We are searching data for your request:
Upon completion, a link will appear to access the found materials.
I use ImageJ to do an analysis of a Westernblot Image.
If everything goes as wanted the workflow is fine. But if I do something wrong creating a lane there is no undo for a lane and also no way to remove all lanes that I can find.
How do I get back to my original image or undo a lane creation?
I'm using 1.49e on Mac OS
ImageJ doesn't have a feature to remove individual lanes. But that shouldn't be a problem. All you have to do is draw the first lane correctly (I'm referring to size). Then press 1. Now, while the selection is still… selected, click inside it, but not on the number (where the cursor becomes hand), and drag it where you want the next lane. And press 2. And so on. Draw all your lanes.
Now, if you misplaced a lane, click the rectangle selection tool, then click on the lane number while the cursor is a hand. It will be selected. If you're having troubles with this, use the shortcuts Ctrl+1 and Ctrl+2 or the menu Analyze - Gels - Select Previous/Next Lane to select between lanes. The cursor turns to normal. Click anywhere inside the lane and drag it on the right place.
If the first lane was bad and you want to remove all lanes, just go to Analyze, Gels and click on Reset.
As a molecular scientist, I perform a lot of western blots. As you may know, western blots are pretty useless unless you can quantify the difference between the lanes. Image J (or FIJI) is frequently utilized to quantify the results from western blots. However, I have always been weary that there are multiple ways to introduce bias into your quantification with this program. I have wondered if the same person were to analyze the same western blot multiple times, would they get the same values each time? To explore this, I decided to utilize an old western blot image and see what different values I could get without trying to be biased.
To start, I utilized the western blot below. Since this was just an exercise of intrigue, i only analyzed the top bands on the last three columns.
This is a really interesting analysis, that I have always wondered about. It makes sense that there is variation between your attempts because of the area around the band, and I wonder if it would be possible to take that into consideration when you are doing western blot analysis. Something like a technical replicate of the area selection? As you mentioned, it is important to correct for noise and that is another added step of variance and potential bias. I am curious if you did the same analysis with a loading control, and then normalized the signal. I realize this was just a quick experiment of intrigue, and it looks like maybe that top band is the control, or a spurious band. Regardless your analysis could shows the consequences of these types of analyses, and it makes
me wonder if the loading control would help. Or if it would just be another layer of area selection bias.
This also brings to light a point that sometimes we rely to heavily on the "statistical significance" of these band intensities. Again, this is all just me observing this band without knowing the scientific relevance (or really the experiment at all), but when I look at those bands I do not see a difference in intensities. Since your analysis says there is, it is important to know if this difference is biologically relevant.
I find this post extremely interesting, my lab is a physiology lab and we try not to do western blots unless we absolutely have to. We recently had to perform a couple (always trying to please grant reviewers) and while I was analyzing the data using ImajeJ I started to wonder the exact same thing. Even though the lanes are parallel once you take the picture you have to take into account the noise, the amount of protein loaded into your gel and as you mentioned the bias created when you are responsible for selecting the correct area for the quantification. Even though differences in the amount of protein can sometimes be detected with the naked eye quantification with the software we must be aware of the bias we introduce.
Like Amanda mentioned all of the quantification goes back to the significance and the magical p<0.05! However, seeing is believing so when we see that your lane 1 has a thinner band when compared to lane 6 can we take that as experimentally relevant? Or will we always be trapped in the paradigm of significance.
This is really an interesting post to me as I've been wondering this for quite a long time ago when I first did Western Blot and was told that I could use the ImageJ to analyze the bands. Actually I've done many blots in past few years, but only a few times when I really used the ImageJ and wanted to see the differences in intensity. This doesn't mean it's not helpful at all but I would recommend use ImageJ to analyze band intensity with caution.
There are a couple of factors that could make ImageJ analysis tricky. First thing is the sensitivity of your developing reagent and films or the imaging system you use. There are multiple choices outside and what we use in our lab is traditional film system and it's not so sensitive to distinguish slight difference. Actually one member of our lab once tried the loading control, where he loaded the same protein lysates at scaler amounts, say, 2ug, 4ug, 8ug,20ug. Then he did the blot and analyzed with ImageJ and he didn't see the intensity increase the same folds as he loaded in the first place. The trends were right, definitely greater intensity with more loading, but it was not two-folds of intensity in 4ug, or 4 folds in 8ug compared to 2ug loading. So I would doubt the amount inference made by ImageJ intensity analysis, let alone performing statistical analysis afterwards under our lab setting.
The second thing I bear in mind is that even if you have highly sensitive system, you should still make sure that what you compare have been done in one experiment and analyzed in one-shot to decrease the possibility of error or bias. As you've shown in the blog about setting the line in the bottom to calculate the value under the curve, this could be tricky too as you might draw quite different lines at different analyzing period. I think this is related to one lecture in our class that you should gather all the data you need first, and then you sit down and analyze it following the plan.
With all that, that's why my advisor always says, "don't use imageJ to analyze the data and show the stats, you need to first persuade me via my eyeball."
Thanks for doing this proof of principle work with ImageJ. I think it is essential that you are able to establish the validity of your measurements before they are used for experimental analysis and draw conclusions from them. These validations are not always privy to the reader following publication and I have wondered about standardization and quantifications with other techniqes. Our program recently had a seminar speaker give a talk on swarming bacteria which merge or do not merge their swarm zones on agar plates with each other based on relatedness of the strains. The "merging" behavior was denoted by whether a "boundary line" was formed between the two swarm zones. Some images of boundary lines were unclear and the distinctions did not seem absolute. The establishment of a reliable and accurate method of determining the population behavior, as you have validated ImageJ here, would have really have boosted my confidence in the speaker's data and conclusions. It's my hope that as we develop as scientists, we will take the approach you have pursue scientific integrity by validating our methods and becoming more knowledgeable about the conclusions we are qualified to make.
With ImageJ quantification of western blots the main issue I ran into, as you point out, is where to draw the line to remove the noise. I was taught to draw the line from the plateau (representing the baseline/background signal) on one side of the peak to the plateau on the other, while the other person analyzing the data was taught to draw the line from the local minima on one side of the peak to the local minima on the other side of the peak. We were sometimes getting very different results so we decided to run a two sample t-test for scores on a few of our westerns to see if there was a difference between my score and his scoring. We found that there was and ultimately we had to standardize our scoring methods. We came up with a protocol to try to encompass as many of the possible variations and circumstances we may encounter and ultimately our scoring was no longer significantly different. However, in the process of creating this protocol we talked to several postdocs and junior PIs who regularly ran westerns and each of them had their own variations on how to perform this analysis. I am unaware of a formal protocol for ImageJ analysis of Westerns, but as it is rarely (if ever) reported in methods how one went about this analysis (other than they performed quantitative analysis using ImageJ) and given the degree to which scorers can vary their methods perhaps a more formal protocol is necessary. If the data themselves are biased no manner of statistics can completely correct for that.
