 So, this morning, I have the pleasure of introducing from Janowa, Alberto Diazpan. Thank you very much, Colin. And thank you for this chance to give a lecture, a second lecture here. So again, as I told you yesterday, in order to convince in you that we are in the era of the extra microscope, reflecting the sentence in the Galilei's letter, today I will focus on nanoscale optical microscopy. I have some topics that I wanted to discuss yesterday at no time that are threat, that are fluorescence recovery effort, photo bleaching, and maybe I will not be able to fully discuss about second harmonic generation and mirror matrix imaging. But in case we can discuss about these topics later or let's see if we have time. Now I would like to touch at this point related to nanoscale optical microscopy and focusing on fluorescence nanoscopy. We will see that keyword fluorescence is relevant in this case. As you have seen yesterday, one of the main, most important objects we are interested in, well, I'm interested in, is the biological cell. Is the biological cell for two reasons. And one reason is related to the fact that if you remember the Feynman lecture of 1959, can be considered a nanomachine. So there is something that you can control or that this machinery controls at this level, nanometer level that influences all the further steps at higher levels. The second issue is related to the fact that cell aggregation and cell work contribute to constitute organ's tissues and our living body and so on. And so try to decipher what's going on in this complicated environment could help in understanding how to treat, for example, diseases or more in general, to live better. Now here you find here this sentence related to cell aggregates, small organisms and organ's tissues. Why? Maybe because I'm getting old, but I would like to move from studies made having cells that people is calling living cells on covers lips, having a strong interaction with glass to cell aggregates, having strong interactions with themselves. And so our idea is to try to export what we know today in fluorescence nanoscopy to large objects when is needed. You don't need a super resolved imaging of the full body, but maybe at a certain time, a certain temporal interval, it would be interesting to have the possibility of penetrating in the full body or organ or tissues recognizing a region and for that region, being quantitative in terms of molecular biology. So this is the idea behind. And at the end of this talk, yesterday, I liked very much this talk from the Othcan group about the portable microscope and I will show you a 50 cent microscope if I remember at the end, that can be plugged in a hyphen or any other device. So nanoscale optical microscopy, I don't know how many of you can recognize from this picture what is behind, I cannot, I couldn't now I know, but if I remove fog here, improving spatial resolution, maybe some of you can recognize the nuclear pores in the nucleus. And so the fact that today, we are able to describe the organization of the nuclear pore at the nanometric level and we can try to start studying what's going on in terms of input and output in the nuclear pore with a very high precision that is a nanometer scale and using probes that are not affecting too much the way the system is functioning, so light, visible light. And I like to bring to your attention this biological system that is chromatin DNA because if you have a look here, you have an object that is interesting because it's densely condensed in the nucleus with a lot of information available and imagine this information are released at the proper time and the proper space with a control that is made directly by the cell and by the organization of the cells. So it's like the case that you have a library with millions of books inside and you are passing from this library with your bike and there is one librarian coming out running to you and bringing to you a specific book open at certain page in order to give you the information for your next step is amazing. If librarian is wrong in the book, in the page or you cannot read very well what is written there, you can be in troubles. So now inside here, there is this great library with a lot of information but then if you get out, you see that you have different level of organization that are of interest, why? Oh, of interest for microscopy, why? Because this extraordinary microscope today allows us to tune as I told you as a radio the spatial resolution. Simply turning a knob, you can move from 200 nanometer down to 180, 20 if you need and then you have to start thinking because what you want to do is working room temperature, KT, Boltzmann constant and living systems. So you can push down to one nanometer but then you have some problems with the activity of your system itself. And so what we can do with the nanoscale optical microscopy is to tune spatial resolution according to our needs or, and I think this is the important issue, according to your question. So before moving to the nanoscale using the microscope, let's start having a question and for this question you need certain resolution and then you move to your instrument. It's not like a washing machine, it's not that you put inside your cells and then you select a program offering you the best resolution and then you get data and then you don't know what to do with this data because you had the appropriate question. Okay, so there is a chain here. Now if you have a look to this optical microscopy scenario again like yesterday, you see again is rather complicated and today we will try to link these two keywords, optics approach and probes approach. Yesterday we were discussing about probes approach, I'm sorry, I didn't discuss about Fred, we can do later but today there is a mix here and within this mix you have in this forest of methods, you have so many methods developed according to different interesting optics or to different questions in biology and in medicine. Now every time there is a new microscope in optical microscopy, there is a question that people is not posing for developments in electron microscopy. Usually when you have electron microscopy people say okay that's fine, we can get a very high resolution for getting that you have to solidify the sample, physically cut the sample, metalize the sample, sending electrons and so on. So forgetting about this but no questions about, for example, living systems. So the very first question is are you able to play with living systems? What about resolution? Fine, multi-dimensionality yes because now you are using light and you have seen in these days or any other mechanism of contrast, thermal or optical and then you try to produce an image. And so you can try to combine all this information in an image and so when you touch your image you get information about fluorescence, about refractive index, about thermal properties, about everything and you can have today this knowledge all together in your image. And then there is something related to versatility of course you don't want too many artifacts and you need to be able to recognize them, sample preparation and people is also asking for about system complexity when they want to use the microscope. It seems that when people is approaching the microscope no one wants to learn something behind the microscope. They simply want to use the microscope and so they want something simple in terms of use. It's not. If you want to get important information you have to design your experiment with your microscope and with your biological question. So nothing is simple. Otherwise everybody can ask. You can buy in the supermarket that you can do but then it's not so simple to use it that you understand what's