Domains into the Carboxyl region. Furthermore, other conserved islands both between groups and within each group were described. Although the role of these last domains is today unknown, they could be a likely target for studies associated with the arenavirus host range.
Simple methodologies that allowed Z protein expression and facilitated its subsequent purification were employed. During the analysis of the Z protein expression in bacteria it was not possible to obtain the product without a fusion protein or tags. Probably, the reason for low Z protein yield during purification is its hydrophobicity, causing it to form relatively insoluble intracellular inclusion bodies that must be denatured. The Borden laboratory [28
] discovered that including Zn+2
in the IPTG-induction media or in the lysis buffers greatly improved yields of recombinant Z protein. Although we included Zn+2
in the purification process, we have not yet tested a purification protocol that uses denaturing buffers.
Several expression strategies were tested, including cell hosts with different biological properties; it was only possible to achieve the goal when Z's amino terminus was fused to a bacterial protein. This fusion stabilized the recombinant protein and improved the expression levels.
In this work the over-expression of three different recombinant variants of Z protein: (1) Tio-Z-V5-His, (2) GST-Z, and (3) His-Z, was achieved. This last fusion protein was obtained from a baculoviral system in an insect cell line, while the other two were obtained from bacterial systems. The three variants of Z protein obtained can be cleaved from their N-fusion protein; Tio-Z-V5-His after proteolysis keeps the V5 epitope and the His tag.
Any of the expressed recombinant Z proteins could be used to obtain polyclonal or monoclonal antibodies that would allow immunological experiments to answer questions about the molecular biology of Junín virus. For example, studies of biological activity in different cell lines, cellular localization and protein interactions. Moreover, this serum could be useful for affinity chromatography designed to allow simple purification of Z protein without tags, or Z protein obtained after proteolysis of its fusion partner.
Tio-Z-V5-His recombinant protein was selected for antibody production, mainly because it showed the best expression level and solubility. This fusion protein was used for the generation of rabbit hyperimmune serum. Purified IgG fraction for subsequent Western blotting and EIA, was also obtained. On the other side, the purified Tio-Z-V5-His was used as substrate for the ENK reaction, in order to obtain Z-V5-His, employed as positive control in different assays.
Tio-Z-V5-His and His-Z fusion proteins can be purified by IMAC, since both peptides have His-Tags. However, this purification method could be counterproductive for the purification of native Z, since it is not known how Co+2
ions could modify Z structure by binding and affecting its RING finger domain. It was observed that the native structure of Z protein contains two Zn+2
atoms, and other transition metals could interfere with these sites [44
On the other hand, the GST-Z fusion protein can be purified by affinity chromatography with glutathione resins, without using immobilized metals. Its expression and purification was optimized with the aim to obtain the viral protein without tags on its native conformation, ready for crystallographic studies. The proteolysis of GST-Z protein with Factor Xa, produced only two peptides: GST and Z. The proteolysis reaction was successful because: (1) in SDS-PAGE analysis, at different times of the proteolysis reaction with Factor Xa, decreasing of GST-Z and increasing of individual GST and Z was observed; and (2) by Western blot analysis, the Z protein identity after the proteolysis reaction was asserted (Figures and ). Consequently, high concentrations of Z protein (necessary for crystallographic studies) could be obtained.
In addition, after GST-Z purification, the presence of two proteins that copurified with GST-Z was observed (indicated with I in the ). A similar phenomenon has been previously reported [45
]. The Western blot, with IgG purified from polyclonal anti-Z serum recognized the GST-Z protein and other copurified peptide (ca. 29 kDa), whiles the control, GST alone, did not generate signal. Nevertheless, apparently this peptide is not a substrate of Factor Xa. In summary, the copurified peptides had high affinity for glutathione resin, so they probably contained GST derived amino acids, and one of them was recognized by the anti-Z serum, so it had Z derived amino acids. To confirm these facts the N-terminus of the copurified peptides, and some peptides obtained by limited proteolysis with chymotrypsin were sequenced (). This assay allowed the identification of structured domains, which were less sensitive to proteolysis, and could be better candidates for crystallographic studies. We found that one of the copurified peptides, the one of approximately 25 kDa, was stable even at high chymotrypsin concentrations. The N-terminus of this peptide shared the sequence with the first six amino acids of GST, indicating that degradation occurred by the carboxyl terminus of the recombinant protein. The same results were obtained after analyzing the N-terminus of the approximately 29 kDa copurified protein, which was also stable after the chymotrypsin treatment. Thus, it was possible that these low molecular weight peptides contained fragments of Z protein at its C-terminus. Two of them were sequenced at the N-terminus because, according to the molecular weight estimated by the electrophoretical migration, the Z protein sequence would be included (indicated with II and III in the ). We could not establish the C-terminal sequence of the two copurified peptides, so we still do not know the specific site of cleavage of GST-Z in order to determine a stable domain. However, the approximate stable polypeptide corresponds to most of GST, which is not of our interest for crystallographic studies.
A subsequent approach to obtain Z protein in a native conformation could take advantage of its interactions with host proteins, such as the promyelocytic leukemia protein (PML) and others; it might be possible to overexpress Z and its interaction partner fused to a suitable tag for affinity purification. This would allow the copurification of the complex by affinity chromatography to later obtain native Z.
For Lassa arenavirus it was demonstrated that expression of Z in mammalian cell lines was sufficient for budding of “pseudo-virions” or Z-containing membranous particles [46
]. Apparently, glycine myristoylation at position 2 of Z sequence is critical for this phenomenon [47
]. This glycine residue is completely conserved in Z protein from all members of Arenaviridae
(). This suggests that all Z homologues within the Arenaviridae
family are myristoylated at this position. Currently, all reports of pseudo-virion budding employed mammalian cell lines. At the present it is not known whether insect cell lines are capable of recognizing this post-translational modification signal present in Z, so during this work we started testing this possibility.
When insect cell lines were used for Z expression, high levels of expression were not achieved (). For this system, Z was His-Tagged so it could be purified by IMAC and then untagged by TEV protease cleavage. Interestingly, Z was detected in the cellular fraction by means of Western blotting with α-Tio-Z-V5-His IgGs, but it was not possible to detect it using an anti α-His Tag antibody (Figures and , resp.). The absence of His-Tag signal could indicate the removal of the His-Tag from the protein, by a yet unidentified cellular process, or possible protein degradation at this particular site. Besides protein Western blot, a purification of the protein by IMAC was attempted, but no retention of protein from a cellular protein extract was observed (data not shown). Interestingly, when the cell culture supernatant fraction was analyzed using a Proteinase K protection assay, the results indicated that Z was included within small lipid vesicles, although other confirmatory assays such as immuno electron microscopy will be required in order to unambiguously confirm this.
The next step will be to further investigate the Z-membrane vesicles obtained from these cell lines. If N-myristoylation of Z in insect cells is proven to be effective, it will be necessary to modify the constructions for purification purposes. The employment of insect cell lines would represent a much safer methodology to obtain virus like particles [50
], which have potential use as vaccine against arenaviruses for which successful treatment has not yet been established.