Description of the cellular system

A genomic locus at which a gene was integrated served as an inducible site of transcription for a 3.4 kb pre-mRNA. Preceding the transgene were 256 lac operator repeats which could be detected with a fluorescently-fused lac repressor protein (CFP, YFP or RFP depending on the color combination needs of the experiment) thus identifying the chromosomal site of integration in living cells. Tetracycline-responsive repeats controlled the transcription from this gene. The mRNA contained a 5’ sequence encoding for cyan fluorescent protein (CFP) bearing a tripeptide peroxisome-targeting sequence, allowing us to monitor the translation of the mRNA via accumulation of the CFP signal in cytoplasmic peroxisomes. For real-time detection of the RNA, we inserted 24 MS2 stem-loops downstream of the open reading frame. These sequences efficiently and specifically bind to a fluorescently-tagged MS2 bacteriophage coat protein. The 3’ end of the transcript consisted of the last intron/exon module of the human beta globin mRNA followed by its terminator. Approximately 200 copies of this gene were stably integrated into a euchromatic site in a human osteosarcoma cell line (U2OS) (1) .

Cells and transfections

Human U2OS osteosarcoma cells containing the integrated gene (clone 2-3-6) were cultured, transfected and transcriptionally activated as previously described (1) . Briefly, cells were transfected by electroporation (Gene Pulser Xcell, BioRad, Hercules, CA) with 2 µg of pTet-On, 2 µg of pSV2-XFP-lac repressor, 40 µg of sheared salmon sperm DNA (Amresco, Solon, OH) and 5 µg of pYFP-MS2-NLS under the control of the ribosomal L30 promoter (Gift from Edouard Bertrand). Cells were plated on coverslips or dishes coated with Cell-Tak (BD Biosciences, Palo Alto, CA). Transcription was induced by the addition of doxycycline (1 µg/ml) to the medium for indicated times. pSV2-YFP-lac-repressor and pSV2-CFP-lac repressor were previously described (1, 2) . pSV2-mRFP-lac repressor was constructed by replacing the coding region of YFP by mRFP1 (3) .

Live cell imaging and single particle tracking

For particle tracking, transcriptionally induced cells were maintained at 37°C in a FCS2 live-cell chamber (Bioptechs Inc., Butler, PA) mounted onto an Olympus BX60 microscope with a 37°C heated Olympus PlanApo 100X 1.4 NA objective lens (Olympus America Inc., Melville, NY). For live cell imaging, cells were maintained in phenol-red free Leibovitz’s L15 medium. Digital images were captured with a 12 bit Imago-QE CCD camera. Exposure times for YFP-MS2 particles were either 333 or 250 msec for 100 frames (2x2 binning) using TILLvisION software with 505 nm illumination from a Polychrome II monochromator (TILL-Photonics, Eugene, OR). The microscope was equipped with filter sets for YFP (41028), CFP/YFP dual band filter provided by TILL-Photonics (part number CFP/YFP+) and CFP custom filter set (composed of Chroma parts D425/40, 455DCLP & HQ460LP) (Chroma Technology Corp., Rockingham, VT).  In experiments at room temperature (22°C), the chamber, lens and medium were not heated. To improve quality, movies were deconvolved using Huygens II Professional with time series option (Scientific Volume Imaging, Hilversum, The Netherlands).
For particle tracking, particles appearing consistently in every frame for more than 8 frames and isolated from other surrounding particles in order to rule out any particle identity confusion in the process of tracking were analyzed in Metamorph using the Motion Analysis and Particle Tracking module (Universal Imaging Corporation, Downingtown, PA). Tracking: >50 particles coming from 10 cells (>1500 displacement steps) were tracked at 37°C; 18 particles from 4 cells at room temperature (>650 displacement steps), 7 transcription sites from 7 cells (>400 displacement steps). The movement was measured per frame and the data was imported into MATLAB (The MathWorks, Natick, MA) for calculations of total distance, instantaneous velocity, plotting of mean square displacement versus time, diffusion coefficients and anomalous diffusion. Diffusion coefficients for two-dimensional Brownian motion of mRNPs were calculated by averaging the square displacements (MSD), as in:
  (4)
<r²>, mean square displacement (MSD) of the mRNP over time (t); All the displacement steps (>1500) from 50 trajectories per time interval were analyzed, and D was equal to the initial slope of the plot MSD versus the time interval (Dt).
The corralled radius for particles exhibiting corralled movement was calculated using:
  (4)
For testing for anomalous diffusion the data was plotted on a logarithmic scale as log(MSD/Dt) versus log(Dt) (5) . In this case of anomalous diffusion, there is an exponential relationship between MSD and Dt which would be presented as a straight line on a logarithmic plot. However, no straight line with a negative slope was observed for tracked particles (not shown). In the case of SPT at room temperature, a reduction in the average diffusion coefficient for diffusive particles tracked at RT was observed, as expected (Fig S3D). As for corralled particles, in our system where the position of a particle is attributed to the center of the pixel in which it is detected, the accuracy of the position where a particle is changing direction is inversely proportional to the number of pixels of the corralled area. When measuring in small areas this could lead to an overestimation of the path traveled. Since in this case there was a 50% reduction in the corralled area at RT (from r=0.9 µm, area=188 pixels at 37°C to r=0.7 µm, area=91 pixels at RT), we consider the change in D of corralled particles in between 37°C and RT to be non significant.
Movies of particles in motion were annotated, using a program written in the laboratory, to aid in the visualization of particle movement over time. The program used the coordinates derived from tracking analysis to generate a glyph that circled the particle in each frame or a line that traced particle movement between frames.
In order to control for detection of different types of movements by SPT, tracking was performed also on cytoplasmic CFP-labeled peroxisomes and characteristics of movements were obtained. These included directed motions which were not found for mRNPs. The movements of CFP-peroxisomes were tracked as with mRNPs and found to be mostly confined oscillatory movements while less than 5% were directed (either uni-directional or bi-directional) with average velocities of up to 1.8 µ/sec (Fig. S3C). These measurements were in correlation to previously described analysis of peroxisome mobility (6, 7) .