going on. And if you have a look from the 80s up to today two years ago, distance between optical microscopy and electron microscopy is something that today is confused, is mixed. So you're approaching this kind of spatial resolution. Now, I have the great privilege of having the possibility of having a victim for some discussion related to super resolution that is having the possibility of talking not too much as I wanted or as I want with Colin Shepard about super resolution issues and not only. But I like to remind that the term super resolution was in some way brought to the attention of the people by another Italian scientist. So we started from Galileo Galilei and now we have Giuliano Torraldo di Francia. Now, apart from going in detail with the paper or with papers by Giuliano Torraldo di Francia I think that is not wrong to think that the message was you cannot do anything against physical laws, the fraction. Simply you have to consider that you have this limitation that is not a limitation is also something that you can get. So it's what you can get with your instrument in terms of space, in terms of time. But what you can do in treating information coming from this physically limited instrument in case you have some other information outside from the direct acquisition channel of your microscope you can merge your information coming from the microscope with information with a priori or an additional set of information and merging them you can take advantage and you can increase your ability for example in getting a better spatial resolution. And if you want to, I think this is a good reading but another very good reading is a recent paper by Colin Shepard that you can find in open access on this journal, simply you can go to this journal and you can download for one month this paper and is an excellent discussion in terms of resolution and super resolution and telling you something about the supersighted ABBE paper and which is the real relationship between the ABBE paper, the formula and citations that this paper is getting when people is talking about super resolution. So now you have your system. In order to describe your system, well, you can have a plot of a intensity along the line and in case you have this frequency or in case you have this behavior or this behavior you can understand that if you want to switch to frequencies you have high frequencies or low frequencies. So you can associate the frequency concept to space and you can think that when your sample has to transmit high frequencies these are high resolution data, spatial resolution data so something changing very fast in space. Distance is very close one to the other and when the system is more relaxed in these changes you have low frequencies. Then you have your system and since we are talking about frequencies you know and you had last week lessons about this you have a cutoff of frequency. Now if for some reason you can introduce data that allow you to have something behind this frequency of cutoff and you go back to space in that case you can have super resolution. Now I don't know if I will mention all the resolution criteria that people is using when it's looking to an image and try to say which is the current resolution. This is relevant and I will show you an example immediately after these few points. It is relevant because now let's assume that you have an appropriate question and this question has its core the fact that you want to learn something at a certain resolution. You want to get information at a certain resolution. If you have this when you look your image you can say something about what's going on that is related to the real resolution of the image. So if you think that one dot can be discriminated in terms of distance with another at the resolution of 10 nanometer you can say something. If you know that you cannot discriminate distances before then I don't know 500 nanometer your conclusion is not the same, right? So you want to be sure that you have something in your hands that allow you to say which is the objective resolution in your image. Well you can and you have seen in last week some models you can have some analytical formulation of the process you can try to model your microscope and you can try to model your system and then you can try to find the solution about the resolution you have. Or more recently, you can have calibration samples that are of the very same family of the samples you're interested in so biological samples. And so for example you can use what are known today as DNA origami. So you're able to place in precise positions fluorescent labels and so this is good for understanding which could be the resolution of your system when you have a biological system under the microscope. It's not completely true because if you remember the slide about chromatin DNA one issue is the one related to a chain where you have spotted your fluorescent probes and then the big mess that you have inside the nucleus also with changes in refractive index and so on so a lot of additional issues but you are close to the best performances of your system. Or what you can do is to have a line profile in your image and using this line profile as in this case you can speculate and say something about the resolution using spiral criterion or Rayleigh criterion to say which is your confidence in saying that you have a certain spatial resolution. This case is interesting but brings you to this, there is one point here. When you perform this line intensity profile it seems that you recognize the object. So you are looking to something that you recognize that the problem you knew before is good because you have some confirmation but maybe cannot tell you something really new about your sample. Then there is another method that is called Fourier ring correlation that can allow you to get an objective evaluation of the current spatial resolution you have with the microscope you are using for the sample you are interested in. In order to have this estimate of your imaging system what you need is you don't need any a prior information. You use your optical system as it is so the system you are currently using. You collect images from your sample and the only constraint is that you need two identical but statistical independent images of the same sample. The very simple way for doing this is collecting two successive images. Today you have detectors that allow you to collect in different position of your detector the two or pixels in your detector the two independent images. So you have some possibilities in this. When you have done this you perform your Fourier transformation and then through correlation here you can find the cutoff frequency which is the good point here is not only that you're able to find the effective cutoff frequency you have to, so it's not simple and it's not again a washing machine. You have to clearly define which is the threshold in your system where the noise is trying to bury the signal you are interested in. The threshold is an issue here and you have to work on this but when you define the threshold you can find your effective cutoff frequency. So now if you try to perform this game using DNA origami this is a conventional imaging and this is the cutoff frequency this low frequency and when you apply a super resolved method you see that you can distinguish them the two dots and but without looking to the image so you are not biased by what you want to find because you know that you have DNA origami so it's easy for you to see clearly that there are two spots but without watching the image you can again find your effective cutoff frequency demonstrating that you had an improvement this is in tune with what you are thinking about your