The way our lab perhaps accounted for this was the standard we went by as far as publishable western data was that, while it is important to do quantitative analysis on westerns, if you cannot see the result qualitatively (i.e. just be able to see that band sizes are different by the naked eye), especially given the degree of variation present in our methods of analysis, your result is much less believable and perhaps ought to be taken with a grain of salt, and thus would not be considered publishable. I would be interested to hear what the practice of the labs of those of you that run westerns is with regards to qualitative vs quantitative analysis.
Westerns were my bread and butter in undergrad and even then quantifying them made me nervous in ways that flow plots still don't now. It always feels like an #overly honest methods post to say that you eyeballed the noise level. Westerns also always worried me with the fact that after a certain point, you lose any higher signal as the film is as developed as it can be, and the currently exposed portion is large enough to limit the area around it that can also react. Westerns are still a great tool, but I do wish we had less direct quantification methods for interpreting them, or were less reliant on those quantifications vs presence/absence or bloody obvious differences.
How to determine band intensity of Western blot? - High School Student in need of help for the Science Fair! (Feb/03/2005 )
I am doing an experiment using the Western Blot for my science fair project.
I heard that in order to produce a graph, the intensity of bands must be evaluated using adobe photoshop 5.5 or 7.0. I was wondering if someone can tell me please, step by step how to go about in obtaining the pixels/intensity of the bands generated using the program?
Thank you! I appreciate the help!
I heard that this can be done with photoshop, but i have no first-hand experience in that. in our lab we use ScionImage from Scion Corp.
You can get this here, though support is only given to paying customers.
but then, maybe someone comes around who uses PS for quantification and can help you out.
I'm not gonna tell you the procedurebut i want to point out about the quality of quantitation. honnestly he single method i've heard consists in loading on the SAME GEL your extracts and a know quantity of proteins 2, 4, 6, 8, etc. with different proteins of different molecular weight and if you use antibodies, to reveal you film with the same exposure time (in practical that says a corevelation). If you want just an aproximate (but not so bad) value you can use photoshop as said by jadefalcon.
First thing you have to do is get a GOOD exposure of your gel. This means you DON'T overexpose. Rule of thumb is you have to be able to see through the bands (i.e. put it up against a piece of paper and see if you can read soem text through it..)
Secondly, ALL the bands you want to compare have to on the same film. The quantification is all relative towards one band you put as 1 (or 100%).
1) In PS, load you your image
2) Then do, Image>Adjust>Invert (this way, you invert the scale)
3) Draw a box, which covers ONE band using the Marquee Tool (do the biggest band first, as the box HAS to be the same size all way through the analysis. This way to don't have to take the box size into consideration)
4) Then do, Image>Histogram and write down the Mean
5) Then drag the box to the next band and repeat
6) Finally, drag the box to measure the Mean from the Background, and subtract the number from all the others.
7) Divide all the numbers with your reference number for the one set as 1 (or 100%)
This is a VERY rough way of doing it, but do give you some hints on the relative increase/decrease of expression.
Just remember when you make a digital image of your film using a scanner, you need to create the highest quality image possible. We scan our films in grayscale and produce a 1200 dpi resolution TIFF file.
Do not make a jpeg of your film, jpegs contain far less image data then tiff files.
Id try out using the Scion image above. I dont have any experience using photoshop to quantify the bands as well.
Load the tiff file into Scion image. Ive never used scion image, but ithe pprinciple i suppose would be the same in any image program. Im sure this method is far from perfect, but It might work for you
First draw a box in an area that has no bands and select analyze->measure. This will tell you the background intensity of your image (mean), and also the area of your box Write these numbers down.
Next, draw a ROI (region of interest) around a band that you want to measure and select analyze-> measure. Again youll get an area measurement, and a mean intensity measurement. Write these numbers down. Repeat this for all the bands you want to measure.
Use care in making your ROIs and I'd experiment drawing ROIs and comparing the values you get from them until you think you can draw them properly and get an accurate measurement.
I might be wrong, but in order to quantify your bands, youll need to multiply the mean intensity and the area together to get the number you want. Do this also for your background. Write these numbers down.
Next subtract the background value (area x mean intensity) from each of your band values (area x mean intensity)
LEt me know if this makes sense. Also if anyone can improve my described method, dont hesitate to do so
I'm not sure how to quantitate the intensity of the bands, but there's free software called ImageJ (from the NIH) that is downloadable from the internet that our lab uses to do densitometry analysis of our Western blots.
Why not provide an exact address.
please note the date of that last post
I suspect you could find the address if you used google? or, alternatively, searched the NIH site?
Another one on western blot quantification and Scion Image (or ImageJ)…
Here is an illustration of the results I get:
With results like the ones illustrated on the image, wouldn’t it be better to take into account the size (area) of the band and not only its intensity .
I have very little experience using Scion Image but, when using the Gelplot Macro, I can only select the different lanes with the rectangular selection tool and the size of the rectangular selection has to stay the same for each lane.
Instead of drawing area of interest, I would like to use the magic wand selection, but this implies thresholding the picture, is it a good idea ?
Do you apply any kind of modification (threshold/enhance contrast/…) on the image prior to the measurements ?
Instead of subtracting a single background value for the whole film (from a rectangular selection taken anywhere on the film), is it a good idea or a waste of time to subtract a background value to each lane (taken just above or just under the corresponding lane) .
Cells are regulated according to the central dogma, which is the flow from genetic information to protein expression. The expression levels of proteins dictate cell fate for example, the expression levels of certain transcription factors regulate skeletal muscle differentiation . Thus, it is critical to quantify protein expression to understand cellular phenomena or potential at the molecular level. Western blotting is the standard method to quantify protein expression . The chemiluminescence method is generally used to detect proteins because of its high sensitivity and ease of use [3, 4]. However, recent developments in proteomics have focused on technologies to detect the expression of multiple proteins simultaneously, as opposed to the chemiluminescence method that detects single proteins. One method for simultaneous detection of multiple proteins which allows more quantitative analysis is fluorescent western blotting [5–7]. However, the detection sensitivity of this method is lower than that of the chemiluminescence method.
In this study, to improve the resolution and sensitivity of fluorescent western blotting, we focused on the fluorescence microscopy step. The microscope’s CCD camera is capable of capturing high-resolution images, but it has a small visual field. Thus, we captured an image with a high resolution and a wide field by merging multiple fluorescence micrographs. In addition, we successfully increased the detection sensitivity to a level comparable to that of the chemiluminescence method by optimizing the filters and fluorescent dyes.
