Stable expression of bacterial transporter ArsB attached to SNARE molecule enhances arsenic accumulation in Arabidopsis

ABSTRACT

Acute and chronic arsenic (As) toxicity is a global health issue affecting millions of people, which leads to inactivation of over 200 enzymes, particularly those involved in cellular energy pathways and DNA synthesis and repair. The fern Pteris vittata acts as a hyperaccumulator of As and may be useful for phytoremediation to reduce disposal risks by utilizing metal-enriched plant biomass in energy and metal recovery. However, these ferns grow in limited environments and its transplantation and transport can be challenging. Therefore, we generated a transgenic Arabidopsis plant as a seed plant model, capable of accumulating As in their vacuole lumen. This was achieved by transforming the As-resistant bacterial As transporter, ArsB, via fusion with a organelle-targeting signal to the vacuolar membrane, N-ethyl-maleimide-sensitive factor attachment protein receptors (SNAREs) protein, VAMP711. In this study, we developed the iVenus assay as a method for detecting whether the N- or C-terminus of a membrane protein is located on the cytoplasmic or exoplasmic side, and from the result of the iVenus assay, we generated the transgenic plant introduced N-terminal end of ArsB with VAMP711, localized to the central vacuolar membrane to accumulate As in the shoot and differentiation zone of root.

KEYWORDS:

• Phytoremediation
• arsenic (As)
• SNARE

View correction statement:

Correction

Introduction

Highly toxic arsenic (As) has two oxidation states in nature: arsenate, As(V), and the more toxic arsenite [As(III)]. Since As(V) is structurally analogous to inorganic phosphate, it readily enters the cell via the phosphate membrane transport system, thereby impairing metabolic reactions¹. Alternatively, As(III) enters the cell via aquaglyceroporins and membrane channels that transport water and small neutral molecules,^2,3 and exerts its toxicity by binding to thiol groups on various proteins.^4,5 Across the world, As contaminates soil and groundwater, entering the food chain through contaminated drinking water and vegetables.⁶ Thus, As contamination is a major health hazard, and reducing its accumulation in groundwater and cultivation soils is critical for human health.

Since As is highly toxic to plants, various plants have developed metabolic systems for its detoxification. After As enters the plant root cells via aquaglyceroporins or phosphate transporters in the form of As(III) or As(V), respectively, the latter is reduced to As(III), which is then detoxified through various pathways. In angiosperms, As becomes detoxified in the cell primarily by being sequestered in the vacuole in the form of phytochelatins (PCs)-As conjugates. AtABCC1 and AtABCC2, members of the ABC transporter superfamily, localize to the vacuolar membrane and transport PCs-As to protect cellular components from toxic As.⁷ Meanwhile, ferns employ an alternate mechanisms for sequestering As into the vacuole. For instance, in the Chinese brake fern (Pteris vittata), identified as the first As-hyperaccumulating plant, As(III) is translocated to the fronds without synthesizing PC-As conjugates, and becomes stored in vacuoles via the vacuole membrane protein PvACR3 (Arsenic Compounds Resistance 3).⁸

One strategy for phytoremediation is to remove hazardous materials, such as heavy metals or metalloids, from polluted soil and water by using plants that hyperaccumulate heavy metals and eliminate As, such as the fern, P. vittata.⁹ However, this fern exhibits an extremely restricted geographic and ecological distribution to warm, damp areas, such as tropical rainforests, while avoiding dry, cold areas. Hence, alternative phytoremediation technologies are needed to eliminate As from contaminated soils in various environments. Previous studies reported that transgenic Arabidopsis plants expressing As(III)-exporter ACR3 from yeast (ACR3), or P. vittata (PvACR3), were capable of tolerating As, however, did not cause its hyperaccumulation.^10,11 Meanwhile, transgenic plants expressing PvACR3;1 increased accumulation of As by 19% in the root, but not in the shoot.^12,13 However, generation of transgenic plants for phytoremediation requires advancement to ensure that As is readily accumulated within the shoots.

