Identification of a retroviral receptor used by an Envelope protein derived by peptide library screening

  1. Anindita Sarangi,
  2. Keith Bupp, and
  3. Monica J. Roth*
  1. Department of Biochemistry, University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School, 675 Hoes Lane, Piscataway, NJ 08854

  1. Communicated by Stephen P. Goff, Columbia University College of Physicians and Surgeons, New York, NY, May 7, 2007 (received for review January 9, 2007)

 
Next Section

Abstract

This study demonstrates the power of a genetic selection to identify a variant virus that uses a new retroviral receptor protein. We screened a random peptide library within the receptor-binding domain of a feline leukemia virus retroviral Envelope (FeLV Env) protein for productive infection of feline AH927 cells. One variant, A5, obtained with altered tropic properties acquired the ability to use the solute carrier protein family 35 member F2 (SLC35F2) as a receptor. The SLC35F2 protein is a presumed transporter of unknown function predicted to encode 8 to 10 transmembrane-spanning regions and is not homologous to any identified retroviral receptor. Expression of the feline SLC35F2 cDNA in nonpermissive cells renders the cells susceptible to infection by A5 virus, with remarkably high titers in the range of 105 infectious units per ml. The human SLC35F2 ORF also functioned as the retroviral receptor, albeit at lower efficiency than the feline homologue. The successful selection of a novel molecule, the SLC35F2 transporter/channel-type protein, as a receptor by the FeLV Env backbone suggests that multipass transmembrane proteins may be particularly suited for use in productive viral entry and fusion. The analysis of retroviral Env libraries randomized in the receptor-binding domain offers a viable means to develop viral vectors targeted to specific cell types in the absence of known targeting ligands.

Tissue specificity is a key factor in defining the safety and efficacy of retroviral vectors. For viral vectors, specificity can be achieved via alteration of the viral–host interactions by targeting entry through alternative host receptor proteins. The results presented here indicate that successful retargeting of retroviral entry can be achieved through an alternative host-cell receptor.

Critical to the question of cell targeting is the range of potential cell-surface proteins that can serve as the reservoir of retroviral receptors. A large number of host-cell receptors have been identified for viruses spanning the orthoretrovirinae subfamily. A correlation between the type of membrane protein serving as the receptor and the retroviral genus can be observed. All known γ-retroviral receptors are multitransmembrane channels or transporters, including the MuLV ecotropic M-MuLV [MCAT (1)] and M813 [SMIT1 (2)]; the amphotropic [PiT-1/2 (3–5)], xenotropic, and polytropic receptors [Xpr1 (6–8)]; as well as the FeLV A [THTR1 (9)], B [PiT1 (10)], and C [FLVCR (11–13)] viral receptors. With these evolutionary observations, it is of great interest as to whether a given retrovirus must remain restricted to use receptors within the broad class of the original receptor or can be altered to use receptors beyond these protein family boundaries.

A variety of approaches have been developed for altering the receptor usage of viral delivery systems. Receptor specificity and binding of the viral Envelope (Env) is mapped to the N-terminal half of the surface (SU) protein through the close interaction of two highly variable regions, VR1 (VRA) and VR2 (VRB) (14–18). Exchange of a defined VR1 segment between the FeLV A and C subgroups altered the viral host range (19). In our studies, codons for 10/11 amino acids within this VR1 FeLV Env receptor-binding region were randomized within a FeLV A/C Env chimera backbone (20). Viral particles bearing this library of >1 million viral VR1 Env variants are selected for productive infection (20–22). This library displays the targeting peptide within the context of the authentic Env backbone, eliminating the problem that preidentified phage display peptides inserted into alternative contexts do not maintain their original properties (23). Alternative approaches have involved the addition of targeting domains (24, 25), peptides (23, 26), or single-chain antibodies (27) within the framework of viral Env or capsids (28–30) or through heterologous intermediates (31). Successful retargeting for Sindbis Env pseudotyped onto lentiviral particles has been achieved with the insertion of the Protein A ZZ domain and sandwiching targeting antibodies to the modified particles (32) or through coexpression of Env proteins and surface-binding molecules (33).

Screening the FeLV random Env library on feline AH927 cells identified the A5 Env isolate, which also efficiently infected human 293T embryonic kidney cells (21). Superinfection interference studies indicated that the receptor for A5 was outside the receptor family used by FeLV A, B, and C and MuLV 4070A viruses (21). The A5 Env therefore provided an excellent system to examine the range of proteins capable of serving as receptors for retroviral Env libraries. In this study, the receptor for the A5 Env isolate was identified as the SLC35F2 protein, a transporter protein of unknown function, through screening a feline AH927 cDNA library.

Previous SectionNext Section

Results

Directed evolution allows for the selection of proteins with new properties. Using this approach, randomization of the FeLV Env receptor-binding domain, followed by selection for productive infection, identified Env isolates with receptor usage outside the interference groups of the known FeLV A, B, and C viruses (21, 22). This result implies that the selection of a receptor protein can occur through screening a single round of viral infection. To definitively prove that alternative viral receptors can be selected, the receptor protein for the A5 Env isolate was identified through cDNA screening.

