Detection of Highly Clonal CD34⁺, Mononuclear Bone Marrow and Peripheral Blood Cells From Patients with Myelodysplastic Syndrome and Age Related Increase of Clonal Hematopoiesis In Healthy Subjects Using a Novel Transcriptional Pyrosequencing Based Clonality Assay

Abstract 977

X-chromosome inactivation occurs in female cells leading to stochastically random silencing of one X-chromosome. Accordingly, investigation of X-chromosome inactivation patterns (XCIP) in healthy female tissues revealed similar ratios of cells carrying either an active maternal or paternal X-chromosome. Genomic XCIP markers like the HUMARA locus have been used to elucidate clonality of myelodysplastic syndromes (MDS). However, analysis of healthy female peripheral blood (PB) cells has shown a strong age-related increase in skewing of XCIP suggesting the frequent manifestation of clonal hematopoiesis in the elderly. Since the occurrence of clonality in blood cells from healthy individuals remains controversially discussed (Swierczek et al, Blood 2008), we developed a novel highly accurate transcription based approach using Pyrosequencing to detect clonal hematopoiesis. In addition, we present the first comprehensive XCIP analysis of CD34 purified (CD34+) and unfractioned bone marrow (BM) cells of MDS patients and healthy females.

BM, CD34+ and PB cells were obtained from patients with MDS (IPSS-low/int-1-risk BM: n=25, CD34+: n=13, PB: n=21, IPSS-int-2/high-risk BM: n=16, CD34+: n=9, PB: n=12) and age related healthy donors (BM: n=19, CD34+: n=15, PB: n=105) after informed consent. Genomic DNA SNP genotyping was carried out in order to screen for heterozygous clonality marker genes located on the X-chromosome, namely BTK, FHL1, IDS, MPP1 and G6PD. After PCR-amplification of informative marker loci from cDNA transcripts, SNP allele ratios, representative for XCIP, were quantified using the PyroMark ID system (Qiagen, Hilden, Germany). Clonality was assumed for samples exhibiting skewed XCIP ratios >80% (allelic ratio of >4:1).

Standard curves from pyrosequencing reactions with predefined allelic ratios revealed strong correlations for assessment of XCIP ratios in all markers (R²>0.99). Furthermore, excellent inter-marker correlations of XCIP ratios from individuals with multiple informative markers were achieved ranging from R²>0.83 to R²<0.98. These findings indicate a balanced inactivation of the whole X-chromosome and corroborate the suitability of our method. Of note, distinctively increased numbers of clonal PB samples from healthy elderly donors (age: 65–93 years) with skewed XCIP ratios in 55% of all cases were detected as compared to young donors with skewed XCIP ratios in 13% and 25% for age groups ranging from 18–40 and 18–65 years, respectively. Analysis of XCIP ratios in MDS samples revealed clonality in 96% of BM, 100% of CD34+ and 90% of PB specimen from MDS low/int-1-risk and 100% of BM, 100% of CD34+ and 92% of PB specimen from MDS int-2/high-risk patients. In contrast, only 44% of BM and 43% of CD34+ samples from healthy elderly (age range 65–93 years) and 20% of BM and 25% of CD34+ samples from healthy young females (age range 18–65 years) were graded as clonal. Furthermore, we observed extremely skewed XCIP ratios of >95% in 56% of BM, 62% of CD34+ and 57% of PB cells from MDS low/int-1-risk and 50% of BM, 67% of CD34+ and 67% of PB cells from MDS int-2/high-risk patients but only in 22% of BM, 29% of CD34+ and 3% of PB cells from healthy elderly subjects.

Our detection of highly clonal cells in MDS patients in 3 hematopoietic cell compartments and frequent detection of extreme XCIP skewing provides strong evidence for the manifestation of a malignant MDS clone even in early progenitor cells. Moreover, detection of non-clonal PB cells in patients with clonal CD34+ cells indicates the presence of residual healthy hematopoietic cells in the peripheral blood. Notably, in our hands, age related increases of XCIP skewing have been detected in PB cells and, for the first time, in CD34+ and BM cells of healthy females suggesting the clonal outgrowth of hematopoietic cells in the early progenitor compartment of healthy elderly. In conclusion, pyrosequencing based clonality assessment has proven to be suitable for quantitative monitoring of clonal and potentially malignant cell populations in hematopoiesis.