Characterization of the Lassa virus GP1 ectodomain shedding: implications for improved diagnostic platforms

Background There is a significant requirement for the development and acquisition of reagents that will facilitate effective diagnosis, treatment, and prevention of Lassa fever. In this regard, detection of early markers of Lassa virus (LASV) infection may improve diagnosis and ultimately successful treatment with antivirals. Characterization of LASV GP1 ectodomain shedding is an important step toward developing sensitive diagnostics to detect circulating levels of this viral glycoprotein in infected patient sera. Results Secretion of GP1 from mammalian cells expressing a native LASV GPC gene was not mediated by proteolytic cleavage, as determined by treatment with a panel of matrix metalloprotease (MMP) inhibitors. The shedding of GP1 was also not the result of over-expression of GPC under the control of a strong intron-A containing CMV promoter, as the soluble component could be immunoprecipitated from supernatants of cells expressing low levels of GPC under the control of an intronless promoter. Cells transfected with GPC retained surface membrane-associated expression of GP1 as determined by immunofluorescence assay, in addition to secreting the glycoprotein. Secreted GP1 derived from GPC expression has a higher content of high mannose N-linked glycosylation than sGP1 expressed independently from the GP2 portion of the protein. Neither GP1 isoform contains sialylated N-glycans, O-linked carbohydrate chains, or galactose-β(1-4)-N-acetylglucosamine commonly present in complex and hybrid N-glycan structures. Conclusion These results demonstrate the non-proteolytic secretory nature of GP1 shedding during expression of the arenaviral glycoprotein complex. This phenomenon parallels shedding of a secretory glycoprotein component in filovirus replication. The glycosylation pattern of soluble GP1 resulting from expression of GPC was different from that of a soluble GP1 construct (sGP1-RRAA-FLAG), highlighting the intricately orchestrated post translational processing of the LASV glycoprotein complex.

patients and as high as 50% during epidemics [4,5]. Presently, there is no licensed vaccine or immunotherapy available for preventing or treating this disease. Although the antiviral drug ribavirin is beneficial, it must be administered at an early stage of infection to successfully alter disease outcome, thereby limiting its utility [6]. Furthermore, there is no commercially available Lassa fever diagnostic assay, thus preventing early detection and rapid implementation of existing treatment regimens (e.g. ribavirin administration). The lack of adequate countermeasures and means of detection, coupled with the severity of disease, contributed to the classification of LASV as a National Institutes of Allergy and Infectious Diseases (NIAID) Category A pathogen and biosafety level-4 (BSL-4) agent.
The LASV genome is comprised of two ambisense, singlestranded RNA molecules, designated small (S) and large (L) [7]. Two genes on the S segment encode the nucleoprotein (NP) and two envelope glycoproteins (GP1 and GP2); whereas, the L segment encodes the viral polymerase (L protein) and RING finger Z matrix protein. GP1 and GP2 subunits result from post-translational cleavage of a precursor glycoprotein (GPC) by the protease SKI-1/S1P [8]. GP1 serves a putative role in receptor binding, while GP2 has the structural features characteristic of class I viral fusion proteins [9].
Recently we reported that expression of wild type LASV GPC in mammalian cells results in the generation of significant levels of soluble GP1 in the supernatants of transfected cells, that is not associated with GP2 [11]. GP1 ectodomain shedding from cells expressing wild type LASV GPC establishes potential new correlates of disease progression and highlights additional opportunities for development of diagnostics targeting the early stages of Lassa fever. In these studies the mechanism of LASV GP1 ectodomain shedding was further elucidated and its characteristics compared and contrasted to a similar phenomenon in filoviruses, as previously reported for Ebola virus (EBOV) [12].

