Purification and Functional Characterization of a Biologically Active Full-Length Feline Immunodeficiency Virus (FIV) Pr50Gag

The feline immunodeficiency virus (FIV) full-length Pr50Gag precursor is a key player in the assembly of new viral particles. It is also a critical component of the efficient selection and packaging of two copies of genomic RNA (gRNA) into the newly formed virus particles from a wide pool of cellular and spliced viral RNA. To understand the molecular mechanisms involved during FIV gRNA packaging, we expressed the His6-tagged and untagged recombinant FIV Pr50Gag protein both in eukaryotic and prokaryotic cells. The recombinant Pr50Gag-His6-tag fusion protein was purified from soluble fractions of prokaryotic cultures using immobilized metal affinity chromatography (IMAC). This purified protein was able to assemble in vitro into virus-like particles (VLPs), indicating that it preserved its ability to oligomerize/multimerize. Furthermore, VLPs formed in eukaryotic cells by the FIV full-length Pr50Gag both in the presence and absence of His6-tag could package FIV sub-genomic RNA to similar levels, suggesting that the biological activity of the recombinant full-length Pr50Gag fusion protein was retained in the presence of His6-tag at the carboxy terminus. Successful expression and purification of a biologically active, recombinant full-length Pr50Gag-His6-tag fusion protein will allow study of the intricate RNA-protein interactions involved during FIV gRNA encapsidation.


Introduction
Genomic RNA (gRNA) packaging during the retroviral life cycle has been a topic of intense investigation over the last few decades. It has become increasingly clear that during this process, two copies of "full-length" unspliced gRNA are preferentially/selectively packaged in a dimeric form into the nascent virus particles over the viral spliced, as well as cellular RNA present in the infected cell [1][2][3][4][5][6][7]. Sequences responsible for selective packaging of unspliced gRNA are called "packaging signal/packaging determinants" or "psi" (ψ) and have been found to be universally present at the 5 end of all retroviral genomes [1][2][3][4][5][6]8]. Interestingly, without any exception, all retroviral ψ sequences have been shown to assume higher order structures containing several physical motifs in the form initiation site (DIS) for FIV gRNA [42,44,45]. Although four of the stem loops (SL1-4) within this region are static in nature, the FIV LRI has been shown to adopt an alternative, yet similarly-conserved, RNA conformation, in which the putative SL5 DIS is occluded, and which may favor translation and splicing over encapsidation [42]. Since FIV gRNA dimerization is initiated at a site within the Gag coding region, such a scenario is likely to confer high specificity to packaging, limited exclusively to the unspliced viral RNA. Moreover, the DIS is a purine-rich sequence and the presence of these purines in the packaging sequences of retroviral gRNA has been proposed to facilitate RNA encapsidation by functioning as a potential Gag binding site [9,[47][48][49][50][51]. Therefore, it is reasonable to hypothesize that in the case of FIV, the palindromic sequence within SL5, in addition to functioning as DIS, could also be involved in RNA-protein interactions during genomic RNA encapsidation by functioning as a potential Gag binding site.
The gag gene of FIV encoding for the precursor polypeptide, Pr50 Gag , is processed by the viral protease into its constituent mature proteins: NH 2 -p15 (matrix; MA), p24 (capsid; CA), p1 (spacer peptide), p13 (nucleocapsid; NC), p2 (spacer peptide)-COOH [26,52]. Like all retroviruses, FIV Pr50 Gag assembles to form an immature capsid and expression of the gag gene alone results in the assembly and release of immature virus-like particles (VLPs) [21,22]. During the assembly of FIV particles, the viral Pr50 Gag protein must specifically select the viral gRNA from a variety of cellular and viral spliced RNAs. However, how Pr50 Gag selectively chooses gRNA for packaging remains unclear. For example, it is not yet fully understood whether the sorting mechanism between gRNA and spliced RNA is only dependent on the inherent binding of the full-length Gag precursor (Pr50 Gag ) or whether there are other steps in the FIV life cycle also responsible in selecting gRNA, such as nucleocytoplasmic export and cellular compartmentalization of gRNA, as has recently been advocated for in HIV-1 [50,[53][54][55][56][57][58]. Therefore, to understand the molecular intricacies involved in FIV gRNA packaging, we must express and purify Pr50 Gag and understand its biological, biochemical, and biophysical attributes.
Only one study in the literature has reported the expression and purification of FIV Pr50 Gag from Escherichia coli (E. coli) [59]. However, its appropriateness for in vitro RNA binding assays, such as band-shift assays and footprinting experiments, has not been established. Therefore, in our attempt to identify Gag binding site(s) on FIV gRNA, as a first step, we sub-optimally over-expressed recombinant full-length Pr50 Gag -His 6 -tag protein in E. coli and purified large amounts from bacterial soluble fractions employing immobilized metal affinity chromatography (IMAC) and high-pressure liquid chromatography (HPLC), as has recently been described for HIV-1, MPMV, and MMTV [13][14][15]60,61]. The purified recombinant full-length Pr50 Gag -His 6 -tag protein showed an intrinsic ability to assemble into virus-like particles in eukaryotic cells. Furthermore, VLPs formed by the recombinant full-length Pr50 Gag -His 6 -tag protein in the eukaryotic cells could package the FIV gRNA, further confirming the biological activity of the purified recombinant protein. The availability of biologically active recombinant full-length Pr50 Gag -His 6 -tag protein should pave the way to further delineate the molecular mechanism(s) during selective FIV gRNA packaging process by establishing the differential binding abilities of the protein to different structural motif(s) on the full-length unspliced gRNA.

FIV Strain and Nucleotide Designations
The FIV Petaluma (34TF10) strain was used for this study, and the nucleotide numbers refer to the GenBank accession number M25381.1 [62].

Prokaryotic FIV Gag Expression Plasmid Construction
The FIV full-length gag sequences from nucleotides 628 to 1980 without codon optimization were commercially synthesized (Macrogen, Seoul, South Korea) with flanking NcoI and XhoI sites at the 5' and 3' ends respectively, to enable cloning into the prokaryotic expression vector pET28b (+) [9,13]. For the ease of cloning, the sequence CTCGAG corresponding to two XhoI sites in the native FIV Gag sequence was modified to CTAGAG (at nucleotide position 1408) and CACGAG (at the nucleotide position 1451), respectively, during gene synthesis to eliminate the XhoI sites while retaining the amino acid sequence. The synthesized FIV full-length gag was cleaved with NcoI and XhoI and cloned into the NcoI and XhoI sites of the pET28b (+). Such a cloning strategy kept the gag sequences in frame with a His 6 -tag at the C-terminus, resulting in the prokaryotic expression plasmid (VP10), which should express recombinant full-length Pr50 Gag His 6 -tag fusion protein with a predicted molecular weight of 50.3 kilodaltons (kDa) as computed by ExPASy-pI/Mw tool. The VP10 clone was sequenced to ensure that there were no point mutations introduced during gene synthesis.

