AbstractNuclear activation of Wnt/尾-catenin signaling is required for cell proliferation in inflammation and cancer. Studies from our group indicate that 尾-catenin activation in colitis and colorectal cancer (CRC) correlates with increased nuclear levels of 尾-catenin phosphorylated at serine聽552 (p尾-Cat552). Biochemical analysis of nuclear extracts from cancer biopsies revealed the existence of low molecular weight (LMW) p尾-Cat552, increased to the exclusion of full size (FS) forms of 尾-catenin. LMW 尾-catenin lacks both termini, leaving residues in the armadillo repeat intact. Further experiments showed that TCF4 predominantly binds LMW p尾-Cat552 in the nucleus of inflamed and cancerous cells. Nuclear chromatin bound localization of LMW p尾-Cat552 was blocked in cells by inhibition of proteasomal chymotrypsin-like activity but not by other protease inhibitors. K48 polyubiquitinated FS and LMW 尾-catenin were increased by treatment with bortezomib. Overexpressed in vitro double truncated 尾-catenin increased transcriptional activity, cell proliferation and growth of tumor xenografts compared to FS 尾-catenin. Serine聽552- 聽alanin substitution abrogated K48 polyubiquitination, 聽尾-catenin nuclear translocation and tumor xenograft growth. These data suggest that a novel proteasome-dependent posttranslational modification of 尾-catenin enhances transcriptional activation. Discovery of this pathway may be helpful in the development of diagnostic and therapeutic tools in colitis and cancer. Introduction尾-catenin is a cytoplasmic protein that participates in intercellular adhesion and Wnt-mediated transcriptional activation (for review see1). Wnt/尾-catenin - induced gene transcription plays a central role in self-renewal, proliferation, differentiation, polarity, morphogenesis, and development2,3,4. Aberrant Wnt/尾-catenin signaling is found in several tumors, including colorectal cancer (CRC)4,5. 尾-catenin signaling is increased in over 90% of CRC due to mutations in either 尾-catenin exon 3 or adenomatous polyposis coli (APC), believed to enhance 尾-catenin stability by reducing degradation6,7. Ultimately 尾-catenin translocates into the nucleus and binds transcription factor TCF4 (T cell factor 4) to drive transcription of Wnt regulated genes6,8,9,10,11.The primary structure of 尾-catenin is composed of N and C terminal regions and a central core of 12 armadillo repeats spanning residues 134鈭?78. Cadherins, APC and TCF family transcription factors bind to 尾-catenin within the core region, whereas GSK3尾 and 伪-catenin bind sites within N terminal amino acids12. Phosphorylation of N terminal sites targets 尾-catenin for degradation in the ubiquitin鈥損roteasome pathway in the cytosol7. Despite the association of N terminal phosphorylation to degradation, the roles of 尾-catenin N and C terminal regions to signaling are less clear. Deletion studies indicate that the N terminal domain is not essential for signaling; rather, its absence may enhance stabilization13. Studies by Funayama et al. indicate that Wnt signaling is increased in Xenopus embryos deficient in the C terminal domain14. In other studies by Cox et al., data indicate that complete deletion of the C terminus reduces signaling15,16. In addition, studies in Drosophila suggest that the C terminal region can be divided into three domains. The region distal to 757 is dispensable for Wnt signaling14,17,18. The region between 710 and 757 enhances signaling but is not essential. Lastly, studies by Mo et al. indicate that the proximal C terminal regions (also known as the transactivation domain) stabilize 尾-catenin armadillo repeats making this domain essential for Wnt signaling. Together, these studies indicate that 尾-catenin regions essential for Wnt signaling include the Armadillo repeat core and proximal C terminal regions (also known as the transactivation domain)19,20. Transgenic studies in mice where endogenous 尾-catenin was replaced by mutant forms with D164A substitution and deleted C-terminus show the ability of this mutant to suppress transcription despite preserved adhesive function21. Studies in Drosophila with mutated Armadillo protein (truncated to only 12 armadillo repeats) show increased transcriptional activation after stimulation22. In general, studies highlight the importance of Armadillo and C terminal transactivation domains in 尾-catenin signaling while suggesting that other regions are dispensible.In humans and rodents, increased PI3K (phosphatidylinositide 3-kinase) activation seen with diminished PTEN (phosphatase and tensin homolog) levels enhances 尾-catenin activation. Studies from He et al. indicate that Akt phosphorylates 尾-catenin at serine 552 located in the open loop within the tenth Armadillo repeat. In PTEN mutant mice (Mx1Cre/PTENfl/fl), increased numbers of epithelial cells with nuclear phospho-尾-cateninSer552 (p尾-Cat552) were detected in small bowel polyps. Staining for p尾-Cat552 co-localized with stem cell markers (Msi-1, phospho-PTEN, 尾-Gal) and increased TCF4 transcriptional activity seen in Mx1Cre/PTENfl/flxTOP-GAL mice23. Using an antibody specific for p尾-Cat552, we detected enhanced nuclear staining in epithelial cells in human colitis and colorectal cancer24. Given that anti-p尾-Cat552 staining identifies an epitope with the core region, we further explored its role in activation of 尾-catenin signaling.Studies here interrogate alterations in 尾-catenin that occur during Wnt/尾-catenin signaling in cancer and mucosal inflammation. The data reveal that post-translational modifications of 尾-catenin in the ubiquitin-proteasome pathway yield a truncated 尾-catenin molecule containing a serine 552-phosphorylated core region without N and C termini. This proteolytic processing of 尾-catenin is required for binding with TCF4 and subsequent transcriptional activation.ResultsLow-molecular weight 尾-catenin predominates in the nuclei of cancer cellsMutations in components of the destruction complex (e.g. APC, 尾-catenin, Axin) are associated with increased nuclear accumulation of 尾-catenin and enhanced Wnt/尾-catenin signaling in tumor cells5. To examine the molecular events occurring during enhanced 尾-catenin signaling in the colon, intestinal epithelial cell (IEC) nuclear isolates from patients with CRC and colitis were assayed by WB (western blot) using antibodies specific for 尾-catenin epitopes at N and C termini as well as within the region of armadillo repeats (core and p尾-Cat552). Data in Fig.聽1A indicate that full-size (FS) (~86-90鈥塳D) 尾-catenin was detected by antibodies specific for N-terminal, C-terminal, core region (armadillo repeats) and p尾-Cat552 epitopes in cytosolic and membrane fractions. Examination of nuclear fractions indicates that levels of FS 尾-catenin detected by C-terminal and core region-specific antibodies were similar in normal and CRC tissue. N-terminal antibody detected enhanced levels of 尾-catenin in this sample, while other WB failed to detect increased N-terminal 尾-catenin (Suppl. Fig.聽S5A) suggesting this finding was inconsistent.Figure 1Nuclear low-molecular weight (LMW) 尾-catenin predominates in colon cancer IEC. Protein isolates from biopsy-derived normal and colorectal cancer (CRC) IEC were probed with antibodies specific for distinct 尾-catenin epitopes. (A) Normal and CRC IEC fraction lysates (cytosolic (Cyto), membrane (Memb) and nuclear (Nucl) were probed sequentially for 尾-catenin antibodies directed to different epitopes as shown. Purity controls were achieved by probing for 伪-tubulin (cytosol), E-cadherin (membrane) and lamin B1 (nucleus). (B) Nuclear lysates from normal and CRC biopsies were probed for phospho-尾-cateninSer552 (p尾-Cat552). WB for lamin B1 serves as loading control. Arrows indicate positions of LMW 尾-catenin. Full size membrane scans for WBs can be seen in Suppl. Fig.聽SS1.Full size imageInterestingly, a smaller, 52鈥?6鈥塳D fragment of 尾-catenin detected by core region and p尾-Cat552antibodies was increased in CRC tissue compared to normal (arrows). Given that both core and p尾-Cat552antibodies identify epitopes within the armadillo repeats, we postulated that 尾-catenin may be cleaved into a low molecular weight (LMW) form in transformed cells.To confirm the presence of LWM 尾-catenin in other CRC tissues, we surveyed epithelial tissue from sporadic CRC as well as metastatic lesions in the liver from primary CRC tumors. Data show that anti-p尾-Cat52 antibody detected FS 尾-catenin in normal colon compared to predominantly LMW 尾- catenin present in relatively high levels in CRC. In most cases, the LMW form of nuclear 尾-catenin detected by anti-p尾-Cat552 antibody was found to the exclusion of higher molecular weight forms (Fig.