Ways to remove autofluorescence
If you experience problems with autofluorescence, the first thing to do is to try and isolate the cause by investigating your sample. When you start your experiments, run an unlabeled control to determine whether there is any contribution to background fluorescence from the staining procedure.
It’s also important to know the spectra of your autofluorescence. This can be determined using spectral lambda scanning. Spectral scanning will help with experiment optimization and avoiding strong peaks in autofluorescence.
Once you have a good idea of the spectra of your autofluorescence, the next stage is to optimize your fluorophore choices. If the spectra of the autofluorescence and your fluorophore overlap very closely then there is a high chance that your signal will be masked by the autofluorescence. In this case, select a fluorophore with spectra far away from your background autofluorescence. For example, if your autofluorescence is in the blue region of the spectrum, move your fluorophore to the green or red.
When selecting fluorophores, choose a modern probe such as Alexa Fluor, Dylight, or Atto. These dyes tend to be brighter, more stable, and have narrower excitation and emission bands, making it easier to select only the signal from your fluorophore. If your microscope can detect them, far red dyes are a very good way to avoid problems with autofluorescence, since these wavelengths are seldom found in biological samples. Being able to detect far red dyes has the added bene fit of being able to expand the number of channels that can be used for multicolor experiments.
With a fluorophore selected, it is important to not blindly run your experiment with the concentrations listed on the tube. More often than not, a titration of the fluor phore will be necessary, testing different concentrations on your sample. This a lows you to see which concentrations give you the best differentiation from your background and the lowest autofluorescence. Use the manufacturer’s instructions as a guide and create a dilution series of the fluorophore to cover a range slightly below and slightly above the recommended usage. Test these dilutions against your sample to work out which will give you the most useful signal.
Optimize your microscope settings
The settings on your confocal microscope play a huge role in which signals are visible in your image, including autofluorescence. Use these settings to your advantage to cut out as much of the autofluorescence signal as possible by adapting the spectral detection. There are different ways to do this depending on the microscope, but the most flexible option is to choose a white light laser coupled with a spectral detector. This means that you can precisely tune the wavelengths of light reaching your sample as well as the wavelengths passing through the detector. This allows for very fine control when selecting which signals are recorded in the captured image, and gives ample opportunity to eliminate autofluorescence.Figure 1: When comparing the spectra of autofluorescence and the fluorophore, pick a fluorophore that does not overlap with the autofluorescence signal.
Treating your sample to avoid autofluorescence
Autofluorescence frequently comes not from the sample, but from the way it is treated prior to imaging. For example, mounting media, tissue culture media and laboratory plastic can all be sources of autofluorescence.
If you are running live-cell imaging experiments, then consider replacing your normal culture medium with pre-warmed phenol red-free medium or a clear buffered saline solution prior to imaging. In addition to the pH indicator phenol red, which is highly fluorescent when excited at 440 nm, culture media and supplements like FBS can contain many proteins and small molecules with fluorescent signals of their own—all of which can add up to a strong autofluorescence background if they are not removed. If you do decide to switch your cells to a new type of medium for live imaging, it’s important to be aware that this could cause unexpected changes in cell behavior and phenotype, so depending on what you are studying you may need to adapt the cells to the new medium first.
It’s also worth measuring autofluorescence from the imaging dishes that you are using with the same microscopy setup to see if this is a source of autofluorescence. If it is, try imaging using culture dishes with glass windows or glass-bottomed dishes that are specially designed to remove
Similar problems can also be seen when imaging biopsies and tissue samples that have been exposed to chemicals. A common example is the digestive tract, which can have large amounts of autofluorescence if it was exposed to antibiotics such as tetracycline that have a strong fluorescent signature. In this case, fluorescence can easily be removed by thorough rinsing with buffered saline or the careful use of solvents.
If you are fixing cells, the fixation method can have a big impact the autofluorescence. Consider alternatives to formalin and glutaraldehyde, and try not to store fixed samples for too long before imaging, because autofluorescence can increase over time.
While omitting autofluorescence through experimental design is optimal, som times it is just not possible. In these cases, there can be computational solutions to your autofluorescence problems. Using your microscopy software or open source solutions such as ImageJ, it is possible to analyze the pixels containing the aut fluorescence and try to subtract this from the overall image . Beware, though, that computational approaches can be complex. There are many different methods and algorithms to choose from, so it’s a good idea to try a few to see which work well. It’s important to remember that these methods can often reduce the strength of your fluorophore signal as well as autofluorescence, so use them carefully and be aware that there is often a tradeoff between increasing contrast with the bac ground and reducing your signal intensity.
Removing autofluorescence after fixation
Once you have your samples prepared and stored in the freezer, it can seem like any autofluorescence is there to stay. While it is easier to eliminate autofluorescence in the sample preparation stages, there are certainly steps that can be taken even at later stages in the process.
There are many chemical treatments that can attenuate autofluorescence signals. Some of these are commercially available, while others can be easily prepared using common lab chemicals such a sodium borohydride, Sudan black B, ammonium ethanol etc. [2,3].
Photobleaching tends to have negative connotations in microscopy, being associated with losing your fluorescent signal after spending a long time hunting for the best image with the laser turned up slightly too high. However, for autofluorescence, bleaching can be your friend. Before adding your fluorophores, you can treat your sample with high intensity LED light to bleach all of the background autofluorescence. Afterwards, your chosen fluorophore should contrast much more strongly against the bleached background .
RESULTS AND DISCUSSION
This laboratory exercise takes 2 or 3 laboratory periods to complete (see Table I for time line). On the first day, liquid LB broth cultures (2 ml) are inoculated from isolated colonies on plates previously streaked out and incubated by the teaching assistant or instructor. At the desired times (either along with the initial inoculation, which is what we have routinely done, or after a specific time of growth), inducer is added to the culture (IPTG for the lac operon in DH5α cells, arabinose for the pGLO plasmid in HB101 cells), and incubation is continued at 37 °C with shaking. After the designated time of incubation (5 or 6 h to overnight), 1.5 ml of culture is transferred to a microcentrifuge tube, and the cells are pelleted by centrifugation at maximum speed for 2 min. Upon resuspension in 1× SDS loading buffer, the samples can be directly loaded into polyacrylamide gels for electrophoresis or stored at −20 °C until use. Electrophoresis and processing can be carried out on the same day, but if so, the transfer of proteins to the PVDF membrane needs to also be done the same day. After transfer, the membrane can be stored at 4 °C for 1 or 2 days, or it can be taken directly to the blocking step followed by overnight incubation with the primary antibody. These steps can all be carried out in one period of about 3 h or divided into 2 days. The washing, incubation with secondary antibody, and development using the chemiluminescent substrate should all be done in 1 day and should also take about 3 h.
