Biomimetic Chemical Sensing by Fluorescence Signals Using a Virus

Article Cite This: ACS Sens. XXXX, XXX, XXX−XXX

pubs.acs.org/acssensors

Biomimetic Chemical Sensing by Fluorescence Signals Using a Virus-like Particle-Based Platform Yuki Kushida,†,§ Yoshiyuki Arai,†,‡ Ken Shimono,§ and Takeharu Nagai*,†,‡ †

Department of Biotechnology, Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita 565-0871, Japan Department of Biomolecular Science and Engineering, The Institute of Scientific and Industrial Research, Osaka University, 8-1 Mihogaoka, Ibaraki 567-0047, Japan § Advanced Research Division, Panasonic Corporation, 3-4 Hikaridai, Seika, Soraku, Kyoto 619-0237, Japan ‡

S Supporting Information *

ABSTRACT: The chemical receptors present in living organisms are promising tools for developing biomimetic chemical sensors. However, these receptors require lipid membranes for functioning under physiological conditions, which prevents their utilization in the production of cell-free in vitro chemical sensing systems. Here, we report the development of a cell-free biomimetic sensing platform using virus-like particles (VLPs) with intact ligand-gated Ca2+ channels and genetically encoded Ca2+ indicator (GECI). We observed that targeting GECI to the plasma membrane was essential for efficient loading GECI in the VLPs. Although the physiological Ca2+ concentration [Ca2+] maintained in the cells was low (∼10 nM), the concentration in the VLPs was high. This prevented the detection of the increase in [Ca2+] caused by binding of the ligand to the receptor. To address this problem, we employed Lyn-R-CEPIA1, which had low affinity for Ca2+, and a membrane targeting sequence. Thus, we succeeded in monitoring the activation of cyclic nucleotide-gated channels (CNG) on the VLPs by measuring the increase in fluorescence of Lyn-R-CEPIA1. Our VLP-based sensing system can act as a fundamental platform for all kinds of ligand-gated channels. KEYWORDS: chemical sensor, virus-like particle, Ca2+ indicator, ligand-gated ion channel, biomimetic sensor

L

and cell-free membrane systems. Virus-like particles (VLPs) have been developed to investigate the mechanism of virus budding and infection,10−12 or to be used as a vaccine.13 VLPs can be produced by expressing the structural protein(s) of the virus in the host cells. These particles have a diameter similar to that of the original virus; however, they are noninfectious, since they lack the genes for viral replication. It has been demonstrated that the VLPs derived from retrovirus bud from the plasma membrane of the host cells14 and carry the membrane proteins expressed by the latter.15,16 In this study, we developed a platform for VLP-based sensors by combining VLPs with a genetically encoded Ca2+ indicator (GECI) and ionotropic receptors. We observed that plasma membrane targeting was essential for packing GECI into VLPs. Additionally, the Ca2+ concentration in the VLPs was unexpectedly high. Based on these results, we employed Lyn-R-CEPIA1, a membrane-targeted GECI with a high Kd value for Ca2+, as an indicator. We monitored Ca2+ influx via ligand-gated channels using the cyclic nucleotide-gated (CNG) channels on VLPs.

iving organisms can sense external and internal environmental changes in the levels of myriad chemicals with high sensitivity and selectivity using chemical receptors. In particular, the olfactory system has a variety of receptors.1,2 In fact, the potential of detecting cancer cells by imaging the neural activity of the transgenic fruit fly olfactory system has been demonstrated.3 Such studies have motivated the researchers to apply similar systems for developing biomimetic sensing devices.4−8 Olfactory receptors are transmembrane proteins, owing to which the presence of a lipid membrane is essential for their functioning under physiological conditions. Thus, in order to utilize the olfactory system for the development of biomimetic sensing devices, either a combination of the olfactory receptor with an artificial lipid membrane or the use of living cells expressing these receptors is essential.4,9 For example, Termtanasombat et al. recently developed a sensor chip with cultured cells expressing an insect odorant receptor and a genetically encoded Ca2+ indicator.4 However, the use of an artificial lipid membrane poses disadvantages, including membrane instability and difficulty handling membrane proteins because of their high hydrophobicity. Additionally, the living cells require continuous culturing, which hinders their use in practical applications, such as onsite devices. Therefore, in order to develop sensing devices using odorant receptors, it was necessary to develop stable, easy to prepare, © XXXX American Chemical Society

Received: August 3, 2017 Accepted: December 19, 2017 Published: December 19, 2017 A

