Improved cell-specificity of adeno-associated viral vectors for medullary thyroid carcinoma using calcitonin gene regulatory elements

Targeted gene therapy using recombinant adeno-associated virus (rAAV) vectors is a potential therapeutic strategy for treating cancer, and tissue-specific promoters may help with tissue targeting. Medullary thyroid carcinoma (MTC) is a disease of the calcitonin secreting thyroid C cells, and calcitonin is highly expressed in MTC tumors compared to other cells. To target MTC cells, we evaluated an rAAV serotype 2 vector (rAAV2-pM+104-GFP) containing a modified calcitonin/calcitonin gene related peptide promoter (pM+104) and a green fluorescent protein (GFP) reporter gene. In vitro transduction experiments comparing the MTC TT cell line with non-MTC cell lines demonstrated that rAAV2-pM+104-GFP infection yielded significantly (p < 0.05) higher GFP expression in TT cells than in non-MTC cell lines (HEK293 and HeLa), and significantly higher expression than in TT cells infected with the positive control rAAV2-pCBA-GFP vector. The rAAV2-pCBA-GFP control vector included a well-characterized, ubiquitously expresses control promoter, the chicken beta actin promoter with a cytomegalovirus enhancer (pCBA). In vivo experiments using a TT cell xenograft tumor mouse model showed that tumors directly injected with 2 x 1010 vg of rAAV2-pM+104-GFP vector resulted in GFP expression detected in 21.7% of cells, 48 hours after the injection. Furthermore, GFP expression was significantly higher for rAAV-pM+104-GFP treatments with a longer vector treatment duration and higher vector dose, with up to 52.6% (q < 0.05) GFP cells detected 72 hours after injecting 1x 1011 vg/tumor. These data show that we have developed an rAAV vector with improved selectivity for MTC.

The calcitonin/calcitonin gene related peptide (C/CGRP) gene is highly expressed in C cells and in MTC [16][17][18][19]. The C/CGRP gene is also expressed in some neuron and immune cells [16,18,20]. However, calcitonin is [21] 1000-fold more highly expressed in C cells and MTC cells, than in these other cell types [16,18,20]. Thus, the calcitonin/calcitonin gene related peptide promoter (pC/CGRP) is a favorable candidate to induce specific gene expression in MTC cells for targeted therapy.
Recombinant adeno-associated virus (rAAV) serotype 2 is a viral vector approved by the FDA for clinical applications [28][29][30]. Wild-type AAV (wt AAV) is non-pathogenic [31] and is able to infect both dividing and non-dividing cells [32,33]. It has been reported that rAAV2, with calcitonin promoter elements, can infect and transduce thyroid cells in vitro [21] with low efficiency. The low efficiency was likely due to use of single-stranded vectors available at the time, and Jiang et al. (2001) suggest they may have included some inhibitory elements in their promotor construct.
As a step to develop a potential novel therapy vector for MTC, with improvements over previous vectors, we engineered a double-stranded rAAV-based gene therapy vector (GTV) to express transgenes under the control of a modified, truncated (765 bp) calcitonin promoter (pM+104) [17]. We now report results from in vitro and in vivo approaches used to evaluate vectors in MTC TT cells in culture, and in a mouse xenograft tumor model system.

Plasmid construction
The full-length calcitonin promoter was amplified by polymerase chain reaction (PCR) from the MTC cell line TT using the forward 5'-CAGGGGTGTCGTGCTAAGAA-3' and reverse 5'-CCAGAATCTCGGGGCTCACCT-3' primers corresponding to -1738 to +125 base pairs, relative to the transcription start site. The PCR product was cloned into the pGEM vector (Promega, Madison, WI, USA) and sequenced at the University of Florida Interdisciplinary Center for Biotechnology Research (ICBR) DNA Sequencing core facility. The full-length calcitonin promoter was cloned into the pGL3 luciferase reporter vector (Promega). A pUC19 plasmid containing two copies of the modified enhancer (-1060 to -905) and one copy of the proximal promoter (-129 to +104) region were synthesized by Life Technologies and cloned into the pGL3 KpnI and NcoI sites.
Drs. George V. Aslanidi and Arun Srivastava (University of Florida) provided adeno-associated virus (AAV) plasmids with the CBA/CMV promoter/enhancer (pCBA) upstream of the GFP gene. The pCBA promoter/enhancer was removed and replaced with the pM+104 promoter/enhancer construct taken from the pGL3 plasmid (described above) using KpnI and NcoI restriction sites.

