I want to thank them both for their tireless enthusiasm for science, and especially Jess for taking me under her wing when I first joined the lab. I also wanted to thank the friendly faces at the Rathskeller, especially Mike and Jorge. Given the importance of OGT in the brain, we further investigated the regulation of the OGT enzyme by phosphorylation.
Introduction
O-GlcNAc Glycosylation
Tools to Study O-GlcNAc Glycosylation
O-GlcNAc Transferase
O-GlcNAcase
O-GlcNAc in Metabolism and Stress
O-GlcNAc in Neurons
Challenges to Studying O-GlcNAc Glycosylation
Conclusion
Generation and Characterization of a Conditional Forebrain-
- Results
- Generation of a Conditional OGT Knockout Mouse
- Loss of O-GlcNAcylation is Accompanied by Progressive Neuronal
- Changes in Anxiety and Deficits in Fear Conditioning Accompany
- Deficits in Additional Memory Tests Suggest Short-term Memory
- OGT cKO Recapitulates AD phenotypes Including
- OGT is Decreased in Human AD Brain
- The Role of Excitotoxicity in Neuronal Death After Loss of OGT . 66
- Discussion
- Conclusion
- Materials and Methods
- Reagents
- Mice
- Immunohistochemistry
- Behavioral Studies
- Microarray
- Chemoenzymatic Labeling
- TUNEL Labeling
- Tau Sedementation Assay
- Tensor Based Morphometry Analysis
- MTT Excitotoxicity Assays
- References
Figure 2.15 Place preference research suggests short-term memory deficits and identifies a possible hyperactive phenotype. 65 Figure 2.26 MTT testing is inconclusive as to whether OGT loss enhances neuronal excitotoxicity to glutamate.
Investigating the Regulation of OGT
Introduction
Understanding the regulation of OGT is an extremely important next step in uncovering how O-GlcNAc glycosylation contributes to cellular signaling in healthy and diseased states. Investigations of OGT enzyme activity and specificity have focused primarily on characterizing the TPR domains of OGT and determining how substrates interact with OGT. Specifically, CaMKIV was shown to dynamically modulate OGT in response to KCl depolarization in a calcium influx-dependent manner in a neuroblastoma cell line ( 3 ).
Although the site of CaMKIV phosphorylation of OGT was not identified, this was the first study to characterize a potential mechanism for OGT activation. This suggests that AMPK modification of OGT may play a role in regulating OGT activity on nuclear proteins and may alter epigenetic factors through histone dynamics. Although these two examples provide valuable insight, they are limited in their scope and do not identify the role of these post-translational modifications of OGT in altering O-GlcNAsylation of specific substrates, or the functional consequences of these changes.
These studies identified several phosphorylation sites, including S993, S994, S481, which are all in or near the C-terminal catalytic domain of OGT. To better understand this phenomenon, we sought to generate phosphorylation-specific antibodies to facilitate the detection of OGT phosphorylation at these sites and to investigate the mechanisms that regulate these phosphorylation events. Phosphorylation-specific antibodies have been widely used to investigate the role of phosphorylation in proteins in response to cell signaling (5).
Finally, a previously developed FRET probe was used to assess changes in OGT activity in living cells in response to cellular stimuli ( 11 , 12 ).
Results
- Neurite Outgrowth
- Generation of Phosphorylation-Specific Antibodies
- OGT Activity Assays
- FRET Probe
After affinity purification, the antibody was concentrated and re-evaluated by dot blots and showed specific binding to only the desired peptides (Figure 3.3A). To confirm that the antibodies recognize OGT, an immunoprecipitation (IP) of exogenously expressed FLAG-OGT with FLAG beads was performed and the purified phosphorylation-specific antibodies were evaluated by Western blot (Figure 3.3C). Demonstration of a significant change in OGT enzyme activity upon mutation of a phosphorylation site would strongly indicate that phosphorylation of OGT at these sites regulates its activity.
First, a previously developed radioactive activity assay was used to determine OGT activity in vitro after overexpression in 293T cells and purification with FLAG beads ( 3 , 8 ). Because we were most interested in the role of this regulatory modification in response to neuronal depolarization, we moved to neurons to see if KCl depolarization affected OGT activity. In an attempt to improve the throughput of the OGT activity assay, a Thorson probe was used to determine UDP concentration, which is a product of OGT enzymatic activity (9, 10).
Optimization of the assay revealed that detergent, DTT, and FLAG beads attenuated the Thorson probe signal, and these normal components of the OGT activity assay were omitted. Unfortunately, after several rounds of optimization, no response to OGT activity was observed using the Thorson sensor (Figure 3.5). To evaluate changes in OGT activity either in response to cellular signaling or modulating post-translational modifications, we sought to use a FRET assay that would respond to dynamic changes in OGT activity in living cells.
After depolarization, a general increase in fluorescence is observed for CFP and YFP, and the ratio of YFP to CFP does not change significantly in the intensity of the raw signal (Figure 3.7).
Discussion
After confirming efficient targeting and expression, N2A cells were transfected with the general FRET probe and FRET efficiencies were analyzed in response to KCl depolarization. Furthermore, treatment with PBS instead of KCl also showed no changes in FRET probe fluorescence. The time course shows no change in FRET efficiency after KCl depolarization (t=0) with control (unglycosylatable) FRET probe.
