DISOPRED, also part of the PSIPRED package, predicted that most of the FUN sequence can be defined as disordered, though the first 50 amino acids of FUN are not predicted to be disordered . In sum, FUN transcript is found in developing leaves and tassels, which fits with the phenotype of the mutant. FUN transcript is also found in developing ears, which is surprising since an ear phenotype was not observed. Higher levels of FUN during leaf development are associated with larger auricles. FUN is predicted to be a disordered protein that localises to the nucleus. Attempts were then made to validate these predictions. Y2H, using the FUN protein as bait, and a cDNA library of immature ears and tassels as prey, retrieved a list of 234 possible interactors. Using the A. thaliana homologues of these proteins in a GO term biological process analysis found enrichment for proteins involved in negative regulation organ, specifically flower, development; and proteins involved in negative regulation of nucleic acid metabolic processes . The A. thaliana homologues were enriched in transcription factor activity and hydrolase activity according to a GO term analysis based on molecular function , and most of these predicted interacting partners were nuclear or cytosol localised. The GRMZM numbers were also run through the GO term analysis prediction software provided by AgriGO and were found to be enriched in GTP and GTPase binding . A table of selected genes returned by the Y2H can be seen in Figure 5-7A. One such gene is ZmDWF1. This brassinosteroid synthesis protein was found to have a synergistic interaction in double mutant analysis with fun . Another gene implicated in the brassinosteroids pathway was BSL1 that is known as an inhibitor of BRI1.
The strongest confidence hit for the Y2H as an interactor for FUN was tubulin,cannabis drying rack ideas but this may be a false positive due to the high concentration of this protein in a cell.In order to test the idea that FUN is nuclear-localized, I carried out a transformation experiment using Nicotiana benthamiana. The entire FUN protein was fused to an N terminus YFP as described in Methods. The FUN-YFP fusion was found to localise to the nucleus in transformed N. benthamiana pavement cells. This result was observed in two separate transformations. Though not all nuclei in the samples were found to fluoresce under YFP excitation, this is likely due to imperfect transformation efficiency and is normal in this kind of experiment109. 25 nuclei expressing YFP were photographed and many more observed during the course of this experiment; YFP expression was not observed in any other subcellular regions. In order to confirm that the YFP expression was nuclear, the leaves were also examined under 405nm excitation. Since the leaves had been infiltrated with DAPI prior to examination, this caused the DNA to fluoresce. YFP fluorescence was shown to overlap with this DAPI fluorescence . As further confirmation, the sample was also examined under bright field and the YFP fluorescence was thus seen to overlap with clearly visible nuclei . This was observed at the microscope, as well as by overlapping micrographs using ImageJ. Close inspection of individual transformed nuclei revealed a nuclear speckle pattern . In order to make an antibody to the FUN protein, the purified protein has to be injected into a living animal and the antibody produced must then be purified. To this end, the third exon of FUN was amplified by primers 53xF/R and cloned into pENTR. Recombining with pDEST17 was unsuccessful. I hypothesised that FUN may be toxic, which would explain why the expression plasmid would not grow, though pENTR would grow, so bacteria were grown at lower temperatures. This was also unsuccessful, so a smaller fragment out of the exon was used, amplified from cDNA with primers An2F/R . This 600bp fragment was successfully cloned into pDEST17. Interestingly, the other fragments attempted that contained the conserved GAKHIL motif did not clone into pDEST17 under the conditions used, which could imply a toxicity of this domain, since all fragments that did not contain the GAKHIL motif were successfully cloned.
The newly made plasmid pDEST17-An2F/R was then cloned into Rosetta cells and grown into a 100ml culture overnight. This culture was used to spike 1 litre of fresh LB. After 4 hours of growth at 37°C this culture reached 0.42 OD600nm and 200μl of 1M IPTG were added. This culture was then shaken at room temperature for 5 hours and spun at 8000 RCF for 15 minutes at 4°C. The resulting pellet was resuspended in 75ml of lysis buffer . This was spun at 12000 RCF for 15 minutes and the supernatant co-incubated with 2.5ml of 50% Ni++ beads in EtOH slurry. The beads were spun and washed several times using wash buffer and were finally eluted using elution buffers at pH 4.4, 3.8 and 3twice each. Most protein was recovered from the second pH 4.4 and first 3.8 elutions as measured by nanodrop. These samples were then run on an SDS gel and the correct size band was observed for An2F/R samples, but nothing was observed for An1F/R. Thus the protein produced by plasmid pDEST17-An2F/R was purified. It was then resuspended in 6M urea by dialysis in a side-A-lyser cassette and sent to Cocalico for injection into guinea pigs. Test-bleeds and pre-bleeds were returned from the company. The pDEST17- An2F/R protein product was blotted onto nitrocellulose and incubated in 2ml of PBS with 2μl of test-bleed or 2μl of pre-bleed for 2 hours before washing and incubating in PBS and anti-Guinea Pig Alkaline Phosphotase fusion antibody. Finally this was incubated in NETN and NBT/BCIP to give an output of purple colour if the antibody is reacting to the protein . Later, second bleeds were returned by the company, and this dot-blot process was repeated, along with the sera from another antibody as a control . Thus, it was confirmed that the guinea pigs were producing appropriate antibody. In order to purify the terminal bleeds, a GST-tagged version