Impact of myo‐inositol trispyrophosphate (ITPP) on tumour oxygenation and response to irradiation in rodent tumour models

Abstract Tumour hypoxia is a well‐established factor of resistance in radiation therapy (RT). Myo‐inositol trispyrophosphate (ITPP) is an allosteric effector that reduces the oxygen‐binding affinity of haemoglobin and facilitates the release of oxygen by red blood cells. We investigated herein the oxygenation effect of ITPP in six tumour models and its radiosensitizing effect in two of these models. The evolution of tumour pO2 upon ITPP administration was monitored on six models using 1.2 GHz Electron Paramagnetic Resonance (EPR) oximetry. The effect of ITPP on tumour perfusion was assessed by Hoechst staining and the oxygen consumption rate (OCR) in vitro was measured using 9.5 GHz EPR. The therapeutic effect of ITPP with and without RT was evaluated on rhabdomyosarcoma and 9L‐glioma rat models. ITPP enhanced tumour oxygenation in six models. The administration of 2 g/kg ITPP once daily for 2 days led to a tumour reoxygenation for at least 4 days. ITPP reduced the OCR in six cell lines but had no effect on tumour perfusion when tested on 9L‐gliomas. ITPP plus RT did not improve the outcome in rhabdomyosarcomas. In 9L‐gliomas, some of tumours receiving the combined treatment were cured while other tumours did not benefit from the treatment. ITPP increased oxygenation in six tumour models. A decrease in OCR could contribute to the decrease in tumour hypoxia. The association of RT with ITPP was beneficial for a few 9L‐gliomas but was absent in the rhabdomyosarcomas.

and stimulate the selection for clonogenic cells with increased hypoxia tolerance. The expansion of these cell clones can, in turn, aggravate tumour hypoxia, thereby establishing a vicious circle of increasing hypoxia and subsequent malignant progression. Translated in the clinic, this vicious circle leads to more local recurrences, locoregional spread, distant metastases and greater resistance to cancer therapies. 7 Particularly, tumour oxygen status has been shown to be important for outcome following radiation; cancer patients with hypoxic tumours have been reported to be at higher risk for radiotherapy failure. 8,9 Considering the compelling link between tumour hypoxia and treatment outcome, efforts have been made to develop effective hypoxia-targeted therapies. [8][9][10][11][12][13][14] Among these, using hypoxia modifier that can improve tumour oxygenation during irradiation represents one of the most attractive strategies. Recently, ITPP (myo-inositol trispyrophosphate), has been suggested to exert such activity. This compound has intrinsic anti-cancer properties as it has demonstrated therapeutic efficacy in various animal models when used alone or combined with chemotherapies. [15][16][17][18][19] ITPP is an allosteric effector that reduces the oxygen-binding affinity of haemoglobin and may thus facilitate the oxygen release by red blood cells. 20,21 This effect as well as a potential vascular normalization through the down-regulation of HIFs/VEGF may alleviate tumour hypoxia. 15,17,18 Starting from these previous studies, the present study was designed to answer the following questions: (i) is the expected increase in

| Tumour models and ITPP treatment
Four mouse tumour models (mouse fibrosarcoma FSaII implanted in C3H mice, mouse mammary tumour NT2 in FVb/Nrj mice, human breast cancer MDA-MB-231 in NMRI nude mice and human cervix squamous cell carcinoma SiHa in NMRI nude mice) and two rat tumour models (rat 9L-glioma in Fischer F344 rats and rat rhabdomyosarcoma in WAG/Rij rats) were used. The origin of cell lines and animals were reported elsewhere. [22][23][24][25] Tumours were inoculated subcutaneously in the thigh of animals according to the protocols described previously. [22][23][24][25] Experiments were performed when tumours reached a diameter of 6-8 mm (in mouse models) or 14-16 mm (in rat models). ITPP (kindly provided by Normoxys Inc) was injected intraperitoneally. Doses ranged from 0.5 to 4.0 g/kg. ITPP solution was prepared by dissolving the compound in injectable water to the target concentration and adjusted to pH = 7 by using a small volume of 0.1 mol/L NaOH. The experimental design included various dosage of ITPP and various treatment schedules to explore which regimen could offer the best oxygenation effect.