To further study these phosphorylation sites, phosphorylation-specific antibodies were generated to aid in the detection of these modified forms of OGT in response to specific cellular stimuli. As the phosphorylation-specific antibodies are still being evaluated, a variety of methods have been used to examine OGT activity in vitro in an attempt to quantify possible changes in mutant OGT activities. In the most robust OGT activity assay, the radioactivity assay, S994A/E mutants were shown to have significantly reduced activity, comparable to that of catalytic dead (D925A) OGT in 293T cells.
It is likely that this assay is not sensitive enough to detect OGT activity in this manner, and although it would greatly improve the throughput of OGT activity assessment, this assay does not appear to be a viable option. Finally, the FRET probe was shown to respond to KCl depolarization, and both PBS treatment and KCl depolarization of a non-glycosyllabile control probe produced no response. Despite the observed change in FRET efficiency after KCl depolarization, the dynamic range of the probe in the current state is limited and would likely require increased FRET probe expression or more rigorous data analysis to serve as a useful assessment of OGT activity in living cells.
Using the RIFRET program, it appears that some regions of the cell have significant increases in FRET efficiency with KCl depolarization; however, it is unlikely that these changes are large enough to effectively serve as a proxy for OGT activity in living cells.
Conclusion
A more acceptable target would be to treat the cells with specific perturbations such as calcium chelators or pathway inhibitors to see if they have an effect on OGT activity rather than point mutations. Both Thorson and FRET probe assays appear to be insufficiently sensitive to reliably assess OGT activity and would need to be further optimized to be useful in this capacity. One possibility is to obtain a second O-GlcNAc fret probe based on CaMKII dynamics, which was recently described (13).
If it could be shown that the activity of mutant (S994A) OGT was not altered by KCl depolarization in neurons or neuron-like cells and that the activity of wild-type OGT was induced by this stimulus, this would support our hypothesis that phosphorylation of OGT at S994 ga activates for KCl depolarization, calcium influx, and CaMKII activation.
Methods
- Neurite Outgrowth Assay
- Generation of Phosphorylation-Specific Antibodies
- Dot Blots and Westerns
- KCl Depolarization of Neurons
- OGT Activity Assays
- FRET Assay
- Thorson Assay
Peptides were conjugated to thyroglobulin (Pierce) and sent to Cocalico laboratories for antibody production. Neurons were blocked with 1 uM tetrodotoxin (TTX) and 10 uM D-AP5 for 12 h before addition of 55 mM KCl. Lysates were spun down at max G for 10 min and supernatants were incubated with FLAG beads for 1 h at 4C, end to end.
An aliquot of FLAG beads was reserved for WB to quantify OGT protein levels in each sample. Activity assays were performed in 40 µl TAB with 1 mM casein kinase II peptide (genscript, PGGSTPVSSANMM) and 0.2-0.4 uCi tritiated UDP-GlcNAc (Perkin Elmer, NET434050UC). Reactions were run at room temperature for 2 h consecutively, quenched with 50 mM formic acid, 500 mM NaCl, and peptides isolated with C18 spin columns (Pierce 89873).
293T, N2A or neurons grown on glass bottom plates (In Vitro Scientific) were transfected with various FRET probe constructs, allowed to grow for 2-3 days and imaged. Images were analyzed in ImageJ (NIH) and the intensities for CFP and YFP (CFPex) were quantified. After immunoprecipitation, OGT was eluted from FLAG beads with 3X FLAG peptide (Sigma, F4799) at 100 µg/ml with 3 elutions of 10 µl each for 30 µl beads.
When ready for reading, UDP-GlcNAc was added to a final concentration of 200 µM and the plates were read at 488 nm.
Carrillo LD, Krishnamoorthy L, & Mahal LK (2006) A cellular FRET-based sensor for beta-O-GlcNAc, a dynamic carbohydrate modification involved in signaling. Carrillo LD, Froemming JA, & Mahal LK (2011) Targeted in vivo O-GlcNAc sensors reveal discrete compartment-specific dynamics during signal transduction. 2013) Diabetic hyperglycemia activates CaMKII and arrhythmias through O-linked glycosylation.
Developing Tools to Study OGT
- Introduction
- Results
- Developing a Cellular OGT Knockout System
- Discussion
- Conclusion
- Methods
- Production of 293T GOT Knockdown Cells
- Chemoenzymatic Labeling with GalT
- Lentiviral Constructs
- Production of Lentivirus
- References
These mice have the first exon of OGT flanked by loxP sites, resulting in excision of the first exon after CRE recombinase expression ( 4 , 5 ). Thus, isolation of cells from these mice and subsequent treatment of the dissociated cultures with CRE-expressing lentivirus would allow efficient knockout of OGT. We characterized the time course of OGT knockout in neurons after addition of CRE recombinase.
To understand why OGT lentivirus was unsuccessful, we generated a catalytically dead form of OGT (D925A) in a lentiviral construct and were successfully able to infect 293T cells. cell line with OGT knockdown does not stably maintain low O-GlcNAc levels. Two different OGT antibodies (DM17, AL28) were used to evaluate OGT protein levels, showing almost complete loss of OGT protein by 10DIV. In an attempt to limit the effects of OGT expression during lentiviral packaging, we first transfected virus-producing cells with the packaging proteins, in the absence of the OGT-containing lentiviral construct.
In this way, translation of OGT would be off during virus production, but after the virus was produced and used to infect neurons, CRE recombinase could be added to excise the stop codon and enable OGT expression. To study the effects of OGT mutations or TPR truncations, a clean background would be necessary. This represents a new strategy to study the effects of OGT loss, which is not normally possible in dividing cells.
Although this system provides a potential source of OGT knockout cells, several challenges had to be addressed before we could use this system experimentally.