In situ hybridization, image analysis and quantification

Cells were transfected with 2 µg of pTet-On, 2 µg of pSV2-YFP-lac repressor and 40 µg of sheared salmon sperm DNA. After the cells adhered to the coverslips, transcription was induced by addition of doxycyline for the indicated times. Cells were fixed and fluorescent hybridization was performed as previously described (8) with a Cy3-conjugated probe against the exonic region of the beta globin module
(5’-CTCATTCTGATGTTTTAAATGATTTGCCCTCCCATATGTCCTTCCGAGTG)
and a Cy5-conjugated probe to the region between the MS2 stem loops (9, 10). Images were acquired with an Olympus BX61 epifluorescence microscope with an internal focus motor and an Olympus UPlanApo 100X 1.35 NA oil objective using a 100 Watt mercury arc lamp for illumination. Digital images were acquired using a Roper Scientific CoolSNAP HQ camera (Roper Scientific, Tucson, AZ) as stacks of 40 images taken with a Z step size of 0.1 µm using IPLab software (Windows v3, Scanalytics, Fairfax, VA) and filter sets 41007 (Cy3) and 41008 (Cy5) (Chroma Technology). For quantification of RNAs all 3-D stacks were first deconvolved using Exhaustive Photon Reassignment (EPR, Scanalytics) which uses a quantitative, constrained-iterative algorithm (11) with an acquired point spread function.
Single molecule mRNA identification and quantification was performed as previously described (12, 13) using a probe which has only one binding site on the transcript and therefore can serve for quantification. The number of probes per particle was determined by calculating the total fluorescent intensity (TFI) of individual RNAs, and dividing that value by the TFI per probe as follows: TFI per probe was calculated for the Cy3-exon probe by collecting images from serial probe dilutions. 5 µl of each probe dilution (ranging from 4 ng/µl to 4 x10^-4 ng/µl) were placed between a coverslip and a slide, onto which 170 nm blue fluorescent beads had previously been dried. Using the beads as markers, the distance between the coverslip and slide was measured (in microns) and the center plane of the dilution was located. A range of interest (256 X 256 pixels) that excluded beads was identified, and a single image was captured at the center plane, using exposure time identical to that used to capture cell images (1000ms). This procedure was repeated three times, each at a different location on the coverslip, for each dilution. The TFI per probe was then obtained by plotting the integrated fluorescence in the total imaged volume against the known number of molecules in that volume (Fig. S2E). The slope of the resulting curve represents the TFI per one fluorescent probe molecule and is further used in the calculations. In parallel, TFI per 3D RNA particle in the FISH images was calculated from deconvolved images with a script written for IPLab. To compensate for the deconvolution, this value was then divided by the number of planes in the point spread function (PSF) used by EPR. Each RNA now had an assigned TFI which was then divided by the TFI/probe molecule calculated from the dilution curve, resulting in the number of RNA molecules in an imaged particle. Particles were selected for analysis after thresholding which was determined by comparing experimental and control cells, where control cells were not transfected with pTet-On. Analysis of 6707 nuclear transcripts from 21 doxycycline-induced cells showed that the majority (~70%) of transcripts hybridized with a single probe indicating that these nuclear particles represent single RNA transcripts and not packets.
The proximity of mRNA transcripts, as detected by FISH, with YFP-MS2 particles was measured using a program written in the laboratory. The result was a histogram of distances between transcripts and particles where a proximity of zero indicated colocalization. The program identified transcripts and particles from the acquired three dimensional images, which were deconvolved, using a threshold and routine segmentation algorithms. Every transcript that spatially overlapped a particle was scored as colocalized. The distance between all other transcripts and the closest particle was calculated by the software and reported in the histogram. 
Color map image volumes were generated, using a program written in the lab, to visualize the number and spatial position of clusters of mRNA. The program identified mRNA clusters from acquired three dimensional images, which were deconvolved, using a threshold and routine segmentation algorithms. The total intensity of each object was calculated and was used to assign a color code for each mRNA cluster that corresponded to the number of mRNA contained in the cluster.

