In Vivo Imaging of HIV Protease Activity in Amplicon Vector-transduced Gliomas

Proteases play key roles in many diseases including infections (1), cancer (2, 3), and autoimmune disorders (4). In recent years, proteases have received substantial attention as therapeutic targets (5), and it is estimated that ∼15% of all new drugs developed target proteases. The ability to evaluate specific enzyme activity in vivo would thus have considerable clinical and basic science applications. For example, real-time in vivo imaging of proteases has been used to improve the early detection of diseases (6, 7), to serve as an in vivo screening tool for drug development, to image the efficacy of protease inhibitors, or to understand how protease activities are regulated in intact micro- and macroenvironments. Although viral proteases serve as important drug targets for viral infections (8), there currently exist no imaging probes that would allow in vivo quantitation of viral proteases.

Here we report the development and validation of an HIV-1 protease (PR) imaging agent based on an activatable near-infrared fluorescence (NIRF) substrate. HIV-1PR belongs to a family of aspartic acid proteases and specifically cleaves VSQNYPIV as well as other related and peptidomimetic sequences with high specificity (9). The protease is initially synthesized in an inactive state as a part of the gag-pol precursor protein and is activated in the final stages of the virus life cycle. Processing of the precursor yields free, active protease as well as the reverse transcriptase (RT) and virion core structural proteins (10, 11). HIV-1PR activation is required for virion maturation, and inhibition of its activity results in immature virus particles containing unprocessed gag-pol precursors that have lost their ability to re-infect cells (12). Imaging in the NIR spectrum (700–900 nm) maximizes tissue penetrance in addition to minimizing the autofluorescence from non-target tissue (13). It offers the possibility of using the fluorescence-tagged reporter probes, which are injectable i.v. to impart molecular specificity. These probes can be physically quenched to minimize signal in the non-activated state and can become brightly fluorescent after activation by specific proteins at the target site (14).

In this study, we have used herpes simplex virus (HSV)-1 amplicon vectors to express HIV-1PR in target cells. Human glioma cells infected with the HIV-1PR amplicon vector in culture and incubated with the HIV-PR-NIRF probe showed specific activation of the imaging probe. Human glioma tumors in nude mice were reliably transduced by intratumoral injection of the amplicon vector, and the HIV-1PR expression could be directly imaged in vivo.

The HIV-1PR-selective NIRF probe was based on a specific peptide substrate GVSQNYPIVGK(FITC)C-NH₂ (9) conjugated to a delivery graft copolymer (15). The HIV-1PR recognizes the bolded region and cleaves the peptide between tyrosine and proline (Fig. 1). Specifically, peptide substrate was conjugated to the lysine side chain of a partially methoxy polyethylene glycol-protected poly-l-lysine graft copolymer (16) through a thiol ether linkage. After conjugation, the monoreactive indocyanine dye, Cy5.5 (Amersham, Arlington Heights, IL) was conjugated to the NH₂ terminus of each grafted HIV-1PR peptide substrate. The probe concentration was quantitated by absorption measurements of Cy5.5. A section of polyethylene glycol-polylysine graft copolymer was modeled using a Metropolis Monte Carlo simulation. Peptide binding to HIV-1PR was modeled using the crystal structure (1YTG; Protein Data Bank). Models were rendered using MSMS (Micheal Sanner, Scripps Institute, San Diego, CA) and Raster3D (17).

