Virus-derived peptide inhibitors of the herpes simplex virus type 1 nuclear egress complex

Herpesviruses infect a majority of the human population, establishing lifelong latent infections for which there is no cure. Periodic viral reactivation spreads infection to new hosts while causing various disease states particularly detrimental in the immunocompromised. Efficient viral replication, and ultimately the spread of infection, is dependent on the nuclear egress complex (NEC), a conserved viral heterodimer that helps translocate viral capsids from the nucleus to the cytoplasm where they mature into infectious virions. Here, we have identified peptides, derived from the capsid protein UL25, that are capable of inhibiting the membrane-budding activity of the NEC from herpes simplex virus type 1 in vitro. We show that the inhibitory ability of the peptides depends on their length and the propensity to form an α-helix but not on the exact amino acid sequence. Current therapeutics that target viral DNA replication machinery are rendered ineffective by drug resistance due to viral mutations. Our results establish a basis for the development of an alternative class of inhibitors against nuclear egress, an essential step in herpesvirus replication, potentially expanding the current repertoire of available therapeutics.

NEC budding at a 10:1 molar ratio of UL25:NEC 20 . Using cryo-ET, we showed that the observed inhibitory effect on the NEC budding correlated with a network of interconnected UL25 stars formed on top of the membranebound NEC layer. We hypothesized that this UL25 network blocked budding by preventing the conformational changes within the NEC necessary to generate membrane deformation and budding. UL25-mediated budding inhibition required the N-terminal α-helix in UL25, residues 45-94 (Fig. 1a), because the UL25∆73 construct, in which the N-terminal ~ 1/2 of this helix was truncated, did not inhibit budding. We then tested whether peptides derived from the N-terminal α-helix of UL25 could inhibit NEC-mediated budding. Here, we show that the UL25-derived peptides inhibit NEC budding and identify structural characteristics essential for inhibition.
UL25∆44 Q72A, which served as a control, was not inhibitory at a 1:1 molar ratio (Fig. 2c), in agreement with our previous findings 20 ; yet at a 1:10 molar ratio of NEC:UL25, it reduced NEC budding by ~ 90% (Fig. 2c). The differences in inhibition between UL25 and UL25-derived peptides are likely due to distinct inhibitory mechanisms.
The ability of the peptides to form α-helical structure correlates with inhibition of NEC-mediated budding. Within capsid-bound UL25, amino acids 48-94 form an α-helix (Fig. 1a). Yet, the circular dichroism (CD) spectra of both the UL25 native(45-94) and the UL25 scr(45-94) peptides lacked any obvious α-helical signature in solution (Fig. 3a, b) and were only ~ 10% α-helical according to the estimates using DichroWeb 23 (Supplementary Table S1). To assess the propensity of the UL25 peptides to form α-helical structure, the CD experiment was repeated in the presence of 30% trifluoroethanol (TFE), which is known to stabilize secondary   Table S1), suggesting that each peptide could, in principle, form an α-helix in the proper environment, for example, when bound to a protein partner such as the NEC. The θ 222 /θ 208 ratios of both peptides either in the absence or the presence of TFE were < 0.9 indicating peptides were forming individual helices rather than coiled coils 25 . The shorter UL25 scr(74-94) peptide adopted a random-coil conformation in solution even in the presence of 30% TFE (Fig. 3c, Supplementary Table 1). Therefore, it may be unable to inhibit NEC budding to a similar extent as the longer UL25 native(45-94) and the UL25 scr(45-94) peptides because it cannot form an α-helix.