Certain bacteria possess a plasmid-determined bacterial resistance system for As and antimony. Specifically, in Escherichia coli¹⁴ this system consisting of five genes (arsR, arsD, arsA, arsB, and arsC),^15–19 and three genes in Staphylococcus species (arsR, arsB, and arsC).^15,20 The well-characterized ArsB functions alone as an As(III) efflux pump using energy supplied by the membrane potential of the cell^16,17 in bacteria, such as E. coli and Staphylococcus species. Additionally, ArsB associated with the ATPase ArsA, pumps out As(III) in an ATP-dependent process.¹⁸

We have previously demonstrated that the Arabidopsis N-ethyl-maleimide-sensitive factor attachment protein receptors (SNAREs), AtVAM3 (SYP22), and SYP121, can be used as organelle-targeting signals to direct the bacterial heavy metal transporter MerC to the vacuolar membrane or plasma membrane in both yeast and Arabidopsis. Moreover, the plants exhibited cadmium and mercury tolerance, as well as an increased accumulation capacity.¹⁹

In the current study we, therefore, aimed to generate a plant with a new vacuolar membrane transporter, other than the ABCC1/2 transporter, to sequester free-As(III) in seed plants. Thus, we fused the vacuolar membrane R-SNARE protein, VAMP711, or GFP to the N-terminal end of ArsB to develop Arabidopsis plants expressing VAMP711-ArsB- GFP (VBG). We observed VBG localization to the central vacuoles and some accumulation of As, without As tolerance in VBG-expressing plants.

Results

ArsB lacking an organelle-targeting sequence localizes to SNX1-endosomes

Although the bacterial arsenic transporter ArsB has been shown to localize to the plasma membrane in bacteria and transport As(III) out of the cell, the specific localization of ArsB in ArsB-expressing plant cells is unclear.²¹ Therefore, to determine the subcellular localization of GFP-ArsB in plant cells, we transiently expressed GFP-ArsB under control of the Cauliflower mosaic virus 35S promoter in various mRFP-tagged organelle marker lines. We found that GFP-ArsB predominantly co-localized with the late endosome marker SNX1-mRFP, and not with the Golgi marker ST-mRFP, trans-Golgi marker mRFP-VHAa1, or vacuolar membrane marker mRFP-AtVam3, indicating that transiently expressed GFP-ArsB was localized in late endosomes in Arabidopsis cells (Figure 1a).

Figure 1. ArsB localizes to the endosome, and N- and C-terminal ends of ArsB are exoplasmic

(a) GFP-ArsB was transiently expressed in five-day-old Arabidopsis seedlings of the marker lines, ST-mRFP, mRFP-VHAa1, SNX1-mRFP, and mRFP-AtVAM3; the root elongation zone was observed by confocal microscopy. (b) For the iVenus assay, ciV-ArsB or ArsB-CiV were transiently expressed with RHM1-niV or AALP-niV in the leaves of Nicotiana benthamiana, and then observed by confocal microscopy. Bars = 10 μm. The topology of endosomal ArsB was determined by estimation.

The N- and C-terminal ends of ArsB are oriented toward the vacuole lumen

To accumulate As in the vacuolar lumen, we next aimed to localize ArsB to the vacuolar membrane by applying organelle-targeting technology as reported in our previous study.¹⁹ Localization to a targeted organelle requires the attachment of a SNARE-localization tag to the cytoplasmic side of the N- or C-terminus of the protein. However, the orientation of the N- and C-terminal ends of ArsB protein was unknown, although both ends should have the same orientation in the cytoplasm and exoplasm since ArsB is a lipophilic protein with 12 transmembrane helices that are crucial for the transport of As(III) or antimonite across the PM, according to the previous report²² and our topology predictions in silico. Therefore, we developed a new assay using the split-Venus, designated auto-interactive Venus (iVenus assay), based on an earlier report describing self-assembly between the N- and C-terminal fragments of Venus,²³ which can be used to examine the orientation of both N- or C-terminal ends of a transmembrane protein. First, we prepared plasmids of the iVenus marker set; the C-terminals of cytosolic protein rhamnose biosynthesis 1 (RHM1) or the vacuole lumen protein aleurain-like protease (AALP) were fused with the N-terminal fragment (M1–I189) of Venus to express RHM1-niV or AALP-niV, which ensured its localization to the cytoplasm, or the vacuole lumen in N. benthamiana (Fig. S1). Next, we attached the C-terminal fragment (E116–K239) of Venus (ciV), carrying the point mutation A207K, to the N- or C-terminus ends of ArsB to express the ciV-ArsB or ArsB-ciV proteins, respectively. These fragments were then subcloned into the plant expression vector pGWB402. To visualize the interaction between ciV-ArsB or ArsB-ciV and two markers in living cells with the Agrobacterium-infiltrated leaves of N. benthamiana, the markers and ciV-ArsB or ArsB-ciV were transiently co-expressed. As a result, both the ciV-ArsB and ArsB-ciV proteins fluoresced when combined with the AALP-niV proteins, but not with RHM1-niV (Figure 1b); however, the expression of marker proteins alone did not cause fluorescence, suggesting that the N- and the C-terminal ends of ArsB were the vacuole lumen.

