Kinetics of recruitment and allosteric activation of ARHGEF25 isoforms by the heterotrimeric G-protein Gαq

Rho GTPases are master regulators of the eukaryotic cytoskeleton. The activation of Rho GTPases is governed by Rho guanine nucleotide exchange factors (GEFs). Three RhoGEF isoforms are produced by the gene ARHGEF25; p63RhoGEF580, GEFT and a recently discovered longer isoform of 619 amino acids (p63RhoGEF619). The subcellular distribution of p63RhoGEF580 and p63RhoGEF619 is strikingly different in unstimulated cells, p63RhoGEF580 is located at the plasma membrane and p63RhoGEF619 is confined to the cytoplasm. Interestingly, we find that both P63RhoGEF580 and p63RhoGEF619 activate RhoGTPases to a similar extent after stimulation of Gαq coupled GPCRs. Furthermore, we show that p63RhoGEF619 relocates to the plasma membrane upon activation of Gαq coupled GPCRs, resembling the well-known activation mechanism of RhoGEFs activated by Gα12/13. Synthetic recruitment of p63RhoGEF619 to the plasma membrane increases RhoGEF activity towards RhoA, but full activation requires allosteric activation via Gαq. Together, these findings reveal a dual role for Gαq in RhoGEF activation, as it both recruits and allosterically activates cytosolic ARHGEF25 isoforms.

Scientific RepoRts | 6:36825 | DOI: 10.1038/srep36825 lacks the N-terminal palmitoylation sites. The plasma membrane localization of both these p63RhoGEF isoforms have been shown to be important for their interaction with Gαq 22,23 .
More recently a third, longer variant encoded by ARHGEF25 was described in the literature 24 . This isoform has a different N-terminus, comprises 619 a.a. in total and is indicated here as p63RhoGEF 619 . It was found in several mass spectrometry based proteomic analyses of human tissues 25,26 . The p63RhoGEF 619 variant is transcribed from two alternative N-terminal exons and, similar to GEFT, does not contain the N-terminal palmitoylation sites. The subcellular localization and functional properties of p63RhoGEF 619 in relation to the other two isoforms have not yet been investigated. Furthermore, the mechanism of activation of ARHGEF25 isoforms in the cytoplasm by Gαq has not yet been elucidated.
Here we use advanced microscopy techniques to characterize the mechanism with which the different ARHGEF25 isoforms (p63RhoGEF 580 , GEFT and p63RhoGEF 619 ) are activated by Gαq coupled GPCR stimulation. We show that although the subcellular localization of the different ARHGEF25 isoforms is strikingly different, they do not show any difference on basal and stimulated GEF activity towards RhoA. P63RhoGEF 580 and p63RhoGEF 619 can activate RhoA and Cdc42 via stimulation of endogenous Gαq coupled histamine-1-receptors (H 1 R). Furthermore we show that p63RhoGEF 619 relocates to the plasma membrane upon activation of Gαq. Using a chemical dimerization system, we are able to show that recruitment of p63RhoGEF 619 to the plasma membrane increases RhoGEF activity towards RhoA, but full activation requires allosteric activation by Gαq. This study highlights the dual function of Gαq, recruitment and allosteric activation, in the activation of cytosolic ARHGEF25 isoforms.

Results
The longer isoform of ARHGEF25 is located in the cytoplasm. P63RhoGEF/GEFT was first identified as a transforming gene product in a retroviral screen for oncogenes 27,28 . Subsequently it was found that the ARHGEF25 locus produces two isoforms, p63RhoGEF 580 (580 a.a.) and its shorter variant GEFT (474 a.a.) 20 . More recently, a third and longer variant was described 24 , p63RhoGEF 619 , which is transcribed from two alternative exons at the N-terminus, and contains 619 a.a. in humans (see Supplemental Fig. 1 for a full sequence alignment of the three isoforms).
To compare the subcellular localization of the different ARHGEF25 isoforms, HeLa cells were transfected with a plasma membrane (PM) marker (Lck-CFP), a cytosolic marker (RFP) and either p63RhoGEF 580 -YFP, p63Rho-GEF 619 -YFP or YFP-GEFT. Confocal images of cells transfected with p63RhoGEF 580 -YFP show that p63Rho-GEF 580 is localized to the plasma membrane, as has been described before 22,23 . In contrast, cells transfected with either p63RhoGEF 619 -YFP or YFP-GEFT show only a cytosolic localization of these isoforms (Fig. 1A,B).