Matrix metalloprotease inhibitors do not affect the secretion of GP1 from LASV GPC expressing cells
The effects of a diverse set of MMP inhibitors on the secretion of a soluble GP1 component from high level expression of LASV GPC in human cells were investigated. None of the inhibitors employed in these studies resulted in statistically significant reduction in the levels of secreted GP1 from HEK-293T/17 cells expressing LASV GPC ( Figure 1A, D, lanes 1 -12), when compared to untreated controls ( Figure 1A, D, lanes 13,14) [p > 0.05, N = 3]. Relative levels of secreted GP1 for each condition were normalized on GPC expression in the presence of 1% DMSO, the solvent for the majority of inhibitors employed in these studies ( Figure 1A, D, lane 13). Inhibitors were added to cells 12 hours after transfection, the earliest time that GP1 could be detected in cell culture supernatants, thereby marking the onset of glycoprotein secretion ( Figure 1B, lane 17). The effects of inhibitors on cellular metabolism were measured by reduction of tetrazolium salt (MTT) and quantitated by A570 absorbance in 96 well plates mimicking the conditions used in the 6 well plate format. With the exception of MMP-2 inhibitor 1 ( Figure 1E, lane 8) [p < 0.001, N = 3] cellular metabolism was not adversely affected by MMP inhibitors and conditions used in these assays. Despite significantly lowered MTT reduction in cells expressing GPC and treated with MMP-2 inhibitor I, the levels of secreted GP1 were equivalent to those in untreated and control treated cells ( Figure 1D, lane 8). Expression of GPC in HEK-293T/17 cells did not adversely affect cellular metabolism, as MTT reduction levels were nearly identical to those obtained with pcDNA vector alone ( Figure 1E, lanes 13 -16).

Secretory pathway inhibitors abrogate secretion of GP1 from LASV GPC expressing cells
The effects of secretory pathway inhibitors brefeldin-A (BFA) and monensin on expression of GPC and secretion of soluble GP1 were investigated. At 12 hours post transfection intracellular levels of GPC and processed GP1 were approximately 70% of those at 36 hours (Figure 2A, lanes 1 and 2), although secreted GP1 was barely detectable in cell culture supernatants ( Figure 2B, lanes 1 versus 2). Addition of BFA at the onset of transfection with GPC resulted in an average of 80% reduction in intracellular levels of GP1, with undetectable secreted glycoprotein at 36 hours (Figure 2A, B, lane 3). Addition of BFA at 12 hours post transfection resulted in higher levels of accumulated intracellular GP1 than when the inhibitor was added at the onset, but still resulted in undetectable glycoprotein secreted into the culture supernatant. Treatment of cell with monensin resulted in overall higher levels of accumulated intracellular protein than in BFA-treated reactions, both when it was added at the onset or 12 hours post transfection (Figure 2A, lanes 7 and 8). Secretion of GP1 was abrogated when monensin was added at the onset of transfection, although very low levels of the glycoprotein could be detected in supernatants of cells treated with the inhibitor at 12 hours ( Figure 2B, lane 8). Overall, treatment with BFA resulted in lowered intracellular GPC expression than with monensin, irrespective of the time of inhibitor addition to culture medium, as assessed by levels of detectable GP1 protein at 36 hours. Inhibition with BFA targets glycoprotein transport within the ER -cis Golgi environment, whereas monensin targets transport within the trans Golgi apparatus. Thus, BFA is known to affect glycoprotein transport more profoundly in mammalian cells than monensin.

Secretion of GP1 occurs in cells expressing high or low levels of LASV GPC
To assess if secretion of LASV glycoprotein is a phenomenon resulting from high level expression of GPC, intracellular and secreted GP1 levels driven by intron-A containing or intronless CMV promoter constructs were analyzed and quantitated. In HEK-293T/17 cells the average level of intracellular GP1 expression derived from intronless CMV promoter constructs was 16% of that obtained with intron-A counterparts ( Figure 3A, lane 3). Expression from the weaker CMV promoter resulted in undetectable levels of secreted GP1, as assessed by western blot analysis of cell culture supernatants ( Figure 3B, lane 3). In an attempt to detect secreted GP1 from lower intracellular GPC expression, LASV GP1-specific mAb L52-74-7A was used to immunoprecipitate (IP) the glycoprotein from cell culture supernatants. One millilitre of each supernatant was subjected to IP with anti-GP1 mAb and Protein G Sepharose, and the resulting immunoprecipitates were detected by western blot. As shown in Figure 3C the entire IP reaction (1) was resolved for control vector (lane 1), intronless GPC (lane 3), and control GP1-TM (lane 4) constructs. Conversely, a fraction (0.025) of each of the intron-A driven GPC and sGP1 constructs were analyzed in the same blots for comparison ( Figure 3C, lanes 2 and 5). The intracellular ratios of GP1 between intron-A driven and intronless GPC constructs were calculated linearly based upon densitometry. Whereas intracellular GP1 levels derived from expression of an intronless GPC construct were, on average, 16% of those obtained with the intron-A containing counterpart, the resulting secreted GP1 was only 0.11%, or ~1/100 of the cellular ratio (Figure 3C, lane 3). A very minor GP1 band could be immunoprecipitated from the supernatants of GP1-TM expressing cells ( Figure 3C, lane 4). The GP1-TM construct contains the LASV GP2 TM domain directly fused to the Cterminus of the GP1 ORF, and the SKI-1/S1P protease cleavage site, RRLL, remains intact. In this conformation the vast majority of GP1-TM is not cleaved, and GP1 is produced almost exclusively as a membrane-anchored form. Expression of LASV glycoprotein constructs in VERO cells generally produced similar results as those obtained with HEK-293T/17 cells ( Figure 3D), with the notable exception of higher secreted levels of GP1 from intronless CMV promoter-driven GPC ( Figure 3D, lane 3). Also, secreted levels of GP1 from a sGP1_RRAA-FLAG construct were approximately twice those observed with intron-A containing GPC ( Figure 3D, lane 2 versus lane 5). In contrast, the same corresponding levels in HEK-293T/17 cells were nearly matched ( Figure 3A, C, lane 5).