Luciferase assays
Luciferase transfections were performed in 96-well plates with cells at 70% cell density, which is equivalent to approximately 5 x 103-5 x 10 4 cells per well. Each well was transfected with 5 μl of serum-free media containing 0.1 μg of luciferase reporter plasmid (pGL3), 0.1 μg pRLrenilla (SV40 or TK promoter) control, and 0.3 μl of FuGene HD transfection reagent (Promega). At 48 hours post transfection (hpt), media was removed, then cells were briefly rinsed in phosphate-buffered saline (PBS) and lysed. All experiments were analyzed using the Dual-Luciferase Reporter Assay System (Promega) following the manufactures protocol. Luciferase activity in each well was measured at two wavelengths (530 and 640 nm) following addition of 100 μl of firefly luciferase substrate then 100 μl Stop & Glo Renilla substrate in a Biotek Synergy HT plate reader with Gen5 data analysis software (Biotek, Winooski, VT, USA) to detect the light output in relative luciferase units (RLU). Results are shown as the mean percentage of firefly luciferase activity (with the experimental promoter) compared to Renilla luciferase activity (with the internal control promoter). The averages of three independent experiments, performed in triplicate, are shown.

AAV vector production
Recombinant AAV vectors containing a GFP gene driven by the pCBA or pM+104 promoters were generated as described [34]. Briefly, HEK293 cells were transfected with a vector genome plasmid containing inverted terminal repeats (ITR), a rep/cap plasmid and a helper plasmid, using polyethylenimine (PEI, linear, Polysciences Inc., Warrington, PA, USA). Vectors were purified 72 hpt by iodixanol gradient centrifugation (Sigma-Aldrich) and ion exchange column chromatography using HiTrap SP HP (GE Healthcare, Piscataway, NJ, USA). Viral titers were determined using nucleic acid dot blot [35].

Recombinant AAV vector in vitro transduction assays
Thirty-thousand cells per well (96-well plate format) were transduced with 2 or 10 x 10 3 vg/cell of AAV2-GFP vector, and experiments were performed in triplicate. For pixel analysis, fluorescent images were captured 48 hpi at 4x magnification using a Biotek Cytation 5 Image Reader (Biotek). Transgene expression was assessed using Image J (National Institutes of Health (NIH), Bethesda, MD, USA) to calculate the total area of green fluorescence per visual field examined for each well (pixel 2 ), and for automatic counting of GFP cells using a threshold of 99.5%.

Animal care
Animals were cared for in accordance with the principles of the Guide to the Care and Use of Experimental Animals. Five or fewer mice of the same sex were housed together in a temperature-controlled room with a 12-hour light cycle and access to standard mouse chow, bedding material, and water ab libitum during the entire experiment. Mice were monitored daily, for general health, weight, and tumor size. Endpoints for euthanasia (CO 2 asphyxiation) were specified to when tumors longest diameter reached 1-cm, to minimize animals pain or distress as a result of tumor implantation. Tumor growth was measured up to several times a week, or daily near the end of the experiment, to ensure that maximal tumor size (1.2 cm) was not exceeded before the animal is euthanized. To reduce animal distress, mice were anesthetized using isoflurane (2-5%) in an inhalation chamber prior to all injection procedures.

Animal use
All animal studies were approved by the University of Florida (IACUC) under protocol #201708136. Five-week old male and female NOD.Cg-Prkdc SCID Il2rgtm1Wjl/SzJ (NSG) mice were obtained from Jackson Laboratories (Bar Harbor, ME, USA). At 6 weeks of age, mice were injected subcutaneously in the hind limb dorsal flank with 1 x 10 7 TT cells. When the diameter of the tumors reached~1 cm, the rAAV2 virus was injected directly into the tumor along with a rhodamine dye (rhodamine B isothiocyanate-Dextran, Sigma-Aldrich). Forty-eight or 72 (when indicated) hours after injection, the mice were euthanized on GD 18 by CO 2 asphyxiation followed by cervical dislocation and whole blood collection by terminal cardiac puncture, and their tumors and livers were harvested and processed for analysis.

