Three-Fragment Fluorescence Complementation Coupled with

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Three-Fragment Fluorescence Complementation Coupled with Photoactivated Localization Microscopy for Nanoscale Imaging of Ternary Complexes Minghai Chen, Sanying Liu, Wei Li, Zhi-Ping Zhang, Xiaowei Zhang, Xian-En Zhang, and Zongqiang Cui ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.6b03543 • Publication Date (Web): 01 Sep 2016 Downloaded from http://pubs.acs.org on September 2, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Nano is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Three-Fragment Fluorescence Complementation Coupled with Photoactivated Localization Microscopy for Nanoscale Imaging of Ternary Complexes

Minghai Chen†,#,§, Sanying Liu†,#,§, Wei Li†, Zhiping Zhang†, Xiaowei Zhang†, Xian-En Zhang‡,*, Zongqiang Cui†,*

†

State Key Laboratory of Virology, Wuhan Institute of Virology, Chinese

Academy of Sciences, Wuhan 430071, China ‡

National Laboratory of Biomacromolecules, CAS Center for Excellence in

Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing 100101, China #

University of Chinese Academy of Sciences, Beijing 100049, China

§

These authors contributed equally to this work

*To whom correspondence should be addressed [email protected] or [email protected].

1

ACS Paragon Plus Environment

ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT Many cellular processes are governed by molecular machineries that involve multiple protein interactions. However, visualizing and identifying multi-protein complexes such as ternary complexes inside cells is always challenging, particularly in the sub-diffraction cellular space. Here, we developed a three-fragment fluorescence complementation system (TFFC) based on the splitting of a photoactivatable fluorescent protein, mIrisFP, for the imaging of ternary complexes inside living cells. Using a combination of TFFC and photoactivated localization microscopy (PALM), namely the TFFC-PALM technique, we are able to identify the multi-interaction of a ternary complex with nanometer-level spatial resolution and single-molecule sensitivity. The TFFC-PALM system has been further applied to the analysis of the Gs ternary complex which is composed of αs, β1, and γ2 subunits, providing further insights into the sub-cellular localization and function of G protein subunits at the single-molecule level. The TFFC-PALM represents a valuable method for the visualization and identification of ternary complexes inside cells at the nanometer scale.

KEYWORDS mIrisFP; TFFC; PALM; ternary complexes; G protein

2


Page 2 of 36

Page 3 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

TOC GRAPHIC

Many cellular processes are governed by multi-protein machineries that rely on protein-protein interactions (PPIs). In particular, many signaling proteins such as G protein function as ternary complexes.1, 2 However, visualizing and identifying multi-protein interactions such as those comprising ternary complexes remains a challenging task. At present, several fluorescence-based methods such as fluorescence resonance energy transfer (FRET),3,

4

bioluminescence resonance energy transfer (BRET),5-7 singlet oxygen triplet energy

transfer-based

imaging

technology

(STET),8

fluorescence

cross-correlation spectroscopy (FCCS),9 linear expression cassette coupled with bimolecular fluorescence complementation (LEC-BiFC)10 and bimolecular fluorescence complementation (BiFC),11-15 are available and used for studying PPIs inside cells. For example, the BiFC assay, which relies on the reconstruction of a reporter fluorescent protein from its two divided non-fluorescent fragments, is very popular for imaging intracellular PPIs. However, these methods usually permit the analysis of interactions between two proteins.

3


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Recently, the three chromophore-based FRET (3-FRET) system16 and the BiFC-based FRET (BiFC-FRET) assay17 were developed for the visualization of ternary complexes within cells. However, these assays have not been widely used because they require skilled personnel, sophisticated microscopes and data analysis. In addition, the optical crosstalk and contamination that occurs during these assays often complicates interpretation of the results. The development of a simple and convenient assay for analyzing ternary complexes inside living cells would therefore have a significant impact.