Previous studies have shown that p63RhoGEF 580 is localized to the plasma membrane because of the palmitoylation of one or more cysteines at its the N-terminus 22,23 . In depth sequence analysis of the first 30 amino acids shows that the N-terminus, including cysteines, of p63RhoGEF 580 is well conserved throughout the whole subphylum vertebrata (Fig. 1C, top). GEFT is essentially a splice variant of p63RhoGEF 580 that does not contain the first 2 N-terminal exons of p63RhoGEF 580 that contain the palmitoylated cysteines, resulting in cytoplasmic localization.
Sequence analysis of the relatively unexplored isoform p63RhoGEF 619 shows that the N-terminus is well conserved among the vertebrate subphylum (Fig. 1C, bottom). There are no obvious PM targeting signals in the N-terminus, which is in agreement with the cytoplasmic localization observed in Fig. 1A. Of note, there seems to be a well-conserved 'GRTD' sequence at amino acid 10-13. Evolutionary sequence analysis (see Methods for details) reveals that both p63RhoGEF 580 and p63RhoGEF 619 are not found outside the vertebrate subphylum, and even so-called early or ancient vertebrate species like the Lancelets 29 and Lampreys 30 do not contain either isoform. Our results fit with a role for ARHGEF25 in smooth muscle (see discussion). To conclude, we find that p63RhoGEF 580 is located at the plasma membrane, whereas the two other isoforms show cytoplasmic distribution.
The basal GEF activity of ARHGEF25 isoforms is low in cells. It has been well established that p63RhoGEF 580 can be activated through G-protein coupled receptors (GPCRs), via direct interaction with the heterotrimeric G-protein subunit Gαq 12,14,22,31 . Activated p63RhoGEF 580 catalyzes the exchange of GDP for GTP on Rho GTPases of the RhoA subfamily 12 . Even though it is located at the plasma membrane, p63RhoGEF 580 has been confirmed to have low GEF activity towards RhoA in vitro 12 and in cells 21 due to the auto-inhibition by its PH domain. To compare the GEF activity towards RhoA of the different ARHGEF25 isoforms in unstimulated cells, we transfected HeLa cells with the previously characterized DORA RhoA biosensor 21 and p63Rho-GEF 580 -RFP, p63RhoGEF 619 -RFP, RFP-GEFT or only a cytosolic RFP (control) (Fig. 2A). The ratio between YFP and CFP fluorescence of the DORA RhoA biosensor, which confers the GTP loading state of RhoA, was determined for control (median = 0.61, 95% CI = 0.58-0.66) p63RhoGEF 580 (median = 0.63, 95% CI = 0.61-0.67), p63RhoGEF 619 (median = 0.69, 95% CI = 0.61-0.73) or GEFT (median = 0.66, 95% CI = 0.61-0.72) ( Fig. 2A,B). The values for p63RhoGEF 580 , p63RhoGEF 619 and GEFT are slightly elevated compared to the control condition, but are in line with low basal activity previously reported for p63RhoGEF 580 21 . These results demonstrate that all different ARHGEF25 isoforms have low basal activity towards RhoA in unstimulated cells.
Both p63RhoGEF 580 and p63RhoGEF 619 activate RhoA and Cdc42. In order to investigate whether the difference in location translates to a functional difference between the ARHGEF25 isoforms, we used DORA FRET biosensors for RhoA 21 , Rac1 32 and Cdc42 33 to measure GEF specificity and activity in stimulated cells. Hela cells were used in this study, because these cells express the endogenous histamine-1 receptor that couple to Gαq. In previous experiments, we have shown that in control conditions, i.e. cells transfected with mCherry 34 or inactive variants of p63RhoGEF 21 , RhoA-GTP levels are not elevated upon stimulation with histamine. When p63RhoGEF 580 -RFP is ectopically expressed in Hela cells, a rapid activation of RhoA is observed after stimulation of the cells with histamine.
