Chemical Exchange Saturation Transfer (CEST) MR imaging has proven to be able to create new, insightful contrasts especially for brain tumors¹. One possible contrast is based on the Nuclear Overhauser enhancement² (NOE) mediated aliphatic proton magnetization transfer (so called exchange-relayed (rNOE)³) occurring at -3.5 ppm from the water resonance in vivo. The signal is attributed to mobile proteins apparent inside the cytoplasm of cells. In addition to the exchange-relayed process this homo-nuclear ¹H-NOE signal it should depend on mobility, quantified by the correlation time τ_c and external field strength B₀ in the following way⁴: $$S_{rNOE}\sim\tau_{c}\left(1-\frac{6}{1+4\omega_{0}^2\tau_c^2} \right)\;\;\;\;\;\;(1)$$ The purpose of this study was to examine if changes of tissue integrity influence the protein mobility and thus rNOEs. We compared the signal of an intact cell environment to an environment of ruptured cells and investigated the dependence of the aliphatic rNOE on B₀ by measuring at different static field strengths.

White matter (WM) was extracted from fresh pig brain from sacrificed pigs. One half was homogenized using a 7 ml Wheaton Dounce (Wheaton, Millville, USA) tissue grinder first with a loose pestle (gap approx. 112 µm) followed by a tight pestle (gap approx. 50 µm). The other half was cut into small pieces (>1mm). Each half was then diluted by Phosphate Buffered Solution (PBS) with pH=7.05 in a ratio of 2:1 (PBS:tissue) to yield a final buffer concentration of 1/15M and 50µM Gadopentetat-Dimeglumin using Magnevist® (Bayer, Leverkusen, Germany) to reduce T₁. The experiment was performed on a 14.1-T NMR spectrometer (Bruker, Germany) using 8 mm tubes at 37°C.
2.5% (w/v) aqueous bovine serum albumin (BSA) at T = 25°C, pH = 6.56 in PBS was measured on a 7-T whole–body MR tomograph (MAGNETOM 7 T; Siemens Healthcare, Germany), a 9.4-T NMR spectrometer (Bruker) and a 14.1-T NMR spectrometer (Bruker).
T₁ was measured using a saturation recovery sequence with 17 recover times. Z-spectra were obtained using saturation by a train of Gaussian-shaped rf pulses (mean amplitude B₁ = 0.6 µT, t_pulse = 15 ms, duty cycle DC = 60%, t_sat = 15 s, 133 unevenly distributed offsets). To correct for influences of direct water saturation, MT, and T₁ the relaxation-compensated metric AREX⁵ was used. AREX(Δω) = (1/Z(Δω)–1/Zref(Δω))/T₁ was calculated using a 2-pool Lorentzian fit of the direct water saturation and MT for Zref(Δω).

There was no significant difference observable when comparing AREX spectra from the homogenate and the native tissue of the WM from pig brain (fig.2). The T₁ values also match within their errors T_1,homogenate = 2.54 ± 0.10 s and T_1,native = 2.47 ± 0.06 s. There is a slight reduction in the chemical exchange (CE) effects downfield from water. The semi-solid MT was reduced by 10% in the homogenized sample. The AREX spectra of BSA acquired at different B0 show that both the aliphatic rNOE and the CE effects of BSA increases non-linearly with decreasing field strength B₀ (fig.3).The invariance of the aliphatic rNOE to the homogenization of the tissue reveals that there is no influence of the macroscopic, cellular structure on the correlation time τ_c of the respective rNOE-CEST signal. In other words, the tumbling of the proteins and peptides is not affected by the tissue structure. Hence only the immediate chemical environment and the structural state of the protein itself alter the CEST signal considerably^6,7. The observed decrease of MT is coherent with the extensive rupture of the semi-solid matrix due to homogenization. The rupture of the cellular membrane should also lead to fragments of membrane that show no additional signal in the AREX spectra. The increase of the rNOE at lower field strengths contradicts theoretical expectation (eq.1). In addition the increase of CE effects as a function of B₀ suggests another origin for the variation in AREX signals, since CE effects have no explicit dependency on B₀ ⁸_. An explanation may be the reduced spectral resolution at lower field strengths and thus, using the same saturation pulse, labeling of more proton species at each frequency offset at lower B₀. Nevertheless the similar behavior of CE and rNOE effects as a function of B₀ disproofs a strong dependendy of the rNOE on B₀.

Through homogenizing of a tissue sample we could show that the CEST signal of the aliphatic rNOE is independent on the cellular structure. Hence mobility of proteins protons contributing to the CEST effects seems to be unaffected by the induced tissue structure change. We could not find a clear indication that the B₀ dependency for rNOEs is different than that of CE effects.