聽1B). This pattern was evident in tumors from all examined regions of the colon (we had tested more than fifty individual biopsies). Examination of liver tissue that harbored metastatic lesions from CRC revealed that LMW nuclear p尾-Cat552 predominated in hepatic metastasis compared to adjacent normal liver where FS 尾-catenin was seen (Suppl. Fig.聽S1A). Furthermore, relatively high levels of LWM 尾-catenin were detected by the anti-p尾-Cat552 antibody in lung, pancreas and primary liver tumor tissues compared to matched normal tissues (Suppl. Fig.聽S1B). In cancer-bearing cirrhotic livers, LMW 尾-catenin levels were greater in tumor (hepatoma) compared to adjacent cirrhotic tissue. We also found that nuclear levels of LMW 尾-catenin were increased in colitis compared to normal IEC and that levels were clearly enriched in colitis-associated cancers (CAC) (Suppl. Fig.聽S1C and D). Of note, LMW 尾-catenin was present in anti-CD45 depleted IEC but not CD45-sorted cell fractions (Suppl. Fig.聽S1E). Together these data indicate that increased levels of LMW p尾-Cat552 are associated with enhanced epithelial Wnt/尾-catenin signaling as seen in cancer and colitis.To confirm the specificity of antibodies detecting LMW 尾-catenin, siRNA knockdown of 尾-catenin in normal human colonic epithelial cells (NCM460) was utilized. As shown in Suppl. Fig.聽S2B and C LMW bands of 尾-catenin detected with core and p尾-Cat552 specific antibodies were significantly reduced after siRNA (small (or short) interfering RNA) treatment. Together, these results indicate that antibodies to core and p尾-Cat552 epitopes identify a LMW form of 尾-catenin present in normal cells and increased in states of enhanced Wnt signaling.Nuclear appearance of LMW 尾-catenin is proteasome-dependentTo examine potential mechanisms for generating LMW 尾-catenin, we tested the hypothesis that 尾-catenin cleavage was dependent on cytosolic proteasome activity. As an attachment of lysine 48 (K48) polyubiquitin chain (poly-Ub) has been shown to target proteins to the proteasome for degradation25,26, we examined its role in 尾-catenin processing. Cytosolic and nuclear fractions of colon cancer cell line HT29 were used for immunoprecipitation (IP) with an antibody specific for the K48 poly-Ub and then probed for 尾-catenin epitopes within N and C termini as well as p尾-Cat552. Data in Fig.聽2A show that the proteasome inhibitor bortezomib27 increased levels of K48 ubiquitinated 尾-catenin in the cytosol. As expected, N and C terminus antibodies detected molecular forms of poly-Ub-尾-catenin at MW 120鈥塳D. These findings were consistent with the predicted molecular weight (MW) of 尾-catenin (88鈥塳D) and attached side chain composed of four ubiquitins (8.5鈥塳D脳4鈥?鈥?4鈥塳D). By comparison, the antibody specific for p尾-Cat552 recognized poly-Ub-尾-catenin of approximately 86鈥?0鈥塳D MW. These data were consistent with the combined MW of LMW 尾-catenin (52鈥?6鈥塳D) and four ubiquitins (34鈥塳D). The detection of K48 polyubiquitinated FS and LMW p尾-Cat552 in bortezomib-treated samples suggests that both FS and LMW p尾-Cat552 are degraded in the proteasome following K48 polyubiquitination. In one of the聽repeated experiments with HT29 cells (as done in Fig.聽2A) we detected an increase in poly-Ub-p尾-Cat552 at 120鈥塳D (seen in over-exposed WB, blue arrow, at Suppl. Fig.聽S3A). Interesting, a聽notable amount of ubiquitinated LMW p尾-Cat552 is detected in nuclear soluble fractions (red arrow at Suppl. Fig.聽S3A). These findings support our hypothesis that serine 552聽 phosphorylation at FS and LMW 尾-catenin is a signal for K48 ubiquitination (confirmed by experiments in RKO cells transfected with mutated serine 552 - alanin聽(Ser552- Ala) 尾-catenin constructs, which will be discussed below) and suggest that further degradation of 尾-catenin may occur in the nucleus. In addition, we found that proteasome inhibition greatly reduced chromatin-bound 尾-catenin levels, especially LMW forms detected by anti-core region and p尾-Cat552 antibodies (Fig.聽2B). In all cases, proteasome inhibition increased cytosolic levels of FS 尾-catenin (Fig.聽2B). In Suppl. Fig.聽S3B are shown results of K48 IP of NCM460 cells after treatment with siRNA to 尾-catenin. The nearly complete attenuation of bands identified by C terminal and anti-p尾-Cat552 antibodies confirms that these derive from 尾-catenin. To confirm that 尾-catenin antibodies detected proteins covalently bound to the ubiquitin chain, denatured proteins were precipitated with anti-K48 antibody and probed with N-terminal and p尾-Cat552 antibodies (Suppl. Fig.聽S3C). WBs revealed bands with the same MW detected in Fig.聽2A.Figure 2聽尾-catenin truncation is proteasome dependent. (A) Cytosolic and nuclear lysates from HT29 cells, treated with bortezomib (Borte) were immunoprecipitated for lysine 48 polyubiquitin chain (K48) and probed for N, C and p尾-Cat552 epitopes of 尾-catenin. (B) HT29 cells treated with bortezomib were fractionated to cytosolic (Cyto), membranous (Memb), soluble (Nucl) and chromatin-bound (Chrom) nuclear fractions. WBs were run sequentially for 尾-catenin antibodies specific for epitopes as indicated. Solid arrows depict decreases in LMW 尾-catenin with bortezomib. Open arrows depict increases in full length 尾-catenin after bortezomib treatment. The braces indicate p尾-Cat552 likely processed in the聽nuclear proteasome. 伪-tubulin is a loading and purity control for cytosolic fraction, histone H3 - for chromatin bound nuclear fraction. Full size membrane scans for WBs can be seen in聽Suppl. Fig.聽SS2.Full size imageTo address whether the proteolytic cleavage of 尾-catenin reported here occurs due to a regulated process, we first identified proteases with the potential to target 尾-catenin (see: http://web.expasy.org/peptide_cutter)28. Next, a number of inhibitors specific for relevant proteases were used to treat HT29 colon cancer cells. Data show that inhibition of chymotrypsin-like activity with epoxomicin29, bortezomib, VR2330 (200鈥塶M), and MG31231 reduced TCF/LEF (T cell factor 4/lymphoid enhancer-binding factor) luciferase activity (Fig.聽3A) as well as levels of LMW p尾-Cat552 in chromatin-bound nuclear fractions (Fig.聽3B). By comparison, inhibitors of proteasome trypsin-like activity (VR23, 1鈥塶M), proline endopeptidase (KYP2047), lysosomal endopeptidases (bafilomycin, chloroquine), cathepsin K or proteinase K (calpeptin) failed to affect TCF/LEF luciferase activity (Fig.聽3A) or LMW p尾-Cat552 levels. These findings support the conclusion that 尾-catenin is specifically cleaved by chymotrypsin-like activity within the proteasome.Figure 3Inhibitors of chymotrypsin-like activity of the proteasome diminish 尾-catenin transcriptional activity. (A) HT29 cells transfected with TCF/LEF reporter were treated with inhibitors as indicated. Luciferase activity is shown in cells incubated with protease inhibitors. Asterisks indicate statistically significant p values compared to control. p鈥?lt;鈥?.001. (B) HT29 cells were treated as in A聽and nuclear chromatin-bound fractions probed for p尾-Cat552. Fibrillarin served as a loading control. Full size membrane scans for WBs can be seen 聽in Suppl. Fig.聽SS3.Full size imageTo address possible concerns that LMW 尾-catenin bands seen on WBs are the result of proteasome activity after cell lysis, human biopsies from normal and colitic patients were divided in half and fractionated into cytosolic, membranous and nuclear protein lysates. One half was fractionated in the presence of epoxomicin. As shown on Suppl. Fig.聽S4, addition of the proteasome inhibitor failed to alter p尾-Cat552 levels detected by WB.TCF4 binds LMW-尾-cateninAs we detected multiple molecular weight forms of nuclear 尾-catenin, we next examined 尾-catenin binding to TCF4. Proteins immunoprecipitated by anti-TCF4 were probed with different 尾-catenin antibodies. Antibodies specific for p尾-Cat52 and core region detected 尾-catenin/TCF4 binding predominantly in CRC, but not antibodies for N and C terminal 尾-catenin epitopes (Fig.聽4A and Suppl. Fig.聽S6A). Reverse co-IPs support these data, as the p尾-Cat552 antibody precipitated TCF4 (Fig.聽4B). The two bands of TCF4 seen in Fig.聽4B may be explained by recent observations from Weise et al. that alternative splicing of TCF4 transcripts generates protein variants (M1/S2)32. In CRC biopsy samples treated with proteasome inhibitor MG132 ex vivo, as well as in HT29 cells treated with bortezomib, we found that proteasome inhibition greatly reduced nuclear levels of p尾-Cat552 bound to TCF4 (Fig.聽4C and D). The effect of proteasome inhibition on above mentioned TCF4 transcriptional activity was supported by results in TCF/LEF luciferase reporter assays in HT29 (Fig.