Use of Human A498 Cells as a Nonpermissive Cell.

A series of nine cell lines were tested for A5/lacZ infection (data not shown and ref. 21). Human A498 kidney cells were selected as recipient cells for the cDNA library because they consistently gave titers of <1 lacZ staining units (lsu) per ml under conditions that yielded titers of 105 on the permissive feline AH927 cells (Table 1). The virus particles consisted of MuLV cores pseudotyped with the FeLV Env derivative. To eliminate the possibility that the low titer observed on A498 cells was because of a post-entry block, rather than a lack of receptor expression, A498 cells were challenged with MuLV particles bearing the amphotropic 4070A Env. Infection of A498 cells with 4070A/lacZ yielded titers of 4.0 × 104 lsu per ml, indicating that there was no post-entry block to infection with identical MuLV viral core particles (Table 1).

View this table:
Table 1.

Characterization of nonpermissive A498 cells


Feline AH927 cDNA Screening.

The identification of the A5 receptor was facilitated by a preexisting cDNA library from feline AH927 cells within the pMX1 retroviral vector (9). The general scheme of the cDNA screening is outlined in Fig. 1. The cDNA library was packaged into MuLV retroviral particles pseudotyped with VSV-G and used to infect the nonpermissive A498 cells. Initial viral challenge with particles bearing A5 Env and packaging GFP-IRES-puro vector (1° screen) yielded 52 puroR colonies, all of which expressed GFP. Subsequent challenge of the puroR colonies with virus-bearing A5 Env and packaging lacZ confirmed infection of 12/20 clones (1° isolates). Positive colonies remained neo S, indicating superinfection interference from the pRVL vector (20), expressing the A5 Env and neo, would not occur.

Fig. 1.
View larger version:
Fig. 1.

Schematic of cDNA library screen. The individual steps involved in screening the retroviral cDNA expression library, including the A5 viral challenges, are outlined.


Initial PCR analysis using Taq polymerase with pMX1 vector primers and sequence analysis of 12 puroR colonies indicated that each cell contained multiple cDNA inserts in the size range of 0.7 to 2.3 kB (data not shown). No single PCR product was common to all puroR cell lines. To further isolate the functional cDNA, the retroviral vectors were mobilized from the puroR clone 42 by transient expression of the MuLV gag-pol genes and VSV-G Env protein. Clone 42 was selected because of its high level of infection by A5/lacZ virus (1.1 × 104 lsu per ml). The transiently expressed virus collected from clone 42 was used for a second round of infection of nonpermissive A498 cells.

Cells were then subjected to a challenge with virus bearing A5 Env and packaging the hygromycin gene (2° screen). Eighteen hygro R colonies were isolated; 14 were subjected to additional challenge by A5/lacZ, and six were expanded yielding titers in the range of 105 A5 lsu per ml (2° isolates). PCR analyses of the inserts using KOD Hot start polymerase and pMX1 vector primers yielded an identical 2.1-kB PCR product containing the cDNA insert in two clones, 42–8 and 42–17.

The Putative A5 Receptor cDNA Encodes the Feline Solute Carrier Protein Family 35 Member F2 (SLC35F2) Protein.

Sequence analysis of the cDNA insert was performed on DNA from both the 42–8 and 42–17 isolates, confirming both contained identical sequences. The library was generated by ligation of BstXI linkers to the cDNA with one nucleotide (C) present between the BstXI linkers and the ATG start of the ORF. The cDNA is 1,935 nucleotides in length, containing an ORF encoding a 374-amino acid protein, a large 791-nucleotide 3′ untranslated region (UTR), and a 21-nucleotide polyA tract at the 3′ terminus. Nucleotide–nucleotide BLAST analysis indicated high homology to the SLC35F2 protein and predicted to encode a transporter of unknown function (Fig. 2). The feline SLC35F2 protein shares 90.6% sequence identity (94.6% homology) with the human SLC35F2 protein (BC039195). The greatest divergence among these two occurs at the N terminus, where 15/35-amino acid differences localize to the first 43 amino acids (Fig. 2). In contrast, the 3′UTRs of the feline and human cDNAs are highly divergent, differing in size (791 vs. 1,622 bases, respectively) and sequence. Sequence homology is limited to a short 74 base region [nucleotides 1,399–1,472 of hSLC35F2 (BC039195.2)] (data not shown). Using the MEMSTAT3 program (34), 10 putative membrane-spanning regions were identified with the N terminus predicted to be cytoplasmic for both the human and feline SLC35F2 proteins. In contrast, TMpred† strongly preferred a model with an extracellular N terminus containing only eight transmembrane regions, eliminating the helices between amino acids 112–121 and 295–304 (Fig. 2).

Fig. 2.
View larger version:
Fig. 2.

