Study On The Clinical Characteristics Of Childhood Hemophagocytic Lymphohistiocytosis And The Pathogenesis Of A New Mutation

Background and objectivesHemophagocytic lymphohistiocytosis?HLH?is a life-threatening syndrome characterized by high gradefever of unknown origin,poor response to broad-spectrum antibiotic treatment,rapid progress,and multiple organ damage.Without effective treatment and supportive care,patients usually undergo death.The current diagnosis and treatment of HLH are based on the HLH-2004 criteria of the International Histiocyte Society.However,the criteria may result in delayed diagnosis,and cannot predict disease severity.HLH comprises primary and secondary forms.Patients with primary HLH have a family history or an identifiable genetic defect while patients with secondary HLH have no genetic defects and are often associated with viral infection,underlying rheumatologic disorders or malignancy.Primary HLH have a clear risk of recurrence and fatal outcome,and allogeneic hematopoietic stem cell transplantation?HSCT?is the only choice to cure.Soit is important to identify patients with primary HLH.In order to improve the early diagnosis and survival of children with HLH,and to establish the diagnostic methods of primary HLH and clarify the mechanisms of HLH,we conducted this study including the following three parts:?1?Analysis of clinical characteristics and gene mutation of children with HLH;?2?study on the diagnostic methods of primary HLH;?3?study on the pathogenesis of new HLH related gene mutation.Methods1 Analysis of clinical characteristics and gene mutation of children with HLH1.1 Analysis of clinical characteristicsof children with HLH:During the last seven years from January 2010 through December 2016,patients with HLH who were admitted in the Hematology/Oncology department of our hospital were enrolled.Patients' data including demographic characteristics,clinical manifestations,regular laboratory findings,serum Th1/Th2 cytokine and T/B/NK cell percentages at the time of diagnosis,and the treatment outcome were collected.Continuous parameters were compared by using Mann-Whitney U-tests;counting parameters were compared by using X2 test.Multivariate analyses to identify prognostic factors for predicting 30-day OS were carried out by using Cox proportional hazard regression model.30-day OS was estimated with the Kaplan-Meier method and compared using the log-rank test.1.2 Gene mutation in children with HLH:Patients'HLH-related genes weresequenced by Sanger sequencing or the Next Generation Sequencing;SNPs were excluded by searching literature and database;functional consequences of mutations were predicted in silico analysis.2 Thestudy on the diagnostic methods of primary HLH:The expression of HLH-related proteins in NK cells were detected by flow cytometry,including:perforin,Munc13-4,syntaxin 11,Munc18-2,SAP,XIAP,Rab27a and AP3?1.NK cells were stained after fixtion and permeabilizaion fluorochome-labeled antibodies aganst above proteins.NK cell degranulation activity was detected by flow cytometry after incubation with K562.Effector-target ratio and time gradient were set to select the optimal experimental conditions;NK cell cytotoxic activity was determined by flow cytometry;the expression of CD3 3 and CD 147 on K562 were measured by flow cytometry,and effector-target ratio gradient was set to select the optimal condition.The reference ranges of these three tests were established according to the results of healthy children.These results were compared among patients with gene mutation,thosewithout gene mutation and the healthy children.3 Study on the pathogenesis of new HLH related gene mutation3.1 Construction of UNC13Dwild-type and mutant lentiviral expression vectors:cDNAs encoding wild-type and mutant human UNC13D were generated by reverse transcription from RNA of patients' peripheral blood.Both WT and mutant cDNAs were independently cloned into TA clone vector.Their