Zebrafish 3-O-Sulfotransferase-4 Generated Heparan Sulfate Mediates HSV-1 Entry and Spread

Rare modification of heparan sulfate (HS) by glucosaminyl 3-O sulfotransferase (3-OST) isforms generates an entry receptor for herpes simplex virus type-1 (HSV-1). In the zebrafish (ZF) model multiple 3-OST isoforms are differentially expressed. One such isoform is 3-OST-4 which is widely expressed in the central nervous system of ZF. In this report we characterize the role of ZF encoded 3-OST-4 isoform for HSV-1 entry. Expression of ZF 3-OST-4 into resistant Chinese hamster ovary (CHO-K1) cells promoted susceptibility to HSV-1 infection. This entry was 3-O sulfated HS (3-OS HS) dependent as pre-treatment of ZF 3-OST-4 cells with enzyme HS lyases (heparinase II/III) significantly reduced HSV-1 entry. Interestingly, co-expression of ZF 3-OST-4 along with ZF 3-OST-2 which is also expressed in brain rendered cells more susceptible to HSV-1 than 3-OST-4 alone. The role of ZF-3-OST-4 in the spread of HSV-1 was also evaluated as CHO-K1 cells that expressed HSV-1 glycoproteins fused with ZF 3-OST-4 expressing effector CHO-K1 cells. Finally, adding further evidence ZF 3-OST-4 mediated HSV-1 entry was inhibited by anti-3O HS G2 peptide. Taken together our results demonstrate a role for ZF 3-OST-4 in HSV-1 pathogenesis and support the use of ZF as a model to study it.


Introduction
Heterogeneous chains of heparan sulfate (HS) are expressed on cell surface as a complex sulfated polysaccharide consisting of repeating disaccharide units of N-acetylglucosamine [GlcN] and glucuronic/iduoronic acid [GlcA/IdoA] [1][2][3]. The disaccharide chains in the HS are covalently linked to a serine residue of specific core protein via tetrasacchride link [GlcA-Gal-Gal-Xyl] making them a hybrid molecule with both protein and sugar component [3]. The chain of HS gets modified by series of multiple enzymes which includes glycosyltransferase, epimerase and sulfotransferases [4]. The first step of modification involves N-deactylation/Nsulfation of the glucosamine unit, followed by C5 epimerization of the glucuronic acid to iduronic acid, and O-sulfation of both residues. The last O-sulfation step is first catalyzed by 2-O sulfotransferase (2-OST), followed by 6-O sulfotransferase (6-OST) and finally 3-O sulfotransferases (3-OSTs) enzyme, which are expressed in multiple isoforms [1][2][3][4]. It is well documented that the modification of HS is responsible for high sequence diversity or heterogeneity which in turn dictates functional specificity and versatility [3,5,6]. Therefore, specific 3-OST isoforms potentially generate unique protein-binding sites within the HS chain which allows HSV-1 entry [5][6][7][8][9][10].
Interestingly, zebrafish (ZF) embryos are known to express multiple isoforms of heparan sulfate modifying enzyme 3-Osulfotransferase (3-OST) [11][12][13]. Cadwallader and Yost (2006b) characterized eight 3-OST family members in ZF via in situ hybridization from early cleavage stage through 48 hr post fertilization with seven genes showing homology to known 3-OST genes in mouse and humans [12]. The exclusive expression of 3-OST-4 in central nervous system in ZF provided us a rationale to examine the role of 3-OST-4 isoforms for a predominantly neurotropic virus HSV-1 [14]. Previously, we reported that enzymatic modifications in HS via 3-O sulfotransferases  isoforms other than ZF encoded 3-OST-4 mediate HSV-1 entry and spread [15][16][17][18]. In this study our goal was to characterize ZF encoded 3-OST-4 isoform for HSV-1 entry and spread. Our results here demonstrate that expression of ZF encoded 3-OST-4 isoform in Chinese hamster ovary (CHO-K1) cells results in susceptibility to HSV-1 entry and spread. In addition we also demonstrate that our G2 peptide, which was isolated and characterized against 3-OST-3 generated HS [19,20], blocks HSV-1 entry into cells expressing ZF-3-OST-4. The functional analogy between human and ZF 3-OST-4 further validates the potential of ZF embryos for studying HSV infection.

Co-expression of ZF 3-OST-2 and 3-OST-4 Enhances HSV-1 Entry
We next evaluated the effect of co-expression of ZF encoded 3-OST-2 and 4 isoforms in the same cells. We reasoned to test this because both 3-OST-2 and 3-OST-4 are widely expressed in central nervous system of ZF [12]. CHO-K1 cells were cotransfected with plasmids expressing ZF 3-OST-2 and 3-OST-4. In parallel, CHO-K1 cells were individually transfected with 3-OST-2 and 3-OST-4 isoforms separately, CHO-K1 cells expressing human 3-OST-2 and pcDNA3.1 was used as positive and negative control respectively. As indicated in Fig. 2C co-expression resulted in higher OD readings indicating the presence of both enzymes generate multiple sites for HSV-1 glycoprotein D (gD) and hence more viral entry.  Next, we evaluated the effect of enzymatic removal of 3-OS HS generated by ZF encoded 3-OST-4 on HSV-1 entry by treating cells with a mixture of heparinase-II and -III (1.5 U ml/L). These enzymes selectively degrade HS chains by cleaving them [21]. For this experiment, CHO-K1 cells expressing ZF 3-OST-4 along with CHO-K1 cells expressing human 3-OST-3 were mock-treated or pretreated with heparinase-II/III before infecting with reporter bgalactosidase expressing HSV-1 gL86 [10]. As a negative control, CHO-K1 cells expressing empty vector pcDNA3.1 was treated with heparinase. As indicated in Fig