FRAP

Cells were transfected with 2 µg of pTet-On, 2 µg of pSV2-RFP-lac repressor, 40 µg of sheared salmon sperm DNA and 5 µg of pYFP-MS2-NLS and plated on 0.17 mm Delta T dishes (Bioptechs). In experiments in which transcription was induced, doxycyline was added for 30 min. For FRAP of YFP-MS2 and pEYFP-nuc (Clontech - BD Biosciences, Palo Alto, CA) no pTet-On was transfected. Images were collected on a Leica TCS SP2 AOBS laser scanning confocal microscope equipped with a 63X, 1.4 NA objective (Leica Microsystems Inc, Exton, PA). Experiments were performed at 37°C, 32°C, 27°C and 22°C using a temperature controlled Delta T4 culture dish system with a heated lid and an objective heater (Bioptechs). Cells were scanned using a 514 nm laser for the detection of YFP-MS2 or a 543 nm laser for detection of RFP-lac repressor protein at the locus.
For each time point, the background taken from a ROI outside of the cell was subtracted from all other measurements. T(t) and I(t) where measured for each time point as the average intensity of the nucleus and the average intensity in the bleached ROI, respectively. One image was collected prior to bleaching and these initial conditions are referred to as T_i = nuclear intensity and I_i = intensity in ROI before bleaching. Ic(t) is the corrected intensity of the bleached ROI at time t (14) :
 
Data from 10 to 15 cells was collected over 3 independent experiments and diffusion coefficients were calculated as previously described (14, 15) . For each different bleach settings, the constants w and K were determined on fixed cells. w is the half width of the laser beam and K is a parameter describing the amount of bleaching.

r is the radius of the bleached ROI and C(r) is the intensity of fluorescence along this radius after bleaching. w and K where determined simultaneously using a non-linear regression in Mathematica (Wolfram Research, Champaign, IL). For imaging of the free YFP-MS2, the laser beam expander was set to 1 and the frame rate was 0.274 seconds per frame. w and K were fitted to 1.13±0.02 and 7±1.4 respectively, using a fixed specimen. For imaging of the mRNPs, the beam expander was set to 3 and the frame rate was 1.635 sec/frame; w=1.39±0.03 and K=3.1±0.08.
The FRAP experiments (15 time-series from different experiments) were then fitted for g and D using a non-linear regression in Mathematica.

D is the diffusion coefficient expressed in µm²/s and g is the fraction of immobile signal. The time series were then averaged and error was calculated.
When mRNPs were bleached, the settings of the laser and the detection threshold were set so that the diffuse MS2-YFP signal was below the detection level allowing for the specific detection of the bright particles. No simultaneous detection and quantification of both free YFP-MS2 and mRNPs could be performed since settings allowing for the detection of free MS2-YFP led to partial saturation of the signal of the mRNP particles.
For the experiment in which H2B-YFP was photobleached, cells transfected with H2B-YFP were bleached in several regions forming a grid pattern and the H2B-YFP structure was followed over time. This was performed for cells with and without ATP depletion treatment.

