Use of genetically encoded, light-gated ion translocators to control tumorigenesis

It has long been known that the resting potential of tumor cells is depolarized relative to their normal counterparts. More recent work has provided evidence that resting potential is not just a readout of cell state: it regulates cell behavior as well. Thus, the ability to control resting potential in vivo would provide a powerful new tool for the study and treatment of tumors, a tool capable of revealing living-state physiological information impossible to obtain using molecular tools applied to isolated cell components. Here we describe the first use of optogenetics to manipulate ion-flux mediated regulation of membrane potential specifically to prevent and cause regression of oncogene-induced tumors. Injection of mutant-KRAS mRNA induces tumor-like structures with many documented similarities to tumors, in Xenopus tadpoles. We show that expression and activation of either ChR2D156A, a blue-light activated cation channel, or Arch, a green-light activated proton pump, both of which hyperpolarize cells, significantly lowers the incidence of KRAS tumor formation. Excitingly, we also demonstrate that activation of co-expressed light-activated ion translocators after tumor formation significantly increases the frequency with which the tumors regress in a process called normalization. These data demonstrate an optogenetic approach to dissect the biophysics of cancer. Moreover, they provide proof-of-principle for a novel class of interventions, directed at regulating cell state by targeting physiological regulators that can over-ride the presence of mutations.

Importantly however, it is beginning to be seen that modulating carcinogenesis and metastasis is not as simple as targeting individual ion channel genes for lossor gain-of function applications. The key parameter can be a complex, non-cell-autonomous physiological state not intrinsically limited to any one specific channel or pump [3,[39][40][41][42][43]. Thus, it is important to exploit amenable model systems to understand how aspects of cancer are regulated by bioelectrical states of tissues in vivo. However, to fully understand and exploit the control of cell functions via bioelectrical signaling requires the development of new techniques that allow ion flux to be influenced in any cell/ tissue of interest with much greater spatio-temporal control than has been possible to date with pharmacological or genetic approaches.
Optogenetics is the expression of light-gated ion translocators and enzymes, with different kinetics and different wavelengths of activation, to control downstream processes, such as ion flux or enzyme activity. It has primarily been used as a precise tool to control neural and muscle excitation/inhibition, and to regulate biochemical processes [44,45]. Extending the use of optogenetics to non-excitable cells [48], our lab has successfully initiated Xenopus tail regeneration by hyperpolarizing cells using the light-dependent H + pump, Archaerhodopsin, thereby reversing the normal, age-dependent loss of regenerative ability [49]. Building on this work and recent data on the bioelectric control of tumorigenesis [43,50], we investigated here the potential applicability of optogenetics to hyperpolarization-induced tumor suppression.
In addition to exhibiting other classic tumor characteristics, these tumor-like structures maintain a depolarized membrane voltage. This bioelectric signature can be used to detect prospective ITLS regions before they become morphologically apparent [42]. Moreover, we have shown that this depolarization is not only a signature but is functionally required for tumorigenesis, by demonstrating that artificially hyperpolarizing oncogene-expressing cells, by expressing any of several ion channels, significantly reduces the incidence of ITLS formation despite high levels of otherwise-sufficient oncogene expression in the same tissue [3,4].
Given the importance of bioelectric signaling, and the need for high-resolution manipulation both for the clinic and for probing the basic biology of this process, we explored whether optogenetics can be used to improve our ability to regulate, in both time and space, the bioelectric signaling that is disrupted during cancer. We hypothesized that spatio-temporally-controllable light-gated ion translocators would enable precise control over ion flux (and therefore membrane voltage (V mem )) in a tissue of interest, and thus we sought to extend optogenetics to the biophysical control of oncogene-dependent tumorigenesis.
To test the ability of optogenetic tools to alter endogenous V mem and thereby suppress ITLS growth and promote tissue normalization, we used Archaerhodopsin (Arch) [69], a light-gated H + transporter that hyperpolarizes cells [49,70]. We confirm here that microinjection of KRAS G12D [71] mRNA, a tumor-inducing KRAS mutant, results in the formation of the ITLSs that we have previously shown to exhibit classic hallmarks of tumors, including histopathology, increased proliferation, lack of differentiation, attraction of vasculature, etc. [3,4,41]. We then demonstrate that co-injection of Arch mRNA and subsequent expression of Arch in KRAS G12D -expressing cells, followed by a 24-hour light-activation of the Arch transporter, reduced ITLS incidence by 32%. We also show that suppression and normalization of ITLSs are not specific to Arch, but can also be accomplished by injection of a different optogenetic reagent, channelrhodopsin-2 (ChR2 D156A ) [72]. Most interestingly, by delaying the activation of ChR2 D156A until ~ stage 35, we were able to convert fully developed ITLSs into normal cells. Thus we demonstrate the utility of optogenetics to suppress ITLS formation and to promote normalization of existing ITLSs into wildtype tissue.