We transfected HeLa cells with p63RhoGEF 580 -RFP, p63RhoGEF 619 -RFP or a cytosolic RFP (control) and one of the three aforementioned Rho GTPase FRET sensors. CFP fluorescence and YFP sensitized emission were followed over time. Cells were stimulated with histamine to activate Gαq via endogenous histamine-1-receptors, and the response was antagonized with mepyramine. Stimulation of cells transfected with either p63RhoGEF 580 or p63RhoGEF 619 showed a reversible increase in YFP/CFP ratio of the DORA RhoA biosensor, with similar kinetic profiles (Fig. 3A). Stimulation of cells in the control condition showed no change in YFP/CFP ratio of the DORA RhoA biosensor. Stimulation of cells transfected with GEFT also showed a similar response on the DORA RhoA biosensor (Supplemental Fig. 2A), which was completely inhibited if cells were incubated with the specific Gαq inhibitor FR900359 35 (Supplemental Fig. 2B). This is in line with earlier results obtained with  p63RhoGEF 580 21 . Hardly any change in YFP/CFP ratio of the DORA Rac biosensor was observed after stimulation of cells transfected with either p63RhoGEF 580 or p63RhoGEF 619 compared to control cells transfected only a cytosolic RFP (Fig. 3B). Interestingly, we found that cells transfected with either p63RhoGEF 580 or p63RhoGEF 619 showed a robust reversible increase in YFP/CFP ratio of the DORA Cdc42 biosensor after stimulation with histamine and mepyramine. Control cells expressing the DORA Cdc42 sensor and a cytosolic RFP did not show any change in YFP/CFP ratio upon stimulation with histamine and mepyramine (Fig. 3C). Overnight incubation with Pertussis Toxin (PTX) 36 to inhibit Gαi activity did not change the response in cells transfected with the DORA Cdc42 biosensor and p63RhoGEF 619 , excluding a Gαi mediated effect (Supplemental Fig. 3A). Incubation of cells transfected with the DORA Cdc42 biosensor and p63RhoGEF 619 with Gαq inhibitor FR900359 abrogated the histamine mediated response on the DORA Cdc42 biosensor, showing that this effect is mediated by the activation of p63RhoGEF 619 via Gαq (Supplemental Fig. 3B).
Hence, despite the difference in their subcellular locations, p63RhoGEF 580 and p63RhoGEF 619 show surprisingly similar activation profiles on the Rho GTPase they can exert their GEF activity on. p63RhoGEF 619 is recruited to the plasma membrane after activation of a GPCR. Since we measured similar Rho GTPase activation profiles with a cytosolic (p63RhoGEF 619 ) and plasma membrane (p63RhoGEF 580 ) located variant of the same GEF, we set out to explore the mechanism behind this unexpected observation in more detail. It is well known that LARG, PDZ-RhoGEF and p115RhoGEF, RhoGEFs that are activated via direct interaction with Gα 12/13 37 , relocate to the plasma membrane upon activation 38 . To investigate whether p63RhoGEF 619 also relocates to the plasma membrane upon activation of Gαq, we used real-time confocal microscopy to inspect the subcellular location p63RhoGEF 619 during stimulation of endogenous histamine-1-receptors. Pilot experiments indicated that the ectopic expression of p63RhoGEF 619 leads to a quick saturation of available binding sites of endogenously activated Gαq. To overcome this, we transfected HeLa cells with p63RhoGEF 619 -YFP, Gαq-CFP and H 1 R-RFP. After stimulation of the cells with histamine we observed a rapid relocation of p63RhoGEF 619 to the plasma membrane, which was reversed upon addition of mepyramine (Fig. 4A). This response was quantified by drawing ROIs in the cytoplasm of individual cells (Fig. 4B). Upon stimulation of cells transfected with YFP-GEFT, Gαq-CFP and H 1 R-RFP we observed a similar reversible relocation of GEFT to the plasma membrane, albeit with a lower amplitude (Fig. 4C, Supplemental Fig. 4A). We observed similar relocation kinetics for p63RhoGEF 619 and GEFT when cells were co-transfected with both constructs at the same time (RFP-GEFT, p63RhoGEF 619 -YFP, Gαq-CFP and H 1 R-untagged) (Supplemental Fig. 4B). From this data we conclude that the ARHGEF25 isoform p63RhoGEF 619 and splice variant GEFT, which reside in the cytosol in unstimulated cells, relocate to the PM upon activation of Gαq mediated signaling.
