AmericanChemicalSociety.com

Results from our laboratory over the past decade have shown that brain microtubule proteins including tubulin, tau and MAP2 are very sensitive to changes in thiol redox state. In addition to oxidation of cysteine residues of these cytoskeletal proteins, we detected S-glutathionylation using an anti-glutathione antibody. This observation from 2004 prompted us to seek better tools to probe these protein thiol reactions/modifications that are easy to use, safe and reliable. To study S-glutathionylation, we are in the process of studying the reactivity of fluorescein- and biotin-labeled derivatives of reduced (GSH) and oxidized (GSSG) glutathione with model proteins.

Two undergraduate students worked on this project lab during the 2008 10 week summer session and continued during the 2008-09 academic year. Considerable progress was made during summer 2008 and now we can synthesize and purify milligrams of singly (F-GSSG) and doubly fluorescein-labeled GSSG (F-GSSG-F). Toward that goal, we reacted GSSG with fluorescein isothiocyanate (FITC) in 0.1 M ammonium bicarbonate (pH 8-8.5) solution rather than in a neutral phosphate buffer. Under the new conditions, FITC reacts completely with excess GSSG thereby eliminating the need to remove unreacted FITC in our purification.

F-GSSG, F-GSSG-F and unreacted GSSG were successfully separated using a solid phase C8 column. Surprisingly, attempts to reduce F-GSSG-F to F-GSH were unsuccessful perhaps owing to the steric hindrance of the two fluoresceins in proximity to the disulfide bond. However F-GSSG was reduced to form F-GSH and GSH. Again using a C8 column, F-GSH was purified from GSH and excess reducing agent and quantitated using a fluorescein standard curve.

With both F-GSSG-F and F-GSH at our disposal, we began to study the reactions of these glutathione analogs with proteins including tubulin, a mixture of heat stable MAPs (predominantly MAP2 and tau), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), lactate dehydrogenase (LDH), alcohol dehydrogenase (ADH), creatine kinase (CK) and papain. Reduced proteins (PSH) can undergo thiol-disulfide exchange with F-GSSG-F to yield oxidized protein (PSSP) and F-GSH. During this process, some mixed disulfide between protein and F-GSH was detected by dot blot and by SDS-PAGE under nonreducing conditions. Addition of a disulfide reducing agent removes the fluorescein-label from all protein samples tested.

2PSH (red) + F-GSSG-F (ox) → PSSP (ox) + 2F-GSH (red) + PSSG-F

This exchange experiment yielded only very low levels of glutathionylated proteins. This is not unexpected because the desired mixed disulfide is merely an intermediate in the thiol-disulfide exchange process.

Much more significant labeling of protein thiols with F-GSH was accomplished by combining reduced protein, F-GSH and an oxidizing agent such as H₂O₂. Labeling was especially pronounced for GAPDH, CK and tubulin.

PSH + F-GSH + H₂O₂ → PSSG-F + 2H₂O

Thus, our concept is valid. Fluorescein-labeled glutathione derivatives will react with proteins and therefore we are encouraged to continue our efforts in this area. During the 2008-09 academic year, the two undergraduate students who worked on the project during summer 2008 continued their work. Among the areas of current study: assaying the effect of F-GSH labeling on enzymatic function (for GAPDH, LDH and ADH) and performing protein digestions to determine if all, or only select cysteines are labeled with F-GSH.

Our focus during summer 2009 was on the synthesis and purification of biotinylated GSH. Though a synthesis and purification of biotinylated GSH (B-GSH) has been published, we were not successful at repeating the published procedure. Two new undergraduate students followed the procedures described above for the GSSG/FITC reactions using 7.5 mM GSSG and limiting concentrations of NHS-biotin (6 mM). The reaction was performed in 0.1 M phosphate pH 8.0 and all NHS-biotin was consumed in 18-24 hours. Biotin is more polar than fluorescein so the final separation of B-GSH from GSH required the addition of 0.1 % acetic acid to the reaction prior to separation on the C8 column and the elution solvents contained 0.1% acid as well.

Biotinylated glutathiones (B-GSH and/or B-GSSG-B) are more challenging to work with because biotin does not have a detectable chromophore. We use an avidin-horseradish peroxidase (avidin-HRP) conjugate to detect biotinylated proteins on a PVDF or nitrocellulose membrane. Avidin-HRP, bound to our biotinylated proteins, is detected when a substrate produces chemiluminescence that we capture on film.

Our most significant accomplishment was to prove that our fluorescein-labeled glutathiones react in an identical manner with model proteins as the biotinylated glutathiones. Biotinylated GSH has been used by several research groups to study protein S-glutathionylation. For all model proteins tested (7 proteins total), we did not observe any significant differences in reactivity with protein thiols. We varied time, concentration and confirmed that reducing agents remove the protein labels.

We have also performed competition studies using unlabeled GSSG. As we increase the percent of unlabeled GSSG, our protein labeling with either fluorescein or biotin decreases by an expected amount. For example if we combine 25% GSSG and 75% labeled GSSG, we observe about 75% of the labeling we would have observed with 100% labeled GSSG. Thus, it appears that we have not altered the “core” structure by modifying the amino-terminus of the glutathione tripeptide.

A manuscript describing the synthesis, purification and introductory work with fluorescein-labeled GSH and GSSG is near completion. The two undergraduates who worked on this project (one funded by PRF) are continuing their research during the 2009-10 academic year.