DOI: 10.1021/acssensors.7b00537 ACS Sens. XXXX, XXX, XXX−XXX

Article

ACS Sensors

■

Next, the detecting solution was added to the samples and mixed. After evaluating the protein concentration, sample volumes containing 0.1 μg protein were mixed with the NuPAGE LDS sample buffer and dithiothreitol (NuPAGE reducing agent; Thermo Fischer Scientific Inc., Waltham, MA, USA), and boiled at 70 °C for 10 min. The samples were loaded on a 5−12% gradient gel (ATTO Corp. Tokyo, Japan), along with a molecular size marker, Precision Plus Dual Color Standards (Bio-Rad Laboratories Inc., Hercules, CA, USA), and subjected to electrophoresis. The gels were stained using Coomassie Brilliant Blue (CBB), EzStain AQUA (ATTO Corp. Tokyo, Japan). Measurement of Fluorescence. The protein samples were diluted to 10 μg/mL and transferred to a 0.35 mL quartz glass cuvette. The fluorescence spectra were recorded using the FP-8500 fluorometer (JASCO Corp., Tokyo, Japan) with 440 ± 5 nm excitation light at 1 nm intervals. Microscopic Observation of the VLPs. The YC3.6-CaaX VLPs were prepared by co-transfecting Gag/pCIneo (11 μg) and YC3.6CaaX/pcDNA3 (3.5 μg) into the HEK293FT cells; the VLPs were collected as follows: The VLP suspension was diluted ∼5−10 times with PBS and transferred onto poly-L-lysine coated glass bottom dishes. The suspension was allowed to settle for more than 30 min at 4 °C in the dark. Next, it was washed twice with 20 mM HEPES buffered Hank’s balanced salt solution without Ca2+ and Mg2+ [H−HBSS (−), pH 7.4; Wako Pure Chemical Industries, Ltd., Osaka, Japan]. The VLPs were observed in H−HBSS (−) using the FV-1000 laser scanning confocal microscope (Olympus Corp., Tokyo, Japan), equipped with a 60× oil immersion lens (numerical aperture, NA = 1.35) and the autofocus system. Samples were irradiated using laser excitation at 458 nm (magnification, 5×; confocal aperture size, 800 nm). The fluorescence intensity of each particle was measured at 475−510 nm (FCFP) and 515−615 nm (FFRET), and the signal ratio (FFRET/FCFP) was calculated as FRET index. For size comparison, images of the fluorescent polystyrene beads, FluoSpheres Carboxylate-Modified Microspheres 0.1 μm, yellow−green fluorescent (Thermo Fischer Scientific Inc., Waltham, MA, USA) were obtained in the same manner. The VLPs with intensiometric GECI (Lyn-R-CEPIA1 or Lyn-OGECO1) were prepared by co-transfection with Gag/pCIneo (11 μg) and the plasmids encoding each GECI (4 μg). The CNG containing VLPs (CNG−VLP) were prepared by co-transfection with Gag/ pCIneo (8 μg), Lyn-R-CEPIA1/pcDNA3 (4 μg), and CNGA2/pcDNA3 (4 μg). The VLPs were collected and transferred on the glass bottom dishes, as described above. Next, they were observed using the IX81 inverted fluorescence microscope (Olympus Corp., Tokyo, Japan) equipped with a 60× oil immersion lens (NA 1.35), xenon lamp (U-RX-T), an electron multiplying charge-coupled device (EMCCD), ImagEM (Hamamatsu Photonics K.K., Shizuoka, Japan), and a filter cube for red fluorescent proteins, U-MRFPHQ (Olympus Corp., Tokyo, Japan). Imaging was performed with an exposure time of 200 ms, EM gain of 200, and was repeated after 10 s intervals.