VBG localizes to the vacuole in differentiated leaf and root cells

The vacuolar R-SNARE VAMP711 is comprised of an N-terminal longin domain, central R-SNARE domain, and a C-terminal transmembrane region from aa 190 to 212 (Figure 2a). The N-terminal cytosolic longin domain of VAMP711 is required for recognition by the t-SNARE SYP22/AtVAM3 to localize to the vacuolar membrane.²⁴ We found that both, the N- and C-termini of ArsB, were exoplasmic (Figure 1). Therefore, we fused VAMP711 and GFP to the N- and C-terminal ends of ArsB (VAMP711-ArsB-GFP), designated as VBG, and generated the VBG-expressing line under the 35S promoter (Figure 2b). The fluorescence of VBG overlapped with that of the central vacuolar membrane marker, mRFP-AtVAM3, in the cells within the differentiation zone of the root, and pavement cells in the leaves (Figure 2c), however, was not merged with mRFP-AtVAM3, or the late endosome marker, mRFP-SNX21, in the cell division zone of the roots (Figure 2d). Taken together, these results suggest that VBG localized to the central vacuole membrane in differentiated leaf and root tissue, but not within the cell division zone.

Figure 2. VAMP711-ArsB-GFP (VBG) localizes to the vacuole membrane

(a) VAMP711 protein consists of three conservative domains. (b) VBG was constructed with VAMP711, ArsB, and GFP under 35 s promoter. Magenta: longin domain of VAMP711, green: R-SNARE domain, deep green: transmembrane region, gray: ArsB, grass green: VAMP711, pale green: GFP. (c) VBG was co-expressed with the central vacuole membrane marker (mRFP-AtVAM3) or late endosome marker (SNX1-mRFP) in the Arabidopsis root, and then observed by confocal microscopy. Bars = 10 μm. (d) Size of VBG-localized organelle was measured as the diameter.

The VBG longin domain is located on the cytoplasmic side

VBG was transiently expressed in Agrobacterium-infiltrated leaves of N. benthamiana,^25,26 and the leaves were treated with endomembrane marker FM4-64, which stains the PM. The fluorescence of VBG protein did not overlap with the plasma membrane marker FM4-64 (Figure 3a), suggesting that VBG was localized in the central vacuole membrane. The N-terminal portion of VBG was estimated to be on the cytoplasmic side as one transmembrane region of VAMP711 was fused to the exoplasmic N-terminus of ArsB (Figure 1). In contrast, the C-terminal portion of VBG was predicted to be on the exoplasmic side as soluble GFP proteins lacking the transmembrane region fused to the exoplasmic C-terminus of ArsB (Figure 1). To confirm these hypotheses, we evaluated the interaction between ciV-VAMP711-ArsB or VAMP711-ArsB-ciV and niV-fused markers using the iVenus assay. As shown in Figure 3, ciV-VAMP711-ArsB and RHM1-niV (cytosolic marker), as well as VAMP711-ArsB-ciV and AALP-niV, were the fluorescent combinations, however, AALP-niV (vacuole lumen marker) and RHM1-niV were not observed to overlap, although fluorescence was not observed in cells expressing each marker separately (Figure 3b). Together these data suggest that the vacuole-targeting signal of VAMP711 in VBG is exposed on the cytoplasmic side. Moreover, the orientation and localization data for VBG, indicate that the VSMP711-tag functions as a vacuole-targeting signal to localize ArsB to the central vacuolar membrane in differentiated leaf and root cells.

