Construction Of Aptamer-based Liquid Crystal Sensor And Its Application In Tumor Marker Detection

Cancer is a major global public health problem and the third leading cause of death in our country and the second leading cause of death in the United States.In the face of high cancer incidence and mortality,effective early diagnosis is undoubtedly an important measure to improve the survival time of cancer patients and reduce their mortality.As a kind of biomarker,serum tumor marker has attracted much attention in the field of early clinical diagnosis of cancer due to its easy sample availability.How to detect tumor markers in serum more quickly,with high sensitivity and wide range has always been an important research content in the field of biological detection.Aptamer is a oligonucleotide synthesized in vitro,which can specifically recognize multiple target substances.The aptamer sensor was studied by using aptamer as the molecular recognition part（sensitive element）and the conversion part（transducer）,which provided an effective way for quantitative detection of tumor markers in serum.Liquid crystal has been widely used in sensor research because of its unique properties of both liquid fluidity and solid anisotropy.In this paper,aptamer liquid crystal sensor was constructed with aptamer as molecular recognition element and liquid crystal molecule as transducer.It was used for detection of three tumor markers in serum,prostate specific antigen（PSA）,carcino-embryonic antigen（CEA）and mucin antigen 1（MUC1）.On this basis,automatic extraction of bright area coverage（Br）feature value of liquid crystal sensor polarizing microscopic image and pattern recognition classification model of liquid crystal sensor polarizing microscopic image based on support vector machine are studied respectively.The ultimate purpose is to reduce the error caused by manual image selection and improve the accuracy of tumor marker detection.The main contents of the paper are as follows:（1）Aptamer-based Liquid crystal sensor construction and PSA detection without labeling.The adaptive liquid crystal sensor was constructed and the Br automatic extraction system for polarizing microscopic image was studied.The apnetic liquid crystal sensor is composed of substrate（lower slide）,copper carrier mesh and upper slide.The optimized conditions of the apnetic liquid crystal molecule 5CB can be induced by DMOAP/APTES-GA-Aptamer and upper slide DMOAP.In the absence of PSA,the liquid crystal molecules in the sensor maintained a vertical orientation and presented a"dark"response image under polarized light microscope.When PSA is present,the specific binding of the aptamer and PSA in the sensor will affect the original vertical alignment of 5CB molecule,resulting in the polarizing microscopic image from"dark"to"bright".Polarizing image processing program established by MATLAB is used for image correction and segmentation,feature extraction and calculation of Brof polarizing microscopic image,eliminating pixel interference in the image,making PSA detection process more rapid and less human error.The sensor has good detection specificity.In the range of 1 fg/m L～1μg/m L,there is a linear correlation between polarized microscope image Br and PSA concentration logarithm(Br=7.86998 Log C_PSA+6.50541),and the linear correlation coefficient R²reaches 0.99413.The limit of detection was 0.58 fg/m L.The effective detection of SERUM PSA was carried out.In addition,AFMwas used to analyze the surface morphology by the substrate,which supported the research conclusion.（2）MUC1 and CEA were detected simultaneously by aptamer-based liquid crystal sensor.On the basis of（1）,the aptamer liquid crystal sensor for simultaneous detection of CEA and MUC1 tumor markers in serum was studied to investigate the sensitivity and specificity of the sensor,the detection effect in actual samples,and the AFM characterization of each molecule modified by the substrate.The results showed that the concentration range of MUC1 detected by the sensor was 1 fg/m L～1μg/m L,and there was a linear relationship between the concentration logarithm and Br.The linear equation was Br=8.16364Log C_MUC1+11.60762,the linear relationship coefficient R²was 0.9828,and the detection limit was 0.47 fg/m L.When the CEA concentration range was 100 fg/m L～100ng/m L,the linear relationship was good(Br=12.00628Log C_CEA+31.33056,R²=0.98406),and the sensor detection limit was 0.038 pg/m L.The sensor showed good specificity for simultaneous detection of MUC1 and CEA.（3）A pattern recognition classification model of liquid crystal sensor polarizing microscopic image based on support vector machine.According to the idea of machine learning,a SVM based pattern recognition classification model was constructed,and GLCM was used to extract texture features from polarized microscopic images.By training the classification model with samples obtained in the study,the computer can identify whether the detection result of the sensor image is below or above the clinical threshold.When a new polarized-light microscopic image is introduced,the classification model can accurately classify it and determine whether the detection result is above or below the clinical threshold of the target.The detection process is very rapid.In addition,by generating confusion matrix,the classification accuracy of the model is high,reaching0.9222（upper limit is 1）,which proves that the model has good application performance.The adaptive liquid crystal sensor constructed in this paper has high sensitivity,specificity,wide linear detection range and good stability.After replacing the adaptor modified by the sensor base,the liquid crystal sensor can also be applied to the detection of other target substances,and the image processing system and pattern recognition model constructed in this study can also be applied to other liquid crystal biosensors with complete the methods.The sensor has low cost,convenient and quick detection process,and can detect tumor markers without additional markers,which has great reference value and development potential in the rapid detection of tumor markers in clinical needs.