聽3A) and NCM460 cell lines (Suppl. Fig.聽S5).Figure 4TCF4 binds proteasome-sensitive LMW-尾-catenin. (A) IEC nuclear fractions from normal (N) and CRC biopsies were immunoprecipitated by anti-TCF4 and probed for p尾-Cat552, core region 尾-catenin and TCF4. (B) Nuclear fractions from CRC biopsies were used for IP with different 尾-catenin antibodies. Precipitated proteins were probed with anti-TCF4 antibody. Control WBs with N and C terminal, core region specific and p尾-Cat552 antibodies can be seen in Suppl. Fig.聽S6D. (C) Normal, CRC and CRC treated with MG132 IEC nuclear fraction were immunoprecipitated by anti-TCF4 and probed for p尾-Cat552. Input and loading control WBs for these samples can be seen on Suppl. Fig.聽S6E. (D) Chromatin-bound nuclear fractions of HT29 cell, untreated and treated with bortezomib (input and loading controls in Fig.聽2B), were immunoprecipitated by anti-TCF4 and probed for anti-p尾-Cat552 and core region 尾-catenin antibodies. Full size membrane scans for WBs can be seen at Suppl. Fig.聽SS4.Full size imageTumor necrosis factor (TNF), Wnt and carcinogenic transformation upregulate chromatin-bound LMW-尾-cateninAs data in Suppl. Fig.聽S1C indicate that IEC nuclear 尾-catenin levels increase in IBD, we examined nuclear accumulation of 尾-catenin following TNF treatment. Cytosolic, membranous, nuclear soluble and chromatin-bound fractions were isolated from NCM460 cells and probed for LMW 尾-catenin. Data revealed that the majority of LMW 尾-catenin associates with the chromatin-bound protein fraction as identified by core region and p尾-Cat552 antibodies (Fig.聽5A). Both LiCl and TNF upregulate chromatin-bound levels of LMW 尾-catenin and TCF4/尾-catenin binding in chromatin-bound protein extracts (Fig.聽5B). Taken together, these data indicate that canonical Wnt signaling and inflammatory cytokines induce 尾-catenin binding to TCF4 in the chromatin-bound fraction of nuclear extracts. WBs of nuclear extracts immunoprecipitated with anti-TCF4 also revealed that LEF-1 binding increased with TNF (Suppl. Fig.聽S6B). Significantly increased binding of p尾-Cat552 to TCF4 was also detected in colon cancer cell lines with different mutations in 尾-catenin signaling pathway - Caco2, SW480, HT29 and HCT116 (Fig.聽5D).Figure 5TNF, LiCl and carcinogenic transformation increase abundance of chromatin-bound LMW-尾-catenin and聽TCF4 binding. (A) NCM460 cells were treated with LiCl and TNF and fractionated as in Fig.聽2B. WBs were probed sequentially with anti-p尾-Cat552 and anti-core region 尾-catenin antibodies. WBs for 伪-tubulin, e-cadherin, laminB1 and histone H3 represent loading controls in sub-cellular fractions. (B) Chromatin-bound fractions from (A) were immunoprecipitated by anti-TCF4 and probed for p尾-Cat552 and core region specific antibodies. (C) Chromatin-bound fractions from NCM460 and indicated colon cancer cell lines were immunoprecipitated by anti-TCF4 and probed for p尾-Cat552. Input control WB was also probed for core region 尾-catenin and fibrillarin. Full size membrane scans for WBs can be seen in聽Suppl. Fig.聽SS5.Full size imageTo confirm that bands detected after TCF4 IP belong to 尾-catenin we used nuclear extracts from NCM460 cells treated with siRNA to 尾-catenin and stimulated with TNF (Suppl. Fig.聽S6C). The significant attenuation of bands identified by anti-core and anti-p尾-Cat552 antibodies confirms that these derive from 尾-catenin.To examine the association of LMW 尾-catenin with active Wnt signaling in normal colonic stem cells, we utilized a well-characterized model of in vitro colonic stem cell expansion published by Hans Clevers and colleagues33. In these cultures, growth of colonic crypt epithelial cells under high Wnt (Methods) conditions promotes expression of stem cell genes whereas low Wnt (Methods) conditions inhibit stem cell expansion/gene transcription. In data presented in Suppl. Fig.聽S7A and B, we show that colonoids grown under high Wnt conditions are noticeably larger and express increased mRNA (message RNA) for genes associated with colonic epithelial stem cells (Lgr5, Axin2, CD44, PCNA) compared to colonoids grown under low Wnt conditions. WB results of p尾-Cat552 show greater levels of p尾-Cat552 localized to chromatin-bound fractions in cells grown under high Wnt compared to low Wnt conditions (Suppl. Fig.聽S7C). Probing WBs with an antibody specific for C terminal 尾-catenin revealed that cells grown in high Wnt had lower levels of FS 尾-catenin compared to cells grown in low Wnt. The absence of C terminal 尾-catenin in chromatin-bound fractions of either low Wnt or high Wnt colonoids was consistent with the notion that the C terminus was cleaved from the 尾-catenin detected in chromatin-bound fractions (Suppl. Fig.聽S7C, upper panel) with anti-p尾-Cat552.Overexpressed double聽truncated 尾-catenin increases 尾-catenin signaling in NCM460 cellsGiven findings that nuclear LMW 尾-catenin levels were increased in colon, pancreas, lung and liver tumors, we suspected that protein cleavage was associated with 尾-catenin transcriptional activity. To test this notion, NCM460 cells were transfected with constructs encoding FS 尾-catenin, and 尾-catenin truncated at N and C termini. The 鈥渄ouble聽truncated鈥?尾-catenin, referred to as 鈭嗏垎 尾-catenin was generated based on the predicted chymotrypsin cutting sites outside of armadillo repeats (see: http://web.expasy.org/peptide_cutter)28. From the total of 28 possible sites flanking N and C termini of the armadillo repeats, we choose high specificity sites 聽tyrosin142 and 聽phenylalanin聽683. To test if treatment with chymotrypsin 聽would generate peptides with molecular weight close to 52鈥?6鈥塳Da we used recombinant 尾-catenin. As seen on Suppl. Fig.聽S8A overnight treatment with chymotrypsin yielded fragments close to this molecular weight. Thus, 鈭嗏垎聽尾-catenin contained amino acids 143 to 683 of 尾-catenin. The 鈭哊142 protein includes the armadillo sequences along with an intact C-terminus (amino acids 143 to 781). All constructs were tagged with His at the N-terminus and Flag at the C-terminus (Fig.聽6C). Results in Fig.聽6A indicate that Flag and His-tagged proteins were detected in cytosolic, membrane and nuclear soluble fractions of cells transfected with FS, 鈭嗏垎 and 鈭哊142 尾-catenin constructs. However, examination of chromatin-bound fractions revealed significant differences in detection patterns of Flag and His-tagged proteins among transfected cells. First, Flag and His-tagged proteins were not detected in chromatin-bound fractions of cells transfected with FS 尾-catenin. Secondly, LMW His-labeled proteins, but not Flag-tagged proteins, were detected in chromatin-bound fractions of聽cells transfected with the 鈭哊142 construct suggesting that cleavage of C terminal sequences occurred prior to translocation. Lastly, detection of both Flag and His-tagged proteins in chromatin-bound fractions of 鈭嗏垎 尾-catenin聽transfected cells suggested this protein localized without further cleavage. These data were confirmed by cobalt (Co2+) sepharose pull down (His tag聽specific) experiment presented on Suppl. Fig.聽S8B, where only 鈭嗏垎 尾-catenin Flag聽tagged protein was detected in chromatin-bound fraction but not FS or 鈭哊142.Figure 6Overexpressed double truncated 尾-catenin increases transcriptional activity. (A) NCM460 cells were infected with FS, 鈭嗏垎 and 鈭哊142 尾-catenin constructs tagged with His (N聽terminus) and Flag (C聽terminus). WBs were probed sequentially with anti-Flag and anti-His antibodies. Fibrillarin served as loading and purity controls for chromatin bound fractions. (B) Total cell lysates of NCM460 cells overexpressing FS 尾-catenin were precipitated with anti-Flag antibody or (Co2+) sepharose (His tag specific). Proteins were resolved on SDS聽PAGE and WBs probed sequentially with N and C聽termini specific anti-尾-catenin antibodies. (C) Schematic representation of 尾-catenin constructs used in (A and B). (D) Chromatin bound fractions from NCM460 cells are shown: vector control, FS, 鈭嗏垎 and 鈭哊89 尾-catenin聽overexpressing cell lines immunoprecipitated with anti-TCF4 and probed with p尾-Cat552 and core region antibodies. Fibrillarin serves as a loading control for input. (E) FS, 鈭嗏垎 and 鈭哊89 尾-catenin overexpressing NCM460 cell lines were co-transfected with TCF/LEF reporter plasmid and 尾-catenin-induced luciferase activities measured. *p鈥?鈥?.002, **p鈥?鈥?.001, ***p鈥?鈥?.003. Flag WB of total cell lysates used in (D and E) (to evaluate levels of overexpressed 尾-catenin) can be seen on Suppl. Fig.聽S8C. Full size membrane scans for WBs can be seen 聽in Suppl. Fig.