Another major concern regarding the analysis of protein interactions is that the functions of protein complexes are closely related to their spatial distribution, and these complexes are often spatially regulated at the nanometer scale inside cells. However, the conventional fluorescence-based methods for PPIs assays face limitations in terms of resolution. Due to the diffraction of light, the best spatial resolution achieved with such methods is ~200 nm, which does not meet the demands of a nanometer-scale assay. In addition, these methods are incapable of imaging and analyzing individual PPIs. Several techniques, such as photoactivated localization microscopy (PALM)18 and stochastic optical reconstruction microscopy (STORM),19 have recently been developed to overcome the resolution limit of fluorescence microscopy. PALM is a method of sequentially localizing sparse subsets of photoactivatable fluorescent proteins to obtain a sub-diffraction-limited 4


Page 4 of 36

Page 5 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

fluorescent image. Based on stochastic switching and sub-diffractive localization of individual fluorescent molecules, PALM has been realized ~20 nm spatial resolution.20,

21

More recently, BiFC systems based on split

PAmCherry122 or mEos3.223 have been developed for the nanoscale imaging of PPIs with PALM. Unfortunately, however, a method for imaging ternary complexes with high resolution is still unavailable.

In this work, we developed a three-fragment fluorescence complementation system (TFFC) based on split mIrisFP for imaging the ternary complexes inside living cells, and for imaging the nanoscale sub-cellular distribution of individual ternary complexes with PALM. TFFC relies on the reconstruction of a reporter fluorescent protein from its three divided non-fluorescent fragments. mIrisFP,

a

recently

developed

monomeric

photoactivatable

and

photoconvertible fluorescent protein,24 was first split at different sites to build the BiFC-PALM systems. Then, the TFFC-PALM system was constructed by splitting mIrisFP into three non-fluorescent fragments. This TFFC-PALM method was applied towards the study of the interactions of the αs, β1, and γ2 subunits of the heterotrimeric Gs protein, which plays important roles in several signal transduction pathways.25, 26 The αsβ1γ2 heterotrimer was imaged inside cells at the single-molecule level, and the results revealed the different distribution patterns and biological functions of the αsβ1γ2 heterotrimer and the β1γ2 heterodimer at the nanometer scale. 5


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

RESULTS AND DISCUSSION

To develop the TFFC system, we first constructed an mIrisFP-based BiFC system. Based on the crystal structure of mIrisFP (PDB accession code 2VVH)27 and sequence alignments with other fluorescent proteins (mCherry and YFP), Three sites were selected among amino acids (aa) 134/135, 150/151, and 165/166, for splitting mIrisFP within the loops of its barrel-like secondary structures to develop the BiFC system (Figure 1a). The site aa 157/158 localized in the β-sheet of the mIrisFP’s secondary structures, was selected as a control split site. The fragment pairs obtained were named mIN134/mIC135, mIN150/mIC151, mIN157/mIC158, and mIN165/mIC166. The coding sequences of these mIrisFP split reporter fragments were inserted into the pcDNA3.1 plasmid to construct the vectors pmIN134, pmIC135, pmIN150, pmIC151, pmIN157, pmIC158, pmIN165 and pmIC166. The well-known interaction of the bJun and bFos protein pair was then used to test the BiFC systems.11 Either bJun or bFos encoding sequences were inserted into pmIN134, pmIC135, pmIN150, pmIC151, pmIN157, pmIC158, pmIN165 and pmIC166, to express bJun-(mIN) N-terminal and bFos-(mIC) C-terminal fusion proteins.

When the four sets of bJun-mIN and bFos-mIC proteins were co-expressed 6


Page 6 of 36

Page 7 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

in Vero cells, the combination of bJun-mIN150 and bFos-mIC151 yielded a strong green BiFC signal from the reconstitution of the mIrisFP molecules (Figure 1b). After the sample was irradiated with a 405-nm laser light for 30 s, the reconstituted green mIrisFP molecules were converted to the red form and a red BiFC signal could then be detected (Figure 1b). The combination of bJun-mIN165 and bFos-mIC166 also yielded moderate green and red BiFC signals (Figure 1b). As a control, we also fused mbFos to mIC151 or mIC166 and co-expressed these proteins with bJun-mIN150 or bJun-mIN165, respectively. mbFos is a mutated bFos known to eliminate the interaction between bFos and bJun, but which has no effect on its interaction with other partners.11 As expected, in the presence of mbFos, only very weak fluorescence signals were visible (Figure 1b). Quantitative analyses confirmed the significant differences for the BiFC signals between these bJun-bFos and bJun-mbFos BiFC systems. The bJun-bFos BiFC signals were ~70-fold higher than bJun-mbFos BiFC signals (Figure 1c). Western blot analyses showed that these fusion proteins were, however, expressed at equivalent levels in the BiFC system (Figure 1d). For the other fragments obtained by splitting mIrisFP at the 134/135 or 157/158 sites, the combination of bJun-mIN134 with bFos-mIC135, and bJun-mIN157 with bFos-mIC158, did not result in any observable BiFC signals (Figure 1e). This meant that the spatial configuration of the mIrisFP could not be reconstituted after the protein was split at the 134/135 or 157/158 sites. These results showed that the mIrisFP could be split 7


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 36

at 150/151 or 165/166 site to develop BiFC system for imaging PPIs.

