| BackgroundCirc RNAs are a large group of noncoding transcripts that form covalently closed continuous loop where the 3’ and 5’ ends are joined together. Circ RNAs are generated from exons, producing exonic circ RNAs, or from introns, producing intronic circ RNAs or ci RNAs. They can also be generated from intron-containing exons,producing exon-intron circ RNAs or EIci RNAs. There is also evidence suggesting that some circ RNAs may bind to RNA-binding proteins to form RNA-protein complexes and regulate their activity. For instance, circ RNAs can bind RNA polymerase II and AGO protein. Levels of circ RNAs can also be regulated by mi RNA-dependent silencing involving Ago2-mediated cleavage. In terms of disease relevance, it is believed that circular ANRIL(c ANRIL) can regulate INK4/ARF expression and increases risks of atherosclerosis. The circular RNAs that bind mi R-7 may also be relevant to the development of Parkinson’s disease and cancer, although this remains to be further explored. Is there have any associated between Circ RNAs and development and progression of tumor? In preliminary experiments, we screened a variety of circ RNAs in tumor cells and non tumor cells and were found that circ-Amotl1(circular angiomotin like 1) in tumor cell lines expression was significantly higher than that in non tumor cell lines. Therefore, our project was focus on the relationship between circ-Amotl1 and development and progression of tumor cells.Main methods:First, stably transfected breast cancer cells MDA-MB-231 with circ-Amotl1 vector and mock control vector, anti circ Amotl1 si RNA-1, si RNA-2 and Olig to study the effect of circ-Amotl1 on growth of breast cancer cells in vitro and in vivo; Second, to understand how circ-Amotl1 functioned in tumorigenesis, we selected uniform cell population by isolating cells from the single colonies formed in agarose. To test the tumor forming capacity of the cells, we performed single cell culture, single cell colony formationassay and injected mice with6-day old colonies, 3-day old colonies and single cells. Finally,by fluorescence in situ hybridization test, immunoprecipitation test, Western blotting test, Northern blotting experiment to investigate the mechanism of circ-Amotl1 promoting the growth of tumor, especially between gene circ-Amotl1 and c-myc contact.Main results1. The effect of circ-Amotl1 on tumor cell growth(1) We measured circ-Amotl1 levels in breast carcinoma biopsies and in adjacent tissues and found that the levels of circ-Amotl1 in tumor specimen were significantly higher than those in the adjacent tissues. The expression of circ-Amotl1 was determined in a number of cancerous and non-cancerous cell lines. We observed that in eight cancerous cell lines( MCF-7, SK-Br-3, HTB-126, MDA-MB-468,MDA-MB-231, Hela, JHH-1, and U87), the levels of circ-Amotl1 was significantly higher than those in the non-cancer cell lines Ha Ca T and MCF-10 A.(2) We generated an expression construct expressing circ-Amotl1 and a mock control,and stably transfected breast cancer cells MDA-MB-231 with this construct. We also found that transfected cells exhibited a higher proliferative capacity and resistance to serum-free growth conditions as compared with the control. These cells formed larger colonies in agarose gel and displayed greater invasive capacity in Matrigel compared with the control.2. Expression of circ-Amotl1 promoted tumor growth(1) Circ-Amotl1- and vector-transfected MDA-MB-231 cells were injected subcutaneously into nude mice. As expected, MDA-MB-231 cells formed tumors slowly.Large tumor masses were excised from the mice seventeen days post-injection. In addition,we detected strong adhesion between the tumor mass and stromal tissues. H&E staining showed extensive invasion between tumor cells and muscle, and between tumor cells and bone. We repeated the tumor formation experiment and obtained similar results. The tumor sections were also subject to TUNEL staining. Extensive cell death was detected in the tumors obtained from the vector-transfected cells, but there were few death cells in the tumors formed by the circ-Amotl1 transfected cells, at significant difference. CD34 staining revealed increased density of blood vessel endothelium in the circ-Amotl1 tumors compared with the control. In addition, we observed signs of tumor cell and blood vessel invasion into bone tissues, resulting in bone tissue degradation.(2) To understand how circ-Amotl1 functioned in tumorigenesis, we selected uniformcell population by isolating cells from the single colonies formed in agarose. This would increase the homogeneous of the cell population. To understand why the growth rates of tumors formed by the circ-Amotl1-transfected cells were unexpectedly high, we performed single cell culture in 96-well tissue culture plates and found that one circ-Amotl1-expressing cell grew to a population of 11,250 cells in 10 days, which represented a 30-fold increase over the vector-transfected cells. The doubling time of the circ-Amotl1-expressing cells was between 15 to 18 hours, a growth rate much faster than any known cancer cells. In the colony formation assay, 12 days after inoculation, we detected a 142-fold increase in number of circ-Amotl1-transfected cells compared to the control.(3) To test the tumor forming capacity of the cells, we cultured single cells in agarose gel for 6 days, which allowed each cell to form a small colony containing approximately219 cells. Each colony was individually injected into nude mice. Mice were also injected with the vectortransfected cells. Each of the 219-cell colonies from one circ-Amotl1 cell began to form tumors 10 days post-injection. By days 14 and 21, some tumors caused skin ulcerations and were sacrificed. All mice were sacrificed on day 28. 17 tumors were obtained from 18 injections. Excluding one technical incident, this represented a nearly100% tumor-take rate using 6day old cell colonies. No tumor formation was observed from control cells, since 107 vector-transfected cells were needed to form a tumor in 28 days.We then injected mice using 3-day old colonies, each of which contained approximately 9 cells. 