A comprehensive evaluation of human plasmacytoid dendritic cells using small volumes of human blood

Abstract

Elucidation of the role played by cells of the innate immune system, particularly the dendritic cell (DC) populations, has led to a precipitous growth in our understanding of immunity to pathogens and foreign antigens. Much of this information has been derived from studies using mouse model systems. However, mice and human DCs differ drastically in the relative distribution of the toll-like receptors (TLRs) critical for immune activation. This is particularly true for the plasmacytoid DCs (pDCs), which are activated almost exclusively through TLR signaling. Variation in this DC subpopulation has been implicated in a number of pathological syndromes; therefore, a thorough understanding of their steady state and activation profiles in human patients is essential. A number of factors, including the relatively low numbers of these cells in blood, have precluded careful analysis in clinical trials. To overcome these limitations, we have developed a technique for studying the steady state and activation profile of pDCs collected from small amounts of human blood. This technique can be performed with 10,000 cells to obtain the immune transcriptome of the pDCs analyzed by quantitative PCR using amplified RNA. In addition, we have used multiplex enzyme-linked immunosorbent assays to measure secreted proteins. We demonstrate the validity of this technique and document its potential for use with blood from human study populations.

FIG. 1.

Differences in RNA yield between cDCs and pDCs. cDCs and pDCs of indicated numbers were lysed and total RNA was isolated. Quantification was performed using Nanodrop spectrophotometer. Each group represents between 5 and 130 individual isolations. *p < 0.05, **p < 3 × 10⁻⁵, as determined using a two-tailed student's t-test.

FIG. 2.

RNA isolation and amplification method does not alter gene profile down to 5000 pDCs. The expression of 10 genes related to pDCs maturation was compared from amplified RNA from the indicated number of pDCs to the RNA from 250,000 pDCs nonamplified. Data are expressed as normalized fold change of ct values from infected cells over NI cells. pDCs were infected with NDV (moi = 0.5) or left untreated for 8 h. After infection, cells were separated into samples containing 1000–250,000 cells for RNA isolation and qRT-PCR. See data analysis section for a detailed description. R² > 0.91 indicates that there is no statistically significant difference.

FIG. 3.

Individual infections and amplification does not alter gene profile down to 10,000 pDCs. The expression of 10 genes related to pDCs maturation is compared from amplified RNA from the indicated number of pDCs to 250,000 pDCs nonamplified. Data are expressed as normalized fold change of ct values from infected cells over NI cells. Between 1000 and 250,000 pDCs were infected with NDV (moi = 0.5) or left untreated for 8 h, before RNA isolation and qRT-PCR. See Data analysis section for a detailed description. R² > 0.47 indicates that there is no statistically significant difference.

FIG. 4.

mRNA expression profiles from various stimuli established for limited numbers of pDCs. In triplicate, 25,000 pDCs were treated for 6 h with media, LPS, CpG, NDV, or PR8. RNA was isolated from each sample for RNA amplification followed by qRT-PCR. Data are expressed as log₂ normalized fold change of ct values from infected cells over NI cells with mean of triplicate indicated and standard error bars.

FIG. 5.

Secreted protein profiles determined from different stimuli from limited numbers of pDCs. In triplicate, 25,000 pDCs were treated for 6 h with media, LPS, CpG, NDV, or PR8. Supernatants were harvested and analyzed using Luminex multiplex system. Data are the mean of triplicate samples from duplicate wells with standard error bars expressed in nanograms/milliliter.