sequences were confirmed by DNA sequencing.And then,both cDNAs were independently subcloned into pGV348 to help generate recombinant lentiviruses.Human embryonal kidney 293T cells were transfected with the pGV348/WT-UNC13D and pGV348/Mut-UNC13D constructs together with Rev,Vsvg and Gag-pol.Lentivial productions were concentrated using the lenti-X concentrator reagent.Viral titer was measured using limiting dilution assay.3.2 Effect of mutant UNC13D on cytotoxic T cell function:The PBMCswere enriched from human peripheral blood using Ficoll-hypaque density gradient centrifugation method,and CTLs were isolated by magnetic bead negative sortingfrom PBMCs.CTLs were activated and expanded by co-culture with anti-human CD3/CD28 beads and recombinant IL-2.We analyzed the growth characteristics of CTLs under the culture condition in vitro.Separated CTLs activated following the methods above,were infected with the lentivirus.Successful transfection and expression of target gene in CTLswere evaluated by fluorescence microscopy,flow cytometry and Western Blot.CTLs degranulation and cytotoxicity activity were used to observe the effect of gene mutation on protein function.Results1 Analysis of clinical manifestations,laboratory findings and gene mutations of children with HLH1.1 Analysis of clinical manifestations and laboratory findings of children with HLH:The onset age peak of most pediatric HLH was between 0 to 4 years old,and those over 10 years were rare,Epstein-Barr virus was the most common pathogen in children with HLH.The overall survival of our patients was 61.0%.The clinical manifestations and laboratory findings of patients were high-grade fever?100%?,hyperferritinaemia?100%?,cytopenias?100%?,liver damage?100%?,bone marrow hemophagocytosis?96.5%?,hepatomegaly?83.2%?,hypofibrinogenemia?76.4%?,splenpmegaly?67.5%?,hypertriglyceridemia?48.7%?,respiratory involvement?41.6%?,superficial lymphadenopathy?21.2%?,skin eruption?18.6%?,digestive system symptoms?15.9%?,jaundice?7.1%?,and central nervous system symptoms?6.2%?.When compared with healthy children,HLH patients had higherlevels of INF-y?40.7?4.6-1477.5?pg/mL vs.3.4?1.4-14.6?pg/mL,P<0.001?,IL-6?604.4?4.6-5000.0?pg/mL vs.3.0?1.7-8.2?pg/mL,P<0.001?and IL-10?210.2?2.4-5000.0?pg/mL vs.3.7?1.0-11.8?pg/mL,P<0.001?,and lower percentages of B cells?12.1?3.0-45.5?%vs.22.2?15.7-32.3?%,P<0.001?and NK cells?4.2?0.2-26.4?%vs.9.8?1.6-15.1?%,P<0.001?,respectively.Alb?28.5g/L?HR=3.672,95%Cl 1.281-10.533,P=0.015?,ChE?4283.0 U/L?HR=15.714,95%CI 2.078-118.840,P=0.008?,IL-10?456.0 pg/mL?HR=2.946,95%CI 1.135-7.744,P=0.027?were independent risk factors of early death inchildren with HLH.The risk of early death was 54-fold increase in patients with three risk factors?HR=54.17,95%CI 7.28-403.17,P<0.001?,and 9.93-fold increase in patients with two risk factors?HR=9.93,95%CI 2.55-36.68,P=0.001?when compared to that in patients with zero to one risk factor.1.2 Gene mutations in children with HLH:Mutations in HLH-related genes were found in 40.2%?33/82?patients,of whom,4 cases were hemizygote,21 cases were monoallelic mutation,4 cases were biallelic mutation,and 4 cases were compound heterozygote.The most frequently detected mutations were UNC13D and LYST?24%?,followed by SXTBP2.Thirty eightgene mutation sites were found,including 4 splice site mutations,1 deep intron mutation,1 frameshift mutation,3 nonsense mutations and 29 missense mutations,of whom 22 were new mutations.Among the missense mutations,13 were predicted to be pathogenic.2 Thestudy on the diagnostic methods of primary HLH:HLH related protein expression in HLH patients and healthychildren were comparable.Patients with PRF1 mutation had significantly lower perforin expression than thosewithoutPRF1 mutation and healthy children?P=0.009 andP=0.021,respectively?.Patients with SH2D1A mutation had significantly lower SAP expression than thosewithoutSH2D1A mutation and healthy children?P<0.001 for both?.Effector-taget ratio of 1:1 and incubation time of 3 hours was the optimal conditions for degranulation assay.The reference ranges of degranulation were 8.4-37.8%.Patients with degranulation related gene mutation had significantly lower degranulation values than those without degranulation related gene mutation and healthy children?9.3?0.5-29.5?