Photo-activation of the hydrophobic probe iodonaphthylazide in cells alters membrane protein function leading to cell death

Background Photo-activation of the hydrophobic membrane probe 1, 5 iodonaphthylazide (INA) by irradiation with UV light (310–380 nm) results in the covalent modification of transmembrane anchors of membrane proteins. This unique selectivity of INA towards the transmembrane anchor has been exploited to specifically label proteins inserted in membranes. Previously, we have demonstrated that photo-activation of INA in enveloped viruses resulted in the inhibition of viral membrane protein-induced membrane fusion and viral entry into cells. In this study we show that photo-activation of INA in various cell lines, including those over-expressing the multi-drug resistance transporters MRP1 or Pgp, leads to cell death. We analyzed mechanisms of cell killing by INA-UV treatment. The effects of INA-UV treatment on signaling via various cell surface receptors, on the activity of the multi-drug resistance transporter MRP1 and on membrane protein lateral mobility were also investigated. Results INA treatment of various cell lines followed by irradiation with UV light (310–380 nm) resulted in loss of cell viability in a dose dependent manner. The mechanism of cell death appeared to be apoptosis as indicated by phosphatidylserine exposure, mitochondrial depolarization and DNA fragmentation. Inhibition by pan-caspase inhibitors and cleavage of caspase specific substrates indicated that at low concentrations of INA apoptosis was caspase dependent. The INA-UV treatment showed similar cell killing efficacy in cells over-expressing MRP1 function as control cells. Efflux of an MRP1 substrate was blocked by INA-UV treatment of the MRP1-overexpressing cells. Although INA-UV treatment resulted in inhibition of calcium mobilization triggered by chemokine receptor signaling, Akt phosphorylation triggered by IGF1 receptor signaling was enhanced. Furthermore, fluorescence recovery after photobleaching experiments indicated that INA-UV treatment resulted in reduced lateral mobility of a seven transmembrane G protein-coupled receptor. Conclusion INA is a photo-activable agent that induces apoptosis in various cancer cell lines. It reacts with membrane proteins to alter the normal physiological function resulting in apoptosis. This activity of INA maybe exploited for use as an anti-cancer agent.


Background
Cells have developed a complex architecture that relies on the compartmentalization of cellular functions within organelles that are bounded by lipid membranes. The plasma membrane constitutes a unique interface between the cytoplasm and extra cellular milieu. While ensuring the physical separation of two very different environments, a constant communication between the cell and its extracellular milieu is established by means of cell surface proteins. Signaling via membrane proteins regulate various cellular functions including cell survival, cell propagation, cell differentiation and cell migration. Therefore membrane proteins are considered prime targets for drugs designed to combat cancer and other diseases. While inhibition of a given cell surface receptor often leads to predictable changes in cells based on the signaling cascade initiated by the receptor, it is unclear how altering cell signaling via a number of receptors would change the cell physiology.
INA is a hydrophobic photo-reactive probe that reacts with transmembrane anchors of membrane proteins upon photo-activation with UV light [1]. INA has been used for the labeling and identification of integral membrane proteins [2][3][4][5][6][7], in the study of membrane dynamics and fusion [8][9][10] and for the detection of protein-membrane interactions [10][11][12][13]. Recently, we have utilized the transmembrane protein anchor reactivity of INA for inactivation of enveloped viruses [14][15][16][17] with preservation of structural integrity for vaccine application. When applied to a variety of enveloped viruses, INA could selectively eliminate functions associated with the hydrophobic domain of the viral envelope required for inducing membrane fusion [16,17]. Based on this premise we hypothesized that treatment of cancer cells with INA-UV would result in inactivation of signaling functions of cellular membrane proteins which in turn would inhibit cell signaling and survival.
With this in mind we conducted the present study to show that INA, a non toxic compound in itself, is highly toxic to a variety of tumor cells including multidrug resistant cells in the presence of UV light. Treatment of cells with INA-UV resulted in inhibition of signal transduction by certain cellular receptors and cell death. The cell death induced by INA-UV showed signs of a classical apoptosis pathway with the involvement of caspases. This study provides evidence that INA can act as a novel and highly active therapeutic agent with a mechanism of action that seems distinct from existing photodynamic therapy compounds.