A pre-formed α-helical peptide moderately inhibits budding whereas a random-coil peptide does not. The inhibitory UL25-derived peptides can form α-helical structure whereas the non-inhibitory peptide does not, implying a correlation between the two properties. However, the inhibitory peptides are also substantially longer than the non-inhibitory one. Therefore, to explore the apparent correlation between the propensity to form an α-helix and the ability to inhibit NEC-mediated budding while ruling out the peptide length as a variable, we tested two heterologous peptides: a 50-amino-acid sequence from HSV-1 glycoprotein B (gB 103-152 ), which lacks any regular secondary structure in the crystal structure 26 and a 53-amino-acid peptide containing a series of well-characterized α-helical EAEKAAK repeats [(EAEKAAK) 7 EAEK] 27 (Fig. 1b). The gB 103-152 peptide was not inhibitory at a 1:1 molar ratio of NEC:peptide (Fig. 2c) and only had a weak inhibitory effect on budding at a 1:10 molar ratio of NEC:peptide, ~ 25%, which was not statistically significant (Fig. 2c). CD analysis showed that this peptide adopted a random-coil conformation in solution even in the presence of 30% TFE (Fig. 3d, Supplementary Table 1). These observations suggest that the ability to form an α-helix is important for the inhibitory activity of the peptides. Budding was tested at 1:0.1, 1:0.3, 1:0.6, 1:1 or 1:10 molar ratios of NEC:UL25 or NEC:peptide. Each experiment was done in three biological replicates, each consisting of three technical replicates. Symbols represent average budding efficiency of each biological replicate relative to NEC220 (100%). Bars show the mean of three biological replicates, with error bars representing the standard error of the mean. Significance compared to NEC220 was calculated using an unpaired t-test. *P value < 0.05, **P value < 0.01, ***P value < 0.001, and ****P value < 0.0001.  Table 1), in contrast to both the UL25 native(45-94) and the UL25 scr(45-94) peptides that became α-helical only in the presence of TFE (Fig. 3a, b). Yet, it was not inhibitory at a 1:1 molar ratio of NEC:peptide and moderately inhibited NEC-mediated budding at a 1:10 molar ratio, by ~ 50% (Fig. 2c). The differences in the inhibitory activities of the three peptides that can form α-helices could be due to different amino acid composition, especially, the charged residue content. The (EAEKAAK) 7 EAEK peptide contains 60% charged residues (glutamates and lysines) whereas the UL25-derived inhibitory peptides contain only 30% charged residues (aspartates, glutamates and arginines). The moderate inhibitory activity of the (EAEKAAK) 7 EAEK peptide at high concentrations could be due to its non-specific interactions with the NEC due to high charge. Alternatively, optimal inhibition may require a peptide that forms an α-helix only in the presence of the NEC, perhaps, in an induced-fit manner, instead of a peptide that exists as a stable, pre-formed α-helix.

Inhibitory peptides do not block NEC/membrane interactions.
Peptides could inhibit NEC-mediated budding by preventing the NEC from binding to membranes. We first tested whether the peptides themselves could bind multilamellar vesicles (MLVs) of the same lipid composition as the GUVs used in the budding assay and found that none did (Fig. 4). This suggested that the peptides do not compete with the NEC for  To assess if the NEC could bind membranes in presence of inhibitory peptides, we performed a co-sedimentation assay with the NEC, MLVs, and either UL25 native(45-94) or UL25 scr(45-94) . UL25∆44 Q72A, which inhibits budding but does not prevent the NEC from binding membranes 20 , was used as a control. Experiments were performed at a 1:10 molar ratio of NEC:peptide or NEC:UL25 because the UL25∆44 Q72A control inhibits NECmediated budding at a 1:10 molar ratio of NEC:UL25. We found that neither peptide prevented the NEC from binding membranes as judged by the same amount of MLV-bound NEC (~ 95%) in the absence or the presence of peptides (Fig. 5). As expected 20 , UL25∆44 Q72A also did not block NEC membrane binding (Fig. 5). Thus, the inhibitory peptides do not block NEC/membrane interactions and, instead, inhibit NEC budding by some other mechanism.

Discussion
Here, we have shown that two peptides derived from the N-terminal α-helix of UL25 efficiently inhibited NECmediated budding in vitro in a dose-dependent manner. The inhibitory propensity of the peptides depended on their length and the propensity to form an α-helix but not on the exact amino acid sequence. Peptide length was more important for inhibition than the sequence because 50-amino-acid UL25 peptides, be they native or scrambled (UL25 native(45-94) and UL25 scr(45-94) ), efficiently inhibited budding whereas the 21-amino-acid native peptide UL25 74-94 did not. Future studies will identify the minimal UL25 peptide length required for inhibition. The inhibitory ability of UL25 peptides also correlated with the propensity to adopt α-helical structure. Although all three UL25-derived peptides formed random coils in solution, the two 50-amino-acid peptides formed α-helices in the presence of TFE, known to stabilize secondary structure, whereas the 21-amino-acid native peptide did not.
To uncouple the propensity to form α-helix from the peptide length, we tested a heterologous 50-aminoacid peptide from HSV-1 gB (gB 103-152 ) that lacks any regular secondary structure in the crystal structure 26 . We confirmed that this peptide formed a random coil and showed that it had a weak inhibitory activity only at a 1:10 molar ratio of NEC:peptide.
