Rational design of true monomeric and bright photoconvertible

Transcription

1 Rational design of true monomeric and bright photoconvertible fluorescent proteins Mingshu Zhang, Hao Chang, Yongdeng Zhang, Junwei Yu, Lijie Wu, Wei Ji, Juanjuan Chen, Bei Liu, Jingze Lu, Yingfang Liu, Junlong Zhang, Pingyong Xu & Tao Xu Supplementary Figures and Tables: Supplementary Figure 1 meos2 but not meos3s tends to form dimer and higher order oligomers. Supplementary Figure 2 Localization of GRM4 and GLUT4 fused with different EosFP variants. Supplementary Figure 3 Sequence alignment of meos2, meos3s, and other monomeric green-to-red PAFPs (Kaede, Dendra2, mkikgr). Supplementary Figure 4 Gel filtration chromatography of meos2 and its mutants. Supplementary Figure 5 Ultracentrifuge analysis of meos2 and its mutants. Supplementary Figure 6 Gel filtration chromatography of EosFP variants and mclavgr2. Supplementary Figure 7 Ultracentrifuge analysis of EosFP variants and mclavgr2. Supplementary Figure 8 Localization of meos3.1 and meos3.2 fusion proteins. Supplementary Figure 9 Excitation and emission spectra of meos2, meos3.1 and meod3.2. Supplementary Figure 10 ph-dependence of meos2, meos3.1 and meos3.2. Supplementary Figure 11 Photoconversion kinetics of meos2, meos3.1 and meos3.2. Supplementary Figure 12 Maturation kinetics of EosFP variants and mclavgr2. Supplementary Figure 13 Expression and folding of meos2, meos3s and tdeosfp when fused with poorly folded polypeptides. Supplementary Figure 14 Photobleaching kinetics of meos2, meos3.1 and meos3.2 at the same laser power. Supplementary Figure 15 Tracking mitochondria dynamics in live HeLa cells. Supplementary Figure 16 Comparing the performance of meos2, meos3.1 and meos3.2 in PALM imaging. Supplementary Figure 17 Observed blinking events of purified meos2, meos3.1 and meos3.2. Supplementary Table 1 The crystallographic summary. Supplementary Table 2 Characteristics of selected meos2 mutations in green form. Supplementary Table 3 Characteristics of EosFP variants and mclavgr2. Supplementary Table 4 Label densities of EosFP variants and mclavgr2. Supplementary Table 5 Performance comparison of EosFP variants and mclavgr2 in super-resolution microscopy. Supplementary Table 6 Primers used in this study. Supplementary Table 7 Parameters used in the data analysis of sedimentation equilibrium experiments. 1

2 Supplementary Figure 1: meos2 but not meos3s tends to form dimer and higher order oligomers. Sedimentation velocity analytical ultracentrifugation of meos2 (a), meos3.1 (b) and meos3.2 (c) at 0.02 mm, 0.05 mm, 0.23 mm and 0.29 mm, respectively. The theoretical molecular weight of monomer meos2, meos3.1 and meos3.2 with 6 His tag is 30 kd. When the protein concentration increases, meos2 tends to form dimer and higher order oligomers, whereas meos3.1 and meos3.2 remain monomer at the highest concentration tested. 2

3 Supplementary Figure 2: Localization of GRM4 and GLUT4 fused with different EosFP variants. (a-d) COS-7 cells expressing metabotropic glutamate receptor 4 (also known as GRM4) fused with meos2 (a), meosfpthermo (b), meos3.1 (c) and meos3.2 (d) were imaged using confocal microscope near the bottom of the membrane. (e-h) GLUT4 (glucose transporter type 4) were labeled with meos2 (e), meosfpthermo (f), meos3.1 (g) and meos3.2 (h) and transfected into L6 cells. meos2 and meosfpthermo caused significant aggregates for GRM4 (ref. 1) and GLUT4, whereas meos3.1 and meos3.2 did not affect the proper subcellular localization of GRM4 and GLUT4, which is mainly dispersed in the plasma membrane for GRM4 and in intracellular vesicles and storage compartments for GLUT4 (ref. 2). 3

