Novel Molecular Diagnosis Of Prostate Cancer: LncRNA383and Gene Fusion USP9Y-TTTY15

Background and ObjectiveCurrent guideline for PCa screening is based on serum PSA which is under warm debate.However PSA protein can be produced and excreted from normal prostate cell, which maycause false positive and false negative issue. It has been proven that single use of serumPSA cannot diagnose PCa precisely. Especially in PSA gray zone （4-10ng/ml）, thepositive biopsy rate is only17%-23%, but false positive biopsy rate is up to65%whichmeans over2/3candidate underwent unnecessary biopsy. Even the negative biopsy resultcannot exclude the possibility of PCa. Repeat biopsy for elevated PSA patient withnegative initial biopsy result cannot promote the positive biopsy rate beyond10%.With thepermission of clinical usage of PROGENSA PCA3by US FDA, molecular diagnosis ofPCa enters in non-invasive RNA-level diagnostic era. As reported, the sensitive ofdiagnosis is67%, and specificity is83%. Meanwhile, the fusion of TMPRSS2gene andERG gene happened only in prostate cancer patient with50%incidence in western.Because of the reasons above all, we need find a new diagnostic biomarker with both highspecificity and high sensitive, to improve the low positive rate of prostate biopsy. We aimto investigate the diagnostic possibility of long-non coding RNA383（lncRNA383） andgenefusion USP9Y-TTTY15which we discovered by RNA-sequencing initially.Materials and MethodsSecond Generation Transcript-sequencing has been used to select non-coding RNA andgenefusion candidates which were differently expressed among PCa tissue and adjacentnormal tissue. The results of transcript sequencing were verified with amplified amount ofsamples by RT-PCR. In the next retrospective study, we recruited the urine sediment fromPCa patient and negative biopsy patient. RT-PCR has been used to investigate thediagnostic value of lncRNA383and genefusion USP9Y-TTTY15in urine. In the nextprospective study, we recruited patients who are going to take the prostate biopsy becauseof elevated serum PSA level. RT-PCR still been used to confirm the value of lncRNA383and genefusion USP9Y-TTTY15in patients’ urine. Logistic regressive analysis have been taken to establish the model of diagnostic panel to improve the sensitive and specificity ofPCa diagnosis.ResultAccording to the result of Second Generation Transcript-sequencing in14pairedprostatectomy specimens; we discovered137differently expressed long non-coding RNA.Two criteria has been used to select the candidate non-coding RNA. First of all,FDR≤0.001and doubled difference expression; secondly, the expression trend should besame in more than10specimens. Finally, we discovered that lncRNA383expressed higherin PCa tissue than adjacent normal tissue. In the next retrospective study,lncRNA383canbe detected in urine sediment. Besides that, it has high value for diagnosis of PCa in urine（AUC=0.858，cut-off=0.877,Se=68.8%,Sp=97.8%）.In the next prospective study,although the efficiency of diagnosis reduced than before（AUC=0.631，cut-off=0.776，Se=85.4%%， Sp=43.5%）, it was also similar as PCA3and even better thanTMPRSS2-ERG.We also found there is no relationship between lncRNA383and serumPSA, DRE, Gleason Score separately. It may be limited the usage of lncRNA383asprostate cancer diagnosis biomarker.Meanwhile, we also found one well-known genefusion-TMPRSS2-ERG and37firstreported genefusions from the result of Second Generation Transcript-sequencing in14paired prostatectomy specimens. The frequency of gene fusion was21.4%both inTMPRSS2-ERG and USP9Y-TTTY15. The fusion happened between exon3of USP9Yand exon4of TTTY15. In the next tissue verifying study, we used54paired PCa tissue andadjacent normal tissue to confirm the result of transcript-sequencing. The frequency was35.2%. Interestingly, the USP9Y-TTTY15fusion transcript did not seem to have any openreading frames （ORF） based on ORF prediction tools such as ‘Six-Frame Translation’,indicating that it may be an ncRNA. In the next retrospective study, we confirmed theexpression of USP9Y-TTTY15in the urine sediment of PCa and negative biopsy patientand the value of prognosis （AUC=0.802, cut-off=0.894,Se=85.7%,Sp=73.3%）. In thenext prospective study, USP9Y-TTTY15maintained the same value of diagnosis（AUC=0.802, cut-off=0.894, Se=85.7%, Sp=73.3%）.Especially in the PSA gray zone, italso have ability to forecast positive biopsy result （AUC=0.774,cut-off=0.845,Se=81.8%，Sp=62.2%）.With the combination with serum PSA and fPSA, the multiplediagnostic model have been established as below: Y=71.354（XUSP9Y-TTTY15）+0.213（XPSA）-1.754（XfPSA）-34.286. The sensitive and specificity of multiple diagnosticpanels reached to66.7%and91.9%separately. Above all, the new multiple diagnosticpanels displayed better than any combinations with PCA3, TMPRSS2-ERG and otherbiomarkers. As an ideal diagnostic model, it will promote the positive rate of prostatebiopsy. ConclusionlncRNA383and genefusion USP9Y-TTTY15can be used as ideal biomarkers individually.Besides that, genefusion USP9Y-TTTY15can also predict the biopsy result in PSA grayzone population. Gleason score≤6PCa can be well forecast by it. In order to increase theefficiency of PCa diagnostic biomarkers, the multiple diagnostic model have beenestablished as below: Y=71.354（XUSP9Y-TTTY15）+0.213（XPSA）-1.754（XfPSA）-34.286.The sensitive and specificity of multiple diagnostic panels reached to66.7%and91.9%separately. Above all, the new multiple diagnostic panels displayed better than anycombinations with PCA3, TMPRSS2-ERG and other biomarkers. As an ideal diagnosticmodel, it will promote the positive rate of prostate biopsy.