| Tumour oxygenation
An L-bandEPR spectrometer (Magnettech, Berlin, Germany) operating at 1.2 GHz was used to evaluate the dynamic change in tumour oxygenation upon ITPP administration. Charcoal suspension (CX0670-1; EM Science, Gibbstown, NJ, USA; 100 mg/mL), used as the oxygen sensor, was introduced intratumourally (about 60 µL for a mouse tumour and 200 µL for a rat tumour) 1 day before the experiment. The charcoal is dispersed over the whole tumour. During EPR recording, animals were anaesthetized with 2% isoflurane in air and their body temperature was maintained at 37 ± 1°C using a circulating warm water system. This anaesthesia regimen was previously demonstrated not to disturb the haemodynamics in rodents. 26 The pO 2 measurements were carried out 15 minutes after the induction of the anaesthesia. The linewidth of the first-derivative EPR spectrum that was the average of five 1-minute scan accumulations was then converted to pO 2 using a calibration curve. 27

| Tumour perfusion
Hoechst 33342 (Sigma) staining was used to assess 9L-glioma perfusion 1 day after completion of the treatment (2 g/kg ITPP once daily for 2 days). Rats were killed exactly 2 minutes after intravenous injection of Hoechst solution (15 mg/kg in saline). Tumour fragments were rapidly excised, embedded in optimal cutting temperature compound and frozen in liquid nitrogen-cooled isopentane. Frozen sections of 5 μm thickness were photographed using a Zeiss Mirax fluorescence microscope and images were analysed using Frida software. The percentage of tumour perfusion was calculated as the ratio of Hoechst-positive area to the total area of tumour sections (no necrosis was histologically detected at this stage of tumour development). 28

| Tumour cell oxygen consumption rate
The impact of ITPP on OCR was assessed on six cell lines (FSaII, SiHa, MDA-MB-231, NT2, 9L-glioma and rhabdomyosarcoma) using a Bruker EMX X-band EPR spectrometer operating at 9.5 GHz and

| Irradiation
The regimen of 2 g/kg ITPP once daily for 2 days was used. Irradiation was performed 2 hours after the second administration of ITPP (time at which tumour oxygenation was shown to be highest). Two rat models (9L-glioma and rhabdomyosarcoma) were employed; rats were randomly divided into four groups: vehicle, ITPP, RT + vehicle and RT + ITPP. Single dose of irradiation was delivered by a 137 Cs irradiator IBL-637 (Oris, France), 20 Gy for rhabdomyosarcoma and 30 Gy for 9L-glioma. Animals anaesthetized with isoflurane (2% in air) were placed on a plexiglass and protected from the beam through a lead block of 3 cm thickness while the tumours were exposed through a hole 25 mm in diameter. The animals were turned midway through the exposure time to enhance the uniformity of dose distribution. Irradiation doses were selected on the basis of their respective radiation sensitivity to ensure a significant growth delay in the irradiation group compared to the untreated one. Treatment effect was analysed based on tumour growth delay assay.
Tumours were measured on the starting day of treatment to determine the initial size and then at least twice a week until the end-point (time at which a tumour doubled its initial diameter).
Clonogenic assays were also performed to evaluate the radiosensitization effect of ITPP on rhadomyosarcoma and 9L-glioma cell lines (Supplementary Materials and Methods).

| Statistical analysis
All results were expressed as mean ± SEM. Differences between groups were analysed using t test or Mann-Whitney test when data were not normally distributed. Log-rank test was used to compare Kaplan-Meier curves. P < 0.05 was considered statistically significant for all tests.