Next, the mIrisFP-based BiFC fluorescent system was tested for the imaging of PPIs at the nanometer scale with the PALM technique. As shown in Figure 2, the red form of the reconstituted mIrisFP was imaged with a 561 nm excitation laser and a 488-nm light for on-switching after photoconversion with 405 nm laser irradiation for 30 s. The BiFC-reconstituted mIrisFP not only maintained the

ability

to

reversibly

photoswitch

between

the

fluorescent

and

non-fluorescent states (Figure 2a), but also held similar photophysical properties to those of native mIrisFP (Figure S1, Video S1 and Figure S2). We then imaged the bJun/bFos complexes using the mIrisFP-based BiFC-PALM system. Single bJun/bFos complexes could be detected in single-molecule images of reconstituted mIrisFP, and each dot represents a single bJun/bFos complex (Figure 2b, Video S2). BiFC-PALM imaging provided a greater level detail and higher resolution (~18 nm molecular organization uncertainty, σ) compared with the traditional method of total internal reflection (TIRF) imaging (Figure 2c,d), acquiring an average of ~900 photons per localization event (Figure 2e). The green form of reconstituted mIrisFP was also imaged using BiFC-PALM, but certain characteristics such as the distribution of σ and the photon number, were inferior to those of the red form (Figure S3a–d). These results demonstrated that split mIrisFP could be used to successfully develop a BiFC-PALM system for the imaging of PPIs at the nanometer scale. 8


Page 9 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

After developing the BiFC-PALM fluorescent system based on mIrisFP, we attempted to split mIrisFP into three non-fluorescent fragments to construct a TFFC system for imaging multi-protein interactions. The schematic principle of the TFFC system is shown in Figure 3a. When the appropriate split sites are chosen, these three fragments can be brought together and reconstituted to emit fluorescence, as a result of the simultaneous and successful association of the three protein units fused to each mlrisFP fragment, thereby allowing the analysis of the ternary complex of interest. Based on the crystal structure of mIrisFP and the results obtained from the mIrisFP-BiFC systems, we designed four schemes for splitting mIrisFP into three fragments, and all split sites were localized within the unstructured flexible loops (Figure 3). The three fragments were denoted fragment 1, fragment 2 and fragment 3. The Fos-Jun-NFAT ternary complex was used as the test model for constructing the TFFC system. NFAT is a transcription factor required for T cell development and for many other cellular processes,28 and is known to form a ternary complex with the Fos-Jun heterodimer to regulate many target genes.28, 29 NFAT1 was fused to the fragment 1, while the basic region leucine zipper (bZIP) domain of Jun (bJun) and Fos (bFos) were fused to fragments 2 and 3, respectively.

Our results showed that when mIrisFP was split at residues 150/151 and 165/166, the three mIrisFP fragments could be reconstituted to recover the 9


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

fluorescence of native mIrisFP via the association of the bFos, bJun, and NFAT1 heterotrimer (Figure 4a–c). Split mIrisFP fragments obtained from the other three schemes did not result in any observable fluorescence (Figure 4d). When mNFAT1 (where three amino acids in NFAT1, R466A/I467A/T533G, were mutated to destroy its interaction with the bJun-bFos heterodimer)17 or mbFos were included in the system, or when any one of the three mlrisFP fragments such as the short middle fragment was absent, no fluorescence signal could be detected (Figure 4a). In addition, co-expression of the three free mIrisFP fragments, mIN150, mIN(151–165), and mIC166, without fusion to the bFos-bJun-NFAT1 heterotrimer, also did not produce a fluorescence signal (Figure S4). The statistical analysis of the complementary fluorescence intensities produced in these assays is shown as Figure 4b. Western blot analyses confirmed that all fusion proteins were expressed at comparable levels (Figure 4c). In summary, the mIrisFP protein was successfully split into three fragments in the region of residues 150/151 and 165/166 to generate a mIrisFP-TFFC system for identifying and imaging the ternary protein interactions.