system. This is a recent paper by Christoph Kramer and he had some questions the paper will be shown in the next slide about chromatin DNA distribution in the nucleus so the big mess that you had there in order to be able to say something about what you are watching here which is the meaning of this image in terms of biological question yet you really need to have a clear view of which is the current resolution you have with your microscope you can try to have a line intensity or to recognize something but it's not so easy in this case they are not filaments that you can recognize these are not DNA origami and so what he decided to do in order to have an idea of the resolution is performing FRC with FRC he knows now that for his conclusion on that set of images he can trust in the fact that he is examining something that in terms of spatial resolution is a spatial resolution of 37 nanometer is this enough for answering his question or not and so you can start analyzing this just for giving you this message I think that is something interesting when you start working in the super result for essence microscopy domain so now making shorter a long story you have your fluorescent molecules at distances closer than the diffraction limit and this is what you get with your best microscope cheap or expensive but this is what you can get and this is what you can get with a super resolved microscope and simply comparing in a visual way these two images you see that this report about what's going on in your system is more close to the original and distribution of the object of course you can speculate and you can say something also here I'm not saying that you cannot say anything at this resolution you can use all your information about the system you can use everything so people has done this for 100 years you can do that but I think that you are facilitated if you are in this domain I will not mention among the super result fluorescent method this approach that is structural delumination microscopy this is an approach developed by Mats Gustavson that was developed in the early times of let's say super resolved microscopy and probably not sure but maybe Rainer Reinsmann will discuss about this method on his Topical Lecture on the 23 but just to tell you that there is a method that allows you to move from this condition to this one so increasing the spatial resolution using fluorescence and using an approach that is in the frequency domain we can discuss in case about this later but I knew we'll have a discussion maybe with Rainer but this is one of the method in the very same scenario of super resolved method I will not talk too much but only showing you another new method that is not related to the super resolved method that are the central core of this lecture and are related with the Nobel Prize given in 2014 and this method I like because I have the excuse to bring to your attention my granddaughter because she has this question can we install polymer chains of swallowing material here she has a problem with diapers so the question is related to how much volume increases when she's providing liquid and other substances to this object now I'm telling you about this because there is a method called the Spunction Microscopy that is a sort of diaper for your biological object this was introduced by Ed Boyden that provided me some slides and the idea is making a souffle with your sample so you have your sample you label your sample with fluorescent molecules then you introduce you start the process of gelification that cross links with the fluorescent molecules then you place then you start adding water as Rainer is doing and the system grows keeping homogeneity in the three directions and having an effect on the cross linked molecules fluorescent molecules this is the following in case they were closer than the diffraction limit now they are physically moved apart and so you have them at distances that you can detect with the diffraction limited system there are a lot of limitations in this application you can say whatever you want in terms of the preparation of the sample always have in mind what you do to your sample when you prepare them for the electron microscope before exhibiting too much criticism on this but the system is a structural chemistry and so you have this object here this dimension and then you have your diaper growing overnight for example you have some speaking about some technicalities you have some technical problem because you can imagine that at the beginning you had a small sample under your microscope and now you have a sample that is five times larger larger than the one before so it's important technically speaking that you are able to perform most hiking probably you cannot use the microscope in the very same way you have to move it up but you can use any microscope any wide field microscope you can have in your lab in order to increase your ability in detecting things that are at distances closer than the diffraction limit at the beginning of the story and you can go through there are only I think few papers one is on science one on cell and one on natural methods and another one on natural nanotechnology so four papers about this method the main difference between the very first paper science paper is that in the science paper they are mentioning specific fluorescent molecules so this was the main drawback when you move to the natural methods paper and specific molecules and they were destroying everything within the sample in the next there is something that is a little bit more dirty in terms of final object and this is that you can use any fluorescent molecule and then you cut links in the sample you don't destroy everything you cut links and so you allow your system to expand this is what we have done we tried we decided to start with this method so this is the pre-expansion and this is what you get after expansion and you see that you have visually you have an increase of resolution and so the just in case you can be interested we started working with this protocol and we found this so this is what you do you have these steps that are labeling gelation and digestion and then you place your sample overnight and then you get a sample increase in terms of volume this fixation for some of you from chemistry or biochemistry interested is this you can perform any labeling and this is the physical object you have in your hands and you have a technical real technical problem because now how can you address the question how many times did you enlarge the sample you need a region in the initial image and the region in the final image and so comparing them you can say something about the expansion it's not so easy to find the very same position when you're performing this series of operation in the system I have to say that this is a student in our lab it became particularly skilled in finding the very same cell from the pre-expansion to the post-expansion and so this is water effect and this is your sample growing and this is the effect so this is the pre-expansion so this is what you can get and this is what you can get after it's not a big increase it's not unlimited resolution but maybe depending on your question and the condition of your sample this can be enough and you can simply use the microscope you have in your lab any microscope again here sorry for this again the image you have seen before and then we decided to have another step here and the further step is to apply a super-resolved method that we will discuss later to the expulsion microscopy object why? because in this case I can expand and then I can move to a super-resolved method that I have as available not pushing too much with the method because I already start