Typical student results for analysis of lac operon control are shown in Fig. 3A. As can be easily seen, there is an increase in signal for the β-galactosidase protein in the presence of the IPTG inducer, but there is still a significant amount of signal in the absence of inducer. This difference is more noticeable for the samples grown for shorter times (5 or 6 h when compared with 10 h). This difference is not due to differences in total protein, as can be seen by comparison of the amount of protein in each lane in an identical, Coomassie Blue-stained gel (Fig. 3B). For overnight cultures, there is less difference in signal for β-galactosidase between the uninduced and induced protein samples (Fig. 4). These results are in agreement with the published findings that control is decreased in stationary phase cultures [ 5 ].
Western blot analysis of β-galactosidase expression in induced and uninduced bacterial cells. A, DH5α cells were incubated in the presence (+) or absence (−) of IPTG inducer as described under “Materials and Methods” for the times indicated at the bottom of the figure. Total protein was recovered and separated by SDS-PAGE, transferred to PVDF membrane, and incubated with mouse anti-β-galactosidase antibody. Detection was performed by chemiluminescence using an anti-mouse-HRP antibody. B, Coomassie Blue-stained gel showing total protein from induced (+) and uninduced (−) cells. The gel was loaded identically as the gel used for the Western blot shown in A. M, marker proteins included in the gel (Bio-Rad Precision unstained markers, with molecular weights from top to bottom of 250, 150, 100, 75, 50, 37, 25, 20, 15, and 10 kDa).
Western blot of β-galactosidase expression in overnight (stationary phase) cultures. As can be seen, there appears to be less difference between the uninduced and induced protein samples.
Results from student Western blots of the arabinose operon are shown in Fig. 5. The first two lanes contain total proteins from cell lysates of overnight cultures grown in the presence (+) and absence (−) of arabinose, the inducer for the operon. The blot has been incubated with polyclonal antibodies against GFP, which is the product of the reporter gene linked with the arabinose promoter in the construct. A strong signal for the size of GFP is observed for the induced sample, whereas no detectable signal is observed in the uninduced lane. The higher molecular weight band is an unrelated protein that cross-reacts with the antibody. Some samples were subjected to immunoprecipitation to concentrate the GFP, as can be observed by the very concentrated GFP band in the IP+ lane (Fig. 5). For the overnight culture grown in the absence of arabinose inducer, a band of the appropriate size is detected, but it is much less intense when compared with the induced sample lane. The band is nearly undetectable in the uninduced sample grown for only 5 h. Variations are observed between samples and may be a result of differences in the length of time of growth until and after induction and small differences in the temperature of incubation. Other post-incubation differences could also lead to variations, including insufficient pelleting or resuspension of the bacterial cells, incomplete denaturation of proteins prior to SDS-PAGE, errors in gel loading, and transfer or post-transfer steps.
Western blot analysis of GFP expression from an arabinose-inducible promoter. The HB101 host strain carrying the pGLO plasmid was grown in the presence (+) or absence (−) of arabinose inducer overnight, total proteins were isolated as before (for the left panel, labeled L for lysate), and samples were separated by SDS-PAGE. Proteins were then transferred to a PVDF membrane and incubated with rabbit anti-GFP antibody as described under “Materials and Methods.” The right panel includes samples grown in the presence (+) or absence (−) of inducer overnight or for 5 h, as indicated, after concentration by IP. M is a marker lane containing labeled proteins (the 30- and 40-kDa markers can be observed). The arrows indicate the ∼29-kDa GFP protein size. The higher molecular weight bands are apparently due to a cross-reacting artifact from the cells.
Weak signals may be due to one or more of the following: insufficient transfer from gel to membrane, incorrect transfer buffers, incubation with antibody was not long enough, or antibody is too dilute.
Too much background can result from using too much antibody or from allowing the membrane to dry out during the incubation steps. Enough solution should be used to keep the membrane completely submersed but without wasting excess solution and antibody. We have found that a pipette tip box lid is an ideal size for holding 20–30 ml of solution. The proper dilution of antibody needs to be empirically determined.
Suboptimal immunoprecipitation conditions may lead to reduced protein signal or an increase in nonspecific bands and background staining following Western analysis. Salt and detergent concentrations in the IP buffer can be adjusted as needed to maximize efficiency and specificity.
Questions for Classroom Discussion
What is the basis of any differences in the amount of β-galactosidase observed in Western blots relative to time of growth?
Does there appear to be more control of gene expression in the logarithmic stage of growth versus stationary phase?
Why should the time when inducer is added to the culture affect expression levels if the cells are grown to the same final density?
How might temperature of culture incubation affect expression what would be some possible reasons for altering the temperature?
Upon comparison, does the ara promoter show greater or lesser control over gene expression when compared with the lac promoter?
How does immunoprecipitation enhance detection levels of proteins?
When would lac promoter constructs have the advantage over ara promoter constructs?
For what applications would the ara promoter be most useful?
What modifications to either promoter might be useful to obtain optimal expression under specific growth conditions?
Results and discussion
Optimization of PCFT expression
In pilot studies where we varied the number of viral particles per cell (multiplicity of infection, MOI) and density of Sf9 cells grown in a 50-ml suspension culture, we found that PCFT expression was better at a MOI of 2 and a density of 2 x 10 6 cell/ml. Fig 1A shows PCFT expression under these conditions, as a function of time after infection. For each timed sample, we determined cell viability and PCFT expression by Western blot. PCFT expression peaked 48 h post-infection with a cell viability of
40%. In summary, the following conditions yielded the best PCFT expression: Sf9 cells infected at a density of 2 x 10 6 cell/ml with a MOI of 2 and harvested 48 h after infection.
A. Sf9 cells in 50 ml suspension culture at a density of 2 x 10 6 cells/ml were infected at a MOI of 2. One-ml whole cell samples were collected at the indicated time points, electrophoresed on a 4–15% Mini Protean TGX Precast SDS-PAGE gel (BioRad), transferred to a PVDF membrane, and immunoblotted using an antibody against the His6 tag of PCFT. PCFT bands were observed at
43 kDa. The highest PCFT expression was observed 48 h post infection. No PCFT expression was observed at the time of infection (0 h) or in uninfected cells after 48 h (uninf). B. Treatment of membrane vesicles with PNGase F under non-denaturing conditions shifted the PCFT band from
39 kDa (each lane treated with PNGase corresponds to a different sample preparation).