To further assess the levels of secreted GP1 from VERO cells expressing GPC under the control of an intronless CMV promoter, supernatants were subjected with IP with mAb L52-74-7A. Western blot analysis revealed that significantly higher levels of sGP1 could be purified from the supernatants of VERO cell expressed intronless GPC than from HEK-293T/17 cells, as expected, based on analysis on non-immunoprecipitated samples (data not shown).

GP1 remains associated with the cell membrane in LASV GPC expressing cells
Expression and localization of LASV GP1 in HEK-293T/17 cells was analyzed by immunofluorescence assay (IFA) with mAb L52-74-7A on fixed and permeabilized, or live, unfixed cells ( Figure 4). All constructs used in LASV glycoprotein expression for IFA were driven by intron-A containing CMV promoters. Expression of GPC on fixed cells was generally diffuse throughout the cytoplasm, with bright distribution of the antigen throughout the peripheral outline of stained cells ( Figure 4A, panels 1 and 3). Expression of GPC on live cells, which were not permeable to the antibody, and thus stained antigen present of the outer leaflet of the membrane, showed primarily bright punctate staining throughout the cellular periphery

Secreted GP1 derived from expression of LASV GPC and sGP1 have disparate glycosylation patterns
The glycosylation profile of secreted GP1 was derived by treatment of immunoprecipitated protein with PNGase-F, Endo-H, or Neuraminidase, followed by western blot analysis of protein migration on SDS-PAGE. The LASV GP1 glycoprotein contains 7 predicted N-linked glycosylation sites that could be accessible to removal by PNGase-F. Secreted GP1 from GPC expression was fully deglycosylated by PNGase-F, as determined by a loss of ca. 20 KDa from the mass of the protein detected on western blots ( Figure 5A, lane 2). The expected mass of LASV GP1 without carbohydrate moieties is 22771.82 Daltons (ExPASy Compute pI/Mw tool [13][14][15]). Expression of sGP1-RRAA-FLAG produced a secreted form of GP1 that was largely deglycosylated by PNGase-F, with an additional moiety at ca. 24 KDa that could not be further deglycosylated with higher levels of enzyme, additional digestion time, or further denaturation (N = 4) ( Figure 5B

Secreted GP1 from GPC and sGP1 expression have identical N-terminal sequences
Immunoprecipitated and PNGase-F deglycosylated GP1 protein from GPC and sGP1-RRAA-FLAG were subjected to SDS-PAGE and transferred to PVDF membranes for Nterminal sequence analysis by Edman degradation. The six N-terminal amino acids from each protein were T-S-L-Y-K-G, thus representing the SPase-cleaved form of the LASV GP1 protein, as previously characterized [8]. , and apply to all panels. Lane designations above A. apply to all panels.

