Genome Signature And Clinical Implication Of Lymphocyte In Acute Lymphoblastic Leukemia Patients Immediately After Total Body Irradiation

Background and purposeBreakage of cellular DNA following radiation occurs in both the nuclear andextra-nuclear DNA. Thus, besides nuclear DNA(nDNA), mitochondrial DNA (mtDNA) isequally affected as an only extra-nuclear genome immediately after radiation. Fewinvestigations ex vivo showed that both gene expression signature and mtDNA damage(such as mutation and depletion) are potential molecular radiation biodosimetry.Unfortunately, both tools lack of direct and strong support from research in vivo.Furthermore, it will help to understanding transcriptional mechanism immediately afterradiation damage by exploring the relation between gene expression signature andbiological effect. Besides, it will provide strong evidence for next step clinical triage.In this study, we taken the acute lymphoblastic leukemia (ALL) patients undergogingtotal body irradiation (TBI) preconditioning as human beings in vivo irradiation model, andbegan our study from two ways, which are gene expression signature and mtDNAalterations (including mtDNA copy number and mtDNA4977bp common deletion (CD)).The aim of section one was to identify candidate genes associated with TBI efficiency andto provide novel insights into the transcriptionally regulated pathways involved TBIpreconditioning before allogenetic transplantation. In section two, we attempted to addressthe mtDNA status in irradiated human peripheral blood lymphocytes in vivo to elucidatewhether alterations in mtDNA can be linked to exposure to total body irradiation.Furthermore, to analysis its potential value as radiation biodosimetry. As above mentioned,our experiments aimed to study genome signature of lymphocyte in acute lymphoblasticleukemia patients immediately after total body irradiation and pointed to some molecularmechanisms in it. Methods1. Section one comprised peripheral blood (PB) samples from six high risked ALLpatients undergoing TBI as pre-transplantation treatment in their first complete remission(CR1) and were treated with two4.5Gy TBI sessions daily.7ml of PB containing EDTAas an anticoagulant was collected from all patients prior to and24h following exposure foreach radiation treatment. For each of the samples, peripheral blood mononuclear cells wereisolated, and total RNA in purified peripheral lymphocytes was prepared and measurement.At this step, two patients were excluded because of low RNA yield in irradiated samples.Then, Cyanine-3(Cy3) labeled cRNA was prepared and hybridized to Agilent WholeHuman Genome Oligo Microarrays according to the manufacturer's instructions. Afterdifferentially expressed genes were identified through Fold-change screening, bioinformaticsanalysis were performed subsequently. Firstly, the random-variance model F-test wasapplied to the24,315probe sets. Both P-value <0.05and false discovery rate (FDR)<10%were considered statistically significant. Secondly, unsupervised hierarchical clusteranalysis was done by using these identified genes. Then, Gene expression profiles wereanalyzed using a method called "Series tests of cluster"(STC), which extracts significantpatterns by calculating the correlations of gene expression and then identifies significantgenes based on their correlations with a specific pattern. Each extracted pattern representeda set of co-expressed genes in further study. GO analysis was applied in order to organizegenes into hierarchical categories and uncover the co-expression network on the basis ofbiological process and molecular function. The co-expression network of gene interaction,representing the critical mRNAs and their targets, was established according to the mRNAdegree. Meanwhile, the significant genes in unique patterns were performed on the basis ofscoring. Finally, the expression levels of12selected genes were validated by quantitativereal-time PCR experiment.2. Section two comprised PB samples from26high risked ALL patients undergoingTBI as pre-transplantation treatment, and after TBI precondition the same as abovementioned,7ml of PB containing EDTA as an anticoagulant was collected from all patientsprior to and24h following exposure for each radiation treatment. Besides,39healthydonors had the same volume of PB collected without ionizing radiation for comparing thedifference of basal mtDNA and CD levels between ALL patients and normal donors before IR. For each of the samples, peripheral blood mononuclear cells were isolated, and DNA inpurified peripheral lymphocytes was prepared. TaqMan probes with conjugated MGBgroups were performed to ensure maximal specificity in real-time PCR reaction.Dose-dependent plasmid-constructed-actin, hypervariable region2(HVR2) and CDstandards were used in each run of real-time PCR. All TaqMan reactions were carried out inan ABI7500Real-Time PCR instrument to explore the mtDNA and CD contents.Results1. Gene expression signature of lymphocyte in acute lymphoblastic leukemia patientsimmediately after total body irradiation①478differentially expressed genes were identified as significant. Unsupervisedhierarchical cluster analysis separated three different dose TBI groups clearly with the478genes in12samples, except for one4.5Gy irradiation sample clustering in0Gy TBItreated group.