from a situation that is two times or three times better than the initial sample and so maybe you can get a further advantage in using a super-resolved method across and expand the sample with respect to the original sample in different colors and so on so you know that so we start so this is just to premise for other methods that I will not discuss now and you have your eye conventional microscope and here where you are theoretically unlimited again limitation comes from the fact now that you want to work room temperature and on living systems you know probably these three guys but what you want to bring to your attention is this sentence there are all the keywords a very short sentence that tells you everything about the motivation there is the development so work started in the 90s there is microscopy and there is the keyword for chemistry that is fluorescence that's the keyword we discussed yesterday about fluorescence and if you have a look to this review paper by one of them, Stefanie there is this sentence that well I found for the first time referred to the fraction barrier that is the fraction barrier is crumbling not overcome it's not surpassed it's not violated it's crumbling it's gretolata now it's difficult to find another verb so we decided to introduce this circumvented the rule is this one you simply need to be able to preclude the simultaneous emission of adiacin so object closer than the fractional limit spectrally identical fluorophores simple if you're able to preclude this you're able to perform superazole fluorescence microscopy but this is relevant also because does not only apply to fluorescence and not fluorescence but you can apply this to any state that you're able to control would be absorption or not absorption if you're able to place in a very specific way something like diaripottermontal something becoming transparent you can work in transmission or you can play with scattering no scattering spin up and spin down is in your hands what you're able to control and what you can get and so when you have your beam waste so beam from your microscope from your illumination you have these objects here and your limitation in terms of resolution is that you don't know which one of these object is responsible from which position is responsible for these fluorescence Mörner and Betzig decided to find a way for convincing molecules not to emit all together but to emit in discrete clusters made of a sparse molecule distribution of molecules this means please don't single together only few of you but in sparse positions so one here one here a distance is larger than my limits in finding you and if you play this game for me it's easier to find your position and when I'm sure that you are the singer and if you continue singing I can refine my knowledge in terms of your localization this is what is that of course I want to determine positions of all of you I have to wait that the first singers stop singing and I ask next singer to start singing and so on and this is with fluorescence I need several steps to do this the other point developed by Stefanelle is something that is an improvement with respect to what you do with confocal microscopy also when you move from wide field microscopy to confocal microscopy to scanning system you are improving the spatial resolution a factor of square root of two all directions with a confocal microscope and if you think what's going on is that you are adding an additional information to your acquisition scheme because when you perform scanning and confocal microscopy you know the position of your laser beam you don't have this knowledge when you perform wide field microscopy and so your resolution is the resolution now here you can add something in case of STED what is you are adding a second knowledge that is related to the fact that we will be using a second beam and we know exactly where this second beam is and which is the region that this second beam is perturbing in order to restrict the signal to the region you are interested in but if you think in terms of channel of information that you have at your disposal you have a diffraction-limited system plus additional information about what you are doing about the overall system you are using for forming your image and so here in case of the Merner-Bethic approach and in case you know that you are dealing with a single molecule that's your target this is the reason why yesterday we discussed more on how to recognize a single molecule so when people are performing single molecule localization microscopy super resolved single molecule localization microscopy behind there is the fact that they are sure that they are dealing with single molecules it's not something again it's not simple it's not that you switch on and you say okay let's let's collect our single molecule imaging and for example also in cases when you want to track single molecules you have to know that this is a single molecule and so you have to use all the knowledge or all the tools we have shown yesterday in order to learn how to recognize a single molecule and to be sure that this is a single molecule and then you can start with your imaging and if you know that the emitter is a single molecule you can refine the knowledge that you have in terms of localizing your molecule of a factor square root of N where N is the number of photons that you are able to collect from your molecule in case your molecule is able to send you an infinite number of photons you are very very precise in terms of localization but think that in case you are in a real case that your molecule is able to provide 10,000 photons 100 times your precision of localization with respect to the conventional approach if you move to the and I will be back of course on this too on the stephanel method you have your second beam now there is an analogy also with confocal microscopy that is related to the fact also in this room you can increase your ability in watching a colleague in this room without spending money please remember I come from Geneva so don't ask me to improve to increase light those here there is an associated cost but how can you play this game can we agree that if you play in this way so you confine your observation to a point in a physical way contrast increases and then if you want to see everything you simply scan and when you design your experiment you know that what you have to pay for this improvement in terms of contrast is that in case some of you is moving too slow or too fast I can see air twice or never but this is part of the design of your experiment so in this case the second beam graphically speaking but not physically speaking is like shrinking the region from which the information is coming impinging the detector that is detecting photons so when you increase the power in the second beam the graphical effect is like shrinking this so you can in the center of this you do nothing and you see the original fluorescence in case there are objects there and so if you write something here what you see is something like this and then when you perform this single molecule you see something like this and again with the other method you see something like this and here you can appreciate immediately something this was your system with some information in the fog now not only you can distinguish that is written IIIT but you can also start thinking that you can count molecules assembling this information region by region so this is not only an improvement in spatial resolution but in your quantitative ability so single molecule approach let's start with this let's try to be faster now you ask the molecules to come to your attention not all together if you want you can start from this these are nice movies