Amplicon plasmids, pHGCX expressing green fluorescent protein (GFP; Clontech, San Jose, CA; from Dr. Yoshi Saeki, Massachusetts General Hospital) and pKSR2 (derived from pHGCX) expressing red fluorescent protein (DsRed2; Clontech) under the immediate-early IE4/5 HSV promoter were used as backbone constructs (Fig. 1). The cDNA sequence encoding a 300-bp HIV-1PR fragment was amplified by PCR using the HIV-1PR cDNA cloned in p83.2 (a gift from Dr. Scadden, Harvard Medical School) as a template. Primer HIVpNhe1f (5′-CTAGCTAGCATGCCTCAGATCACTCTTTGGC-3′) and primer HIVpBamHInr (5′-CGCGGATCCCTAAAAATTTAAAGTGCAGCCAATCTG-3′) amplified a PCR fragment of 300 bp that was cloned downstream of the cytomegalovirus promoter in a NheI- and BamHI-cleaved pHGCX amplicon, resulting in the A-HIV-1PR construct. To generate the amplicon bearing the DsRed2 reporter gene and a PR-GFP fusion protein, the cDNA sequence encoding the GFP was amplified from pEGFP-N3 (Clontech) by PCR using the primers eGFPHindIIIf (5′-CCCAAGCTTTATTTTCTTCGATGTCCCTG-3′) and eGFPXhoIr (5′-CCGCTCGAGCGGCCGCCCTTGGACCAGATA-3′). The cDNA sequence encoding a 300-bp HIV-1PR fragment was amplified by PCR as above using primer HIVpNhe1f and HIVpHindIIIr (5′-CCCAAGCTTCTAAAAATTTAAAGTGCAGCCAATCTG-3′). A 700-bp GFP fragment flanked by HindII1f and XhoIr sites was ligated together with the 300-bp HIV-1PR fragment flanked by NheI and HindII1f downstream of the cytomegalovirus promoter into NheI/XhoI-digested pKSR2 generating the amplicon A-HIV-1PR-GFP.

All amplicon plasmid constructs were packaged as HSV amplicon vectors by using a helper virus-free packaging system (18). For this purpose Vero 2-2 cells were transfected by using Lipofectamine (Invitrogen, Carlsbad, CA) with a mixture of the amplicon DNA to be packaged, plasmid DNA containing the ICP27 gene, and a bacterial artificial chromosome DNA that includes the entire HSV genome deleted for the sequences containing the DNA cleavage-packaging signals (ori-pac sequences) and the essential gene, ICP27. Amplicon vector stocks were harvested 60 h later, freeze-thawed three times, sonicated, and purified by brief centrifugation at 1000 × g for 10 min. Amplicon titers, transducing units (TUs)/ml were determined by infecting 293T/17 cells/well in 24-well plates and then counting the GFP- or red fluorescent protein-positive cells at ×10 magnification 18 h after infection.

HIV-1 protease (HIV-1 HXB2 KIIA, from Drs. Davis, Stahl, Wingfield, and Kaufman) was obtained through the AIDS Research and Reference Reagent Program (National Institute of Allergy and Infectious Diseases, Bethesda, MD). The HIV-1PR-NIRF probe was tested for activity by incubating 1 μg of protease in 100 μl with 0.2 μm HIV-1PR-NIRF probe and incubated at 37°C for 1 h. Clinical protease inhibitor, indinavir (a gift from Dr. Marty Hirsch, Department of Immunology and Infectious Diseases, Harvard School of Public Health), at concentrations ranging from 0.1 to 10 μm, was incubated with 1 μg of protease for 10 min at room temperature and then incubated with the HIV-1PR-NIRF probe in a final volume of 100 μl at 37°C for 1 h. Fluorescence was measured on a Spectra Max Gemini fluorescence multiwell plate reader (Molecular Devices, Sunnyvale, CA) with an excitation wavelength of 675 nm and an emission wavelength of 694 nm.

African green monkey kidney Vero 2-2 cells provided by Dr. Rozanne Sandri-Goldin (University of California, Irvine, CA) and Gli36 human primary glioma cells provided by Dr. Anthony Campagnoni (UCLA, Los Angeles, CA; Ref. 19) were grown in DMEM with 10% fetal bovine serum (Sigma Chemical Co., St. Louis, MO) at 37°C in a humidified atmosphere with 5% CO₂ and 1% penicillin/streptomycin (Invitrogen, Grand Island, NY). To generate Gli36 cells stably expressing HIV-1PR, cells were cotransfected with pHIV-1PR plasmid and pBabePuro plasmid (containing the gene for puromycin resistance; Ref. 20) using Lipofectamine (Invitrogen, Carlsbad, CA). Gli36HIV-1PR-positive clones were selected using 1 μg/ml puromycin in the growth medium, expanded, and checked for HIV-1PR mRNA expression by RT-PCR, as described below. The clone with the highest level of mRNA for HIV-1PR, Gli36HIV-1PR, was used for further studies.

