Mycoplasma-free human ccRCC 786-O and human prostate adenocarcinoma LNCaP cells (ATCC, Manassas, VA, USA) were propagated in RPMI-1640 medium supplemented with 10% FBS (Invitrogen, Burlington ON, Canada). The 786-O cell identity was verified by STR analysis (ATCC). Murine RCC RAG cells and human glioblastoma LN-18 cells (ATCC) were maintained in Eagle's MEM and Dulbecco's MEM, respectively, supplemented with 10% FBS.

The shRNA vector for human CA9 (CA9 shRNA) and the non-effective negative scrambled control were purchased from Origene (Rockville, MD, USA). Transfection of human ccRCC 786-O cells was performed using 12% Fugene (Promega, Madison, WI, USA). Cells stably transfected with shCA9 or the scrambled control were selected with 1.0 µg/ml puromycin (Sigma-Aldrich, Oakville, ON, Canada).

RAG, 786-O, and LN-18 cells (250 per well) were seeded onto 6-well plates in 3 ml of medium and allowed to adhere by incubation at 37°C and 5% CO₂ for 4 h. For LNCaP, 1000 cells were seeded per well and allowed to adhere for 24 h.

For the AEBS treatment, a stock solution of 4-(2-aminoethyl)benzene sulfonamide (AEBS, 33 mM, Sigma-Aldrich) was prepared fresh with H₂O and filter-sterilized using a 0.2 µm syringe filter. AEBS at concentrations ranging from 3.3 µM to 3300 µM was added to 100 µl of H₂O to 3 ml media. Solvent controls were also included. The medium was aspirated and replaced with 3 ml of the appropriate drug solutions in media in duplicate wells. After 24 h of incubation, the media were aspirated, and 3 ml of fresh medium was added to each well.

Ionizing radiation (IR) of cells was performed using a Varian Linear Accelerator (LINAC) generating six MV X-rays (Varian Medical Systems, Inc., Palo Alto, CA, USA). Cells plated in duplicate 6-well plates were irradiated at a distance of 100 cm in a 16 cm by 20 cm field. A 19-mm-thick acrylamide sheet was placed on the plates as a build-up region. Thermoluminescent dosimeters were used to measure and calibrate the dose. The cells received from 1 to 8 Gy of 6 MV X-ray radiation.

Following treatment, the cells were incubated for an additional 6 days (786-O), 7 days (RAG), or 12 days (LNCaP and LN-18), after which the medium was aspirated and the colonies were stained with crystal violet (0.25% in 95% ethanol) for 10 min. Colonies of 50 cells or more were counted. Survival was expressed as a percentage of the corresponding untreated controls, and IC₅₀ values were calculated using CalcuSyn software version 1.2 (Biosoft, Cambridge, UK). Experiments were repeated three times.

Confluent RCC cells in 6-well plates were washed twice with 3 ml of PBS. One ml of 0.9% saline (adjusted to pH 8.0 with NaOH) containing 0.15 mg/ml phenol red was added to each well. Deionized water (100 µl) with or without AEBS at various concentrations was added to the wells. Starting 5 sec after the addition of saline to the cells, the absorbance at 565 nm was measured using a Powerwave HT spectrophotometer (BioTek, Winooski, VT, USA) at 1 sec intervals for 20 sec. The relative absorbance of a particular well throughout the 20 sec was determined as a percent of the no cell control average at that time interval (A/A _{no cells}).

Cultured cells were lysed using RIPA lysis buffer. Lysate (40 µg) was resolved on a 10% SDS-PAGE gel and transferred onto nitrocellulose membrane. Primary antibodies used were CA9 (1:1,000 Epitomics, Burlingame, CA, USA) and β-actin (1:2,000, Sigma-Aldrich). Secondary HRP-conjugated antibody (1:200, Dako, Carpinteria, CA, USA) was used in conjunction with chemiluminescence detection.

All protocols for animal studies were reviewed and approved by the institutional Animal Research Ethics Board (AUP# 12-09-37). Per group, 7–10 female inbred nude (Balb/c nu/nu) mice (Charles River, St. Constant, QC, Canada) 5 weeks of age were used. 786-O parental, shCA9 or scrambled control 786-O cells [1–3×10⁶ in 50% (v/v) Matrigel] were injected subcutaneously into the right flank of each mouse. Tumor size was measured every three days using Vernier calipers, and the tumor volume was determined using the formula π/6(length x width x height) until the largest tumor reached 400 mm³. The mice were sacrificed 7–12 weeks after tumor cell injection, when the tumors were dissected, weighed, fixed in formalin and embedded in paraffin.