Discussion
A panel of soluble MMP inhibitors with a wide range of specificities was employed in the characterization of protease-mediated GP1 shedding from mammalian cells expressing LASV GPC. A broad spectrum hydroxamic acid inhibitor of MMPs, GM 6001 (inhibits MMP1, 2, 3, 8,9), metal ion chelators (EDTA, EGTA, 2,2 dipyridyl), a specific inhibitor of chymotrypsin-like serine proteases (α 1antichymotrypsin), a cell surface aminopeptidase inhibitor (Bestatin), an inhibitor of thermolysin-like protease and metalloendopeptidases (Phosphoramidon), a broad MMP inhibitor (phenanthroline), TNF-α convertase (TACE) protease inhibitor-1 (TAPI-1), and specific inhibitors of MMPs (MMP2, 9), were added to growth medium of LASV GPC-expressing mammalian cells in an attempt to determine their role in the release of soluble GP1. None of the inhibitors showed statistically significant inhibition of GP1 shedding from HEK-293T/17 cells expressing LASV GPC, although most compounds were tested at concentrations significantly greater than reported IC 50 values (Table 1). These results strongly suggest that GP1 is released from cells as a soluble and uncoupled protein arising from the intracellular cleavage and processing of the GPC complex. Further, it can be proposed that the 1:1 stoichiometry of GP1 and GP2 in cells expressing GPC does not result in a 1:1 ratio of assembled tripartite complex on the cell surface. It is possible that tripartite complex assembly in the ER results in retention of a fraction of GP2 in favour of GP1 secretion. This model is partly supported by our previous studies that established a critical role for GP1 as a chaperone for correct processing and secretion of GP2 [11]. These studies also suggested that cell surface expression of GP2 without associated GP1 was unlikely; GP2 expression resulted in very low levels of intracellular protein, and it was undetectable on the cell surface (unpublished data). Cytotoxicity of MMP inhibitors was monitored by MTT uptake assays. MMP-2 inhibitor I resulted in significantly cytotoxicity compared to untreated controls (p < 0.001, N = 3), but the resulting release of GP1 from cells was not significantly affected ( Figure 1A, D, E, lane 8), indicating that de novo protein synthesis and cellular secretory pathways were not targeted by the inhibitor.
The mammalian expression system employed in these studies was originally developed for high level production of recombinant LASV proteins for diagnostic assay development [11]. High level expression of soluble LASV GPC resulted in both secretion of glycoprotein precursor uncleaved by SKI-1/S1P, and high levels of sGP1. Although CMV promoter strength has not been compared to that of the LASV promoter, it is reasonable to suggest that very high level expression of recombinant LASV genes in mammalian cells could result in artefactual processing, secretion, and stoichiometry of glycoprotein complex components. It is evident that high level expression of soluble GPC in HEK-293T/17 cells driven by CMV plus intron-A overwhelms cellular SKI-1/S1P protease, as approximately one half of the secreted glycoprotein is uncleaved [11]. In order to address the possibility that overexpression of GPC resulted in artefactual secretion of GP1 we expressed the glycoprotein from an intronless CMV-driven promoter. Expression of GPC from CMVintron-A driven vectors resulted in 8 -10 fold higher levels of cleaved intracellular GP1 than from intronless counterparts. Low levels of sGP1 in the supernatants of cells expressing intronless GPC were detected only when the glycoprotein was immunoprecipitated. An 8 -10 fold dilution of culture supernatant from cells expressing sGP1 still resulted in detection of significant levels of secreted glycoprotein, whereas similar levels of immunoprecipitated GP1 from intronless GPC expression required 400 fold higher volume of supernatant. These results suggest that overexpression of GPC results in disproportionately enhanced secretion of sGP1 when compared to lower expression levels of the glycoprotein. Generally, expression and secretion levels of GP1 in the HEK-293T/17 line paralleled those obtained with VERO cells, a common line used for the propagation of Lassa virions and analysis of LASV gene expression ( Figure 3D).
Association of GP1 with the cell membrane as a component of the tripartite glycoprotein complex was verified by staining live cells overexpressing GPC. After extensive cell washing to remove sGP1, followed by staining with a LASV GP1-specific mAb, cell surface associated GP1 could be readily detected. The staining pattern was less diffuse than with fixed and permeabilized cells overexpressing GPC, with bright fluorescence primarily associated with the cell membrane. Thus, a fraction of the GP1 protein associates with GP2 and SSP on the cell membrane of GPC-transfected cells, whereas a soluble, un-associated component is shed into the cellular milieu.