MATERIALS AND METHODS

Plasmid Construction. cDNA of the human immunodeficiency virus type 1 (HIV-1) gag gene with mutations for Rev-independent expression17 was obtained from GeneArt Strings (Thermo Fischer Scientific Inc., Waltham, MA, USA), and subcloned into pCI-neo (Promega Corp., Madison, WI, USA) using EcoRI and SalI. The R-CEPIA1 cDNA was synthesized based on the sequence of the R-CEPIA1er cDNA, but without the ER-targeting signal,18 and subcloned into pcDNA3 using BamHI and EcoRI using the In-Fusion cloning technique (Takara Bio USA Inc., Mountain View, CA, USA). YC3.6/ pcDNA3 and YC3.6-CaaX/pcDNA3 (coding for plasma membranetargeted YC3.6) were synthesized as described previously.19 The gene of YC3.6-CaaX contained DNA encoding CaaX motif of Ki-Ras, which was fused to the 3′-terminus of the gene of YC3.6 through a linker sequence (GTGGSGGSGT). To construct O-GECO1/pcDNA3, O-GECO1 in the modified pcDNA3 (ref 20) was digested with BamHI and EcoRI and the O-GECO1 fragment was purified using gel electrophoresis. The purified product was subcloned into the BamHI− EcoRI digested site of pcDNA3. Lyn-R-CEPIA1/pcDNA3 and Lyn-OGECO1/pcDNA3 were constructed by subcloning the annealed oligonucleotides, (Lyn_Fw and Lyn_Rv, see Supporting Information), encoding the N-terminal sequence of Lyn kinase (MGCIKSKGKDSA), into the KpnI and BamHI sites of R-CEPIA1/pcDNA3 and O-GECO1/ pcDNA3, respectively. R-CEPIA1−CaaX/pcDNA3 was constructed by ligating fragments of R-CEPIA1 and CaaX prepared by PCR into the BamHI and EcoRI sites of pcDNA3. The former was prepared using pcDNA3_Fw and R-CEPIA1_Rv primers with R-CEPIA1/ pcDNA3 as a template, and digested with BamHI. The latter was prepared using YC_SmaI_CaaX_Fw and pcDNA3_Rv primers with YC3.6-CaaX/pcDNA3 as a template, and the product was digested with SmaI and EcoRI. The mouse CNGA2 cDNA (Genbank ID, NM_007724) was amplified from the mouse brain cDNA library (GenoStaff Co., Ltd., Tokyo, Japan) using primers CNGA2_cloning_Fw and CNGA2_cloning_Rv. The sequences for XhoI, NotI, and Kozak (ccacc) were added to the product using PCR with primers XhoI_CNGA2_Fw and CNGA2_NotI_Rv. The product was subcloned into pcDNA3. Preparation of VLPs. The HEK293FT cells (Thermo Fischer Scientific Inc., Waltham, MA, USA) were cultured in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich Co. LLC., St. Louis, MO, USA) supplemented with 10% fetal calf serum, nonessential amino acid solution (Thermo Fischer Scientific Inc., Waltham, MA, USA), and penicillin−streptomycin mixture (Thermo Fischer Scientific Inc., Waltham, MA, USA) at 37 °C in 5% CO2 atmosphere. One day before transfection, 4.6 × 106 cells were subcultured in a T-75 flask (Corning Inc., Corning, NY, USA). The next day, Gag/pCI-neo (12 μg) was transfected into the cells using Lipofectamine 2000 (Thermo Fischer Scientific Inc., Waltham, MA, USA), according to the manufacturer’s instructions. After 4−6 h, the medium was replaced with fresh medium and the culture was incubated for 2 days. Next, the culture was centrifuged at 2000 × g for 10 min and the supernatant was collected. The supernatant was filtrated using a 0.45 μm filter to remove any cell debris. For VLP collection, the medium was centrifuged at 110 000 × g for 1 h at 4 °C using the CS100FNX centrifuge with swing rotor (Hitachi Koki Co., Ltd., Tokyo Japan). The pellet was washed with phosphate-buffered saline (PBS, 1 mL). Next, it was resuspended in PBS (100 μL) and stored at 4 °C. Dynamic Light Scattering (DLS). The size distribution of VLPs was measured using DLS, ELSZ-1000 (Otsuka Electronics Co. Ltd., Osaka, Japan). The measurements were performed at 25 °C, at a scattering angle of 165°, and signals were accumulated 70 times per measurement. The particle size distribution was calculated according to the manufacturer’s instructions (Marquardt method). SDS-PAGE. Total protein of the VLP suspension was evaluated using the Bicinchoninic Acid (BCA) Protein Assay kit (Takara Bio Inc., Shiga, Japan). The VLP suspension and bovine serum albumin (BSA, used as a standard) were mixed with equal volumes of the lysis buffer [NaCl (137 mM), sodium deoxycholate (2%), Triton-X100 (2%), SDS (1%), EDTA (2 mM), and Tris (20 mM, pH 7.6)].

