To maintain sorting efficiencies at relatively high levels (>80%) and high yields and purities of sorted samples the differential pressure of the core and the sheath fluids can be increased but cannot be >1. Slow sort rates while maintaining optimal differential pressure of flow http://www.selleckchem.com/products/Y-27632.html stream improves efficiency of sorts and the overall yield of intact nuclei. However the greatest variable in our sorting was the origin of the tissue. For example TNBC sorted more efficiently than did PDA samples for both FF and FFPE samples. Gating based on DNA content provides a robust quantitative measure for identifying and sorting tumor populations from samples of interest. For example the 3.0N population sorted from a FF PDA sample was detected 3 years later in an FFPE sample from the same tissue (Figure 5).
The ploidy and the relative distribution of each population present in a biopsy can be recovered by fitting the G0/G1 and G2/M peaks as Gaussian curves and the S phase distribution as a Gaussian broadening distribution. The DNA content histograms from tumor tissue are frequently suboptimal (broad c.v��s, high debris and aggregation) and often complex (multiple overlapping peaks and cell cycles) with frequent skewing and non-Gaussian peak shapes. This is even truer for FFPE specimens that often contain higher levels of damaged or fragmented nuclei (debris) resulting in events usually most visible to the left of the diploid G1 peak and that fall rapidly to baseline (Figures S4,S9). For reproducible phase measurements we typically acquire 10,000 events.
However if a substantial proportion of events are from debris or aggregates, the total number of events acquired must be correspondingly higher in order to assure the required minimum number of intact single nuclei for accurate curve fitting. Different reports have shown that tumor cells can be efficiently sorted from FFPE samples with DNA content based assays and used for genomic analysis , . These studies have typically relied on PCR based assays including SNP arrays. These assays have limited resolution based on the ability to distinguish homozygous from partial copy number losses, the mapping of breakpoints and focal amplicons, and in the number of genes and loci interrogated. Furthermore SNP arrays typically require the preparation of platform-specific reduced complexity samples for optimal results limiting the utility of DNA prepared from each sorted sample.
In contrast our methods use whole genome templates that are compatible with a wide variety of high definition assays including aCGH and NGS. For aCGH analysis, short DNAse 1 digestion of genomic DNAs extracted directly from sorted nuclei or with amplified DNAs from FF or FFPE samples provides uniform templates for labeling . The resolution of our assays with purified sorted samples enables discrimination of single copy loss from homozygous loss and the AV-951 mapping of amplicon and deletion boundaries in each tumor genome.