Photo-activation of INA inhibits cell viability
INA is a hydrophobic compound that partitions in the membrane of cells. Upon irradiation of cells with UV light (310-380 nm) the azido moiety is converted into a highly reactive nitrene that covalently binds membrane proteins. This process leads to the selective inactivation of functions associated with those proteins in the hydrophobic domain of the membrane as has been previously shown for isolated cell membranes [18] and enveloped viruses [14][15][16][17]. In this study, we examined the effect of INA-UV on the viability and membrane associated functions of different tumor cell lines. We treated a human breast cancer cell line MCF7 with varying concentrations of INA and irradiated with UV light. As seen in figure 1A (Table 1).

Photo-activation of INA induces apoptosis
Next we asked whether the loss of viability seen by INA-UV treatment was due to apoptosis induced in these cells. Apoptosis is characterized by a variety of distinct cellular changes like phosphatidylserine exposure, mitochondrial depolarization and DNA fragmentation. MCF7 cells treated with various concentrations of INA-UV were analyzed for apoptosis markers like mitochondrial depolarization using DiOC 6 dye (figure 2A) or phosphatidylserine exposure via Annexin V staining (figure 2B). A dose dependent induction of apoptosis was seen in these cells. To further characterize whether the cells were undergoing apoptosis we used the TUNEL assay to detect DNA fragmentation. As seen in figure 2C the cells treated with INA-UV also showed DNA fragmentation confirming the apoptotic mechanism of cell death.

Apoptosis mediated by INA-UV is caspase dependent
Caspases are cysteine proteases that are key mediators of apoptosis [19]. The involvement of caspases in the apoptotic pathway can be studied by the inhibition of apoptosis via the pan-caspase inhibitor Z-Val-Ala-Aspfluoromethylketone (ZVADfmk) and the cleavage of caspase specific substrates like PARP [20,21]. To determine whether apoptosis mediated via INA-UV treatment was caspase dependent, we treated SupT1 cells with ZVADfmk (40 μM) prior to INA-UV treatment. As seen in figure 3A and 3B INA-UV mediated apoptosis at low concentrations was inhibited by ZVADfmk confirming the role of caspases. Interestingly higher concentrations were not inhibited by ZVADfmk suggesting a different mechanism at very high concentrations. In order to monitor caspase activation in SupT1 cells 24 h post treatment with INA-UV, the cells were labeled FITC-VAD-FMK which binds irreversibly with active caspases and analyzed by flow cytometry. As seen in figure 3C, high activation of caspases is observed at 5 μM which correspond to the IC 50     tions a caspase-independent apoptotic process may be involved.

INA-UV induced cell killing is not affected by multidrug resistance proteins
An important problem that arises during chemotherapy is the emergence of drug resistant cells [22]. A major contributor of that phenomenon is an increased efflux of the drugs facilitated by proteins such as the P-glycoprotein (Pgp) or the multidrug resistance protein (MRP) [23][24][25].
To test whether INA was affected by this phenomenon we used 293/MRP1 cells, a stably transfected 293 cell line that continuously expresses the MRP1 gene at high levels [26].
Indeed, those cell lines exhibited a much lower sensitivity to the commonly used chemotherapeutic agent doxorubicin ( figure 5A). However, the IC 50 of INA-UV cell killing was insensitive to the presence of the multidrug resistance associated gene MRP1 (figure 5B, table 1). Similarly, INA-UV treatment was not significantly affected by Pgp overexpression (table 1).