that you can find on the websites and that tells you what you are doing when you are collecting photons or in general information from a source of information when you perform this game and you have a limited channel in terms of view you can start from this relationship if you want if you like but what is relevant here is that if you analyze this in this case so you are the error or the uncertainty that you have and you consider your weight of your channel and the amount of information you are collected you can come you can go here and that is something that reminds you what is known as the a formula at the end about your uncertainty that you have in determining the position of an object in space and this what happens here is that you have these single molecules that are at distances larger than the diffraction limit this is the profile you have distribution that you have and if you continue collecting photons from this you can refine this information in this way again keyword is in the preparation of the sample and in the fact that you are sure that you are collecting data from a single molecule from a single fluorescent source if not you cannot say anything these are Betzig and Arald S some way when they developed the microscope I think Arald S house was because I think it was not allowed to build the microscope in the room and instead of watching the system in this way like when you watch the sky they were able in convincing molecules to emit not all together but one by one they were able to detect their position in a very precise way get in the advantage that you know today so these are molecules convinced emitting not all together and at the end you have an improvement in terms of localization precision for each of these objects that allows you to have at the end what we call today right or wrong that is a super resolved image so this was the paper by Betzig Betzig and Das in 2006 and I bring to your attention this fact that this date submission date March 2006 you will see some other paper related to this topic in the very same year but please remember this data and you also can see the public acceptance date and publication date so what they did what they did was to collect an increased number of photons from single molecules when you collect this number you can get this precision of localization when you collect this you can improve and you can improve again this was the scheme so you have your molecules coming not all together you can put them together getting a conventional image or I'm sorry I go to the next this larger and if instead of putting all together in a single image you put here something related to your localization precision that is given by this relationship at the first approximation a better description of the position of the objects in the sample here I don't mind all these parameters now I've shown before what they are but the relevance one are N that is the number of photons collected from the single molecule and unfortunately B that is background so our photons coming from something that you don't know that you are not interested in and then there are some technicalities about pixelation and about the system you are using in the very same year but submitted in June there was this paper by Sam S not Arald S but Sam S playing the very same game on the cell Betsy and Arald S were very bright and clever in finding their sample so they decided to use a very thin sample 100 nanometer using for excitation not a conventional microscope but a tiered microscope so confining the excitation in a very small region and the reason was that they wanted to eliminate as much as possible background Sam did the very same game in a biological cell in a physical journal publication was later a submission was later and submission too in the very same year Sho Wei Zhuang published a similar paper but here you can see that this was submitted in July the very same year and here the mechanism is different is not controlling two states like bright and dark but two colors doesn't matter the final result you want to have is that you don't want current acquisitions in something that is disturbing your localization precision so it doesn't matter if you're switching across colors or bright and dark and she performed this game and so this is what you got before and this is what you get after the application of this what you have to pay 40,000 frames and this number of points to be localized 40,000 frames because I'm asking all the singers not to sing all together one here, one here, one here then stop, then start again at the end I mix and they get sung and so I need this number of frames we can discuss about the stop criterion in collecting frames when do you stop collecting frames when you recognize the object this means so my question now is if you recognize the object and you stop when you recognize the object why did you perform imaging you already knew what you wanted to have to be a criterion for this and I want to bring to your attention this that maybe is the only thing I can tell this morning about label free that is there is this recent paper by Bachman group that is a super resolution that is using intrinsic fluorescence so it's a label free method for your sample you can exploit some intrinsic fluorescence they decided to call this in a different way so photo localization microscopy instead of single molecule doesn't matter the only thing behind is that the auto fluorescence is in some way driven by the dense local density of DNA so take care in preparing your sample and in defining your experiment because it's not that you have auto fluorescence from all the DNA you have at your disposal but you have auto fluorescence also as function of the availability of detectable auto fluorescence that is related to the density discuss about this method I want to discuss about this localization this is your sample you are very good in inserting your labels now you are very precise in localizing their position and you get something that is close to your sample good same sample you are not so good for some reasons in localizing very well your labels could be that your molecules undergo photo bleaching faster and so you are not able to collect a large number of photos and so you get this that is close to this but not sure that information is the same or you are poor in labeling and precise in localization and you get this or your molecules are moving while you are performing your frames 40 thousands you can imagine that even for a fixed sample you can have some motion or if you want to go to the living sample you have to take into account this but this is not so negative as it seems because in the last 10 15 years people became skilled in particle tracking and so with the living sample and with this method you can find a solution for using your tracking ability in repositioning the objects where it has to be so nothing is completely lost due to the motion of the molecules then you have to decide how precise you want to be so the number of photos you collect or the number of frames and you have different images at your disposal so you are sorry for but you can see that here you have some molecules and your and this is network of your molecules here this looks like a conventional image with the exception that you know their position with a precision of 15 nanometers we can say we are pointless painters and you can decide if you want to spend your 2,000, 40,000 frames, 200 I don't know how many frames in order to get a view that is not this one but is this one to answer your biological question if this is relevant do that if for your question this is enough no reason for collecting 40,000 frames and I've shown images coming from satyr skeleton filaments