Injection of KRAS G12D results in ITLS formation
To study the usability of V mem -altering optogenetic tools in oncogene-mediated tumorigenesis, we expressed a human oncogene in Xenopus embryos. Injection of the oncogene KRAS G12D [71] into Xenopus embryos (1 blastomere at the 16-cell stage) induced ITLSs ( Figure 1A), which have previously been shown to exhibit many of the defining hallmarks of their mammalian counterparts.

ChR2 D156A activity alters membrane voltage of Xenopus embryonic cells
To allow modulation of V mem via light activation of a channelrhodopsin-2 (ChR2 D156A ) channel, ChR2 D156A mRNA was injected into 1 cell of a 2-cell embryo, allowing the uninjected side to serve as an internal control ( Figure 1). ChR2 D156A is a non-specific cation channel that, at physiological pH, passes mostly protons, but there is also significant Na + and K + flux [45] [73]; it was selected because of the low incidence of side effects in Xenopus embryos [48]. Because of the extremely low ion concentration of the surrounding medium, light activation of this channel is predicted to hyperpolarize those cells due to efflux of cations; Table 1 gives the internal and external ion concentrations. Injected embryos, raised to stage 18, were exposed to blue light then imaged using the membrane voltage-sensitive dye DiBAC 4 (3), a semiquantitative method that has been extensively used to monitor relative resting-potential differences among cells in vivo [49,54,[74][75][76][77][78]. As predicted, we observed that the injected right half of the embryo was much dimmer, indicating relative hyperpolarization compared to the uninjected left side.

Arch and ChR2 D156A activities reduce KRAS G12Dinduced ITLS incidence
Co-injection of mRNAs for KRAS G12D and either Arch or ChR2 D156A into one cell of a 16-cell stage embryo resulted in the expression of the corresponding proteins as early as 4 hours post injection (data not shown). Embryos expressing light-activated ion channels were exposed to light for 24 hours: arch-expressing embryos were stimulated by green light, 555 nm, irradiance of 1 mW/mm 2 for 500 ms followed by 1.5s in darkness, while ChR2 D156A -expressing cells were stimulated by blue light: 470 nm light of 2.4 mW/mm 2 for 10 ms every 30 seconds ( Figure S2, 3A; [48]). To study ITLS prevention, embryos were exposed to the light beginning four hours post injection, or, approximately stage 9; to induce normalization of tumors, exposure was from stage 28 to stage 35. Two sets of controls were used for comparison: embryos injected with only KRAS G12D and un-stimulated embryos expressing the oncogene and either Arch or ChR2 D156A (light by itself does not affect V mem of cells that do not express light-gated channels, and tumor incidence does not vary among KRAS-only injected embryos kept in dark, ambient light, or blue/red optogenetic exposure [data not shown]).
Consistent with our previous finding that hyperpolarization suppresses ITLS formation, our experimental data show that Arch activity significantly reduces the number of embryos that develop KRAS G12D ITLSs by 32% compared to KRAS G12D -expressing embryos lacking Arch (χ 2 = 7.1, p = 0.007) ( Figure 3B). Similarly, the activity of ChR2 D156A was able to significantly lower KRAS G12D ITLS incidence by 31.4% (χ 2 = 6.8, p = 0.009) ( Figure 3B). Together, these data show that light-activated Arch and ChR2 D156A can reduce ITLS formation by KRAS G12D in non-excitable tissues, suggesting that lightdependent activity of optogenetic reagents is a V memaltering modality with efficacy in the suppression of ITLS formation.

