Introduction to microscopy
Zacharias Jansen designed the first microscope but he didn’t observe any biology. Robert Hooke is the first to
describe and coin the phrase “cell” when observing a slice of cork using a microscope power of 30X. Hooke observed
box- shaped structures, which he called “cells” as they reminded him of the cells, or rooms, in monasteries.
In 1973 Antony van Leeuwenhoek used a microscope to see cells in blood, sperm and other biological material. This is
the start of cell biology. It then took 150 years before cells were finally acknowledged as the basic units of life.
In 1880 august Kohler had worked out light source and condenser position to obtain the best image projection.
Ernst abbe made the clear difference between magnification and resolution and made the formula that describes the
resolution of light based on the diffraction of light.
To image you need objectives and hardware, mostly Germans have been contributing. Ernst Leitz introduced a
revolving mount (turret) for 5 objectives onto a microscope. Ernst Abbe himself designs apochromatic objective that
brings red, yellow and blue into one focus. So that you can image more colour than one.
Along with the developments in light microscopy there were also developments in electron microscopy. As Abbe
realised that the resolution of light microscopy is physically limited in the 1930 people started looking into electron
microscopy.
Ultimately the power of the electron microscope was not realized until the 1950’s when ultramicrotomes were built.
These instruments could slice pieces of biological materials as thin as 500 angstroms.
In the last 50 years there have been four noble prices in the area of microscopy. One is the discovery of the green
fluorescence protein which has been important in many biological discoveries.
In 2017 three guys got the Nobel price for developing cryo-electron microscopy for the high-resolution structure
determination of biomolecules in solution.
Resolution and magnification
Magnification is how much bigger a sample appears to be under the microscope than it is in real life. Magnification is
determined by objective magnification and eyepiece magnification.
Resolution is the ability to distinguish between two points on an image – the amount of detail. Resolution is
determined by the numerical aperture of the objective and the wavelength of the elementary particles used.
Magnification often scales with resolution. Often, the higher the magnification
the objective has also the higher numerical aperture has. But this is not always
the case.
Plan Apo has a magnification of 40X with a numerical aperture of 1.3. The plan
Fluorite has a 60X magnification with only a numerical aperture of 0.85 and
thus a lower resolution. So if you want to see more detail it would be better to
use the lower magnification.
Ernst Abbe equation presented in 1873 that the only two factors that influence d
(resolution (distance between two molecules that you can still resolve)) is the
wavelength (lambda) and 2NA (two times the numerical aperture). In practice, the best NA you are going to achieve
is 1.4, the wavelength we use is 500nm (visible light spectrum) equals to a resolution of 200nm.
With a conventional microscope you can get a resolution of 200nm.
Once light has passed through an objective it will not go back to a single point due to the
diffraction of light. Once light passes a small slit or objective it can never be focussed back
onto a single point because of the electromagnetic elementary behaviour of photons. What
happens is that you get this airy disk. You have a point source, small source of light (a
molecule) that emits light. A molecule is only 1nm. This goes through an objective and
because of diffraction it will never go back to one point, it will displace a disk, a distribution of photons onto a
camera. The histogram of these photons is called a point spread function. This point spread function determines the
resolution of the microscope.
The point spread function is the response of a microscope to a point source. The size of the disk is about 200nm. Any
sample that you image is convoluted with the point spread function. So you have a ground truth and is convoluted
,with a PSF and that is what you observe with a microscope. That is the amount of detail that you can be aware of.
The smaller the PSF is (dependent on the NA and wavelength), the more detail you can observe.
The numerical aperture is the light capturing ability of a objective. The more light you can
catch, the more information you have and the better you can know where the light is coming
from, the more you can focus it back on your camera. Image a is a low NA objective and so
you have the ability to only capture a certain portion of diffraction pattern and you get a big
airy disk on your camera and thus a low resolution. The resolution is low because each point
on your sample will give a blurry dot on the camera. Image c has a high NA objective and can capture light from a
high angle and thus a better focus effect on the camera
The light spectrum that we use for fluorescent microscopy is from 200nm to 800nm but you can only see from 450
and 700. The wavelength of light determines the diffraction. If you have photons with a longer wavelength they will
have a bigger diffraction pattern on the camera which will result into a bigger PSF. Smaller wavelengths resolve into a
smaller diffraction pattern and results in a better resolution.