Photoactivation

Oligo 1: ATAGGATCCACCATGCCAAAAAAG
Oligo 2: CCTCGCCCTTGCTCACCATAGCGGCCGCGTAGATGCCGGAGTTTGC
Oligo 3: GCAAACTCCGGCATCTACGCGGCCGCTATGGTGAGCAAGGGCGAGG
Oligo 4: ATACTCGAGTTTACTTGTACAGCTCGTCC

The MS2 coat protein cDNA was PCR amplified from the MS2-GFP construct (9) with oligos 1 and 2 using Tli DNA polymerase (Promega, Madison, WI) to give DNA1. The photoactivable GFP cDNA was amplified from pPA-GFP N1 (16) using oligos 3 and 4 giving DNA2. DNA1 and 2 were mixed and used as a PCR template for oligos 1 and 4. The PCR fragment obtained was digested with BamH1 and Xho1 and ligated into pcDNA3.1 (Invitrogen, Carlsbad, CA) giving rise to pMS2-paGFP1. Human U2OS osteosarcoma cells containing the integrated gene (clone 2-3-6) were stably transfected with pMS2-paGFP1 and grown under selection of 100 µg/ml hygromycin B (Sigma) and 1.5 mg/ml geneticin (Gibco, Grand Island, NY). Clone 10, which gave a reasonable expression level of the photoactivable protein, was chosen and is referred to as “clone 2-3-6-PA”.
Clone 2-3-6-PA was transiently co-transfected as described above with pTet-On and RFP-lac repressor. Cells were imaged at room temperature on a Leica TCS SP2 AOBS laser scanning confocal microscope (beam expander 1) using dual excitation 488 and 543 nm to simultaneously follow the paGFP and the RFP signals. Activation of GFP fluorescence at the transcription site was performed using one full power pulse of a 405 nm laser for 1.635 seconds (beam expander 6). The fluorescent signal was then followed in a time series imaged every 1.635 sec.
On each post-activation frame, concentric arcs centered at the transcription site were quantified for their averaged intensity per pixel. This measurement allowed monitoring of the diffusion of mRNPs released from the transcription site over time. Values were corrected for the background measured on the preactivated image and the relative intensity was plotted over the distance from the transcription site. Bleaching of paGFP during imaging was not detectable as concluded from several control experiments. Since the UV activation pulse was directed at an area larger than the transcription site itself, the first flow of molecules reaching a defined position relative to the source, contained in addition to the mRNPs released from the gene, the free mRNPs that surrounded the transcription site at the time of activation. We used this to measure the speed of propagation of the peak which can be directly linked to the diffusion coefficient.

D was calculated from 10 different combinations of time points and the average with standard error is displayed.

Measurements of diffusion coefficients at different temperatures

We performed FRAP at 4 different temperatures (22°C, 27°C, 32°C and 37°C) on free YFP-MS2 and mRNPs. The different diffusion coefficients where plotted as function of the temperature.
The viscosity of the medium (h) can be expressed in terms of the diffusion constant (D), the Stokes radius (r) and the temperature (T), following the Stokes-Einstein equation:

The relative viscosity (h_t1/ h_t2) between two different temperatures T₁ and T₂, was calculated using the diffusion values from FRAP experiments performed at 37°C and 22°C and the following simplification of the Stokes-Einstein equation.


Inhibitors of physiological processes

A variety of inhibitors were added to transfected cells in order to test for changes in mRNP mobility, but had no effect: ribavirin (100 µM; decreases GTP pool), cycloheximide (75 µg/ml; inhibits protein synthesis), geldanamycin (2 µM; inhibits src-kinase and HSP-90), leptomycin B (10 nM; inhibitor of crm1 export), cytochalasin D (5 µg/ml; inhibitor of actin polymerization), latrunculin B (12 µM; inhibitor of actin polymerization), BTS (20 µM; inhibitor of skeletal myosin S1), ML-7 (1 µM; specific inhibitor of myosin light chain kinase (MLCK)), H-7 (300 µM; MLCK and nonspecific protein kinase inhibitor), monastrol (inhibitor of kinesin Eg5), calcium ionophore A323187 (10 µM). DMSO or ethanol (used to solubilize some of the inhibitors), 2 mM D-glucose had no effect on mobility.