聽SS6.Full size imageInterestingly, there were 75kDa bands seen in the FS panel identified by both Flag and His tags. We propose these bands represent 尾-catenin fragments where the N聽terminus has been cleaved (identified by Flag tag) or where the C-terminal sequence has been cleaved (identified by His tag). Both of the residual peptides weighed approximately 75kD. To confirm these interpretations, we performed pull down experiments in which total cell lysates were immunoprecipitated with anti-Flag antibody or precipitated with His tag聽specific Co2+ sepharose beads. Sequential WBs with N and C聽terminus聽specific antibodies (Fig.聽6B) revealed full size (~90kD) peptides and those where cleavage of C or N termini resulted in 75kD fragments. The remaining peptides contained armadillo repeats with N or C termini respectively. A single 聽~90kD band was detected in anti-Flag IP probed with N terminal Ab whereas two bands were seen in His聽specific sepharose beads precipitates probed with N terminal antibody聽(~90聽and ~75kD). (A diagram of these peptides is shown in聽Fig. 6C). On the right panel聽of Fig.聽6B, Flag IP of lysates were probed with C terminal 聽antibody. Again we see full size 尾-catenin (88鈥塳D) along with a ~75kD fragment containing armadillo and C terminal peptides. In the lane on the right panel of Fig.聽6B showing His聽specific Co2+ sepharose precipitate, the C聽terminal WB shows a single ~90kD full size band without a ~75鈥塳D band containing C terminal sequences because this peptide was absent (see diagram in Fig.聽6C). These latter data were consistent with the interpretation that the N terminal peptide containing His tag was lost after cleavage leaving an intermediate fragment containing armadillo and C terminal (Flag-tagged) peptides. The data provide further support for the notion that N and C terminal regions of 尾-catenin are sequentially cleaved prior to nuclear translocation.To examine the effect of 尾-catenin truncation on TCF4 binding and transcriptional activation, NCM460 cells were transfected with FS, 鈭嗏垎 and 鈭哊89 尾-catenin聽 (a molecule with聽first 89聽 amino聽acids truncated). Cells were transfected with 鈭哊89 尾-catenin to allow comparisons to mutated, 鈥渃onstitutively active鈥?尾-catenin13,34,35. TCF4 IP studies showed that levels of p尾-Cat552 bound to TCF4 were highest in 鈭嗏垎尾-catenin cells compared to 鈭哊89 or FS 尾-catenin (Fig.聽6D). Similarly, TCF4/LEF luciferase activity was highest in cells transfected with 鈭嗏垎 compared to 鈭哊89 or FS 尾-catenin (Fig.聽6E). Total cell lysate WBs (Suppl. Fig.聽S8C) show equal expression of tagged 尾-catenin in the cell lines used in Fig.聽6D and E.To investigate the ability of 鈭嗏垎 尾-catenin to increase expression of 尾-catenin target genes, mRNA expression from NCM460 transfected with FS and 鈭嗏垎 尾-catenin was compared to vehicle control. As presented in Suppl. Fig.聽S8D elevated levels of 尾-catenin target gene mRNA (Axin2, Lgr5, CD44, ASCL2, Ki97 and cyclin D1) were detected in cells overexpressing 鈭嗏垎 尾-catenin. c-myc mRNA expression wasn鈥檛 changed. Interestingly, increases in c-myc and cyclin D1 protein levels (WB) were greater (Suppl. Fig.聽8E) than mRNA changes. This might reflect an importance of post-translational regulation of c-myc and cyclin D1 expression by reduced degradation of c-myc and nuclear accumulation of cyclin D1. c-myc protein is degraded after K48 ubiquitination in normal cell, while in cancer it is stabilized and accumulated in cytosol36. Alterations in cyclin D1 turnover in cancer can lead to nuclear accumulation of cyclin D1 independent of changes in cyclin D1 mRNA expression. In normal, cyclin D1 protein is translocated in cytosol where it is utilized by proteasome37.Together, these data indicate that LMW 尾-catenin efficiently localizes to chromatin-bound fractions where it binds TCF4 and drives TCF/LEF transcriptional activity.Double聽truncation and serine聽552 phosphorylation of 尾-catenin increases tumor invasivenessTo further investigate whether 尾-catenin truncation conveys enhanced cell invasiveness, NCM460 cells transfected with control, FS 尾-catenin, 鈭哊89 尾-catenin and 鈭嗏垎尾-catenin vectors were examined in methylcellulose colony formation and proliferation assays. Findings in Suppl. Fig.聽S9A and B show that cells transfected with 鈭嗏垎聽尾-catenin invaded and proliferated at higher levels compared to those transfected with 鈭哊89 or FS constructs. These findings suggest that double聽truncated 尾-catenin promotes cellular responses associated with tumor invasiveness.Phosphorylation of 尾-catenin at serine聽552 has been reported to affect nuclear translocation23,38. To examine the impact of 尾-catenin 聽serine聽552 phosphorylation on localization to chromatin聽bound fractions, RKO tumor cells were transfected with distinct 鈭嗏垎 尾-catenin constructs containing wild type (FS and聽螖螖) or Ser552- Ala substituted (FS聽552A and聽鈭嗏垎聽552A) sequences. RKO colon cancer cells were used to avoid mutations in the 尾-catenin signaling pathway. All constructs were Flag聽tagged at the C聽terminus. As seen in Fig.聽7A, Flag聽tagged 鈭嗏垎 尾-catenin was prevalent in cytosol, membrane, nuclear and chromatin-bound fractions whereas Flag聽tagged 鈭嗏垎552A was detected in membrane and nuclear fractions without appreciable levels seen in cytosol or chromatin-bound fractions suggesting serine聽552 phosphorylation was needed for efficient localization to nuclear chromatin. Analyses of TCF/LEF transcriptional activity showed that FS and 鈭嗏垎 尾-catenin聽transfected cells exhibited 12 and 16-fold higher luciferase activities, respectively, compared to vector-only controls or alanine-substituted FS聽552A or 鈭嗏垎聽552A expressing cells (Fig.聽7B). Thus, chromatin聽localization of p尾-Cat552 correlated with transcriptional activity.Figure 7尾-catenin serine聽552 phosphorylation enhances translocation to chromatin-bound fraction and increases xenograft tumor growth. (A) RKO cells were transfected with FS, 鈭嗏垎, FS聽552A and 鈭嗏垎聽552A 尾-catenin and fractionated. WBs were probed with anti-Flag antibody. Fibrillarin served as loading and purity controls for chromatin-bound fractions. (B) RKO cell lines overexpressing wild type聽 and mutated 尾-catenin were co-transfected with TCF/LEF reporter plasmid and luciferase assay performed. *p鈥?鈥?.002, **p鈥?鈥?.001, ***p鈥?鈥?.005. (C) RKO cells were treated with 20nM epoxomicin and cytosolic lysates precipitated with K48 specific antibody. WBs were developed with Flag聽tag antibody. Actin served as a loading control. (D)聽Xenograft mice were injected with RKO cells transfected with control and overexpressing 尾-catenin constructs. The graph represents volumes of developed tumors. *p鈥?鈥?.016; **p鈥?鈥?.04; ***p鈥?鈥?.003. n鈥?鈥?. Full size membrane scans for WBs can be seen in Suppl. Fig.聽SS7. (E)聽Proposed mechanism of 尾-catenin transcriptional activation.Full size imageTo examine the impact of serine聽552 phosphorylation on 尾-catenin ubiquitination in the cytosol, RKO cells (transfected as in Fig.聽7B) were treated with epoxomicin to block cytosolic proteasome activity. Probing of anti-K48 poly-Ub IPs revealed substantial levels of Flag聽tagged wild type聽FS and 鈭嗏垎聽 proteins. In contrast, cells transfected with聽serine 552聽substituted sequences (FS聽552A or 鈭嗏垎聽552A) failed to yield ubiquitin-bound proteins in epoxomicin聽treated cells (Fig.聽7C). Taken together with other data presented here (Fig.聽2A, Suppl. Fig.聽S3A) these findings suggest that serine聽552 phosphorylation promotes ubiquitination and targets 尾-catenin to the cytosolic proteasome for specific cleavage prior to nuclear translocation (Fig.聽7E).To examine the ability of 鈭嗏垎 尾-catenin and 聽serine聽552 phosphorylation to promote tumor growth, we utilized a xenograft mouse model. RKO cells transfected with vector, FS, 鈭嗏垎, FS聽552A and 鈭嗏垎聽552A were injected into nude mice. Data presented in Fig.聽7D show that RKO cells transfected with 鈭嗏垎 尾-catenin grew larger than RKO cells transfected with wild type FS or Ser552- Ala - mutated FS or 鈭嗏垎 constructs. Colony formation and proliferation assays (Suppl. Fig.聽S10A鈥揅) confirmed significant advantages of overexpressed 鈭嗏垎 尾-catenin compared to wild type FS or mutated constructs. Together, these findings support the notion that 尾-catenin serine聽552 phosphorylation and cleavage at both N and C termini promote tumor aggressiveness. In the cartoon model in Fig.