Intranuclear Protein Transduction through a Nucleoside Salvage Pathway*

Regulation of gene expression by intranuclear transduction of macromolecules such as transcription factors is an alternative to gene therapy for the treatment of numerous diseases. The identification of an effective intranuclear delivery vehicle and pathway for the transport of therapeutic macromolecules across plasma and nuclear membranes, however, has posed a significant challenge. The anti-DNA antibody fragment 3E10 Fv has received attention as a novel molecular delivery vehicle due to its penetration into living cells with specific nuclear localization, absence of toxicity, and successful delivery of therapeutic cargo proteins in vitro and in vivo. Elucidation of the pathway that allows 3E10 Fv to cross cell membranes is critical to the development of new molecular therapies. Here we show that 3E10 Fv penetrates cells through a nucleoside salvage transporter. 3E10 Fv is unable to penetrate into cells deficient in the equilibrative nucleoside transporter ENT2, and reconstitution of ENT2 into ENT2-deficient cells restores 3E10 Fv transport into cell nuclei. Our results represent the first demonstration of protein transport through a nucleoside salvage pathway. We expect that our finding will facilitate a variety of methods of gene regulation in the treatment of human diseases, open up new avenues of research in nucleoside salvage pathways, and enhance our understanding of the pathophysiology of autoimmune diseases.

The ability to regulate gene expression through intranuclear delivery of macromolecules would significantly impact the treatment of a multitude of human diseases. Effective macromolecular therapy is dependent upon molecular delivery vehicles to circumvent the plasma membrane barrier and facilitate intracellular transport of cargo molecules. The single chain Fv fragment of the 3E10 anti-DNA autoantibody (3E10 Fv) has recently been harnessed as a novel molecular delivery vehicle due to its specific nuclear localization and apparent lack of toxicity (1). 3E10 Fv and Fv fusion proteins readily transduce across cell membranes and penetrate into cell nuclei, and 3E10 Fv has successfully delivered biologically active proteins such as Hsp70 (2) and p53 (3) into living cells in vitro. Moreover, 3E10 Fv mediated full-length p53 protein therapy in vivo (4). The pathway that carries 3E10 Fv across cell membranes and into cell nuclei, however, has not been identified.
Previous studies implicated DNA binding as important in 3E10 Fv transduction into cell nuclei. Specifically, mutations that abrogate DNA binding by the antibody render it incapable of cellular penetration (5). The association between cellular penetration and DNA binding distinguished 3E10 Fv from other protein transduction domains and implied that nucleoside salvage pathways might be involved in 3E10 Fv transport. Both concentrative (CNT) 3 and equilibrative (ENT) nucleoside salvage transporters mediate the uptake of nucleobases and nucleosides by mammalian cells (6). Since any major role of CNTs in 3E10 Fv transport was excluded by previous studies that demonstrated 3E10 Fv penetration into COS-7 cells that lack endogenous CNTs (2, 7), we examined the role of ENTs in 3E10 Fv transport.
Plasmids-A construct for expression of 3E10 Fv in the X-33 strain of Pichia pastoris, pPICZ␣A-Fv, was generated as described previously (9).
Purification of 3E10 Fv-3E10 Fv was purified from the supernatant of P. pastoris transfected with pPICZ␣A-Fv as described previously (9).
Cellular Penetration Assays-Purified 3E10 Fv was exchange-dialyzed into PBS prior to application to cells. After dialysis, 10% fetal calf serum was added to the buffer. Control buffer was PBS with 10% fetal calf serum. For adherent cell lines (COS-7, PKNTD/ENT1, and PKNTD/ENT2), 50 l of control buffer or 3E10 Fv was added to cells on 96-well plates for 1 h. After incubation with 3E10 Fv, the antibody fragment was removed, and cells were washed, fixed in chilled 100% ethanol, and stained with the 9E10 ␣-Myc antibody as described previously (9). For non-adherent cells (K562 and CEM/ENT1), cell pellets composed of ϳ200,000 cells were resuspended in 100 l of control buffer or 10 M 3E10 Fv and allowed to incubate with intermittent shaking at 37°C for 1 h. Cells were then centrifuged at 100 ϫ g for 2 min and washed three times with PBS. Next, cells were spread on glass slides and allowed to dry overnight. Cells were then fixed in chilled 100% ethanol for 10 min, * This work was supported by a Veterans Affairs grant (to R. H. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  washed three times in PBS, and stained with the 9E10 ␣-Myc antibody.
