Recently, Chemical Exchange Saturation Transfer (CEST)^1-3 has shown great potential in detecting low concentration metabolites and proteins. In CEST, proton exchange accumulates the effect of the perturbation of an exchangeable small spin pool in a larger pool, coupled with a strong sensitivity enhancement. Traditionally, the perturbation is saturation or excitation of exchangeable proton spins. In this work, we expanded this to an excitation-based perturbation specific for heteronuclear coupled exchangeable protons and acquired MRI images of ¹⁵N-containing urea.

Phantoms containing ¹⁵N labeled urea solutions at various pH values were prepared. MRI experiments were performed on a vertical bore 17.6 T (750 MHz) scanner (Bruker, Ettlingen, Germany). A solution broadband inverse (BBI) probe was used for NMR experiments, while a 23 mm heteronuclear volume coil (¹H/¹⁵N) was used for MRI imaging. The images were acquired using a RARE sequence with TR/TE = 10s/8.0ms, RARE factor = 32, slice thickness = 4 mm, matrix size 128×128 and FOV 2.0×2.0 cm².

As illustrated in Fig. 1, the pulse sequence contains repeated label transfer modules (LTMs)^4,5 consisting of two steps: (i) a proton spin echo sequence with flip back to the longitudinal axis in which the total evolution time has length of 2τ=1/J_NH. Evolution under scalar coupling is alternatively engaged or disengaged by turning a heteronuclear refocusing pulse on or off, respectively, during the proton refocusing pulse,⁶ resulting in a sign discrimination for protons that are scalar coupled to ¹⁵N. (ii) an exchange transfer time (t_exch). This LTM is repeated multiple times (n_LTM) to achieve maximum CEST signal.

Using the above experimental setup, two images were acquired, one for reference without a pulse on the ¹⁵N channel (I_off, Fig. 2a), the second with a 180° pulse on the ¹⁵N channel (I_on, Fig. 2b). The relative difference image $$$ I_{diff}=(I_{off}-I_{on})⁄I_{off}\times 100\% $$$ is shown in Fig. 2c. In order to obtain maximum CEST signal, the parameters τ, t_exch and n_LTM were optimized (Fig. 3). The urea proton exchange rates effects were also measured, which depend on the pH (Table 1). To test the effects of concentration, heteronuclear CEST MRI images of samples of 1M, 250 mM, 100 mM, and 25 mM ¹⁵N labeled urea and 1M nonlabeled urea were acquired at pH = 5.0. The data in Fig. 4 show that there is no visible effect on nonlabeled urea. Using this scheme, we can thus selectively image ¹⁵N labeled samples.

For τ, a value of 1/2J = 5.6 ms should give a maximum transfer. The optimal value is expected to be affected by R_2,app, i. e. on the pH value (Fig. 3a), and for fast exchanging protons negligible signal was observed. A transfer time τ of 4.5 ms was found suitable for most pH values using a t_exch of 100 ms and an n_LTM of 24.

During t_exch, the labeled urea protons exchange with bulk water protons. The exchange rates are relatively slow at higher pH and longer exchange times are required to get a maximum CEST image signal (Fig. 3b). At lower pH values, a shorter exchange time suffices. An exchange time of 100 ms worked for most phantoms.

As expected, a larger number of LTMs leads to higher CEST signal until a plateau (Fig. 3c).

The requirement for efficient transfer would be that $$$J \gg k_{exch}$$$, while k_exch is as fast as possible. In this experiment, a pH of 5.0 with the exchange rate of 35 Hz was found to be the condition for a maximum signal difference for ¹⁵N in urea. At pH 4.1, the exchange rate was 222 Hz and almost no CEST signal was observed. At pH values higher than 5.0, the exchange rate reduced further and a relatively small CEST signal was observed. If the pH would increase further, base catalysis would become active and the CEST effect would first increase before going down again at much higher pH values.We demonstrated a model phantom experiment of MRI imaging of heteronuclear CEST transfer. Compared to imaging ¹⁵N directly, this scheme enhances the ¹⁵N signal in two ways. First, the gyromagnetic ratio (γ) of ¹H is 10 times that of ¹⁵N. By using proton excitation and detection, the sensitivity is increased by a factor γ^5/2. Second, the extra proton-based enhancement due to heteronuclear CEST was a factor of 5.