We also tested a heterologous 53-amino-acid peptide containing a series of well-characterized α-helical EAEKAAK repeats [(EAEKAAK) 7 EAEK] 27 . We found that, unlike the inhibitory UL25-derived peptides, this peptide was α-helical in solution even in the absence of TFE and that it had a modest inhibitory effect only at a 1:10 molar ratio of NEC:peptide. This heterologous α-helical peptide is highly charged and forms an amphipathic α-helix with a charged face, rich in glutamates and lysines, and a hydrophobic face, rich in alanines ( Supplementary Fig. S2). In contrast, the inhibitory UL25-derived peptides are less charged and form α-helices with a less pronounced amphipathic character (Supplementary Fig. S2). The moderate inhibitory activity of the (EAEKAAK) 7 EAEK peptide at high concentrations could thus be due to its non-specific interactions with the NEC. Alternatively, it is possible that a stable, preformed α-helix may not be able to form appropriate inhibitory contacts with the NEC and that for efficient inhibition, an α-helix of a specific length and amino-acid content . Peptides do not bind synthetic liposomes. NEC but not peptides bind multilamellar vesicles (MLVs) of the same composition as the GUVs used in the in-vitro budding assay. PreScission protease, which does not bind membranes, was used as a negative control. Each experiment was done in three biological replicates, each consisting of two technical replicates. Symbols represent average binding efficiency of each biological replicate relative to NEC220 (100%). Bars show the mean of three biological replicates, with error bars representing the standard error of the mean. Significance compared to NEC220 was calculated using an unpaired t-test. ****P value < 0.0001. www.nature.com/scientificreports/ should form only in the presence of the NEC, in an induced-fit manner. Therefore, future design of NEC peptide inhibitors should take both of these characteristics into consideration. The mechanism by which the inhibitory peptides inhibit budding is yet unclear. Our data suggest that the inhibitory peptides do not prevent the NEC from binding the membranes. Therefore, we hypothesize that the peptides may, instead, inhibit the membrane-budding activity of the NEC by some other mechanism, for example, Figure 5. Peptides do not block membrane binding by the NEC. Co-sedimentation of NEC220 alone and in the presence of either UL25 native(45-94) , UL25 scr(45-94) , or UL25∆44 Q72A was assessed by a co-sedimentation assay in the presence and absence of MLVs. The amount of co-sedimentation was quantified with Coomassie staining. Bands from representative gels (full gel images are shown in Supplementary Fig. 1) used for quantification are boxed and colored as follows: pink-NEC (UL31: 34 kDa and UL34: 25 kDa); green-either UL25 native(45-94) (5 kDa) or UL25∆44 Q72A (57 kDa); blue -UL25 scr(45-94 ) (5 kDa). Each experiment was done in three biological replicates, each consisting of two technical replicates. Symbols represent average binding efficiency of each biological replicate. Bars show the mean of three biological replicates, with error bars representing the standard error of the mean. Significance compared to NEC220 in the presence of MLVs was calculated using an unpaired t-test. ****P value < 0.0001. www.nature.com/scientificreports/ by blocking NEC/NEC interactions required for budding or by preventing the conformational changes within the NEC required for budding (Fig. 6).

Scientific Reports
Our previous study suggested that UL25 inhibited NEC-mediated budding by oligomerizing into a network of interconnected stars on top of the membrane-bound NEC layer, as observed by cryo-ET 20 . We hypothesized that the N-terminal helices linked the neighboring UL25 molecules into a network of stars by forming antiparallel helical bundles that sterically blocked the NEC from undergoing conformational changes necessary for membrane budding. Both the N-terminal helix and the C-terminal core of UL25 (Fig. 1a) were important for inhibition. However, the inhibitory UL25 peptides in our current study were composed of only the N-terminal helix. Therefore, they likely bind the NEC differently from UL25 and use a different mechanism to inhibit budding. Being much smaller than UL25, the inhibitory peptides can interact with the NEC at sites otherwise inaccessible to the UL25 protein. Alternatively, NEC/peptide interactions may block NEC/NEC interactions necessary for the assembly of the hexagonal NEC coat thereby blocking the induction of negative curvature and, ultimately, budding (Fig. 6). Additional experiments, such as fluorescent labeling of the peptides, are necessary to show peptide binding to membrane-bound NEC.
Our results suggest that inhibitory peptides should be of a certain length, moderately charged, and unstructured in solution yet capable of forming α-helical structure, presumably, upon binding to the NEC. Further structural analyses of the inhibitory peptides in the presence of the NEC are necessary to narrow down the optimal characteristics of the inhibitory peptides, including their delivery to their intracellular target. The length of the inhibitory peptides identified in this study, 50 amino acids, presents challenges for intracellular delivery. Most therapeutic peptides are ~ 20 to 30 amino acids in length, which increases the likelihood of crossing the plasma membrane, yet many require additional alterations of their physical properties to do so efficiently 28 . Furthermore, peptides are susceptible to proteolytic degradation and can exhibit off-target effects in cells, both of which become more problematic with increasing length, reducing the chances of binding the intended target. Therefore, determining the minimal inhibitory length of these peptides will be essential for transitioning from in-vitro to in-vivo studies. Lastly, the NEC is located within the nucleus, which means that the inhibitory peptides must not only reach the cytoplasm but also enter the nucleus. A recent peptide cellular localization assay showed that peptides smaller than 40 kDa can diffuse through the nuclear pore 29 , suggesting that the NEC-derived inhibitory peptides described here could do so as well.