4 Supplementary Figure 3: Sequence alignment of meos2 (ref. 3), meos3s, and other green-to-red PAFPs (Kaede 4, Dendra2 (ref. 5), mkikgr 6 ). Mutations from meos2 to meos2-na are marked by a red frame, mutations to enhance fluorescence are marked by a blue frame, the Tyr121 mutation site are marked by a light blue frame, and the chromophores are marked by a green frame. 4

5 Normalized Absorbence (mau) Supplementary Figure 4: Gel filtration chromatography of meos2 and its mutants meos2 meos2-i102n meos2-y189a meos2-na Elute Volume (ml) Overlaid size exclusion chromatography of meos2, meos2-i102n, meos2-y189a, and meos2-na (approximate 3-4 mg/ml) in PBS using a Superdex /300 GL column. The elute volume of meos2-i102n or meos2-na is approximate 16.7 ml, and the elute volume of meos2 is approximate 15.9 ml, together with the much wider peak of meos2, indicating that meos2-i102n and meos2-na have better monomeric characteristic. meos2-y189a has the similar gel filtration properties with meos2. 5

6 Supplementary Figure 5: Ultracentrifuge analysis of meos2 and its mutants. Sedimentation equilibrium analytical ultracentrifugation of meos2 (a), meos2-i102n (b), meos2-y189a (c) and meos2-na (d). Data sets (blue dots) were fitted globally to the monomeric (red line) and dimeric (green line) molecular weights (lower panels). The residual differences between the predicted distribution and the data are shown in upper panels. (b) and (d) don t have a dimeric fit curve because they had a very high K d that beyond the limit of our instrument. 6

7 Normalized Absorbence (mau) Supplementary Figure 6: Gel filtration chromatography of EosFP variants and mclavgr meos2 meos3.1 meos3.2 meosfp meosfpthermo mclavgr Elute Volume (ml) Overlaid size exclusion chromatography of meos2, meos3.1, meos3.2, meosfp, meosfpthermo and mclavgr2 (approximate 3-4 mg/ml) in PBS using a Superdex /300 GL column. The elute volume of meos3.1 and meos3.2 are approximate 16.7 ml, and the elute volume of meos2 is approximate 15.9 ml, indicating that meos3.1 and meos3.2 have better monomeric characteristic. 7

8 Supplementary Figure 7: Ultracentrifuge analysis of EosFP variants and mclavgr2. Sedimentation equilibrium analytical ultracentrifugation of meos3.1 (a), meos3.2 (b), mclavgr2 (c), meosfp (d), and meosfpthermo (e). Data sets (blue dots) were fitted globally to the monomeric (a-c, red line) and dimeric (d and e, green line) molecular weights. The residual differences between the predicted distribution and data are shown in upper panels. 8

9 Supplementary Figure 8: Localization of meos3.1 and meos3.2 fusion proteins. HeLa cells expressing meos3.1-β-actin, meos3.1-α-actinin, meos3.1-mito, meos3.1-src-n15, meos3.1-h2b, and meos3.1-tfr (left panels), meos3.2-β-actin, meos3.2-α-actinin, meos3.2-mito, meos3.2-src-n15, meos3.2-h2b, and meos3.2-tfr (right panels) show no disturbed localization. Scale bars: 10 μm. 9

10 Normalized Fluorescence Normalized Fluorescence Supplementary Figure 9: Excitation and emission spectra of meos2, meos3.1 and meos3.2. a b green state meos2 meos3.1 meos red state meos2 meos3.1 meos Wavelength (nm) Wavelength (nm) (a) The excitation and fluorescence emission spectral profiles of green species meos2, meos3.1 and meos3.2 display similar maxima at around 506 nm and 519 nm, respectively. (b) The red species of meos2, meos3.1 and meos3.2 also exhibit identical spectral profiles that have maxima at around 573 nm (excitation) and 584 nm (emission), respectively. 10