The whole cell samples used for the initial expression investigations yielded two close bands (
43 kDa) in Western blots. N-linked glycosylation of two asparagine residues within the first extracellular loop has been observed in recombinant human PCFT expressed in mammalian cells or Xenopus laevis oocytes [26, 31]. In these expression systems, treatment with PNGase F shifted the PCFT band in Western blots from
35 kDa. Insect cells can add compact, relatively homogenous α1–6 fucosylated Man3GlcNAc2 sugar moieties  of
16 kDa per glycosylation site . Therefore, we suspected that the band with slower mobility in our blots corresponds to glycosylated PCFT localized at the plasma membrane, whereas the faster band corresponds to non-glycosylated PCFT. Isolation of membrane vesicles and solubilization yielded samples highly-enriched in the 43 kDa species (Fig 1B control). Consistent with glycosylation, PNGase treatment shifted the slower mobility band to a band of higher mobility (Fig 1B). In our hands, PCFT bands obtained from Sf9 whole cell lysates correspond to approximate molecular weights of 43 for glycosylated PCFT and 39 for deglycosylated PCFT.
Detergent screening for solubilization of PCFT
DDM has been widely used as a detergent to solubilize and study membrane proteins, including MFS transporters [34–43]. In pilot experiments to identify detergents suitable for solubilization of PCFT from Sf9 membranes we tested nine different detergents (concentrations in % w/v), including nonionic (0.5–2% DDM, 1–2% UDM, 0.5–2% DM, 1.25% and 4.5% OG, 1.25% NG, 1% DMNG) and zwitterionic detergents (0.5% AZ, 1% and 2.5% CHAPS and 1% FS-12). PCFT in Sf9 membranes was solubilized with these detergents, centrifuged at high-speed to separate solubilized proteins (supernatant) from unsolubilized material (pellet), separated by SDS-PAGE, and visualized by Western blot. The efficiency of solubilization was assessed by comparing the intensity of the PCFT band after solubilization to that of the control band (unsolubilized, non-centrifuged sample). PCFT in the Sf9 membranes was solubilized quantitatively (
90 to 100%) with 1 or 2% DDM, 1% FS-12 or 0.5% AZ (Fig 2). DDM consistently yielded very high efficiency in solubilization (> 70%). Based on this and its widespread use, we used DDM for all subsequent preparations.
Nine different detergents were used at the indicated concentrations to analyze solubilization of PCFT from membranes isolated from Sf9 cells. After a 2-h incubation, solubilized supernatants and pellets were analyzed using 4–15% MiniProtean TGX Precast SDS-PAGE gels, transferred to PVDF membranes and immunoblotted for detection with an antibody against the His6 tag of PCFT. First lane (Total protein) is the total amount of PCFT in Sf9 crude membranes before solubilization, and the second lane is a sample without detergent. (A) Non-solubilized pellets of the initial screen, (B) solubilized supernatants of the initial screen, and (C) solubilized supernatants with increased OG and CHAPS concentrations. See text for detergent abbreviations.
We enriched PCFT from solubilized membrane proteins by liquid chromatography based on the affinity of its His6 tag for transition metals. The PCFT eluted from a TALON Co 2+ resin was significantly cleaner than that eluted from a resin containing immobilized Ni 2+ (data not shown). The His6-tagged PCFT eluted from the Co 2+ resin with 200 mM imidazole migrated at ca. 43 kDa in a 4–15% MiniProtean TGX Precast SDS-PAGE gel and showed a high degree of purity (Fig 3).
PCFT was reconstituted in liposomes as described in Experimental Procedures. Purified protein and reconstituted PCFT were subjected to SDS-PAGE (4–15% MiniProtean TGX Precast gel). (A) Protein staining (Stain-free gel, BioRad) and (B) Western blot of the same gel analyzed using an antibody against the His6 tag of PCFT. Lane 1: purified protein eluted from the Talon Co 2+ resin lane 2: PCFT reconstituted in proteoliposomes.
Size-Exclusion Chromatography (SEC) analysis
DDM-solubilized PCFT purified by immobilized Co 2+ affinity chromatography was purified further by SEC, with an overall yield of pure PCFT of
0.9 mg/l culture. Based on PCFT’s elution volume of 11.4 ml and the elution profiles of protein standards, the calculated molecular mass of the PCFT-DDM complex was
280 kDa (Fig 4) consistent with a large amount of detergent, as has been observed previously for other 12 transmembrane segment transporters . Alternatively, our purified PCFT could be an oligomer, but this seems unlikely because we have shown that the monomer is the human PCFT structural/functional unit in membranes of mammalian cells and frog oocytes .
(A) Elution profile of PCFT solubilized in DDM. Elution volumes of standard proteins are indicated as follows: thyroglobulin (T, 669 kDa), ferritin (F, 440 kDa), aldolase (A, 158 kDa), conalbumin (C, 75 kDa), ovalbumin (O, 44 kDa). Blue dextran (BD, 2 MDa) was used for void volume determination. (B) PCFT fraction corresponding to the peak PCFT elution fraction was analyzed by protein staining (BioRad stain-free imaging). (C) The partition coefficients (Kav) of the standard proteins are plotted against the log of their molecular weights to calculate the size of the PCFT-DDM complex, yielding an apparent size of 280 kDa.
Specific uptake of folic acid in Sf9 cells.
The uptake of 3 H-folic acid in Sf9 cells expressing PCFT and uninfected cells was measured over 10 min at pH 5.5 . The time course of folic acid uptake was not examined in the present study, but Fig 5A shows that the uptake in cells expressing PCFT was significantly higher than in uninfected cells, and was reduced in the presence of a 200-fold excess of cold (unlabeled) folic acid (one-way ANOVA with Dunnett’s multiple comparison test, P < 0.0001). These data indicate that PCFT expressed at the plasma membrane in Sf9 cells is functional. Fig 5B shows that the uptake measured at pH 5.5 over 10 min was concentration dependent, with a Km for 3 H-folic acid uptake of 1.94 ± 0.20 μM (n = 3). This Km is similar to that reported in mammalian cells (1.7 μM in HEK 293 cells) and X. laevis oocytes (1.3 μM).
(A) 10-min uptake of 500 nM 3 H-folic acid (FA) by PCFT-expressing Sf9 cells determined at pH 5.5. Data are means ± SD. The value in Sf9 cells expressing PCFT (PCFT) was significantly different from that of uninfected cells (Control) (1-way ANOVA with Tukey’s multiple comparison test, P ≤ 0.0001, ****). Uptake was reduced significantly in the presence of a 200-fold excess of unlabeled folic acid (PCFT + Ex FA) (1-way ANOVA with Tukey’s multiple comparison test, P ≤ 0.0001, ****). The difference between the PCFT + Ex FA and Control was not significant (ns). (B) Concentration dependence of the 3 H-folic acid uptake in PCFT-expressing Sf9 cells (PCFT) and uninfected cells (Control). Data was fit using the Michaelis Menten equation (Graphpad Prism, San Diego, CA).