Light-initiated ChR2 D156A activity normalizes fully developed KRAS G12D ITLSs
We next tested the utility of optogenetics in the conversion of existing ITLSs into normal tissue. Similar to the suppression experiments, ChR2 D156A was co-expressed with KRAS G12D ; in these experiments, however, stimulation of the light-dependent channel was delayed until stages between 28-35 when ITLSs were fully developed. Tadpoles were then scored for presence or absence of ITLSs when they reached stages 45-47 ( Figure 4A). Stimulation of ChR2 D156A -expressing cells within the tumors resulted in 31% more embryos having normalized their tumors -compared to their injected but un-stimulated counterparts (χ 2 = 8.6, *p = 0.003) ( Figure 4B). These results demonstrate that ChR2 D156A increases the number of oncogene-induced ITLSs that are normalized.
Moreover, optogenetic stimulation had a larger effect than several drugs considered promising anti-tumor therapies in human cancer medicine: the highly-selective MEK 1 inhibitor Selumetinib; the potent Pi3K inhibitor Pictillisib; and the inhibitor of oncogenic B-RAF kinase activity Vemurafenib (grey bars, Figure 4B).

ChR2 D156A activity regulates muscle marker expression in KRAS-induced ITLSs
To determine whether there is a link between ChR2 D156A activation and the expressions of tumor-specific transcriptional markers, we examined chd15 (a satellite cell marker expressed in differentiating myoblasts) and myod1 (a developing muscle cell marker expressed in newly formed somites and involved in differentiation). Cl -11.1 30

Figure 2: Set up used to deliver spatio-temporally precise light stimulation of optogenetic ion-translocators expressed in
Xenopus embryo ITLSs. We customized a Nikon AZ100 dissection scope for in vivo optogenetics [87] by replacing the epifluorescence illumination source and light guide with a Spectra4 LED illuminator connected to the scope via fiber optic cable (the black light guide that passes in front of the vacuum source). The light is passed through a pinhole to set spot size diameter (located behind the oculars, not visible in this image), then enters the scope and goes through an 80/20 splitter that allows the user to view the specimen on the monitor even while the LED is on. Finally, the light passes through the 5 × objective lens which further reduces spot diameter and aims the spot at the sample. An automated Ludl MAC6000 XY stage (that can also be manipulated manually by a joystick) allows multiple embryos to be exposed repeatedly to the activating wavelength of light. Up to thirty embryos are loaded into a slide-mounted PDMS "chip" designed to use microfluidics to hold embryos in place [100]; the chip is held to the slide by a vacuum and 0.1 × MMR is circulated by a peristaltic pump. The optogenetics components and the microscope are all controlled by NIS Elements.
We chose these two markers because the KRAS G12D induced rhabdomyosarcoma arises from skeletal muscle tissue, and because these markers signal the presence of rhabdomyosarcoma in human samples [79][80][81]. As expected from comparison with KRAS G12D -induced human tumors, ITLSs observed growing in a transgenic line of animals that drive muscle-specific GFP fluorescence [82] revealed the presence of ectopic muscle cells ( Figure 5). Using qPCR, we next found that both muscle markers were significantly up-regulated in tadpoles with KRAS G12D tumors ( Figure 6, green bars) compared to uninjected controls ( Figure 6, blue bars); however, the expressions of myod1 and cdh15 became normalized to control levels when ChR2 D156A was activated in tadpoles ( Figure 6, red  . Taken together, these data show that optogenetic modulation of ion flux and resting potential reduces the expression of tumor markers as well as normalizes the tissue structure.