11 Normalized fluorescent intensity Normalized fluorescent intensity Normalized fluorescent intensity Normalized fluorescent intensity Normalized fluorescent intensity Normalized fluorescent intensity Supplementary Figure 10: ph-dependence of meos2, meos3.1 and meos a b c meos2 Green form Fitting curve 1.0 meos3.1 Green form Fitting curve 1.0 meos3.2 Green form Fitting curve pka= ph value pka= ph value d e f meos2 Red form Fitting curve meos3.1 Red form Fitting curve pka= ph value meos3.2 Red form Fitting curve 0.5 pka= ph value 0.5 pka= ph value 0.5 pka= ph value Characterization of the ph dependence of green and red states of meos2, meos3.1 and meos3.2. Plot of fluorescent intensity at 525 nm for green state meos2 (a), meos3.1 (b), meos3.2 (c) and 590 nm for red state meos2 (d), meos3.1 (e), meos3.2 (f) as a function of ph ranging from

12 k on (s -1 ) Supplementary Figure 11: Photoconversion kinetics of meos2, meos3.1 and meos meos2 meos3.1 meos3.2 Linear fit of meos2 Linear fit of meos3.1 Linear fit of meos nm intensity (W/cm 2 ) The on rate k on of meos2, meos3.1 and meos3.2 are positive correlated with the activation intensity in the range of W/cm 2. The slope of meos2 (0.068 s -1 W -1 cm 2 ) and meos3.2 (0.069 s -1 W -1 cm 2 ) are similar, whereas meos3.1 (0.073 s -1 W -1 cm 2 ) is slightly more sensitive to 405-nm laser. n = 4 for each group of cells. Error bars represent s.e.m. 12

13 Supplementary Figure 12: Maturation kinetics of EosFP variants and mclavgr2. Maturation of meos2, meos3.1, meos3.2, meosfpthermo, mclavgr2 at 37 C, and meosfp at 28 C. Maturation half-time comparison (a) of various PAFPs. (b-g) The maturation profiles of meos2 (b), meos3.1 (c), meos3.2 (d), meosfpthermo (e), mclavgr2 (f), meosfp (g) and tdeosfp (h) can be fitted by a single exponential function with time constants of 30 (n = 8), 23 (n = 10), 20 (n = 15), 39 (n =4), 79 (n = 3), 96 (n = 4 ) and 35 (n = 4) min, respectively. Error bars represent s.e.m. 13

14 Supplementary Figure 13: Expression and folding of meos2, meos3s and tdeosfp when fused with poorly folded polypeptides. It has been reported 7 that FPs with good maturation properties can folds well even when fused to poorly folded polypeptides including partially soluble proteins. To compare meos3s with meos2 and tdeosfp of their expression and maturation properties, we expressed meos3.1, meos3.2, meos2 and tdeosfp as C-terminal fusions with human HCCR1(NM_ , aa ), an insoluble protein when expressed in Escherichia coli, which is mainly found in inclusion body 8. (a) SDS-PAGE of non-induced strain (N) and induced strain (I) expressing HCCR1-EosFPs. HCCR1 fused with meos3.1 and meos3.2 has much higher whole-cell expression level than HCCR1-mEos2, whereas HCCR1-tdEosFP has the lowest. Black arrow indicates the molecular weight of HCCR1-mEos2, HCCR1-mEos3.1 and HCCR1-mEos3.2; red arrow indicates the molecular weight of HCCR1-tdEosFP. The bands of different sizes were used as loading control. (b) Fluorescence histograms (number of cells having indicated fluorescence) obtained by flow cytometry of liquid cultures of E. coli BL21(DE3) expressing HCCR1-EosFPs. In consistent with (a), HCCR1 fused with meos3.1 and meos3.2 has much higher fluorescence as compared to HCCR1 fused with meos2 and tdeosfp. Please note that meos3.1 has the highest brightness than the other three EosFPs. 14

15 Nomalized fluorescent intensity in green state Nomalized fluorescent intensity in red state Supplementary Figure 14: Photobleaching kinetics of meos2, meos3.1 and meos3.2 at the same laser power. 1.0 a meos2 meos3.1 meos b Time (s) Time (s) Photobleaching curves of the green state (a) and the red state (b) of meos2 (black lines), meos3.1 (blue lines) and meos3.2 (red lines) under confocal condition using the same laser power (488 nm laser: 24 μw; 543 nm laser: 64 μw). Averaged decay half-time of the green state fluorescence intensity is 14 s for meos2, 6 s for meos3.1, and s for meos3.2. Averaged decay half-time of the red state fluorescence intensity is 48 s for meos2, 36.7 s for meos3.1 and 48 s for meos3.2. n = 10 for each group of cells. Error bars represent s.e.m. 15