Efficient NEC function is required for successful replication of herpesviruses. To our knowledge, this is the first study reporting a peptide-based approach targeting NEC budding. Our work provides a starting point for the rational design of peptide-based inhibitors of the NEC and nuclear egress. Both the NEC and UL25 are conserved across all herpesviruses, which raises an intriguing possibility that UL25-derived peptides could also be used to inhibit NEC homologs from other herpesviruses.

Materials and methods
Peptides. All peptides used in this study were purchased from Peptide 2.0 with the exception of the α-helical (EAEKAAK) 7 EAEK peptide, which was cloned, expressed and purified as described below.
Cloning. All primers used in cloning are listed in Supplementary Table S2. The DNA sequence encoding the α-helical [(EAEKAAK) 7 EAEK] peptide was subcloned into the prokaryotic expression vector pGEX-6P-1 that encodes a N-terminal GST-tag followed by a PreScission Protease cleavage site in frame with the BamHI restriction site within the multiple cloning site. The (EAEKAAK) 7 EAEK peptide DNA sequence was obtained as a gBlock gene fragment (IDT) and subjected to restriction digest cloning with BamHI and XhoI into the pGEX-6P-1 vector (the gene fragment sequence is listed in Supplementary Table S2) to create the pED39 [(EAEKAAK) 7 EAEK peptide] plasmid.
Expression and purification of UL25∆44 Q72A. The plasmid encoding HSV-1 UL25∆44 Q72A (pED03) was expressed and cells harvested as described above for NEC220 except that only 100 µg/mL kanamycin was used. UL25∆44 Q72A was purified as previously described 20 . Expression and purification of (EAEKAAK) 7 EAEK. The plasmid encoding the peptide sequence was expressed and cells harvested as described above for the NEC220 with the exception that only 100 µg/mL ampicillin was used. Cells were resuspended in lysis buffer (50 mM Na HEPES pH 7.5, 500 mM NaCl, 1 mM TCEP, and 10% glycerol) plus Complete protease inhibitor (Roche) and lysed using a microfluidizer 110S (Microfluidics). The cell lysate was centrifuged at 13,000×g for 35 min, and the supernatant passed over a Glutathione sepharose 4B (GE Healthcare) column, which was subsequently washed with lysis buffer. The GST-tag was cleaved on the glutathione sepharose column for 16 h using the PreScission Protease described above. The peptide was eluted off the column with lysis buffer. As a final purification step, the peptide was purified with size-exclusion chromatography using a Superdex 75 column (GE Healthcare) equilibrated with gel filtration buffer (20 mM Na Co-sedimentation assays. Co-sedimentation of either peptides alone or NEC and peptides (1:10 molar ratio of NEC:peptide) to multilamellar vesicles (MLVs) was performed as previously described 14 . MLVs were prepared in a 3:1:1 molar ratio of 1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC):1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS):1-palmitoyl-2-oleoyl-sn-glycero-3-phosphate (POPA) (Avanti Polar Lipids). Background signal in the absence of liposomes is due to protein aggregation during centrifugation. Each experiment was done in three biological replicates with two technical replicates. For peptides alone, the reported values represent the average binding of each biological replicate of peptide to MLVs relative to the NEC (100%). For experiments measuring NEC/MLV binding in the presence of peptides, reported values represent the percentage (0-100%) of co-sedimentation of either NEC220 or peptides in the absence or presence of MLVs. The standard error of the mean is reported for each measurement.
GUV budding assays. Giant unilamellar vesicles (GUVs) were prepared as previously described 14 . For The total volume of each sample during imaging for all experiments was brought to 100 μL with the gel filtration buffer and the reaction was incubated for 5 min at 20 °C. Samples were imaged in a 96-well chambered cover-glass. Images were acquired using a Nikon A1R Confocal Microscope with a 60 × oil immersion lens at the Tufts Imaging Facility in the Center for Neuroscience Research at Tufts University School of Medicine. Quantification was performed by counting vesicles in 15 different frames of the sample (~ 300 vesicles total). Each experiment was done in at least three biological replicates, each with three technical replicates. Prior to analysis, the background was subtracted from the raw values. All raw values are given in Supplementary Table S3. The reported values represent the average budding activity relative to NEC220 (100%). The standard error of the mean is reported for each measurement. Significance compared to NEC220 was calculated using an unpaired one-tailed t-test against NEC220. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.