②STC extracted three patterns of change in gene expression. Pattern1included77genes that were progressively induced from4.5to9Gy (P=2e-7). Pattern2included64genes that were progressively repressed from4.5to9Gy (P=5e-4). Pattern3included85genes that were repressed only at9Gy (P=5e-3).③The differentially expressed genes in the patterns were classified into differentfunctional categories according to Gene Ontology (GO) project for biological process. Theresult indicated, genes in pattern1indicated neurotransmitter transport, proteolysis,epidermal growth factor receptor signaling pathway, and some metabolic process. Theprimary GO categories for genes in pattern2are involved in DNA recombination,ubiquitin-dependent protein catabolic process, and stress response so on.④The STC-pathway analysis for regulated genes in distinct patterns found hematopoieticcell lineage, T cell receptor signaling pathway, and primary immunodeficiency in pattern2,while allograft rejection and graft-versus-host disease (GVHD) pathway in pattern3.⑤Gene co-expression network method indicated ubiquitin-dependent protein catabolicprocess and transcription pathway.8genes (CYP4F11, PCNT, LOC286467, PSMD14,SENP3, CIZ1, GZMA, EPS8) were found fairly tight conjunction with others. Of which,PSMD14, SENP3were involved in ubiquitin-dependent process. ⑥The reliability of the data obtained from the microarray was verified by performingquantitative real-time PCR on12representative genes. All genes tested responded with thesimilar general pattern as seen on the microarrays.2. Mitochondrial DNA alterations of peripheral lymphocytes in acute lymphoblasticleukemia patients undergoing total body irradiation therapy①No statistically significant difference was found for logarithm of basal mtDNA andCD level between healthy donors and patients with ALL.(P＞0.05).②Changes of mtDNA content and CD ratio after TBI in patientsmtDNA content was significantly increased in4.5and9Gy irradiation groupscompared with0Gy group (4.5Gy vs.0Gy=0.043,9Gy vs.0Gy=0.025), as well as CDratio reduced in4.5and9Gy irradiation groups compared with0Gy group (4.5Gy vs.0Gy=0.038,9Gy vs.0Gy=0.011).③Relative change of mtDNA and CD in lymphocytes from each patient after TBIThe increase in mtNDA content was average1.87and2.13times individually after4.5and9Gy TBI (P <0.001, Friedman test). Meanwhile, decrease in CD was0.78and0.61when4.5,9vs.0Gy cohorts respectively (P <0.001, Friedman test). Moreover, significantdifference was observed in mtDNA copy (P=0.041) and CD ratio (P <0.001) in eachpatient when comparing9Gy vs.4.5Gy exposure.④Relation between mtDNA and CD level after irradiationWhen CD values were segregated in two populations (the lower third against the twoupper thirds of the distribution), a modest inverse correlation was found reachingsignificant level for mtDNA content at different dosage (P=0.037for4.5Gy,0.048for9Gy).⑤Older people contained higher mtDNA content in general in the age range of19-56in our experiment.Conclusions1. Our small cohort research shown that differentially expressed gene cluster is morefitting to distinguish high dose (9Gy) TBI samples than average dose (4.5Gy) irradiationsamples. This method points to some possibility for molecular radiation biodosimetrydistinguishing average and high dose. 2. Bioinformatics method indicated ubiquitin-dependent protein catabolic process andGVHD pathway at transcriptional level. We focused on these two pathways and discussedtheir association with TBI preconditioning. These results provide a proof of principle foridentifying candidate genes associated with TBI sensitivity and efficiency. Moreover, ourstudy may provide further validation of the therapy and help understand the complicatedmechanisms induced by TBI preconditioning.3. Our results suggest that radiation immediately increased mtDNA content anddeclined common deletion ratio in peripheral lymphocytes of ALL patients, which may beconsidered as potential radiation biodosimetry.