HSV-1 Glycoprotein Expressing Effector Cell Mediates Fusion with ZF Encoded 3-OST-4 Expressing Target Cell
We next examined the role of ZF 3-OST-4 in HSV-1 spread by using HSV-1 glycoprotein mediated cell-to-cell fusion assay. We purposefully used wild type CHO-K1 cells because they lack endogenous glycoprotein D (gD) receptor required for HSV-1 entry. [10]. To quantify the HSV-1 glycoprotein induced cell fusion between 3-OS HS cells modified by 3-OST-4 and HSV-1 glycoproteins a luciferase reporter gene assay was performed [22,23]. Wild type CHO-K1 cells were transiently transfected with each of four glycoprotein plasmids: pPEP98 (gB), pPEP99 (gD) pPEP100 (gH), and pPEP101 (gL), as well as, the plasmid pT7EMCLuc that expresses a luciferase reporter gene was considered ''effector'' cell. In parallel ''target'' cells were transfected with a 3-OST plasmid expressing ZF encoded 3-OST-4 and the plasmid pCAGT7, which expresses T7 RNA polymerase to induce expression of the Luciferase gene. For a negative control, cells were transfected with T7 RNA polymerase and control plasmid pcDNA3.1. The cells expressing human 3-OST-4 and T7RNA polymerase served as a positive control. As shown in Fig. 4 a high amount of fusion occurred in ZF encoded 3-OST-4 expressing cells (red bars) compared to the negative control (yellow bar). Clearly, the 3-OS HS generated by ZF encoded 3-OST-4 is capable of mediating cell fusion as well. These results reinforce our findings that CHO-K1 cells expressing ZF 3-OST-4 allow cell fusion to occur, and thus potentially could facilitate spread of HSV-1 in a ZF model. We also evaluated if enzymatic removal of ZF 3-OST-4 cells affects HSV-1 glycoprotein mediated cell fusion. In this experiment, CHO-K1 cells expressing ZF 3-OST-4 were treated separately with heparinase I and heparinase II (1.5 U ml/ L) or mock treated with 1 6 PBS for 45 minutes before mixing cells with effector cells expressing HSV-1 glycoproteins; gB, gD, gH-gL. The two population, effector and target cells were mixed in equal 1:1 ratio and co-cultivated. As shown in Fig 4 cells ZF 3-OST-4 treated with heparinase-II/III showed significant reduction in fusion ( Fig. 4; blue bars) as well as reduced polykaryocytes formation (Fig. 4B, panel c), while mock treated cells forms multinucleated giant cells or polykaryocytes (Fig. 4B, panels a and b). Cartoon in Fig. 4C panel a, demonstrates that mock treated ZF 3-OST-4 cells were able to fuse with effector cells, while heparinase treatment to ZF3-OST-4 cells resulted no fusion (Fig. 4C, panel b).

Phage Display Library Screening Derived Anti-3-OS HS (G2) Peptide Blocks HSV-1 Entry in ZF 3-OST-4 Cells
We finally tested 12-mer G2 peptide that was isolated and characterized for binding to human 3-OST-3 generated heparan sulfate (HS) and blocking its role in viral entry [19,20]. The CHO-K1 cells expressing ZF 3-OST-4 were pre-treated with G2 peptide or a control peptide (Cp) before infecting them with HSV-1 reporter virus. CHO-K1 cells expressing human 3-OST-3 treated with G2 peptide was used as a control. As indicated in the Fig. 5, the pre-treatment of ZF-3-OST-4 expressing cells resulted in loss of HSV-1 entry.

Cell Fusion Assays
A cell-to-cell fusion assay described previously was used [23]. CHO-K1 cells were grown in 6-well plates to subconfluent levels. The so-called ''target'' cells were transfected with ZF encoded plasmids expressing 3-OST-4 and the luciferase gene. The ''effector'' or virus-like cells were co-transfected with plasmids expressing HSV-1 glycoproteins gB, gD, gH, and gL, and T7 RNA polymerase. In either case, the total amount of DNA used for transfection was kept constant. After 16 h, target and effector cells were mixed in a 1:1 ratio and then replated in 24-well dishes. The activation of the reporter luciferase gene as a measure of cell fusion was examined after 24 h. To demonstrate sensitivity to heparinase treatment target CHO-K1 cells expressing ZF 3-OST-4 were treated with a 1:1 mixture of heparinase-II/III for 2 h prior to mixing with the effector cells. Target cells were mock treated with the buffer alone to serve as a control.