Eukaryotic FIV Gag Expression Plasmid Construction
A pcDNA3 (Thermo Fischer Scientific, Waltham, MA, USA)-based eukaryotic expression vector expressing the recombinant full-length FIV Gag (Pr50 Gag -His 6 -tag) was created by amplifying the FIV gag gene from VP10 using primers OTR1384 and OTR1385. OTR1384 (5'CCG CTC GAG GCC GCC ACC ATG GGG AAT GGA CAG GGG CG 3'; nucleotides 628-647) is the forward primer containing 3 non-viral nucleotides (bold) followed by an XhoI site (italized) and a Kozak sequence (underlined) upstream of gag sequences. OTR1385 (5'CCG CTC GAG TTA gtg gtg gtg gtg gtg gtg TAA ATC CAA TAG TTT CTC CTC CAT T 3'; nucleotides 1977-1953) is the reverse primer containing 3 non-viral nucleotides (underlined) followed by an XhoI site (italicized), termination codon (bold), and His 6 -tag sequence (lower case) downstream of the gag sequences. Similarly, a full-length FIV Gag without His 6 -tag was created by amplifying the FIV gag gene from VP10 using primers OTR1384 and OTR1386. OTR1386 is the reverse primer (5'CCG CTC GAG TTA TAA ATC CAA TAG TTT CTC CTC CAT T 3'; nucleotides 1977-1953) containing 3 non-viral nucleotides (underlined), followed by an XhoI site (italicized) and a termination codon (bold) downstream of Gag. Amplifications were setup using the Phusion High-Fidelity PCR Kit (New England Biolabs, Ipswich, MA, USA), and standard PCR conditions were followed: initial denaturation at 98 • C for 1 min, followed by 30 cycles of denaturation at 98 • C for 30 s, annealing at 61 • C for 30 s, extension at 72 • C for 30 s, and a final extension at 72 • C for 10 min. PCR amplified products with His 6 -tag (amplicon size 1379 base pairs) and without His 6 -tag (amplicon size 1361 base pairs) were digested with XhoI and cloned into the XhoI sites of the pcDNA3-a eukaryotic expression vector to generate AD2 and AD3, respectively. For the efficient nuclear export of Gag mRNA, the MPMV constitutive transport element (CTE; [63]) was cloned downstream of the gag stop codon, as described previously [14,15]. Both AD2 and AD3 clones were confirmed by sequencing.

Bacterial Strains and Culture Media for Cloning and Expressing Full-Length Recombinant FIV Pr50 Gag -His 6 -Tag Protein
The ligated mixtures during the cloning process were transformed into the DH5α strain of E. coli using Luria-Bertani (LB) media in the presence of 50 µg/mL of kanamycin. For bacterial protein expression, VP10 was transformed into the BL21(DE3) strain of E. coli and cultured in an LB medium containing kanamycin. Analytical preparations of full-length recombinant Pr50 Gag His 6 -tag fusion protein were prepared as described previously [14,15].

Eukaryotic Expression of Recombinant Full-Length FIV Pr50 Gag Protein
Expression of full-length FIV Pr50 Gag both with and without His 6 -tag (AD2 and AD3 respectively) was tested in transient transfections of human embryonic kidney (HEK) 293T cells using a calcium phosphate kit, as described previously [14,15]. Briefly, transfections were performed in triplicate in 6-well plates using 4 µg of full-length FIV Gag eukaryotic expression plasmids (AD2 and AD3) along with 2 µg of an FIV-based transfer vector, MB87 [36]. For measuring transfection efficiency, a secreted alkaline phosphatase expression plasmid (pSEAP) was used at a concentration of 100 ng per well. The supernatants from the transfected cultures were harvested roughly 60 h post transfection and cellular debris was cleared by low speed centrifugation. The VLPs were pelleted by subjecting the clarified supernatant to ultracentrifugation at 70,000× g with a 20% (w/v) sucrose cushion. The pelleted VLPs were resuspended in a TNE buffer (50 mM Tris-HCl, pH 7.4; 100 mM NaCl; 1 mM EDTA pH 8.0). Cellular and viral RNAs were extracted using TRIzol, following the manufacturer's recommendations.

Transmission Electron Microscopy (TEM) of Eukaryotically-Expressed Virus-Like Particles (VLPs)
VLP formation by the recombinant FIV Pr50 Gag protein with and without His 6 -tag (AD2 and AD3, respectively) following transfection in 293T cells was observed using transmission electron microscopy (TEM; FEI Tecnai Biotwin Spirit G2). Briefly,~60 h following transfection, cells were harvested, pelleted, and washed with 0.1 M phosphate buffered saline (PBS). The fixation, ultrathin (95 nm) sectioning of the embedded samples and staining for TEM was performed as described previously [14,15]. Packaging of FIV RNA into the VLPs was monitored by RT-PCR, as described previously [14,15], but with modifications. Briefly, RNA preparations were subjected to DNase treatment to eliminate any contaminating DNA that may have been carried forward from the transfected cultures. Following DNase treatment, amplification on both cytoplasmic and viral RNA preparations were performed using transfer vector (MB87) specific primers, OTR660 (5' GAG GAC TTT TGA GTT CTC CCT TGA GGC 3'; nucleotides 230-256) and OTR662 (5' AGC AGG AGT TCT GCT TAA CAG CTT TC 3'; nucleotides 440-415, amplifying a 210 base pair region between U5 and UTR), to ensure that the RNA preparation were not contaminated with any plasmid DNA. After ensuring that RNA preparations were devoid of any contaminating plasmid or genomic DNA by PCR, they were converted into cDNA using random hexamers (5'NNNNNN3') and MMLV reverse transcriptase (Promega, Madison, WI, USA). Finally, cDNAs were amplified with the same vector-specific primers as described above (amplicon size 210 base pairs) to assess the ability of Pr50 Gag VLPs to package transfer vector (MB87) RNA. Amplifications were setup using PCR master mix (Promega, USA) and standard PCR conditions were followed: initial denaturation at 94 • C for 2 min, followed by 30 cycles of denaturation at 94 • C for 45 s, annealing at 65 • C for 45 s, and extension at 72 • C for 45 s. An in-house SYBR Green-based qPCR assay was developed to quantitate the relative packaging of the FIV transfer vector (MB87) RNA by VLPs using the same primer set (OTR660/OTR662; 400 nM) and Solis BioDyne 5× Hot FirePol EvaGreen qPCR Supermix (Solis BioDyne, Tartu, Estonia). Expression of the FIV-specific signal was normalized to the endogenous β-actin expression by using spliced β-actin-specific primers [64] at a concentration of 100 nM each using the 2 −∆∆CT method. Appropriate primer concentrations for the assay were determined empirically by conducting test assays, and the primer concentration that gave a single, narrow peak upon melt curve analysis with the maximum amplification was chosen. To circumvent any non-specific signal, the annealing and extensions were performed at 72 • C for 40 cycles using the ABI QuantStudio 7 or 3 machines (Applied Biosystems, Foster City, CA, USA)