The NIRF probe was tested in cell culture using Gli36HIV-1PR cells and Gli36 cells with or without infection with HIV-1PR amplicon vector at a multiplicity of infection (MOI) of 1. Cells were grown to confluence in a 24-well dish, and the medium was replaced with 200 μl of fresh medium containing 0.2 μm HIV-1PR-specific NIRF probe. After incubation for 1 h at 37°C, cells were washed three times and then visualized under the fluorescence microscope (Axiovert) or analyzed by flow cytometry with the FACS Calibur system (BD Bioscience).

Cells were grown on coverslips in a 24-well dish and infected with amplicon vectors as described above. After 48 h, fluorescence of the cells was analyzed by laser scanning microscope 5PASCAL (Carl Zeiss Germany). A ×40 oil immersion objective (numerical aperture, 1.3) was used for scanning cells with step (pixel) size of 101 μm in the X plane and 7.4 μm in the Y plane. The pinhole setting was 60 μm, which yielded a theoretical thickness (full width at half maximum) of ∼0.05 μm.

Gli36 tumor cells in mid-log phase were harvested by trypsinization, and single-cell suspensions of 5 × 10⁶ cells in DMEM were injected s.c. into the lateral abdomen of nude mice 4–5 weeks of age. Five-seven days later, when tumors had grown to 2–7 mm in diameter, mice received intratumoral injections of either A-HIV-1PR amplicon vector (2.5 × 10⁸ TUs/ml) or control amplicon vector, HGCX (2.6 × 10⁸ TUs/ml), in a total volume of 20 μl, with intratumoral manipulation of the needle to ensure spread of virus. Forty-eight h after viral injections, animals bearing tumors were sacrificed by intracardiac perfusion with 10 mm sodium phosphate buffer (pH 7.3), followed by 4% paraformaldehyde in PBS. Tumors were removed, sectioned on a sledge microtome (10-μm slices), and visualized for GFP fluorescence.

Total RNA was isolated from Gli36HIV-1PR cells and Gli36 cells using Trizol reagent (Life Technologies, Inc.) as recommended by the manufacturer. Four-five μg of total RNA were mixed with random hexamers, first-strand synthesis RT buffer, and multiscribe RT and RNase inhibitor in 20 μl. The RT reaction was performed at 25°C for 10 min and subsequently at 48°C for 2 h. The cDNA synthesized by RT was denatured at 94°C for 5 min, and 1 μg of resulting cDNA was used to detect HIV-1PR-specific transcripts by PCR amplification using HIV-1PR-specific primers HIVpNhe1f and HIVpBamHInr that amplified a final product of 300 bp. The PCR was performed under the following conditions: 94°C for 3 min, 30 cycles at 94°C for 45 s, 55°C for 30 s, 72°C for 1 min, and a final extension at 72°C for 10 min. The reaction was performed in a final volume of 50 μl.

