Background/Purpose:
Currently, classification and treatment decisions in lupus nephritis (LN) are
largely based on renal histology. Transcriptome analysis may accurately
differentiate types of renal involvement, and better inform treatment outcome
and prognosis. Single-cell RNAseq is an emerging technology that allows for
resolution of differential gene expression at the single-cell level, thus
facilitating the detection of expression changes that would be unnoticed in
conventional bulk-tissue RNAseq. Our objectives were to determine if
single-cell RNAseq can differentiate among various cell types in LN kidney, and
distinguish between LN class IV and V.

Methods:
RNAseq was performed on ~2 mg kidney tissue from consented clinically indicated
renal biopsies in 5 SLE patients, using Fluidigm C1 Integrated Fluidic Circuits
and the C1 Single-Cell Auto Prep. cDNA libraries were prepared using the
Nextera XT DNA Library Prep Kit followed by HiSeq 2500 (Illumina) sequencing.
Sequence-read alignments were conducted using the STAR aligner, and uniquely
mapped reads were summarized at the gene level using the featureCounts
software. Differential expression analysis was performed using the DESeq2
package for R version 3.2.0.

Results:
A total of 155 single cells were characterized from LN kidney biopsies (n=36
pure class IV and n=108 pure class V). An average of 3500 genes/cell were
identified (Fig. 1a), with 50% of reads mapping to the reference genome (Fig.
1b). Through transcriptome analysis and utilization of standard lineage
markers, major renal resident and infiltrating cell types including podocytes,
endothelial cells, tubular cells, macrophages, dendritic cells, and lymphocytes
could be identified (Fig. 1c). Comparison of gene expression between cells from
class IV and class V biopsies revealed that class IV podocytes and endothelial
cells expressed significantly higher levels of complement C7 than class V cells
(p=.001 and .01, respectively). Podocytes from class IV kidney expressed significantly
higher IL-1b (p<.01) and TLR3 (p<.001) as well as IFN-γ receptor
(p<.001) and TNF receptor (p<.05). Similarly, endothelial cells from
class IV kidney demonstrated significant upregulation of several inflammatory
molecules including NF-κB (p=.005), CXCL8 (p<.05), and IL-1b
(p<.05).  Proximal tubular cells from class IV patients expressed significantly
higher levels of CXCL12 (p<.05) than those isolated from class V patients.

Conclusion:
These findings demonstrate that single-cell RNAseq is feasible and informative in
cell specific transcriptome analysis from fresh renal biopsy tissue in SLE. In this
pilot study, single-cell RNAseq correlates well with histologic classification,
and shows changes in gene expression that could directly drive or be influenced
by disease pathogenesis. If confirmed, this novel approach may identify new therapeutic
targets and track clinical responses.