聽7E we propose a sequence of 尾-catenin transcriptional activation based on our findings in this study. In this model, activation induces phosphorylation at FS 尾-catenin serine聽552 prior to ubiquitination (with K48 poly-Ubi) and partial cleavage in the cytosolic proteasome. Truncated 尾-catenin is ubiquitinated again with K48 poly-Ubi and translocated into nucleus. In the nucleus, de-ubiquitinated LMW p尾-Cat552 binds TCF4 in chromatin-bound fractions to activate transcription. We further suspect that the remaining ubiquitinated LMW p尾-Cat552 goes to the nuclear proteasome for complete degradation.DiscussionData presented here are consistent with a novel view of 尾-catenin processing during Wnt-activated signaling. In states, where Wnt/尾-catenin signaling is increased (colorectal cancer, pancreatic cancer, lung cancer, liver cancer and colitis), we found increased levels of a serine 552聽phophorylated LMW form of 尾-catenin in the nucleus suggesting that 尾-catenin processing had occurred. Biochemical studies revealed that FS 尾-catenin was ubiquitinated in the cytosol by K48 poly-chain (Ub-尾-catenin) (Figs聽2A and 7C) prior to cleavage. The failure to detect LMW 尾-catenin by antibodies directed against residues in N or C termini suggests that 尾-catenin undergoes partial site-directed cleavage outside the core armadillo repeat region. The detection of FS p尾-Cat552 in the membrane and cytosol but not the nucleus (Figs聽1A and 6A) suggests that the cleavage event occurs prior to nuclear translocation. Specific enzyme inhibition studies indicated that chymotrypsin-like activity within the cytosolic proteasome was responsible for 尾-catenin cleavage (Fig.聽3). Detection of LMW poly-Ub p尾-Cat552 in bortezomib-treated cells suggests that聽serine聽552 phosphorylation may be a signal for a second step of K48 poly-ubiquitination. This view was supported by studies showing Ser552- Ala mutated 尾-catenin was not ubiquitinated (Fig.聽7C). The failure to detect LMW p尾-cat552 in the chromatin bound nuclear fraction of cells treated with bortezomib (Fig.聽2B) suggests that K48 ubiquitinated LMW p尾-Cat552 was degraded in the proteasome as was FS 尾-catenin. The ubiquitinated FS 尾-catenin form is likely retained in the cytosol and unable to translocate to the nucleus. (This process may be analogous to other signaling pathways such as NFkB39). In NF魏B signaling, the inhibitor I魏B伪 is degraded in the proteasome which frees NF-魏B to translocate into the nucleus and regulate gene transcription40). In contrast, ubiquitinated 鈭嗏垎聽尾-catenin might translocate in the nucleus, where we propose de-ubiquitinating enzymes generate LMW p尾-Cat552 that localizes to the chromatin-bound fraction (Figs聽2B, 6A and 7A). Once activated, LMW p尾-Cat552 binds TCF4 and drives transcription in chromatin-bound fractions (Figs聽5,6 and Suppl. Fig.聽S6鈥?a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"/articles/s41598-017-18421-8#MOESM1\">S8). It is also possible that K48聽poly-ubiquitinated p尾-Cat552 is degraded in the nucleus by the nuclear proteasome (Fig.聽2B, marked with a brace, Suppl. Fig.聽3A). We diagramed this proposed model in the cartoon in Fig.聽7E.The detection of relatively high levels of nuclear LMW p尾-Cat552 in colon cancer cells suggests that this pathway does not preclude the well-characterized degradation pathway altered by mutations in APC and 尾-catenin. In these cells, a high level of 鈥渟tabilized鈥?full length 尾-catenin is available for phosphorylation and cleavage in the pathway presented here (Figs聽5D and 7E). In fact, overexpression of the LMW p尾-Cat552 led to higher levels of TCF/LEF luciferase activity (Fig.聽6E) and tumor growth (Fig.聽7D) suggesting this form may induce higher levels of 尾-catenin聽induced transcription in transformed cells. Together these data suggest a novel step in 尾-catenin signaling involving post-translational modifications (phosphorylation and cleavage). Although levels of LMW p尾-Cat552 were pronounced in transformed cells, detection of LMW p尾-Cat552 in colitis, high Wnt colonoids and TNF聽treated normal cell line indicated that 尾-catenin cleavage likely occurs in non-transformed cells as well (Suppl. Fig.聽S1C,D; Fig.聽5, Suppl. Fig.聽S7C).The cytosolic proteasome has been linked to 尾-catenin degradation following phosphorylation of serine 45 by CK1 (facilitated by APC) and then phosphorylation of serine 33/37 by GSK3尾41. N terminal phosphorylation targets 尾-catenin for ubiquitination by E3 ligase which directs Ub-尾-catenin to the proteasome7,42. Data regarding which type of ubiquitin modification targets 尾-catenin for degradation under this condition聽 are limited. Most likely, it is mono-ubiquitination, as shown by Aberle et al.7. Data presented here indicate that K48 poly-ubiquitination targets FS 尾-catenin and LMW p尾-Cat552 for degradation (partial cleavage) in the cytosolic proteasome (Fig.聽2 and 7C). FS p尾-Cat552 was detected in the cytosol (Figs聽1, 5, Suppl. Fig.聽S7C) as was LMW p尾-Cat552. We did not detect FS p尾-Cat552 in chromatin-bound fractions (Figs聽5A, 6A, Suppl. Fig.聽S7C and S8B) suggesting that it is cleaved prior to binding TCF4. The detection of chromatin and TCF4-bound LMW p尾-Cat552 but not FS 尾-catenin is consistent with the notion that LMW p尾-Cat552 translocates to nuclear chromatin after cleavage in the cytosol (Fig.聽7A).The findings reported here support the conclusion that 尾-catenin undergoes an important cleavage step essential for signaling in the nucleus. This event appears to be most prevalent in transformed tissue where 尾-catenin activation is increased related to mutations in the canonical Wnt/尾-catenin pathway (APC, 尾-catenin, Axin, etc.). The present findings unveil an important nuance of this pathway, that 尾-catenin undergoes a significant modification on its way to nuclear translocation and TCF4 binding. Careful studies were conducted to support the conclusion that generation of LMW 尾-catenin resulted from regulated chymotrypsin-like activity in the proteasome (Fig.聽3). Given the importance of this pathway in cancer development and metastasis, these findings may be helpful in the development of novel diagnostic and therapeutic tools in colitis and cancer. These data also enhance understanding of mechanisms operating in benign disorders associated with increased 尾-catenin signaling such as human colitis.Materials and MethodsThe antibodies used are listed in Suppl. Table聽S1.Reagents used: calyculin A (Calbiochem, San Diego, CA); bortezomib and VR23 (Selleck Chemicals, Houston, TX); TNF (Peprotech, Rocky Hill, NJ); MG132, epoxomicin, bafilomycin, chloroquine, cathepsin K inhibitor I, calpeptin, KYP2047 and polybrene (Sigma, St. Louis, MO); zeocin and puromycin (Invitrogen, Carlsbad, CA); siPORT NeoFX transfection reagent (Thermo, Rockford, IL); jetPRIME transfection reagent (Polyplus, New York, NY); luciferase reagent (Promega, Madisson, WI); recombinamt 尾-catenin (Abcam, Cambridge, MA, cat # ab63175).Human biopsy samplesFor all human studies, informed consent was obtained from every patient and samples were coded. Collection of all material was approved by the Northwestern University or University of Kentucky Institutional Review Boards, in accordance with their guidelines and regulations. Human colonic biopsy specimens were obtained from patients undergoing diagnostic or surveillance colonoscopy for known or suspected ulcerative colitis (UC) or cancer. Specimens were collected from Northwestern Memorial Hospital (Chicago) and the University of Kentucky and Good Samaritan Hospitals (Lexington, KY). For comparison and ex vivo stimulation, biopsy specimens were obtained from healthy patients undergoing routine colon cancer surveillance. Colorectal cancer (CRC) specimens were obtained from patients undergoing surgery. Collection of all patient materials for this study was approved by Institutional Review Board protocol (IRB #13鈥?559-F3R). Pancreas and lung biopsy samples were obtained from Northwestern Memorial Hospital (Chicago). Liver biopsy samples were homogenized in the laboratory of Professor Josep Llovet and semi-purified protein frozen at 鈭?0鈥壜癈 delivered overnight in dry ice.Human biopsy epithelial cell isolationHuman colon epithelia samples were delivered from the operating room in ice cold PBS. Samples were washed once with ice cold PBS and incubated at 4掳C with rotation in PBS with 10mM DDT and 50nM calyculin A for 30min, 4掳C. The samples were centrifuged at 300鈥塺pm for 5鈥塵in. Cells were frozen in liquid nitrogen and stored at 鈭?0掳C until use.Experiment ex-vivoHuman colon biopsies from CRC patients were treated as above. The tissue was divided into two parts and treated with MG132 (in HBSS) as indicated (Fig.聽4C). Samples were incubated at 4掳C with slow rotation for 2hrs, centrifuged at 300鈥塺pm for 5min and supernatant removed. Samples were frozen and stored at 鈭?0掳C until use.AnimalsAll animal experiments were approved by the University of Kentucky Institutional Animal Care and Use Committee and were conducted in accordance with their regulations and guidelines. Male athymic nude mice (5鈥? weeks old) were purchased from Taconic (Hudson, NY, USA). To establishe CRC xenografts, mice were subcutaneously injected with tumor cells (1鈥壝椻€?06/mouse) in a 1:1 mixture of media and Matrigel聽(Corning Inc.,聽Corning, NY)聽 (n鈥?