Comparison of feline and human SLC35F2 homologues. (Upper) Schematic comparison of the feline and human SLC35F2 cDNAs, including the ORF and 3′ untranslated region (3′UTR). (Lower) Protein sequence alignment of the predicted feline and human SLC35F2 proteins. The human protein ID is AAH39195.1 (sequence accession no. BC039195). Amino acid identities are marked by dots. Areas shaded in gray correspond to central transmembrane helix segments predicted by using the MEMSTAT3 program with an N-terminal inside loop (34).


Expression of the 1935 Nucleotide Feline SLC35F2 cDNA in A498 Cells Confers Susceptibility to Infection of Virus-Bearing A5 Env.

Experiments were carried out to confirm that the feline SLC35F2 cDNA functions as the A5 Env receptor. Initial experiments examined whether the feline SLC35F2 cDNA was represented in multiple primary and secondary isolates because it was isolated from the two cell lines 42–8 and 42–17. Using primers specific to the feline SLC35F2, the genomic DNA from nine independent primary isolates and five of the clone 42 secondary isolates were PCR-amplified (Fig. 3). A product corresponding to the predicted 538-bp species was observed in all 14 isolates, which was not detected in the nonpermissive parental A498 cells or control reactions lacking input genomic DNA. This result confirmed that the cDNA was present in independent primary isolates.

Fig. 3.
View larger version:
Fig. 3.

PCR analysis of the primary and secondary cDNA isolates. Primers within the feline SLC35F2 ORF (42fs and 42–17ext) were used to detect the feline SLC35F2 cDNA within primary and secondary isolates. The arrow marks the position of the predicted 538-bp product. Control cells (nonpermissive A498) lacking cDNA library and PCR performed in the absence of genomic template DNA are included. M, marker DNA. The predicted molecular weights are indicated at the right.


Direct confirmation that the feline SLC35F2 cDNA encoded the A5 viral receptor was obtained by reintroduction of the cloned cDNA into nonpermissive cell lines. Two cell lines were examined, the parental A498 cells and human prostate DU145 cells, which have been shown to have low viral titer when challenged with A5/lacZ. The feline SLC35F2 cDNA was cloned from the PCR product back into the pMX1 vector and introduced into the nonpermissive cells by infection of VSV-G pseudotyped retroviral particles packaging the feline SLC35F2 cDNA. Cell lines in the presence and absence of the feline SLC35F2 gene were challenged with A5 virus packaging lacZ. Infection of both the A498 and DU145 cells was at background levels in the absence of the feline SLC35F2 gene (Fig. 4 A and C). Introduction of the feline SLC35F2 cDNA rendered both cells susceptible to infection, with titers of A5/lacZ virus on A498/feline SLC35F2 cells reaching 3.7 × 105 lsu per ml and on DU145/SLC35F2 1.8 × 105 lsu per ml. These results demonstrate that the feline SLC35F2 cDNA encodes the receptor for the A5 virus. Thus, Env isolates identified by library screening can select alternative host-cell proteins to serve as receptors for gene delivery.

Fig. 4.
View larger version:
Fig. 4.

Introduction of feline SLC35F2 cDNA renders cells permissive to A5 infection. Feline SLC35F2 cDNA was introduced into nonpermissive cells by retroviral infection of VSV-G-pseudotyped particles packaging the pMX1/42–17 cDNA. All cells are challenged with A5/lacZ virus. (A) Nonpermissive A498 cells. (B) A498 cells expressing feline SLC35F2 cDNA. (C) DU145 cells. (D) DU145 expressing feline SLC35F2 cDNA. Titers of A5/lacZ virus (lsu per ml of virus) are indicated in each box.


Human SLC35F2 Can Function as an A5 Receptor.

The A498 cells, the nonpermissive host used in the cDNA library screening, are a human cell line. The question was asked whether the human SLC35F2 analogue could function as a receptor for A5 infection. The hSLC35F2 ORF lacking the 3′UTR was obtained as a synthetic cDNA clone. To compare the receptor function of the human and feline SLC35F2 proteins, the human SLC35F2 ORF was substituted into the pMX1 vector while maintaining the feline 3′UTR sequence. The human SLC35F2 vector was introduced into A498 cells by using VSV-G pseudotyped retroviral particles. Positive cells expressing human SLC35F2 were selected by challenge with A5/pGIP and puromycin selection. A498/pMX1/human SLC35F2 cells were then challenged with A5/lacZ (Table 2). Titers of A5/lacZ on cells enriched for functional human SLC35F2 were 3.4 × 103, ≈4-fold lower than cells expressing the feline SLC35F2 protein. Maximal titer remained on feline AH927 cells, expressing the endogenous feline SLC35F2 gene (8.5 × 105). These results indicate that the human SLC35F2 ORF can function as a receptor and cannot account for the lack of infection of A498 cells. The functionality of the human protein is in agreement with the initial infection of A5 virus on the human 293T cells (21). However, the titer of A5 virus on cells expressing human SLC35F2 optimized for A5 infection is lower than that observed for cells expressing feline SLC35F2.

View this table:
Table 2.

Expression of feline and human SLC35F2 in pMX1 vector backbone


Viral-Binding Studies and Expression of SLC35F2 on the Cell Surface.