Nucleoside Transporter Inhibition Assay-Nitrobenzylmercaptopurine riboside(NBMPR) was purchased from Sigma, and a stock solution of 100 mM NBMPR in Me 2 SO was prepared. To control for the effects of Me 2 SO in cell culture, Me 2 SO was added to control buffers not containing NBMPR. The concentration of Me 2 SO in all control and experimental buffers was 0.1%. COS-7 cells were pretreated for 30 min with control buffer (PBS ϩ 10% fetal calf serum) or buffer containing 10 or 100 M NBMPR. Buffers were then replaced with control buffer or 10 M 3E10 Fv in the presence or absence of 10 or 100 M NBMPR for 1 h. Cells were then washed, fixed, and stained with the 9E10 ␣-Myc antibody.
Microscopic Images-As described previously, images of cells were acquired with an Olympus IX70 inverted microscope with RC reflected light fluorescent attachment and MagnaFire SP Digital Imaging System (Olympus, Melville, NY) (10). The scale bar in cell images in Figs. 1-3 ϭ 5 m.

RESULTS
ENT1 and ENT2 mediate equilibrative nucleoside transport in mammalian cells, and both ENT1 and ENT2 activity is inhibited by high concentrations of NBMPR (8). We therefore tested NBMPR for inhibition of 3E10 Fv transport. Purified 3E10 Fv (constructed with His 6 tag for purification and Myc tag for identification) migrated as a single ϳ30-kDa protein on SDS-PAGE (Fig. 1A). Transduction of 3E10 Fv into COS-7 cells was confirmed by incubating cells with 10 M 3E10 Fv for 1 h at 37°C followed by Western blot analysis of cell lysates (Fig. 1B) or immunocytochemical staining of cells (Fig. 1C, top panels). Western blot analysis of cell lysates demonstrated the presence of an ϳ30-kDa Myc-tagged protein inside cells treated with 3E10 Fv (Fig. 1B, lane 2), which indicated penetration of the full-length antibody fragment. Furthermore, immunocytochemical staining confirmed nuclear localization by 3E10 Fv (Fig. 1C, top right panel), consistent with previous confocal microscopy and immunocytochemical studies on the antibody (2, 10). Next, COS-7 cells were pretreated for 30 min with control buffer or buffer containing 100 M NBMPR prior to a 1-h incubation with 10 M 3E10 Fv in the presence or absence of NBMPR. Subsequent immunocytochemical staining of the cells demonstrated that 100 M NBMPR suppressed nuclear penetration by 3E10 Fv (Fig. 1C, bottom left panel), which suggested that ENT1 or ENT2 is involved in 3E10 Fv transport.
In an attempt to resolve which of the ENTs was linked to 3E10 Fv transport, 3E10 Fv penetration into COS-7 cell nuclei was tested in the presence of a lower dose of NBMPR to take advantage of the different K i of NBMPR for ENT1 and ENT2 (0.4 nM versus 2.8 M, respectively) (8). At 10 M, NBMPR ENT1 activity is completely inhibited, whereas ENT2 retains moderate activity. In contrast to the distinct inhibition of 3E10 Fv transport provided by 100 M NBMPR (Fig. 1C, bottom left  panel), 3E10 Fv successfully penetrated COS-7 cell nuclei in the presence of 10 M NBMPR (Fig. 1C, bottom right panel). This result suggested that ENT2, not ENT1, mediated transport of 3E10 Fv. The decreased nuclear staining intensity in cells treated with 3E10 Fv ϩ 10 M NBMPR when compared with cells treated with 3E10 Fv alone likely reflects partial inhibition of 3E10 Fv transport due to the expected Ͼ50% suppression of ENT2 activity by 10 M NBMPR.