16 Supplementary Figure 15: Tracking mitochondria dynamics in live HeLa cells. Mitochondria in live HeLa cells was marked by meos3.2-mito. After illuminated by 405 nm laser for 30 s in the rectangular area, meos3.2-mito was distinctly photoconverted into the red form, which could be used for long term scanning. Scale bars: 10 μm left and 5 μm right. 16

17 Supplementary Figure 16: Comparing the performance of meos2, meos3.1 and meos3.2 in PALM imaging. PALM and TIRF (lower right corner) images of BS-C-1 cells expressing β-actin fused with meos3.1 (a) and HeLa cells expressing α-actinin fused with meos3.2 (b). Magnified images (c, e) of the rectangular regions in (a, b) display an enhanced resolution over TIRF images (d, f). (g) Distribution of total photon numbers per burst in HeLa cells expressing β-actin fused with meos2 (n = 6), meos3.1 (n = 8), and meos3.2 (n = 8). (h) Distribution of localization errors in HeLa cells expressing β-actin fused with meos2 (n = 6), meos3.1 (n = 8) and meos3.2 (n = 8). Scale bars are 2 μm (a, b) and 500 nm (c, d, e, f). 17

18 Supplementary Figure 17: Observed blinking events of purified meos2, meos3.1 and meos3.2. (a) Observed blinking distribution for purified meos2 (n = 9,457), meos3.1 (n = 15,841) and meos3.2 (n = 15,939) proteins immobilized on a coverslip. 49% of meos2, 53% of meos3.1 and 55% of meos3.2 do not have a second localization within 100 nm. (b) Distribution of measured on-time as well as the 1/e decay value of off-time obtained by single exponential fitting (inset). Error bars represent s.e.m. 18

19 Supplementary Table 1: The crystallographic summary. Data collection statistic Space group P a, Å 72.4 b, Å c, Å Resolution Å 2.2 Total observations 508,410 Unique reflections 46,553 Completeness, % 99.5 Rmerge* 0.161(0.531) I/ Ϭ(I)* 22.3(8.1) Refinement Rcryst Rfree No. of protein atoms 7,149 No. of waters 525 rms deviation from ideality Bond lengths, Å Bond angles, Average B factors, Å 2 All atoms Chromophore 9.37 rmsd bonded B, Å * values in parentheses are for the highest resolution shell. 19

20 Supplementary Table 2: Characteristics of selected meos2 mutants in green form. Fluorescent EC MW protein (M -1 cm -1 QY Brightness ) (KD) I102N 36, , / Y121R 23, , / Y121K 16, , ± 31 M Y189A 67, , ± 5 M meos2-na 38, , / K d EC: extinction coefficient; QY: quantum yield. Brightness is calculated as extinction coefficient quantum yield/1,000. MW: molecular weight in sedimentation equilibrium state. K d was measured by nonlinear least-squares fitting using the software package (Microcal Origin) supplied by Beckman and is presented as fitted value ± standard deviation. / represents a very high K d that was beyond the limits of our instrument. 20