Full activation of RhoGEF activity requires heterotrimeric G-protein signaling. The observation that p63RhoGEF 619 relocates to the plasma membrane argues for a dual role of activated Gαq where it both recruits p63RhoGEF 619 to the plasma membrane and relieves it from its allosteric autoinhibition by the PH domain. To separate these events we made use of the rapamycin heterodimerization system 39 . In this approach, addition of rapamycin will recruit p63RhoGEF 619 to the plasma membrane without allosteric activation by Gαq. Subsequent activation of a GPCR will result in allosteric activation without a contribution of relocation. By simultaneously monitoring the RhoGEF activity with the DORA RhoA biosensor we can quantify the contribution of both events to RhoGEF activity. To this end, HeLa cells transfected with the DORA RhoA biosensor, Lck-FRB-ECFP(W66A) and p63RhoGEF 619 -fkbp12-RFP were sequentially treated with rapamycin, histamine and mepyramine (Fig. 5A). After adding rapamycin, a gradual increase in YFP/CFP ratio of the RhoA biosensor is observed, most likely due to some constitutive activity of p63RhoGEF 619 at the plasma membrane towards RhoA. (Fig. 5B,C red trace). Subsequent stimulation with histamine fully activates Gαq, leading to the allosteric release of the inhibitory PH domain of p63RhoGEF 619 , and rapid and large change in YFP/CFP ratio on the RhoA biosensor. The histamine-mediated activation of RhoA is then fully reversed upon addition of mepyramine (Fig. 5B,  red trace, Fig. 5D). A similar experiment with the splice variant GEFT (RFP-fkbp12-GEFT) revealed an identical pattern in YFP/CFP ratio changes on the RhoA biosensor (Supplemental Fig. 5A). In contrast, when we recruited the isolated catalytic domain of ARHGEF25 (cDH-fkbp12-RFP) with rapamycin, we observed a large and fast change in ratio of YFP/CFP of the RhoA biosensor (Fig. 5B,C, black trace, Fig. 5D). The ratio did not change further upon stimulation with histamine or mepyramine, indicating that relocation of the isolated catalytic GEF domain of ARHGEF25 to the plasma membrane is sufficient for full activation of RhoA as previously reported 21 . Recruitment of an fkbp12-RFP serves as a control since this unit has no RhoGEF activity. Indeed, recruitment of fkbp12-RFP does not lead to any change in YFP/CFP ratio of the RhoA biosensor, whereas addition of histamine leads to a small reversible change in FRET ratio that can most likely be explained by the activation of the endogenous Gαq mediated GEF Trio 40 (Fig. 5B, grey trace, Fig. 5C). Of note, when we compared the raw ratios of the RhoA biosensor in all conditions, we observed only an increased basal YFP/CFP ratio of the RhoA biosensor in cells transfected with cDH (Supplemental Fig. 5B). This increased basal ratio has been observed before by us 21 , and is likely due to the random interactions of the constitutive active cDH construct and the RhoA biosensor in the cytoplasm. The raw FRET ratio data also shows that the maximum YFP/CFP ratio change induced by rapamycin in the cDH condition is very comparable (25%) to the ratio change induced by rapamycin + histamine in the p63RhoGEF 619 /GEFT conditions (Supplemental Fig. 5B).
Together, from these results we conclude that recruitment of p63RhoGEF 619 increases RhoGEF activity but that maximal stimulation requires allosteric activation via heterotrimeric G-protein signaling.

Discussion
The two undisputed gene products of ARHGEF25, p63RhoGEF 580 and p63RhoGEF 619 , display obvious differences in subcellular localization when expressed in unstimulated HeLa cells. Five putative palmitoylation sites at the N-terminus of p63RhoGEF 580 target this isoform to the plasma membrane 22,23 , while p63RhoGEF 619 resides in the  cytoplasm. Interestingly, despite this difference in localization, we observed no difference in basal GEF activity towards RhoA. Surprisingly, both isoforms reversibly activate RhoA with similar magnitudes upon stimulation of Gαq coupled GPCRs. Closer inspection of p63RhoGEF 619 revealed that this isoform dynamically relocates to the plasma membrane during GPCR stimulation on a seconds timescale. Of note, the RhoGEF Trio, which has a DH/PH domain homologous to p63RhoGEF, has been shown to relocate to the plasma membrane after activation of Gαq at much slower timescales 41 . These observations resemble the well-known relocation mechanism of the Gα 12/13 mediated RhoGEFs p115RhoGEF, PDZ-RhoGEF and LARG 38 .