We next used the mIrisFP-based TFFC fluorescent system to image multi-protein interactions at the nanometer scale with the PALM technique. The red form of mIrisFP-TFFC PALM imaging is shown in Figure 5. The TFFC-reconstituted mIrisFP not only maintained a reversible photoswitching 10


Page 10 of 36

Page 11 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

character (Figure 5a), but also held similar photophysical properties to those of native mIrisFP (Figure S2). We first imaged the bJun-bFos-NFAT1 heterotrimer using TFFC-PALM. Single bJun-bFos-NFAT1 heterotrimers could be detected in single-molecule images of reconstituted mIrisFP, and the spots with apparently higher fluorescence represent the bJun-bFos-NFAT1 heterotrimers (Figure 5b, Video S3). The bJun-bFos-NFAT1 heterotrimer exhibited a discontinuous dot distribution in the high-resolution PALM image (Figure 5c). TFFC-PALM imaging provided a greater level of detail with higher resolution (~18 nm molecular organization uncertainty, σ) compared with TIRF imaging (Figure 5c, d), acquiring an average of ~770 photons per localization event (Figure 5e). Similarly to the results obtained with BiFC-PALM imaging, the green form of mIrisFP-TFFC could also be used for TFFC-PALM, although this form was again inferior to the red form (Figure S5a–d). These results demonstrate that the mIrisFP TFFC fluorescent system is a powerful tool for imaging multi-protein interactions at the nanometer scale.

With the mIrisFP-based TFFC-PALM fluorescent system, it is possible to identify and image ternary complexes with a sub-diffraction-limit spatial resolution. To test-drive this TFFC-PALM method, we next applied it towards the imaging of the Gs protein αβγ heterotrimers, as they play important roles in signal transduction. Heterotrimeric G proteins are composed of α and βγ subunits. The βγ complex only dissociates when denatured.30-32 Upon receptor 11


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

activation the GTP-bound α is disassociated from the βγ dimer and transmits signals to downstream molecules, and the GDP-bound α can then reassociate with the βγ dimer to reconstitute the heterotrimer. Although the sub-cellular localization, function and regulation of the subunits of the G protein heterotrimers have been extensively studied,33-35 the actual ternary αβγ complex has not yet been directly observed and investigated in cells because of the absence of the appropriate methodology. Here, by using our TFFC-PALM fluorescent system, we were able to visualize the αsβ1γ2 heterotrimer in cells at the single-molecule level. As shown in Figure 6a, the interactions among the αs, β1, and γ2 subunits resulted in the reconstitution of the three mIrisFP fragments, resulting in fluorescence emission in the TFFC-PALM system. The αsβ1γ2 heterotrimer showed a discontinuous small dot distribution pattern on the cell membrane (Figure 6a and Figure S6a). Each dot in the high-resolution image represented a single αsβ1γ2 heterotrimer complex. As a control, when αs was mutated (denoted as αsm) to destroy its interaction with β1γ2,36 no fluorescence signal could be detected in the TFFC-PALM system (Figure S7).

We also visualized the β1γ2 heterodimer using the mIrisFP-based BiFC-PALM system. As shown in Figure 6b, the β1γ2 heterodimer, which produced BiFC signals from the interaction between the β1 and γ2 units, was imaged at a single-molecule scale. The β1γ2 heterodimers were present in two 12


Page 12 of 36

Page 13 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

forms, namely clusters and small dots, and these were primarily located in the cytoplasmic zone (Figure 6b,c and Figure S6b), which differs from the cell membrane localization observed for the αsβ1γ2 heterotrimers.

