Generation of Arabidopsis lines with a red fluorescent marker for endoplasmic reticulum using a tail-anchored protein cytochrome b₅-B

ABSTRACT

The endoplasmic reticulum (ER) is a multifunctional organelle that performs multiple cellular activities in eukaryotes. Visualizing ER using fluorescent proteins is a powerful method of analyzing its dynamics and to understand its functions. However, red fluorescent proteins with both an N-terminal signal peptide (SP) and a C-terminal ER retention tetrapeptide (HDEL) often cause mislocalization to vacuoles or extracellular spaces when they are constitutively expressed in Arabidopsis. To obtain a red fluorescent ER marker, we selected Arabidopsis cytochrome b₅-B (Cb5-B), a tail-anchored (TA) protein on the ER membrane. Its localization is determined by the transmembrane domain (TMD) and tail domain at the C-terminus. We fused the TMD and the tail domain of Cb5-B to the C-terminus of a red fluorescent protein, tdTomato (tdTomato-CTT). When tdTomato-CTT was constitutively expressed under the ubiquitin10 promoter in Arabidopsis, the fluorescent signal was exclusively detected at the ER by means of the reliable ER marker SP-GFP-HDEL. Therefore, tdTomato-CTT can accurately visualize the ER in stable Arabidopsis lines. Additionally, transient assays showed that tdTomato-CTT can also be used as an ER marker in onion, rice, and Nicotiana benthamiana. We believe that TA proteins could be used to generate various organellar membrane markers in plants.

The endoplasmic reticulum (ER) is an essential organelle that synthesizes and regulates the quality of proteins and lipids, calcium homeostasis, and carbohydrate metabolism.¹ The ER structure is composed of an elaborate polygonal network of interconnected tubules and flattened cisternae, with the nuclear envelope enclosing the nucleus. Additionally, some plants (e.g., Brassicaceae species) develop ER-derived spindle-shaped structures, called ER bodies, that accumulate β-glucosidases, which act as a defense mechanism against herbivores.² Numerous proteins have a signal peptide (SP) attached to the N-terminus and an ER retention signal, His-Asp-Glu-Leu (HDEL), attached to the C-terminus for localization to the ER. These signals are useful for visualizing the ER when connected with a green fluorescent protein (GFP). Therefore, transgenic Arabidopsis thaliana lines (GFPh) that stably express SP-GFP-HDEL under a cauliflower mosaic virus 35S promoter are widely used to better understand the structure, dynamics, and functions of the ER in plants.^3,4 ER visualization markers with other fluorescent proteins, such as a red fluorescent protein (RFP), are also valuable to the analysis of plant ER. For example, SP-mRFP-HDEL are often used for colocalization assays, along with GFP, to track unknown proteins through transient expression.⁵ However, we noticed that unlike SP-GFP-HDEL, the RFP variants with both the SP and the HDEL sequences often mislocalize to vacuoles or extracellular spaces when they are constitutively expressed in Arabidopsis (Figure 1). These results suggest that the HDEL peptide does not sufficiently retain RFP in the ER in stable transgenic Arabidopsis lines.

Figure 1.

Subcellular distribution of SP-mRFP-HDEL in Arabidopsis.

Arabidopsis transgenic lines, where both SP-mRFP-HDEL and SP-GFP-HDEL were constitutively expressed. (a) Mis-localization of SP-mRFP-HDEL to the vacuole and the extracellular spaces. (b) Mis-localization of SP-mRFP-HDEL to the extracellular spaces. An arrow and arrowheads indicate the vacuole and the extracellular spaces, respectively. SP-GFP-HDEL was used as an ER marker. Scale bars = 10 μm.