Sequencing of GP1 protein from expression of GPC or sGP1 revealed identical N-termini through six amino acids, suggesting that both forms of the glycoprotein result from cleavage of pre-GPC by SPase at amino acid 58, and that the SSP functions as the secretory signal.   [16], the possibility of this form of linkage in sGP1-RRAA-FLAG exists. This type of modification is most commonly found in plant and some insect glycoproteins, however. Computational analysis of the LASV GP1 protein sequence using OGPET v1.0 (University of Texas, El Paso [17] revealed the possibility of an O-GalNAc (mucin type) glycosylation in GP1 at Ser 80 (SnISTnHL), with a predictive score of 0.599833 (cutoff = 0.4247). Additionally, NetGlycate 1.0 Server [18], a predictor of glycation of ε amino groups of lysines in mammalian glycoproteins, attributed a score of 0.914 to Lys5 in LASV GP1, which is significantly above the threshold of 0.5. Lectin-binding assays did not aid in resolving the glycosylation pattern of the 24 KDa form of sGP1-RRAA-FLAG. The lectins PNA (Peanut agglutinin) and DSA (Datura stramonium agglutinin), which recognize the core disaccharide galactose (1-3) N-acetylgalactosamine, and Gal-(1-4)GlcNAc in complex and hybrid glycans, in O-glycans, and GlnNAc in O-glycans, respectively, did not bind to the 24 KDa sGP1 form. A higher degree of N-linked glycosylation heterogeneity on LASV GP1 was obtained in Endo H digestion profiles. Endo H removes only high mannose and some hybrid types of N-linked carbohydrates from glycoproteins, and will not cleave more complex glycan structures [19,20]. Soluble GP1 from GPC expression revealed a more heterogeneous distribution of high mannose and hybrid glycosylation than were present in sGP1-RRAA-FLAG, indicating a higher content of mannose and/or hybrid structures. Conversely, it appears that sGP1-RRAA-FLAG has a significantly higher content of complex glycan structures, as determined by reduced EndoH activity on the glycoprotein. These results were confirmed by the differential staining of both GP1 forms by GNA; soluble GP1 from GPC expression was heavily stained with GNA, whereas sGP1-RRAA-FLAG showed significantly dimmer reactivity. Densitometry analysis showed a ratio of 2.27 ± 0.014 in GNA staining intensity of GPC-derived GP1 versus sGP1-RRAA-FLAG (N = 3). The marked differences in glycosylation of each form of LASV GP1 could be attributed to a potentially different translational and post-translational modification profile and timeline in expressing cells. The process of translation, ER translocation, SPase and SKI-1/S1P cleavage, ER retention of glycoproteins, and assembly of the LASV tripartite complex (GP1+GP2+SSP) has been somewhat dissected, and constitutes a finely orchestrated process in the ER and Golgi of mammalian cells. The correct processing and assembly of the tripartite complex is essential for subsequent binding of LASV virions via GP1 to the α-dystro-glycan cellular receptor, and initiation of viral entry via the fusion peptide on the N-terminus of GP2, a process largely mediated by the presence of the intact SSP peptide. Thus, it is conceivable that LASV GPC-derived glycoproteins remain accessible to ER resident glucosyltransferases, and the homologous lectins calnexin and calreticulin longer than sGP1-RRAA-FLAG, resulting in an overall different endpoint glycosylation pattern. Glucosyltransferases act as one of the primary protein folding quality control mechanisms in the ER, in concert with lectins, peptidyl-propyl-isomerases, and chaperones that are tightly regulated by nuclear transcription mediated by ER membrane resident proteins (e.g. ATF6) that sense high levels of protein misfolding in this organelle [21]. The formation of hybrid and complex glycans takes place in the medial and trans Golgi compartments. Therefore, proteins residing in the ER which have not passed through the Golgi cannot possess hybrid or complex N-linked structures. Thus, the intricate assembly process for the LASV GPC complex would be expected to differ from that of single glycoprotein expression from the sGP1-RRAA-FLAG construct. The kinetics of GP1 secretion from expression of GPC and sGP1-RRAA-FLAG do not differ significantly, as the onset for detection of each form of the glycoprotein in the extracellular medium occurs at approximately 12 hours post transfection (with CMV+intronA driven constructs). Post-translational processing and secretion of sGP1-RRAA-FLAG is not necessarily faster than GPC expressed GP1, but it is executed differently. It is noteworthy to point that such findings contrast with previously reported results by our group that showed clear and complete removal of high mannose glycans from sGP1 and sGP2 generated with the same expression constructs and cell type [11]. Such contrasting results point to subtle differences in culture conditions that could significantly affect resulting glycosylation patterns. Although the HEK-293T/17 clone used in both studies was obtained from the same source (American Type Culture Collection [ATCC]), the studies were performed with different lots of culture media and supplements, namely fetal bovine serum. It has been extensively reported that subtle differences in cell culture environments, including changes in medium composition can result in significant differences in glycosylation patterns of secreted glycoproteins. Such impact on glycosylation is a primary focus of product quality control in many glycoprotein-based biopharmaceutical products, such as therapeutic antibodies, erythropoietin, and reproductive hormones [22][23][24][25][26]. Lastly, treatment of secreted GP1 from GPC and sGP1-RRAA-FLAG expression did not result in observable shifts with neuraminidase. Neuraminidase (Acetyl-neuraminyl hydrolase; Sialidase) catalyzes the hydrolysis of α2-3, α2-6, and α2-8 linked Nacetyl-neuraminic acid residues from glycoproteins and oligosaccharides [27,28]. These results were further confirmed by the lack of binding to both forms of GP1 by the lectins MAA (Maackia amurensis agglutinin) and SNA (Sambucus nigra agglutinin), which are specific for α2-3 and α2-6 galactose, respectively. Thus, it appears that LASV GP1 is not synthesized with terminal sialic acid residues in its carbohydrate profile, irrespective of the glycoprotein precursor.