something that you can easily recognize something that is flat something that is on a coverslip really not a revolutionary environment but a relaxed environment for the cell but now we are more interested in for example in aggregator cells yes because if you want to understand now we have a tool that allows you to find a position in some case dynamics related to single molecules so why not start using these for aggregator cells and not at the interface with the glass this is a tumorous ferrite so we decided to couple two methods I did not discuss yesterday about this method but this is a method that is usually I will comment later about this is cutting your sample with a plane of illumination collecting information 90 degrees apart why people is doing this why we are doing this because as you have seen in the Betzig slide the major enemy is background and so when I'm trying to collect information collecting from this direction I have a lot of background coming from adjacent planes for a big object so I'm in troubles with my precision localization but if I am able to reduce in a significant way background I can improve my ability in precision localization so there is a a way for doing this with some drawbacks but it works and then within this plane of illumination thickness is from one micron to two microns can be even sharper you can apply here super resolution detection and so when you navigate through your sample with this plane up to 100 now today 200 nanometer you can collect single molecules plane by plane single molecule position and then you can get a localization position of 28 nanometers looks good but is bad is bad because these 28 nanometers are lost in 2,000 nanometers of the thickness of the plane of illumination is good having this xy but ok so from Shouwaiz Wang came the suggestion of using some asymmetry in the system and you can imagine that this is the thickness of the let's imagine this, this is the thickness of the plane of illumination and now you have you are 90 degrees apart ok and now you see signature from your molecules, single molecules and when you see that due to the illumination that you are using now so some astigmatism introduced in the system when you see around the shaped signature this is a molecule in the focal region in the focal position but when the molecule is moving apart from the focus of the lens it becomes stretching one direction or in the other direction in ellipsoidal shape you perform a calibration you have this curves of calibration and they allow you to find the position within the plane of illumination with a better precision that the conventional you can get with your microscope and so you can perform also z-axis super resolution like in this case 65 for molecules that are in the membrane of a nanocapsule doesn't matter what it is but it is an artificial sample small then when you go in your thick object you are in troubles why? because in order to perform this calibration you can be you can try not to understand the problem and so you can say ok come on let's elaborate using beads great but beads provide an enormous number of photons are available your sample is real so when you have a thick sample our suggestion is that you go inside you find a region where you have a subresolved cluster of objects and you perform there your calibration and so you have the real situation also in terms of release photon in this case the situation is worst so it is 63 141 along the z-axis but you can call this again super-resolved imaging and so you can find the region in your cluster of cells and in the region of interest you can have your super-resolved image localized in a precise way and so you can integrate yesterday we were discussing about two photon excitation and you see that here you used as first step photo-activation of your single molecules I'm sorry that I missed telling you what you didn't ask and I missed to tell you how you convince molecules to emit not all together what you're using are photo-activatable molecules in photo-activatable molecules you can control states dark and bright so when you decide to have them bright you shine some light this induces a photoconversion and starting from that point you have them bright how to have a sparse set of molecules reducing you need the photophysical knowledge of your molecules but reducing a lot the photo-activation intensity so if you have a very poor photo-activation your probability of photo-activating is poor and the probability of having the photo-activated the distance is larger not glossarized but the distance is larger than the threshold limit is higher when you perform this game in a thick object using single-photon photo-activation we have the effect we have discussed yesterday so you have a lot of scattering first of all because of the wavelength and then you have a lot of noise or photo-activation from position that you don't like but what about using two-photon photo-activation in this case you are moving to the red you can penetrate more in your sample and even if you have a scattering because the sample is scattering due to properties a large sample is a scattering sample you can be homogeneous in the thickness of the plane of illumination due to the fact that you are killing scattering that you are not able to kill when you perform single-photon photo-activation so ok so this is what happens this is the single-photon two-photon now when you have a scattering sample scattering sample this is the condition with single-photon and two-photon we did this and you can perform this game in terms of photo-activation in terms of photo in terms of two-photon excitation is nice but we discussed yesterday about the fact that two-photon excitation is not the best way for a stratinphoton from frozen molecule because the efficiency is not very high there are a lot of other properties but not the efficiency in stratinphoton now I move to the last part and I change I move from single-molecal localization super-resolution to targeted super-resolution so the stephanel method is where I use the knowledge of the position of the first excitation beam and of a second beam that I call depletion beam we will see about this immediately again you can find I don't know why this time I decided to find these movies but they are available and they are nice and different websites these cartoons what about the method where excitation spectrum for essence and you remember that is independent from the region where you are performing excitation this is your probability of emission you can see this also as the possibility of the molecule to explore energy states and coming back to the ground state these are the possibilities let me say that we can say that this is green and this is red just to have some keyword that we can use when you have so you have your excitation you bring the molecule to the excited state when they are in the excited state they do something for 10-12 seconds telephone call so there is energy some energy lost and so you have in case they are fluorescent you have an emission at a wavelength that is longer than the excitation that's everything that we discussed yesterday fine now when your molecules are in the excited state since they are fluorescent molecules is natural for them with the quantum mill that you can define yesterday to jump down from this terrace to the ground state they will do that and they will follow most of them they will follow this route and so since you are interested in them and the way they go down also your acquisition of information is confined in this region down from the terrace in this position you set there your camera and you collect information and you see them when they jump down now what happens if someone goes to the terrace