To further characterize the αsβ1γ2 heterotrimer, cells were treated with cholera toxin during the TFFC-PALM imaging assay. Cholera toxin can cause Gs protein to abnormally maintain activation, forcing Gαs to dissociate from the Gβγ heterodimer.37 During the assay, Vero cells were transfected with a combination of pαs-mIN150, pβ1-mIN(151–165) and pγ2-mIC166 and then treated with 3µg/mL of cholera toxin at 5 h after transfection. After 16–18 h of further incubation, the cells were imaged with PALM. The results showed that in cholera toxin-treated cells, the quantity of αsβ1γ2 heterotrimers showing TFFC fluorescence decreased significantly compared to that in untreated cells, while the BiFC fluorescence signals from the β1γ2 heterodimers showed no apparent difference between the cholera toxin-treated and untreated cells (Figure 6d). These results suggested that cholera toxin inhibited the formation of the αsβ1γ2 heterotrimer, while it had no obvious influence on the formation of the β1γ2 heterodimer. In our experiments, we also attempted to use cholera toxin to reduce pre-existing TFFC signals arising from the αsβ1γ2 interaction. Vero cells with TFFC fluorescence signals arising from the αsβ1γ2 heterotrimers were treated with cholera toxin for various time periods (e.g. 20 min, 40 min or 60 min), and then analyzed with PALM imaging. No obvious decrease in TFFC 13


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

fluorescence signals was apparent in the cells treated with cholera toxin compared with untreated cells (Figure S8). As cholera toxin causes Gαs to dissociate from the Gβγ heterodimer, a reasonable explanation for this result maybe that the mIrisFP TFFC system is irreversible, similar to most BiFC systems.12, 38

The development of a convenient assay for the visualization and identification of ternary complexes at the nanometer scale will facilitate significant insights in protein interaction studies. In this study, we constructed the TFFC fluorescent system based on mIrisFP for imaging ternary complexes inside living cells. By taking advantage of the photoactivatable properties of mIrisFP, we are able to visualize ternary complexes at the single-molecule level by combining the mIrisFP-based TFFC system with the PALM technique. Our TFFC-PALM fluorescent system therefore provides a powerful tool for the direct visualization and identification of ternary complexes inside cells with higher spatial resolution compared with traditional TIRF imaging. The cleavage sites in mIrisFP used to construct the TFFC system also provides a good reference point for the development of TFFC systems using other fluorescent proteins. The reconstituted form of mlrisFP comprising the three split mIrisFP fragments retains the photophysical properties of native mIrisFP, which validated its feasibility for use in the TFFC-based PALM imaging technique. We believe this breakthrough in the creation and application of TFFC systems 14


Page 14 of 36

Page 15 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

will promote the study of multi-protein interactions inside cells. Of course, there are possible limitations for the TFFC system, such as inability to quantify interaction strength between protein partners in the complexes, and the orientation influence of the fragments as fusion to the N or C-terminus of the interaction proteins. During this study, several mIrisFP-based BiFC-PALM systems were also built for visualizing and identifying individual PPIs inside cells. These systems also add tools to the current suite of BiFC-PALM systems, which, until now, only comprised the recently developed PAmCherry1- and mEos3.2-based BiFC-PALM systems,22, 23 thereby providing alternative tools for PPIs analyses at high resolution.

The TFFC-PALM fluorescent system and BiFC-PALM were also used to study the heterotrimer subunits of Gs protein, which play important roles in cell signaling pathways. Our results provided the previously unobserved localization patterns for the αsβ1γ2 heterotrimer and the β1γ2 heterodimer at the single-molecule level. The αsβ1γ2 heterotrimer showed a uniform small dot distribution pattern on the cell membrane, while the β1γ2 dimers were primarily localized in the cytoplasm in clusters or small dot forms. The different localization patterns observed for the αsβ1γ2 heterotrimer and β1γ2 heterodimer correlate with their known biological functions, and thus further indicate that the αs subunit plays key roles in the sub-cellular translocation of Gs protein during signal transduction. 15


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Cholera toxin treatment significantly decreased the formation of TFFC fluorescence dots comprising the αsβ1γ2 heterotrimer, while such treatment had no obvious influence on the formation of β1γ2 heterodimer. These observations further confirm that signal activation causes the αs to dissociate from the β1γ2 subunit, indicating the important role of αs in the recycling of the αsβ1γ2 heterotrimer and β1γ2 heterodimer. After the αsβ1γ2 subunits were observed to form a stable interaction in the mIrisFP TFFC system, the TFFC signals emitted as a result of the αsβ1γ2 interaction could not be reduced by cholera toxin treatment. One of the explanations for this result may be that the mIrisFP TFFC system is irreversible, as most BiFC systems.12, 38, 39 Reversible TFFC system allows to study the dynamic combination and dissociation of protein complexes. Further efforts should be dedicated to developing a reversible TFFC system, though the irreversible TFFC system offers significant advantages for the screening of weak or transient interacting partners.