DISCUSSION
Given the instructive role of bioelectric parameters in orchestrating cell behavior towards adaptive pattern formation [53,[83][84][85], our lab is interested in using modulation of resting potential in vivo as an approach to understanding misregulation of developmental signaling as occurs in cancer, birth defects, and other disease conditions. Previous work has shown that tumor-like structures induced in Xenopus laevis embryos by mammalian oncogenes (Figure 1A) exhibit many of the properties of tumors: overproliferation, expression of known tumor markers, attraction of vasculature, and histological disorganization [3,40]. They also possess a depolarized potential relative to adjacent cells that are not part of the ITLS -a property long-known to be associated with cancer cells [24,86], that was recently shown to be an instructive factor, not merely a read-out; actively preventing this depolarization can prevent the appearance of ITLSs [4]. In this study, we extend our ability to control tumorigenic pathways by expanding the use of optogenetics to this new system, and we exploit the temporal and cellular resolution afforded by light-gated ion channels to remove ITLSs after they have already been formed.
We have previously demonstrated the functionality of Arch in Xenopus embryonic and larval tissue: light activation of this hyperpolarizing H + pump restores tail regeneration capability at non-regenerative stages [49]. Here, we show that light activation of another light-gated channel (ChR2 D156A ) also induces hyperpolarization in nonexcitable Xenopus cells ( Figure 1C). This differs from mammalian neurons where high Na + concentration in the medium favors influx of the cation through channels, resulting in depolarization; Xenopus embryos grow in very-low extracellular Na + medium, thus opening of cation channels results in hyperpolarization [87].
Channelrhodopsin variant D156A was chosen because of many optogenetic constructs tested [48] it caused the fewest unwanted side effects on embryogenesis. While ChR2 does exhibit some desensitization under illumination, this was not a problem in this application because high spiking rates are not used to encode information in this kind of developmental bioelectricity context [45]. Moreover, exposure of ChR2 D156A was limited to 10 ms pulses, followed by 30s of darkness, i.e. a regimen than minimizes desensitization [88] and the shift away from a mixture of positive cations to primarily proton flux [89]. Thus, we are confident that stimulation of ChR2 D156A led to the hyperpolarization we measured due to efflux of a mixture of positive cations. We found that the surface expression of Arch and ChR2 D156A coupled with the non-toxic, high temporal-resolution light regimen, makes the optogenetic approach (Figure 2) suitable for targeting tumorigenesis in vivo.
We successfully exploited light-induced hyperpolarization of Arch expressing cells to decrease ITLS incidence among embryos expressing the KRAS G12D oncogene ( Figure 3A, 3B). Importantly, compared to expressing constitutively-open hyperpolarizing channels, the light regimen was delivered for only 24 hours, was temporally more precise and convenient, and the technique is equally effective at suppressing ITLSs. To rule out any Arch H + -pump-specific ITLS suppression effects, we also employed light activation of the ChR2 D156A nonspecific cation channel in similar experiments ( Figure 3A), thereby generating hyperpolarizing current by a different protein and a different mechanism, finding that passive cation efflux, like active H + -efflux, reduced the number of embryos with KRAS G12D -induced ITLSs ( Figure 3B). The results obtained using the two distinct optogenetic channels that are not restricted to one type of ion flux, are consistent with V mem per se being the instructive factor in regulating oncogene-mediated tumorigenesis [4,10]. We also demonstrate for the first time that lightgated channels expressed in cells of tadpoles with KRAS ITLSs can be non-invasively activated for a duration sufficient to significantly increase the frequency of ITLS normalization, an exciting finding suggesting the possibility of remission-induction in addition to prevention (Figure 3, 4). Moreover, we show that the performance of this non-invasive optogenetic stimulation is superior to other anti-tumor agents we tested, including those that have been shown to be promising in human cancer medicine: Selumetinib, a highly-selective MEK 1 inhibitor; Pictillisib, a potent Pi3K inhibitor; and Vemurafenib, a potent inhibitor of oncogenic B-RAF kinase activity, (grey bars, Figure 4B). While these are the first data testing these compounds in the Xenopus tumor assay, we cannot rule out that subsequent research could identify a treatment regime that would allow these compounds to be more efficacious.
Our qPCR data ( Figure 6) indicate that light-induced hyperpolarization of tumors also involves normalization of the expression of tumor markers, although it must be kept in mind that mRNA data do not necessarily reflect a linear correspondence to the presence of protein. Together, the data show that manipulation of bioelectric cell state, a powerful, tractable regulator of cancer cell normalization and reprogramming, is possible using optogenetics, thus introducing a new class of biomedical strategies for tumor treatment. The recent development of pharmacological approaches to render existing ion channels light-sensitive [90][91][92] suggests a next-generation approach that would not require introduction of transgenes (optogenetic channels) into the target tissue.
In summary, we report here the first use of optogenetics as a temporally precise regulator of V mem to suppress and normalize oncogene-induced ITLSs. More broadly, this tool will advance the study of resting potential as another important component of the microenvironment that is so crucial for cancer initiation and progression [93][94][95][96][97]. Our implementation of light-gated bioelectrical associated with KRAS mutations express the muscle markers myod1 and cdh15. Using qPCR (see Methods for details), we found that expression of myod1 and cdh15 was likewise up-regulated in tadpoles with KRAS G12D -induced ITLSs. Moreover, activation of ChR2 D156A in these tadpoles significantly lowered myod1 and cdh15 expressions, down to baseline levels (χ 2 test compared to controls, *p < 0.01). signaling in vivo highlights another opportunity for the optogenetic toolbox to extend beyond excitable cells. Moreover, our data suggest a light-based therapeutics strategy that couples gene therapy with optogenetics to counteract tumorigenesis and promote regression in vivo.