ATP depletion 

Energy depletion was obtained by addition of 6 mM 2-deoxyglucose (2-DG; Sigma) and 10 mM Na-azide (Sigma) to the medium. Washing out the inhibitors was performed by perfusion of medium without inhibitors. Reduction in ATP levels was observed as the loss of mitochondrial staining with Rhodamine-123 (Sigma, St. Louis, MO). In addition, intracellular ATP levels were monitored in U2OS cells undergoing the same ATP depletion treatments using a Bioluminescent Somatic Cell Assay Kit (Sigma). ATP content of U2OS cells was found to be ~3*10^-15 moles/cell (somatic cells contain approximately 2*10^-15 moles ATP/cell). ATP levels dropped to 46% of initial values after 10 min incubation with 2-DG and azide and continued to drop with time. In addition, the fluorescent signal of both YFP-NLS and YFP-MS2 was found to increase in the cytoplasm following ATP depletion at 37°C or incubation of cells at at 4°C for 1 hr, indicating a block in the energy-dependent import pathway of NLS containing proteins from cytoplasm to nucleus. ATP depletion did not cause the induction of apoptosis as assayed using an annexin V detection kit (United States Biological, Swampscott, MA). For live cell imaging experiments, cells were imaged for 30 frames 333 msec exposures before and after energy depletion. Interestingly, changes in mRNP mobility were caused by treatment with 2-DG alone and not by azide alone, in agreement with the finding that the nucleus is the cellular compartment most susceptible to ATP depletion by glycolytic inhibition, showing a 80% drop in ATP levels (17) . The mobility of stalled mRNPs after ATP depletion was restored once the inhibitors were washed out.
In experiments using Triton X-100 on living cells, the detergent (0.02% in the medium) was perfused into the heated chamber and mRNPs were followed on either a widefield microscope or confocal microscope as described above. Mean intensities in the nucleus were then calculated for each time point.
For labeling of chromatin in living cells, cells were incubated with Hoechst 33342 (2.5 µM) for several minutes before ATP depletion. In these experiments cells were transfected with: pSV2-CFP-SC35, pSV2-CFP-Sp100, dsRed-coilin (gift from E. Bertrand) or H2B-YFP (gift from J. Swedlow).

Changes in nuclear volume and immunofluorescence for PLA₂

To verify that the change in nuclear size observed upon 2-DG treatment was due to nuclear shrinkage and not to change in the 3-dimensional nuclear shape, cells were transfected with H2B-YFP and 75 planes spanning the whole nucleus were collected on a Leica TCS SP2 AOBS laser scanning confocal microscope. Z stacks were collected for the same cells before and after 2-DG treatment. Nuclear volumes were reconstructed and measured from the Z stacks using Imaris 4 (Bitplane, Zurich, Switzerland). These reconstructions showed that 2-DG treatment caused global nuclear shrinkage.
Since nuclear translocation of phospholipase A₂ (PLA₂) has been implicated in nuclear shrinkage during hypoxia (18) and in ATP depletion (19) , we tested whether PLA₂ moves to the nucleus immediately after ATP depletion. Two forms of PLA₂ were tested: cytoplasmic PLA (cPLA₂) and Ca⁺² independent PLA (iPLA₂). Cells were treated with 2-DG and Na-azide for 0, 1, 10, 30, 60 and 120 min and then fixed on ice in 4% paraformaldehyde (plus metabolic inhibitors) for 20 min followed by 0.5% Triton X-100 for 3 min. After washes in PBS and blocking in 5% BSA (Sigma) cells were labeled with either rabbit polyclonal anti-cPLA₂ (N-216) (Santa Cruz Biotechnology, Santa Cruz, CA) or anti-iPLA₂ (Cayman Chemical, Ann Arbor, MI) for 45 min. Following washes in PBS the cells were incubated with donkey anti-rabbit IgG-FITC F(ab’)₂ (Jackson Immunoresearch Laboratories, West Grove, PA) for 45 min, followed by DNA stain with Hoechst 33342. Although immunofluorescence of cPLA₂ showed nuclear translocation only after 2 hrs post ATP depletion, no change was seen in the sub-cellular distribution of iPLA₂ (Fig. S5, C to J).

Electron microscopy

For TEM of ATP depleted cells, cells were treated with 2-DG and Na-azide for 30 min and then fixed with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer containing the metabolic inhibitors. Samples were dehydrated through a graded series of ethanol and embedded in LX112 resin (LADD Research Industries, Burlington VT).  Ultrathin sections were cut on a Reichert Ultracut UCT. Sections were stained with 4% uranyl acetate in 40% ETOH for 25 minutes, followed by lead citrate for 3 minutes then viewed on a JEOL 100CX transmission electron microscope at 80kv (JEOL USA Inc., Peabody, MA).