Expression of Recombinant FIV Pr50 Gag -His 6 -tag Protein in E. coli
Large scale expression of the recombinant Pr50 Gag -His 6 -tag protein was initiated using a starter culture with kanamycin (50 µg/mL), which was inoculated with a single colony of bacteria (BL21 (DE3)) and grown overnight at 37 • C. The following day, this starter culture was used to inoculate 500 ml LB media in 2 liter baffled flasks containing kanamycin (50 µg/mL) and 1% glucose, as described previously [14,15]. The cultures were allowed to grow to an OD 600 of 0.6, and then cultures were induced with isopropyl β-d-1-thiogalactopyranoside (IPTG), as described previously [14,15]. The induced cultures were grown for another 4 h at 28 • C and then pelleted by centrifugation at 6300× g for 15 min at 4 • C. The bacterial pellets were stored at −80 • C until further use.

Affinity Purification and Size Exclusion Chromatography of Recombinant FIV Pr50 Gag -His 6 -Tag Protein
The recombinant Pr50 Gag -His 6 -tagged protein was purified based on the principle of IMAC, as described previously [14,15]. Briefly, the frozen bacterial pellets were lysed using the CelLytic B buffer supplemented with benzonase, lysozyme, and EDTA-free protease inhibitor. Following lysis, the protein was purified on a Ni 2+ -immobilized HisTrap fast flow (FF) 5 mL column. An Amicon Ultra 15 device with a 30 kDa cut-off was used for concentrating the eluted protein. Coomassie Brilliant Blue staining and western blotting of the concentrated fractions were performed to establish the quality of the protein. Finally, the concentrated protein was loaded onto a Superdex 200 increase 10/300 GL column for further purification to homogeneity. Fractions that gave a peak were collected and then analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The absorbance ratio at 260 and 280 nm ensured the purity of the protein after size exclusion chromatography. Finally, aliquots of purified Pr50 Gag -His 6 -tag fusion protein were pooled and stored at −80 • C.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE) and Western Blotting
The Pr50 Gag -His 6 -tagged protein was monitored for expression and purification using SDS-PAGE and western blotting. Protein samples were run on a 4%-12% ExpressPlus TM PAGE gel, electrophoresed under reducing conditions, and stained using Coomassie Brilliant Blue, as described previously [14,15]. For western blot analysis, duplicate gels were transferred onto nitrocellulose membranes and probed with an FIV α-p24 monoclonal antibody (Biodesign International, Saco, Maine, USA) and an α-His 6 monoclonal antibody-horseradish peroxidase (HRP) conjugate, as described previously [14,15]. In vitro assembly of VLPs was performed as described recently for MPMV and MMTV full-length Gag [14,15]. Briefly, the purified recombinant Pr50 Gag -His 6 -tagged protein expressed in bacteria was resuspended in an assembly buffer (50 mM Tris (pH 7.4), 1.0 M NaCl) at a concentration of 2 mg/mL. Next, protein in the assembly buffer was mixed with yeast tRNA at a nucleic acid to protein ratio of 4% (w/w) and placed in a Slide-A-Lyzer®10 K dialysis cassette G2 (Thermo Scientific). The dialysis cassette containing the mixture of protein and yeast tRNA was dialyzed against 20 mM Tris (pH 7.4) containing 150 mM NaCl and 10 mM dithiothreitol (DTT) overnight at 4 • C. Following dialysis, 8-10 µL of protein and yeast tRNA was spotted on a carbon coated formvar grid (Proscitech, Queensland, Australia), air dried, and stained with 1% uranyl acetate for observation using TEM.