For mice in the IR group, 21–27 days post injection, when tumors were palpable, the animals were anaesthetized using isoflurane and positioned in sterile, acrylamide cylinders connected to a portable anaesthetic machine. Cylinders were transported to the treatment area, where they were positioned at a distance of 100 cm to the source and irradiated with 6 Gy in a 2 cm by 2 cm field using a Varian Linear Accelerator generating 6 MV X-rays. A 5-mm-thick sheet of superflab bolus material served as a build-up region. Animals in non-irradiated control groups were anaesthetized for a similar time period.

Animals in the AEBS-treatment groups received 50 or 200 µg/ml AEBS in the drinking water supplied fresh every two days starting two days before IR. No adverse effects of the treatment were observed.

The serum levels of VEGF were determined using a mouse VEGF ELISA kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's instructions on a BMG Labtech SpectroStar Nano multi-well plate reader. Immunostaining of 4 µm-thick tumor xenograft sections for CD31 and subsequent image analysis was performed as previously described (7) resulting in the microvessel density expressed as endothelial length (in µm) per mm².

Trifluoroacetic acid (2 µl) and methanol (500 µl) were added to 100 µl mouse serum, vortexed for 10 sec and centrifuged at 5000 rpm for 10 min. Resulting supernatant (300 µl) was removed and blown down to dryness with nitrogen. The sample was reconstituted in 200 µl of methanol/water (1:1) containing 25 µg/ml phenylalanine (internal standard) and filtered using a 13-mm syringe filter (0.2 µm GHP membrane). The sample (5 µl) was run on the LC-MS (Agilent 6340 Ion Trap coupled to an Agilent 1200 HPLC, Agilent Technologies Inc, Mississauga, ON, Canada) at the McMaster regional centre for mass spectrometry. Analysis was performed using multiple reaction monitoring on AEBS and phenylalanine with the transition at 201-184 (m/z) and 166-120 (m/z), respectively. Control serum was spiked with AEBS at 1–16 µg/ml.

Values are expressed as the mean ± the standard error of the mean. Where appropriate, results are presented with 95% confidence intervals (CI). Dependent on whether the data were normally distributed or not, parametric (Student's t-test) or nonparametric methods (Mann-Whitney U-test) were used with a p-value <0.05 indicative of statistical significance.

Protein expression of CA9 was determined in the lysates of human prostate adenocarcinoma LNCaP, human ccRCC 786-O, murine RCC RAG and immortalized human embryonic kidney HEK-293 cells by western blot analysis (Fig. 1A). CA9 was detectable in the 786-O and RAG cells, but not in the LNCaP or HEK-293 cells. Clonogenic survival experiments demonstrated that RAG cells were more sensitive to IR than 786-O cells (p<0.05), whereas LNCaP cells were the most sensitive among the cell lines tested (Fig. 1B). Human glioblastoma LN-18 cells exhibited similar tolerance to IR as the RAG cells. The 786-O cells displayed significantly decreased survival at a dose of 2 Gy and above (p<0.001), whereas LNCaP cells displayed significantly decreased survival at all radiation doses (p<0.001). The calculated IC₅₀ values for each cell line are presented in Table I.

To investigate the significance of CA9 expression on RCC radiosensitivity, human ccRCC 786-O subclones stably expressing shRNA specific for CA9 were generated. These shCA9 cells showed 92% knockdown of CA9 protein expression compared to the respective control cells (scrambled shRNA) (Fig. 2A). Clonogenic survival of the 786-O scrambled control and shCA9 cells was determined after IR with increasing doses of 1–8 Gy and clearly demonstrated that knockdown of CA9 confers sensitivity to IR (p<0.001). The IC₅₀ value decreased by >50%, from 3.64 Gy (95% CI: 3.27–4.05) in the scrambled control cells to 1.81 Gy (95% CI: 1.44–2.27) in the shCA9 cells (Fig. 2B). To further demonstrate the effect of CA9 knockdown in vitro, the activity of the CA9 enzyme was measured using phenol red as an indicator. In comparison to the scrambled control cells, shCA9 cells had significantly decreased acidification capacity (p<0.001) and experienced 51% of the absorbance change observed in the scrambled control cells (Fig. 2C).