Calculation of the mass added by the MS2 proteins

Analysis of the human genome shows an average of 5-12 introns per gene (20) . The exon junction complex (EJC) is composed of at least 7 proteins representing an RNA binding complex of ~400 kDa deposited at each exon-exon junction (21) . This represents a range of at least 2.8 MDa in the mass of an mRNP formed on an mRNA encoded by a gene containing 1 intron and a gene containing 8 introns. The EJC proteins accompany the mRNAs through the export process. In addition, a multitude of hnRNPs and other nuclear factors are part of the mRNP. The 24 MS2 repeats present in the 3’UTR of our reporter gene bind an average of 33 MS2 coat proteins representing a complex of 1.4 MDa (10) . There is one intron in the gene that would add another 409 kDa. The mass of our reporter mRNP bearing 33 YFP-MS2 coat proteins is thus within the range of the naturally occurring mRNP masses in human cells. The 30 kb Balbiani ring mRNA forms a mRNP of 50 nm in diameter and represents an extreme case of a very large mRNA assembled into a compact mRNP (22) .

Theoretical Diffusion coefficients of nuclear mRNPs

Using the Stokes-Einstein equation the predicted diffusion coefficient can be calculated:

K = Boltzman’s constant (1.38*10^-23 J/K or 1.38*10^-23 N*m/K)
T = absolute temperature (273.15 (0°C) + 37°C = 310K)
h = nucleoplasmic viscosity (6.6 cP (23) ; 1cP (centiPoise) = 1/1000 Poiseuille = 1/1000 N*sec/m²)
r =  molecular radius of mRNA transcript. 1 nt = ~ 0.3 nm, so for a 2.8 kb the total length is = 2800 * 0.3 = 840 nm.
If the mRNA travels as a circular unit, with a circumference of 2pr, then r = 133 nm or 13.4*10^-8 m.
D = [(1.38*10^-23 N *m/k)*(310K)]/[(6*3.14)*(6.6/1000 N *sec/m²)*(13.4*10^-8 m)]
D = 2.56*10^-13 m²/sec = 0.25 µm²/sec

However, since the RNA travels as an mRNP complex composed of many types of hnRNPs, the radius of the complex is actually larger. If we estimate four times the radius of a bare RNA (similar to the estimate used by Politz et al. (24) in their evaluation of the predicted coefficients of the total poly(A) population in the nucleus), then the adjusted diffusion coefficient is D = 6.4*10^-14 m²/sec = 0.064 µm²/sec. The average diffusion coefficient (0.04 µm²/sec) obtained by particle tracking is in reasonable agreement with the theoretical D (0.064 µm²/sec) calculated using the Stokes-Einstein equation.

Supplemental Results and Discussion

Observations in living cells

Sequential imaging of cells containing YFP-MS2 labeled mRNPs revealed a relatively immobile transcription site versus released, mobile nucleoplasmic mRNPs. Later time points post-induction (1-3 hrs post dox) revealed dynamic cytoplasmic mRNP particles in contrast to more constrained peroxisomes (Movies 3 and 4). As observed in the movies, mRNPs appeared to move randomly throughout the nucleus after release from the transcription site (Movies 9 and 10). To further emphasize the random diffusion nature of the mRNP movements previous to nuclear export, the fate of particles released from a transcription site in close proximity to the inner nuclear membrane was observed. The mRNPs dispersed throughout the nucleoplasm in this case as well, even though their export would have been facilitated by direct access to the nuclear pores (25) . Furthermore, we did not find any indication for accumulation of mRNPs in intra-nuclear domains for significant periods of time that would be suggestive of the existence of sites for post-transcriptional processing. In a few cases (less than 1% of total detected mRNPs), mRNPs remained relatively immobile throughout most of the imaging period (~40 sec), although their movement could resume at any moment. Pausing might imply functionally significant interactions with nucleoplasmic constituents or hindrance by nuclear structures. In addition, minutes after mild permeabilization of the nuclear membranes with Triton X-100 (0.02%) the nucleoplasm was devoid of all mRNPs while the signal at the transcription site remained longer (Fig. 3, E to G), indicating that stable attachments of the released mRNPs with the environment did not occur. Finally, mRNPs did not penetrate into the nucleolus but rather appeared to “bounce” off it (data not shown).

Segregation of chromatin and nuclear domain during ATP depletion

Restructuring of chromatin regions was found to occur after energy depletion, as mentioned in the text. In order to ascertain the composition of the inter-chromatin domains in relation to known nuclear protein distributions, H2B-YFP was compared to specific nuclear markers for: speckles (CFP-SC35), Cajal bodies (dsRed-coilin), PML bodies (CFP-Sp100) and the gene locus (CFP-lac repressor). Energy depletion demonstrated that the dense chromatin structures were segregated from all of the nucleoplasmic domain markers (Movies 19 and 20 and data not shown). However, concentrated mRNPs colocalized with the speckles (Movies 21 to 23) but not with Cajal bodies or PML bodies. Noteworthy, is that some Cajal bodies exhibited increased mobility in these nuclear domains, as previously described (26) .

