Fibrosis and metastatic disease represent major causes of morbidity and mortality worldwide, yet current imaging modalities lack the molecular specificity needed for early detection and quantitative monitoring. This dissertation describes the development, characterization, and translational validation of hProCA32.Collagen, a next-generation protein-based MRI contrast agent engineered for collagen-targeted molecular imaging. hProCA32.Collagen was designed by conjugating a collagen type I-targeting peptide to the C-terminal of human parvalbumin through a flexible linker, achieving exceptional relaxivity values (r₁ = 34 mM⁻¹s⁻¹, r₂ = 50 mM⁻¹s⁻¹ at 1.4 T) and high collagen binding affinity (Kd = 0.42 μM) while maintaining superior metal selectivity (10⁴-10¹² fold preference for Gd³⁺ over Ca²⁺ and Zn²⁺). A scalable expression and purification platform was established, achieving >95% purity with an optimized refolding protocol, yielding 49% recovery, and the formation of intact disulfide bond structure. Comprehensive biophysical characterization demonstrated robust metal coordination, pH stability, and formulation conditions supporting clinical translation. In preclinical models of lung cancer, hProCA32.Collagen enabled the detection of tumors as small as 0.6 mm and multi-organ metastases (lung, adrenal, kidney) with a 3- 4-fold higher contrast-to-noise ratio than clinical Gadovist and superior performance over high-resolution CT, which failed to detect sub-millimeter adrenal and renal lesions. In fibrotic disease models, hProCA32.Collagen successfully differentiated fibrosis stages in bleomycin-induced pulmonary fibrosis and COPD. In contrast, quantitative CT showed no significant differences despite histology demonstrating high collagen expression in fibrotic regions of the lungs. Preliminary radiomics analysis revealed distinct imaging signatures that separate cancer from fibrosis, based on texture and intensity features. In the PCK rat model of polycystic kidney disease, hProCA32.Collagen provided a 3- to 5-fold greater enhancement than Gadovist, mapping
striated fibrotic remodeling and individual cyst burden with strong histological correlation. This work establishes hProCA32.Collagen as a transformative agent for non-invasive fibrosis quantification and early metastasis detection, addressing critical gaps in current diagnostic imaging. The analytical qualifications, safety profiling, and multi-disease validation presented herein provide a comprehensive foundation for IND filing and clinical translation toward precision molecular imaging in oncology and fibrotic diseases.