To date, the known members of the SLC35 subfamilies A–E localize to the lumen of the endoplasmic reticulum and the Golgi. However, for the SLC35F2 protein to function as a viral receptor, initial binding is presumed to occur on the cell surface. To examine this, viral-binding studies were performed. Virus was bound to permissive and nonpermissive cells at 4°C to limit endocytosis and subsequently incubated with a monoclonal Ab (C11D8) recognizing a conserved SU epitope (36). The antibody–virus–receptor complex was visualized by using FITC-conjugated secondary antibodies and FACS analysis (Fig. 5). Several critical observations were obtained in these studies. Binding of A5 virus to control nonpermissive A498 cells was not detected (Fig. 5 A, green), confirming the low level of surface expression of the endogenous human SLC35F2. Introduction of the pMX1/feline SLC35F2 construct into A498 cells resulted in a clear signal for virus binding (Fig. 5 A, red). This shift was not observed in experiments lacking the primary antibody (data not shown). Binding of A5 to feline AH927 cells expressing endogenous feline SLC35F2 yielded a broad peak with a mean fluorescent intensity (1.24) lower than that observed for feline SLC35F2 expressed on A498 cells (1.56) (Fig. 5 B). Surprisingly, binding of A5 virus on cells expressing the human SLC35F2 from the pMX1 vector showed extremely strong association of viral particles on the cell surface (Fig. 5 A, black). Virus binding showed an inverse relationship with viral titer. The titers of the virus presented in Table 2 correspond with the samples used in the binding studies. The highest viral titer (8.5 × 105, Table 2) was observed on the feline AH927 cells, which displayed the lowest fluorescence intensity. Expression of the human SLC35F2 in A498 cells reproducibly displayed extremely high A5 binding fluorescence intensity (100), but maintained titers of 3.4 × 103 (Table 2). Expression of the feline SLC35F2 in A498 cells showed titers and binding intermediate of these two (1.4 × 104, Table 2; mean fluorescence of 1.56). In all cells studied (A498 and AH927), the feline and human SLC35F2 proteins were on the cell surface accessible for virus binding.

Fig. 5.
View larger version:
Fig. 5.

Expression of SLC35F2 on the cell surface. A5 virus binding was detected by flow-cytometric analysis using the C11D8 mAb recognizing the FeLV Env backbone (36). Data are representative of six independent experiments. (A) A5 virus binding to human A498 cells. Green, nonpermissive A498 cells; red, A498 cells expressing the pMX1/felineSLC35F2 construct; black, A498 cells expressing the pMX1/human SLC35F2 construct. (B) Binding to permissive feline AH927 cells. Blue, control assay performed in the absence of C11D8 antibody; green, control binding performed in the absence of virus; red, complete binding reaction performed with A5 virus plus C11D8 antibody.


Previous SectionNext Section

Discussion

The results of these experiments establish that retroviral infection can be retargeted to productively enter cells through selection of an alternative host-cell protein receptor by using a library screening process. Host–receptor interactions can be altered through randomization of a short 11-amino acid sequence within the FeLV Env backbone. Several independent screens of these libraries have identified, in at least two cases, use of host-cell receptors outside the known receptor interference pattern of the FeLV Envs (21, 22). The studies presented in this manuscript identify the receptor for one of these Env isolates (A5) as the SLC35F2 protein. These results highlight the potential of the system, indicating the number of host-cell proteins that can serve as receptors for retroviral gene delivery can be expanded.

To date, all known receptors for the murine and feline leukemia viruses have used multitransmembrane receptors. The maintenance of selected alternative receptors within this class of membrane proteins implies a functional advantage for productive viral entry. Viral entry is a carefully orchestrated series of events of which binding is only the first step. Productive infection requires coordinated conformational changes yielding release of the SU protein, exposure of the fusion peptide, and subsequent fusion of membranes ultimately yielding content transfer of the virus into the cell. These events are spatially and temporarily regulated. The use of multitransmembrane receptor proteins may facilitate the transition between receptor binding and generation of the fusion pore through providing the optimal spatial proximity needed between the host and viral proteins and membranes. Thus, attempts to retarget γ-retroviral infection to single-pass transmembrane cell-surface proteins, rather than multipass transmembrane proteins might not be productive. It would be of interest to correlate the site of fusion (i.e., acidic endosomes) with the class of receptors used by different viruses.

The function of the SLC35F2 protein remains unknown. The SLC35 family is a member of a larger drug/metabolite transporter superfamily (DMT; transporter classification 2.A.7) (37). Members of the SLC35 family encode nucleotide sugar transporters and have been divided into six subfamilies (A–F). For each of the SLC35 subfamilies A–E, the function of at least one member has been defined. The SLC35A1 transports CMP-Sia, SLC35A2 transports UDP-Gal (38, 39), SLC35A3 transports UDP-GlcNAc (38, 39), SLC35C1 encodes a GDP-Fuc transporter (39), and SLC35D1 transports UDP-GlcA/UDP-GalNac (40). SLC35B3 encodes a 3′ phosphoadenosine 5′-phosphosulfate (PAPS) transporter, involved in formation of sulfated proteoglycans and glycoproteins (41). There are five members of the SLC35 subfamily F (F1-F5); however, no function has been identified for any of the members. Homology searches have reported that feline SLC35F2 is related to BAD70612 (or GI:16416383), a putative anthocyanin-related membrane protein 1. However, there is no phenotypic or functional linking of this protein to anthocyanin transport.