■

RESULTS AND DISCUSSION Concept of a VLP-Based Sensor. We designed a VLPbased sensor system, illustrated in Figure 1. Among various VLPs, we selected the HIV-1-derived VLPs, since it has been reported that the retroviral VLPs carry membrane proteins of the host cells.15,16 Recently, they have been used for extracting proteins from the mammalian host cells.21 The HIV-1 VLPs can be produced by expressing Gag in mammalian cells. Gag is a precursor for the formation of structural proteins of viruses, and upon mutation, it enables high-level production of VLPs without the need of any other viral protein.17 First, in order to validate the production of HIV-1 Gag VLPs, we collected the culture medium of the gag-transfected HEK293FT cells and purified the VLPs using centrifugation. We measured the size distribution of VLPs by DLS and confirmed the existence of VLP particles in the suspension (mean radius, 151 ± 1 nm; n = 3). The size of VLPs in our study corroborated with the size of the reported HIV-1 Gag virions B

DOI: 10.1021/acssensors.7b00537 ACS Sens. XXXX, XXX, XXX−XXX

Article

ACS Sensors

First, the VLPs were lysed using the lysis buffer and the total protein content was measured using the BCA method. After suitably diluting the samples, their fluorescence was measured using a fluorometer at 440 nm. YC3.6-CaaX VLPs showed an emission spectrum with two distinct fluorescence peaks, which was consistent with that of purified YC3.6 (ref 19). On the other hand, these peaks were not observed in the YC3.6 VLPs under similar conditions (Figure 2D). Therefore, we concluded that targeting the Ca2+ indicator to the plasma membrane enabled it to be efficiently packaged into the VLPs. Next, we elucidated whether Ca2+ concentration change in VLP was detectable by YC3.6-CaaX. We observed YC3.6-CaaX VLPs with a confocal microscope. The apparent size of them was larger (∼500 nm) than that DLS indicated, but it should be due to the limitation of the optics, as 100 nm beads appeared with similar size under the same condition (Figure 3A) and some aggregations were observed as well. Moreover, there are some variation in the brightness and Ca2+ concentration in VLP reflecting in the difference of fluorescence between the CFP and FRET channel. The HIV-1 VLPs possess a membrane derived from the plasma membrane of the host cells. Thus, we hypothesized that ionomycin, a Ca2+ carrier, could be used to forcibly change the concentration of Ca2+ in the VLPs and make it similar to that of the external solution. We sequentially added EGTA (2 mM) and CaCl2 (10 mM) to the samples. EGTA, when added along with ionomycin, induced a decrease in the FRET index (which reduced to 70% of its original value) and showed a decrease in the Ca2+ concentration of VLPs induced by EGTA chelation. Subsequently, upon addition of Ca2+ (10 mM) along with ionomycin, the value of FRET index increased; however, it did not exceed the basal value (Figure 3C). This was a surprising result since we observed that, initially, the Ca2+ concentration in VLPs was so high that it resulted in the saturation of YC3.6-CaaX. It suggested the concentration of Ca2+ in VLPs was higher than its normal concentration in cells. On the other hand, these changes were

Figure 1. Schematic representation of a VLP-based chemical sensor.

observed using electron microscopy.22 Next, we explored whether Ca2+ indicators could be packaged into the VLPs. YC3.6 is a Förster resonance energy transfer (FRET)-based genetically encoded Ca2+ indicator. It consists of two fluorescent proteins, cyan fluorescent protein (CFP) and circularly permutated Venus (Figure 2B), as the donor and acceptor, respectively.19 At first, we investigated which (cytosolic or plasma membrane-targeted) GECI was more preferable, since the HIV-1 VLPs contained both, the host cytosolic as well as membrane proteins. We synthesized the VLPs containing either YC3.6 or YC3.6-CaaX (YC3.6, plasma membrane targeted), by co-transfecting them in HEK293FT cells along with gag. We compared the SDS-PAGE profiles of the YC3.6 VLPs, YC3.6-CaaX VLPs, and the VLPs without GECI. We observed bands of similar density and molecular weight (56 kDa) in all the samples, which corresponded to the molecular weight of the HIV-1 Gag protein (Figure 2C). Combining these results with those of DLS, we concluded that the pellet mainly consisted of the Gag VLPs. Next, we estimated the amount of indicator present in the VLPs by measuring the fluorescence.