INA-UV treatment blocks the MRP1 induced efflux
MRP1 is a complex transmembrane protein with 17 membrane spanning domains that mediates efflux of various substrates from the cytoplasm of the cell to the extracellular milieu. Calcein is an anionic fluorescent probe that

INA-UV treatment induces apoptosis
acts as a substrate for MRP1 and determination of its cellular accumulation and efflux allows the investigation of MRP1 activity [27]. As observed in figure

INA-UV affects CXCR4 signaling
CXCR4 is a seven transmembrane G protein-coupled chemokine receptor that is over-expressed in a variety of

INA-UV induced apoptosis in SupT1 cells is mediated by caspases
tumors and involved in tumor metastasis [28]. In cell membranes, blocking the signaling by INA-UV treatment of another G protein-coupled receptor, human chorionic gonadotropin (hCG) has been previously documented [18]. We wished to determine whether INA-UV treatment would alter signaling via CXCR4. Binding of the CXCR4 ligand SDF-1α to its receptor induces a calcium flux in cells, which is monitored by ratio of fluorescent signals at 340 and 380 nm excitation, respectively, and 510 nm emission using the fluorescent calcium indicator dye, Fura 2 pre-loaded into the cells [29,30]. Upon addition of SDF1α we observed calcium flux in the control SupT1 cells but not in cells pretreated with INA-UV (figure 7) suggesting an inhibition of CXCR4 signaling. By contrast, INA-UV treatment had no effect on the action of the calcium ionophore 4-bromo A-23187 that directly mediates calcium flux across membranes. When added to SupT1, in both cases the cytosolic concentration of calcium rose to a similar level as was shown by the increase of the fura 2 ratio signal (Figure 7). Hence the effect of INA-UV was at the level of CXCR4 signaling and not a general effect on intracellular calcium accumulation.

INA-UV does not inhibit IGF1 signaling
Growth factor receptors including EGFR and IGF1R are often overexpressed in tumor cells. Overexpression of IGF1R is associated with increased survival and prolifera-tion of tumor cells. Hence we wished to determine whether INA-UV had an effect on IGF1R signaling and whether it was related to the apoptosis seen in previous experiments. The binding of IGF1 to its receptor activates the PI3kinase pathway leading to Akt phosphorylation at serine 473 [31], which has been reported to be involved in apoptosis inhibition. As can be seen on figure 8, the pretreatment of MCF7 cells with INA-UV leads to an amplification of Akt phosphorylation induced by IGF1 compared to untreated cells. While DMSO partially contributed to this effect, further amplification was observed upon irradiation in the presence of INA. Interestingly we observe that INA-UV treatment alone, in the absence of IGF1, can partially induce Akt phosphorylation in MCF7 cells. This effect requires the reaction to transmembrane proteins with INA, since it is not observed with UV and DMSO treatment alone. These results suggest that INA-UV may in fact activate signaling via IGF1R contrary to the results seen with other receptors. These results underscore the complex physiological outcome of INA interaction with cellular receptors and the diversity in the signal transduction mechanism by different receptors.

Treatment affects the mobility of membrane proteins
Our efforts to identify the mechanism of action of INA-UV mediated apoptosis induction suggest a direct effect on membrane proteins. Although CXCR4 and MRP1 func-