Of interest is the subcellular localization of the known SLC35 family members that predominate in the lumen of the endoplasmic reticulum and the Golgi (38, 41). For the SLC35F2 protein to function as a viral receptor, the primary contact point is assumed to be on the cell surface. The viral-binding assay confirms the presence of the SLC35F2 protein on the cell surface. The signal observed in the binding assay can result from either high-level expression of the receptor on the cell surface, high affinity of the Env with the receptor, or a preferred Env conformation, which stabilizes the binding of the monoclonal antibody in the assay (42). Interestingly, the apparent virus titer and efficiency of binding did not directly correlate. Tight receptor binding by human SLC35F2 may inhibit SU conformational changes required for Env activation and could lead to higher receptor levels detected via Env binding while resulting in lower infection efficiencies. The host proteins interacting with the receptor as well as the pathway of receptor recycling can also affect the viral titer.

Expression profiles of SLC35F2 have been deposited within the National Center for Biotechnology Information/National Library of Medicine/National Institutes of Health Gene Expression Omnibus databank, including two large-scale mouse studies (GDS1322 and GDS592) and one of the human transcriptome (GDS596). High expression was found in the small intestine and epididymis (GDS1322), with closest cluster analysis with Spfh2 (579608), a membrane protease subunit. Expression has also been high in the mouse-fertilized egg and blastocyst (GDS592) and transformed cell lines, including those of Daudi and Raji Burkitt's lymphoma. Although this differential expression pattern does not shed light on the protein's function, it may impart specificity for gene delivery systems.

The Env library screen used in this study allows for an unbiased selection of retroviral entry on uncharacterized cells. The repertoire of cell-surface proteins expressed on the target cell influences the selection process, and thus the selection of the target cell is of critical importance. Receptor expression, both in tropism and abundance, defines a critical component in the specificity of the retargeted viral particle (43). Although receptors might not be uniquely expressed on one tissue, their abundance can be regulated. Proteins expressed at high levels on the cell surface may have a higher potential to be selected by the random Env library. Biologically relevant potential target receptor proteins may be abundantly expressed on specific tissues. For example, the BCRP multidrug resistance protein is expressed on hematopoeitic stem cell populations (44) as well as transformed cells (45). Multidrug resistance proteins are multipass transmembrane proteins that have the potential to serve as possible viral receptors. The experiments presented in this manuscript raise the possibility that these biologically relevant proteins could be targeted as viral receptors to allow for gene and drug delivery to cells through overexpression on the host-cell type.

Previous SectionNext Section

Materials and Methods

Cell Lines and Plasmids.

The introduction of plasmid CeB (46) expressing the gag-pol gene into human TE671 cells producing TECeB [F2 isolate; (47)] and TELCeB6 cells expressing the Ψ+ lacZ gene plus CeB (46) were previously described. The maintenance of the human DU145, human A498, human TE671, TECeB, TELCeB6 cells, human 293T, 293TCeB, and feline AH927 were as previously described (21, 22). TELCeB6/4070A cells were generated by infection of TELCeB6 cells with NCAC-Am (18). All viral infections were performed in the presence of 8 μg/ml polybrene.

TECeBF2/A5/GIP and TECeBF2/A5/Hygro cells were generated in two steps. Ten micrograms of pRVL/A5/neo (21) plus 10 μg of pHIT-G (48) were introduced into 293T/CeB cells by using a calcium phosphate method (Mammalian Transfection kit; Stratagene, La Jolla, CA), treated with 10 mM sodium butyrate (20), and released virus used to infect TECeBF2 cells. G418R cells were maintained as a population. The plasmids pGIP [GFP-IRES-puro (49)] or pBabeHygro were introduced into the cells by infection of the G418R TECeBF2/A5 population. 293TCeB cells (10-cm plate) were cotransfected with either 10 μg of pGIP or pBabeHygro plus 10 μg pHIT-G (48) using calcium phosphate. Virus released was used to infect the TECeBF2/A5 population and selected for puroR or hygroR.

Primary Screening Feline AH927 cDNA Library for A5 Infection.

