Date of Award

4-10-2020

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

First Advisor

Meggan Mackay, MD, MS

Abstract

Introduction

Cognitive dysfunction (CD) and depression are highly prevalent in systemic lupus erythematosus (SLE). Interferon-alpha (IFNa) contributes to SLE pathogenesis, associates with these phenotypes independent of SLE and stimulates the kynurenine/tryptophan (KYN/TRP) pathway, causing an increase in quinolinic acid (QA) relative to kynurenic acid (KA) that predisposes to neurotoxicity. This imbalance is associated with depression, CD, and changes in brain structure and function. We hypothesize that IFNa contributes to SLE-mediated CD and depression through activation of the KYN/TRP pathway.

Methods

This is a cross-sectional study of 74 SLE subjects and 74 healthy controls (HC) without history of focal neurologic or primary neuropsychiatric disorders. SLE subjects were categorized by disease activity. All subjects had serum levels of KYN, TRP, QA, and KA measured by high-performance liquid chromatography (HPLC), and were assessed using the Automated Neuropsychological Assessment Metrics (ANAM), a 2x2 array, a Spatial Navigation Task (SNT), and behavioral tests. Additionally, RNA sequencing (RNAseq) analysis compared global gene expression profiles in whole blood samples of SLE subjects and HC. A subset of 14 SLE subjects and 5 HC underwent brain imaging [fluorodeoxyglucose-positron emission tomography (FDG-PET) and diffusion tensor imaging (DTI)] as part of a separate, ongoing longitudinal imaging study.

Results

Serum ratios of KYN/TRP and QA/KA were elevated in SLE compared to HC (p< 0.001). There were no differences in KYN/TRP or QA/KA ratios among SLE disease activity groups (Inactive, Stable Active, or Flare). SLE subjects performed worse than HC on 4/5 ANAM tests and the 2x2 array (all p< 0.03), and Inactive SLE performed the worst on most tasks. Multivariable modeling revealed that an elevated QA/KA ratio had a significant, negative impact on Match-to-Sample (MTS) scores, while increased disease activity had a positive impact on MTS scores. An elevated QA/KA ratio trended towards slightly higher odds of clinically significant depression in SLE (p=0.088), and no association was found in multivariable modeling. Differential gene expression analysis yielded 933 genes differentially expressed in SLE with ≥2-fold change from HC; of the top 100 most differentially expressed genes identified in whole genome expression principle component analysis, 70 were IFN stimulated genes (ISG). 82 of the 933 differentially expressed genes in SLE correlated significantly with QA/KA ratios (p< 0.05) and 46% (38/82) were ISG. A Targeted-Modular IFN score derived for each SLE subject comprised of 3 type I ISG (CCL8, CXCL10, LAP3), positively correlated with both the KYN/TRP and QA/KA ratios (p< 0.001), and these associations were stronger in anti-double stranded DNA antibody (anti-dsDNA) negative subjects (p< 0.006). An elevated QA/KA ratio correlated with reduced metabolism on FDG-PET in the premotor cortex (BA 6) of 14 SLE subjects (p=0.021), and reduced white matter microstructural integrity in fronto-parietal pathways (p=0.018) in 13 of these subjects.

Conclusions

Elevated serum KYN/TRP and QA/KA ratios indicate KYN/TRP pathway activation in SLE. An increased QA/KA ratio, after controlling for other factors, associated with poor performance on a task of working memory and visuospatial processing. Its lack of a significant relationship with depression in this SLE cohort may be due to the paucity of subjects with severe depression. Unsupervised gene expression analyses demonstrate an enrichment for ISG in SLE subjects and gene expression of many ISG correlate with KYN/TRP and QA/KA ratios. Supervised analysis of ISG and metabolite ratios yielded a Targeted-Modular IFN score, composed of 3 type I ISG with plausible relationships to neuroinflammation. The positive association between the Targeted-Modular IFN scores and KYN/TRP pathway activation was stronger in anti-dsDNA negative subjects, a subgroup of SLE subjects in which IFNa may be pathogenic. An elevated QA/KA ratio was associated with brain functional and structural deficits in a small subgroup of SLE subjects compared to HC in regions associated with working memory performance, consistent with the cognitive testing results in the entire cohort. These results need to be assessed longitudinally, and validated in a separate cohort. Overall, these findings support further study of the KYN/TRP pathway as a potential marker for brain functional and structural deficits, and cognitive and behavioral abnormalities, in SLE.

Download

Share

COinS