Effects of temperature on diffusion coefficients

As discussed in the text, diffusion coefficients decreased in a temperature-dependent manner. FRAP of mRNPs in induced cells or free nuclear YFP-MS2 in uninduced cells measured at 37°C, 32°C, 27°C and 22°C showed a linear decrease in D from 0.09 to 0.049 µm²/sec and from 1.84 to 1.35 µm²/sec, factors of 1.83 and 1.36, respectively (Fig. S4, A to H). Similarly, different dextrans in the nucleus were observed to linearly decrease their diffusion coefficients by factors of between 1.45-1.86 from 37°C to 10°C (23) . This linear decrease shows that no energy dependent binding or motion are involved in the mobility of these molecules. The diffusion coefficients of these molecules are solely dependent on the terms of the Stokes-Einstein relation and so the linear decrease in the diffusion coefficeients is solely attributed to the temperature dependence of aqueous solutions viscosity. The nucleoplasm is not a homogenous solution; therefore, nuclear viscosity will have different values in different microenvironments. The diffusion coefficients that we observe by FRAP are the integration of the many diffusion coefficients of individual particles present in microenvironments of different viscosities and can be represented as the ratio of apparent viscosities at different temperatures. For YFP-MS2, h₂₂/ h₃₇ = 1.35 ±0.06 and for mRNPs, h₂₂/ h₃₇ = 1.77 ±0.14. This difference means that mRNPs and the free YFP-MS2 protein do not have access to the same nucleoplasmic environment. Indeed, using SPT we did find that the movement of some mRNPs was corralled, implying that they cannot access all of the nucleoplasmic space.

Comparison of diffusion coefficients measured for mRNAs

The rules of movement for mRNAs in the nucleoplasmic space have been studied for more than a decade. It was proposed by Zachar et al. (27) that mRNAs  diffuse through the interchromatin space (then termed: extrachromosomal channel network) on the way to the nuclear pores. This study was performed in Drosophila, looking by FISH at the dispersal of mRNAs transcribed from the suppressor-of-white-apricot gene. This observation was later extended to human cells (28) . Subsequent approaches have tried to address the movement characteristics of the total mRNA population in the nuclei of living cells. Fluorescence correlation microscopy (FCS) of fluorescein-labeled oligo(dT) probes revealed that nuclear mRNAs move at different rates with diffusion coefficients ranging from 0.01-10 µm²/sec (24) . One third of these diffusing oligo(dT)-poly(A) mRNAs were a slow moving fraction with average diffusion rates of <0.59 µm²/sec. In a further study using photo-activation of a caged oligo(dT) probe the calculated average diffusion coefficient was 0.6 µm²/sec (29) . Similar numbers were obtained by FRAP of GFP-tagged poly(A)-binding protein II (PABP2) (30) . Recently, tagging of poly(A) RNA with oligo(U) probes and measurements of diffusion coefficients in the nucleoplasm and speckles by FRAP, resulted in D= 0.04  µm²/sec and 0.03  µm²/sec, respectively (31) . Our measurements of single nuclear mRNPs had a diffusion coefficient range of 0.01-0.09 µm²/sec, falling into the lower range of nuclear diffusion coefficients measured by Politz et al. (24) and in agreement with Molenaar et al. (31) . FRAP measurements of the mRNPs showed a D of 0.09 µm²/sec. These results are similar to the diffusion coefficients measured for single cytoplasmic particles which were in the range of 0.02-0.1 µm²/sec (10) . The diffusion coefficient of fluorescein-tagged U7 snRNA, was measured in the nucleoplasm of Xenopus oocytes and found to be 0.26 µm²/sec (32) , the diffusion coefficient of large Balbiani ring mRNPs was estimated to be either 0.08 or 0.12 µm²/sec (33) and the diffusion rates of HCMV-IE mRNA are ~0.13 µm²/sec (34) . Different methods may detect different populations of mRNPs and the results depend on several factors such as the sensitivity of the technique used, actual volume in which measurements were made, total length of experiments and affinity of the label to the RNA.
In the work reported here, three independent techniques yielded diffusion coefficients falling all into the same range of measurements. The advantage of using the MS2 system is the extremely high affinity of the MS2 coat protein to the mutant stem-loop we are using (U => C in position -5) which has a K_d ~10^-11 M (35) . Therefore the exchange of MS2 proteins on the mRNA is unlikely and does not affect the measurements. This assumption cannot be made using other techniques. The Kd of PABP2 is 500 times greater (36) . PABP2 can exchange from a bound to a free nucleoplasmic form and so it is possible that slow diffusing mRNPs would not be detectable as PABP2 exchange could be faster than diffusion.