Full-Length FIV Pr50 Gag Protein, Both with and without a His 6 -Tag Can Be Expressed in Eukaryotic Cells, Capable of Making Virus-Like Particles (VLPs) That Can Package FIV gRNA
In this study, our goal was to express and purify a full-length FIV Pr50 Gag -His 6 -tag fusion protein that could be used for RNA-protein interaction studies, to help establish the role of FIV Gag in gRNA packaging. However, there has been some concern that the presence of a positively-charged His-tag may favor interactions with the nucleic acids on its own [61]. Therefore, as a first step, we cloned the FIV gag (Pr50 Gag ) both in the presence (AD2; Figure 1A) and absence (AD3; Figure 1A) of the His 6 -tag in a eukaryotic expression vector. Using these expression plasmids, we developed a two-plasmid genetic complementation assay to monitor the packaging of FIV unspliced sub-genomic RNA ( Figure 1B). Briefly, the assay was based on the premise that the FIV sub-genomic transfer vector RNA (MB87; Figure 1A; [36]) contains the already-established packaging determinants of FIV [37,38,43] which could be recognized by the FIV full-length Pr50 Gag protein, facilitating its encapsidation/packaging in the nascently forming virus particles. Towards this end, both the His 6 -tag (+) and His 6 -tag (−) FIV Pr50 Gag expression plasmids (AD2 and AD3, respectively) were individually co-transfected along with the FIV transfer vector, MB87, in 293T cells ( Figure 1B). Electron micrographs taken from different fields of the transfected 293T cells revealed VLP production both in the presence and absence of the His 6 -tag. As shown in Figure 2, an electron dense protein layer was observed underlying the plasma membrane from which spherical shaped VLPs could be seen budding when FIV Pr50 Gag expression plasmids with His 6 -tag (AD2) (Figure 2A-D) or without His 6 -tag (AD3) ( Figure 2E-H) were transfected in 293T cells. In mock experimental controls (cells that were transfected with the void expression plasmid pcDNA3), as expected, no budding of VLPs could be observed ( Figure 2I-J). The size of the majority of VLPs produced was calculated to be~100-120 nanometers (nm) in diameter ( Figure 2), which is slightly smaller than the mature FIV particles (125-150 nm) and consistent with the release of immature and spherical FIV particles from eukaryotic cells, as reported earlier [65,66]. These results demonstrate that we were successful in expressing the recombinant full-length FIV Pr50 Gag protein both with and without His 6 -tag in eukaryotic cells, and the expressed proteins could make VLPs. Therefore, cloning His 6 -tag at the C-terminus of Gag did not interfere with the expression of the recombinant full-length FIV Pr50 Gag or the production of VLPs. transfected with the void expression plasmid pcDNA3), as expected, no budding of VLPs could be observed ( Figure 2I-J). The size of the majority of VLPs produced was calculated to be ~100-120 nanometers (nm) in diameter ( Figure 2), which is slightly smaller than the mature FIV particles (125-150 nm) and consistent with the release of immature and spherical FIV particles from eukaryotic cells, as reported earlier [65,66]. These results demonstrate that we were successful in expressing the recombinant full-length FIV Pr50 Gag protein both with and without His6-tag in eukaryotic cells, and the expressed proteins could make VLPs. Therefore, cloning His6-tag at the C-terminus of Gag did not interfere with the expression of the recombinant full-length FIV Pr50 Gag or the production of VLPs. Next, supernatants from the transfected cultures were collected, and the immature viral particles were pelleted by ultracentrifugation. RNA from immature viral particles was isolated and subjected to RT-PCR to monitor the ability of the His (+) and His (−) Pr50 Gag proteins to package FIV subgenomic RNA. Towards this end, RNA preparations were DNase-treated and cDNAs prepared, as described in Materials and Methods. The use of conventional RT-PCR revealed that immature viral Next, supernatants from the transfected cultures were collected, and the immature viral particles were pelleted by ultracentrifugation. RNA from immature viral particles was isolated and subjected to RT-PCR to monitor the ability of the His (+) and His (−) Pr50 Gag proteins to package FIV sub-genomic RNA. Towards this end, RNA preparations were DNase-treated and cDNAs prepared, as described in Materials and Methods. The use of conventional RT-PCR revealed that immature viral particles formed by both His (+) and His (−) Pr50 Gag proteins could package RNA from the sub-genomic transfer vector, MB87, in an efficient manner ( Figure 1C; upper panel). To assess the relative ability of the His (+) and His (−) Gag proteins to encapsidate FIV subgenomic RNA, an in-house SYBR Green-based qPCR assay was developed in which β-actin was used as an endogenous control, as described in Materials and Methods. The relative transfer vector (MB87) RNA packaging efficiency (RPE) by the His (+) and His (−) Gag proteins was determined by calculating the ratio of the transfer vector (MB87) RNA packaged in the VLPs to the transfer vector (MB87) RNA being expressed in the cytoplasm. Using this assay, essentially a similar ability of the two proteins to package MB87 RNA was observed ( Figure 1C; lower panel). This observation confirmed that cloning the His 6 -tag at the C-terminus did not interfere with the expression of the recombinant full-length FIV Pr50 Gag or production of VLPs, nor did it negatively affect its ability to package FIV unspliced sub-genomic RNA. This is also consistent with the earlier observations made in the case of MPMV and MMTV, where VLPs produced in eukaryotic cells by full-length Gag both in the presence or absence of His 6 -tag were able to package nearly equal amounts of their respective gRNAs [14,15]. Finally, a close corroboration in the functional capacities of recombinant full-length FIV Pr50 Gag proteins (both with and without His 6 -tag) was further supported by the fact that when sequences for AD2 (with His 6 -tag) and AD3 (without His 6 -tag) were investigated using the online ExPASy-Compute pI/Mw tool to calculate the theoretical isoelectric point (pI) of the proteins, predicting almost no difference (0.07) in the expressed recombinant proteins (AD2 with His 6 -tag: pI: 8.92 versus AD3 without His 6 -tag: pI: 8.99). particles formed by both His (+) and His (−) Pr50 Gag proteins could package RNA from the subgenomic transfer vector, MB87, in an efficient manner ( Figure 1C; upper panel). To assess the relative ability of the His (+) and His (−) Gag proteins to encapsidate FIV subgenomic RNA, an in-house SYBR Green-based qPCR assay was developed in which β-actin was used as an endogenous control, as described in Materials and Methods. The relative transfer vector (MB87) RNA packaging efficiency (RPE) by the His (+) and His (−) Gag proteins was determined by calculating the ratio of the transfer vector (MB87) RNA packaged in the VLPs to the transfer vector (MB87) RNA being expressed in the cytoplasm. Using this assay, essentially a similar ability of the two proteins to package MB87 RNA was observed ( Figure 1C; lower panel). This observation confirmed that cloning the His6-tag at the C-terminus did not interfere with the expression of the recombinant full-length FIV Pr50 Gag or production of VLPs, nor did it negatively affect its ability to package FIV unspliced sub-genomic RNA. This is also consistent with the earlier observations made in the case of MPMV and MMTV, where VLPs produced in eukaryotic cells by full-length Gag both in the presence or absence of His6tag were able to package nearly equal amounts of their respective gRNAs [14,15]. Finally, a close corroboration in the functional capacities of recombinant full-length FIV Pr50 Gag proteins (both with and without His6-tag) was further supported by the fact that when sequences for AD2 (with His6-tag) and AD3 (without His6-tag) were investigated using the online ExPASy-Compute pI/Mw tool to calculate the theoretical isoelectric point (pI) of the proteins, predicting almost no difference (0.07) in the expressed recombinant proteins (AD2 with His6-tag: pI: 8.92 versus AD3 without His6-tag: pI: 8.99).