21 Supplementary Table 3: Characteristics of EosFP variants and mclavgr2. Fluorescent EX EM QY EC Brightness pka t 1/2 MW Oligomerizati Ref. protein (nm) (nm) (M -1 cm -1 ) (kda) on [K d ] meos2 (G) , (0. 41, meos2 (R) ) (46,000) (30) (6.4) 42.8 Weak dimer 20 ± 3 µm 3 meos3.1 (G) , meos3.1 (R) , Monomer / this work meos3.2 (G) , meos3.2 (R) , Monomer / this work tdeosfp (G) (506) (516) (0.91) (65,000) (59) (5.7) - tdeosfp (R) (569) (581) (0.62) (48,000) (30) Tandem Dimer 3 meosfp (G) meosfp (R) (505) (516) (0.70) (72,000) (50) (5.7) - (571) (581) (0.55) (41,000) (23) Weak dimer 233 ± 53 µm 3 meosfptherm o (G) meosfptherm o (R) , , Weak dimer 483 ± 77 µm this work mclavgr2 (G) mclavgr2 (R) (488) (504) (0.77) (19,000) (15) (8.0) - (566) (583) (0.53) (32,000) (17) (7.3) Monomer / 9 EX: maximum excitation; EM: maximum emission; QY: quantum yield; EC: extinction coefficient. Brightness is calculated as extinction coefficient quantum yield/1,000. t 1/2 : photobleaching half-time; MW: molecular weight measured by sedimentation equilibrium. K d was measured by nonlinear least-squares fitting using the software package (Microcal Origin) supplied by Beckman and is presented as fitted value ± standard deviation. / represents a very high K d that was beyond the limits of our instrument. Values in parentheses represent reported values taken from references. indicates not measured. 21

22 Supplementary Table 4: Label densities of EosFP variants and mclavgr2. Fluorophore Np, typical Total Label density obtained (µm -2 ) detected molecule photons per number Mean Median Max burst meosfp (30 C) 419, ,680 meosfpthermo 2,289, ,286 3,125 11,564 meos2 1,653, ,649 2,059 20,843 mclavgr2 2,624, ,774 2,795 29,253 tdeosfp * 8,636, ,945 4,407 30,615 meos3.1 5,348, ,110 5,577 44,933 meos3.2 7,310, ,865 5,640 59,495 * We divided the label density of tdeosfp by a factor of two to reflect the fact that two copies of the EosFP label one copy of the partner. These numbers are provided for a rough comparison of the probes maximum label density only, not as an absolute limit on probe density, which varies with construct, transfection conditions, growing conditions, cell line, cell and even regional of the cell 3. 22

23 Supplementary Table 5: Performance comparison of EosFP variants and mclavgr2 in super-resolution microscopy. Oligomerization Maturation ph Stability Total photons per burst after switching Label Density meosfp Weak dimer, may not be suitable for labeling membrane proteins Slow (at 28 C) ph stable Medium (371) Very low at 30 C Weak dimer, may not be meosfpthermo tdeosfp suitable for labeling membrane proteins Tandem dimer, causes incorrect localization when fused with many standard targets (i.e., tubulin, histones) Medium Not measured Medium (374) Low N/A ph stable High (499) Medium mclavgr2 Monomer Slow 50% quenched at ph 7.2 Low (334) Low Weak dimer, not suitable meos2 for labeling membrane proteins Medium ph stable Medium (379) Low meos3.1 Monomer Fast ph stable High (451) High meos3.2 Monomer Fast ph stable High (473) High Green means favorable properties. It is seen that meosfp cannot be used at 37 C and gives a very low label density at 30 C. Although this has been corrected in meosfpthermo, meosfpthermo is not good at photon budget and has a low performance in label density. In addition, meosfp and meosfpthermo can form weak dimmer, which may cause aggregates for membrane proteins (see Supplementary Fig. 2). tdeosfp gives the highest photon budget, however it causes incorrect localization for many standard targets (i.e., tubulin, histones, intermediate filaments and gap junctions) 3. mclavgr2 is slow in maturation, partially quenched at neutral ph, has low photon budget and label density. meos2 is not suitable for labeling membrane proteins and has a low label density. meos3.1 and meos3.2 are the only probes that are green in all five properties and give the best overall performance. 23