| Impact of ITPP on tumour oxygenation and treatment schedule optimization
A first screening of pO 2 evolution upon ITPP administration was conducted on various tumour models to assess the effect of this compound on tumour oxygenation. The first test with 9L-glioma and FSaII showed a significant increase in tumour pO 2 , quickly within 2 hours after the injection of a single dose of 2 g/kg ITPP (P = 0.0002 and P = 0.0014, respectively) ( Figure 1A,B). The oxygenation increase observed in 9L-glioma (83.6%) was much larger than in FSaII (29.6%). This effect was maintained for 1 day before slowly returning to baseline levels. In the next model, rhabdomyosarcoma, another administration of 2 g/kg ITPP was added on the following day that further enhanced the effect ( Figure 1C). The second increase in pO 2 after the second dose of ITPP was also found in SiHa model ( Figure 1D). Given the moderate oxygenation effect on the first four models, we applied a doubled dose of ITPP (4 g/kg) to mice bearing NT2 tumours. However, only a limited increase in pO 2 (from 6 to maximum 12 mm Hg) was obtained ( Figure 1E). Similarly, we observed the accumulative effect in MDA-MB-231 model upon the second injection; however, 2 hours after the first injection, the group of 4 g/kg had no advantage over that of 2 g/kg ( Figure 1F).
Rhabdomyosarcoma was then used to optimize the treatment schedule of ITPP. To investigate if lower doses of ITPP could induce a similar increase in oxygenation, various doses of ITPP ranging from 0.5 to 2 g/kg were injected once a day within 4 consecutive days.
The results showed an obvious relationship between the dose and the response (Figure 2A). Dose of 2 g/kg offered the optimal effect; however, the third and fourth doses did not induce any additional effect. The effect of once-daily and twice-daily treatment was then compared. In the twice-daily regimen, the second dose of the day was given at 6 hours after the first one, the treatment was lasting for 3 days. So overall, the animals of this group received six doses of 2 g/kg in 3 days; whereas, the animals of the once-daily treatment group received four doses of 2 g/kg in 4 days. No difference between two groups was observed ( Figure 2B). Hence, 2 g/kg once daily for 2 days was considered as the optimal regimen of ITPP treatment. This optimal schedule was finally verified on 9L-glioma.
The result showed the most elevated tumour oxygenation at 2 hours after the second injection of ITPP as expected ( Figure 2C).

| Contributing factors to increase in oxygenation: Effect of ITPP on tumour perfusion and OCR
Tumour perfusion 1 day following the optimal treatment (2 g/kg of ITPP once daily for 2 days) was assessed on 9L-glioma ( Figure 3).
Areas of perfusion, corresponding to areas stained by Hoechst 33342, were not significantly different between the groups with and without ITPP treatment. The impact of this compound on OCR of cancer cells was then studied. As shown in Figure 4, exposure to 10 mmol/L ITPP in 2 hours significantly inhibited OCR in all six cell lines. No further benefit was found when expanding the incubation time to 6 hours (data not shown). Interestingly, the PI3K inhibitor LY294002 also induced a comparable effect at the similar timing.

| Radiosensitization effect of ITPP
To investigate the radiosensitization effect of ITPP, we combined the optimal treatment schedule of ITPP (2 g/kg once daily for 2 days) with irradiation of 20 and 30 Gy on rhabdomyosarcoma and 9Lglioma, respectively. The results of tumour growth delay assay are presented on Figure 5. In both models, ITPP monotherapy did not affect tumour growth, as the times to reach the end-point were