Figure 3. N-terminus of VBG is located on the cytosolic side and C-terminus is on the vacuole lumen side

(a) VBG was transiently expressed in Nicotiana benthamiana leaf, which was then treated with the plasma membrane marker FM4-64(b) In the iVenus assay, ciV-VAMP711-ArsB or VAMP711-ArsB-ciV were transiently expressed with RHM1-niV or AALP-niV in the leaves of N. benthamiana, and then observed by confocal microscopy. Bars = 10 μm. The topology of VAMP711-ArsB was determined by estimation. Pale green: Venus, magenta: longin domain of VAMP711, gray: ArsB, grass green: VAMP711, pink: RHM1, cyan: AALP.

VBG-expressing plant exhibits arsenic accumulation but not arsenic tolerance

We then examined the tolerance of VBG-expressing plants against As by measuring their root length following growth in 1/2 MS medium containing 0, 5, or 10 μM of As(III). No difference in As(III) tolerance was detected between the wild-type (WT) and three distinct VGB lines (Figure 4a). Next, we measured the arsenic concentration in VBG-expressing plants by ICP-OES. Although only one VBG line (VBG #4-5) accumulated a some higher concentration of As(III) in the shoots than the WT (Figure 4b), other VBG-expressing lines (VBG #2-1, VBG #3-1) tended to accumulate more As(III) in the shoot and root compared to WT plants (Figure 3b and Figure3c). Further, we measured the transcript levels of the VBG-expressing lines by real-time RT-PCR and observed that the VBG #4-5 exhibited the highest expression level of VBG transcript (Fig. S2) and As accumulation in shoot.

Figure 4. Growth assay of VBG plant at various As(III) concentrations

(a) VBG plants were grown on medium containing 0, 5, or 10 mM As(III), and the length of their primary root was measured. Data are shown as the mean ± SD *P < .001 (Student’s t-test; n > 20). Col-0 and VGB plants were grown in control medium for 4 days and then transferred to medium containing 10 μM As(III). After 10 days, the plants were harvested and As concentration was quantified in the shoot (b) and root (c) by ICP-OES. Data are shown as the mean ± SD *P < .001 (Student’s t-test; n > 4).

VBG-expressing plants exhibit high accumulation of various elements

Lastly, we measured the concentrations of other elements (Ca, Cu, Fe, Mg, Mn, Zn, S, and P), all of which, save for Cu and S, were found to be higher in the shoot of VBG-expressing plants compared to WT plants (Figure 5).

Figure 5. Various microelement concentrations in lines expressing VBG under As(III) stress

Col-0 and VBG plants were grown in control medium for 4 days and then transferred to medium containing 10 μM As(III). After 10 days, the plants were harvested and concentrations of elements (Ca, Cu, Fe, Mg, Mn, Zn, S, and P) were quantified in the shoot by ICP-OES. Data are shown as the mean ± SD *P < .001 (Student’s t-test; n > 4).

Discussion

Usage of VAMP711-tag led to generation of ArsB-introduced transgenic lines

The use of the VAMP711-tag led to generation of plant transgenic lines expressing ArsB. However, it is noted that we failed to generate transgenic Arabidopsis lines expressing GFP-ArsB as in the past studies, likely due to the toxicity of GFP-ArsB protein on plant cells. In the present study, we obtained transgenic plants expressing the chimerical construct VAMP711-ArsB-GFP, indicating that fusion with VAMP711 stabilized the structure of ArsB, thereby reducing its toxicity in plants.

Central vacuolar sequestration of As into VBG plants

VBG fluorescence did not merge with mRFP-AtVAM3 nor mRFP-SNX1 in root cell division zone (Figure 2d), indicating that VBG protein may reside in the small vacuole or an aberrant compartment induced by the chimerical construct in the root division zone.²⁷ In contrast, unlike the endosomal localization of GFP-ArsB during its transient expression (Figure 1a), VBG localized to the membrane of central vacuole (Figure 2c) since the VAMP711-tag functions as a transport tag for the vacuole membrane in differentiated leaf and root tissues. Furthermore, VBG plants showed a tendency to accumulate As in the shoots, suggesting that the bacterial As transporter ArsB is functional in seed plants. To the best of our knowledge, this is the first study to demonstrate that accumulation of As increased by localizing free-As transporter to the vacuolar membrane in seed plants.

VBG plants did not show increased tolerance to As, compared with WT, which may be explained by root health being affected by VBG preferentially localizing to the membrane of the small vacuole or aberrant compartment, with a low capacity for As accumulation in root division zone cells (Figure 2d), but not in the central vacuole membrane. The As transporter VBG is illustrated in the intracellular map in Fig. S3.