鈥? per group). Tumor dimensions were measured using a caliper and tumor volumes were calculated as mm3鈥?鈥壪€/6脳(larger diameter)脳(smaller diameter)2,43.Cell cultureNCM460 cells (normal derived colon mucosa cells) were received by a cell licensing agreement with INCELL Corporation (San Antonio, TX), and were routinely propagated under standard conditions in M3:10A medium聽supplemented by 10% of fetal bovine serum (FBS) with addition of the conditioned medium (33%) from previously cultured NCM460 cells44. Cells were treated overnight with 1ng/ml of TNF or with 20mM of LiCl, harvested the next morning and fractionated. For experiments in Suppl. Fig.聽S6, bortezomib was added at 50nM for 8 hrs.SW480 cells (ATCC, CCL-228) were cultured in Leibovitz鈥檚 L-15 medium with 10% FBS under standard conditions. HT29 (ATCC, HTB-38), RKO (ATCC, CRL-2577), Caco2 (ATCC, HTB-37) and HCT116 (ATCC, CCL-247) cells were cultured in Dulbecco鈥檚 Modified Eagle鈥檚 Medium (DMEM) with 10% FBS under standard conditions. HT29 cells in log phase were treated with a series of specific inhibitors of proteasomal chymotrypsin-like activity (epoxomicin, 20nM; bortezomib, 50nM; VR23, 200nM; and MG312, 50nM, trypsin-like activity of the proteasome (VR23, 1nM), proline endopeptidase (KYP2047, 50nM), lysosomal endopeptidases (bafilomycin, 0.5nM; chloroquine, 30渭M), inhibitors of cathepsin K, 100nM, and proteinase K (calpeptin, 100nM). The cells were incubated under standard conditions for 8 hrs with proteasome inhibitors and overnight with other inhibitors. Cells were harvested and fractionated.siRNAsiRNA to human 尾-catenin was obtained from Ambion (Thermo, Rockford, IL;聽siRNA ID: s438). siRNA was introduced in NCM460 cells with siPORT NeoFX transfection reagent. Forty eight hours after transfection cells were harvested and used for experiments. siRNA nonsense mix (Santa Cruz, Dallas, TX) was used as a control.Lentiviral constructs and transductionsThe pLV lentiviral plasmids encoding FS, 鈭?9 尾-catenin, 鈭哊142 and 鈭嗏垎 尾-catenin with His tag on N聽terminus and Flag tag on C-terminus were made based on human 尾-catenin pcDNA3.1 neo (gift from Dr. Eric Fearon: Addgene plasmid # 16828)31 with primers listed in Suppl. Table聽S2. The constructs obtained were cloned into pLV-EF1a-MCS-IRES-GFP-Puro (pLV) (Biosettia Inc. San Diego, CA). 鈭哊89 pcDNA3.1 neo 尾-catenin with C聽terminal Flag tag was a gift from Eric Fearon (Addgene plasmid # 19288)34. The insert was also re-cloned into a pLV vector. Primers are listed in Suppl. Table聽S2. VSV-G pseudotyped lentivirus stocks were made in the DNA/RNA Delivery Core, SDRC (Chicago, IL). NCM460 cells were infected in the presence of 1渭g/ml polybrene (Sigma St. Louis, MO) and stable cell lines generated. Cells were maintained under selection pressure with 5渭g/ml of puromycin. Expression levels of FS and 鈭嗏垎聽聽尾-catenin in infected NCM460 cells were assessed by WB for Flag tag.The reporter construct containing TCF/LEF luciferase (TCF/luc) was generated in the DNA/RNA Delivery Core, SDRC at Northwestern University (Chicago, IL) by inserting six copies of the TCF/LEF response element in the lentiviral pGF1 vector (System Bioscience, Mountain View, CA). The cells were infected with TCF/luc viral particles as described above. Luciferase activity was detected with luciferase reagent.pcDNA3.1zeo 尾-catenin expressing constructs and RKO cells transfectionsWild type FS 尾-catenin insertion was re-cloned in pcDNA3.1 with zeocin resistance from pcDNA3.1 expressing C聽terminal FLAG tagged FS 尾-catenin (gift from Dr. Eric Fearon: Addgene plasmid # 16828)34 and pcDNA3.1 expressing C聽terminal FLAG tagged FS 尾-catenin with serine 552 substituted to alanine (FS聽552A) (gift from Dr. Dexing Fang38). Double truncated (鈭嗏垎 and 鈭嗏垎聽552A) and 鈭哊142 尾-catenin constructs were created with primers listed in Suppl. Table聽S2. RKO cells were transfected with vector or with 尾-catenin constructs by using jetPRIME transfection reagent according to the manufacturer鈥檚 instructions (Polyplus, New York, NY). Stable cell lines were established under selection pressure of 400鈥壩糶/ml of zeocin.NCM460 cell proliferation assaysProliferation assays were performed by using the CyQUANT庐 Cell Proliferation Assay Kit (Invitrogen, Carlsbad, CA) according to the manufacturer鈥檚 instructions.RKO cell proliferation assaysCells were placed in 96 well plates at 104 per well (6 wells for each cell line). Cells were stained with trypan blue and viable cells were counted on a hemocytometer 5 times every 24hrs after initial 48hrs of incubation.ColonoidsColonic crypts were isolated from C57BL6 mice via collagenase digestion and embedded in Matrigel as previously described45. Colonoids were established in WENR media (100ng/ml Wnt3a and 1渭g/ml R-spondin)33. After 48hrs, media were changed to reduced Wnt3a and R-spondin (25ng/ml Wnt3a and 250ng/ml R-spondin) for 5 days.Subcellular protein fractionationThe subcellular protein fractionation (human epithelial cells) protocol was modified from described procedures46. All buffers used contained Protease Arrest鈩?protease inhibitor cocktail (G-Biosciences, St. Louis, MO), and phosphatase inhibitor cocktail I and II (Sigma, St. Louis, MO) at 1:100. Human epithelial cells were homogenized in glass homogenizer in buffer I (50mM Tris-HCl pH 7.4, 100mM NaCl, 0.01% digitonin), lysates were passed through a 250渭m tissue strainers (Thermo, Rockford, IL), centrifuged at 4掳C for 10鈥塵in at maximum speed in table top centrifuge. The supernatants were collected and used as the cytosolic fraction. Pellets were re-suspended in buffer II (50mM Tris-HCl pH 7.4, 2% Triton X100, 100mM NaCl) and incubated on ice for 30min, then centrifuged as above. The supernatants were used as the membrane/organelle fraction. Pellets were dissolved in buffer III (50mM Tris-HCl pH 7.4, 0.25% DDM (n-Dodecyl-D-maltoside), 100mM NaCl), and with 2U of Benzonase (Sigma, St. Louis, MO) per 100碌l of lysate and incubated for 30鈥塵in at room temperature (RT). Following centrifugation the supernatants were used as nuclear fractions.Cultured cells and colonoids were fractionated according to Pierce manufacturer鈥檚 protocol (Subcellular Protein Fractionation Kit, Thermo, Rockford, IL). In particular, chromatin-bound protein extraction was performed after nuclear soluble proteins isolation. Nuclear Extraction Buffer (NEB-company formulation) with 5渭l of 100mM CaCl2 and 3渭l of micrococcal nuclease (300 units) per 100渭l, supplemented with protease and phosphatase inhibitors, was added to the pellet. After vortex for 15sec at highest setting mixture was incubated at room temperature (RT) for 15min and centrifuged at 16,000脳g (highest setting of microcentrifuge) for 5min. Supernatant was collected and used as chromatin bound fraction of nuclear proteins.Protein concentration was measured by BCA assay (Thermo, Rockford, IL).The purity of the fractions was confirmed by WB with anti-伪-tubulin, anti-e-cadherin, anti-laminB1, anti-histoneH3, and anti-fibrillarin antibodies (see Suppl. Table聽S1).Immunoprecipitation and western blotting500碌g of nuclear soluble protein or 200碌g chromatin-bound protein fractions, 2碌g of TCF4 primary antibody coupled to agarose beads (Pierce Co-IP Kit, Thermo, Rockford, IL) were added in each IP聽reaction and the mixture was incubated overnight聽at 4oC. The beads were washed according to Pierce protocol and the proteins were eluted from agarose, acetone precipitated, and resolved on SDS-PAGE followed by WB. In experiments presented in Fig.鈥?a data-track=\"click\" data-track-label=\"link\" data-track-action=\"figure anchor\" href=\"/articles/s41598-017-18421-8#Fig2\">2A and聽in Fig.