Background/Purpose

Lupus flare reflects an increase of disease activity that is associated with significant morbidity and accumulation of tissue damage. Prediction and prevention of lupus flare is an important goal of lupus research and, ultimately, clinical management. Our aim was to identify perturbations in molecular and cellular mechanisms that precede and coincide with lupus flare.

Methods

PBMC samples, clinical and laboratory data were collected longitudinally from 19 patients meeting ACR criteria for SLE. From 2 to 8 visits/patient (total 90 data points) were analyzed. All patients experienced at least one SELENA-SLEDAI flare. Patients did not receive biologic agents during flare visits, but therapy was otherwise uncontrolled. Transcriptional profiles were obtained for each visit using Affymetrix HG U133Plus 2.0 GeneChips (by Novo Nordisk). SELENA-SLEDAI, BILAG and physician global assessments were compared to identify a point of maximum disease activity (flare). The flare point was set as day 0 for each patient over a 2 year interval. The remaining visits were arranged by time before or after flare. An mRNA profile for each transcript was established using the smoothing-splines mixed effect model (Berk M., 2012). All fitted models were classified by hierarchical clustering, and obtained clusters were studied using the gene-enrichment profiler database (Xavier lab‎; Benita Y. et al., 2010) and the DAVID v6.7 database to link functionally-related transcripts. The selected cell-specific or function-related genes were refitted and dynamic changes over one year intervals preceding and following flare were characterized.

Results

From 22190 transcripts, the models were successfully fit for 3189 transcripts. Clustering identified 9 major patterns over the observed time course. Most striking was the rise of B cell-related transcripts, including IRF4, SPIB, CD19, CD22, and CD79b, as early as 180 days before lupus flare. Subsequently, transcripts expressed in the myeloid lineage [mannose receptor (MRC1/CD206), CLEC10A/CD301, GPR137B] and those linked to lysosomal function [N-acetylgalactosaminidase (NAGA), acid phosphatase 2 (ACP2), scavenger receptor class B (SCARB2), and endolyn (CD164)] were decreased. B cell-related transcripts rapidly declined immediately prior to flare, while a distinct transcript cluster, including SIGLEC5/CD170, FPR1, IL1R2, SLC11A1, CR1, C1RL, emerged and coincided with flare. A tissue enrichment profiler identified those transcripts as highly expressed in blood neutrophils. The mean level of neutrophil-related transcripts correlated with SLEDAI and BILAG scores at time of flare (R=0.49 and R=0.44, respectively; p<0.001 for both).

Conclusion

Despite the clinical heterogeneity that characterizes flare in individual patients, a common sequence of molecular events was observed preceding and during lupus flares. Strikingly, changes in the B cell population were observed as early as 6 months before flare. Similarly, perturbations in myeloid lineage transcripts occurred prior to flare. An increase in neutrophil-related transcripts is an indicator of lupus flare and correlates with disease activity at time of flare.

---

« Back to 2014 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/b-cell-and-neutrophil-related-transcripts-predict-and-characterize-a-lupus-flare/