The feline AH927 cDNA library expressed within the retroviral pMX1 backbone was a gift of Julie Overbaugh (Fred Hutchinson Cancer Research Center, Seattle, WA) (9). One hundred nanograms of the AH927 cDNA library was introduced into Ultramax DH5α FT chemically competent cells (Gibco–BRL Life Technologies, Gaithersburg, MD) and plated onto six (245 × 245-mm) plates containing LB carbenicillin (100 mg/ml), yielding 1.22 × 107 colonies. Colonies (5 gm wet weight) were scraped with a rubber policeman, and plasmids were CsCl purified. To assemble virus particles, ten 10-cm plates of 293T/CeB cells were pretreated with chloroquine (12.5 μg/ml) (20), and 10 μg of the AH927 cDNA library plus 10 μg of pHIT-G (48) were introduced by using calcium phosphate (Stratagene) overnight. Cells were induced with sodium butyrate (20), and the media were filtered (0.45-μm filters) and used to infect ten 10-cm plates of nonpermissive A498 cells. Cells were split 1:1. Three days after infection, 20 A498 plates expressing the feline AH927 cDNA were challenged with A5/GIP virus (10 ml viral supernatant/10-cm plate A498 cells) and selected for growth in 2.5 μg/ml puromycin. Fifty-two puroR/GFP+ colonies were identified. Secondary challenge used virus expressed from the TELCeB6/A5 cell line containing the A5 Env and packaging the lacZ gene. Titration analyses by limiting dilutions on gridded plates were performed as previously described (20).

Mobilization of cDNA for a Second Round of Screening.

The cDNAs integrated within primary isolate #42 were mobilized into viral particles. Two 10-cm dishes of #42 (GFP+/lacZ−) were each treated with 10 μg of pHIT-G plus 10 μg of pCgp using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA) as suggested by the manufacturer. The media were changed after 6 h and treated with 10 mM sodium butyrate (20), and 20 ml of supernatant was concentrated by centrifugation at 2,700 × g using Vivaspin 20 concentrators (Vivascience; 300,000-MW cutoff) for 20 min at 4°C to 1 ml final volume for infection of fresh A498 cells in a 35-mM dish. The pMX1-cDNA inserts and the pGIP should be mobilized. The titer of GIP transferred was 1.51 × 102, indicating efficient mobilization of viral vectors bearing a packaging signal. Virus (77 ml) was collected from nine 10-cm plates of TECeBF2/A5/Hygro and concentrated by ultracentrifugation at 15,000 × g at 4°C overnight. The bottom 1 ml of supernatant was used to infect A498 cells in a 35-mm gridded dish and then selected under 50 μg/ml hygromycin; 15/18 HygroR colonies were challenged with virus isolated from TELCeB6/lacZ/A5. Fourteen colonies were positive for lacZ staining. Six isolates were further expanded (42–1, 42–8, 42–9, 42–13, 42–14, and 42–17). TELCeB6/lacZ/A5 challenge yielded titers in the range of 1.4 to 2.9 × 105 lsu per ml.

PCR and Sequence Analysis of cDNA Inserts.

Genomic DNA was extracted by using the DNeasy Tissue Kit (QIAGEN, Valencia, CA) for PCR amplification. PCR analysis of the cDNA insert within pMX1 was performed with primer MX11 (5′-GTGGACCATCCTCTAGACTGC-3′) (12) and MX14RBam (5′-CGGGATCCCTTTTATTTTATCGTCGACCACTGTGC-3′). Analysis of the primary screen was performed by using Taq polymerase. PCR analysis of the inserts from the 2° screen was performed with MX11/M14 primers with KOD Hot start polymerase (Novagen, Madison, WI) in the presence of 3% DMSO with 35 cycles of 94°C, 15 sec/60°C, 30 sec/68°C, 3 min and yielded a 2.1-kB PCR product from isolates 42–8 and 42–17. PCR analysis of cDNAs was also performed by using primers 42fs (5′CCTTCTTGTCCAGCTACAAGG) and 42–17ext (below) hybridizing to SLC35F2 by using KOD polymerase.

Sequence and Amino Acid Analysis.

Sequencing was performed by using the MX11 and MX14 vector primers plus internal primer 42–17ext (5′-CTTGGTCCTTCTTGGAGCTTC-3′) and 42–3prime (5′-TGAAAATAGAACAGAAGACAGGG-3′). The 42–17 cDNA insert sequence has been deposited in the GenBank (accession no. DQ860275).

Receptor cDNA Cloning.

Full-length feline SLC35F2 cDNA was subcloned back into the pMX1 backbone through digestion and ligation of the 42–17 2.1-kb PCR product and pMX1 vector with BstXI, generating pMX-1/cDNA42–17. This vector was packaged into retroviral particles through introduction of pMX-1/cDNA42–17 (2 μg) plus pHIT-G (1 μg) into 293T/CeB (35-mm plate; 2–3 × 105 cells) using the calcium phosphate method/10 mM sodium butyrate induction as described above. After 48 h, the viral supernatant was filtered (0.45-μm filters), and 2 ml was used to infect 4 to 6 × 105 DU145 and A498 cells in 35-mm plates. Four days after infection, cells were split 1:2 and challenged with 1 ml of A5/lacZ virus, concentrated 33-fold by ultracentrifugation at 15,000 × g in a SW27 rotor at 4°C. Two days after A5/lacZ challenge, the cells were fixed and stained for β-galactosidase activity.