As an additional approach to identifying the equilibrative nucleoside transporter(s) involved in 3E10 Fv transduction, we examined 3E10 Fv transduction into the CEM/ENT1 cell line that lacks ENT2 (11). 10 M 3E10 Fv was applied to the CEM/ ENT1 cells for 1 h. As a positive control, 10 M 3E10 Fv was also applied to K562 leukemia cells that express both ENT1 and ENT2 (12). As predicted, immunocytochemical staining of the 3E10 Fv-treated K562 cells demonstrated penetration of the antibody fragment into ϳ100% of the cells ( Fig. 2A, right panel). CEM/ENT1 cells treated with 3E10 Fv, however, showed no staining (Fig. 2B, right panel). This result demonstrated that the absence of ENT2 significantly impaired transduction by 3E10 Fv into cells. Taken together, the inhibition of 3E10 Fv transport by high concentrations of NBMPR and failure of 3E10 Fv to penetrate a cell line lacking ENT2 strongly support a role of ENT2 in 3E10 Fv transport.
To verify that ENT2 facilitates 3E10 Fv intranuclear protein transduction, experiments were performed on nucleoside transporter-deficient PK15 cells (PKNTD) with either ENT1 or ENT2 reconstituted through stable transfection and expression of ENT1 or ENT2 cDNA (8). PKNTD/ENT1 cells treated for 1 h with 10 M 3E10 Fv showed no evidence of nuclear penetration (Fig. 3A, right panel), similar to the results obtained with the CEM/ENT1 cells. By contrast, 3E10 Fv-treated PKNTD/ENT2 cells exhibited distinct nuclear staining (Fig. 3B, right panel), which indicated that restoration of ENT2 to the nucleoside transporter-deficient cells significantly augmented nuclear penetration by 3E10 Fv. This result confirmed that the presence of ENT2 facilitates nuclear penetration by 3E10 Fv and verified protein transduction of the antibody fragment through the ENT2-mediated nucleoside salvage pathway.

DISCUSSION
The nucleoside salvage pathways have been studied in detail, but protein transport through or related to nucleoside salvage has not been previously described. The precise interaction between 3E10 Fv and ENT2 remains to be determined, but it is tempting to speculate that 3E10 Fv may be carried into cells by virtue of its binding to nucleosides or nucleobases that are subsequently transported into cells by ENT2. Alternatively, 3E10 Fv may mimic the structure of a nucleoside or nucleobase that is recognized and transported into cells by ENT2. Elucidation of the mechanism by which ENT2 facilitates 3E10 Fv transport should yield further insights into both protein transduction and nucleoside salvage pathways. Furthermore, since ENT2 is located in both plasma and nuclear membranes, it will be important to ascertain whether ENT2 facilitates transport of 3E10 Fv across both cellular and nuclear membranes or whether another pathway is involved in nuclear penetration (13).
With regard to molecular therapy, the linkage between ENT2 and nuclear penetration by 3E10 Fv reported here further  establishes 3E10 Fv as a novel molecular delivery vehicle that is distinct from other protein transduction domains previously described. Endosomal localization by cell-penetrating peptides limits their role in molecular therapy (14,15), but the identification of 3E10 Fv transport through ENT2 provides a rationale for future studies on the use of 3E10 Fv in delivering molecules such as small interfering RNAs, antisense oligonucleotides, and transcription factors to cell nuclei. It also generates an impetus to determine whether toxic cell-penetrating antibodies utilize a nucleoside salvage pathway in cellular penetration since inhibition of nucleoside transporters might then be investigated as a means of limiting tissue damage by cytotoxic autoantibodies in certain autoimmune diseases. The discovery of intranuclear protein transduction by 3E10 Fv through the ENT2-mediated nucleoside salvage pathway has profound implications for cell biology, pharmacology, and medicine.