Figure 2. (A) Hydrodynamic radius measurement of the VLPs using DLS. Results of three measurements are shown. (B) DNA constructs with Ca2+ indicators. (C) SDS-PAGE profile of the VLPs. M, molecular size marker; 1, YC3.6-CaaX VLPs; 2, YC3.6 VLPs; 3, VLPs without GECI. Asterisk indicates the band corresponding to Gag protein (56 kDa). (D) Fluorescence spectra of the YC3.6 and YC3.6-CaaX VLPs. VLPs (0.1 μg) were suspended in PBS and excited at 440 ± 5 nm. The data represent mean ± SEM (n = 3). C

DOI: 10.1021/acssensors.7b00537 ACS Sens. XXXX, XXX, XXX−XXX

Article

ACS Sensors

Figure 3. Response of YC3.6-CaaX in VLPs. (A) Images of the YC3.6-CaaX VLPs and the fluorescent PS beads (0.1 μm) acquired using a confocal microscope. Imaging was performed using 405 nm laser excitation, which detected CFP at 475−510 nm, FRET at 515−615 nm, and beads at 515−615 nm. Scale bars indicates 4 μm. (B) The response of YC3.6-CaaX in VLP upon changing the external concentration of Ca2+. Error bars represent standard error of the mean (SEM). 116 particles were analyzed from a twice repeated experiment. (C) The response of YC3.6-CaaX in VLP observed by forcefully changing the internal Ca2+ concentration. Error bars represent standard error of the mean (SEM). 200 particles were analyzed from 3-times repeated experiment. (Right) Trail of fluorescence change observed with CFP and FRET channels in a single representative particle.

hardly observed without the addition of ionomycin (Figure 3B), with few exceptions which might be leaky particles with deficient membrane, which was also included in analysis for Figure 3B. Thus, we validated that YC3.6-CaaX reflected the Ca2+ level changes in VLPs. In order to observe Ca2+ influx under such conditions, we used another Ca2+ indicator with a higher Kd value. R-CEPIA1 has been recently reported as a Ca2+ indicator, used for monitoring the Ca2+ concentration in organelles with high Ca2+ levels, such as the ER (Kd = 455 μM).18 We hypothesized that the high Kd value of R-CEPIA1 would also enable us to monitor the changes in the Ca2+ concentration of the VLPs. We constructed membrane-targeted R-CEPIA1 by fusing it with the Lyn sequence following the strategy employed to make R-GECO1.2 localize on the plasma membrane, which was developed from R-GECO1 in a similar manner to R-CEPIA1 (ref 20). We prepared VLPs containing R-CEPIA1 or Lyn-RCEPIA1. Without the Lyn sequence, no fluorescent particles could be observed under the microscope, even after the addition of ionomycin and CaCl2 (10 mM; data not shown). This result also suggested the indicators needed to localize the plasma membrane as we concluded above with YC3.6. We observed the VLPs containing Lyn-R-CEPIA1 and sequentially added Ca2+ and ionomycin to the solution. We observed a graded increase in the averaged fluorescence upon addition of Ca2+ and ionomycin (Figures 4A,B). This stepwise increase seemed to be caused by including at least 2 kinds of particles in analysis; one had a membrane that was Ca2+ permeable or deficient, the others had an intact membrane so that they were responsive only when Ca2+ added along with ionomycin. Indeed, some responsive particles just by changing the extraparticular Ca2+ concentration did not show any increase of fluorescence intensity, which suggest that those apparent fluorescent particles had a permeable membrane comparable to the ionomycin effect (Figure S1A). Furthermore, such particles did

Figure 4. (A) Representative images of VLPs with Lyn-R-CEPIA1 in the basal solution (t = 0 s), after Ca2+ addition (t = 70 s), and after ionomycin addition (t = 190 s). Scale bar indicates 5 μm. (B) Changes in the fluorescence of VLPs with intensiometric GECI. CaCl2 (2 mM) and ionomycin (5 μM) were added at the time points indicated by the arrows. (Red line indicates Lyn-R-CEPIA1, black line indicates Lyn-OGECO1). Error bars represent SEM [2212 particles (for Lyn-R-CEPIA1), and 267 particles (for Lyn-O-GECO1), each from twice repeated experiments].

not show any decrease of fluorescence intensity by reduction of extraparticular Ca2+ (Figure S1B). It suggested that Lyn-RCEPIA1 reached saturation rather a stable state of Ca2+ flux. There were also some particles showing a stepwise increase of the intensity like the averaged trail; however, with our optics, we could not reveal whether that was the response of a single particle or a mixture of several particles. In contrast, the D

