Practical image acquisition
Perhaps one of the most confusing aspects of starting to use a confocal microscope is selecting and adjusting the controls – which ones should one adjust to get the best image? And by how much?
Before attempting to answer these questions, ask a couple of different questions – why are you collecting an image on the microscope and what do you expect to see in your image? More specifically, are you collecting images for quantitation? Is your image to be used for the cover of a journal? Do you have co-localisation? Is the protein localised in the nucleus or in the cytoplasm? By answering questions like these, which microscope controls to adjust (and by how much) will become more obvious.
One comment for new users of confocal microscopes - “you get nothing for free”. This means when one adjusts one control to improve an image (or the data in the image), something else will get worse. Good confocal microscopy is all about understanding the limitations and advantages of each of the adjustments in relation to the questions you are asking without compromising the integrity of your specimen.
The eternal triangle
Remember we mentioned that you “get nothing for free” when doing confocal microscopy. The “Eternal Triangle” describes this conundrum particularly well:
If maximum Speed, Resolution and Sensitivity are represented at the apexes of an equilateral triangle, then all possible combinations of these three parameters are contained within the space occupied by that triangle.
Essentially, this means that if one needs to obtain the maximum image acquisition speed, then either resolution and/or sensitivity must be compromised. This might be necessary for imaging fast moving objects in the specimen.
If maximum sensitivity is required to image weakly fluorescent or faint specimens, then either speed and/or resolution must be compromised.
If the objects of interest are very small and maximum resolution is required, speed and/or sensitivity must be compromised.
Unfortunately, this is the nature of confocal microscopy. It is up to you, the microscopist, to decide which parameters are most important and which ones are of least significance to the questions you are asking about your specimen. This will then guide you towards choosing the best image acquisition parameter values for imaging your specimen.
What is important?
For whatever reason you are using a confocal microscope, it is essential that, by using the microscope, you are not altering the very thing you are trying to measure. Intense laser light has the ability to alter biological molecules and structures; local heating from the laser may modify the environment and cause the sample to move; live cells can respond to light and some organelles (chloroplasts) and cells in the retina are designed to react to light. It is essential that experimental design takes into account these possible effects.
Except in special circumstances, photodamage of biological molecules must be avoided at all costs. This will often require lowering the illuminating laser power or using a laser wavelength that minimises any photodamage.
Photobleaching is the irreversible destruction of a fluorescent molecule so that it no longer fluoresces. Photobleaching results when a fluorescent molecule is in its excited semi-stable state and is then hit by a second photon causing permanent damage to the molecule. This mostly occurs when using higher laser powers for longer times. Apart from decreasing the strength of any fluorescent signal, biologically reactive molecules are often produced as a consequence of photobleaching, thereby potentially altering the very events one is trying to measure. Reducing the laser power (and pixel dwell time) can minimise photobleaching. It has been calculated that 150μW of laser power at the sample is sufficient to fully saturate all fluorescent molecules in that sample (Pawley Handbook), thereby rendering them more likely to be damaged irreversibly.
- Signal:Noise Ratio
To image a specimen in the confocal microscope, a minimum number of photons must be collected at the detector. These detected photons are converted to photoelectrons and that signal is then amplified and displayed as a pixel intensity value in our image. These values are usually referred to as our signal. Inherent in this process are electrons that are also produced within the detectors and amplifiers. However these electrons are not directly related to the photons emanating from our specimen and are, therefore, usually referred to as noise within our microscope system. While the absolute signal is important, it is the ratio of signal to noise (often shown as S/N ratio) that is critical. Sometimes it is easier to reduce the noise in a detector system (perhaps by cooling the detector) than to increase the absolute signal emanating from our sample. Such a system would give a higher S/N ratio and, therefore, more robust data and a better image.
One of the most difficult tasks when teaching confocal microscopy is to convince trainees of the importance of using appropriate control samples. Such samples are critical, especially when using new antibodies or looking at new specimens. Why is this?
- Simply by adjusting the amplifier microscope controls, it is possible to make a negative sample look positive. A good procedure to guard against this possibility is to image a “positive” sample and then, without adjusting any microscope controls, image a suitable “negative” control. Open both images on the same computer monitor, side by side, and compare the images. If the “positive” looks bright and the “negative” image appears dim, then it is likely you have some specific signal. If the images look of similar intensity, then it is unlikely any specific signal exists in the “positive” sample. Comparisons like this also allow a very objective means to evaluate labelling protocols when titrating antibodies.
- Many samples will show some level of autofluorescence. It is essential, at least when starting with a new specimen, to image a totally unstained sample. By imaging this unstained sample together with a “positive” and “negative” sample all collected at the same microscope settings, the levels of autofluorescence, non-specific signal and specific signal can all be evaluated.
- If no signal can be seen in a “positive” sample, it may be necessary to label and image a sample known to contain the molecules or proteins of interest (a positive control).
The Confocal Pinhole
The major function of the confocal pinhole is to block or reject out-of-focus light from reaching the detector. If the confocal pinhole is small, out-of-focus light emanating from just above and from just below the focal plane is rejected by the pinhole. If the pinhole diameter is increased in size, more out-of-focus light from above and below the focal plane can reach the detector. If the pinhole is opened to its maximum value, the confocal microscope can produce images that are similar to those produced by a widefield, epifluorescence microscope.
One of the most difficult tasks in confocal microscopy can be actually focussing on your sample. It is always good practice to first focus on the sample using brightfield illumination. Additionally, using widefield, epifluorescence microscopy is helpful because out-of-focus light is usually easily seen even if the sample is grossly out of focus. However, when in confocal mode, if the sample is just slightly out of focus and the confocal pinhole is closed to a small diameter, it is possible that no light at all will reach the detectors. When no signal is visible it is nearly impossible to decide what the problem is - (1) Which direction to adjust the focus controls? (2) Is there any fluorescence on the sample? (3) Do the detector amplifiers need to be increased? or (4) Has the sample moved out of the scan area? Simply opening the confocal pinhole to the maximum diameter will allow any out-of-focus light to be visible and will increase the total signal reaching the detector. When some signal can be seen, it is then simply a matter of refocussing and then improving the quality of the image by adjusting other controls. When this has been done, slowly close the pinhole while continuing to adjust other controls.
Good confocal imaging is usually achieved by starting with some signal and then gradually improving the quality (and often the intensity) by making small, incremental adjustments.
Most commonly, to achieve the best resolution, a confocal pinhole diameter of 1.0 Airy Units is used for imaging. Reducing the pinhole size to less than 1.0 Airy Units will give better resolution. However, because of the significantly lower signal level, this is rarely done when imaging biological specimens.