聽4B, 500鈥壜礸 of cytosolic and nuclear soluble protein or 300鈥壜礸 of chromatin-bound protein and 2渭g of K48 or聽尾-catenin聽antibodies (尾-catenin antibodies were coupled to agarose beads as anti-TCF4 (see above)) were used for each IP reaction. The mixture was incubated overnight at 4掳C. 20碌l of Protein A/G Plus agarose (Santa Cruz, Dallas, TX) were added to the mixture and incubation continued for another 30min at 4掳C with GEntle rotation. Agarose beads were washed four times with ice cold RIPA buffer (20% in PBS) and re-suspended in LDS NuPAGE sample buffer (Invitrogen, Carlsbad, CA) with 10% 2-mercaptoethanol. The samples were boiled and resolved with SDS-PAGE, followed by WB.For WB, proteins were transferred on Immobilon FL (Millipore, Billerica, MS) by semi-dry transfer (Bio-Rad, Hercules, CA) and membranes blocked in Pierce Protein-Free T20 blocking buffer (Thermo, Rockford, IL) for 1hr, and incubated overnight at 4掳C in 1:1000 primary antibody solution. Membranes were extensively washed, incubated in 0.02渭g/ml secondary antibody for 1hr., washed again and developed using West Pico, Dura or Femto reagent (Thermo, Rockford, IL).Cobalt sepharose pulldown of total lysates and chromatin-bound fractions from FS, 鈭?42 and 鈭嗏垎 尾-catenin overexpressing NCM460 cell lines was performed according to manufacturer鈥檚 protocol (Thermo, Rockford, IL). For each reaction 500渭g of lysate and 20渭l of sepharose were used. Precipitates were run on SDS-PAGE and WB probed with anti-Flag antibody.Denatured protein IP1mg of cytosolic or nuclear protein lysate was used to precipitate with 1:4(v:v) of 100% trichloroacetic acid. After 10min incubation at 4掳C samples were centrifuged and pellets were washed with ice-cold acetone. The pellets were dried, reconstituted in RIPA buffer and protein concentration measured by BCA assay. IP was performed as above.All WB and IP experiments were repeated at least three times.Real time PCRTotal RNA was isolated from colonoids or cultured cells using the RNeasy Mini Kit (Qiagen, Valencia, CA) and reverse transcribed using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). Real time PCR used the ABI Step OnePlus real-time PCR system and Power SYBR green PCR master mix (Applied Biosystems). Primers were designed by Primer Express software 3.0 (Applied Biosystems) based on nucleotide sequences from the National Center for Biotechnology Information data bank (Suppl. Table聽S3). For each sample, glyceraldehyde-3-phosphate dehydrogenase was used as the internal reference. All assays were performed in triplicate and fold changes were calculated using the 螖螖CT method.Chymotrypsin cleavage of recombinant 尾-cateninDDT was added (final concentration of 5mM) to 5碌g of recombinant 尾-catenin dissolved in 25% glycerol, 50mM Tris-HCl, 150mM NaCl, 0.25mM DDT pH7.5. Protein was incubated 20min at 50鈥?0鈥壜癈 for 20鈥塵in. Reduced protein mixture was cooled to RT and iodoacetamide was added to a final concentration of 15mM. Mixture was Incubated in dark for 15min at RT. Volume was adjusted to 50碌l with 100mM Tris-HCl, 10mM CaCl2 (pH 8.0) and 2ng/碌l chymotrypsin, sequencing grade (Promega, Madison, WI) added to sample. After overnight incubation at 25掳C protein was precipitated with acetone (1:10鈥塿/v, 30min at 鈭?0掳C) and resolved at SDS-PAGE along with intact聽5碌g of 尾-catenin. Gel was stained with silver according manufacturer protocol (Fast Silver Kit, G-Bioscience, St. Louis, MO).Statistical analysisAll experiments were performed at least three times. Results are expressed as mean鈥壜扁€塖.E.M. where applicable. A two-tailed Student鈥檚 t-test was used to compare the intergroup. Differences between groups were considered statistically significant at p 0.05.Data availabilityAll data generated or analyzed during this study are included in this published article (and its Suppl. Information files). References1.Gottardi, C. J. Gumbiner, B. M. Distinct molecular forms of beta-catenin are targeted to adhesive or transcriptional complexes. J Cell Biol 167, 339鈥?49 (2004).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 2.Chien, A. J., Conrad, W. H. Moon, R. T. A Wnt survival guide: from flies to human disease. J Invest Dermatol 129, 1614鈥?627 (2009).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 3.Logan, C. Y. Nusse, R. The Wnt signaling pathway in development and disease. Annu Rev Cell Dev Biol 20, 781鈥?10 (2004).CAS聽 Article聽 PubMed聽Google Scholar聽 4.Clevers, H. Nusse, R. Wnt/beta-catenin signaling and disease. Cell 149, 1192鈥?205 (2012).CAS聽 Article聽 PubMed聽Google Scholar聽 5.Bienz, M. Clevers, H. Linking colorectal cancer to Wnt signaling. Cell 103, 311鈥?20 (2000).CAS聽 Article聽 PubMed聽Google Scholar聽 6.Morin, P. J. et al. Activation of beta-catenin-Tcf signaling in colon cancer by mutations in beta-catenin or APC. Science 275, 1787鈥?790 (1997).CAS聽 Article聽 PubMed聽Google Scholar聽 7.Aberle, H., Bauer, A., Stappert, J., Kispert, A. Kemler, R. beta-catenin is a target for the ubiquitin-proteasome pathway. EMBO J 16, 3797鈥?804 (1997).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 8.Korinek, V. et al. Constitutive transcriptional activation by a beta-catenin-Tcf complex in APC鈭?鈭?colon carcinoma. Science 275, 1784鈥?787 (1997).CAS聽 Article聽 PubMed聽Google Scholar聽 9.Nelson, W. J. Nusse, R. Convergence of Wnt, beta-catenin, and cadherin pathways. Science 303, 1483鈥?487 (2004).ADS聽 CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 10.Brantjes, H., Barker, N., van Es, J. Clevers, H. TCF: Lady Justice casting the final verdict on the outcome of Wnt signalling. Biological chemistry 383, 255鈥?61 (2002).CAS聽 Article聽 PubMed聽Google Scholar聽 11.Verheyen, E. M. Gottardi, C. J. Regulation of Wnt/beta-catenin signaling by protein kinases. Dev Dyn 239, 34鈥?4 (2010).CAS聽 PubMed聽 PubMed Central聽Google Scholar聽 12.Poy, F., Lepourcelet, M., Shivdasani, R. A. Eck, M. J. Structure of a human Tcf4-beta-catenin complex. Nat Struct Biol 8, 1053鈥?057 (2001).CAS聽 Article聽 PubMed聽Google Scholar聽 13.Munemitsu, S., Albert, I., Rubinfeld, B. Polakis, P. Deletion of an amino-terminal sequence beta-catenin in vivo and promotes hyperphosporylation of the adenomatous polyposis coli tumor suppressor protein. Mol Cell Biol 16, 4088鈥?094 (1996).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 14.Orsulic, S. Peifer, M. An in vivo structure-function study of armadillo, the beta-catenin homologue, reveals both separate and overlapping regions of the protein required for cell adhesion and for wingless signaling. J Cell Biol 134, 1283鈥?300 (1996).CAS聽 Article聽 PubMed聽Google Scholar聽 15.Cox, R. T., Pai, L. M., Kirkpatrick, C., Stein, J. Peifer, M. Roles of the C terminus of Armadillo in Wingless signaling in Drosophila. Genetics 153, 319鈥?32 (1999).CAS聽 PubMed聽 PubMed Central聽Google Scholar聽 16.Cox, R. T. et al. Membrane-tethered Drosophila Armadillo cannot transduce Wingless signal on its own. Development 126, 1327鈥?335 (1999).CAS聽 PubMed聽Google Scholar聽 17.Pai, L. M., Orsulic, S., Bejsovec, A. Peifer, M. Negative regulation of Armadillo, a Wingless effector in Drosophila. Development 124, 2255鈥?266 (1997).CAS聽 PubMed聽Google Scholar聽 18.Funayama, N., Fagotto, F., McCrea, P. Gumbiner, B. M. Embryonic axis induction by the armadillo repeat domain of beta-catenin: evidence for intracellular signaling. J Cell Biol 128, 959鈥?68 (1995).CAS聽 Article聽 PubMed聽Google Scholar聽 19.Mo, R. et al. The terminal region of beta-catenin promotes stability by shielding the Armadillo repeats from the axin-scaffold destruction complex. J Biol Chem 284, 28222鈥?8231 (2009).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 20.Gottardi, C. J. Peifer, M. Terminal regions of beta-catenin come into view. Structure 16, 336鈥?38 (2008).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 21.Valenta, T. et al. Probing transcription-specific outputs of beta-catenin in vivo. Genes Dev 25, 2631鈥?643 (2011).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 22.Blauwkamp, T. A., Chang, M. V. Cadigan, K. M. Novel TCF-binding sites specify transcriptional repression by Wnt signalling. EMBO J 27, 1436鈥?446 (2008).CAS聽 PubMed聽 PubMed Central聽Google Scholar聽 23.He, X. C. et al. PTEN-deficient intestinal stem cells initiate intestinal polyposis. Nat Genet 39, 189鈥?98 (2007).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 24.Lee, G. et al. Phosphoinositide 3-kinase signaling mediates beta-catenin activation in intestinal epithelial stem and progenitor cells in colitis. Gastroenterology 139, 869鈥?81, 881 e861鈥?69 (2010).25.Chau, V. et al. A multiubiquitin chain is confined to specific lysine in a targeted short-lived protein. Science 243, 1576鈥?583 (1989).ADS聽 CAS聽 Article聽 PubMed聽Google Scholar聽 26.Finley, D. et al. Inhibition of proteolysis and cell cycle progression in a multiubiquitination-deficient yeast mutant. Mol Cell Biol 14, 5501鈥?509 (1994).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 27.Chen, D., Frezza, M., Schmitt, S., Kanwar, J. Dou, Q. P. Bortezomib as the first proteasome inhibitor anticancer drug: current status and future perspectives. Curr Cancer Drug Targets 11, 239鈥?53 (2011).