| DISCUSSION
Myo-inositol trispyrophosphate has demonstrated therapeutic efficacy in a wide range of animal models [15][16][17][18][19] and shown safety in a Phase I human clinical trial (http://normoxys.com/clinical-trial-re sults/). Using the OxyLite system, Raykov et al found that a single ITPP injection induced an increase in partial pressure of oxygen for almost 1 week in pancreatic tumour xenografts. 16 In another study, Kieda et al observed a similar effect in B16 melanoma model and 4T1 mammary tumour model. 15 Based on the previous results, we decided to explore the potential benefit when combining ITPP with radiotherapy. First, we assessed the effect of ITPP on tumour oxygenation in a larger panel of tumour models and tried to define the optimal dose/schedule regimen leading to a maximal increase in tumour oxygenation. For that purpose, we used EPR oximetry, 31,32 a non-invasive and highly sensitive technique that can provide the quantitative absolute values of pO 2 in vivo and is able to repeat the measurements at the same site over long periods of time. Our results obtained by EPR indicated that ITPP treatment quickly increased tumour oxygen levels in the six tested models and reached the maximum effect at 2 hours post-administration. Such elevated pO 2 was sustained for at least 1 day and then gradually decreased ( Figure 1). Interestingly, a cumulative effect was found when the second ITPP administration was given with a 24-hours interval, but additional administration did not lead to further increase in tumour pO 2 (Figure 2). It should be noted that the effect of ITPP was rather moderate compared to those previously reported. 15,16 This may be because of the difference in response of different tumour models but also because of the methods used to measure oxygen tension as the previous studies were carried out using OxyLite. OxyLite provides dynamic oxygen measurement at a single point inside tissue, whereas EPR reports pO 2 on a larger volume as the oxygen sensor is dispersed over the whole tumour ( Figure S2).
Regarding the mechanism, ITPP is an allosteric effector that reduces the oxygen-binding affinity of haemoglobin and thus facilitates the oxygen release by red blood cells. 20,21 In addition, long-   Interestingly, ITPP has been shown to inhibit PI3K, 15 and the inhibition of PI3K pathway has been proposed to reduce the cell OCR. 33,34 We suggest that the effect of ITTP on OCR could be comparable to other inhibitors of PI3K such as LY294002. We observed that both compounds induced similar effects on the OCR of six cell lines at the same timing. Our results demonstrated that ITPP was an inhibitor of tumour cell respiration and identified the inhibition of OCR as another contributing mechanism to the ITPP-induced increase in oxygenation. At this stage, the mechanism supporting the effect of ITTP on OCR remains unknown and the hypothesis of an effect mediated by an inhibition of the PI3K remains to be validated in future studies. Of note, it has been suggested theoretically 35 and F I G U R E 4 Impact of ITPP on oxygen consumption rate in vitro. *P < 0.05 when comparing the treated group with the control group (n = 3-8/group) F I G U R E 5 Effect of ITPP treatment (2 g/kg once daily for 2 days) on tumour growth of rhabdomyosarcoma and 9L-glioma. Top panel: Bar graphs showing the time for tumours to reach the end-point. Bottom panel: Kaplan-Meier curves showing the percentage of tumours that did not reach the end-point. For the cured tumours, the ending day of experiment was taken as the value of tumour growth time. "ns" = not significant. (n = 6-10/group) experimentally 36 that in order to alleviate tumour hypoxia, decreasing oxygen consumption should be more efficient than increasing oxygen delivery. Interestingly, in the study of Kelly et al, blockade of PI3K pathway was found to reduce OCR and to increase tumour oxygenation despite no change in overall perfusion. 33 This result was quite in line with what we observed in the present study. A possible limitation of our approach is that OCR was measured in vitro.