Quantitative PCR of vector nucleic acid isolated from tumor tissue
At the time of tissue harvest, a slice of~30-50 mg tumor tissue was excised from the part of tumor bearing the traces of the rhodamine using a razor blade, and preserved along with liver lobes from the same mice at -20˚C with RNAlater solution (Ambion, Austin, TX, USA) for future processing. Tissue samples stored in RNALater solution (Ambion) were used for RNA and DNA purification with RNeasy Plus and DNeasy Blood & Tissue kits (QIAGEN, Germantown, MD, USA) correspondingly. Samples were finely minced on an ice chilled block by scalpel blades and split equally for RNA and DNA isolation. Complementary DNA (cDNA) was synthesized using iScript Reverse Transcription Supermix (Bio-Rad Laboratories, Hercules, CA, USA) with an RT cycle profile: 25˚C, 5 min; 46˚C, 20 min; 95˚C, 1 min; 4˚C hold. Realtime PCR (qPCR) was conducted using SsoAdvanced™ Universal SYBR1 Green Supermix (Bio-Rad Laboratories) and primers, targeting amplicon of about 120 bp: CommonGFP_For: 5'-TGCTTGTCGGCCATGAT -3' and CommonGFP_Rev2: 5'-AACTACAAGACCCGCGC -3'. PCR conditions were set according to the manufacturer of the real-time PCR amplification device (Bio-Rad CFX96), protocol CFX2stepAmp.prcl: 95˚C, 2 minutes; 2 Step PCR, 39 cycles: 95˚C,10 seconds, 55˚C, 30 seconds; Melting Curve Analysis. For quantitative analysis, standard dilutions of plasmid DNA samples were used that included the same GFP gene as in rAAV vectors.

Immunohistochemistry
Tumors were sliced in half along a plane that aligned perpendicular to the approximate injection site. Following fixation with 4% paraformaldehyde (PFA), samples were embedded in paraffin and sectioned with a Micron HM325 microtome (Walldorf, Germany). Slides containing 5 μm thick sections were deparaffinized and rehydrated through a series of ethanol (EtOH) /water washes. Following a 95˚C 25-minute antigen retrieval in Trilogy unmasking solution (Cell Marque, Rocklin, CA, USA). Slides were serum treated for one hour at room temperature with 2% normal goat serum (Vector Laboratories, Burlingame, CA, USA). Chicken antibody to GFP (Abcam, Cambridge, MA, USA) was applied at 1:750 overnight at 4˚C. Positive staining was detected with a one-hour incubation in 1:500 Alexafluor goat anti-chicken 488 (Invitrogen, Carlsbad, CA, USA) mounted in Vectashield (Vector Laboratories) with 4', 6-diamidino-2-phenylindole (DAPI) as counterstain.

Histology
The tumor and liver tissues were removed from the mice, cut in half, and placed inside of 20ml scintillation vials. The tissues were immersion fixed overnight in 4% paraformaldehyde (PFA) before rinsing with 1x phosphate buffered solution (PBS) and transferred to a 25% sucrose solution to equilibrate for three days at 4˚C. Once equilibrated, prepared tissues were embedded in an optimal cutting temperature (OCT) medium, and frozen in a 2-methylbutane/dry ice slurry. The sections were cut at 7μm thickness, thaw mounted onto positively charged slides and cover slipped in Vectashield mounting media containing DAPI (Vector Laboratories). GFP and DAPI in tissue sections were visualized using the Olympus BX43 upright microscope and DP80 camera mount, using Olympus cellSens software (Olympus, Pittsburgh, PA, USA).
For processing in hematoxylin and eosin (H&E), the frozen tumors were cut into 7μm thick sections, thaw mounted onto positively charged slides. Slides were rehydrated to water through a graded ethanol series (100% EtOH twice for 2 minutes each, 95% EtOH for 3 minutes, 70% EtOH for 1 minute, and water twice for 1 minute each). In the staining process, the slide was placed in hematoxylin for 1 minute and 30 seconds, rinsed in water for 15 seconds then 1 minute, placed into a defining solution for 45 seconds, water for 1 minute, bluing solution for 1 minute, water for 1 minute, 80% EtOH for 1 minute, Eosin for 1 minute, and 100% EtOH three times for 1 minute each. The slide was then mounted in Vectashield with DAPI as a counterstain (Vector Laboratories) and examined on a Leica DM2500 microscope and Retiga 4000R camera mount, using Q imaging software (Leica Microsystems, Morrisville, NC, USA).