Tail-anchored (TA) proteins are a unique class of membrane proteins that are composed of three domains: an N-terminal cytosolic domain, a single hydrophobic transmembrane domain (TMD) near the C-terminus, and a short hydrophilic tail domain at the C-terminus.⁶ The intracellular localization is determined using the structural features of the TMD and the tail domain. TA proteins are distributed and inserted into various organellar membranes, including the ER membrane, via post-translational mechanisms such as the guided entry of tail-anchored proteins (GET) pathway and the ER membrane complex (EMC).^7,8 One such TA protein that is localized to the ER is cytochrome b₅ (Cb5), an electron-transfer protein that is a part of the redox system.^9,10 Among the five Cb5s (Cb5-A to Cb5-E) in Arabidopsis, Cb5-B, Cb5-C, Cb5-D, and Cb5-E are located on the ER membrane.^11,12 Therefore, we hypothesized that the TMD and the tail domain of Cb5 could be used to generate Arabidopsis transgenic lines to stably visualize the ER with red fluorescence.

We used Arabidopsis Cb5-B (At2g32720) as a donor for acquiring the ER-targeting domains as we previously demonstrated its localization to ER.¹² The TMD and tail region of Cb5-B were fused to the C-terminus of tdTomato, which emits a strong red fluorescence (Figure 2(a)). The construct named tdTomato-CTT (Cb5-B’s TMD and Tail region) was placed downstream of the promoter for ubiquitin 10 (UBQ10, At4g05320), due to its ubiquitous expression in Arabidopsis.

Figure 2.

Arabidopsis transgenic lines stably expressing tdTomato-CTT.

(a) Schematic images of tdTomato-CTT constructs. Cytochrome b₅-B (Cb5-B) contains three domains: cytosolic domain at the N-terminus, a transmembrane domain (TMD) near the C-terminus, and the tail domain at the C-terminus (upper). The C-terminal region of Cb5-B including the TMD and tail (CTT) were fused with tdTomato through the linker amino acids (Gly-Gly-Ser-Gly-Gly) (lower). (b) Immunoblot analysis of tdTomato-CTT using α-RFP antibodies in the microsomal fractions extracted from the seedlings of the 14-day-old Arabidopsis tdTomato-CTT lines and the tdTomato-CTT/GFPh lines. RHD3 was used as the control for microsomal membranes.¹³ (c) Phenotypes of 24-day-old tdTomato-CTT lines. Scale bar = 1 cm.

Immunoblot analyses demonstrated that tdTomato-CTT proteins were highly accumulated in the microsomal fraction of tdTomato-CTT line #1 (Figure 2(b)). This line did not show any phenotypic changes compared with the WT under normal growth conditions (Figure 2(c)). In tdTomato-CTT line #1, the tdTomato signal was observed in the reticular network and the membranes of ER bodies in both roots and cotyledons (Figure 3(a–d)). This suggests that tdTomato-CTT was localized on the ER membrane in Arabidopsis lines. We also established Arabidopsis lines that constitutively expressed both tdTomato-CTT and SP-GFP-HDEL (tdTomato-CTT/GFPh), to confirm the localization of tdTomato-CTT to ER. Immunoblot analysis showed that tdTomato-CTT/GFPh lines #4, #6, and #9 had high tdTomato-CTT protein expression (Figure 2(b)). We then compared the intracellular distribution patterns between tdTomato-CTT and SP-GFP-HDEL in tdTomato-CTT/GFPh line #9, and found both fluorescent signals (red and green) in the same reticular network of roots and cotyledons (Figure 3(e,f)). Additionally, the high-intensity GFP signals of the ER bodies were engrailed with the tdTomato signals (Figure 3(e,f), arrowheads). The tdTomato-CTT fluorescent signal was not detected in the vacuolar- or plasma membranes (Figure 3). These results indicate that, unlike SP-mRFP-HDEL, tdTomato-CTT exclusively localizes to the ER of Arabidopsis lines. Expression of tdTomato-CTT did not affect ER morphology, although previous reports suggested that overexpression of some transmembrane proteins, like calnexin, increased the flat sheets of ER.¹³

Figure 3.

