Screening And Identification Of Phage Antibody Of Single Chain Variable Fragment Of Human Hepatocellular Carcinoma

There have been two methods to get single chain variable fragment antibody (scFv) against hepatocellular carcinoma(HCC scFv).The first, HCC scFv can be redeveloped from monoclonal antibody (mAb) against HCC (HCC mAb). Because HCC scFv and its parent mAb can respond to the same epitopes of HCC antigen, the biological characteristics of HCC scFv depend on its parent mAb, but it is not a really new antibody.The second, HCC scFvs can be found from HCC scFv library constructed by immuning mice with human hepatocellular carcinoma lines (HCC lines). It is an important approach to obtain new HCC scFvs by panning the scFv library containing all scFv to all HCC antigens in theory.In recent years, HCC mAb has been successfully redeveloped into HCC scFv, it can inhibit HCC in vitro and be fused into a scFv-immunotoxin protein. Now the attention of research for HCC scFv was payed to construct HCC scFv library, there were some HCC scFv libraries containing 2.14× 10~6(2001, Yang DH), 5.3 × 10~7(2004,Yang DH), 2× 10~5(2002 Li SH), 1.1 × 10~6(2004, Nian H) different clones.It was a rontine method for panning and identication HCC scFv that the positive clones for HCC antigen in library were gragually enriched by many selection rounds of binding-washing-growing scFv library, and then clones from the last selection round were respectively selected for assay. Up to now, the characters of HCC scFvs wereonly identified by human HCC lines and normal liver cells lines (NLC lines) (2002 Li SH)or HCC lines and other cancer cells lines (OCC lines)(2004, Nian H), but no valube HCC scFv has been found.At present, above methods of panning and identification for HCC scFv have some defects, such as random pick with blindness, larger workload, smaller assayed clones once, It can remove HCC scFvs binding NLC lines, but not exclude them binding OCC lines to assay them by HCC and NLC lines. Simultaneously, it can remove them binding OCC lines, but not get rid of them binding NLC lines to assay them by HCC and OCC lines. Due to difference between HCC lines and HCC tissue, there are differences between their epitopes, so the HCC scFvs which have higher specificity for HCC1 lines do not always have the same specificity for HCC tissues. In order to get HCC scFvs with higher specificity for human HCC tissues. We established colony lift assay, performed large-scale panning for clones of the HCC scFv library constructed in our laboratory by combining colony lift assay with random pick, HCC scFvs were identified with three kinds of HCC lines (SMMC-7721, Bel-7402 and HepG-2), carcinoma of stomach cell lines (SCG-7901), and cancer tissues of liver, large intestine cell lines (CaC02), NLC lines (L-02). Higher HCC antigen-specific scFv were identified further with human tissues of normal liver. HCC, carcinoma of stomach, large intestine, esophagus and lung.Part 1 The phage display HCC scFv library panned by colony lift assay and random pickObjective: To establish colony lift assay and screen rough HCC scFv library by colony lift assay and random pick and analyse the efficiencies for them. Method: The positive clones for HCC antigen in HCC scFv library were enriched by 1-4 selection rounds (SR) with HCC lines, and then panned rough by colony lift assay and random pick. The clones assayed in every selection rounds and the positive clones among assayed clones were calculated.Result: The positive clones for HCC antigen in 1 to 4 selection rounds accounted respectively for 1% (9/912), 5.1% (15/296X 12.1% (30/248), 64.5% (169/262) for colony lift assay, and 0% (0/96), 4.2% (4/96), 8.3% (8/96), 64.4% (121/188) for random pick. There was no significantly difference between two methods (SRI, p=1.0, SR2, X2=0. 01, p=0. 93, SR3, X:>=1.0, p=0. 32 and SR4, XL>=0. 01, p=1.0). It suggested that the rates of positive clones for the same specimen by colony lift assay were consistent with random pick. When colony lift assay was used to screen HCC scFv library, it did not need to pick library clones from a plate, and all library clones growing singly in a plate could be detected at the same time. However, random pick was used to screen this library, it had to pick and detect library clones one by one from a plate. So the latter had to need obviously larger workload than the former. 