Expression of Full-Length Recombinant FIV Pr50 Gag -His6-tagged Fusion Protein in Prokaryotic Cells
Having established the expression and functional characterization of the recombinant FIV Pr50 Gag -His6-tagged fusion protein in eukaryotic cells, we next embarked on expressing and purifying this protein from E. coli. Towards this end, the FIV full-length Gag was cloned into the prokaryotic expression plasmid, pET28b(+), to generate VP10 (Figure 3), as described in Materials and Methods.

Expression of Full-Length Recombinant FIV Pr50 Gag -His 6 -Tagged Fusion Protein in Prokaryotic Cells
Having established the expression and functional characterization of the recombinant FIV Pr50 Gag -His 6 -tagged fusion protein in eukaryotic cells, we next embarked on expressing and purifying this protein from E. coli. Towards this end, the FIV full-length Gag was cloned into the prokaryotic expression plasmid, pET28b(+), to generate VP10 (Figure 3), as described in Materials and Methods. Such a cloning strategy should generate a fusion protein expressing the full-length FIV Pr50 Gag containing a LEHHHHHH tag at the C-terminus (Pr50 Gag -His 6 -tag), with a predicted molecular weight of 50.3 kDa. The transformation of VP10 plasmid DNA into BL21 (DE3) bacterial cells allowed high level expression of the recombinant protein when induced by IPTG. containing a LEHHHHHH tag at the C-terminus (Pr50 Gag -His6-tag), with a predicted molecular weight of 50.3 kDa. The transformation of VP10 plasmid DNA into BL21 (DE3) bacterial cells allowed high level expression of the recombinant protein when induced by IPTG. As a first step, we monitored the expression of VP10-transformed BL21 (DE3) cultures in total bacterial lysates at 0, 2, 4, and 6 hours at 37 °C. Briefly, total protein lysates were prepared for each time point from both IPTG-induced as well as uninduced cultures and examined by SDS-PAGE. As shown in Figure 4, a discrete band of approximately 50 kDa could be detected exclusively in cultures from 2, 4, and 6 h post IPTG induction (lanes 5, 7, and 9), and not in cultures at 0 hour (lane 3), uninduced cultures (lanes 4, 6, and 8), or in cultures transformed with only the expression plasmid without any FIV Pr50 Gag sequences (pET28b alone; lane 2). These observations demonstrate that a recombinant full-length FIV Pr50 Gag protein could be expressed in bacteria. However, they did not show whether the recombinant protein was present in the soluble fraction from which it could be purified or if it confined to inclusion bodies, as has been previously reported in the case of MPMV [67].  As a first step, we monitored the expression of VP10-transformed BL21 (DE3) cultures in total bacterial lysates at 0, 2, 4, and 6 hours at 37 • C. Briefly, total protein lysates were prepared for each time point from both IPTG-induced as well as uninduced cultures and examined by SDS-PAGE. As shown in Figure 4, a discrete band of approximately 50 kDa could be detected exclusively in cultures from 2, 4, and 6 h post IPTG induction (lanes 5, 7, and 9), and not in cultures at 0 hour (lane 3), uninduced cultures (lanes 4, 6, and 8), or in cultures transformed with only the expression plasmid without any FIV Pr50 Gag sequences (pET28b alone; lane 2). These observations demonstrate that a recombinant full-length FIV Pr50 Gag protein could be expressed in bacteria. However, they did not show whether the recombinant protein was present in the soluble fraction from which it could be purified or if it confined to inclusion bodies, as has been previously reported in the case of MPMV [67].

FIV Pr50 Gag -His6-tagged Fusion Protein is Expressed in the Soluble Fraction in Bacteria
To establish that the recombinant full-length FIV Pr50 Gag -His6-tag protein was expressed in soluble fractions, an overnight bacterial culture of VP10-transformed in BL21(DE3) bacterial cells was induced with IPTG and allowed to grow at 37 °C for 2, 4, and 6 h. Cells harvested at various time points were pelleted and processed as described in the Materials and Methods section. Cell debris and/or inclusion bodies, if any, and other insoluble material, were removed by centrifugation, and the soluble fraction of the full-length FIV Pr50 Gag -His6-tag fusion protein obtained from different time points was analyzed by SDS-PAGE and immunoblotting. SDS-PAGE analysis following Coomassie Brilliant Blue staining revealed a characteristic band of ~50 kDa consistent with the anticipated size of the recombinant FIV Pr50 Gag -His6-tag protein from IPTG-induced cultures ( Figure 5A; lanes 5, 7, and 9). Next, we established the uniqueness and recombinant nature (Pr50 Gag -His6-tag fusion protein) of the bands observed on Coomassie Brilliant Blue staining by western blotting using HRPconjugated α-His6 as well as FIV α-p24 monoclonal antibodies (Figure 5B,C; lanes 3, 4, and 5, respectively). Immunoblotting employing both α-His6 as well as FIV α-p24 monoclonal antibodies of soluble fractions from VP10 uninduced cultures exhibited some expression of FIV Pr50 Gag -His6-tag fusion protein ( Figure 5B; lane 2 and Figure 5C; lane 2). Such low-level expression in the uninduced culture could possibly be due to the leaky nature of the T7 promoter of the expression plasmid, a phenomenon that has been reported earlier [68]. In order to inhibit promoter leakiness, we opted to express the Pr50 Gag -His6-tag fusion protein from VP10 following IPTG induction using sub-optimal conditions, such as low temperature (28 °C) instead of 37 °C and growth in the presence of glucose, as described previously [68]. Such a strategy has been successful in suppressing the leaky nature of the T7 promoter of the expression plasmid during MMTV Pr77 Gag expression [15]. Consistent with

