Purpose:To estimate the exchange rate of creatine (Cr) CEST and to evaluate the pH sensitivity of guanidinium (Guan) CEST in the mouse brain. Methods: Polynomial and Lorentzian line-shape fitting (PLOF) were implemented to extract the amine, amide, and Guan CEST signals from the brain Z-spectrum at 11.7T. Wild-type (WT) and knockout mice with the guanidinoacetate N-methyltransferase deficiency (GAMT −/− ) that have low Cr and phosphocreatine (PCr) concentrations in the brain were used to extract the CrCEST signal. To quantify the CrCEST exchange rate, a two-step Bloch-McConnell (BM) fitting was used to fit the CrCEST line-shape, B 1 -dependent CrCEST, and the pH response with different B 1 values. The pH in the brain cells was altered by hypercapnia to measure the pH sensitivity of GuanCEST.Results: Comparison between the Z-spectra of WT and GAMT −/− mice suggest that the CrCEST is between 20% and 25% of the GuanCEST in the Z-spectrum at 1.95 ppm between B 1 = 0.8 and 2 μT. The CrCEST exchange rate was found to be around 240-480 s −1 in the mouse brain, which is significantly lower than that in solutions (∼1000 s −1 ). The hypercapnia study on the mouse brain revealed that CrCEST at B 1 = 2 μT and amineCEST at B 1 = 0.8 μT are highly sensitive to pH change in the WT mouse brain.
Conclusions:The in vivo CrCEST exchange rate is slow, and the acquisition parameters for the CrCEST should be adjusted accordingly. CrCEST is the major contribution to the opposite pH-dependence of GuanCEST signal under different conditions of B 1 in the brain. K E Y W O R D S amide CEST, amine CEST, chemical exchange saturation transfer (CEST), concentration, creatine CEST (CrCEST), exchange rate, guanidinium CEST (GuanCEST), high spectral resolution (HSR) CEST, polynomial Lorentzian line-shape fitting (PLOF), two-step Bloch-McConnell (BM) fitting Ziqin Zhang and Kexin Wang contributed equally to this work.