and decides to push all of them down on another route except one or two of them that what you see are only the ones that will jump down in the region where you are focused you don't see all the others because they have been pushed out jumping to the rocks or somewhere there and this is what you can do with the second beam you have a second beam here why you select a region in the emission spectrum because you need to use an energy that can be explored by the molecule so you could also put this second beam here or you could put this second beam here but the problem you have if you place this second beam closer to the excitation curve is that you increase the probability of it exciting so you are pushing people down and you are helping people coming up you don't want to do this in a region where you can simply push them down here so this is what you do with the second beam now if you have a look to this process with two round shaped beams you shine the excitation and you collect fluorescence one now you switch on a second beam in the red that is the one responsible of pushing them through another pathway you see that the function of the intensity of this second beam you have a decrease of the senior fluorescence that you see in the green you are depopulating in an artificial way excited states I need only one definition that is intensity of saturation saturation to zero that is the intensity of the second beam that you need for depopulating 50% the excited state this is only the definition method so just to tell you what is IS if you have a look to resolution and we will see immediately after this you find that the strength of your shrinking so the resolution that influences resolution while you are scanning is function of this ratio I divided by IS I is the intensity of the second beam is what you put is what is in your hands like the length and like your laser like everything you have in your hands but IS for the first time in such a relationship related to spatial resolution the property of the molecule you are working with is something related to that molecule or that wavelength is not related to the optics you are using so in case you have friends able to produce for you molecules having a very poor very small IS you can use this laser pointer as the second beam for producing super resolution unfortunately it is not you need a very high intensity I will be back on this topic since you are still pointless painters this is your brush with a conventional microscope with a confocal microscope and this is your brush with a stead microscope why? because instead of providing the second beam as round shaped I provide the second beam as donut shaped donut shaped means to push molecules down in well known spatial region so I know exactly from which position I do this game and I also know exactly that I will do nothing at the molecules that are, I try to do nothing at the molecules that are at the center of the donut the only requirement is that you have at least one zero at the center in terms of intensity then if you shine your donut beam to the wall and you try to measure the dark part system that you are using is the fraction limited and even if you increase your intensity you measure to 100 nanometers because you are using a different but what matters is the distribution of the intensity that is the one that is influencing the resolution that you can get from your profile now think about this even if I'm late but I can manage here you have fluorescent molecules and you are using the second beam for perturbing if you want in the near field whatever you can have in mind locally fluorescence conditions what are you doing I will be back on this later not too much later but later what are you doing what does it mean perturbing you are offering to your fluorescent molecule another possibility to deactivate and going to the ground state you remember yesterday when we discussed about lifetime about the rates of fluorescence not radiative we are offering another possibility we will see the influence of this and then if you want to use the ABBE formula you have this D that is ruled by this relationship until here then you add this element and when you switch off your second beam I you have the original formula so you can use this for trying to estimate which is the final resolution you can get the best is using FRC pointless painters again and the reason why I like this method since ever is because when I switch on a second beam I immediately have the effect I want to have like with the radio when I want to switch to another channel I simply turn it up this is the image you can get and beads, 40 nanometer beads using a confocal microscope and then you switch on the second beam and immediately get this no processing no frames to be collected simply switched on a second beam and when you go into the cell you can play the very same game you can move across your system improving locally or improving over the resolution spatial resolution in your system switching on a second beam and if you were this was you got from synaptic connection this is what you can get even better you can perform this in the real time, real time means the delayed time that you have in confocal microscopy due to the scanning you can have these in colors and along the third axis but now I want to bring your attention to this point confocal means I0, second beam off then I switch on the second beam I'm using this number just to give you an idea of the power, it doesn't matter depends on the real condition of this hour 50 milliwatts on the backfocal plane and I get this improvement 100 milliwatts I get this improvement this is fine comes from the description we had before now let me skip this because I want to now let's try to find ways for reducing the power of the second beam for some simple reasons at least two one is an economic reason if I'm able to reduce the intensity of the second beam I'll need cheap lasers the other one is that it's true that I'm working in the infrared on the tail of the emission that I'm telling the story that systems do not absorb that wavelength in the red and so on but it's not true it's not true that they do not absorb at all but it's true that I'm using an enormous amount of light there so this large amount of light multiplied by something that is not exactly zero makes something this means I can re-excite and when I re-excite I increase the probability of bringing the molecule to the photo bleaching condition that we discussed yesterday so never for essence forever you understand immediately you're interested in super-resolution and in telling at which distances are two objects and one of them is disappearing due to photo bleaching your experiment is off doesn't matter so you really have to take care of this so you want to find ways for reducing so sorry for this at least I'm not interested in this now but I'm interested in telling you about the current method developed and that is the current method of Ilaria Testa in her new group in Stockholm she's also looking for people there and she comes from our lab very proud about this why not using the donut shape beam for inducing photo-switching instead of depletion so the process of depletion is very low probability but the process of photo-switching due to the characteristic of the molecule has a very high probability or very high cross-section and so you really need less light to do this there are some drawbacks one of the drawbacks that I see maybe you know better than me about this is that since you are using photo-switching and the time scale is not picosecond like in the depletion but it's microseconds the system is more sensitive to changes in the environment and so I have to use a smart second a smart