Introduction

The main obstacle in application of phosphorus-containing MR contrast agents for in vivo measurements is high natural ³¹P signal from the surrounding tissue. We would like to address this problem by presenting a metal-free probe based on phosphorus-containing polymer of methacrylate-type (p(TMPC)), which is traceable by ³¹P MRI and MRS, because its signal can be reliably separated from the others related to biological phosphorus-containing compounds. This is possible due to the presence of phosphorothioate group, ensuring large chemical shift of the polymer signal in the ³¹P MR spectra from Larmor frequencies of the tissue. This proof of principle study is comparing two polymers with the same backbone, but one bearing a phosphoester group (p(MPC)), which is commonly present in biological phosphorus-containing compounds, and the other containing a phosphorothioate group (p(TMPC)), which, in turn, is extremely rare in living organisms¹. High phosphorus concentration in the p(TMPC) polymer together with the high frequency shift compared to the p(MPC) probe would enable imaging with sharp ³¹P signal boundary of our area of interest in the organism. We visualized ³¹P polymers at 4.7T scanner, using surface ¹H/³¹P custom-made dual coil intended for small laboratory animals.

Methods

Both presented polymers – poly[O-(2-(methacryloyloxy)ethyl) O-(2-(trimethylamoniumyl)ethyl) phosphorothioate] (p(TMPC)) and 2-(methacryloyloxy)ethyl 2-(trimethylammonio)ethyl phosphate p(MPC) were synthesized by controlled radical polymerization technique (RAFT) of the corresponding zwitterionic monomer. The phosphorothioate group (P=S) in the p(TMPC) polymer provides a chemical shift different from the ³¹P MR signal of the naturally occurring phosphorus-containing compounds, represented herein by the reference p(MPC) probe with a phosphoester group (P=O). MR imaging and spectroscopy were obtained on 4.7T scanner using home-made ¹H/³¹P RF surface coil, which is tuned for both ¹H and ³¹P nuclei, so that anatomical reference acquired by ¹H MR could be obtained. Firstly, ¹H MR imaging (RARE, repetition time/echo time TR/TE=2500/12ms; field of view FOV=10cm; scan time ST=1min) was applied in three planes for probes localization. MR properties were than assessed by ³¹P MRS/MRI measurement of phantom (c_P=100mmol/L; V=1.4ml in H₂O). Single pulse sequence was used for obtaining ³¹P T₁ relaxation times (TR=200–4000ms; ST=30min–10h) and for frequency offset assessment of both contrast agents with different scan times (TR=200ms; ST=20s–3h). ³¹P relaxation times T₂ were measured using Carr–Purcell–Meiboom–Gill (CPMG) spin lock sequence (TE=2–1200ms; ST=1h 23min). For ³¹P MRI chemical shift imaging (CSI) sequence was optimized (TR=500ms, ST=15min–3h, FOV=4.0cm; resolution 2.5x2.5x0.02mm³). ¹H/³¹P MRI overlapping and ³¹P image processing were obtained using imageJ software. MR spectra were evaluated using homemade MATLAB script. ¹H relaxation times T₁ and T₂ were obtained using 1.5T relaxometer (c_P 10–100 mmol/L; V=240 µl in H₂O). AlamarBlue Assay was used for testing cytotoxicity of the polymer on primary human fibroblasts (HF) and rat mesenchymal stem cells (rMSC).

Results

¹H MR T₁/T₂relaxation times (2510.0ms/2067.2ms; c_P=100 mmol/L), were found to be close to those from clear water. ³¹P MR T₁/T₂ relaxation times (2018.3/119.9ms; c_P=100mmol/L) are adequate for further MR experiments. ³¹P MR spectroscopy showed a chemical shift of 56.07ppm between the peaks from polymers with phosphoester and phosphorothioate groups, respectively (Fig.1). This shift made it possible to acquire ³¹P images from both phantoms separately during one CSI measurement by frequency selection (Fig.2). Signal-to-noise ratio calculated from ³¹P spectra results in 13.1–195.1 (ST=2min–3h) for probe with phosphorothioate bond and 12.5–163.7 for its reference. Imaging analyses results in SNR of 6.3–13.6 (ST=15min–3h) and 3.1–5.4, respectively. Higher SNR in both ³¹P MRS/MRI obtained from probe with phosphorothioate group favours it over the reference. Cytotoxicity testing confirms that the polymer is not significantly influencing cells viability even at the concentration reaching 10mg/ml.

Discussion

High phosphorus concentration and SNR obtained from small volume within a short scan time is beneficial for further application in various in vivo animal models, where low probe signal is gained and short acquisition time is needed. Large chemical shift of the probe from frequencies of biological ³¹P spectra gives a unique possibility to distinguish its signal even at in vivo conditions. Our preliminary imaging and spectroscopy results, together with excellent cell viability, indicate that p(TMPC) polymer could serve as an efficient theranostic agent. Results obtained from ¹H MR relaxometry shows that the polymer did not significantly affect water relaxation times, thus reference ¹H images will be free of artefact originated from the polymer.

Conclusion

We present a proof of principle results of a novel ³¹P MR of phosphorus-containing contrast agent based on chemical shift. The MR experiments proved high sensitivity of p(TMPC) probe for ³¹P MRI/MRS and ability to differentiate its ³¹P signal from biological phosphorus signals.

Acknowledgements

The authors acknowledge financial support from the Ministry of Health of the Czech Republic (grant # NU20-08-00095). The study was also supported by the Charles University, GA UK No 358119; Charles University, First Faculty of Medicine; Institute for Clinical and Experimental Medicine IKEM, IN00023001; Ministry of Education of the Czech Republic through the SGS project no. 21332/3012 of the Technical University of Liberec.