FIV Pr50 Gag -His 6 -Tagged Fusion Protein Is Expressed in the Soluble Fraction in Bacteria
To establish that the recombinant full-length FIV Pr50 Gag -His 6 -tag protein was expressed in soluble fractions, an overnight bacterial culture of VP10-transformed in BL21(DE3) bacterial cells was induced with IPTG and allowed to grow at 37 • C for 2, 4, and 6 h. Cells harvested at various time points were pelleted and processed as described in the Materials and Methods section. Cell debris and/or inclusion bodies, if any, and other insoluble material, were removed by centrifugation, and the soluble fraction of the full-length FIV Pr50 Gag -His 6 -tag fusion protein obtained from different time points was analyzed by SDS-PAGE and immunoblotting. SDS-PAGE analysis following Coomassie Brilliant Blue staining revealed a characteristic band of~50 kDa consistent with the anticipated size of the recombinant FIV Pr50 Gag -His 6 -tag protein from IPTG-induced cultures ( Figure 5A; lanes 5, 7, and 9). Next, we established the uniqueness and recombinant nature (Pr50 Gag -His 6 -tag fusion protein) of the bands observed on Coomassie Brilliant Blue staining by western blotting using HRP-conjugated α-His 6 as well as FIV α-p24 monoclonal antibodies ( Figure 5B,C; lanes 3, 4, and 5, respectively). Immunoblotting employing both α-His 6 as well as FIV α-p24 monoclonal antibodies of soluble fractions from VP10 uninduced cultures exhibited some expression of FIV Pr50 Gag -His 6 -tag fusion protein ( Figure 5B; lane 2 and Figure 5C; lane 2). Such low-level expression in the uninduced culture could possibly be due to the leaky nature of the T7 promoter of the expression plasmid, a phenomenon that has been reported earlier [68]. In order to inhibit promoter leakiness, we opted to express the Pr50 Gag -His 6 -tag fusion protein from VP10 following IPTG induction using sub-optimal conditions, such as low temperature (28 • C) instead of 37 • C and growth in the presence of glucose, as described previously [68]. Such a strategy has been successful in suppressing the leaky nature of the T7 promoter of the expression plasmid during MMTV Pr77 Gag expression [15]. Consistent with our expectations on Coomassie Brilliant Blue staining, the recombinant Pr50 Gag -His 6 -tagged fusion protein was observed exclusively in the IPTG-induced cultures ( Figure 6A; lanes 5, 7, and 9) and not in the uninduced cultures ( Figure 6A;  lanes 4, 6, and 8). Immunoblotting of soluble fractions from VP10 using α-His 6 monoclonal antibody also confirmed the lack of expression of the recombinant full-length FIV Pr50 Gag -His 6 -tag protein from uninduced cultures ( Figure 6B; lane 2). These results clearly demonstrate that a FIV Pr50 Gag -His 6 -tag protein can be expressed in the soluble fractions of bacterial lysates employing sub-optimal culturing conditions, with the maximal expression at 4 hours ( Figure 6A; lane 7 and Figure 6B; lane 4). Despite the fact that the recombinant FIV full-length Pr50 Gag -His 6 -tag expression could be observed between 2 and 6 hours following induction by IPTG ( Figure 6A; lanes 5, 7, and 9 and Figure 6B; lanes 3, 4, and 5), for further large scale purification, we decided to express the protein sub-optimally only (28 • C) and for 4 h post IPTG induction to circumvent any possible accumulation and/or confinement, as has been reported in the case of the MPMV Gag polyprotein by culturing bacteria at 37 • C for extended periods of time [67]. our expectations on Coomassie Brilliant Blue staining, the recombinant Pr50 Gag -His6-tagged fusion protein was observed exclusively in the IPTG-induced cultures ( Figure 6A; lanes 5, 7, and 9) and not in the uninduced cultures ( Figure 6A; lanes 4, 6, and 8). Immunoblotting of soluble fractions from VP10 using α-His6 monoclonal antibody also confirmed the lack of expression of the recombinant full-length FIV Pr50 Gag -His6-tag protein from uninduced cultures ( Figures 6B; lane 2). These results clearly demonstrate that a FIV Pr50 Gag -His6-tag protein can be expressed in the soluble fractions of bacterial lysates employing sub-optimal culturing conditions, with the maximal expression at 4 hours ( Figure 6A; lane 7 and Figure 6B; lane 4). Despite the fact that the recombinant FIV full-length Pr50 Gag -His6-tag expression could be observed between 2 and 6 hours following induction by IPTG ( Figure  6A; lanes 5, 7, and 9 and Figure 6B; lanes 3, 4, and 5), for further large scale purification, we decided to express the protein sub-optimally only (28 °C) and for 4 h post IPTG induction to circumvent any possible accumulation and/or confinement, as has been reported in the case of the MPMV Gag polyprotein by culturing bacteria at 37 °C for extended periods of time [67].

Purification of the Recombinant Full-Length Pr50 Gag -His6-tagged Fusion Protein from Soluble Fraction by Immobilized Metal Affinity Chromatography (IMAC)
Following the confirmation that the FIV recombinant full-length Pr50 Gag -His6-tagged fusion protein is indeed present in the soluble fraction, we embarked on purification of the protein by IMAC, as has recently been accomplished for the HIV-1, MPMV, and MMTV full-length Gag proteins ( [13][14][15]; see Materials and Methods for details). During IMAC purification, we made sure to use nondenaturing buffer conditions (especially 1.0 M NaCl concentration) that enable the binding of the protein to the column and at the same time avoid protein aggregation and precipitation. Briefly, soluble fraction from the bacterial lysates were applied to the Ni-immobilized HisTrap fast flow (FF) column, which allowed the protein to bind to the column. Consequently, no protein could be visualized in the flow through from the HisTrap column or in the 50 mM immidazole wash ( Figure  7A; lanes 2, 3, and 4). Next, the HisTrap-column bound protein was eluted using different concentrations of imidazole (100 mM, 250 mM, and 500 mM) with the maximum protein eluting at 250 mM imidazole ( Figure 7A; lanes 5, 6, and 7). IMAC purification was able to get rid of most of the bacterial proteins present in the soluble fraction ( Figure 7A; compare lane 2 with lane 6). Following IMAC purification, the purity of the recombinant FIV Pr50 Gag -His6-tag fusion protein was further established by immunoblotting using HRP-conjugated α-His6 ( Figure 7B; lane 4) as well as FIV α-p24 monoclonal antibodies ( Figure 7C; lane 4). Taken together, these observations validate that the recombinant FIV full-length Pr50 Gag protein is indeed fused in a frame with His6-tag, which facilitated its retention on the HisTrap column and its elution into a purer form ( Figure 7A; compare lane 2 with lane 6; Figure 7B; lane 4; Figure 6C lane 4). The IMAC-purified Pr50 Gag -His6-tagged fusion protein, upon immunoblotting with the FIV α-p24 antibody, still revealed additional faint bands ( Figure 7B; lane 4 and Figure 7C; lane 4), which could be attributed to the possible degradation of the recombinant protein. Following the confirmation that the FIV recombinant full-length Pr50 Gag -His 6 -tagged fusion protein is indeed present in the soluble fraction, we embarked on purification of the protein by IMAC, as has recently been accomplished for the HIV-1, MPMV, and MMTV full-length Gag proteins ( [13][14][15]; see Materials and Methods for details). During IMAC purification, we made sure to use non-denaturing buffer conditions (especially 1.0 M NaCl concentration) that enable the binding of the protein to the column and at the same time avoid protein aggregation and precipitation. Briefly, soluble fraction from the bacterial lysates were applied to the Ni-immobilized HisTrap fast flow (FF) column, which allowed the protein to bind to the column. Consequently, no protein could be visualized in the flow through from the HisTrap column or in the 50 mM immidazole wash ( Figure 7A; lanes 2, 3, and 4). Next, the HisTrap-column bound protein was eluted using different concentrations of imidazole (100 mM, 250 mM, and 500 mM) with the maximum protein eluting at 250 mM imidazole ( Figure 7A; lanes 5, 6, and 7). IMAC purification was able to get rid of most of the bacterial proteins present in the soluble fraction ( Figure 7A; compare lane 2 with lane 6). Following IMAC purification, the purity of the recombinant FIV Pr50 Gag -His 6 -tag fusion protein was further established by immunoblotting using HRP-conjugated α-His 6 ( Figure 7B; lane 4) as well as FIV α-p24 monoclonal antibodies ( Figure 7C; lane 4). Taken together, these observations validate that the recombinant FIV full-length Pr50 Gag protein is indeed fused in a frame with His 6 -tag, which facilitated its retention on the HisTrap column and its elution into a purer form ( Figure 7A; compare lane 2 with lane 6; Figure 7B; lane 4; Figure 7C