Conclusions
ZF model expresses wide range of heparan sulfate modifying 3-O sulfotransferases (3-OST) enzymes including 3-OST-4 isoform during physiological development [11][12][13]24]. Interestingly both 3-OST-2 and 3-OST-4 isoforms are predominantly expressed in different regions of ZF brain and there is a possibility that HSV-1 being neurotropic virus may exploit one or both these isoforms during neuronal invasion in ZF model. Our previous study has already indicated the role ZF 3-OST-2 isoform for HSV-1 entry [18]. In this study we characterized 3-OST-4 isoform for HSV-1 entry and spread.
At the beginning of our study ZF encoding 3-OST-4 region was successfully cloned into empty vector pcDNA3.1. The resultant construct allowed us to successfully express 3-OST-4 in resistant CHO-K1 cells (Fig. 1D), which became susceptible to HSV-1 entry upon ZF 3-OST-4 expression (Fig. 2). In addition, CHO-K1 cells expressing 3-OST-4 allowed cell-to-cell fusion as an indicator of HSV-1 spread. Both the events of HSV-1 entry and spread were HS dependent as evident from enzymatic treatment of cells which resulted in significant decrease in HSV-1 infection (Fig. 4). Further, we provided evidence that by blocking ZF modified HS by using phage display derived anti-3-OS HS (G2) peptide HSV-1 entry was significantly reduced. Interestingly co-expression of both 3-OST-2 and 3-OST-4 resulted in higher viral entry. ZF 3-OST-2 and 3-OST-4 are highly expressed in forebrain, hindbrain, and olfactory epithelium of central nervous system [12]. Taken gB, gD, gH, and gL while heparinase treatment significantly blocks ZF 3-OST-4 mediated fusion. The target CHO-K1 cells were transfected with plasmids expressing ZF 3-OST-4 and luciferase reporter gene. The effector CHO-K1 cells were transfected with HSV-1 glycoproteins gB, gD, gH, and gL, and T7 RNA polymerase. CHO-K1 effector cells expressing control plasmid without HSV-1 glycoproteins were used as a negative control. In addition, target CHO-K1 cells expressing ZF 3-OST-4 were treated with heparinase II/III (1.5 U/ml) (blur bar) or left untreated (red bar) for 1 hr prior to co-cultivation with effector CHO-K1 cells expressing four HSV-1 essential glycoproteins (gB, gD, gH-gL; 0.5 mg DNA each glycoprotein). A luciferase reporter assay was performed 24 h after the two cell populations were mixed together. Cell fusion was measured in relative luciferase units ( together our findings strongly suggest that ZF 3-OST-4 mediates HSV-1 entry and cell-fusion similar to human 3-OST-3 and 3-OST-4 isoforms. It is also very interesting that HSV-1 entry was enhanced with the co-expression of 3-OST-2 and -4 isoforms, which means their co-expression naturally in ZF brain can result in higher infection. We also provide additional new information that virus entry via ZF 3-OST-4 isoform was inhibited by anti-3-OS HS peptide (Fig. 5) as suggested in proposed model (Fig. 6). Our results not only extend the list of 3-OSTs receptors for HSV-1 entry into ZF model [24] but also provide new information related  to the mechanism needed to infect brain tissues of ZF. While ZF 3-OST-2 is widely expressed in CNS, ZF-3-OST-4 is expressed in various regions of CNS and also in the eye tissues [12]. Similar cell and tissue specific 3-OST-2 and 3-OST-4 expression profiles for human and mouse model has been suggested [25,26]. Therefore, ZF embryo model to study HSV-1 infection could be very useful for various reasons. For instance, fine alterations of GAG modification is a dynamic event during zebrafish development which is regulated as the ZF embryos age [27]. ZF embryos to adult form might show variability in susceptibility to HSV-1 infection, which in turn could shed new light on how the modifications within HS play a role in viral infectivity. Similarly, HSV-1 tropism in ZF embryo may very well be guided by 3-OSTs expression especially in the brain or in the eye tissues. In this regard our phage display generated anti-3-OS HS (G2) peptide [19] will be useful to study its ability to affect HSV-1 entry and spread in live ZF embryos. Alternatively, a micro-injection based strategy using anti-3-OS HS peptides fused with cargo-nanoparticles will be useful to enhance efficacy of anti-HSV-1 activity in targeted specific tissue rich in HS sulfation. Overall, our data confirms the role of ZF 3-OST-4 in HSV-1 infection and supports the use of ZF model to study HSV-1 infection. Figure S1 The transfection efficiency of both zebrafish encoded 3-OST isoforms (3-OST-2 and 3-OST-4) was verified via co-transfection with GFP (pGFP-N1) expressing plasmid (panel a: bright field; panel b: GFP expression and panel c: overlay). Zeiss Axiovert 100 inverted microscope was used for imaging. (JPG)