Animal husbandry
Xenopus laevis eggs were fertilized in vitro, and embryos were cultured according to standard protocols [98], in 0.1 × Modified Marc's Ringers (MMR; pH 7.8) with 0.1% Gentamicin. Xenopus embryos were housed at 14-18°C and staged according to Nieuwkoop and Faber [99]. All experimental procedures involving the use of animals for experimental purposes were approved by the Institutional Animal Care and Use Committees (IACUC) and Tufts University Department of Lab Animal Medicine (DLAM) under the protocol number M2014-79.

Microinjection
Fertilized Xenopus embryos were transferred into mesh-bottomed dishes with 3% Ficoll and injected with capped, synthetic mRNAs (made using the Ambion Message Machine kit) dissolved in water at the stages indicated. The doses per cell were KRAS G12D [71] 40pg; Arch [69], 60pg; and ChR2 D156A [72] 50pg. Two hours after injection, embryos were transferred into 0.75 × MMR for 45 minutes before they were washed and cultured in 0.1 × MMR until desired stage was reached. Injected embryos were stimulated with the appropriate wavelength of light and irradiance before or after ITLSs fully form (stages [28][29][30][31][32][33][34][35]. Embryos were scored for the presence of ITLSs using bright field microscopy as described in [3,41,42].

Light stimulation
During ITLS suppression experiments, an optogenetic set up (Figure 2) was used so that each embryo, in a microfluidic chip placed on a motorized stage, individually received a light regimen with the following parameters: Arch -580 nm wavelength, 1 mw/ mm 2 irradiance, 500 ms on, 1.50s off; Ch2R (D156A) -450 nm wavelength, 2.4 mw/mm 2 irradiance, 10 ms on, 30s off.