CONCLUSION In conclusion, we developed the method TFFC-PALM for studying cellular multi-protein interactions with high spatial resolution. The TFFC-PALM fluorescent system represents a valuable tool for the imaging of ternary complexes at the single-molecule level. Application of TFFC-PALM towards the study of the Gs protein αsβ1γ2 heterotrimer also provides insights into the 16


Page 16 of 36

Page 17 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

sub-cellular localization and function of G protein subunits at the single-molecule level.

METHODS

Construction of plasmids: The mIrisFP protein was split at four positions between aa 134/135, 150/151, 157/158, and 165/166 to construct the BiFC systems. The fragment pairs obtained were named mIN134/mIC135, mIN150/mIC151, mIN157/mIC158, and mIN165/mIC166, which correspond to the aa sequences of 1-134 and 135-226, 1-150 and 151-226, 1-157 and 158-226, and 1-165 and 166-226, respectively. For the construction of pbJun-mIN134, pbJun-mIN150, pbJun-mIN157 and pbJun-mIN165 plasmids, bJun was amplified by PCR from plasmid pbJun-iRN97.14 Kozak sequences and a HA tag were included in the forward primer and inserted into the corresponding sites of pcDNA3.1 (+) using NheI and KpnI restriction enzymes. mIN134, mIN150, mIN157 and mIN165 were amplified by PCR from the mIrisFP sequence (Sangon, Shanghai) and inserted into the corresponding sites of pcDNA3.1-bJun using BamHI and EcoRI restriction enzymes. The coding regions were connected by the linker sequence GGGGSGGGGS. To construct the plasmids of pbFos-mIC135, pbFos-mIC151, pbFos-mIC158 and pbFos-mIC166, bFos was amplified by PCR from plasmid piRC98-bFos.14 Kozak sequences and a flag tag were included in the forward primer and 17


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

inserted into the corresponding sites of pcDNA3.1 (+) using NheI and KpnI restriction enzymes. mIC135, mIC151, mIC158 and mIC166 were amplified by PCR from the mIrisFP sequence and inserted into the corresponding sites of pcDNA3.1-bFos using BamHI and EcoRI restriction enzymes. The coding regions were also connected by the linker sequence GGGGSGGGGS. The plasmids of pmbFos-mIC151 and pmbFos-mIC166 were constructed by replacing bFos with mbFos at the same sites of pbFos-mIC151 and pbFos-mIC166. mbFos was amplified by PCR from plasmid piRC98-mbFos.14 We chose the Fos-Jun-NFAT1 ternary complex as the test model to construct the mIrisFP-TFFC systems because of their well-known strong interactions. We split mIrisFP into three non-fluorescent fragments. Fragment 1 was fused to NFAT1 (Sangon, Shanghai) or (R466A/I467A/T533G) NFAT1 (named mNFAT1), fragment 2 was fused to bJun, and fragment 3 was fused to bFos or mbFos. For the construction of plasmids pβ1-mIN150 and pmIC151-γ2, β1 and γ2 were PCR-amplified from random-primed HeLa cDNA and fused to the fragments mIN150 and mIC151, respectively. For the construction of pαs-mIN150, pβ1-mIN(151-165) and pγ2-mIC166 plasmids, αs (Sangon, Shanghai), β1 and γ2 were fused to the N-terminus of mIN150, mIN(151-165) and mIC166, respectively. The primers used in this study are listed in Table S1. All of the sequences were verified by DNA sequencing. 18


Page 18 of 36

Page 19 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Cell culture and transfection: Vero cells were grown in Dulbecco’s Modified Eagle’s Medium (DMEM) containing 10% fetal calf serum, 100 U/ml penicillin, and 100 µg/ml streptomycin at 37°C under 5% CO2 in a humidified incubator. The cells were seeded the day before transfection in 35-mm glass bottom cover slips at 70-80% confluency. Transfection of the Vero cells with the plasmids was conducted using 2µg of each component with Lipofectamine 2000 (Invitrogen; United States) following the manufacturer’s instructions. For the experiment characterizing Gs subunits, 3 µg/ml cholera toxin was added to the medium at 5 h after transfection. The transfected cells were incubated at 37°C (5% CO2) for 8-12 h followed by 10-12 h at 27°C (5% CO2) before imaging. For PALM imaging, cells were fixed in fresh 4.0% PFA for 30 minutes at room temperature and changed to PBS buffer after fixation.