Cells were lysed 48 h after infection with A-HIV-1PR-GFP amplicon vector, in a buffer containing 50 mm Tris (pH 8.0), 150 mm NaCl, and 1% NP40, supplemented with single-strength proteinase inhibitor Complete-Mini cocktail (Boehringer-Mannheim, Indianapolis, IN.). The lysates were passed through a 25-μm needle and centrifuged at 10,000 × g for 30 min at 4°C. The total amount of protein for all samples was determined by using the Coomassie Plus protein assay reagent (Pierce, Rockford, IL) and a BSA standard (Bio-Rad, Hercules, CA). An equal amount of total cell protein (60 μg) was denatured and separated by SDS-PAGE and transferred to nitrocellulose membrane (Immobilon) in a transfer buffer (25 mm Tris, 192 mm glycine, pH 8.3) by using a Bio-Rad Transblot Cell for 3 h at 0.5 mA at 4°C. The membranes were washed in TBS-T buffer [10 mm Tris-HCl (pH 7.5), 150 mm NaCl, and 0.05% Tween 20] and blocked with 10% nonfat dried milk in TBS-T for 1 h at room temperature, washed with TBS-T, and incubated with the rabbit antibody to GFP (Clontech; final dilution, 1:1000) for 1 h at room temperature. After three washes in TBS-T, membranes were incubated for 30 min with a 1:5000 dilution of donkey anti-rabbit IgG peroxidase-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) in TBST. After washing, as described above, the blots were developed by using enhanced chemiluminescence (ECL) reagents (Amersham Life Sciences). Membranes were then exposed to film (30 s to 30 min). Films were scanned, and the bands were analyzed for their integrated densities by using NIH Image 1.62 software.

Anesthetized mice (n = 6) were imaged 24 h after i.v. injection of 2.5 nmol of HIV-1PR-specific NIR fluorochrome (corresponding to 2.5 mg of probe/kg body weight). The imaging system consisted of a light-tight imaging chamber equipped with a 150-W halogen white light and an excitation bandpass filter (610–650 nm; Omega Optical, Brattleboro, VT) to excite Cy5.5. Fiberoptic cables and light diffusers resulted in a relatively spatially homogenous photon source over the imaging area. A CCD camera equipped with a f/1.2 12.5–75 mm zoom lens and an emission long pass filter at 700 nm (Omega Optical) was used to detect the fluorescence. Image analysis was performed using CMIR Image. Elliptical regions of interest were selected manually from the tumor center and tumor periphery for Gli36HIV-1PR tumors and for the entire Gli36 tumors, as well as for representative adjacent normal tissue (thigh). The mean and SD of pixel values were recorded. Contrast-to-noise ratios were calculated for each animal. Significance of differences among groups was determined using a one-tailed Student’s t test.

A schematic of the amplicon constructs and the vector packaging strategy used in this study are shown in Fig. 1, A and B. Computer modeling demonstrated the HIV-1PR binding and activation of the HIV-1PR NIRF graft copolymer (Fig. 1, C–F). A section of the probe in the inactive (Fig. 1,C) and activated (Fig. 1,D) forms are represented in a space-filling model. The binding of the Cy5.5-peptide substrate to the HIV-1PR (Fig. 1,E, green) results in the cleavage of the Cy5.5 peptide substrate at the cognate cleavage sites Tyr-Pro. Upon cleavage, the residues COOH-terminal to proline, containing Cy5.5, were released from the graft copolymer, thus resulting in fluorescence activation (Fig. 1 F).

The infectivity of the HSV amplicon vectors for human glioma Gli36 cells was assessed by infecting the cells with A-HIV-1PR or A-HIV-1PR-GFP at an MO1 of 1 and assessing for either GFP or DsRed2 fluorescence 18 h after infection. As shown in Fig. 2,A, ∼90% of the cells were infected with both amplicon vectors. Confocal microscopy performed 48 h after infection with A-HIV-1PR-GFP construct revealed that the protease-GFP fusion protein was localized in the cytoplasm in a punctuate pattern, whereas DsRed2 was distributed evenly throughout the cytoplasm and the nuclei (Fig. 2,B). Total protein from the infected cells was harvested, fractionated by denaturing SDS-PAGE, and immunoblotted using an antiserum against GFP. The expression of the 28 kDa protein corresponding in size to that of GFP was present in A-HIV-1PR-infected cells (Fig. 2,C, Lane 1). In A-HIV-1PR-GFP-infected cells, a 40-kDa protein corresponding to the size of HIV-1PR-GFP-fusion protein was seen in addition to the cleaved GFP protein and a 34-kDa protein of intermediate size (Fig. 2 C, Lane 2). These results indicate that HIV-1PR protease can be expressed in Gli36 cells in culture via HSV amplicon vectors. Cleavage of some HIV-1PR-GFP fusion protein into a 28-kDa GFP and a 34-kDa protein indicates the autocatalytic cleavage of the HIV-1PR.