RNA extraction and cDNA synthesis
RNA extraction was achieved using RNeasy Mini Kit (50) (QIAGEN 74104). Tadpoles (collected n = 10 per Eppendorf tube, three biological replicates) were put into 10 volumes of RNAlater solution. RNAlater-stabilized tissues were then removed from the reagent using forceps and placed into 2 ml Eppendorf tubes with 600 ul of B-ME containing Buffer RLT for disruption and homogenization. Disruption and homogenization of tissue was achieved using MICROSONTM XL 2000 ULTRASONIC LIQUID PROCESSOR. The lysate was then centrifuged for 3 minutes at full speed and the supernatant (lysate) was put into a new microcentrifuge tube. After adding and pipet-mixing 1 volume of 70% Ethanol into the lysate, 700 ul of the resultant sample was transferred to RNeasy spin column. Following the steps outlined in the RNeasy MiniHandbook 04/2006, we completed series of treatments with buffer RW1, RPE and RNase-free water to obtain total RNA. RNA yield and quality were assessed by spectrophotometry (ND-1000, NanoDrop) and gel electrophoresis, respectively, to assess integrity of 28S and 18S RNA.
Reverse transcription was performed using ThermoScript RT-PCR System (Life Technologies). Each in vitro reverse transcription reaction was performed using 1 μg of total RNA and 50 μg of oligo(dT)20 primers (Life Technologies). RNA and primers were mixed, denatured for 5 min at 65°C, and placed on ice before adding the reaction mix according to the manufacturer's instructions. Reverse transcription reaction was carried out at 50°C for 45 min. The reaction was terminated by incubating at 85°C for 5 min, followed by RNA degradation using 1 μg of RNase H for 20 min at 37°C. The complementary DNA (cDNA) was stored at −20°C until use. The quality and quantity of cDNA were validated using Advantage 2 PCR kit (Clontech) on cDNA samples using Orinithine Decarboxylase (ODC) primers.

Quantitative real-time PCR
Primers were designed using Genius for myogenic differentiation 1 (Myod1) and myogenic differentiation 1 (Cdh15). ODC, a widely used endogenous control for Xenopus, was used to normalize target gene expression. The PCR specificity was verified by BLAST (http:// blast.ncbi.nlm.nih.gov/Blast.cgi) using the National Center for Biotechnology Information X. laevis reference sequence. Desalted primers were obtained from Invitrogen by Life Technologies as follows: MyoD1-Forward www.impactjournals.com/oncotarget CCGAGGGCAGTCCCTGTT; MydoD1-Reverse TGGGA CAGTTGAGTGCAGG; Cdh15-Forward ACAATCGT CCAGTGTTTGTGC; Cdh15-Reverse GTTCAGCATT GTCTGTCCTTGG. For each primer pair, standard curve primer analysis was performed using serial dilutions of cDNA from control embryos [1 (undiluted), 10−1, 10−2, 10−3]. Formation of primer-dimer and amplification specificity was assessed by efficiency and melt curve analysis. The cDNA from validated RNA was used to perform RT-qPCR assays. For each biological sample, three technical replicates were run in each RT-qPCR experiment. Each treatment contained five biological replicates. Triplicate negative controls lacking template were also run for each cDNA sample for each reaction. PCRs were assembled manually. Samples were prepared by adding 1 μl of cDNA (diluted 1:5 in ddH2O), 10 μl of 2 × Power SYBR Green PCR Master Mix (Applied Biosystems), and 0.5 μl of each primer (diluted to 10 μM) in a final volume of 20 μl. Reactions were incubated in 96-well MicroAmp Optical Reaction plates at 95°C for 10 min followed by 40 cycles at 95°C for 15 s and at 60°C for 1 min in a StepOnePlus qPCR instrument (Applied Biosystems). The RT-qPCR data were analyzed using the StepOne software v.2.3, and ΔΔCT values were calculated (Applied Biosystems). Fold change of target genes relative to the amount of the control gene ODC was calculated as 2^-ΔΔCT.

Statistics
Following the appropriate light regimens, stimulated construct-expressing embryos were compared -for ITLS incidence -to their expressing but un-stimulated counterparts using a χ 2 test (α = 0.01).

Pharmacological agents
The stages selected for the antineoplastic drugs mirror those of the light treatment. Embryos with ITLS between St. 28-35 were selected, treated with the drugs, and scored for ITLS between St. 45-47. Concentration of the drugs and their effects on control embryos vs embryos with ITLS are given below. Following the manufacture's instruction we prepared stock solutions in DMSO of Selumetinib, Pictilisib, and Vemurafenib in concentration of 100 mM, 50 mM, and 25 mM, respectively. Embryos were then exposed in 0.1 × MMR for the stages indicated to: Selumetinib -100 nMm, Pictillisib -1 µM, and Vemurafenib -1 µM. All three compounds were obtained from Selleckchem.