Discussion
INA is a highly hydrophobic molecule that partitions with high specificity in cellular membranes [32]. This high hydrophobicity along with its photoactivable property makes it a specific probe for transmembrane anchors of membrane proteins. We have exploited this specificity towards transmembrane anchors to inactivate a variety of viral membrane proteins for design of new vaccine candidates [15,17]. In this study we studied the effects of INA-UV treatment on whole cells. Our results indicate that treatment of cell lines with INA by itself was non toxic; however, in the presence of UV light the treatment mediated loss in viability of numerous cancer cells.
The photoactivation of INA is mediated by its azido group that is sensitive to UV irradiation. Upon UV irradiation, a highly reactive nitrene radical is formed that covalently reacts with transmembrane proteins in its vicinity [1]. Such covalent modifications of proteins has been shown to result in the complete loss of infectivity of several enveloped viruses [14][15][16][17] and correlates with the loss of membrane fusion function of the viral fusion proteins [16,17]. The transmembrane segments of the fusion proteins of retroviruses [33] and influenza virus [34] are indeed critical for the full fusion process to take place. Although replacement of the transmembrane anchor of influenza virus hemagglutinin with a lipid-anchor obliterated its full fusion capacity, the lipid-anchored hemagglutinin still promoted the hemifusion phenotype leading to the mixing of the outer layers of the viral and target cell membranes [34]. ing the apoptotic pathway. However, at higher concentrations of INA, caspase activation is decreased and accordingly ZVADfmk becomes ineffective in preventing mitochondrial depolarization and PS presentation suggesting a caspase independent pathway of apoptosis as seen with other treatments like hexaminolevulinate PDT of lymphoma cells [35]. A caspase independent cell death referred to as sub apoptosis has been previously docu-  mented [36]. The death inducing stimulus might be such that apoptosis can be achieved through release of apoptosis inducing factors (AIF) from mitochondria without the participation of caspases [37].

INA-UV treatment blocks SDF1α induced CXCR4 signaling
As the dosage of INA is increased, more targets within the transmembrane region and possibly other membrane compartments are likely to be reached by the treatment and this can alter the mechanisms inducing the death of the treated cells. If activated INA will covalently react with any protein that is deeply anchored in the lipid bilayer, the consequence of this covalent modification will depend on the function of this portion of the particular protein.
Understanding the effect of INA on cellular proteins is important for determining the mechanism of apoptosis induction by INA and its development as an anti cancer agent.
It has been previously reported that INA-UV treatment of cell membranes induces inactivation of gonadotropin receptor by uncoupling of the response of adenylate cyclase to gonadotropins [38]. While binding of chorionic gonadotropin and the luteinizing hormone to gonadotropin receptor were preserved, the binding failed to induce stimulation of the adenylate cyclase pathway. On the other hand, this pathway was shown to be still functional when stimulated directly with NaF. The luteineizing hormone receptor is a member of the G protein-coupled receptor, a family of receptors displaying seven predicted transmembrane helices. CXCR4, another member of this family, has recently been shown to be overexpressed in many cancer cell lines [39,40]. This overexpression is largely due to the hypoxic conditions in the tumor environment [41] and can favor the metastasis of cancer cells through its ligand SDF1α [42][43][44]. We show here that INA-UV treatment blocks CXCR4 mediated calcium signaling generated by SDF1α stimulation. At the same time the calcium gradient is preserved in the INA-UV treated cells as it is still sensitive to the effect of calcium ionophores indicating that the integrity of the treated cell membranes is not compromised. These data indicate a direct inactivation of CXCR4 receptor signaling by INA-UV treatment.
Similarly, the activity of MRP1 a member of the ABC transporters superfamily that along with Pgp is involved in the drug resistance phenotype in various cancers [45] was also affected by INA-UV treatment. Although no crystal structure is yet available for this protein, ABC transporters are thought to be composed of clusters of predicted transmembrane helices [46]. Overexpression of drug resistance genes makes cells several orders of magnitudes less sensitive to conventional chemotherapeutic agents. The mechanism of action is not clearly understood but the working hypothesis of "hydrophobic vacuum cleaner" has been proposed [7,45] whereby the hydrophobic chemotherapeutic agents while partitioning into the membrane will interact with the transporter within the lipid domain of the bilayer and be pumped outwards.  Akt phosphorylation was assessed by Western detection and quantified. The different samples correspond to MCF7 cells untreated, exposed or not to 1% DMSO, INA at the indicated concentrations and/or IGF1. All the samples exposed to DMSO or INA were irradiated with UV. The Akt phosphorylation signal was normalized to the signal obtained with MCF7 exposed only to IGF1. The Western signal obtained with GAPDH on the same samples is shown for loading control. pSer473-AKT GAPDH IGF1 stimulation, which was further enhanced upon incubation with IGF1. This suggests that INA treatment may activate IGF1R consistent with reports that a mutation in the transmembrane anchor of IGF1R can result in activation of the receptor [47]. The effects of INA can be interpreted through intramolecular effects via the chemical modification introduced by the covalent addition of a hydrophobic moiety in the transmembrane segment of a protein. However protein-protein and protein-lipid interactions are also likely to be affected by this modification. Proper signaling within cells relies on a dynamic reorganization upon stimuli of the signaling receptors within different domains of the membrane that are thought to assemble and disassemble for the signaling to proceed [48]. We show here by FRAP analysis that INA-UV treatment considerably reduces the mobile fraction of proteins within plasma membranes. Whether this is due to a partial aggregation of membrane proteins and/or a reorganization of domains to accommodate the enhanced hydrophobicity needs to be further studied. The enhanced basal activation of Akt following INA and light treatment might also be a result of receptor immobilization/redistribution. Furthermore, the ability of IGF1 to stimulate the tyrosine kinase IGF1R receptor is considerably amplified by INA treatment. IGF1R has been shown to relocalize in membrane "raft" microdomains in MCF7 cells upon stim-ulation with IGF1 [49] and activation of Akt also appears to rely on its membrane redistribution [50] in membrane "raft" microdomains [51]. Nevertheless Akt activation induced by INA treatment does not prevent the cells to undergo apoptosis suggesting the mechanism of cell death induced by INA is independent of Akt signaling pathway.
In this report we show the potency of INA as a novel and efficient photoactivable chemotherapeutic agent. The activity of this treatment is strictly dependent on activation by UV light and is mediated by the covalent reaction of INA with membrane embedded domains of proteins. Unlike conventional photodynamic sensitizer that are dependent on reactive oxygen species for activity, reaction of INA with membrane proteins has been shown to be increased under hypoxic conditions [52]. Such hypoxic conditions are common in tumors micro environment and present a major challenge for other photosensitizers. Furthermore, while INA can be directly activated by UV, an equivalent activation can be obtained through energy transfer processes called photosensitization using a variety of chromophores as photosensitizers [6]. The result presented here show that INA is a very potent light activatable therapeutic agent whose targets and mechanism of action are very different from existing PDT agents. Those