24 Supplementary Table 6: Primers used in this study. Primer A69V-F A69V-R Y189A-F Y189A-R I102N-F I102N-R I157I/V&H158D/E-F I157I/V&H158K/R-F Before157-R Eos-pRseta-F Eos-pRseta-R mclavgr2-prseta-f mclavgr2-prseta-r tdeos-prseta-f tdeos-prseta-r Eos-N1-F Eos-N1-R Eos-C1-F Eos-C1-R mclavgr2-n1-f mclavgr2-n1-r Src-F1 Src-F2 Src-F3 Src-R Lifeact-F Lifeact-R H2B-nhe1-N1-F H2B-xho1-N1-R Actinin-nhe1-N1-F Actinin-xho1-N1-R Sequence GCAACAGGGTATTCGTCGAATATCCAGACCACATACAAGACTATTT GAATACCCTGTTGCCGTAATGG GTCAAGTTACCAGGCGCCCACTTTGTGGACCACTGCATTGAG GCCTGGTAACTTGACACCCTTCTC AGACGGGGGCATTTGCAACGCCAGAAACGACATAACAATGGA GCAAATGCCCCCGTCTTC GAGTGCTGACGGGTGATRTAGASATGGCTTTGTTGCTTGAAGGAAA GAGTGCTGACGGGTGATRTAARGATGGCTTTGTTGCTTGAAGGAA A ATCACCCGTCAGCACTCCATCAC ATATGGATCCATGAGTGCGATTAAGCCAGAC ATATGAATTCTTATCGTCTGGCATTGTCAGGCAA ATATGGATCCATGGTGAGCAAGGGCGAGG ATATGAATTCTTACTTGTACAGCTCGTCCATG AGCACAGTGGCGGCCGGGATCCAGTGCGATTAAGCCAG GTGACACTATAGAATAGAATTCTTATCGTCTGGCATTG ATATGGATCCAAGTGCGATTAAGCCAGACATG ATATGCGGCCGCTTATCGTCTGGCATTGTCAGGCAA CTAGGCTAGCCACCATGAGTGCGATTAAGCCAGACATG CCGAGATCTTCGTCTGGCATTGTCAGGCA ATATGGATCCAGTGAGCAAGGGCGAGGAGAC ATATGCGGCCGCTTACTTGTACAGCTCGTCCATG CGGGAATTCCCACCATGGGATCCTCAAAGAGTAAGCCTAAGGACC CAAGCC AAGAGTAAGCCTAAGGACCCAAGCCAGAGACGGGGAGGTACCGG AGGAAG CGGGGAGGTACCGGAGGAAGTGCGATTAAGCCAGACATGAAGATC CTAGGCGGCCGCTTTATCGTCTGGCATTGTCAGGC AATTCCCACCATGGGTGTCGCAGATTTGATCAAGAAATTCGAAAGC ATCTCAAAGGAAGAACGG GATCCCGTTCTTCCTTTGAGATGCTTTCGAATTTCTTGATCAAATCT GCGACACCCATGGTGGG CTAGGCTAGCCACCATGCCTGAACCGGCAAAATC CTAGCTCGAGCTTGGAGCTGGTGTACTTGGTGAC CTAGGCTAGCCACCATGGACCATTATGATTCTCAGCAAAC CTAGCTCGAGGAGGTCACTCTCGCCGTACAGC 24

25 Supplementary Table 7: Parameters used in the data analysis of sedimentation equilibrium experiments. Mutant Solvent density Partial specific volume Molecular weight (Da) meos ,972.9 I102N ,973.7 Y189A ,880.7 meos2-na ,881.6 meos ,859.7 meos ,873.7 meosfp ,956.6 meosfpthermo ,984.7 mclavgr ,

26 References: 1. Iacovelli, L. et al. Mol. Pharmacol. 65, (2004). 2. Bai, L. et al. Cell Metab. 5, (2007). 3. McKinney, S. A., Murphy, C. S., Hazelwood, K. L., Davidson, M. W. & Looger, L. L. Nat. Methods 6, (2009). 4. Mizuno, H. et al. Mol Cell 12, (2003). 5. Gurskaya, N. G. et al. Nat Biotechnol 24, (2006). 6. Habuchi, S., Tsutsui, H., Kochaniak, A. B., Miyawaki, A. & van Oijen, A. M. PLoS One 3, e3944 (2008). 7. Pedelacq, J. D., Cabantous, S., Tran, T., Terwilliger, T. C. & Waldo, G. S. Nat. Biotechnol. 24, (2006). 8. Liu, B. C., Huo, J. R., Liu, Y. F., Fan, G. C. & Wang, Q. M. Chin. J. Cell Mol. Immunol. 27, (2011). 9. Hoi, H. et al. J. Mol. Biol. 401, (2010). 26