%vs.17.5?4.9-33.8?%,P<0.001;9.3?0.5-29.5?%vs.19.8?8.4-36.8?%,P<0.001?.The effector-taget ratioof 10:1 was the optimal condition for NK cell cytotoxic activity assay.NK cell cytotoxic activity was associated with percentage of NK cells?P = 0.014,r = 0.438?,and the reference ranges were 10.8-74.2%.NK cytotoxic activity of patients in acute phase of HLH were significantly lower than those of patients in remission and healthy children?2.9?0.6-7.7?%vs.7.5?2.8-25.0?%,P<0.001;2.9?0.6-7.7?%vs 9.8?3.1-23.3?%,P<0.001?,while children with PRF1 and/or degranulation related gene mutations had lower NK cytotoxic activity than healthy children?21.9?3.8-43.1?%vs.39.6?10.9-74.2?%,P=0.003?.3 Study on the pathogenesis of a new HLH related gene mutation3.1 Construction of UNC13Dwild-type and mutant lentiviral expression vectors:Results from gel electrophoresis and sequencing both showed that wild-type and mutant cDNAs were successfully amplified and ligated to the lentiviral expression vector.The lentiviral particles carrying wide-type and mutant gene were obtained from culture supernatant of 293T cell line.Titer for concentrated lentiviral solution was more than 108 based on the limiting dilution assay.3.2 Effect of mutant UNC13D on cytotoxic T cell function:CTLs could be activated and expanded when co-culture and stimulated with anti-human CD3/CD28 beads and IL-2 in culture medium,and more than 120-fold of amplification could be obtained.Lentiviruscontaining the target genes could be used to successfully infect the CTLs.Results from flow cytometry and Western Blot both showed that wild type and mutant protein were successfully expressed in CTLs.Degranulation in P1:Control 9.8,UNC13D-WT 13.8,UNC13D-Mut 8.2;in P2:Control 9.1,UNC13D-WT 16.0,UNC13D-Mut 6.7;in P3:Control 8.3,UNC13D-WT 15.2,UNC13D-Mut 6.9.Cytotoxic activity in P1:Control 59.7%,UNC13D-WT 97.4%,UNC13D-Mut 47.7%;in P2:Control 48.5%,UNC13D-WT 93.8%,UNC13D-Mut 48.2%;in P3:Control 55.4%,UNC13D-WT 95.2%,UNC13D-Mut 44.2%?Conclusions1.The onset age peak of most pediatric HLH is between 0 to 4 years old,patients over while 10 years are rare.Epstein-Barr virus is the most common pathogenic factors.2.Fever,cytopenias,hyperferritinaemia,EBV positiveand liver damage are the significantfeatures of children with HLH.Patients with these features should be suspected of HLH,and warranted relevant laboratory examination.3.Serum IL-10 and IFN-? levels can be used to early diagnose HLH.Alb?28.5g/L,ChE?4283.0 U/L and IL-10?456.0 pg/mL are the three independent risk factors of early death.4.40.2%of pediatric HLH harbores gene mutations,in which UNC13D and LYSTare the most frequently detected genes with mutation,followed by STXBP2.Most patients wereheterozygous,which they cannot be used solely to diagnose the primary HLH based on the results of gene sequencing alone.5.Flow cytometry assay of HLH-related proteins,degranulation and cytotixic activity of NK cells are successfully established.Low expression of perforin and SAP expression suggeststhe presence of PRF1 and SH2D1A gene mutation.Patients with normal NK cell degranulation and cytotoxic activity can be exclueded as primary HLH,while thosewith abnormal NK cell degranulation and cytotoxic activity should be highly suspected ofprimary HLH.6.UNC13D:c.2295₂298delGCAG,p.Glu765Aspfs*27 is a pathogenic mutant site and lead to encode a nonfunctional protein.7.Successful establishment of CTL-based in vitro validation method for pathogenetic mutation site.8.The diagnosis of primary HLH should be made according to gene sequencing combined with protein expression,functional assay,and mutation site pathogenetic validation approches.