IGF1 mediated signal transduction
Signaling was measured by following the insulin-like growth factor 1 (IGF1) dependent phosphorylation of Akt protein [31]. MCF7 cells were plated in a 12 well plate in RPMI. The medium was replaced with serum free RPMI for overnight incubation. INA was added from 100× stock solution in DMSO for each sample and the samples were irradiated with a UV lamp as described above. IGF1 (R&D systems, Minneapolis, MN) was then added to the medium at a final concentration of 40 ng/ml and the samples were incubated for 10 minutes at 37°C. The cells were then scraped on ice in PBS. The cells were lysed in RIPA buffer supplemented with 2 mM sodium orthovanadate. Proteins from the lysate were separated by electrophoresis and subjected to Western blot analysis for the detection of phosphorylated Akt.

Western blot analysis
Upon blotting, the membranes were incubated for 1 h in Odyssey blocking buffer (LICOR, Lincoln, NE). The blots were then incubated with the appropriate primary antibody for 2 h at room temperature in Odyssey blocking buffer containing 0.2% Tween-20. The following dilutions were used: 1/1000 for phospho-Akt, 1/4000 for Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and 1/500 for PARP. Upon four washes of 10 min each with 0.1% Tween-20 in PBS (PBST), the membranes were incubated one hour with anti rabbit 800 and anti mouse 700 (LI-COR, Lincoln, NE) at a dilution of 1/5000 in Odyssey blocking buffer with 0.2% Tween-20, and washed four times for 10 min with PBST. Immunoreactiv-ity was detected by using an Odyssey infrared imaging system (LI-COR, Lincoln, NE).