Quantitative analysis of tumor images
To quantify GFP intensity in data images of tumor central sections, we used ImageJ to measure the total area of GFP pixels (mean, gray value, and standard deviation) under a limited threshold of approximately 99.5%. Two to five images per central section were averaged, per tumor. The average GFP pixels 2 per image for uninfected tissue images was subtracted from mean values for rAAV infected tissue images. GFP-expressing cells were auto picked and counted using ImageJ software [37]. Data images were converted to 8-bit and a Gaussian Blur (sigma = 3) was applied. The Bernsen Auto Local Threshold counting method (using the following parameters: radius = 8 pixels, parameter_1 = 30) was used [38,39]. GFP cells larger than 155 pixel 2 , with a circularity between 0.4 and 1.0, and DAPI-stained nuclei larger than 95 pixel 2 with a circularity between 0.1 and 1.0, were counted.

Statistical analysis
Values reported are means, and error bars represent the standard error of the mean (SEM). Pairwise differences between means were evaluated using Student's T-test, and resulting p-values are reported. Two-tailed T-tests were used, except when comparing treatments that differed in dose or treatment duration, for which one-tailed T-tests were used. One-way ANOVA was used for multiple comparisons followed by post hoc pairwise T-tests. T-test p-values < 0.05 were considered statistically significant and are indicated in the figures with asterisks using the following convention: � = < 0.05, �� = < 0.01, ��� = < 0.001. Circled asterisks follow the same convention, but indicate both significant and Bonferroni corrected p-values (p c -values), corrected for multiple comparisons. Statistical analyses were done using Microsoft Excel V. 16.16.15 (2018) and GraphPad Prism V.7 software (La Jolla, California, USA).

Calcitonin promoter-mediated luciferase gene expression
As a step toward developing a potential novel therapy vector with tissue specific expression in MTC cells, we utilized key elements of the calcitonin promoter. We modified the pGL3 firefly luciferase gene reporter plasmid to include the full-length pC/CGRP (Fig 1). Using transient transfection with pGL3 plasmids, we tested for luciferase gene expression in MTC and non-MTC human cell lines. Co-transfection experiments of pGL3-pC/CGRP with renilla luciferase gene control vectors pRL-SV40 were used to measure the relative firefly luciferase activity (RFLA) in human cell lines. The MTC-derived cell line TT transfected with pGL3-pC/CGRP yielded a RFLA (19.5% +/-7.5 of Renilla luciferase activity), which was significantly (p c < 0.05) higher than nine non-MTC cell lines tested, including cell lines derived from human follicular thyroid tissue, K1 (2.48% +/-0.78) and Nthy-ori-3.1 (1.4% +/-1.95) (Fig 2A).
These data show that pC/CGRP confers MTC cell-specific gene expression, and the pM +104 mini promoter improves that specificity.

In vitro transgene expression of rAAV-pM+104-GFP vectors in the MTC TT cell line
Next, we identified which rAAV serotypes show high levels of transgene expression in TT cells. First, rAAV serotypes 1-6, each with the chicken beta promoter and cytomegalovirus enhancer control region (pCBA) upstream of the GFP gene, were tested to determine which serotype best promoted transgene expression in MTC TT cell line. AAV2 was the only serotype to give GFP expression significantly higher (p < 0.01) using higher vector titer (10,000 vg/cell) than with infections at lower titer (2,000 vg/cell) (Fig 3). The rAAV2-pCBA-GFP vector infected at high titer (10,000 vg/cell) produced significantly (p c < 0.05) higher RFLA than the other serotypes tested (Fig 3). Thus, the rAAV2 capsid form was selected for use with pM+-104-GFP vector genomes.