It is well known that hydrophobic interactions predominate over the proper and improper folding of proteins. Further, the hydrophobic areas of high mannose N-glycans are closely associated with promotive effects on protein folding, possibly depressing the hydrophobic interactions unfavorable to the proper folding of nascent polypeptides in the E.R. It has been established by Jitsuhara et al. [29], Narhi et al. [30], and others that higher oligomannose chains of glycosylated proteins promote protein folding more strongly than lower ones [31]. Of noteworthy mention is the folding of influenza virus hemagglutinin with a truncated N-glycan Glc 1 Man 5 GlcNAc 2 moiety that normally interacts with the lectin calnexin in the E.R. is markedly depressed [32]. The E.R. resident lectins calnexin and calreticulin tether monoglucosylated high mannose N-glycans from nascent proteins, resulting in suppression of aggregation, thus promoting protein folding [33]. Many studies have established the correlation between proper protein glycosylation with the acquisition and retention of glycoprotein biological activities. It is reasonable to expect that in the arenaviral tripartite complex the proper folding of the individual glycoprotein subunits, and the interaction with the SSP, have critical roles in establishing functional binding of virions to host cell receptors and subsequent fusion and viral entry. Our own previous LASV GP1 protein expression studies in multiple biological systems have resulted in markedly different results. Although expression of full length GP1 in E. coli was achievable, the lack of glycosylation in the prokaryotic system required the presence of sodium dodecyl sulfate (SDS), reducing agents (e.g. dithiothreitol [DTT]), and the zwitterion morpholino-ethanesulfonic acid (MES) in the purification scheme in order to yield non-aggregated protein [34]. Further processing of bacterial expressed GP1 protein for analytical assays that required removal of SDS and DTT readily resulted in aggregation. In contrast, expression of multiple forms of LASV GP1 in mammalian cells always resulted in glycosylated, soluble protein, although aggregation of the standalone glycoprotein remained an issue [11]. Further characterization of post-translational processing of the arenaviral glycoprotein complex may be of crucial importance in the establishment of viral infectivity correlates, as well as pave the way for identification of disease susceptibility markers, and development of effective vaccines and therapeutics.
The present studies affirm the transition of GP1 from the ER to the Golgi in its path toward secretion, and that the secretory component of GPC follows the same route. Complete or nearly complete inhibition of GP1 secretion in supernatants of cells transfected with GPC and treated with BFA and monensin is indicative of a secretory pathway mediated phenomenon. This conclusion is further strengthened by the lack of inhibition of GP1 secretion from GPC-expressing cells treated with a wide range of MMP inhibitors. Despite secretion of GP1 from GPCexpressing cells, a surface membrane-associated GP1 component was specifically visualized on live cells. These observations are indicative of a secreted GP1 component, in addition to the presence of a cell surface expressed, functional tripartite glycoprotein complex. A proposed model outlining the secretory pathway of LASV GP1 is graphically displayed in Figure 6.