The effect of CA9 knockdown on the in vivo growth after subcutaneous injection of shCA9 or scrambled control cells (1×10⁶) into nude mice (n=7/group) was determined. A tumor take rate of 64% was achieved. The tumor volume from mice injected with the shCA9 cells was not significantly different from mice injected with cells transfected with the scrambled control shRNA (Fig. 3A and B). However, when the subcutaneous tumors were also irradiated 21 days after cell injection, we observed that IR of the shCA9-transfected 786-O cells led to decreased tumor growth (p<0.001) and a 78.7% decrease in tumor volume at sacrifice (Fig. 3B). IR of the scrambled control-injected mice led to a 20.7% reduction in tumor volume after 12 weeks (Fig. 3A). Western blot analysis of the tumor homogenates showed that CA9 remained reduced in the shCA9-injected mice (53%) at sacrifice, 12 weeks after tumor cell injection (Fig. 3C).

AEBS is a known inhibitor of CA9's enzymatic reaction with a Ki of 33 nM (8). However, to the best of our knowledge, AEBS has not been used previously to inhibit CA9 in vivo, in contrast to acetazolamide, a similar sulfonamide. In the presence of 33 µM AEBS, the human ccRCC 786-O cells exhibited a significant decrease in clonogenic survival after IR when compared to the untreated control (Fig. 4A, p<0.001). Similarly, the more radiation-sensitive murine RAG cells also exhibited a decrease in clonogenic survival after IR when treated with the same concentration of AEBS for 24 h (Fig. 4B, p=0.018). To further demonstrate the effectiveness of AEBS in vitro, the acidification of the extracellular environment of 786-O cells was determined (Fig. 4C). In comparison to the untreated control 786-O cells, incubation with either 33 µM or 33 nM AEBS caused significantly less acidification (p<0.001), leading to a decrease of 70 and 36%, respectively. AEBS was not cytotoxic; by clonogenic survival the IC₅₀ values for AEBS amounted to 1,360 and >5,000 µM in the 786-O and RAG cells, respectively (data not shown).

To investigate whether a hypofractionated regimen of radiation delivery could increase the survival difference between CA9-inhibited and non-inhibited RCC cells, 786-O cells expressing scrambled control shRNA or shCA9 or with and without the addition of AEBS to the medium were subjected to either one treatment of 6 Gy or three treatments of 2 Gy for three consecutive days. Cells receiving AEBS or expressing shCA9 had significantly reduced survival (p<0.001) after receiving a single dose treatment of 6 Gy compared to the fractionated radiation treatment of three doses of 2 Gy (Fig. 5).

We determined the effect of CA9 inhibition on the in vivo growth after subcutaneous injection of 786-O cells (3×10⁶) into nude mice (n=20/group). A tumor take rate of 85% was achieved. Mice were treated with either AEBS starting 25 days after cell injections or received IR on day 27 and were sacrificed 24 days later. One untreated control group was included and one group of mice received both IR and AEBS. The tumor volume in mice treated with AEBS was significantly smaller than that in the control mice (Fig. 6A; p=0.03 ANOVA). When the subcutaneous tumors were also irradiated 27 days after cell injection, we observed that the combination treatment led to decreased tumor growth compared to either AEBS or IR alone (p<0.0005 and p=0.04, respectively). At sacrifice, this led to an average increase of a mere 12% in tumor volume from the time of IR on day 27 to sacrifice on day 51 in the combination group (Fig. 6B). In comparison, the tumor volumes of the untreated control mice had increased in size by almost 3-fold. Protein levels of CA9 in the mouse tumors did not differ between treatment groups (Fig. 6C). Mass spectrometry showed that measurable and CA9-inhibitory levels of AEBS were achieved in mouse serum (Fig. 6D). Levels reached 26.5 µM 18 h after starting the treatment and remained similar throughout the experiment (data not shown).

Using immunohistochemical staining of the subcutaneous tumors for CD31, an endothelial marker, we also demonstrated a decrease in the microvessel density in the irradiation-treated mice, with or without AEBS, with a linear microvessel length decrease of 55.6 and 64.5% compared to the control (Fig. 7A and B, p=0.04 and p=0.03). Serum VEGF levels at endpoint were significantly decreased in the AEBS-treated and irradiated mice compared to the control mice (Fig. 7C).