Expression of the Human SLC35F2.

Human SLC35F2 synthetic cDNA clone (synthetic construct FLH119351.01L, National Center for Biotechnology Information no. AY893139) was obtained from DNA Resource Core/Dana–Farber/Harvard Cancer Center. The human SLC35F2 ORF was introduced into the pMX-feline SLC35F2 construct containing the 3′UTR using overlapping PCR. The four primers used were 1–5′Bam (5′-CCGGATCCCAGTGTGGTGGAAAGCATGGAGGC A G A C TCGCCAGCGGG-3′), 2–3′overlap (5′-GCACCCCTTCTTGTCCAGCTA CAACAAGACAGCAGAGTGGGTC-3′), 3–5′overlap (5′-GACCCACTCTGCTGTCTTGTTG TAGCTGGACA A G A A GGGGTGC-3′), and 4–3′ Bam (5′-CCACCGGATCCAGTCCAATGCTACG-3′), where the BamHI sites are underlined, the human SLC35F2 sequence is in bold, and the feline SLC35F2 3′UTR is in italics.

pMX1/human SLC35F2 was transiently introduced into 293TCeB cells in the presence of pHITG, and the resultant viral supernatants were used to infect A498 cells. A498 cells expressing the pMX1/human SLC35F2 were challenged with A5/pGIP virus and grown in 2.5 μg/ml puromycin. The population of puro R cells was further titrated with A5/lacZ virus.

FACS Sorting.

FACS sorting with C11D8 antibody (36), recognizing the FeLV SU (50), was performed as previously described (35) with the following modifications. One milliliter of viral supernatant was incubated at 4°C with 1 × 106 cells. Cells were washed with PBS plus 5% FBS, and the secondary antibody was FITC-conjugated anti-mouse antibody (1:100) (α-mouse IgG, whole molecule; Sigma–Aldrich, St. Louis, MO). Cells were analyzed by FACS sorting on a Coulter Cytomic FC500 Flow Cytometer Core Facility at University of Medicine and Dentistry of New Jersey/Environmental and Occupational Health and Safety Services.

Previous SectionNext Section

Acknowledgments

We thank Xuejun Ma for his input and John Kerrigan for assistance in the analysis of transmembrane domains. This work was supported by National Institutes of Health Grant RO1 CA49932 (to M.J.R.).

Previous SectionNext Section

Footnotes

  • *To whom correspondence should be addressed. E-mail: roth{at}umdnj.edu

  • Author contributions: K.B. and M.J.R. designed research; A.S. and K.B. performed research; A.S., K.B., and M.J.R. analyzed data; and M.J.R. wrote the paper.

  • The authors declare no conflict of interest.

  • Data deposition: The feline SLC35F2 cDNA sequence has been deposited in the GenBank database (accession no. DQ860275).

  • ↵ † Hofmann, K., Stoffel, W. (1993) Biol Chem Hoppe-Seyler 374:166 (abstr.).

  • Abbreviations:
    Env,
    Envelope;
    FeLV,
    feline leukemia virus;
    lsu,
    lacZ staining units;
    SLC35F2,
    solute carrier protein family 35 member F2;
    SU,
    surface.
Previous Section
 