Subcellular distribution of tdTomato-CTT in Arabidopsis lines.

(a-d) The roots (a and b) and cotyledons (c and d) of 8-day-old Arabidopsis tdTomato-CTT line #1 were observed.(e and f) The roots (e) and cotyledons (f) of the 8-day-old Arabidopsis tdTomato-CTT/GFPh lines, which express both tdTomato-CTT and SP-GFP-HDEL, were observed. Scale bars = 10 μm. Arrowheads indicate ER bodies (b, d, e, and f).

To assess whether tdTomato-CTT functions as an ER marker in multiple plant species, we transiently expressed it in onion (Allium cepa) epidermal cells using the particle bombardment method. As shown in Figure 4(a), the tdTomato signal was detected in the perinuclear region and cell periphery, and colocalized completely with the signal of the ER marker with SP-CFP-HDEL, which contained a cyan fluorescent protein (CFP). Similarly, the tdTomato-CTT fluorescent signal colocalized with that of SP-CFP-HDEL in rice (Oryza sativa) protoplasts (Figure 4(b)), when we introduced tdTomato-CTT in rice using a polyethylene glycol-mediated protoplast transformation method. The tdTomato-CTT signal did not overlap with that of Venus-OsRac1, a small GTPase that localizes in the plasma membrane¹⁴ (Figure 4(c)), further confirming the ER-specific localization of tdTomato-CTT. When we introduced tdTomato-CTT in Nicotiana benthamiana (N. benthamiana) by the agrobacterium-mediated infiltration assay, the tdTomato-CTT signal in the reticular network was colocalized with that of SP-GFP-HDEL (Figure 4(d)). Our results identify tdTomato-CTT as an useful red fluorescence ER marker in various plant species, for both transient expression and stable transformation. TA proteins are distributed to the ER as well as the nuclear envelop, mitochondrial outer membrane, chloroplast outer membrane, peroxisomal membrane, plasma membrane, and even the chloroplast thylakoid membrane,⁶ suggesting that the use of TA proteins would be valuable to generate various membrane markers with fluorescent proteins. We hope that our findings will contribute to the advancement of cell biology in plants.

Figure 4.

Subcellular distribution of transiently expressed tdTomato-CTT.

(a) tdTomato-CTT was co-expressed with SP-CFP-HDEL (ER marker) in onion epidermal cells using the particle bombardment method. Scale bar = 100 μm. Arrowhead indicates the nuclear envelope. (b and c) tdTomato-CTT was co-expressed with SP-CFP-HDEL (b, ER marker) or Venus-OsRac1 (c, plasma membrane marker) in rice protoplasts. Scale bars = 10 μm. Arrowheads indicate the nuclear envelope. (d) tdTomato-CTT was co-expressed with SP-GFP-HDEL in the epidermal cells of N. benthamiana (tobacco) using the agrobacterium-mediated infiltration assay. Scale bar = 10 μm.

Materials and methods

Plant materials

The Columbia (Col-0) accession of Arabidopsis thaliana (L.) Heynh was used in this study. Transgenic Arabidopsis GFPh lines expressing SP-GFP-HDEL were described in Matsushima et al. (2003).¹⁵ Plants were cultivated in half MS medium or in soil at 23°C under continuous light (60 μmolm⁻² s⁻¹).

Plasmid preparation and generation of Arabidopsis transgenic lines

To produce the tdTomato-CTT, the Cb5-B region (102–134 aa) containing the TMD and tail was amplified using PCR (forward primer: 5ʹ-GGAGGATCAGGAGGAGACAAGACCTCTGAATTCAT-3ʹ, reverse primer: 5ʹ-CTACCCTGATTTGGTGTAGA-3ʹ) and was then combined with tdTomato, also amplified using PCR (forward primer: 5ʹ-ATGGTGAGCAAGGGCGAGGA-3ʹ, reverse primer: 5ʹ-TCCTGATCCTCCCTTGTACAGCTCGTCCATGC-3ʹ), at the N-terminus via the linker sequence (Gly-Gly-Ser-Gly-Gly). Then, tdTomato-CTT was introduced into the destination vector, pB2BW7-UBQ10pro, which contained the Arabidopsis promoter UBQ10,¹⁶ using the Gateway system (Invitrogen).