Supporting References

1. S. M. Janicki et al., Cell 116, 683 (2004).
2. T. Tsukamoto et al., Nat Cell Biol 2, 871 (2000).
3. R. E. Campbell et al., Proc Natl Acad Sci U S A 99, 7877 (2002).
4. M. J. Saxton, K. Jacobson, Annu Rev Biophys Biomol Struct 26, 373 (1997).
5. M. J. Saxton, Biophys J 66, 394 (1994).
6. E. A. Wiemer, T. Wenzel, T. J. Deerinck, M. H. Ellisman, S. Subramani, J Cell Biol 136, 71 (1997).
7. M. Schrader, S. J. King, T. A. Stroh, T. A. Schroer, J Cell Sci 113, 3663 (2000).
8. P. Chartrand, E. Bertrand, R. H. Singer, R. M. Long, Methods Enzymol 318, 493 (2000).
9. E. Bertrand et al., Mol Cell 2, 437 (1998).
10. D. Fusco et al., Curr Biol 13, 161 (2003).
11. W. A. Carrington, Proceedings of the SPIE 1205, 72 (1990).
12. A. M. Femino, F. S. Fay, K. Fogarty, R. H. Singer, Science 280, 585 (1998).
13. Y. Shav-Tal, S. M. Shenoy, R. H. Singer, Eds., Visualization and quantification of single RNA molecules in living cells (Cold Spring Harbor Laboratory Press, 2004).
14. R. D. Phair, T. Misteli, Nature 404, 604 (2000).
15. D. Axelrod, D. E. Koppel, J. Schlessinger, E. Elson, W. W. Webb, Biophys J 16, 1055 (1976).
16. G. H. Patterson, J. Lippincott-Schwartz, Science 297, 1873 (2002).
17. C. D. Gajewski, L. Yang, E. A. Schon, G. Manfredi, Mol Biol Cell 14, 3628 (2003).
18. K. Shinzawa, Y. Tsujimoto, J Cell Biol 163, 1219 (2003).
19. A. M. Sheridan et al., J Biol Chem 276, 29899 (2001).
20. M. Sakharkar, F. Passetti, J. E. de Souza, M. Long, S. J. de Souza, Nucleic Acids Res 30, 191 (2002).
21. G. Dreyfuss, V. N. Kim, N. Kataoka, Nat Rev Mol Cell Biol 3, 195 (2002).
22. B. Daneholt, Curr Biol 9, R412 (1999).
23. I. Lang, M. Scholz, R. Peters, J Cell Biol 102, 1183 (1986).
24. J. C. Politz, E. S. Browne, D. E. Wolf, T. Pederson, Proc Natl Acad Sci U S A 95, 6043 (1998).
25. G. Blobel, Proc Natl Acad Sci U S A 82, 8527 (1985).
26. M. Platani, I. Goldberg, A. I. Lamond, J. R. Swedlow, Nat Cell Biol 4, 502 (2002).
27. Z. Zachar, J. Kramer, I. P. Mims, P. M. Bingham, J Cell Biol 121, 729 (1993).
28. R. W. Dirks, K. C. Daniel, A. K. Raap, J Cell Sci 108 ( Pt 7), 2565 (1995).
29. J. C. Politz, R. A. Tuft, T. Pederson, R. H. Singer, Curr Biol 9, 285 (1999).
30. A. Calapez et al., J Cell Biol 159, 795 (2002).
31. C. Molenaar, A. Abdulle, A. Gena, H. J. Tanke, R. W. Dirks, J Cell Biol 165, 191 (2004).
32. K. E. Handwerger, C. Murphy, J. G. Gall, J Cell Biol 160, 495 (2003).
33. O. P. Singh, B. Bjorkroth, S. Masich, L. Wieslander, B. Daneholt, Exp Cell Res 251, 135 (1999).
34. S. P. Snaar, P. Verdijk, H. J. Tanke, R. W. Dirks, J Cell Sci 115, 321 (2002).
35. H. E. Johansson et al., Proc Natl Acad Sci U S A 95, 9244 (1998).
36. E. Wahle, A. Lustig, P. Jeno, P. Maurer, J Biol Chem 268, 2937 (1993).