References

  1. ↵
    1. Albritton LM,
    2. Tweng L,
    3. Scadden D,
    4. Cunningham JM
    (1989) Cell 57:659–666.
    CrossRefMedlineISI
  2. ↵
    1. Hein S,
    2. Prassolov V,
    3. Zhang Y,
    4. Ivanov D,
    5. Lohler J,
    6. Ross S,
    7. Stocking C
    (2003) J Virol 77:5926–5932.
    Abstract/FREE Full Text
  3. ↵
    1. Johann SV,
    2. Gibbons JJ,
    3. O'Hara B
    (1992) J Virol 66:1635–1640.
    Abstract/FREE Full Text
  4. ↵
    1. vanZeijl M,
    2. Johann SV,
    3. Closs E,
    4. Cunningham J,
    5. Eddy R,
    6. Shows TB,
    7. O'Hara B
    (1994) Proc Natl Acad Sci USA 91:1168–1172.
    Abstract/FREE Full Text
  5. ↵
    1. Miller DG,
    2. Edwards RH,
    3. Miller AD
    (1994) Proc Natl Acad Sci USA 91:78–82.
    Abstract/FREE Full Text
  6. ↵
    1. Tailor CS,
    2. Nouri A,
    3. Lee CG,
    4. Kozak C,
    5. Kabat D
    (1999) Proc Natl Acad Sci USA 96:927–932.
    Abstract/FREE Full Text
  7. ↵
    1. Yang Y,
    2. Guo L,
    3. Xu S,
    4. Holland C,
    5. Kitamura T,
    6. Hunter K,
    7. Cunningham J
    (1999) Nat Genet 21:216–219.
    CrossRefMedlineISI
  8. ↵
    1. Battini J,
    2. Rasko J,
    3. Miller AD
    (1999) Proc Natl Acad Sci USA 96:1385–1390.
    Abstract/FREE Full Text
  9. ↵
    1. Mendoza R,
    2. Anderson MM,
    3. Overbaugh J
    (2006) J Virol 80:3378–3385.
    Abstract/FREE Full Text
  10. ↵
    1. Takeuchi Y,
    2. Vile RG,
    3. Simpson G,
    4. O'Hara B,
    5. Collins MKL,
    6. Weiss RA
    (1992) J Virol 66:1219–1222.
    Abstract/FREE Full Text
  11. ↵
    1. Quigley JG,
    2. Yang Z,
    3. Worthington MT,
    4. Phillips JD,
    5. Sabo KM,
    6. Sabath DE,
    7. Berg CL,
    8. Sassa S,
    9. Wood BL,
    10. Abkowitz JL
    (2004) Cell 118:757–766.
    CrossRefMedlineISI
  12. ↵
    1. Quigley JG,
    2. Burns CC,
    3. Anderson MM,
    4. Lynch ED,
    5. Sabo KM,
    6. Overbaugh J,
    7. Abkowitz JL
    (2000) Blood 95:1093–1099.
    Abstract/FREE Full Text
  13. ↵
    1. Tailor CS,
    2. Willett BJ,
    3. Kabat D
    (1999) J Virol 73:6500–6505.
    Abstract/FREE Full Text
  14. ↵
    1. Barnett AL,
    2. Wense DL,
    3. Li W,
    4. Fass D,
    5. Cunningham JM
    (2003) J Virol 77:2717–2729.
    Abstract/FREE Full Text
  15. ↵
    1. Battini JL,
    2. Heard JM,
    3. Danos O
    (1992) J Virol 66:1468–1475.
    Abstract/FREE Full Text
  16. ↵
    1. Fass D,
    2. Davey RA,
    3. Hamson CA,
    4. Kim PS,
    5. Cunningham JM,
    6. Berger JM
    (1997) Science 277:1662–1666.
    Abstract/FREE Full Text
  17. ↵
    1. Heard JM,
    2. Danos O
    (1991) J Virol 65:4026–4032.
    Abstract/FREE Full Text
  18. ↵
    1. Peredo C,
    2. O'Reilly L,
    3. Gray K,
    4. Roth MJ
    (1996) J Virol 70:3142–3152.
    Abstract
  19. ↵
    1. Rigby MA,
    2. Rojko JL,
    3. Stewart MA,
    4. Kociba GJ,
    5. Cheney CM,
    6. Rezanka LJ,
    7. Mathes LE,
    8. Hartke JR,
    9. Jarrett O,
    10. Neil JC
    (1992) J Gen Virol 73:2839–2847.
    Abstract/FREE Full Text
  20. ↵
    1. Bupp K,
    2. Roth M
    (2002) Mol Ther 5:329–335.
    CrossRefMedlineISI
  21. ↵
    1. Bupp K,
    2. Sarangi A,
    3. Roth MJ
    (2005) J Virol 79:1463–1469.
    Abstract/FREE Full Text
  22. ↵
    1. Bupp K,
    2. Roth MJ
    (2003) Hum Gene Ther 14:1557–1564.
    CrossRefMedlineISI
  23. ↵
    1. Work L,
    2. Buning H,
    3. Hunt E,
    4. Nicklin S,
    5. Denby L,
    6. Britton N,
    7. Leike K,
    8. Odenthal M,
    9. Drebber U,
    10. Hallek M,
    11. Baker A
    (2006) Mol Ther 13:683–693.
    CrossRefMedlineISI
  24. ↵
    1. Hartl I,
    2. Schneider RM,
    3. Sun Y,
    4. Medvedovska J,
    5. Chadwick MP,
    6. Russell SJ,
    7. Cichutek K,
    8. Buchholz CJ
    (2005) Gene Ther 12:918–926.
    CrossRefMedlineISI
  25. ↵
    1. Masood R,
    2. Gordon E,
    3. Whitley M,
    4. Wu B,
    5. Cannon P,
    6. Evans L,
    7. Anderson W,
    8. Gill P,
    9. Hall F
    (2001) Int J Mol Med 8:335–343.
    MedlineISI
  26. ↵
    1. Hall FL,
    2. Gordon EM,
    3. Wu L,
    4. Zhu NL,
    5. Skotzko MJ,
    6. Starnes VA,
    7. Anderson WF
    (1997) Hum Gene Ther 8:2183–2192.
    MedlineISI
  27. ↵
    1. Peng K,
    2. Donovan K,
    3. Schneider U,
    4. Cattaneo R,
    5. Lust J,
    6. Russell S
    (2003) Blood 101:2557–2562.
    Abstract/FREE Full Text
  28. ↵
    1. Bupp K,
    2. Roth M
    1. Cid-Arregui A,
    2. Garcia Carranca E
    (2000) in Viral Vectors: Basic Science and Gene Therapy, eds Cid-Arregui A, Garcia Carranca E (Eaton, Natick, MA), pp 379–399.
  29. ↵
    1. Bupp K,
    2. Roth MJ