To test the specificity of the NIRF probe, it was incubated with purified HIV-1PR and visualized at 675 nm excitation and 694 nm emission. The native HIV-1PR-NIRF probe was essentially optically “silent”, generating <180 AU, whereas the HIV-1PR activated probe generated over 7-fold higher fluorescence (Fig. 3,A). To test whether the activation of the HIV-1PR-NIRF probe could be inhibited, the protease was incubated with different concentrations of a clinically used inhibitor, indinavir, and then incubated with the HIV-1PR-NIRF probe. This resulted in inhibition of protease activity and thus a decrease in NIRF fluorescence in a dose-dependent manner (Fig. 3 B).

To characterize the stably transfected GFP-positive, Gli36HIV-1PR cells (Fig. 4,A), RT-PCR analysis for the HIV-1PR message was performed on the total mRNA isolated from these cells. The Gli36HIV-1PR cells expressed this message, whereas the control cells (Gli36) were negative for HIV-1PR mRNA expression (Fig. 4,B). We also tested the activation of HIV-1PR-NIRF probe in cultures of Gli36-HIV-1PR cells and Gli36 cells. The protease-expressing cells showed dramatically increased NIR fluorescence as compared with the control Gli36 cells (Fig. 4 C). These results show that the HIV-protease NIRF probe combination retains specificity in mammalian cells.

To demonstrate that HIV-1PR can be delivered into the tumors in vivo with HSV amplicon vectors and imaged using the NIRF probe, we first performed experiments in culture on Gli36 cells infected with A-HIV-1PR amplicon vector. Cells were infected with an A-HIV-1PR or control HGCX vector at an MOI of 1 and incubated with the NIRF probe 36 h after infection. The cells infected with the protease-encoding vector showed a significantly higher NIRF fluorescence as compared with the HGCX-infected cells (Fig. 5, A and C). At a higher magnification, NIRF signal was found mostly in the cytoplasm, again in a punctuate pattern, in A-HIV-1PR-infected cell as compared with the primarily nuclear GFP signal seen in both A-HIV-1PR- and HGCX vector-infected cells (Fig. 5,B). The NIRF probe was specifically activated in Gli36 cells infected with A-HIV-1PR amplicon (Fig. 5 C).

To determine whether the HIV-1PR could be used as a marker transgene for in vivo imaging, nude mice bearing lateral abdomen Gli36 gliomas (5–7 mm in diameter) were injected with HIV-1PR amplicon vector (2.5 × 10⁵ TUs in 20 μl) into the left tumor and HGCX vector (2.4 × 10⁵ TUs in 20 μl) into the right tumor, and 36 h later HIV-1PR-NIRF probe (40 nmol in 150 μl of saline) was administered systemically i.v. Subsequent NIRF imaging showed significantly higher (>4 fold) signal from HIV-1PR amplicon-infected Gli36 tumors as compared with HGCX-infected flanking tumors (Fig. 6, A–C). Sections of HIV-1PR amplicon vector-injected tumors assessed histologically by GFP fluorescence verified the infection of tumor cells by the vector (Fig. 6 D). These results show that HIV-1PR delivered via viral vectors can be imaged in live animals.

In this study we have investigated a novel approach for imaging HIV-1PR. We have shown that this method can be used to image cells infected with vectors, e.g., HSV amplicon vectors expressing HIV-1PR. In addition, HIV-1PR, virally delivered into tumors (e.g., gliomas), was directly imaged in vivo in mice. Tumors infected with the protease-expressing vector showed a markedly higher NIRF signal when compared with tumors infected with control vectors.

