Splenic LLPCs depend on BAFF in the context of B-cell depletion in NZB/W mice. (A) Gating strategy for sorting ASCs from NZB/W mice for single-cell analysis. ASCs from 27-week-old NZB/W mice were sorted after dump-staining with CD3/CD4/Gr1/Cd11b by a 2-step procedure and separated into CD138^highB220⁻ and CD138^highB220^low fractions. ASCs were collected from the spleen (CD138^highB220⁻, CD138^highB220^low) and bone marrow (CD138^highB220⁻) from 2 untreated and 4 anti-CD20–treated mice at the nadir of B-cell depletion. (B) Heatmap representation of multiplex single-cell RT-PCR performed with the Fluidigm Dynamic Array on splenic CD138^highB220^low and CD138^highB220⁻ populations using diagnostic genes allowing the discrimination of PBs from LLPCs as in Figure 3A. (C) Percentage of individual cells expressing n PC genes in each population (CD138^highB220^low [gray], CD138^highB220⁻ cells [blue], CD138^highB220⁻ anti-CD20 PCs [purple], BM-PCs [green]). (D) Cumulative percentage of individual cells expressing n or more PC genes in each group. (E) Comparison at the nadir of B-cell depletion of CD138^highB220⁻ cell numbers in the spleens of mice. Analysis from 2 independent experiments. Significant differences were estimated by 1-way analysis of variance, Dunn test for multiple comparisons (* P < .05; *** P < .001; **** P < .0001). Mean values are indicated.