1
It's a Small, Small World: A Brief History of Biological Correlative Microscopy

Christopher J. Guérin1, Nalan Liv2, and Judith Klumperman2

1 VIB Bioimaging Core, Ghent, VIB Inflammation Research Center, Ghent and Department of Molecular Biomedical Research, University of Ghent, Belgium

2 Section Cell Biology, Center for Molecular Medicine, University Medical Center Utrecht, The Netherlands

1.1 It All Began with Photons

Light microscopy (LM) is arguably the oldest technology still used in scientific research today. Until the mid-1600s, the world of structures smaller than about 400 microns was unseen and unknown. While the principles of using lenses to magnify were known as far back as Euclid (c. 300 BCE) [1], microscopy had to await technical developments in the manufacture of lenses and the casings to hold and position them, before they could be used to extend the power of human visual resolution. The earliest published description of a biological sample viewed using a simple one lens microscope was probably in 1658's Scrutinium pestis physico-medicum [2] written by a German friar Athanasius Kircher. In this manuscript he describes the presence of "little worms" in blood that he associates with disease; thus anticipating the germ theory by almost 100 years.

Around the same period, Dutchman Antonie van Leeuwenhoek used his single-lens microscope to examine samples of mold, bees, and lice, and reported these and other observations to the Royal Society in a series of letters beginning in 1673. It was when he went on to look at samples of blood, tooth plaque, and sperm that he observed that individual small structures that moved of their own volition! When he reported his observations in a letter to the Royal Society in London in 1676, they were met with great skepticism. In 1677, a delegation was sent to determine if he was brilliant or demented. Having vindicated his observations, he was elected to the Royal Society in 1680. However, while the best of van Leeuwenhoek's microscopes had an impressive maximum magnification of 260 times, their resolving power was limited to about 1.4 µm [3].

Although simple one-lens microscopes like Van Leeuwenhoek's were impressive, a Dutch inventor by the name of Cornelius Jacobszoon Drebbel brought a new device to London [4] even earlier (1619), a two-lens microscope that possessed higher magnification capacity than the Van Leeuwenhoek instrument since it was based on the principle that in a two-lens microscope the total magnification of the lenses was multiplicative [5]; although the resolution was limited by optical aberrations.

Using a microscope very much like Drebbel's, but with an improved source of illumination, the Englishman Robert Hooke was able to see details in pieces of plants, animals, and insects that had previously been unknown. For example, he observed that a piece of cork bark was composed of many small rectangular compartments. They reminded him of the small rooms that monks slept in. He called them cells, a name we still use today; had he called them chambers we might be studying chamber biology instead. He published these observations as well as the first recorded attempt to make measurements using a microscope in his 1665 book Micrographia[6]. These early studies of the invisible world of cells represent the birth of modern microscopy.

In the eighteenth and nineteenth centuries, microscopes became progressively more powerful, lens design was improved to remove aberrations, and innovations such as the use of polarized illumination were introduced. In the 1880s the German scientists Ernst Abbe and August Valentin Köhler working with Carl Zeiss brought together a sophisticated lens design [7] and improved illumination methods [8, 9] to create microscopes that could resolve subcellular structures. Abbe was the first to mathematically calculate the limits of microscope resolution using photons [10]. His calculations showed that the wavelength of visible light and the angle from which the diffracted light is collected defined the limits for microscopic resolution. Thus, the Abbe diffraction barrier of 188 nm was elucidated, and this would remain the limit of light microscopy until the advent of super-resolution techniques some 125 years later.

1.2 The Electron Takes Its Place

In the 1920s, while light microscopy still had to fully exploit its resolution possibilities, a young French physics student was pondering the theories of Einstein, in particular the nature of electrons, and wondering if they had a wavelength. His name was Louis de Broglie and the equation describing the wave nature of electrons was at the heart of his PhD thesis [11]. In a triumph of early career achievement his thesis secured him the 1929 Nobel prize in physics! Being a theoretician, he had no practical use for his work and went on to the next equation. Fortunately, there were more practically minded physicists who did see the use of the wave nature of electrons. Ladislaus Marton in Brussels, and Ernst Ruska, Max Knoll, and Ernst Brüche in Berlin developed simultaneous prototype transmission electron microscopes, which proved that not only did electrons have a wavelength but also that they could be focused by electromagnetic lenses and used in the same manner as light was used in optical microscopy [12]. Ruska theorized that under the right conditions these microscopes could achieve a resolution of 2Ä, which was proved correct almost 40 years later [13].

Biologists rejoiced at the news that smaller subcellular structures could finally be resolved; however, it came at a price. Specimens had to be imaged in high vacuum and radiation damage from the strong electron beam was intense. Despite that, Marton published the first biological electron micrograph of a sample of Drosera intermedia, sundew, in the journal Nature in 1934 [14]. While this was a breakthrough, the actual resolution of electron micrographs would be insufficient to produce useful scientific data for another 20 years. So until almost the 1960s, electron microscopes were like the optical microscopes of the seventeenth century, largely curiosities.

1.3 Putting It Together, 1960s to 1980s

Although both light and electron microscopy continued to improve, it wasn't until the 1960s that researchers tried to combine the two imaging techniques. When searching the early literature for correlative microscopy publications, it becomes obvious that the term as we now use it, to indicate light and electron microscopic studies on the same area of the same sample, has evolved over time. The earliest references are frequently studies of the same tissue or sample type but not necessarily on the same specimen; thus, they are more comparative than truly correlative. The earliest paper that we have found that imaged a sample in a light microscope with a similarly prepared sample in an electron microscope is from the pioneering work of Keith Porter, where chick embryonic fibroblasts were cultured on a formvar substrate, fixed and imaged (Figure 1.1) [15]. This was only done as a proof of principle for developing EM techniques, though, and no attempt was made to draw conclusions from any correlation. A correlative study from 1960 by Goodman and Morgan was performed on separate cell cultures and published as two papers, one for light [16] and one for transmission electron microscopy (TEM)[17].

Figure 1.1 The first micrograph to compare a sample imaged with a light microscope; 1) and an electron microscope; 2), was published by Keith Porter in 194515. While not truly correlative, e.g. of the same specimen, this did demonstrate that samples prepared with the same procedures could be imaged using multiple methods.

Reproduced with permission of ROCKEFELLER UNIVERSITY PRESS via Copyright Clearance Center ©1945.

Other correlative studies from 1969 [18] and 1970 [19, 20] used biopsy samples that had been divided and processed for either light or electron microscopy, and then extrapolated between the morphological findings in each. Additional studies of correlative microscopy went a step further and used the same sample but adjacent sections. In 1970, Watari and coworkers published a study of the islets of Langerhans using adjacent resin-embedded sections [21], and in 1979, Hyde et al. used the same block to first cut thick sections and inspect them by LM, then selected areas were cut out from these samples, and thin sectioned for TEM [22]. A very early attempt to combine immunohistochemistry with TEM was published in 1974 by Bordi and Bussolati [23].

In 1980, Gonda and Hsu combined LM, scanning electron microscopy (SEM) and TEM to study developing mouse blastocysts [24]. These early studies, although not meeting the criteria for correlative microscopy that we use today, were examples of researchers trying to use multiple microscopy methods to bridge the resolution gap between photons and electrons.

It was probably the 1967 article by McDonald, Pease, and Hayes [25] that examined sectioned rabbit tissues by LM and SEM, that marks the first use of correlative microscopy with the specific purpose of adding the extra resolution available in the EM to the LM data (Figure 1.2). A 1969 paper by McDonald and Hayes used fixed,...