356 positive clones (223 clones for colony lift assay, 133 clones for random pick) were obtained from 2194 assayed clones (1718 clones for colony lift assay, 476 clones for random pick) by two methods.Conclusion: (T)we established a new method of colony lift assay for panning HCC scFv library. Colony lift assay and random pick had identical efficiencies. The former also was an effective method in panning HCC scFv library. ?All clones growing singly in a plate could be detected by colony lift assay at the same time, it had some characteristics of low level workload, high efficiency and fast pace, ?large-scale clones of HCC scFv library could be sreened once, and then successful probability of obtaining scFv specific for HCC would be improved, by colony lift assay or combination of it with rendom pick.Part 2 Screening of HCC antigen-specific of phage display HCC scFvs and sequencing of their genesObjective: To screen the higher specific scFvs for HCC antigen with SMMC-7721,Bel-7402 and HepG-2, amplify their genes and analyse their gene sequences.Method: The titers of HCC scFv, generated by 356 positive clones, binding SMMC-7721, Bel-7402, HepG-2, L-02, SCG-7901 and CaCO2 were detected by ELISA. The genes of high specific scFvs for HCC antigen were amplified by PCR and their gene sequences were analysed.Result: 10 higher specific phage scFvs for HCC antigen were obtained from HCC scFvs generated by 356 positive clones. The lowest titers of them binding SMMC-7721.Bel-7402 or HepG-2 were less than 1 of 128 to 1 of 256 of those binding to L-02, SCG-7901 or CaCO2. 9 of 10 genes of higher specific scFvs were successfully amplified by PCR, the sizes of their genes were about 750 bp.. six of nine gene sequences except three genes were successfully obtained, five genes were the same sequence except one. All gene sequences contained Sfil and Notl. Linker. Tag E sequences. These results indicated that these clones carried the genes of specific scFv of HCC lines.Conclusinon: ?The clones expressing high specfic HCC scFvs existed in low frequency in HCC scFv library, large-scale screening of library clones would be an importment way to enhance the propability to obtain really valuble high specfic HCC scFvs,.?. 10 higher specific scFvs for HCC antigen were obtained, the lowest titers of them binding to HCC lines were less than 1 of 128 to 1 of 256 of those binding to L-02, SCG-7901 or CaCO2. 9 of 10 genes of higher specific scFvs were amplified by PCR, the sizes of their genes were about 750 bp. (3) Squence analysis confirmed that 2 different gene squences of scFvs specific for HCC lines were successfully screened.Part 3 Identification of specificity of HCC antigen-specific scFvs for tissues of human liver cancerObjective: To study whether the specific scFvs to HCC lines would have specificity for tissues of human heptocellular carcinoma.Method: It was detected by immunohistochemistry whether the scFvs with high specific for hepatocellular carcinoma cell lines had responded to tissues of humanhepatocellular carcinoma, the tissues of normal liver, carcinoma of stomach, large intestine and esophagus were used as control. The rates of positive responses were calculated and compared with control. Moreover, the rates of the positive response of tissues of HCC to HCC scFvs were compared with that of AFP mAb. Result: The rates of positive responses to HCC scFv for the tissues of hepatocellular carcinoma, liver cirrhosis, hepatitis, carcinoma of stomach, carcinoma of large intestine, carcinoma of esophagus and carcinoma of the lung, normal liver were respectively 35.5%(22/62)> 6.7%(2/30), 6.7%(2/30), 5.0%(l/2(^ 5.0%(l/20)^ 0% (0/20X 0%(0/20)> 0% (0/12) .it was significantly higher in HCC tissues than in tissues of normal liver (X2=4.48, P=0.034), liver cirrhosis (X2=7.28, P=0.007). hepatitis (X2=7.28, P=0.007), carcinoma of stomach (X2=5.53, P=0.019), large intestine (X2=5.53, P=0.019), esophagus (X2=7.98, P=0.005) and lung (X2=7.98, P=0.005). There were not significantly differences among other tissues. The results suggested that HCC scFv panned by us had high specific for HCC tissues and low specific for other tissues. There were 56.5% (35/62) positive responses to AFP mAb for the same 62 HCC tissues, it was significantly higher than HCC scFv (X2=6.01, PO.05). It indicated that this HCC scFv was probably different from AFP mAb. But the other one had no specific reactivity with HCC tissues. Conclusion: 1. Of 2 high specific HCC scFvs, there was a difference between them for rates of positive reactivity with HCC tissues. One of them still had specific reactivity with HCC tissues, except for other one. 2. The rate of positive reactivity of high specific HCC scFv with HCC tissues was lower than that of AFP mAb. It probably had no relation to AFP mAb. 3. It was helpful to obtain valuble HCC scFvs that the combiation of HCC lines with HCC tissues was used to identify HCC scFvs.