The notion that the wavelength of a particle determines the resolution that you can obtain with a microscope is true
for many microscopes, it is even true for radio telescopes. Radio waves that we observe from the big bang are
extremely large but also there the resolution is determined by the length of those radio waves. This is also true for
smaller ones like electron microscopes. Electrons have a way shorter wavelength, single nanometres as compared to
hundred nanometres. With an electron microscope you can obtain a way higher resolution than with light
microscopy. The basic principle of light and electron microscopy are not even that different.
You have a source that produces the elementary particles, which in the case of
light microscopy is photons and in the case of electron microscopy is electrons.
Then you need something to focus it. In light microscopy this is easy and
manageable because you can focus photons with an objective, this is not possible
with electrons because they are immediately stopped by objectives so you use
magnetic lenses that focusses the electrons onto the sample and back onto the
detective. The resolution of electron microscopes is a hundred times bigger than
conventional microscopes and thus you can see an immense amount of detail like
mitochondria and other cellular structures.
Phase contrast microscopy
Even without any fluorescence or labels there is quite a bit that we can learn from cells. By simply shining light on
cells (white light) will give a diffraction pattern. Light will bounce of at certain cellular structures and this is what gives
a contrast. Bright filed microscopy is simply shining light onto a cell and already you can see the nucleus, some
vesicles, the front and back of the cell. With this you can check the
proliferation, whether they move, are they close together etc. So
already without labels you can learn quite something. Even more if
you not only shine light but mess a little bit with the light such as
the angle of entrance changes you can see even more, this is called
phase contrast microscopy or differential interference contrast
microscopy and you can see even more detail. Depending on how
you interact with the light path you get different types of information.
You can mess a little bit with the light path and create different interference patterns
and get different types of information from the microscope. This is called phase
contrast microscopy.
Fluorescence microscopy
Cells are transparent, we don’t see the molecules. We can detect the biological
structures by fluorescent labelling.
Fluorescence is the emission of light by a substance that has absorbed light or other electromagnetic radiation. In
most cases, the emitted light has a longer wavelength, and therefore lower energy, than the absorbed light/radiation.
Any elementary process there is some energy loss going on, in this case in the form of e dissipation.
,Fluorescence is a property of a material in which it is possible to excite an electron, to hit the photons into an
electron and put them in a higher state, the electron will almost immediately fall back
into its original state, thereby lose energy again and a photon is lost.
We have an electron that is circulating the core of an atom (proton and neutrons). The
photon bounces it into a higher energy state (different circle), and almost immediately
(nanoseconds) it falls back into its original state and there you get the photon and
difference in colour.
Jablonski diagram shows two states, ground state and excited state. The arrows indicate what
happens to the electron. You have excited light that puts the electron from the ground state to the
excited state. Then you have internal energy dissipation and then it goes back. And that is the
photon you get back from the sample.
The difference in wavelength between the excitation light and emission light is the stroke shift.
For any fluorescence material you can draw this histogram. You can draw the distribution of the
excitation photons. This material is excited maximally at 500nm. And each
material has a emission histogram. For this material the peak emission
photons are about 540nm but there are also photons with 600nm coming back from the
sample. Stoke shift is the difference between excitation and emission peak.
Types of fluorophores
Organic dyes are typically used for fixed material. In the course of
the past 30 years there have been dozens organic dyes being developed. The
characteristics of each dye is that they excite and emit on a different wavelength. So if you
want to do a multicolour experiment in a fluorescent microscope you can use three
different dyes so you can set the colours apart in a microscope and do a triple labelling.
If you look at such dyes they all have something in common. They are all constructed
around synthetic aromatic organic chemicals. Aromatic means that you have a ring
structure. Fluorescence is simply a property of a molecule for a electron to get excited.
These organic dyes have this property. Stability has been optimized for those dyes. Simply having a dye does not
mean anything, if you want to label you need to bind an organic dye to a biological
macromolecule like a protein or nucleic acid. You have also dyes that localize within a specific
structural region without any specific biological macromolecule, these happen to have
binding molecules for the cytoskeleton, mitochondria, Golgi apparatus, er, nucleus. There are
also dyes that can monitor dynamic processes and localized variables, the colour then can
depend on pH concentrations, reactive oxygen species, cellular integrity, endocytosis,
membrane fluidity, protein trafficking. For any biological process there are probably dyes that respond to it and
fluoresce in a specific colour.