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 28.Wilkins, M. R. et al. Protein identification and analysis tools in the ExPASy server. Methods Mol Biol 112, 531鈥?52 (1999).CAS聽 PubMed聽Google Scholar聽 29.Kim, K. B., Myung, J., Sin, N. Crews, C. M. Proteasome inhibition by the natural products epoxomicin and dihydroeponemycin: insights into specificity and potency. Bioorg Med Chem Lett 9, 3335鈥?340 (1999).CAS聽 Article聽 PubMed聽Google Scholar聽 30.Pundir, S., Vu, H. Y., Solomon, V. R., McClure, R. Lee, H. VR23: A Quinoline-Sulfonyl Hybrid Proteasome Inhibitor That Selectively Kills Cancer via Cyclin E-Mediated Centrosome Amplification. Cancer Res 75, 4164鈥?175 (2015).CAS聽 Article聽 PubMed聽Google Scholar聽 31.Lee, D. H. Goldberg, A. L. Proteasome inhibitors: valuable new tools for cell biologists. Trends Cell Biol 8, 397鈥?03 (1998).CAS聽 Article聽 PubMed聽Google Scholar聽 32.Weise, A. et al. Alternative splicing of Tcf7l2 transcripts generates protein variants with differential promoter-binding and transcriptional activation properties at Wnt/beta-catenin targets. Nucleic Acids Res 38, 1964鈥?981 (2010).CAS聽 Article聽 PubMed聽Google Scholar聽 33.Sato, T. et al. Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett鈥檚 epithelium. Gastroenterology 141, 1762鈥?772 (2011).CAS聽 Article聽 PubMed聽Google Scholar聽 34.Kolligs, F. T., Hu, G., Dang, C. V. Fearon, E. R. Neoplastic transformation of RK3E by mutant beta-catenin requires deregulation of Tcf/Lef transcription but not activation of c-myc expression. Mol Cell Biol 19, 5696鈥?706 (1999).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 35.Naishiro, Y. et al. Morphological and transcriptional responses of untransformed intestinal epithelial cells to an oncogenic beta-catenin protein. Oncogene 24, 3141鈥?153 (2005).CAS聽 Article聽 PubMed聽Google Scholar聽 36.Vervoorts, J., Luscher-Firzlaff, J. Luscher, B. The ins and outs of MYC regulation by posttranslational mechanisms. J Biol Chem 281, 34725鈥?4729 (2006).CAS聽 Article聽 PubMed聽Google Scholar聽 37.Kim, J. K. Diehl, J. A. Nuclear cyclin D1: an oncogenic driver in human cancer. J Cell Physiol 220, 292鈥?96 (2009).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 38.Fang, D. et al. Phosphorylation of beta-catenin by AKT promotes beta-catenin transcriptional activity. J Biol Chem 282, 11221鈥?1229 (2007).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 39.Palombella, V. J., Rando, O. J., Goldberg, A. L. Maniatis, T. The ubiquitin-proteasome pathway is required for processing the NF-kappa B1 precursor protein and the activation of NF-kappa B. Cell 78, 773鈥?85 (1994).CAS聽 Article聽 PubMed聽Google Scholar聽 40.Wu, Z. H. Shi, Y. When ubiquitin meets NF-kappaB: a trove for anti-cancer drug development. Curr Pharm Des 19, 3263鈥?275 (2013).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 41.Orford, K., Crockett, C., Jensen, J. P., Weissman, A. M. Byers, S. W. Serine phosphorylation-regulated ubiquitination and degradation of beta-catenin. J Biol Chem 272, 24735鈥?4738 (1997).CAS聽 Article聽 PubMed聽Google Scholar聽 42.Amit, S. et al. Axin-mediated CKI phosphorylation of beta-catenin at Ser 45: a molecular switch for the Wnt pathway. Genes Dev 16, 1066鈥?076 (2002).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 43.She, Q. B. et al. 4E-BP1 is a key effector of the oncogenic activation of the AKT and ERK signaling pathways that integrates their function in tumors. Cancer Cell 18, 39鈥?1 (2010).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 44.Moyer, M. P., Manzano, L. A., Merriman, R. L., Stauffer, J. S. Tanzer, L. R. NCM460, a normal human colon mucosal epithelial cell line. In vitro cellular developmental biology. Animal 32, 315鈥?17 (1996).CAS聽 Article聽Google Scholar聽 45.Miyoshi, H. Stappenbeck, T. S. In vitro expansion and genetic modification of gastrointestinal stem cells in spheroid culture. Nat Protoc 8, 2471鈥?482 (2013).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 46.McCarthy, F. M., Burgess, S. C., van den Berg, B. H., Koter, M. D. Pharr, G. T. Differential detergent fractionation for non-electrophoretic eukaryote cell proteomics. J Proteome Res 4, 316鈥?24 (2005).CAS聽 Article聽 PubMed聽Google Scholar聽 Download referencesAcknowledgementsThis work was supported by Merit Review [Award # IO1CX001353 to TAB] from the United States (U.S.) Department of Veterans Affairs Clinical Sciences Research and Development Program; the National Institutes of Health [2R01DK095662-06A1 to TAB and R01CA175105to Q-BS]; National Institute Of Diabetes Digestive Kidney Diseases [U01DK085507 to LL], and the Training Program in Oncogenesis and Developmental Biology through the National Cancer Institute [NCI T32 CA080621, to support EMB]. We would like to thank the Northwestern University Medical Center, University of Kentucky Hospital and Department of Veterans Affairs Medical Center at Lexington clinic staff as well as technical and nursing support staff of the endoscopy labs for their assistance in obtaining tissue samples.Author informationAffiliationsDepartment of Internal Medicine, Division of Gastroenterology, University of Kentucky, Lexington, KY, USATatiana Goretsky,聽Emily M. Bradford,聽Olivia F. Lamping,聽Patrick C. Keller聽 聽Terrence A. BarrettMarkey Cancer Center, University of Kentucky, Lexington, KY, USAQing Ye,聽Tianyan Gao聽 聽Qing-Bai SheLouisiana State University Health Sciences Center, New Orleans, LA, USATomas VanagunasINCELL Corporation, San Antonio, TX, USAMary Pat MoyerNorthwestern University, Chicago, IL, USAPreetika SinhThe Mount Sinai Hospital, New York, NY, USAJosep M. LlovetStowers Institute for Medical Research, Kansas City, MO, USALinheng LiDept of Pathology and Laboratory Medicine, University of Kansas Medical Center, Kansas City, KS, USALinheng LiAuthorsTatiana GoretskyView author publicationsYou can also search for this author in PubMed聽Google ScholarEmily M. BradfordView author publicationsYou can also search for this author in PubMed聽Google ScholarQing YeView author publicationsYou can also search for this author in PubMed聽Google ScholarOlivia F. LampingView author publicationsYou can also search for this author in PubMed聽Google ScholarTomas VanagunasView author publicationsYou can also search for this author in PubMed聽Google ScholarMary Pat MoyerView author publicationsYou can also search for this author in PubMed聽Google ScholarPatrick C. KellerView author publicationsYou can also search for this author in PubMed聽Google ScholarPreetika SinhView author publicationsYou can also search for this author in PubMed聽Google ScholarJosep M. LlovetView author publicationsYou can also search for this author in PubMed聽Google ScholarTianyan GaoView author publicationsYou can also search for this author in PubMed聽Google ScholarQing-Bai SheView author publicationsYou can also search for this author in PubMed聽Google ScholarLinheng LiView author publicationsYou can also search for this author in PubMed聽Google ScholarTerrence A. BarrettView author publicationsYou can also search for this author in PubMed聽Google ScholarContributionsT.G. designed the study, performed and analyzed the experiments, interpreted data and wrote the manuscript. E.M.B. conducted human colonoid study, designed, performed and analyzed Real-time PCR experiment. Q.Y. performed and analyzed the xenograft experiments. O.F.L. performed and analyzed Real-time PCR experiment. T.V. contributed in Fig. 5. M.P.M. provided NCM460 cell line and reviewed the manuscript. P.C.K. contributed in Supplemental Figure S1. P.S. provided technical assistance in Figures 1 and 4. M.L. provided liver biopsy samples. Q.-B.S. performed and analyzed the xenograft experiments and contributed in critical revision of the manuscript. TGao and L.L. contributed in critical revision of the manuscript for important intellectual content. T.A.B. designed the study, interpreted data, wrote the manuscript and supervised study. All authors reviewed the results and approved the final version of the manuscript.Corresponding authorCorrespondence to Terrence A. Barrett.Ethics declarations Competing Interests The authors declare that they have no competing interests. Additional informationPublisher\'s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Electronic supplementary material Supplemental materialsRights and permissions Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article鈥檚 Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article鈥檚 Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Reprints and PermissionsAbout this articleCite this articleGoretsky, T., Bradford, E.M., Ye, Q. et al. Beta-catenin cleavage enhances transcriptional activation. Sci Rep 8, 671 (2018). https://doi.org/10.1038/s41598-017-18421-8Download citationReceived: 11 May 2017Accepted: 11 December 2017Published: 12 January 2018DOI: https://doi.org/10.1038/s41598-017-18421-8 Naushad Ali, Charles B. Nguyen, Parthasarathy Chandrakesan, Roman F. Wolf, Dongfeng Qu, Randal May, Tatiana Goretsky, Javid Fazili, Terrence A. Barrett, Min Li, Mark M. Huycke, Michael S. Bronze Courtney W. Houchen Scientific Reports (2020) CommentsBy submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate. Sign up for the Nature Briefing newsletter 鈥?what matters in science, free to your inbox daily.