We report a new bead based flow cytometry assay to measure in real time, the affinity and kinetics of nucleotide binding to small GTPases, using GST-fusions of Rab7 and RhoA as test cases. The particular features of the assay that assure broad utility are the facts that: (1) it circumvents the need for radiolabeled analogues; (2) only femtomoles of purified protein are needed per assay; (3) the assay may be run on a conventional flow cytometer; (4) the assay is quick and measures binding in real time without large numbers of wash steps; and (5) it may be run with or without magnesium.
The assay enables the sensitive identification of comparative binding differences between GTPases and their nucleotides, though as executed here does not establish the absolute kinetic and affinity constants for the GTPase. The nucleotide binding pocket of GTPases is nearly always occupied by GDP except in association with molecular chaperones that promote exchange and stabilize the nucleotide free form [19
]. Consequently, considerable effort must be taken to purify small GTPases free of nucleotide and magnesium as described by Simon et al. [11
]. Since the procedure of stripping nucleotide is laborious and accompanied by loss of activity, it is advisable only in situations where it is critical to determine absolute constants. The dissociation constant and kinetic rates determined by flow cytometry are not directly comparable to those determined in nucleotide free systems where a Kd
= 6 nM for Rab7 binding to methylanthraniloyl (mant) GTP was reported [11
]. Instead, our values agree well with those measured under conditions allowing nucleotide exchange for BODIPY FL analogues and using fluorescence spectroscopy [21
]; in which case the nucleotide binding affinities (Kd
) for Cdc42, Rac1, RhoA, and Ras ranged between =10–100 nM. The use of a reduced Mg2+
system diverges from the normal in vivo environment that GTPases reside in. Mg2+
are important for securing the nucleotide into the binding pocket and Mg2+
drastically slows nucleotide off rates [41
]. Because the flow cytometry based measurement reflects exchange of endogenous nucleotide for fluorescently labeled BODIPY FL GTP, we found it useful to remove Mg2+
ions, which tend to limit this exchange (c.f. ). Nevertheless, as shown here, the applicability of our assays is independent of Mg2+
, but can be used when desired or necessary.
The GST-Rab7 binding affinities for several nucleotides and nucleotide analogues yielded three interesting results. First, although Rab7 has a 3-fold higher affinity for GDP Kd
= 54 nM than for GTP Kd
= 150 nM, however, this is a relatively small difference considering that the cell and tissue concentration of GTP is in the micromolar range (19–276 µM with erythrocytes having the most variable levels) and 2- to 9-fold higher than GDP [45
]. Thus, in the cellular context it makes sense that the nucleotide bound state must be regulated by accessory proteins to preclude purely concentration dependent binding. Second, recent work by Korlach et al. [21
] suggests that BODIPY FL nucleotides may act analogous to guanine nucleotide exchange factors (GEFs) and aid displacement of endogenous, unlabeled nucleotide. However, because the bound BODIPY FL nucleotides were completely competed off with unlabeled nucleotides we suggest that the BODIPY moiety is not a major factor in nucleotide exchange or binding. Third, Rab7 is sensitive to substitution at the γ-phosphate position. A recent paper by Zhang et al. suggests that during nucleotide exchange on Ras the incoming nucleotide (GTP) attacks the nucleotide binding pocket of Ras through its β- and γ-phosphate groups to displace GDP [43
]. The observed sensitivity of GST-Rab7 to changes in the γ-phosphate region of the nucleotide may imply a conserved mechanism for nucleotide exchange among the Ras-related GTPases.
Unlike many other GTPases, GST-Rab7 exhibited a marked inability to bind BODIPY FL GTP when the fluorophore was attached to the γ-phosphate as in BODIPY FL GTP-γ-NH. Relative to RhoA, we find that GST-Rab7 exhibits a 90% decrease in binding affinity to BODIPY FL GTP- γ-NH, while GST-RhoA has no preference. This result is nicely explained by structural differences in the nucleotide binding pockets of the two GTPases with Rab7.
GTPases play a crucial role in regulation of a multitude of cellular functions. An understanding of the specific differences between GTPases, and their nucleotide binding will become increasingly important as we try to gain insight into mechanisms of disease. We have presented a rapid and sensitive assay for detection of nucleotide binding dynamics. With only femtomoles of protein required for individual measurements, and without complicated purification steps this assay approach can be used to quickly investigate and compare nucleotide binding of individual mutant GTPases, make comparisons of GTPases involved on the same pathway, and using modified beads for high throughput screening, identify effector molecules and or small molecule inhibitors and activators (PubChem AID # 757–61, and AID 764; http://nmmlsc.health.unm.edu/assays.shtml