DOI: 10.1021/acssensors.7b00537 ACS Sens. XXXX, XXX, XXX−XXX

Article

ACS Sensors

upon 8-Br-cGMP addition, although a similar fluorescence increase was invoked in the CNG VLPs by Ca2+ (Figure 5C). These results showed that the increase in fluorescence of the VLPs reflected an increase in the Ca2+ concentration, which was mediated via CNG present on the VLP membrane. To estimate the sensitivity of the CNG VLPs, we gradually increased 8-BrcGMP. The number of brightly fluorescent VLPs increased as concentration of 8-Br-cGMP got higher, and 10 μM and more of 8-Br-cGMP induced significant fluorescence increase (Movie S2, Figure S2). In conclusion, we developed a VLP-based chemical sensor, with a Ca2+ indicator and a cation-selective ion channel, by using membrane targeted GECI with high Kd. We monitored the Ca2+ influx via CNGA2, an ionotropic receptor, similar to the insect odorant receptors. In addition, the CNG channels transduce the binding of ligands to the odorant receptors in the vertebrate olfactory system.23 Therefore, we have proposed a versatile system by developing a biomimetic sensing device with odorant receptors. We should point out that there are still issues to tackle in order to improve this system. As shown in Figure 5A, the fraction of responsive VLPs was relatively small. Moreover, some particles seemed to have Ca2+ permeable membrane, or initially saturated without extraparticular Ca2+, and such particles tended to show no change to 8-Br-cGMP. For the ideal sensor, we have to attempt to make homogeneously responsive VLPs. Applying stable cells of the receptor to be utilized for this system would be helpful. For further investigation, we elucidated R-CEPIA1-CaaX as an indicator for a VLP-based sensor. As a result, R-CEPIA1CaaX also worked in VLPs, even seemed superior (Figure S3). Considering the difference of lipid domain localization by Lyn and CaaX motives, researchers should choose the more suitable indicator according to receptor to be employed. We could not understand the reason behind high Ca2+ concentration in the cytosol, which resulted in the saturation of a GECI, such as YC3.6. Gag expression possibly affects Ca2+ regulation in cells. The strong relationship between HIV-1 replication and cytosolic Ca2+ has been demonstrated.25 In addition, the matrix proteins of some viruses have been shown to increase the cytosolic Ca2+ level in the HEK293T cells, by affecting the endogenous Ca2+ channels present on the plasma membrane.26 The immature matrix protein of HIV-1, Gag, may regulate the level of Ca2+ during viral replication using a similar mechanism.

increase in fluorescence was not observed in the case of Lyn-OGECO1 (Kd = 1.5 μM)20 VLPs upon ionomycin addition (Figure 4B). Therefore, we concluded that Lyn-R-CEPIA1 could be a useful indicator for our VLP-based sensor because of its high Kd value. Monitoring Activation of the Ligand-Gated Channels on VLPs. In order to explore whether it was possible to observe the Ca2+ influx via ion channel activation using an agonist, we investigated the functioning of the cyclic nucleotide-gated ion channels (CNG) in the VLPs as a model system. The CNG channels are ligand-gated nonselective cation channels, as with the insect odorant receptors.23 They allow Ca2+ to permeate across them by binding to cyclic nucleotides (such as cAMP and cGMP) present inside the cells.24 We prepared the CNG VLPs by co-transfecting the mouse CNGA2 gene with Gag and Lyn-R-CEPIA1 into the HEK293FT cells. We observed the CNG VLPs under a microscope and sequentially added Ca2+ and 8-Br-cGMP (membrane permeable cGMP, 300 μM). We found that 8-BrcGMP induced a drastic increase in the fluorescence of some particles (Figure 5A, Movie S1). By averaging all the particles in

Figure 5. Effect of CNG activation on VLPs. (A) Representative images of VLPs with Lyn-R-CEPIA1 in the basal solution (t = 0 s), after Ca2+ addition (t = 70 s), and after 8-Br-cGMP addition (t = 190 s). Scale bar indicates 5 μm. Representative VLPs showing FI increase upon the addition of 8-Br-cGMP are indicated using arrows. (B) Mean fluorescence change of VLPs with CNG, after addition of CaCl2 (2 mM) and 8-Br-cGMP (300 μM). Each solution was added at the time points indicated using arrows. The data represent mean ± SEM (n = 1215) (C) FI change of Lyn-RCEPIA1 in CNG VLP and VLP without CNG (Mock VLP) upon addition of Ca2+ (white), and 8-Br-cGMP or buffer addition (gray). Addition of 8-Br-cGMP to CNG VLP showed significant increase of FI (1215 particles) compared to the controls (adding buffer to CNG VLP, 1423 particles, and adding 8-Br-cGMP to Mock VLP, 1753 particles) (p < 0.01, Student’s t-test). The experiment repeated 3 times. The data represent mean ± SEM.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssensors.7b00537. Primer lists, Analysis of Ca 2+ responsive VLPs, Comparison of R-CEPIA1-CaaX with Lyn-R-CEPIA1 (PDF) Response of CNG-VLP (AVI) Response of CNG-VLP to gradual increase of agonist (AVI)