of the protein was made. The pENTR-An2F/R was therefore combined with pDEST-15,grow trays 4×4 transformed into Rosetta cells and induced and spun down as for the pDEST17- An2F/R purification step. This pellet was resuspended in NETN and put through a French press twice at ~1000lbs of pressure before 2x 20s of sonication to shear DNA. Triton X-100 was then added to 0.5%. This cell lysate was then spun at 12,000 RCF for 15mins and the supernatant was added to 2ml of 50% glutathione-sepharose beads equilibrated in 10ml NETN. After co-incubation during which the GST-tagged protein should bind to the beads, the beads are spun down and washed several times with 0.2M Borate pH8. The washed beads are then transferred to a column and cross-linked to the beads using DMP solution before washing with 0.1M glycine-Cl pH2.5 to remove non-covalently linked molecules and a final wash with 1xTBS . The terminal bleeds were thawed and 1/10 volume 10xTBS was added. This TBS-serum was then incubated in a column containing beads bound to GST for 30min at 4°C. Thus any antigens specific to GST that might be in the guinea pig sera would bind to these beads and the eluate from this column should be free of such contaminants. This eluate was incubated in the column described in the preceding paragraph and incubated at 4°C for 1 hour. Thus any antibodies specific to the FUN fragment should bind to these beads. The column was washed several times with TBS and 0.1 x TBS to remove any unbound molecules. Finally the column was eluted with 0.1M glycine-Cl pH2.5. 6ml of this 0.1M glycine-Cl pH2.5 was added to the column, and each sequential 800μl was collected separately. There were two guinea pigs murdered for this experiment designated 13 and 14. The 2nd elution for GP13 and the 3rd elution for GP14 had the highest concentrations of protein as measured by nanodrop. Purified antibody GP14-3 was then used for immunoblots on a transverse section of normal immature tassel, cross section of normal immature tassel and on normal SAM tissue.
Tissue was sectioned from wax using a microtome. Sections were laid on a slide, and the wax dissolved with Histoclear. Histoclear was then washed off with 100% EtOH and then the tissue was slowly rehydrated step-wise with lower concentrations of EtOH until finally pure water. The slides were then boiled in citrate buffer and left to stand for 10min. Slides were then incubated in PBS and then blocking solution . Next, the slides were incubated in 2ml of blocking solution and 2μl of antibody overnight, before incubating with the secondary antibody GP-AP, washing, and finally NETN and NBT/BCIP. Slides were examined under a light microscope and signal was observed at the base of the pedicel in the longitudinal section , and in the L2 in the transverse section , of the tassel. A similar localisation pattern was observed by in situ for TS1 RNA35. No signal was observed in the longitudinal section of the SAM, though it is possible that there is signal at the base of the axillary meristem . As a control, a longtitudinal section of a fun tassel was also used for an immunoblot . A similar localisation pattern was observed in this mutant control as was seen in the normal tassel casting doubt on the reliability of the wild-type localisation patterns since the fun mutation causes a truncation of the protein before the region that the antibody was designed to . This apparent non-specificity prompted a western blot using antibody GP14-3. No band was observed at the predicted size of 90kDa, and a 30kDa band was observed in both mutant and normal samples . Due to this lack of specificity, it was decided to make another antibody, using the entire third exon of FUN. The pENTR clone containing the 53xF/R insert was amplified by Phusion with primers AV53 and AV54 to create a DNA fragment containing the third exon of FUN flanked by appropriate restriction sites. The fragment was then cut with XhoI and BamHI, purified by phenol:choloroform extraction and ligated into the His-tagged expression vector pet. After verification by sequencing, this plasmid was transformed into Rosetta cells, grown in a 1L culture and induced with IPTG and purified with Ni++ beads as described above. This time, most protein was eluted at the 3rd pH3 elution. As before, this was dialysed into 6M urea and sent to Cocalico for injection into Guinea Pigs 15 and 16. In a dot plot, the test bleeds from GP15 and 16 showed strong reactivity against the antigens sent to Cocalico, no reactivity with control purified Tru1 protein, and a little reactivity with purified Bd1 protein, possibly due to the His tag on this protein. The secondary bleeds showed similar results. Recombination of pENTR-53xF/R and pDEST15 was unsuccessful, so pgex5x-2, a GST-fusion expression plasmid, was used instead. pENTR-53xF/R was amplified by Phusion with primers AV63 and AV64 for blunt end cloning. This PCR fragment was incubated with Polynucleotide Kinase to remove phosphoryl groups at the 3’ end and phosphorylate the 5’ end. pgex5x- 2 was cut with SmaI and incubated with Shrimp Alkaline Phosphotase to remove both 5’ and 3’ phosphate groups. The cut plasmid was then ligated to the PCR product with T4 ligase at 16°C overnight. Sequencing showed successful recombination, so the pgex5x-2-53xF/R plasmid was transformed into Rosetta cells and grown in a 1litre culture. This culture was induced with IPTG and made into a column, as described above. The pellet and supernatant of this Rosetta-pgex5x-2-53xF/R were used in a dot plot against an anti-GST antibody to test if the protein made was soluble, which it was .The pgex5x-2-53xF/R column was then used to purify sera from GP15 and GP16 .This western also found a >100kDa band in WT crude extract which was not present in fun crude extract, and a ~34kDa band that was present in both WT and fun crude extract . Crude extract of protein was carried out by grinding 2g of tissue in liquid nitrogen and adding 1ml of loading buffer , which was centrifuged before loading into gel.