28 days post- injection, some tumors began to cause skin ulcerations and the mice were sacrificed. On day 31, all mice were sacrificed. 17 tumors were obtained from 20 injections. In the controls, mice were injected with different number of cells, none of which produced signs of tumor formation. Considering a 85% success rate of colony harvest and transfer, this represented 100% rates of tumor formation injected from the 3-day colonies.(4) Furthermore, we injected mice with single cells: 1 cell per injection. 20 days after the injection, some cells formed visible tumors, some tumors were excised 32 and 35 days after the injection. All mice were sacrificed on day 40, and we obtained 60 tumors from 180 injections. Considering 80% success rates of cell harvest and transfer, this represented a41% tumor-take rate using single-cell subcutaneous injections.These results suggested that every circ-Amotl1-transfected cell had the potential to form an individual tumor. In the parental MDA-MB-231 cells, 1x107 cells typically require30 days post-injection to form any detectable tumor. This suggests that the tumorigenicity of MDA-MB-231 cells have been reprogrammed by circ-Amtol1 overexpression. The most notable characteristic was the fast proliferation rate of every cell without sign of differentiation. This cancer burgeoning cells(CBC) may represent a small population of cells in cancers, but have not been recognized.3. Mechanism of circ-Amotl1 promote tumor growthTo study how circ-Amotl1 may trigger this highly tumorigenic phenotype, we examined potential interaction between circ-Amotl1 and c-myc, which is a transcript factor regulating expression of many oncogenes.(1) To explore the possibility that circ-Amotl1 could pull-down c-myc and vice versa,we tested whether the probe used for Northern blot had the capacity to pull-down circ-Amotl1. While the total inputs of circ-Amotl1 and the probe were similar), streptavidin beads pulled-down significantly higher levels of circ-Amotl1 compared to the controls.Using this probe, much higher levels of c-myc were pulled-down.(2) We analyzed levels of some c-myc targets in the circ-Amotl1- and vector-transfected cells and found that a number of c-myc targets were up-regulated in the circ-Amotl1-transfected cells cultured at subconfluence and full confluence. We analyzed the effect of circ-Amotl1 on the interaction of c-myc to the promoters and found that ectopic expression of circ-Amotl1 enhanced the binding affinity of c-myc to the promoters including HIF-1, Cdc25 a, ELK-1, and JUN, using the primers amplifying these promoters.(3) To validate the effect of c-myc in mediating circ-Amotl1 function, we transfected the c-myc null cells HO15.19 with circ-Amotl1 or a control vector. Although the transfected cells expressed significantly higher levels of circ-Amotl1, there was no difference in rates of cell proliferation, confirming the role of c-myc in mediating circ-Amotl1 functions. To validate these results, we performed silencing assays. Transfection with the si RNAs targeting circ-Amotl1 decreased the levels precipitated by anti-c-myc antibody, while the levels of circ-Amotl1 was silenced by the si RNAs. The probe which specifically targeted circ-Amotl1 was shown to pull-down significantly lower levels of circ-Amotl1 m RNA and c-myc protein when circ-Amotl1 was silenced. Finally, silencing circ-Amotl1 significantly down-regulated c-myc targets compared to the control oligo.(4) We isolated total RNA from cytoplasm and nuclei and measured the levels of circ-Amotl1. While the m RNAs of GAPDH and linear Amotl1 m RNA were found mainly in cytoplasm, higher levels of circ-Amotl1 were detected in nuclei than incytoplasm. Significantly higher levels of the over-expressed circ-Amotl1 were also detected in the nuclei relative to the cytoplasm. Western blot using fractionated samples showed a similar distribution of c-myc. Silencing circ-Amotl1 produced an opposite pattern of c-myc distribution. Confocal microscopic analysis detected mainly cytoplasmic c-myc in the vector-transfected cells, but high intensity of nuclear distribution of c-myc in the circ-Amotl1-transfected cells. Using a probe specifically recognizing circ-Amotl1, we detected co-localization of circ-Amotl1 and c-myc in the nuclei of the circ-Amotl1 cells.Silencing circ-Amotl1 facilitated cytoplasmic distribution of c-myc. Taken together, our study showed that the ectopic expression of circ-Amotl1 induced nuclear translocation of c-myc. It has been reported that nuclear c-myc is readily degraded. In our study, high levels of c-myc were translocated to the nucleus, but we did not detect decreased level of total c-myc. This suggested that c-myc was localized to the nucleus of circ-Amotl1 cells,but co-localized with circ-Amotl1, preventing degradation.Conclusions:1. The levels of circ-Amotl1 in tumor specimens were significantly higher than those in the adjacent tissues.2. MDA-MB-231 transfected by circ-Amtol1 vector has a higher proliferative capacity which had no reported before. The most notable characteristic was the fast proliferation rate of every cell without sign of differentiation.3. Large tumor masses were excised from the mice seventeen days post-injection. In addition, we detected strong adhesion between the tumor mass and stromal tissues. 100%rates of tumor formation injected from the 3-day colonies and 6-day colonies. Single cell injection assay results suggested that every circ-Amotl1-transfected cell had the potential to form an individual tumor.4. High levels of c-myc were translocated to the nucleus, but we did not detect decreased level of total c-myc. This suggested that c-myc was localized to the nucleus of circ-Amotl1 cells, but co-localized with circ-Amotl1, preventing degradation. |
PDF Full Text Request