Fluorescent proteins are typically used for living cells. Fluorescent proteins have been derived from the jellyfish
Aequorea Victoria. Deep sea creatures, due to the absence of light, are known to produce their own light.
Fluorescent proteins don’t produce their own light. The jellyfish has a protein that using ATP can produce photons
which can be converted by fluorescence proteins to produce the colours. They realised that if we have a fluorescent
protein we can fuse it to any protein of interest and then visualize the distribution of the protein in the cell.
You fuse it genetically to your protein of interest, you express it and then you can see it in the
cells. This way you can label all kinds of different proteins and make this multicolour images to
get an idea of what is going on in the cells.
The most remarkable thing about this fluorescent protein is that it fluorescence without any
cofactors. There is no need for ATP, calcium etc. there is nothing necessary for GFP to fluoresce
except that you have an intact protein and some light. The true fluorescent part of GFP is three aminoacis SYG. This
tripeptide is responsible for the fluorescent part. It is no coincidence that there are two aromatic rings in GFP as
there are also aromatic rings in the organic dyes.
, The GFP chromophore is formed in an autocatalytic cyclization of the tripeptide
65SYG67 sequence, it does not require a cofactor, and is followed by an oxidation
of the intrinsically formed structure. This means that GFP fluoresces in the
absence of any other proteins, substrates, or cofactors. This makes it very
convenient, you can express this protein in all kind of cells.
GFP is formed like a barrel, it has 11 beta sheets organized in a beta barrel. Inside
the beta barrel is where the chromophore is buried. If you want to mess with the
fluorescent spectrum you need to get to the tripeptide. Different mutations in the
tripeptide where aromatic rings still exist gives different colours
fluorescent.
You need to make constructs, fuse the GFP to the c or n terminus of
your protein of interest and then express it in the cells.
Quantum dots are tiny particles or nanocrystals of a semiconducting
material with diameters in the range of 2-10 nanometers. They can
be functionalized by conjugation to biological molecules. Size is
comparable to an antibody. Quantum dots are not often used. It does not penetrate a cell by itself. Quantum dots are
fluorescent by itself, extremely stable and they are very electron dense. You can also use them to enhance the
contrast in an electron microscope. Quantum dots can be used in correlated light and electron microscopy. Same as
with the organic dyes, if you just throw them into cells, nothing will happen. You have to conjugate quantum dots to
biological macromolecules like an antibody so that you can specifically see structures in the cells.
Photobleaching is a photochemical alteration of a dye or a fluorophore molecule such that it
becomes permanently unable to fluoresce. Caused by cleaving covalent bonds or non-specific
reactions between the fluorophore and surrounding molecules. Any fluorescence material
have a certain stability, each time that you excite a population of molecules a portion of the
molecules will stop being fluorescents. The electrons in these molecules cannot be excited
anymore by other photons. Behind the microscope you will see that the intensity of the signal
becomes less. When we do experiments we go for the most stable fluorophores. Fluorescein is
the least stable, it bleaches the fastest. Alexa fluor 488 is the most stable. Sometimes you just have to deal with the
fact that it is fading. Organic dyes are the smallest but you have to conjugate them to macromolecules. Qdots absorb
the best. For GFP you need a plasmid because it is a protein. Qdots do not photobleach.
Labeling strageties
Just throwing dyes or fluorescent proteins onto cells in not going to work, you need to
label them.
In situ hydridistaion means that you have a single stranded DNA or RNA piece which
you conjugate to a fluorophore. The single stranded DNA or RNA piece is
complementary to something that is present in the cell. If you then fix your cells,
denature all the DNA and you incubate the cell with complementary DNA that is fused
to a fluorophore you can use that to label specific sequences.
In a tissue sample you can visualize specific transcripts. If you want to know where the transcript is of your protein of
interest you can use in situ hydirdistation and you can see particular regions where transcripts are positive.
Immunocytochemistry use antibodies and something with fluorescence. You can do a one-step or a two-step (direct
or indirect) immunofluorescence.
In direct you have an antibody that recognizes your specific antigen and that antibody is already conjugated to a
fluorescent dye. So you only have to incubate your specific cell with that antibody and you can already visualize your
protein of interest.
In a two step you have your primary antibody which is not conjugated yet to a
fluorophore but then in a second incubation step you come with a secondary antibody
and label such that you get fluorescence in your sample.
In direct double labelling you label one antibody with a green dye and another which
conjugates to another protein with a red dye.