Together, these data strongly suggest dual modalities in LASV GPC expression; the formation of a fully functional tripartite complex for display on the surface of infectious virions, and the secretion of a soluble GP1 component with as of yet uncharacterized function. The fate of unassociated GP2 during expression of LASV GPC was not investigated in these studies, but it warrants analysis, as the stoichiometry of the glycoprotein complex depends on the ratio of the available individual components. Dolnik et al. [12] have partly characterized the role of a soluble GP component in EBOV, and have proposed a role in the infectivity of the virus in vivo. According to Dolnik et al., the secreted EBOV GP component acts as an immune decoy, by targeting the humoral immune system and possibly removing from circulation antibodies with innate neutralizing activity against the virus. This mechanism would provide circulating infectious virions a clearer path toward infection of new target cells, with enhanced virulence. Also, Wahl-Jensen et al. have further delineated the role of secreted EBOV GP in activation of primary target cells [35,36]. These studies revealed that rigid presentation of the EBOV glycoprotein, as it's displayed on the surface of virions is critical for activation of proinflammatory cytokines and chemokines, whereas soluble glycoprotein failed to elicit any response. Additionally, EBOV GP(1,2) in viral particle-associated form mediated endothelial cell activation and decreased cell barrier function, whereas sGP, the major soluble glycoprotein of EBOV, demonstrated an anti-inflammatory role by protecting the endothelial cell barrier function.
Although a similar mechanism has not been ascribed to arenaviruses, our work presents a platform upon which to investigate this phenomenon in vivo. The secreted GP1 component resulting from GPC expression has not been detected in the serum of LASV infected patients, and further work will be performed by our group in an attempt to test the feasibility of this hypothesis. If the phenomenon of GP1 shedding is present in vivo the glycoprotein should be preferentially detected in the early stages of arenaviral infection, when circulating antigens are present at the highest levels. Soluble GP1 in serum could be separated from virions by selective size filtration, thus providing a simple and easily deployable platform for identification of the soluble glycoprotein component. If confirmed, arenaviral glycoprotein shedding could result in the development of more sensitive and earlier detection diagnostics, as well as establish new correlates of pathogenesis and immune evasion by this class of hemorrhagic fever viruses.

Conclusion
The shedding of GP1 from cells expressing GPC results from secretion of a cleaved and un-associated fraction of the protein. This phenomenon is not the result of cleavage of GP1 from the tripartite complex expressed on the cell surface, and it is not mediated by cellular MMPs. Instead, secretion of GP1 occurs when GPC is processed into GP1 and GP2 in the ER, and a significant fraction of this protein is shuttled through the secretory pathway and is released into the extracellular medium. Both forms of secreted GP1 share N-terminal sequences, but have dissimilar glycosylation profiles that suggest a differential Schematic representation of arenaviral glycoprotein expression pathway and secretory forms Figure 6 Schematic representation of arenaviral glycoprotein expression pathway and secretory forms. Arenaviral glycoprotein (GP) mRNA is transcribed from plasmid DNA in the host cell (1), and is directed to the rough endoplasmic reticulum (RER) for translation by associated ribosomes. The nascent polyprotein is directed to the RER lumen by the unique arenaviral SP, where SPases cleave the SP from the GP polyprotein precursor (2). In the RER the GP precursor is cleaved by SKI-1/S1P to yield GP1 and GP2 subunits (3). Wild type GP is possibly comprised of covalently linked, or non-covalently associated GP1 and GP2 subunits, and associated SP (3). Glycosylation of GP1 and GP2 is mediated by glycosyltransferases in the RER or in the Golgi cisternae (4). Membrane bound vesicles transport proteins to the cis face of the Golgi where final processing and assembly of the GP complex occurs. Fully assembled and glycosylated GP complex or subunits are packaged in exocytic vesicles which emerge from the trans face of the Golgi and are directed toward the plasma membrane. Glycoprotein complex is anchored to the cell membrane via the transmembrane domain of GP2 (5a). A complexed heterodimeric GP1-GP2 is displayed in 5a. Extracellular GP1 originating from GPC expression is a secreted glycoprotein isoform (5b). The positions of BFA and monensin inhibitory action along the secretory pathway are indicated. The role of LASV GP2 in the secretory pathway of GP1 was not elucidated in these studies (black box). A graphic representation of each relevant component in the outlined pathway is shown in the legend box.
retention and glycoprotein processing timeline in the ER and Golgi compartments.