In vivo transgene expression of rAAV-pM+104-GFP vectors in MTC xenograft tumors
The rAAV2-GFP vectors were next tested in a TT xenograft mouse model system to determine if the MTC could be targeted in vivo. Tumors were induced through subcutaneous injection of TT cells into the dorsal flank of six-week old male and female NOD.Cg-Prkdc SCID Il2rgtm1Wjl/SzJ (NSG) mice. About 8 weeks after tumor initiation, tumors reached about 1 cm in diameter and were infected via direct injection. Tumors infected with 2 x 10 10 vg/tumor of rAAV2-pM+104-GFP for 48 hours. Infected tumor tissue showed normal MTC tumor morphology and vascularization, calcitonin staining (a hallmark for C-cells and MTC), and GFP expression (Fig 5). GFP expression was quantified as relative fluorescence (GFP pixels 2 ) in TT cells treated using an rAAV2-pCBA-GFP vector (a non-tissue specific vector) with capsid serotypes one through six. Two infection dose levels were tested, 2,000 vg/cell and 10,000 vg/cell. Asterisks indicate a significant ( �� = p < 0.01) difference in GFP expression, when comparing the two dose levels for AAV serotype 2, using a one-tailed T-test. The circled asterisk indicates a significant ( � = p c < 0.05) increase in GFP expression at the higher dose level (10,000 vg/cell), when comparing AAV2 with to all other serotypes tested, using pairwise one-tailed T-tests, following ANOVA. https://doi.org/10.1371/journal.pone.0228005.g003 Improved tissue-specific targeting of medullary thyroid carcinoma using AAV To compare pM+104 to pCBA for GFP expression in our TT xenograft mouse model, 1-cm tumors were injected with 2 x 10 10 vg of either rAAV2-pM+104-GFP or rAAV2-pCBA-GFP, and then tumors were processed for GFP expression analysis 48 hours later. Tumors injected with rAAV2-pM+104-GFP showed significantly (p = 0.018) higher GFP intensity (1835.2 +/-86.4 pixels 2 ) than tumors injected with rAAV2-pCBA-GFP (1284.4 +/-282.3 pixels 2 ) (Fig 6A). Tumors treated with pM+104-GFP vector were found to contain 22.9 times more (p = 0.0097) vRNA molecules per vector genome detected than pCBA vector-treated tumor tissue (0.3149 +/-0.2642, and 0.0183 +/-0.0309) (Fig 6B). Vector nucleic acid was not detected off-target in the liver tissue harvested from vector treated animals.  HeLa, and HEK293) infected at an MOI of 10,000 vg/cell with the non-tissue specific control vector rAAV2-pCBA-GFP, and with test vector rAAV2-pM+104-GFP were analyzed. GFP cells were automatically counted using ImageJ, and the two vectors were compared in a given cell type, using two-tailed T-tests. Asterisk indicates significant (p = 0.0126) difference between the non-tissue specific control vector (rAAV2-pCBA-GFP) and the test vector (rAAV2-pM+104-GFP) in TT cells. Asterisk in a circle indicates a significant ( � = p c < 0.05) difference between TT and all cell types tested, in the number of GFP cells detected following infection with rAAV2-pM+104-GFP. Comparisons were made using pairwise two-tailed T-tests, following ANOVA. These data demonstrate that rAAV2-pM+104 vectors can promote transgene expression in MTC xenograft tumors, and support the hypothesis that pM+104 promotes higher levels of gfp gene and GFP protein expression in MTC TT tumor cells than with the ubiquitously expressed positive control vector, rAAV-pCBA.    Improved tissue-specific targeting of medullary thyroid carcinoma using AAV following vector injection. Transduction efficiency was measured by counting detectible GFPexpressing tumor cells (green cells in Fig 7). GFP cell counts are reported as percentages of DAPI stained nuclei (Fig 8).
These data show dose-and infection-duration dependencies on transduction efficiency with rAAV2-pM+104 vectors in MTC tumor cells. The greatest number of GFP cells were detected with a 1 x 10 11 vg dose for 72 hours (52.6% +/-14.2), suggesting that even higher dosages and longer infection durations may improve transduction in tumors.