It was previously shown that recruitment of the isolated catalytic DH domain of p63RhoGEF from the cytosol to the plasma membrane leads to robust activation of RhoA 21 . By synthetically recruiting p63RhoGEF 619 to the plasma membrane we could separate the effects of plasma membrane recruitment of p63RhoGEF 619 on the one hand, and allosteric activation by Gαq on the other hand. P63RhoGEF 619 mediated GEF activity towards RhoA is slightly enhanced when it is artificially recruited to the plasma membrane, but full GEF activity requires stimulation by Gαq. This reveals a double role for Gαq in cells, where it both recruits and allosterically activates P63RhoGEF 619 .
Interestingly, we also found GEF activity towards the Cdc42 of both p63RhoGEF isoforms, while we did not find evidence of Rac activation. Although Cdc42 and Rac activity of both p63RhoGEF 580 42 and p63RhoGEF 619 24 has been described before, these results are controversial 12,20 . Our results with the Gaq inhibitor FR900359 indicate that the Cdc42 activity most likely originates, directly or indirectly, from Gαq mediated activity.
The ARHGEF25 gene and its protein products are only found in the vertebrate subphylum. It seems likely that ARHGEF25 gene arose from a split gene event of either Trio 40 or Kalirin 43 , which are both found in vertebrate predecessors, and contain a DH/PH cassette that is highly homologous in sequence to ARHGEF25. Interestingly, recent efforts with proteomics approaches 25,26 have found expression of both p63RhoGEF 619 and p63RhoGEF 580 specifically in human tissues that also contain smooth muscle, including the brain (vasculature), fetal heart, uterus, reproductive tract, urinary tract, gastrointestinal tract and the retina. The fact that smooth muscle tissue has evolved only in vertebrates 44 , and p63RhoGEF function has been extensively linked to vascular smooth muscle cell function 16,17,28,45 and hypertension 18,19,46,47 , provides strong evidence for co-evolution and a functional connection between p63RhoGEF and smooth muscle tissue. Which of the ARHGEF25 isoforms is expressed in vascular smooth muscle tissue, whether they can be specifically targeted and if one of isoforms is main cause of hypertensive pathogenicity, should be subject of future studies.
In conclusion, we compared subcellular localization and functional output of two isoforms of the gene ARHGEF25, p63RhoGEF 580 and p63RhoGEF 619 . Interestingly, while the two isoforms display remarkable differences in subcellular localization, p63RhoGEF 580 is located at the plasma membrane, and p63RhoGEF 619 located in the cytosol, they activate the RhoGTPases RhoA and Cdc42 to a similar extend upon Gαq mediated GPCR stimulation. We found that p63RhoGEF 619 relocates to the plasma membrane upon activation of Gαq, very similar to the mechanisms of Gα 12/13 mediated RhoGEFs. Synthetic recruitment of p63RhoGEF 619 to the plasma membrane slightly increases its GEF activity towards RhoA, but full activation requires allosteric activation by Gαq. The resemblance of signaling output between a cytoplasmic and membrane located GEF is striking. Future studies may reveal whether signaling output of the two variants is similar under physiological conditions.