Confocal fluorescent microscopy, image acquisition and data analysis: The cells were imaged using an UltraVIEW VOX Confocal system (PerkinElmer, Co.) with a 60X, 1.4 numerical aperture (NA), oil immersion objective lens. The green fluorescence from the mIrisFP green channel was excited at 488 nm. The red fluorescence from the mIrisFP red channel was excited at 561 nm after the sample was irradiated with a 405-nm laser light (10 mW cm-2) for 30 s. The nuclei were stained using Hoechst 33342, which was excited at 405 nm. The original fluorescence images were acquired using the 19


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 36

same exposure time in the same channel. For fluorescent quantitative analysis, the green and red fluorescence intensities were acquired from their original images by subtracting their background fluorescence. The background values were the average fluorescence intensities obtained from a region (50 × 50 pixel2) without green and red signals. The average fluorescence intensities of approximately 60 fluorescent cells (from randomly selected 10-15 images) were used for the relative BiFC and TFFC efficiency calculations.

Western blot analysis: Western blot analysis was performed to determine the expression levels of various fusion proteins. The cell lyses were subjected to sodium

dodecyl

sulfate-polyacrylamide

gel

electrophoresis

and

then

transferred onto polyvinylidene fluoride (PVDF) membranes. The PVDF membranes were incubated with specific antibodies against the HA or Flag epitopes

of

the

fusion

proteins

(Abcam),

followed

by

horseradish

peroxidase-conjugated goat anti-rabbit or goat anti-mouse IgG (Abcam). The signal was detected with a chemiluminescent detection system (BioRad).

Super-resolution imaging and data analysis: BiFC-PALM and TFFC-PALM imaging were conducted using a Nikon TiE inverted microscope equipped with a 100X, 1.49 NA, oil immersion objective lens (NIKON, A1 MP STORM) and an Andor-897 EMCCD (Andor). Three lasers with emission wavelengths of 561 nm (Coherent, 50 mW), 488 nm (Coherent, 100 mW) and 405 nm 20


Page 21 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

(Coherent, 100 mW) were available for fluorescence excitation, photoswitching to the on state and green-to-red photoconversion. The mIrisFP channel was collected with continuous 488-nm and 405-nm laser irradiation or continuous 561-nm and 488-nm laser irradiation after photoconversion by a 405-nm laser. The lasers were slowly adjusted for optimal photoswitching rates. Fluorescent spots above the threshold were acquired by subtracting background (the background was defined as the fluorescence intensity of non-fluorescent region in the cell). Localization precision was determined as previously described.40,

41

Super-resolution image reconstruction was performed using

Insight 3 software, and then, the images were reanalyzed by a Gaussian fitting routine. Total internal reflection (TIRF) illumination was used in all PALM imaging experiments.

21


ACS Nano

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. Construction of mIrisFP-based BiFC systems. (a) The cleavage sites (arrows) of mIrisFP are indicated. (b) Green and red mIrisFP BiFC signals were detected in Vero cells for the cleavage sites between aa 150/151 and 165/166 for mIrisFP fragments based on the interactions between bJun and bFos or mbFos. The green form of mIrisFP was generated with 488-nm laser excitation, and the red form of mIrisFP was generated by 405-nm laser irradiation for 30 s prior to image acquisition with 561-nm laser excitation. (c) Quantitative analysis of the BiFC efficiency in (b). (d) Comparable expression of the fusion proteins in (b) determined by western blotting with anti-HA and anti-Flag antibodies. (e) No mIrisFP BiFC signal was detected when mIrisFP 22


Page 22 of 36

Page 23 of 36

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

was split into fragments at the aa sites 134/135 and 157/158. The nucleus was stained with Hoechst 33342. Scale bars: 10 µm. All data are given as the mean ± S.D. (n = 60). The statistical significance was evaluated using a two-tailed Student’s t-test. *** indicates p

Three-Fragment Fluorescence Complementation Coupled with

Recommend Documents