Discussion
We report a novel rAAV vector (rAAV2-pM+104-GFP) with increased transgene expression in the MTC cell line TT over the non-MTC cell lines tested, demonstrating tissue specification. We are able to use and rAAV2-GFP vector with a miniature (~765 bp) modified calcitonin promoter/enhancer construct (pM+104) to give strong and specific expression in MTC cells in vitro and in vivo. The small size of this modified calcitonin promoter, compared to the full calcitonin promoter region, facilitates packaging of larger therapeutic genes in an rAAV vector.
We predict that rAAV2 serotype vectors showed higher transduction efficiency in TT cells over all other serotypes tested (AAV1-6) due to the cell-surface receptors displayed on TT cells. For example, the combination of heparan sulfate proteoglycan (HSPG) as a primary receptor, with αVβ5 integrin as a co-receptor is unique for AAV2 cell entry and transduction, not shared by the other serotypes tested [40] In a related study (data not shown), we ablated the known rAAV2 heparin binding site [41,42] through capsid mutagenesis and abolished rAAV-pM+104-GFP vector transduction in TT cells.
In these studies, we have demonstrated robust dose and duration-dependent GFP expression efficiency with rAAV-pM+104-GFP vector in MTC tumor xenografts, revealing the potential for further optimization of vector administration to improve the expression of therapeutic genes in tumor tissue. We observed over 52.6% of tumor cells expressing GFP in rAAV-pM+104-GFP-treated (1 x 10 11 vg per tumor) tumor tissue at 72 hours. This relatively low percentage of GFP cell, though, may be due to the fact that MTC tumor cells can be multinucleated, and cells that express lower levels of GFP, below our threshold of detection, are not counted.
A previous report examining rAAV transduction in TT cells showed 40% infection efficiency in vitro, but with adenovirus type 5 (Ad-5) co-infection [21]. In our studies we administer rAAV alone (without the rAAV infectivity boost that comes with Ad co-infection), and in vitro and in vivo, thus demonstrating improved infectivity of the vector with better clinical relevance (as Ad is not approved for use in gene therapy), and we have expanded the model into animal studies.
Although rAAV2 has been introduced in the clinic, and is a favorable serotype for preclinical studies in mice, it is not a favored serotype to treat diseases in regions of the body that are outside of the central nervous system (CNS), and that do not have immune privilege. Strategies have been proposed to overcome rAAV2 neutralizing antibodies (Nab), which can limit the transduction efficiency of therapeutic vectors, including using high vector doses, pre-dosing with empty capsids to saturate the Nab, and suppressing the B cell response prior to drug administration [43]. Other rAAV serotypes and vector modifications will need to be tested for transducing MTC cells in humans. Furthermore, additional pre-clinical studies are needed using the mouse xenograft MTC model described in this report, but with systemic (tail vein) vector administration. Off-target infection using the GFP vectors described in this study, would be detected as GFP expression in non-MTC cells, and non-tumor tissues. Such studies will allow us to determine if vectors can target MTC cells in vivo and to assess vectors for offtarget infection, as AAV2 is known to infect neurons, which express C/CGRP at low levels. AAV capsid peptide inserts, using peptides know to target MTC, may be useful to further narrow vector specificity to MTC [44,45].
Multiple strategies for gene therapy have been proposed for treatment of thyroid carcinomas including targeting of specific pathways, reintroduction of the sodium iodine symporter, immune modulation, and gene-directed enzyme/prodrug therapy (GDEPT) [46]. As the RET gene is commonly mutated in both hereditary and sporadic MTCs, drug treatment inhibiting the RET pathway has been used [47][48][49]. Improved, specific methods of inhibiting RET could be delivered using rAAV vectors. Many GDEPT methods not only induce cell death in targeted cells, but result in bystander killing of surrounding cells due to the release of toxic metabolites [20]. The rAAV2-pM+104 vector reported here transduced up to 52.6% of TT cells in vivo, thus a GDEPT method, as well as multiple injections per tumor and increased vector dosage may be beneficial in targeting tumors without 100% transduction efficiency.
In conclusion, we report a novel rAAV vector for in vivo transgene expression in MTC. Strong and specific expression was achieved in MTC cells due to the use of the modified calcitonin promoter used in our rAAV2 vectors, and continued expression of calcitonin in medullary carcinomas was confirmed [50,51]. We speculate that a gene therapy that utilizes the calcitonin promoter has the potential to treat MTC across clinical subtypes. With further modification, and the introduction of therapeutic genes, rAAV vectors could emerge as a useful tool in targeting and treating of MTC.