Functionality of lipid reconstituted PCFT.
Affinity purified PCFT was concentrated to 0.6 mg/ml and reconstituted in unilamellar liposomes as indicated under Experimental Procedures. The protein staining of SDS-PAGE gel and Western blot analysis in Fig 3 show the presence of PCFT in the proteoliposomes. PCFT function was demonstrated by the 30-s uptake of 300 nM 3 H-folic acid (Fig 6), where PCFT-proteoliposomes showed significantly higher uptake of 3 H-folic acid than liposomes (t-test: L vs. PCFT-PL P = 0.008, r 2 = 0.86 one preparation, n = 3). These results demonstrate reconstitution of functional PCFT in liposomes.
The 30-s uptake of 300 nM 3 H-folic acid (FA) into unilamellar PCFT-proteliposomes (PCFT-PL) was measured at pH 5.5. Unilamellar empty liposomes (L) served as control. The uptake in PCFT-PL was significantly higher than that in liposomes (t-test: L–PCFT-PL P = 0.008, r 2 = 0.86 one preparation, n = 3).
Peroxisome fission is important for degradation
To analyze whether peroxisome fission is important for glucose-induced selective peroxisome degradation (macropexophagy), we analyzed this process in wild-type H. polymorpha as well as in two mutant strains (dnm1 and pex11) that are strongly impaired in peroxisome fission. 19 , 20 In line with earlier observations, the levels of the peroxisomal marker protein alcohol oxidase (AO) gradually decreased in the wild-type control upon induction of macropexophagy by glucose ( Fig.ꀚ and B ). However, in both dnm1 and pex11 cells, no significant reduction of AO protein levels was observed. A similar result was obtained in the atg11 control strain, which is defective in selective pexophagy. 21 These results suggest that a reduction in peroxisome fission affects glucose-induced macropexophagy.
Figureਁ. Reduced peroxisome degradation in H. polymorpha and S. cerevisiae fission mutants. (A) Pexophagy was induced by glucose in H. polymorpha cells grown for 20 h on methanol. Equal volumes of cultures were loaded per lane. Western blots decorated with anti-alcohol oxidase (α-AO) antibodies show no significant reduction in AO levels in the peroxisomal fission mutants dnm1 and pex11, similar to the atg11 control, which is blocked in pexophagy, whereas AO levels gradually decreased in the wild-type control as expected. (B) Densitometry quantification of the blots shown in (A). The amount of AO protein present at t = 0 h was set to 100%. The bar represents the standard error of the mean (SEM). (**p < 0.01). (C) Western blot analysis showing thiolase levels of S. cerevisiae cells grown on oleate and subsequently diluted into SD(-N) medium to induce peroxisome degradation. Samples were harvested at different time points, and equal amounts of protein were loaded per lane. There is no significant decrease of thiolase in the dnm1∆ vps1∆ double mutant where the peroxisomal fission is completely blocked. A similar result was obtained for the atg1 control strain. In the wild-type and the single vps1 and dnm1 deletion strains degradation occurred. (D) Quantification of thiolase blots shown in (C). The level of thiolase protein at t = 0 h was set to 100%. Levels were adjusted to the loading control glucose-6-phosphate dehydrogenase (not shown). The bar represents the SEM (**p < 0.01). (E) Constitutive peroxisome degradation is reduced in H. polymorpha fission mutants. Cells were grown on methanol for 16 h. Western blots were prepared using crude extracts and anti-GFP antibodies to detect Pmp47-GFP and GFP degradation products. The data show that the levels of the cleaved fusion protein (lower band) are reduced in pex11 and dnm1 cells, relative to the wild-type control. Equal concentrations of protein were loaded per lane. (F) Quantification of blots shown in (E). The levels of full-length Pmp47-GFP proteins were arbitrarily set to 100%. The levels of cleaved GFP (lower band) are indicated as percentage of the full-length fusion protein. The bar represents the SEM (*p < 0.05 **p < 0.01).
To rule out that the observed block in pexophagy is not related to the fission defect in H. polymorpha dnm1 or pex11 cells, but related to a direct function of fission proteins in pexophagy, we performed a control study using Saccharomyces cerevisiae. Different from H. polymorpha, in baker’s yeast, Dnm1 and Vps1 have redundant functions in peroxisome fission. Consequently, single dnm1 and vps1 mutants are only partially affected in peroxisome fission and thus can be analyzed for a direct function of these proteins in pexophagy. 22 As shown in Figureꀜ and D , single S. cerevisiae dnm1 and vps1 mutants were not blocked in glucose-induced pexophagy, whereas cells of the dnm1 vps1 double mutant, which has a major peroxisome fission defect, were impaired in peroxisome degradation.
Using an H. polymorpha strain that produces the peroxisomal membrane protein Pmp47 fused to green fluorescent protein (Pmp47-GFP), we analyzed constitutive peroxisome degradation in methanol-grown cells of H. polymorpha wild type and both dnm1 and pex11 mutant strains using western blot analysis and anti-GFP antibodies ( Fig.ꀞ ). As expected, in extracts of wild-type cells in addition to the band representing the full-length Pmp47-GFP fusion protein, a faster migrating band consisting of cleaved GFP was also evident, indicative of constitutive pexophagy. The ratio of the amount of cleaved GFP relative to the full-length fusion protein was reduced in dnm1 and pex11 cells compared with wild-type controls ( Fig.ꀟ ), indicating that constitutive autophagy of peroxisomes is also affected in H. polymorpha pex11 and dnm1 cells. 23 Summarizing these data indicates that in yeast peroxisome fission is required for pexophagy.
Intra-peroxisomal protein aggregates affect growth and cause oxidative stress
Next, we addressed whether peroxisome fission acts in quality control of the organelles. For this we took advantage of earlier observations that production of a mutant variant of catalase, designated Cat mut , produced in wild-type H. polymorpha (i.e., also producing the endogenous catalase protein) forms enzymatically inactive protein aggregates in the peroxisomal lumen. 13 Peroxisomes containing these protein aggregates were used as a model for aberrant peroxisomes in this study.
First, we tested whether the presence of peroxisomal protein aggregates had physiological consequences. As shown in Figureꀪ cultures of the Cat mut strain showed a reduced yield. Remarkably, the specific catalase activities in cell extracts of cells of the Cat mut strain (185 U/mg) were enhanced relative to that of the wild-type control (130 U/mg). ROS measurements indicated that at all time points examined, the cells containing peroxisomal protein aggregates had enhanced ROS levels relative to the wild-type control ( Fig.ꀫ ).
