Imaging tumour pH and carbonic anhydrase activity using magnetic resonance imaging


Year:

Session type:

Ferdia Gallagher1, Mikko Kettunen1, Sam Day1, Rebekah McLaughlin1, De-En Hu1, Jan Ardenkjær-Larsen4, Rene in ‘t Zandt5, Pernille Jensen5, Magnus Karlsson5, Klaes Golman5, Mathilde Lerche5, Kevin Brindle1

1Cancer Research UK, Cambridge Research Institute, Cambridge, UK, 2Department of Biochemistry, University of Cambridge, Cambridge, UK, 3Department of Radiology, University of Cambridge and Addenbrooke's Hospital, Cambridge, UK, 4GE Healthcare, Amersham, UK, 5Imagnia AB, Malmo, Sweden

Abstract

Background

Tissue acid-base balance is tightly controlled: the main buffer is bicarbonate (HCO3-) which is interconverted into carbon dioxide (CO2) by the enzyme carbonic anhydrase (CA). Tumours have an acidic extracellular pH and CA expression can be induced by hypoxia. However, the spatial distribution of pH cannot currently be imaged in humans. Dynamic Nuclear Polarisation (DNP) is a technique for increasing the sensitivity of 13C-MRS and we show here that H13CO3- can be hyperpolarised using this method and then used to image tissue pH and determine CA activity in vivo.

Method

Hyperpolarised 13C-labelled bicarbonate was injected into mice bearing murine lymphoma tumours: pH was calculated using the Henderson-Hasselbalch equation from the concentrations of H13CO3- and 13CO2. This was validated using an established pH probe: 3-APP. CA activity was demonstrated by saturating the 13CO2 signal: the relative decrease in the H13CO3- signal is dependent on the amount of CA present.

Results

The pH of the lymphoma tumours, calculated using hyperpolarised H13CO3-, was 6.72 ±0.13, which agreed with the pH calculated using 3-APP. The spatial distribution of pH within the animals confirmed that the tumours were acidic. Saturation of 13CO2 showed that the reaction catalyzed by CA was ~8-fold faster than the CA-independent reaction.

Conclusion

We have demonstrated that pH can be imaged in vivo using hyperpolarised bicarbonate and that this system is dependent on CA activity. This technique exploits an endogenous buffer which is present at high concentration in humans and therefore could be used for clinical imaging in the future.

Background

Tissue acid-base balance is tightly controlled: the main buffer is bicarbonate (HCO3-) which is interconverted into carbon dioxide (CO2) by the enzyme carbonic anhydrase (CA). Tumours have an acidic extracellular pH and CA expression can be induced by hypoxia. However, the spatial distribution of pH cannot currently be imaged in humans. Dynamic Nuclear Polarisation (DNP) is a technique for increasing the sensitivity of 13C-MRS and we show here that H13CO3- can be hyperpolarised using this method and then used to image tissue pH and determine CA activity in vivo.

Method

Hyperpolarised 13C-labelled bicarbonate was injected into mice bearing murine lymphoma tumours: pH was calculated using the Henderson-Hasselbalch equation from the concentrations of H13CO3- and 13CO2. This was validated using an established pH probe: 3-APP. CA activity was demonstrated by saturating the 13CO2 signal: the relative decrease in the H13CO3- signal is dependent on the amount of CA present.

Results

The pH of the lymphoma tumours, calculated using hyperpolarised H13CO3-, was 6.72 ±0.13, which agreed with the pH calculated using 3-APP. The spatial distribution of pH within the animals confirmed that the tumours were acidic. Saturation of 13CO2 showed that the reaction catalyzed by CA was ~8-fold faster than the CA-independent reaction.

Conclusion

We have demonstrated that pH can be imaged in vivo using hyperpolarised bicarbonate and that this system is dependent on CA activity. This technique exploits an endogenous buffer which is present at high concentration in humans and therefore could be used for clinical imaging in the future.