Further Purification of the IMAC-Purified Pr50 Gag -His6-tag Protein by Gel Filtration/Size Exclusion Chromatography
After having confirmed the recombinant nature of the Pr50 Gag -His6-tagged protein eluted following IMAC purification, the protein was concentrated using an Amicon ® Ultra 15 centrifugal device (30 kDa cut-off membrane). Next, the concentrated protein was further purified by gel filtration/size exclusion chromatography under non-denaturing conditions using a Superdex 200 10/300 GL column, as described previously [14,15]. As in the case of IMAC purification, any possible protein aggregation and precipitation was avoided by using a high salt concentration (1.0 M NaCl) in the gel filtration buffer. On the basis of absorbance at 280 nm, which suggested elution of the purified recombinant protein, 500 µL fractions were collected and analyzed by SDS-PAGE and immunoblotting.
As shown in Figure 8A, fractions 27-32 of the eluted protein corresponded to a distinct peak at A280. Analysis of these fractions by SDS-PAGE and Coomassie Brilliant Blue staining of the gel revealed that the distinct peak consisted of variable amounts of pure FIV Pr50 Gag -His6-tag fusion protein ( Figure 8B). These fractions (28)(29) containing the purest form of the protein (without any bands other than of Pr50 Gag ) were combined and again concentrated using Amicon® Ultra 15 centrifugal columns (30 kDa cut-off membrane) to bring the protein to a final concentration of ~2 mg/mL.
The concentrated protein was further analyzed by immunoblotting using HRP-conjugated α-His6 as well as FIV α-p24 monoclonal antibodies. The results depicted in Figure 8C   Western blot analyses on the same soluble fractions (before and after IMAC purification) using α-His 6 (B) and FIV α-p24 (C) monoclonal antibodies, respectively, using equal amounts of protein.

Further Purification of the IMAC-Purified Pr50 Gag -His 6 -Tag Protein by Gel Filtration/Size Exclusion Chromatography
After having confirmed the recombinant nature of the Pr50 Gag -His 6 -tagged protein eluted following IMAC purification, the protein was concentrated using an Amicon ® Ultra 15 centrifugal device (30 kDa cut-off membrane). Next, the concentrated protein was further purified by gel filtration/size exclusion chromatography under non-denaturing conditions using a Superdex 200 10/300 GL column, as described previously [14,15]. As in the case of IMAC purification, any possible protein aggregation and precipitation was avoided by using a high salt concentration (1.0 M NaCl) in the gel filtration buffer. On the basis of absorbance at 280 nm, which suggested elution of the purified recombinant protein, 500 µL fractions were collected and analyzed by SDS-PAGE and immunoblotting.
As shown in Figure 8A, fractions 27-32 of the eluted protein corresponded to a distinct peak at A 280 . Analysis of these fractions by SDS-PAGE and Coomassie Brilliant Blue staining of the gel revealed that the distinct peak consisted of variable amounts of pure FIV Pr50 Gag -His 6 -tag fusion protein ( Figure 8B). These fractions (28)(29) containing the purest form of the protein (without any bands other than of Pr50 Gag ) were combined and again concentrated using Amicon®Ultra 15 centrifugal columns (30 kDa cut-off membrane) to bring the protein to a final concentration of~2 mg/mL.
The concentrated protein was further analyzed by immunoblotting using HRP-conjugated α-His 6 as well as FIV α-p24 monoclonal antibodies. The results depicted in Figure 8C corroborated well with the SDS-PAGE analysis ( Figure 8B), demonstrating that the pooled protein fractions contained a pure FIV Pr50 Gag -His 6 -tag fusion protein. Finally, we estimated the purity of the purified protein by measuring the A 260 /A 280 ratio spectrophotometrically, which gave a value of 0.62, suggesting that the protein purity was greater than 95% and contained only an inconsequential amount of nucleic acid impurities, if any. In terms of yield, a one-and-a-half-liter bacterial culture yielded~16 mg of protein following IMAC purification, from which two milligrams of the purest form of recombinant Pr50 Gag -His 6 -tag fusion protein could be recovered following gel filtration/size exclusion chromatography. measuring the A260/A280 ratio spectrophotometrically, which gave a value of 0.62, suggesting that the protein purity was greater than 95% and contained only an inconsequential amount of nucleic acid impurities, if any. In terms of yield, a one-and-a-half-liter bacterial culture yielded ~16 mg of protein following IMAC purification, from which two milligrams of the purest form of recombinant Pr50 Gag -His6-tag fusion protein could be recovered following gel filtration/size exclusion chromatography. (C) Western blot analyses on the pooled fractions (fractions 28 and 29) of the purified protein using α-His6 and FIV α-p24 monoclonal antibodies, respectively, using equal amounts of protein.