The SP-mRFP-HDEL fragment was inserted into the EcoRI-HindIII site of pBI121, from SP-mRFP-HDEL/pBI221,⁵ for its constitutive expression in Arabidopsis.

The plasmids were transformed into Arabidopsis Col-0 (WT) or the GFPh lines, using the Agrobacterium-mediated floral dip method.

Particle bombardment

Plasmid DNA was introduced into onion epidermal cells using particle bombardment with a helium-driven particle accelerator (GIE-3 IDERA, TANAKA). The fluorescence was analyzed using a confocal laser scanning microscope (TCS-SP5, Leica).

Protoplast transformation

Protoplast extraction from rice suspension cultures and transformation was performed according to a method in Nagano et al. (2016).¹⁷ After incubating for 12 h to 18 h at 30°C, the protoplasts were observed using a confocal laser scanning microscope (TCS-SP5, Leica).

Agrobacterium-mediated infiltration assay

Agrobacterium tumefaciens strain GV3101 carrying the tdTomato-CTT or SP-GFP-HDEL was cultured overnight and diluted to OD₆₀₀ = 0.1 with distilled water, and was simultaneously infiltrated into the abaxial air space of four-week-old N. benthamiana using a 1 mL syringe. Fluorescence was analyzed using a confocal laser scanning microscope (LSM 800, Zeiss).

Confocal microscopy

The Leica TCS-SP5 was used to observe onion epidermal cells and rice protoplasts. CFP and Venus were excited using 458 nm and 514 nm argon (Ar) lasers, respectively. tdTomato was excited using a 543 nm helium (He)-neon (Ne) laser. The Olympus FV10i was used to detect fluorescence in Arabidopsis lines. GFP was excited using a 473 nm Ar laser, and tdTomato was excited using a 559 nm He/Ne laser. The Zeiss LSM 510 and 800 were used to observe Arabidopsis lines and N. benthamiana. GFP was excited using a 488 nm laser, and tdTomato was excited using a 561 nm laser.

Extraction of microsomal fractions and immunoblot analysis

Raw 14-day-old Arabidopsis plants were homogenized in buffer (50 mM Tris-HCl [pH 7.5], 2 mM EDTA [pH 7.5], 150 mM NaCl, 5 mM MgCl₂, 10% glycerol, and protease inhibitor cocktail [Roche Diagnostics, Basel, Switzerland]). The homogenates were centrifuged at 8,000 g (4°C, 20 min). The supernatants were centrifuged at 100,000 g (4°C, 1 h), using an ultracentrifuge. The pellets were resuspended in homogenizing buffer with 0.5% CHAPS, at 4 °C with rotation. After centrifugation at 15,000 rpm (4°C, 20 min), the supernatants were subjected to immunoblot analysis. α-RFP antibodies were purchased from MBL. The α-RHD antibody was previously descried in Ueda et al. (2016).¹⁸

Funding Statement

This research was supported by a Grant-in-Aid for Japan Society for the Promotion of Science (JSPS) Fellows Grant Nos. [07J03838 and 11J08349] to M.N.; JSPS KAKENHI Grant Nos. [26850232, 17K15412, and 20K05966 to M.N., No. 15H05776 to I.H.-N., and Nos. 15KT0151, 16K07397, and 19K06732 to H.U., No. 18K19164 to M.K.-Y.], and by the Hirao Taro Foundation of KONAN GAKUEN for Academic Research [to I.H.-N. and H.U.].

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Author contributions

M.N. and H.U. designed the research, performed the experiments and wrote the manuscript. All authors reviewed and approved the manuscript.