Although the reduced OCR of cancer cells exposed to ITPP was correlated to the increase in tumour oxygenation in terms of timing, in vitro results may not translate perfectly to in vivo system. 37 It will thus be very interesting in the future to assess the impact of ITPP on OCR in vivo. For this purpose, 17  Finally, the anti-tumour properties of ITPP were investigated on rhabdomyosarcoma and 9L-glioma whose radiosensitivity has been shown to be correlated with oxygenation level. [42][43][44] When used as a single therapy, ITPP showed no influence on tumour growth in these two models. This result was not in accordance with those observed previously on melanoma, hepatoma and pancreatic cancer where tumour growth was dramatically delayed thanks to long-term treatment with weekly ITPP. [15][16][17] However, differences in treatment schedules make the comparison of the data in the present paper with those published earlier very tenuous. We then combined ITPP treatment with irradiation to explore if the increase in oxygenation induced by ITPP could lead to a radiosensitization. We did not observe any benefit from the association ITPP + RT in rhabdomyosarcoma. In 9L-glioma model, a trend towards an increase in the response was found when combining ITPP with irradiation; however, the response was highly heterogeneous. Indeed, some 9L-gliomas were completely cured, whereas the other tumours did not respond any better. In comparison with our previous study, 45 carbogen breathing was able to radiosensitize a majority of 9L-gliomas and the degree of response was significantly correlated with oxygen level. Such correlation could not be found herein, suggesting that ITPP-induced increase in oxygenation may not be the determinant factor affecting tumour outcome. In fact, many factors and mechanisms may contribute to the ultimate efficacy of radiotherapy.
Besides oxygenation status, ability of repairing DNA damage and cancer cell repopulation following irradiation are also believed to be an important cause of treatment failure. [46][47][48] If ITPP actually works via a molecular pathway related to PI3K signaling, then the radiosensitizing effect of ITPP would be more complicated than just a simple decrease in hypoxia. PI3K pathway is a key regulator of various cellular functions from cell proliferation to cell survival and is implicated in all major mechanisms of radioresistance. 49,50 Several studies have indicated the strong involvement of PI3K pathway in repairing the radiation-induced DNA double-strand breaks through DNA-dependent protein kinase. [51][52][53] In non-small cell lung cancer, treatment using PI3K/Akt inhibitors could change the apoptotic potential of cancer cells and counteract cell survival, resulting in a better radiosensitivity. 54,55 However, this effect was found only in tumours and cell lines with high level of PI3K/Akt activation. Similarly, targeting PI3K could only promote radiation-induced apoptosis in breast cancer cell lines in which this pathway is overstimulated. 56 Regarding the therapeutic efficacy of ITPP, it should be emphasized that our study is not the first that reported the disappointing result of ITPP.
The recent work combining ITPP with radiotherapy on mice bearing GL261 57 and another one using long-term ITPP treatment on rats bearing RG2 glioblastoma 58 have both pointed out the failure of this compound. The fact that not all tumours could benefit from ITPP treatment suggests that the anti-cancer properties of ITPP may be based on a specific signaling which is not ubiquitously expressed. To verify this hypothesis, another study will be needed to identify the main pathway and key elements that are critical for driving the action of ITPP. To mimic more closely clinical irradiation protocols, it would be interesting to use fractionated irradiation instead of a single irradiation dose as used in the present study.
In summary, our data consistently demonstrated the increased tumour oxygenation in six animal models upon ITPP administration.
We also showed in a proof of concept experiment that the enhancement in oxygen level likely resulted from a decrease in oxygen consumption rather than an increase in oxygen perfusion at least at the early stage. The increase in tumour oxygenation induced by ITPP only partly radiosensitized one of the two investigated models. Taking our finding together with the previous reports from the literature, ITPP possesses to some extent potential characteristics that can be beneficial to cancer treatment. However, to take full advantage of its capacity and to move further into clinical trials, a complete picture on the mode of action of this compound is mandatory.
Future studies should focus on the underlying mechanism of ITPP and on how the oxygenation effect would be involved in the ultimate therapeutic efficacy of ITPP.

ACKNOWLEDG EMENTS
This work was funded by Normoxys Inc. (Boston, MA, USA). LBAT performed the experiments, analysed data and drafted the manuscript. AH and SD contributed the study design, data analysis and critical revision. BJ and TTCP participated in critical revision. BG supervised the study, contributed to study design and critical revision.

CONFLI CT OF INTEREST
The authors confirm that there are no conflicts of interest.