■

image, the CNG VLPs showed a graded increase in fluorescence upon stimulation with an agonist. Ca2+ and 8-BrcGMP induced 0.41 ± 0.03- and 0.62 ± 0.04-fold increase in fluorescence, respectively (Figure 5B,C). The fraction of 8-BrcGMP-responsive VLPs was 17.1 ± 8.7% (mean ± SD among 3 experiments). On the other hand, when we added the buffer solution without 8-Br-cGMP, the increase in fluorescence was not observed (Figure 5C). Moreover, when the VLPs did not contain CNG, they did not display any increase in fluorescence

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Yuki Kushida: 0000-0001-6073-2893 Yoshiyuki Arai: 0000-0003-2862-0829 Notes

The authors declare no competing financial interest. E

DOI: 10.1021/acssensors.7b00537 ACS Sens. XXXX, XXX, XXX−XXX

Article

ACS Sensors

■ ■

(19) Nagai, T.; Yamada, S.; Tominaga, T.; Ichikawa, M.; Miyawaki, A. Expanded dynamic range of fluorescent indicators for Ca(2+) by circularly permuted yellow fluorescent proteins. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 10554−10559. (20) Wu, J.; Liu, L.; Matsuda, T.; Zhao, Y.; Rebane, A.; Drobizhev, M. Improved orange and red Ca 2 + indicators and photophysical considerations for optogenetic applications. ACS Chem. Neurosci. 2013, 4, 963−972. (21) Eyckerman, S.; Titeca, K.; Van Quickelberghe, E.; Cloots, E.; Verhee, A.; Samyn, N.; De Ceuninck, L.; Timmerman, E.; De Sutter, D.; Lievens, S.; Van Calenbergh, S.; Gevaert, K.; Tavernier, J. Trapping mammalian protein complexes in viral particles. Nat. Commun. 2016, 7, 11416. (22) Briggs, J. A. G.; Simon, M. N.; Gross, I.; Kräusslich, H.-G.; Fuller, S. D.; Vogt, V. M.; Johnson, M. C. The stoichiometry of Gag protein in HIV-1. Nat. Struct. Mol. Biol. 2004, 11, 672−675. (23) Kaupp, U. B.; Seifert, R. Cyclic nucleotide-gated ion channels. Physiol. Rev. 2002, 82, 769−824. (24) Dzeja, C.; Hagen, V.; Kaupp, U. B.; Frings, S. Ca2+ permeation in cyclic nucleotide-gated channels. EMBO J. 1999, 18, 131−144. (25) Papp, B.; Byrn, R. A. Stimulation of HIV expression by intracellular calcium pump inhibition. J. Biol. Chem. 1995, 270, 10278− 10283. (26) Han, Z.; Madara, J. J.; Herbert, A.; Prugar, L. I.; Ruthel, G.; Lu, J.; Liu, Y.; Liu, W.; Liu, X.; Wrobel, J. E.; Reitz, A. B.; Dye, J. M.; Harty, R. N.; Freedman, B. D. Calcium Regulation of Hemorrhagic Fever Virus Budding: Mechanistic Implications for Host-Oriented Therapeutic Intervention. PLoS Pathog. 2015, 11, e1005220.

ACKNOWLEDGMENTS We thank Shuhei J. Yamazaki for providing technical support. REFERENCES