donut intensity that is adapted to the different condition changing in order to get the very same shrinking point by point so this is my point about this so if you play this game that is again an open-access paper you can find on my cross-preservation technique you can play a game like the one done by Laria Testa with nanoscopy or living brain slices using low light levels but now even if I have again some other things to tell you but let me bring you again to this aspect perturbation is there any chance you have to reduce the intensity of the second beam? I mean if you reduce the intensity of the second beam the probability you have to push the molecule down to the ground state decreases this means that graphically you can imagine that you have the molecules that you are interested in that are passing in the center you do nothing but other molecules through periphery let's have this graphical in mind can we use now time for getting information about their position? well this is what you do with children when there is a thunderstorm if you count so if you use time you can get information from the distance so special information and you can find this you can really be very sharp in this because you can also take care of the temperature for example of the environmental conditions in doing this your counting and calculation fine what about the fluorescent molecules? well which is the property which is the characteristic that I am affecting using such a perturbation I give the molecule one additional chance to go to the ground state the final effect is that lifetime fluorescence is short-lived and so I send the excitation thunderstorm I start counting and when I do not collect photos arriving too early because they come from the perturbation in the donut the ones coming with the right lifetime are the ones I am doing nothing that comes from the center of the donut and so I can use this temporal information for so this is what is going on in terms of lifetime center and periphery for building my image this is what I can get when I use this gated acquisition this means practically speaking if you need 600 mW for getting 40 nanometer resolution using this approach you can need 50 nanometers not 600 for getting the very same resolution nothing more sorry for being fast here this is 50 mW for doing the same you can perform this also using to photon excitation but now I have a guy from Katania in my lab and most of us are from Genoa and we don't like wasting energy when we spend a lot of money for producing it and Luca Lanzano immediately came in tune with us the point is now in order to improve resolution to have the very same resolution the reduced amount of intensity for the image I select photons with the time of arrival time instead of not collecting them why not classifying them with the time of arrival and so we decided to use this approach and in collecting photons instead of removing them from the image we removed them from the final image but we collected them in order to improve the resolution about their space position so this is the steady image this is the confocal image and since we are performing this game modulating excitation and collecting in a correlated way when we have a correlated signal this comes from the background and so we are able to have the background shape that we can subtract so we can further improve so this is the confocal this is gated and this is what is called split I have no time my chairman is watching to the clock just two minutes but not yes I think 30 also I want to bring to your attention this to photon because it's interesting to go to photon excitation you can use also there are a lot of people using adaptive optics I like all of them but we don't use it at the moment now with two photons you can go deep now you need a second beam or shrinking if this is a second beam as a different wavelength there is a different distortion when you bring this deep in your system with respect to the excitation beam you can calculate this you can do something I'm not saying you cannot do anything but what about using the very same wavelength or exciting and depleting in this case I'm sure that I'm bringing in the very same position the dulled beam and the excitation beam and this is what we did and we got this result how did we manage so in two photons I have a red beam when I have this red beam with very short pulses I have a very high density of photon that allow me to induce to photon excitation so 100 femtosecond pulse width and I can induce the two photon excitation because in this temporal window I have a very large amount of photon if I broaden the pulse width down to 200 picosecond I don't have this chance this chance is very poor because I have a not very high density of photon so I cannot induce to photon excitation but I have the appropriate number of photon for inducing a depletion and so I use the very same wavelength that is in the red for performing depletion with a branch of the beam at 100 picosecond and with the other one 100 femtosecond for inducing to photon excitation you can find some information about developments in these papers and I think that I'm at the end but may I ask you 3 minutes more just to move to something that is cheaper than stead and can give you an improvement like the Sponcho Microscopy that is around 2.5 times in terms of resolution this comes I will be very short is very simple now one of the advantages of super resolved methods is that in several commercial microscopes and also in your own setup there is an improved technical ability in managing different optical and electronic components because they come with the super resolution for example the second beam now we decided to use the second beam not for the plating but the second beam very cheap with normal laser the very same we use for excitation for exciting so you have excitation with your round shape beam excitation with your donut shape beam and then if you subtract what you are doing in terms of excitation you can be sharper than exciting only with a round shape beam subtraction is complicated and the factor that you use is not so easy to be found we discussed a lot with Colin with many other in the lab and as you can see here a different solution came out then there was a guy one of the researchers Ksenia Korobcheskaya she is married with a robotic guy Chinese guy and this guy had a problem in his robot he had to find a way for subtracting information while the robot was walking and terrain was changing so it was something differential so the robot was moving and he had to adapt the robot to the terrain so it wasn't something in a subtractive way but since the range changes in an unpredictable way it couldn't use blocked alpha a blocked factor of subtraction use the dynamic one that is this one and we decided to use the very same for subtracting the effect of the donut beam to the second image what is nice what I like from this approach is that you can use this method to any method so you can use this for fluorescence for reflection for to photon, for second harmonic and you get an improvement in terms of contrast or resolution that is not very high but for some questions could be something that can help you in the siphoning battle what's going on in your image in a very cheap way that is managing the second beam and you can perform this also in the field axis for those of you interested in correlative nanoscopy there are also some approaches combining atomic force microscopy instead but we will not discuss about them now in case you're interested we can discuss later or in the coffee break so I think that I can stop here more recent time and waiting for your questions or discussion in the coffee break time that's pretty much clear