CXCR4 chemokine induced signaling
SupT1 cells were treated with 15 μM INA and labeled with 5 μM of Fura 2 for 30 minutes at room temperature. Calcium flux was monitored in a FluoroMax-3 fluorimeter from Horiba Jobin-Yvon (Edison, NJ, USA). The cells were resuspended at 10 6 /ml and the fluorescence emission was recorded at 510 nm using simultaneous excitation at 340 and 380 nm. The calcium flux was triggered by the addition of the chemokine SDF1α (PeproTech, Rocky Hill, NJ) at a final concentration of 35 ng/ml. Complete equilibration of the intracellular and extracellular calcium concentration was achieved by the addition of the ionophore 4bromo A-23187 at a final concentration of 5 μM.

Apoptosis
Apoptosis was detected in cells 24 h post treatment using various assays. Mitochondrial depolarization was detected by staining with 10 nM 3,3'-dihexyloxacarbocyanine iodide (DiOC 6 ) for 30 min at 37°C followed by flow cytometry. Annexin V-FITC was used to detect PS exposure on cells using ApoAlert kit (BD bioscience). DNA fragmentation was detected by terminal uridine deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) using In situ cell death detection kit (Roche). For PARP cleavage analysis, SupT1 cells were treated with various amounts of INA or doxorubicin. 24 hours upon treatment, the cells were lysed with RIPA buffer for electrophoresis and Western detection. CaspACE™ FITC-VAD-FMK (Promega) was used following manufacturers instruction to detect caspase activation by FACS analysis 24 h post treatment.

FRAP measurement and analysis
Fluorescence recovery after photobleaching (FRAP) was performed as previously described [54] using a Zeiss LSM 510 (Carl Zeiss, Jena, Germany) confocal laser scanning microscope. HeLa cells were plated on 35 mm glass bottom dishes (MatTek, Ashland, MA) and transfected with CCR5-GFP (a kind gift from Dr Santos-Manes [55]) 24 hours prior to confocal analysis as described previously [54]. INA-UV treatment was performed as described above. The cells were then submitted to FRAP while kept at physiological conditions of 37°C and 5% CO 2 in a stage incubation system (Incubator S; PeCon GmbH, Erbach, Germany). A 488 nm Ar + laser line was used for excitation and emission light was collected with a 500-550 bandpass filter. A 40×/1.3 NA oil immersion objective lens was used with a zoom factor of 4. The detector pinhole was opened slightly to acquire an optical section of 2 μm thickness. This allowed more light to be collected for better quantification. Three pre-bleach images were acquired to determine the rate of non-purposeful photobleaching. Photobleaching was performed by increasing the transmission of the laser to 100% for 20-50 iterations to optimize the extent of bleaching. Following photobleaching 8-10 images were acquired at one second intervals and then the acquisition rate was changed to ten second intervals to follow the recovery to completion. A total of 20-40 images were acquired. FRAP analysis was performed using the MIPAV (CIT/NIH, Bethesda, MD) software package using a 1D diffusion FRAP model to retrieve the mobile fraction [56]. Data were automatically corrected with background subtraction, as well as normalization for the non-purposeful photobleaching rate calculated from the whole cell membrane.

MTS assay
The cells were plated a day before in a 96 wells plate at a density of 2 10 4 /well. The cells were then subjected to UV irradiation in the presence of INA. A day after the treatment, the cells were submitted to a 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4sulfophenyl)-2H-tetrazolium, inner salt (MTS) assay (CellTiter 96 ® AQ ueous One, Promega) as per the manufacturer's instructions. For IC 50 calculation, seven independent measurements were performed and analyzed together. IC 50 values were computed by fitting a four-parameter nonlinear regression model with R statistical software [57]. The standard error (SE) represents the 95% confidence interval.
part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract NO1-CO-12400. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organization imply endorsement by the United States Government.