(1) Bushdid, C.; Magnasco, M. O.; Vosshall, L. B.; Keller, A. Humans Can Discriminate More than 1 Trillion Olfactory Stimuli. Science (Washington, DC, U. S.) 2014, 343, 1370−1372. (2) Hallem, E. A.; Carlson, J. R. Coding of Odors by a Receptor Repertoire. Cell 2006, 125, 143−160. (3) Strauch, M.; Lüdke, A.; Münch, D.; Laudes, T.; Galizia, C. G.; Martinelli, E.; Lavra, L.; Paolesse, R.; Ulivieri, A.; Catini, A.; Capuano, R.; Di Natale, C. More than apples and oranges - Detecting cancer with a fruit fly’s antenna. Sci. Rep. 2015, 4, 1−9. (4) Termtanasombat, M.; Mitsuno, H.; Misawa, N.; Yamahira, S.; Sakurai, T.; Yamaguchi, S.; Nagamune, T.; Kanzaki, R. Cell-Based Odorant Sensor Array for Odor Discrimination Based on Insect Odorant Receptors. J. Chem. Ecol. 2016, 42, 716−724. (5) Fukutani, Y.; Hori, A.; Tsukada, S.; Sato, R.; Ishii, J.; Kondo, A.; Matsunami, H.; Yohda, M. Improving the odorant sensitivity of olfactory receptor-expressing yeast with accessory proteins. Anal. Biochem. 2015, 471, 1−8. (6) Glatz, R.; Bailey-Hill, K. Mimicking nature’s noses: From receptor deorphaning to olfactory biosensing. Prog. Neurobiol. 2011, 93, 270−296. (7) Goldsmith, B. R.; Mitala, J. J.; Josue, J.; Castro, A.; Lerner, M. B.; Bayburt, T. H.; Khamis, S. M.; Jones, R. a; Brand, J. G.; Sligar, S. G.; Luetje, C. W.; Gelperin, A.; Rhodes, P. a; Discher, B. M.; Johnson, a T. C. Biomimetic Chemical Sensors using Nanoelectronic Read out of Olfactory Receptor Proteins. ACS Nano 2011, 5, 5408−5416. (8) Hamada, S.; Tabuchi, M.; Toyota, T.; Sakurai, T.; Hosoi, T.; Nomoto, T.; Nakatani, K.; Fujinami, M.; Kanzaki, R. Giant vesicles functionally expressing membrane receptors for an insect pheromone. Chem. Commun. 2014, 50, 2958−2961. (9) Funakoshi, K.; Suzuki, H.; Takeuchi, S. Lipid Bilayer Formation by Contacting Monolayers in a Microfluidic Device for Membrane Protein. Anal. Chem. 2006, 78, 8169−8174. (10) Balliet, J. W.; Bates, P. Efficient infection mediated by viral receptors incorporated into retroviral particles. J. Virol. 1998, 72, 671− 676. (11) Bavari, S.; Bosio, C. M.; Wiegand, E.; Ruthel, G.; Will, A. B.; Geisbert, T. W.; Hevey, M.; Schmaljohn, C.; Schmaljohn, A.; Aman, M. J. Lipid Raft Microdomains: A Gateway for Compartmentalized Trafficking of Ebola and Marburg Viruses. J. Exp. Med. 2002, 195, 593−602. (12) Lampe, M.; Briggs, J. A. G.; Endress, T.; Glass, B.; Riegelsberger, S.; Kräusslich, H.-G.; Lamb, D. C.; Bräuchle, C.; Müller, B. Doublelabelled HIV-1 particles for study of virus−cell interaction. Virology 2007, 360, 92−104. (13) Lua, L. H. L.; Connors, N. K.; Sainsbury, F.; Chuan, Y. P.; Wibowo, N.; Middelberg, A. P. J. Bioengineering virus-like particles as vaccines. Biotechnol. Bioeng. 2014, 111, 425−440. (14) Ivanchenko, S.; Godinez, W. J.; Lampe, M.; Kräusslich, H.-G.; Eils, R.; Rohr, K.; Bräuchle, C.; Müller, B.; Lamb, D. C. Dynamics of HIV-1 Assembly and Release. PLoS Pathog. 2009, 5, e1000652. (15) Zemanová, L.; Schenk, A.; Hunt, N.; Nienhaus, G. U.; Heilker, R. Endothelin receptor in virus-like particles: Ligand binding observed by fluorescence fluctuation spectroscopy. Biochemistry 2004, 43, 9021−9028. (16) Willis, S.; Davidoff, C.; Schilling, J.; Wanless, A.; Doranz, B. J.; Rucker, J. Virus-like particles as quantitative probes of membrane protein interactions. Biochemistry 2008, 47, 6988−6990. (17) Schneider, R.; Campbell, M.; Nasioulas, G.; Felber, B. K.; Pavlakis, G. N. Inactivation of the human immunodeficiency virus type 1 inhibitory elements allows Rev-independent expression of Gag and Gag/protease and particle formation. J. Virol. 1997, 71, 4892−4903. (18) Suzuki, J.; Kanemaru, K.; Ishii, K.; Ohkura, M.; Okubo, Y.; Iino, M. Imaging intraorganellar Ca2+ at subcellular resolution using CEPIA. Nat. Commun. 2014, 5, 4153. F

DOI: 10.1021/acssensors.7b00537 ACS Sens. XXXX, XXX, XXX−XXX