Figureਂ. The effect of peroxisomal protein aggregates on growth and ROS levels. (A) Final optical densities of wild-type, dnm1 and pex11 cells, producing or not producing Cat mut , upon growth on methanol as sole carbon source for 40 h. Cell were extensively precultivated in glucose medium and subsequently shifted to medium containing methanol. Final optical densities are expressed as adsorption at 660 nm. The bar represents the SEM (*p < 0.05). (B) ROS levels in cells at different time points after the shift of glucose-grown cells to methanol medium. The mean intensity was measured by FACS. Cat mut cells show an enhanced ROS production relative to wild-type controls. The bar represents the SEM (*p < 0.05 **p < 0.01).
Intraperoxisomal protein aggregates are removed by fission and degradation
To substantiate whether the protein aggregates induce peroxisome fission, we introduced Cat mut in the H. polymorpha atg1 mutant, in which autophagic degradation of peroxisomes is blocked. 24 In order to visualize peroxisomes we introduced the fluorescent peroxisomal membrane marker Pmp47-GFP. Fluorescence microscopy analyses of cells, cultivated for 16 h on methanol, revealed that the Cat mut -producing atg1 strain contained, in addition to the normal organelles, relatively small organelles that were not observed in the atg1 parental strain ( Fig.ꀺ ). This result was confirmed by electron microscopy analysis of KMnO4-fixed cells ( Fig.ꀻ ) and also showed the presence of aggregates in the small organelles ( Fig.ꀻ ). Subsequent analysis, however, revealed that small organelles were not evident by fluorescence microscopy of wild-type cells producing Cat mut ( Fig.ꀺ ), whereas electron microscopy revealed that these cells harbored fewer small aggregate-containing peroxisomes relative to the atg1 strain producing Cat mut . Apparently, the small separated organelles were rapidly removed in the wild-type background. This process was further analyzed by electron microscopy. These analyses revealed that once formed, aggregates migrated to the periphery of the organelle where they were included in the buds that subsequently split off from the mother organelle ( Fig.ꀼ ).
Figureਃ. Microscopy analysis of H. polymorpha atg1, atg1-Cat mut and wild-type-Cat mut cells. Cells were grown on glycerol/methanol for 16 h. Peroxisomal membranes are marked by Pmp47-GFP. (A) Relative to the atg1 control, the atg1-Cat mut strain contains multiple small peroxisomes, which was not evident in wild-type-Cat mut cells. The bar represents 1 µm. The peroxisomal phenotype was confirmed by electron microscopy (B). Note the presence of protein aggregates in many of the organelles in atg1-Cat mut cells of which fewer are observed in wild-type-Cat mut cells. The bar represents 0.5 µm. (C) Ultrathin sections of KMnO4-fixed atg1-Cat mut cells showing different stages of the formation of a small peroxisome containing a protein aggregate. The bar represents 0.2 µm. (D) Electron micrographs, showing details of dnm1.atg1 cells (left panel) and pex11.atg1 cells (right panel) producing Cat mut , demonstrating the presence of aggregates in the enlarged peroxisomes in these cells. The bar represents 0.2 µm. P, peroxisomes M, mitochondria N, nucleus V, vacuole.
To further address their degradation, we cultivated the H. polymorpha Cat mut strain on a mixture of glycerol/methanol and subsequently administered 1 mM of the protease inhibitor phenylmethanesulfonylfluoride (PMSF) to the culture to reduce the rate of degradation of autophagic bodies in the vacuole. Electron microscopy analysis of these cells showed that at the onset of the experiment vacuoles in Cat mut -producing cells contained very low numbers of autophagic bodies ( Fig.ꁊ and B ). However, after 2 ( Fig.ꁌ and D ) and 4 h ( Fig.ꁎ and F ) of PMSF administration the accumulation of aggregate containing organelles in the vacuole was evident. These structures were not observed in the wild-type control. The presence of catalase protein was confirmed by immunocytochemistry ( Fig.G and H ). These data suggest that peroxisomes with protein aggregates are delivered to the vacuole for autophagic degradation. In conclusion, we showed that protein aggregates, which accumulated in the peroxisomal lumen are removed by concerted fission and degradation events.
Figure. Catalase aggregates are degraded by autophagy. H. polymorpha wild-type and Cat mut cells were cultivated on glycerol/methanol for 12 h and subsequently treated with 1 mM PMSF (t = 0). Electron micrographs of KMnO4-fixed Cat mut cells [at t = 0 h (B)], [at t = 2 h (D) and t = 4 h (F)] revealed progressive accumulation of autophagic bodies containing peroxisomes with protein aggregates that are not observed in wild-type cells (A, C and E). Immunocytochemical localization of catalase shows that labeling is confined to peroxisomes of wild-type cells (G) and is also abundant in the vacuole of Cat mut cells (H). M, mitochondria N, nucleus P, peroxisomes V, vacuole. The bar represents 0.5 µm.
Degradation of aggregates is reduced in dnm1, pex11 and atg11 cells
Clearly, the separation of small aggregate-containing peroxisomes is different from the fission events that participate in normal peroxisome proliferation in wild-type cells. Therefore, we analyzed if this fission process depends on Dnm1 and Pex11. 19 , 20 To this end, corresponding deletion stains were constructed that produced Cat mut N-terminally fused to GFP (GFP-Cat mut ) to fluorescently mark the protein aggregates. Fluorescence microscopy analysis of cells also producing the DsRed-SKL peroxisomal matrix marker protein revealed that small green fluorescent spots were present in peroxisomes of both mutant strains, as well as in the wild-type control ( Fig.ਅ ). To analyze the possible presence of GFP-Cat mut in the vacuole, the red vacuolar marker FM 4 was used instead of DsRed-SKL ( Fig.ਆ ). Fluorescence microscopy analysis confirmed that green fluorescence was observed only infrequently in vacuoles of dnm1 and pex11 cells relative to the wild-type control ( Fig.ਆ ). Western blot analysis using crude extracts of these cells confirmed reduced cleavage of GFP-Cat mut relative to that in the wild-type cells producing GFP-Cat mut ( Fig.ਇ ). The reduced degradation of the aggregates also affected growth of the strains as lower growth yields on methanol were observed in both dnm1 and pex11 cells for the strains producing Cat mut ( Fig.ꀪ ).
Figureਅ. Visualization of catalase protein aggregates. H. polymorpha wild-type, dnm1 and pex11 cells, producing DsRed-SKL and GFP-Cat mut , were grown on methanol for 16 h. The peroxisomes contained GFP spots in the peroxisomal lumen, which represent the protein aggregates.