In order to construct p63RhoGEF 619 -RFP, a plasmid containing the following nucleotide sequence was ordered (Eurofins): 5′-ggactcagatctcgagctcaagcttcgaattctgcagtcga cggtaccaccatgaagcccccggaccgc cccgcccctggccgcac tgaccggatactgggggtcatggggggcatgctgcgcgcatgcgccctccctgggcaggaggggcccccaaggagaagccctctagggttggtgggtaccgagccagagtctgaacgtacggagggagatcacagaagggatcgcgaacatgaggtcctcgccggggctctgcagcccga atcctattccattgcgggcagtgaggggagtatatcggcttctgctgcctccggtctggctgccccctctggccccagctctggcctcagctctggcccctgttccccaggccc cccagggcccgtcagtggcctgaggagatg gttggatcattccaaacattgtctcagtgtgg aaactgaggcagacagtggtcaggcagga-3′. Cutting the plasmid with HindIII and PshAI (restriction sites in bold and underlined) and ligating it into a plasmid containing p63RhoGEF 580 -RFP, cut with the same enzymes, resulted in p63RhoGEF 619 -RFP. P63RhoGEF 619 -YFP was obtained by swapping the RFP for YFP with AgeI and BsrGI. p63RhoGEF 619 -fkbp12-RFP was obtained by cutting p63RhoGEF 619 -RFP with HindIII and BsrGI and ligating fkbp12, cut with the same enzymes from N1-fkbp12-mTurquoise2, in between the sequence for p63RhoGEF 619 and RFP.
The Lck-FRB-ECFP(W66A) was a kind gift from M. Putyrski 39 . The DORA Rho GTPase sensors were described previously; DORA-RhoA 21,50 , DORA-Rac1 sensor 32 and the DORA-Cdc42 sensor 33 . N1-fkbp12-mTurquoise2 was a kind gift from Kevin Batenburg. Plasmids constructed in this study will be made available through Addgene: https://www.addgene.org/Dorus_Gadella/ Cell Culture & Sample Preparation. HeLa cells (American Tissue Culture Collection: Manassas, VA, USA) were cultured using Dulbecco's Modified Eagle Medium (DMEM) supplied with Glutamax, 10% FBS, Penicillin (100 U/ml) and Streptomycin (100 μg/ml). Cell culture, transfection and live cell microscopy conditions were performed as previously described 21 . Widefield microscopy. Ratiometric FRET measurements were performed using a previously described wide-field fluorescence microscope 21 . Typical exposure times ranged from 50 ms to 150 ms, and camera binning was set to 4 × 4. Fluorophores were excited with 420 nm light (slit width 30 nm) and reflected onto the sample by a 455DCLP dichroic mirror and CFP emission was detected with a BP470/30 filter, and YFP emission was detected with a BP535/30 filter by rotating the filter wheel. All acquisitions were corrected for background signal and bleedthrough of CFP emission in the YFP channel (around 55% of the intensity measured in the CFP channel).

Confocal microscopy.
Localization and relocation experiments were performed using a Nikon A1 confocal microscope equipped with a 60x oil immersion objective (Plan Apochromat VC, NA 1.4). The pinhole size was set to 1 Airy unit (<0.8 μm). 457 nm, 514 nm and 561 nm laser lines were reflected onto the sample by a 457/514/561 dichroic mirror. To avoid bleed-through, images were acquired with sequential line scanning modus. CFP, YFP and RFP emission was detected using BP482/35, BP540/30 and BP595/50 emission filters respectively. Data Analysis. ImageJ (National Institute of Health) was used to analyze the raw microscopy images.
Further processing of the data was done in Excel (Microsoft Office) and graphs and statistics were conducted using Graphpad version 6.0 for Mac, GraphPad Software, La Jolla California USA, www.graphpad.com. The non-parametric two-tailed Mann-Whitney test was used to determine p values. We list exact p-values where possible, unless p values < 0.1e-3, since GraphPad does not give exact p-values in these situations.
Boxplots were generated online, using the website http://boxplot.tyerslab.com/, showing individual data points, the median (center line) and 95% confidence intervals (notches). Box limits indicate the 25 th and 75 th percentiles as determined by R software and whiskers extend 1.5 times the interquartile range (IQR) from the 25 th and 75 th percentiles. The notches are defined as +/−1.58 * IQR/ √(n) and represent the 95% confidence interval for each median.
The p-values and 95% confidence intervals are determined to aid the interpretation and are not used as thresholds to determine biological significance 51 .
To compare the intensities of the GEF, PM marker and cytosolic marker, line plots in Fig. 1 were made by normalizing the fluorescent intensities in the different channels to the maximum value. Sequence Analysis. Protein sequences from other species were obtained from -and evolutionary analysis of p63RhoGEF 580 and p63RhoGEF 619 was performed using a combination of BLAST (NCBI), Ensembl genome browser 52 and the UCSC human genome browser 53 . Sequences were aligned and visualized using Jalview 54 .