Transient expression of LASV gene constructs
Recombinant LASV protein expression was analyzed in HEK-293T/17 or VERO cells transiently transfected with mammalian expression vector DNAs, which were prepared using the PureLink HiPure plasmid filter midiprep kit (Invitrogen, Carlsbad, CA). The negative control vector pcDNA3.1(+):intA was included in all transfections. Briefly, 1 × 10 6 cells were seeded per well of a Poly-D-Lysine-coated 6-well plate in 2 mL of cDMEM. After overnight incubation in growth conditions cells were transfected with unrestricted recombinant plasmid DNAs using Lipofectamine™ 2000 (Invitrogen), according to the manufacturer's instructions. Four μg of each plasmid DNA were used per transfection. Transfections were incubated for the times described in results, at 37°C, 5% CO 2 , 90% RH, after which cell culture supernatants were collected and clarified by centrifugation. To prepare cell extracts from transfected cultures, cell monolayers were carefully washed twice with Ca ++ -and Mg ++ -free PBS, pH7.4, collected by gentle dislodging, transferred to 1.5 mL polypropylene tubes, and lysed for 10 minutes in a mammalian cell lysis buffer comprised of 50 mM Tris buffer, pH 7.5, 1 mM EDTA, 0.1% SDS, 0.5% deoxycholic acid, 1% Igepal CA-360, and a protease inhibitor cocktail (Sigma Aldrich, St. Louis, MO), according to the manufacturer's instruc-tions. The insoluble fraction was pelleted by centrifugation at 14,000 × g for 10 minutes, and the supernatants were transferred to fresh tubes. Protein concentration was determined for each sample by A280 with A260 subtraction, and verified using a Micro BCA™ Protein Assay Kit, as outlined by the manufacturer (Thermo Scientific).

Metalloprotease and secretory pathway inhibition assays
Metalloprotease inhibitors were purchased from Sigma (1,10-Phenanthroline monohydrate, 2,2'-Dipyridyl Rea-gentPlus ® , Bestatin hydrochloride, Phosphoramidon disodium salt, α 1 -Antichymotrypsin, EDTA, EGTA) or Calbiochem (MMP-2 Inhibitor I, MMP-9 Inhibitor I, InSolution™ GM 6001, InSolution™ TAPI-1, Captopril), and resuspended in 100% DMSO (Sigma), or with cell culture medium when appropriate, at 100× concentration. Inhibitors were diluted in growth medium for analysis, at a final DMSO concentration of 1%. For inhibition assays HEK-293T/17 cells were transiently transfected with plasmid DNAs in 6-well plates, as described above, for 12 hours prior to the addition of metalloprotease inhibitors. Transfection medium was then removed and cells were washed twice with 1× PBS, pH7.4, followed by addition of 2 mL of fresh growth medium containing inhibitors. Plates were returned to the growth chamber and reactions were harvested between 42 -45 hours post transfection by collecting supernatants and generating whole cell protein extracts, as described above. The Golgi secretory pathway inhibitors brefeldin A and monensin were purchased from Sigma, and diluted in growth medium to 15 μg/mL and 10 μM, respectively. Inhibitors were added to cells at the time of transfection with pcDNA3.1(+):intA or pcDNA3.1(+):intA:LASV GPC, or at 12 hours. Supernatants and cell extracts were harvested at 36 hours post transfection for western blot analysis.

Western blot and densitometry analyses
Expression of LASV GP1 in cell extracts and culture supernatants was confirmed by Western blot analysis using anti-LASV GP1 mAb L52-74-7A and a horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (H+L) secondary antibody. Briefly, 10 μg of total cell protein in 10 μL (~1 × 10 5 cell equivalents), or 26 μL of cell culture supernatant were resolved on 10% NuPAGE Novex Bis-Tris gels, according to the manufacturer's specifications (Novex, San Diego, CA). All samples in these studies were denatured and reduced in SDS-PAGE buffer containing DTT. Proteins were transferred to 0.45-μm nitrocellulose membranes, blocked, and probed in 1× PBS, pH 7.4, 5% non-fat dry milk, 0.05% Tween-20, and 0.1% thymerosal. Membranes were then incubated in LumiGlo chemiluminescent substrate (KPL) and exposed to Kodak BioMax MS Film. Developed films were subjected to high resolution scanning for densitometry analysis. Quantification of band intensity was performed using National Institutes of drafted the manuscript. RFG contributed to the experimental design and provided critical review of the manuscript.