The Recombinant Full-Length Pr50 Gag -His6-tagged Fusion Protein Can Assemble in Vitro to Form Virus-Like Particles (VLPs)
The ability of retroviral Gag proteins to self-assemble in vitro and form VLPs for lentiviruses such as HIV-1 and FIV, and other retroviruses, such as MPMV and MMTV, has been reported earlier [13][14][15]59,60,69]. Therefore, as a logical next step, to further validate the purity and biological activity of the recombinant full-length FIV Pr50 Gag -His6-tag fusion protein expressed in prokaryotic cells, we monitored the ability of this protein to form VLPs by assembling itself in vitro. Since the presence of nucleic acids has been found to be necessary for in vitro assembly and VLP formation [13,59,60,69],  The ability of retroviral Gag proteins to self-assemble in vitro and form VLPs for lentiviruses such as HIV-1 and FIV, and other retroviruses, such as MPMV and MMTV, has been reported earlier [13][14][15]59,60,69]. Therefore, as a logical next step, to further validate the purity and biological activity of the recombinant full-length FIV Pr50 Gag -His 6 -tag fusion protein expressed in prokaryotic cells, we monitored the ability of this protein to form VLPs by assembling itself in vitro. Since the presence of nucleic acids has been found to be necessary for in vitro assembly and VLP formation [13,59,60,69], these experiments were carried out in the presence of 4% w/w yeast tRNA, as described previously [13][14][15]. To circumvent the aggregation of the purified recombinant FIV Pr50 Gag -His 6 -tag fusion protein, the protein was mixed with yeast tRNA in a high salt buffer (1 M NaCl) and dialyzed against a buffer with relevant physiological salt concentration. As an ancillary control, the purified recombinant FIV Pr50 Gag -His 6 -tag fusion protein without yeast tRNA was also dialyzed in a similar fashion. Dialysis was carried out overnight at 4 • C, following which samples were collected individually from the dialysis cassette, and 10 µL of the 500 µL assembly reaction mixture was spotted on formvar carbon coated grids, negatively stained, and visualized by TEM.
The electron micrographs of the structures assembled in vitro by the recombinant FIV Pr50 Gag -His 6 -tag fusion protein were taken from different fields. As expected, the formation of VLPs in the form of compact electron-dense spherical rings of approximately 89 nm in size confirmed the in vitro assembly process when the recombinant FIV Pr50 Gag -His 6 -tag fusion protein was mixed with yeast tRNA (Figure 9A-D). On the other hand, as reported earlier [59], the recombinant FIV Pr50 Gag -His 6 -tag fusion protein alone (without yeast tRNA) could not be assembled in vitro to form VLPs ( Figure 9E-F). However, the size (~89 nm) of the in vitro assembled FIV Pr50 Gag -His 6 -tag particles observed in this study was comparatively larger than the earlier reported size (~33 nm) of FIV Gag in vitro assembled particles [59]. However, both these observations are consistent with other published reports, where the smaller size of the in vitro assembled FIV Pr50 Gag -His 6 -tag particles is more similar to observations made in a number of other retroviruses, including HIV-1, showing a smaller size of the in vitro assembled particles when compared to the Gag particles assembled in vivo [13,59,60,65,69]. Taken together, these results demonstrate that the purified recombinant FIV Pr50 Gag -His 6 -tag fusion protein has the intrinsic ability to assemble in vitro to form VLPs when incubated with nucleic acids. these experiments were carried out in the presence of 4% w/w yeast tRNA, as described previously [13][14][15]. To circumvent the aggregation of the purified recombinant FIV Pr50 Gag -His6-tag fusion protein, the protein was mixed with yeast tRNA in a high salt buffer (1 M NaCl) and dialyzed against a buffer with relevant physiological salt concentration. As an ancillary control, the purified recombinant FIV Pr50 Gag -His6-tag fusion protein without yeast tRNA was also dialyzed in a similar fashion. Dialysis was carried out overnight at 4 °C, following which samples were collected individually from the dialysis cassette, and 10 µL of the 500 µL assembly reaction mixture was spotted on formvar carbon coated grids, negatively stained, and visualized by TEM. The electron micrographs of the structures assembled in vitro by the recombinant FIV Pr50 Gag -His6-tag fusion protein were taken from different fields. As expected, the formation of VLPs in the form of compact electron-dense spherical rings of approximately 89 nm in size confirmed the in vitro assembly process when the recombinant FIV Pr50 Gag -His6-tag fusion protein was mixed with yeast tRNA (Figure 9A-D). On the other hand, as reported earlier [59], the recombinant FIV Pr50 Gag -His6tag fusion protein alone (without yeast tRNA) could not be assembled in vitro to form VLPs ( Figure  9E-F). However, the size (~89 nm) of the in vitro assembled FIV Pr50 Gag -His6-tag particles observed in this study was comparatively larger than the earlier reported size (~33 nm) of FIV Gag in vitro assembled particles [59]. However, both these observations are consistent with other published reports, where the smaller size of the in vitro assembled FIV Pr50 Gag -His6-tag particles is more similar to observations made in a number of other retroviruses, including HIV-1, showing a smaller size of the in vitro assembled particles when compared to the Gag particles assembled in vivo [13,59,60,65,69]. Taken together, these results demonstrate that the purified recombinant FIV Pr50 Gag -His6-tag fusion protein has the intrinsic ability to assemble in vitro to form VLPs when incubated with nucleic acids.

Conclusions
In this study, we successfully expressed and purified a full-length recombinant FIV Pr50 Gag protein with a C-terminus His 6 -tag in a bacterial expression system. The resultant purified recombinant protein had the ability to assemble in vitro to form VLPs. In vivo VLPs formed by the recombinant FIV Pr50 Gag with and without a His 6 -tag showed the ability to specifically recognize and package FIV sub-genomic RNA at equal levels. The ability to express and purify a full-length recombinant Pr50 Gag -His 6 -tag fusion protein will further improve our understanding of the molecular mechanisms underlying various steps in retroviral replication, such as RNA dimerization, packaging, and virus assembly. In vitro binding assays using the purified full-length recombinant Pr50 Gag -His 6 -tag fusion protein could further delineate the mechanism of Gag-RNA interactions during FIV gRNA packaging. Furthermore, the availability of a purified biologically active full-length Pr50 Gag -His 6 -tag fusion protein could allow cryo-EM studies to understand the structural interactions involved in the complex process of Gag multimerization and assembly.