On the other hand, root-to-shoot translocation of essential metals (Fe, Mn, Cu, and Zn) is inhibited in Arabidopsis growing on medium containing As.²⁸ However, among the concentrations of eight elements (Ca, Cu, Fe, Mg, Mn, Zn, S, and P) measured in the shoot and root, most elements, save for Cu and S, showed higher concentrations in VBG plants than in WT plants (Figure 5), indicating that sequestration of As into the vacuole decreased the cytosolic As concentration in VBG plants, thereby increasing the root-to-shoot translocation of many nutrients.

In contrast, hyperaccumulators with high tolerance may be generated by leaf-specific expression of VBG while avoiding As accumulation in the root. Additionally, expressing the bacterial ATPase subunit ArsA in VBG plants may be effective for achieving hyperaccumulation and high tolerance of As.

Materials and methods

Plant materials

Arabidopsis thaliana ecotype Columbia (Col-0) was used as the WT species in all experiments. The following transgenic marker lines were used in accordance with previous studies: ST-mRFP,²⁹ mRFP-VHAa1,²⁹ SNX1-mRFP,³⁰ and mRFP-AtVAM3.³¹ Six-week-old Nicotiana benthamiana plants grown on soil in a short-day light cycle (8 h light/16 h dark) at 22°C were used for Agrobacterium-mediated transient gene expression^25,26and the iVenus assay.

Generation of transgenic Arabidopsis lines expressing VAMP711-ArsB-GFP in the WT background

The cDNA fragments of ArsB and VAMP711 were prepared by amplification from the plasmid pI258 or Arabidopsis cDNA with primer ArsB-N-f/ArsB-C-r and VAMP711-N-f/VAMP711-C-r (Table 1). The resulting PCR fragments were cloned into the pENTR vector to generate the pENTR-VAMP711-ArsB construct. After confirming the sequences, these inserts were introduced into pGWB405 by LR cloning to generate the pGWB405-VAMP711-ArsB construct. Next, transgenic Arabidopsis expressing VAMP711-ArsB-GFP was generated by Agrobacterium-mediated floral dip transformation. Positive transformants were selected based on their kanamycin resistance. Homozygous T3 plants were used for all analyzes.

Arsenic tolerance assays

Arabidopsis seeds were surface-sterilized and germinated on 1/2 Murashige–Skoog (MS) gellan gum plates (1/2 MS salts, 1% sucrose, pH 5.7, solidified with 1.2% gellan gum) with 0, 5, or 10 μM As(III). All plates were incubated at 4°C in the dark for 2 days to synchronize their germination. The plates were then placed vertically, and the plants were grown under white light at 22°C under a 16-h/8-h (light/dark) photoperiod. Homozygous mutant plants were selected from the F₂ segregating population based on genotyping and marker fluorescence. The root length was determined after several days of growth. All experiments were performed in a minimum of three independent biological replicates.

Total RNA isolation and cDNA synthesis

Total RNA was extracted from 6-day-old WT plants using the NucleoSpin RNA kit (TAKARA, Shiga, Japan) following the manufacturer’s instructions. The extracted total RNA was used for cloning of VAMP711, RHM1, and AALP and quantitative reverse transcription (RT)-PCR. For cDNA synthesis, 1 μg of total RNA was reverse-transcribed using ReverTra Ace qPCR RT Master Mix (TOYOBO, Osaka, Japan).

Plasmid constructs

All constructs were generated by Gateway cloning technology. To generate the ArsB, RHM1, AALP, and VAMP711-ArsB without the ArsB stop codon, their cDNA fragments were PCR-amplified and cloned into pENTR/D-TOPO vectors (Supplementary Table S1 and Table S2). Recombination with the Gateway recombinant cloning technology was performed to assemble multiple DNA fragments to generate the various constructs.

Quantitative RT-PCR

ArsB and UBQ10 cDNAs were amplified at an annealing temperature of 55°C for 40 cycles. The following primer pairs were used in qRT-PCR: ArsB-qRT-f/ArsB-qRT-r for ArsB and UBQ10-F/UBQ10-R for UBQ10.