Figureਆ. Visualization of catalase protein aggregates. Identical experiment as shown in Figureਅ , using wild-type, dnm1 and pex11 cells, producing GFP-Cat mut stained with the vacuolar marker dye FM 4. GFP fluorescence was frequently observed in the vacuoles of wild-type cells. GFP fluorescence was also observed in the vacuoles of the dnm1 and pex11 cells, albeit at reduced numbers compared with the wild-type control. The bar represents 1 µm.
Figureਇ. Autophagic degradation of GFP-Cat mut is reduced in pex11 and dnm1 cells. (A) Western blot analysis using crude extracts of cells described in Figureਅ , decorated with anti-GFP antibodies. All strains contain both the full-length GFP-Cat mut protein together with GFP (arrow), due to cleavage of the fusion protein. Pyruvate carboxylase (Pyc1) was used as a loading control. (B) Quantification of the levels of GFP-Cat mut and GFP protein of the blots shown in (A). The level of full-length GFP-Cat mut is set to 100%. The bar represents the SEM (**p < 0.01). (C) Quantification of GFP fluorescence in the vacuole. The percentage of cells containing GFP fluorescence in the vacuole was calculated for wild-type, dnm1 and pex11 cells containing PAOXGFP-Cat mut . Per strain 2 samples of each 100 cells were counted The bar represents the SEM (*p < 0.05).
To strengthen the observation that Dnm1 and Pex11 are indeed required for the separation of the small aggregate-containing peroxisomes, we also performed electron microscopy analysis of dnm1 atg1 and pex11 atg1 double-mutant cells producing Cat mut . As evident from Figureꀽ , these cells harbor large peroxisomes that contain protein aggregates.
Degradation of the aggregates was also reduced upon deletion of ATG11, a gene involved in selective peroxisome degradation. In atg11 cells producing GFP-Cat mut numerous green spots were observed ( Fig.ꂊ ). However, vacuoles did not accumulate GFP as observed in the wild-type background (compare Fig.ਆ and Fig.ꂋ ). The block in autophagic degradation was furthermore evident from western blot analysis of these cells showing that cleavage of GFP from the GFP-Cat mut fusion protein was strongly reduced ( Fig.ꂌ ).
Figureਈ. Autophagic degradation of GFP-Cat mut is reduced in atg11 cells. (A) In H. polymorpha atg11 cells producing DsRed-SKL and GFP-Cat mut and grown on methanol for 16 h, most GFP-Cat mut spots colocalize with the peroxisomal matrix marker DsRed-SKL. Some of the green spots do not colocalize with DsRed-SKL. Most likely this is the result of the asymmetric peroxisome fission process, resulting in the formation of small aggregate-containing peroxisomes that contain no or very little DsRed-SKL. (B) Vacuolar staining of atg11 GFP-Cat mut cells with FM 4 dye. GFP fluorescence in the vacuoles is reduced in atg11 cells relative to the wild-type control (see Fig.ਆ ). The bar represents 1 µm. (C) Western blots decorated with anti-GFP antibodies fail to demonstrate the GFP cleavage product (arrow) in atg11 cells producing GFP-Cat mut .
This error usually means that your image planes are larger than the maximum supported size.
The original ImageJ only supports image planes with 2 gigapixels (2^31 = 2147483648 pixels in case of a square image, the maximum allowed is 46340 x 46340 pixels) or less. If your data has extremely large image planes—e.g., 50000 x 50000 pixels—you may need to analyze region by region. One way to do this is using the “Crop on import” feature of the Bio-Formats plugin.
If you are using Bio-Formats to open a file, however, the size limit is a bit more complicated. Instead of using short as in ImageJ, Bio-Formats store data in byte when reading planes. If the source image is in 16 bit or in 32 bit (4 bytes, eg. floating point TIFF), the maximum pixel numbers allowed per plane will be 1/2 (1 gigapixels) or 1/4 (0.5 gigapixels), respectively.
ImageJ2 supports larger image planes internally, but uses the original ImageJ user interface by default, which once again limits visualization to 2 gigapixels. The ImageJ2 team is working to lift these size restrictions see imagej/imagej#87.
Plant materials and growth conditions
Arabidopsis (Columbia-0) and the T-DNA insertion mutant line (SALK_056011, locus At4g17740) were obtained from the Arabidopsis Resource Center (Columbus, OH). Seedlings were grown on 1/2 MS medium containing 3.0% sucrose (pH 5.7) for 4 weeks under the conditions of 16 h light / 8 h dark cycle and 20 μmol photons m − 2 s − 1 light intensity during the light periods.
Blue native PAGE and 2D SDS-PAGE
Chloroplasts were extracted from 50 wild-type and 50 atctpa-mutant plants. Blue native gel electrophoresis was performed as described previously [25, 38]. For 2D SDS-PAGE, the blue native gel lanes were excised with a razor blade and incubated in 2 × SDS sample buffer containing 2.5% (vol / vol) β-mercaptoethanol (β-ME) for 20 min at 75 °C, then for 20 min at 25 °C. Lanes with denatured proteins were placed on top of 12% SDS gels, then subjected to the second dimensional separation.
For immunoblotting, protein samples were separated on 12% SDS gels and transferred to nitrocellulose membranes (BioTrace™ NT nitrocellulose, Mexico) followed by a western blot analysis. After blocking with 5% milk, the membranes were subsequently incubated with primary antibodies generated against the indicated proteins and detected using the Super Signal™ West Pico PLUS Chemiluminescent Substrate kit (Thermo Scientific, USA).
Yeast two-hybrid assay and growth curves analysis
The yeast two-hybrid assay was performed using the Split Ubiquitin System (DUAL membrane, Dualsystems Biotech) as described previously [39, 40]. The mature D1 (amino acids 1–344) and pD1 (amino acids 1–353) were cloned into pCCW-STE vector (encoding the Cub-LexA-VP16 fragment) as the bait for interaction assay. CP43, CP47 and D2 were cloned into pDSL-Nx vector (encoding the NubG fragment) as the prey for the assay. Yeast strain NMY32 was co-transformed with the bait and prey constructs, respectively. The interactions were determined by the growth of yeast cells on agar plates with Synthetic Defined (SD) medium lacking Trp, Leu, His, and adenine (SD-Trp Leu His Ade, FunGenome) without or with 2 mM 3-amino-1, 2, 4-Triazole (3-AT). The growth curves of liquid cultured yeast cells were obtained by measuring the absorbance at 600 nM (OD600). Six colonies of each transformation were cultured in SD-Trp Leu His Ade medium containing 30 μg / ml kanamycine. The OD600 values were recorded at various time points. pDSL-Nx vector was used as the negative control, and NubI was used as the positive control.
ImageJ (https://imagej.nih.gov/ij/) was employed to qualify the distribution ratio of PSII subunits among different subcomplexes.