Introduction

Liver fibrosis is a common pathway shared by all progressive chronic liver diseases regardless of the underlying etiologies. With liver biopsy being the gold standard in assessing fibrosis degree, there is an unmet clinical need to develop noninvasive imaging tools that can directly and repeatedly quantify fibrosis. Type I collagen is a desirable target for molecular imaging as its excessive deposition is specific to fibrosis. Molecular MRI with peptide-based gadolinium (Gd)-containing probes targeted to type I collagen, such as EP-3533 and CM-101, have previously been shown to detect and stage fibrosis in animal models of liver fibrosis^1-3. The rising concerns about the long-term safety of Gd-based probes have underscored a pressing urgency to develop Gd-free MRI contrast agents⁴. Unlike Gd, manganese (Mn) is an essential nutrition element and cellular Mn levels are tightly regulated in the human body. Here, we evaluate the feasibility of molecular MRI using a collagen-targeted Mn-based MRI probe, Mn-CBP8, for imaging liver fibrosis.

Methods

Collagen binding peptide (CBP) has 3 primary amines³. The three amines were acylating using the mono-anhydride of CyDTA, resulting in peptide functionalism with 3 CyDTA chelators. Next, the peptide-chelate conjugate was reacted with manganese(II) chloride. Mn-CBP was isolated after purification using RP-HPLC. Male C57BL/6 mice were treated with carbon tetrachloride (CCl₄) or vehicle (olive oil) for 6 or 12 weeks to induce liver fibrosis or to serve as controls, respectively (N = 3 – 5 / group). Two additional groups of mice were fed a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD) (N = 6) or standard chow (N = 3) for 6 weeks. Animals were anesthetized with isoflurane (1.5%) and imaged with a 4.7 Tesla Bruker MRI scanner. 3D T₁-weighted FLASH images were acquired before and up to 60 minutes post intravenous administration of 10 µmol/kg Mn-CBP8. In the 12-wk CCl₄/vehicle groups, molecular MRI with EP-3533 (10 µmol/kg) was also performed as a positive comparison for Mn-CBP8 in the same animals. Mice were euthanized 90 minutes p.i. and liver tissue was collected and analyzed for hydroxyproline content and histology. Pharmacokinetics, biodistribution, and elimination of Mn-CBP8 radiolabeled with Mn-52 were evaluated in normal mice at 1d and 7d post-injection.

Results

Increased liver fibrosis can be detected by red staining on Sirius Red in 6-wk and 12-wk CCl₄ mice as compared to the controls (Fig. 1A). Mn-CBP8 enhanced MR showed higher liver signal enhancement in the CCl₄ mice compared to the age-matched vehicle-treated controls (Fig. 1B). At 45 min post-injection, Mn-CBP8 induced significantly greater liver-to-muscle contrast to noise ratio (ΔCNR) in CCl₄ mice (Fig. 1C). Both collagen-targeted probes, Mn-CBP8 and EP-3533 induced significantly greater percentage signal change (%SI) in the livers of 12-wk CCl₄ mice as compared to controls, while similar %SI over time was seen between Mn-CBP8 and EP-3533 in the control groups (Fig. 1C), indicating comparable pharmacokinetics between the Mn-based and Gd-based collagen probes. Similarly, liver fibrosis induced by CDAHFD was confirmed by elevated hydroxyproline content and detected by ΔCNR after Mn-CBP8 injection (Fig. 1D). Biodistribution analysis after [52Mn]Mn-CBP8 confirmed predominant renal clearance of the probe (Fig. 1F).

Conclusion

The ability of this Mn-based collagen probe to detect liver fibrosis was demonstrated in vivo using a CCl₄ and CDAHFD mouse models of liver fibrosis. Mn-CBP8 shows similar pharmacokinetics and sensitivity to type I collagen as compared to Gd-based EP-3533 with renal clearance. This study opens the door to studying manganese-based collagen-targeted probes for imaging fibrosis with in vivo efficacy and safety that is poised for clinical development.

Image/Figure:

Click to view full size

Image/Figure Caption:

Figure 1 A) Total collagen assessed by collagen proportional area (CPA) in Sirius Red stained slides as a fibrosis measure in control and CCl₄ groups. B) T1-weighted images acquired before and 45 min after injection of Mn-CBP8 (10 µmol/kg) in control and CCl₄ groups. C) Change in the liver to muscle contrast to noise ratio (ΔCNR) at 45 mins post-injection of Mn-CBP8 in control and CCl₄ groups. D) Percentage signal change (%SI) time courses after Mn-CBP8 or EP-3533 injection in each group. D) Change in the liver to muscle contrast to noise ratio (ΔCNR) at 45 mins post-injection of Mn-CBP8 or EP-3533.  E) Total collagen as assessed by liver hydroxyproline (Hyp) and mean DCNR at 30-40 mins post-injection of Mn-CBP8 in the CDAHFD model and naïve mice. F) Percentage of injected dose per gram of tissue (%ID/g) in representative organs 1 day and 7 days after injection of Mn-PyC3A to normal mice. All data are shown as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

Author