The ability to image gene expression in vivo is critical to assess the efficacy of novel targeted vectors and to quantitate their therapeutic effects. In prior work, we have developed protease-specific NIRF imaging probes that are activated by endogenous proteases (6). We have thus imaged endogenous proteases [up-regulated in tumors (6)] and cardiovascular disease (21) and have been able to evaluate protease inhibitor treatments (22). To directly image transgenes, however, it is essential to develop protease/substrate systems that are not ubiquitously expressed in mammalian systems. This can be a challenging task because of the evolutionary similarities of protease families. We thus decided to investigate viral proteases such as HIV1-PR. The specificity of HIV-1PR has been the subject of numerous studies (23). Kinetic, modeling, and substrate studies have revealed that an extended β conformation of the substrate sequence within the HIV-1PR binding site is essential for altering the confirmation on the surface of the HIV-1PR dimer for optimal processing (9). We adapted the VSQNY*PIV sequence as a unique HIV-1PR substrate and modified it terminally for use in our delivery system. We observed that the resultant NIRF probe was not activated by any endogenous proteases present in human glioma cells but was specifically activated by HIV-1PR when this protease was introduced into cells via stable transfection or vector infection.

The HIV-1PR-specific NIRF probe was designed based on a synthetic graft copolymer (partially methoxy-polyethylene glycol-modified poly-l-lysine) originally developed as a long-lasting drug carrier with circulation times in excess of 24 h in rodents and 20 h in humans (24),^. The graft copolymer is slowly internalized into proliferating tumor cells by endocytosis (25) and has also been used for chemotherapeutic drug delivery (16). Fig. 1 summarizes the structure of the graft copolymer as well as the interaction of the grafted PR substrates with HIV1-PR. Upon cleavage, the NH₂-terminal fragment of Cy.5.5 is released still attached to PIV-Cy5.5, which is brightly fluorescent. The latter can be detected by a variety of means, such as intravital microscopy (16), mesoscopic fluorescence-mediated tomography (26), or surface-weighted reflectance imaging (27), such as done in the current study. NIRF imaging used in this study, however, limits the applicability of the technique to semiquantitative imaging of objects at depths of only 5–6 millimeters. Imaging optical signatures deeper in tissues requires the application of advanced excitation-detection schemes and the use of tomographic principles for combining data acquired at different projections. With fluorescence-mediated tomography imaging, it is currently possible to detect femtomolar concentrations of Cy5.5 with millimeter spatial resolution deep inside animals (28), and it is possible to penetrate human tissues up to ∼10 cm (29).

The developed methodology has several practical applications, the most obvious ones include quantitation of transgene expression in vivo and/or direct imaging of viral proteases in infected subjects. With respect to the former, imaging in the NIR (700–800 nm) offers some advantages over using photoproteins that fluoresce in the visible (450–610 nm) range. Foremost, NIR light travels through tissue more efficiently, thus enabling seeing in tissues deeper than the 1–2 mm currently available for GFP. NIR also has the added advantage of much lower tissue autofluorescence, and mathematical light reconstructions enable true three-dimensional quantification of substrate conversion in vivo (30). The second major application (direct imaging of HIV1-PR during viral replications) should also be feasible with the developed probes. Although we were able to test this in cell culture using HSV amplicon vectors, we were not able to test this in vivo in the current study because of the inherent difficulties of accessing infected mice. However, it is anticipated that scale up of the synthetic protocol will allow us to test the efficacy in higher mammals using clinical detection systems.

In conclusion, we show that HIV-1PR can be imaged in transfected tumors in culture and in vivo. It should now be possible to engineer therapeutic proteins that can be activated by HIV-1PR expression for tumor therapy, for example, by induction of tumor cell apoptosis with simultaneous visualization of gene expression in vivo.

We thank Dr. Qing Zeng for help with animal imaging.