Agrobacterium-mediated transient expression in Arabidopsis

GFP-ArsB was transiently expressed in intact Arabidopsis seedlings using the Agrobacterium-mediated transient expression system, AGROBEST (Agrobacterium-mediated enhanced seedling transformation).³²

Topology prediction

The web-versions of two different topology prediction methods were used to estimate membrane topology of ArsB and these were: MEMSAT³³ and phobias.³⁴

iVenus, a self-assembled split fluorescent protein analysis

iVenus (auto-interactive Venus) is a split fluorescent protein method that can be used to examine the orientation of both ends of a protein. Venus shows strong fluorescence intensity, and fast and efficient maturation properties. We, therefore, developed a modified split fluorescent protein method in which the engineered self-assembling fragments of only a yellow fluorescent protein reconstitute and gives a fluorescent signal when the two domains co-localize in the same cellular compartment. The sites at which the N-terminus (niV) and C-terminus (ciV) domains split were selected as described by Ohashi et al.²³ The N-terminus Venus (M1–I189) was fused with marker proteins for the vacuole lumen, AALP; the cytoplasm, RHM1, and C-terminus Venus (ciV) fragments (E116–K239) were fused with the N- or C-terminus of VBG. A point mutation of A to K at residue 206 was introduced into the ciV tag to reduce dimer formation.³⁵ The niV-fused marker protein and ciV-fused VBG were transiently expressed in the leaves of Nicotiana benthamiana by Agrobacterium infiltration, and the leaves were examined using a Leica TCS SP8 confocal microscope (Wetzlar, Germany).

Vector constructs for iVenus

The ciV or niV fragments were amplified by PCR and inserted into the plasmid pGWB402, and digested by XbaI or AfeI/SacI to generate the binary vectors pGWB402NciV or pGWB402 CciV and pGWB402 CniV.

FM4-64 staining

Tobacco leaf epidermal cells were stained by infiltration of leaves with 10 µM of the red fluorescent FM4-64[N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino)phenyl)hexatrienyl)pyridiniumdibromide].

Confocal microscopy

GFP, TagRFP, and Venus fluorescence signals, as well as differential interference contrast images were obtained using a Leica TCS SP8 laser scanning microscope, and the captured images were processed using Leica LAS X or ImageJ software (NIH, Bethesda, MD, USA). Imaging parameters for GFP and Venus were excitation at 488 nm and emission detection at 500–530, and FM4-64 TagRFP and mRFP, were excitation at 522 nm and emission detection at 574–594 nm, respectively.

Element analysis with ICP-OES

Element analysis of Arabidopsis plants was performed as described by Uraguchi et al. Briefly, Arabidopsis Col-0 and transgenic Arabidopsis lines were grown on 1/10^th of modified Hoagland medium with 1.5% agar for 10 days. The seedlings were then transferred to control plates or plates containing 5 μM As(III) and incubated for an additional 4 days. Shoots and roots from each plate (15 plantlets per individual line) were separately pooled. The concentrations of elements in the samples were quantified by inductively coupled plasma-optical emission spectroscopy (ICP-OES; iCAP7400Duo, Thermo Fisher Scientific, Waltham, MA, USA).

Accession numbers

The following Arabidopsis Genome Initiative locus identifiers were used for genes mentioned in this manuscript: VAMP711(AT4G32150), UBQ10 (At4g05320), AALP (At5g60360), and RHM1 (At1g78570).

Author contributions

Yusuke Deromachi, Shimpei Uraguchi, Kazuhiro Kuga, and Kohji Nishimura performed the experiments and analyzed the data; Tomoko Hirano and Masa H. Sato designed the experiments and supervised the work; Tomoko Hirano drafted the paper. The authors declare no competing financial interests. Correspondence and requests for materials should be addressed to Tomoko Hirano ([email protected]).

Highlight

The first arsenic (As) accumulator was developed using the bacterial As transporter ArsB, and vacuole-targeting signals VAMP711 in a seed plant.

Acknowledgments

We thank T. Nakagawa (Shimane University, Japan) for providing pGWB402, pGWB405, pGWB406 and pGWB460, and T. Uemura (Ochanomizu University, Japan) and T. Ueda (National Institute for Basic Biology) for providing ST-mRFP and VHA-a1-mRFP, and A. Miyawaki (the RIKEN BSI, Japan) for providing Venus. We thank G. Endo for providing the plasmid pI258.

Supplementary data:

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was supported by the Ministry of Education, Culture, Sports, Science and Technology Japan; a-grant-in-aid for Scientific Research [No. 17K08200].