Date of Award

2026

Embargo Period

3-18-2028

Document Type

Dissertation - MUSC Only

Degree Name

Doctor of Philosophy (PhD)

Department

Cell and Molecular Pharmacology and Experimental Therapeutics

College

College of Graduate Studies

First Advisor

Richard Drake

Second Advisor

Tammy Nowling

Third Advisor

Peggi Angel

Fourth Advisor

Lauren Ball

Fifth Advisor

Jim Oates

Sixth Advisor

Bethany Wolf

Abstract

Systemic Lupus Erythematosus (SLE) is a chronic autoimmune disease that disproportionately affects women of childbearing age. Nearly half of affected individuals develop Lupus Nephritis (LN), a severe renal manifestation linked to kidney failure and increased mortality. Current clinical monitoring relies heavily on urine biomarkers, yet these lack specificity for glomerular injury and frequently require confirmatory biopsy, underscoring the need for sensitive, non-invasive alternatives. Because protein glycosylation is dynamically remodeled during inflammation and tissue injury, we hypothesized that disease-specific renal N-glycan alterations occur in LN and are reflected in biofluids, enabling non-invasive detection of renal pathology.

To define the spatial glycomic architecture of LN, we performed high-resolution MALDI mass spectrometry imaging (MALDI-MSI) on formalin-fixed paraffin-embedded renal biopsies using sublimation-based matrix application for robust in situ N-glycan detection. Across pilot (HC n=4, DN n=3, LN n=4) and validation cohorts (HC n=10, DN n=10, LN n=10), 122 N-glycan compositions were identified. Distinct disease-specific remodeling patterns emerged: healthy kidneys were enriched in bisected species, diabetic nephropathy (DN) in biantennary and sialylated glycans, and LN in highly fucosylated and pauci-mannose structures. While whole-biopsy profiling revealed global class-level shifts, glomerulus-restricted analysis sharpened disease discrimination, underscoring the value of compartment-resolved spatial glycomics.

To determine whether renal signatures translate to circulation, matched serum and urine samples from healthy controls (n=50) and LN patients (n=50) were analyzed. N-glycan profiles distinguished LN from controls in both biofluids. In fourteen LN patients with matched biopsy, serum, and urine samples, strong correlations were observed between renal tissue and urinary glycans, indicating that urinary glycans directly reflect histologic remodeling. Diagnostic performance was evaluated in an independent cohort of 404 urine samples (LN n=175, SLE n=115, HC n=114). Sparse partial least squares discriminant analysis and five supervised machine learning models achieved near-perfect discrimination of LN (AUC 0.99). Finally, in a longitudinal LN clinical trial cohort, several urinary N-glycans demonstrated dynamic changes at 3 months associated with clinical improvement, supporting their potential utility for monitoring treatment response.

Collectively, these findings establish spatial N-glycomics as a precision framework linking renal tissue remodeling to urinary biomarkers and support the development of glycan-based, non-invasive diagnostics for LN.

Rights

Copyright is held by the author. All rights reserved.

Download

Available for download on Saturday, March 18, 2028

Share

COinS