Identification of a Novel MPL Loss of Function Mutation in a Patient with Cyclic Thrombocytopenia and Characterization of This Syndrome

Abstract

Introduction: Cyclic thrombocytopenia (CT) is a rare disease characterized by periodic fluctuations in platelet counts. The etiology of this disease has not been fully elucidated. Here, we present molecular characterization of CT in a 53-year-old male patient who has platelet counts ranging between 1 x 109/L to >400 x 109/L during a cycle period of 40 days which has occurred over several years. Methods: Blood transcriptome profiles and plasma cytokine levels were investigated in 24 samples that were sequentially collected every 3-4 days to cover two complete cycles. Total RNA was extracted after lysing red blood cells. 3SEQ (3′-end RNA sequencing for expression quantification) was performed for transcriptome profiling. SAMSeq quantitative analyses were conducted to identify differentially expressed genes. Three screening criteria were applied to this data set to identify a panel of exclusive platelet-specific genes. Plasma TPO (thrombopoietin)/testosterone/ estradiol/CD41a/CD42b levels were detected by ELISA. A total of 62 cytokines were screened using a luminex immunoassay. MPL (TPO receptor) gene was Sanger-sequenced using blood DNA and mRNA templates. DNA was extracted from the patient's hair follicles to exam the presence of the identified MPL mutation, and in-silico predictions were conducted to evaluate the MPL mutation. The wild-type (WT) and c.1210G>A mutant MPL expression constructs were generated and transfected into Ba/F3 cells for stable MPL expression. TPO-stimulated, IL-3 independent growth of MPL-expressing Ba/F3 cells were assessed to determine the function of MPL proteins. Results: Plasma TPO levels cycled between 6 to 2745 pg/mL during the sampling period; TPO levels inversely mirrored the fluctuation of platelet counts and preceded the latter by 3-4 days. Additional plasma cytokines demonstrated either an inverse correlation (e.g. FasL, GM-CSF, and TNF-β, etc.), or well synchronized correlation (e.g. BDNF, PDGF-BB, and RANTES etc.), with platelet counts. The patient was negative for platelet autoantibodies. Transcriptome analysis revealed 977 genes with expression that positively correlated with platelet count changes. Unsupervised clustering stratified these genes into subgroups, including: i) platelet-specific genes, which precede platelet count changes by 3-4 days; ii) platelet-modulated genes, which follow platelet count changes and are regulated by coagulation factors (e.g. F2) and platelet-contained growth factors (e.g. PDGF-BB and EGF); and iii) neutrophil-specific genes, which precede platelet count changes by 7-10 days. Among platelet-specific genes, 34 genes were identified to have >50-fold induction over a cycle, and are present in the platelet transcriptome [Rowley JW, 2011]. A heat map of these exclusive platelet genes is shown in Figure 1A. A novel MPL c.1210G>A heterozygous mutation, which leads to a p.Gly404Arg substitution, was identified in both blood and hair follicle DNA samples of the patient. This substitution resides in a highly conserved site and is predicted to be deleterious. In addition, the mutation creates a novel splice site, which results in the insertion of a 36 bp intron fragment in approximately 15% of MPL transcripts. In vitro analysis confirmed that MPL c.1210G>A is a loss-of-function (LOF) variant. Ba/F3 cells that express WT MPL exhibited IL-3 independent growth in the presence of TPO; however, the Gly404Arg mutant failed to support TPO-stimulated growth (Figure 1B). Conclusion: We describe a patient with CT who exhibits fluctuations of multiple cytokines including an inverse correlation between TPO and platelet counts, and profound gene expression changes in neutrophil- and platelet-specific gene expressions which precede platelet count fluctuations. These data suggest that the cyclic megakaryopoiesis and thrombopoiesis underlies pathogenesis of the disease. Moreover, the identified novel LOF MPL allele may contribute to the disease pathogenesis. The biologic and molecular annotation of this case presents a unique opportunity to examine the transcriptional regulation of platelet homeostasis. Figure 1 (A) Heat map of 34 genes that are exclusively expressed in platelet. (B) Growth of Ba/F3 cells expressing WT or c.1210G>A MPL in IL-3 free medium containing TPO (50 ng/mL). Average viable cell numbers of duplicate experiments were plotted with SD as error bars. Figure 1. (A) Heat map of 34 genes that are exclusively expressed in platelet. (B) Growth of Ba/F3 cells expressing WT or c.1210G>A MPL in IL-3 free medium containing TPO (50 ng/mL). Average viable cell numbers of duplicate experiments were plotted with SD as error bars. Disclosures No relevant conflicts of interest to declare.

Cyclic thrombocytopenia with statistically significant neutrophil oscillations

“…The latter modeling work has suggested that a defective TPO-TPO receptor interaction may contribute to the pathogenesis. Recently, Zhang et al 26 described an interesting case of CT due to a heterozygous loss-of-function mutation of TPO receptor, further suggesting that CT may be due to abnormalities with the TPO receptor. It is fair to say, however, that these previous modeling attempts to understand the mechanistic origins of cyclical thrombocytopenia are not capable of fully explaining the observations of statistically significant oscillations with identical periods in both platelets and neutrophils of patient 1.…”

A Blueprint for Identifying Phenotypes and Drug Targets in Complex Disorders with Empirical Dynamics

“…We began by evaluating the clinical characteristics of individuals with CTP identified by our team at the Stanford Medical Center. A total of n = 3 individuals were included in the study: subject A has a previously reported novel c-mpl heterozygous germline mutation affecting TPO receptor function, 33 subject B carries the same mutation but shows no clinical presentation of the disease, and subject C has CTP. Subject A visited the Stanford Medical Center clinic every 3–4 days over a period of 84 days and complete blood counts were taken; subject C visited the clinic every 3–4 days for a total of 49 days.…”

“…To distinguish phenotypic heterogeneity in CTP beyond the behavior of TPO and platelets, we employed ELISA and a 62-plate Luminex immunoassay to quantify TPO and plasma cytokine concentrations, 33 respectively, in all three of our subjects. We have previously reported classes of cytokines cycling with identical periods in each subject through Lomb-Scargle 37 , 38 periodogram analysis.…”

“…To demonstrate the feasibility and usefulness of empirical dynamics for our case study of CTP, we analyzed data on platelets, TPO, and more than 60 cytokines measured in time series for three individuals: subject A, who carries a heterozygous germline mutation resulting in a loss of c-mpl function, 33 his father, subject B; and an unrelated individual with CTP, subject C. The father-son relationship between subjects A and B is particularly interesting because, despite sharing the c-mpl mutation, subject B has never shown symptoms of CTP. This makes subject B a useful healthy (non-CTP) control and evidence of a possible pre-disease state.…”

Novel cytokine interactions identified during perturbed hematopoiesis

“…[7][8][9] Recently, a 53-year-old male patient with CTP (cycle period of 40 days with platelet oscillations between 1 × 10 9 /L and > 400 × 10 9 /L) was found to have a heterozygous germline mutation resulting in a loss of of c-mpl function. 10 Previous spectral analysis identified statistically significant oscillations in platelet and neutrophil count and in and cytokine concentrations over time. 11 Using sampling data from this patient, we endeavored to uncover the many relationships between more than 60 cytokines and platelets using standard and state-of-the-art statistical and mathematical techniques.…”