1
The Past as Prologue

Flow cytometry, like most scientific developments, has roots firmly grounded in history. In particular, flow technology finds intellectual antecedents in microscopy, blood cell counting instruments, and the inkjet technology that was, in the 1960s, being developed for computer printers. It was the coming together of these three strands of endeavor that provided the basis for the development of the first flow cytometers. Because thorough accounts of the history of flow cytometry have been written elsewhere (and make a fascinating story for those interested in the history of science), we cover history here in just enough detail to give readers a perspective as to why current instruments have developed as they have.

Microscopes have, since the 17th century, been used to examine tissue sections and cells. Particularly since the end of the 19th century, stains have been developed that make various cellular constituents visible; in the 1940s and 1950s, fluorescence microscopy began to be used in conjunction with fluorescent stains for nucleic acids in order to detect malignant cells. With the advent of antibody technology and the work of Albert Coons in linking antibodies with fluorescent tags, the use of fluorescent stains gained wider and more specific applications. In particular, cell suspensions or tissue sections are now routinely stained with antibodies specific for antigenic markers of cell type or function. The antibodies are either directly or indirectly conjugated to fluorescent molecules (most commonly fluorescein). The cellular material can then be examined on a glass slide under a microscope, and the fluorescence of the cells can be excited and observed (Fig. 1.1). The fluorescence microscope allows us to see cells, identify them in terms of both their physical structure and their orientation within tissues, and then determine whether and in what pattern the cells fluoresce when stained with one or another of the specific stains available. In addition, a microscope can also be fitted with a camera or photodetector, which will then record the intensity of fluorescence arising from the field in view. The logical extension of this technique is image analysis cytometry, which digitizes the output to allow precise quantitation of fluorescence intensity patterns in detail (pixel by pixel) within that field of view.

Fig. 1.1. The optical path of a fluorescence microscope. In this example, the filters and mirrors are set for detection of fluorescein fluorescence.

Modified from Alberts et al. (1989).

The development of hybridoma cultures for the production of monoclonal antibodies (for which César Milstein and Georges J.F. Köhler were awarded the Nobel Prize in 1984) led to a vast increase in the number of cellular components that can be specifically stained and used to classify cells. Whereas monoclonal antibody techniques are not directly related to the development of flow technology, their invention was a serendipitous event that had a great impact on the potential utility of flow cytometric systems.

In 1934, Andrew Moldavan in Montreal took a first step from static microscopy toward a flowing system. He suggested the development of an apparatus to count red blood cells and neutral-red-stained yeast cells as they were forced through a capillary on a microscope stage. A photodetector attached to the microscope eyepiece would register each passing cell. Although it is unclear from Moldavan's paper whether he actually ever built this cytometer, the development of staining procedures over the next 30 years made it obvious that the technique he suggested could be useful not simply for counting the number of cells but also for comparing their characteristics.

In the mid-1960s, Louis Kamentsky took his background in optical character recognition and applied it to the problem of automated cervical cytology screening. He developed a microscope-based spectrophotometer (on the pattern of the one suggested by Moldavan) that measured and recorded ultraviolet absorption and the scatter of blue light ("as an alternative to mimicking the complex scanning methods of the human microscopist") from cells flowing "at rates exceeding 500 cells per second" past a microscope objective (Fig. 1.2). Then, in 1967, Kamentsky and Myron Melamed elaborated this design into a sorting instrument that provided for the electronic actuation of a syringe to pull cells with high absorption/scatter ratios out of the flow stream. These "suspicious" cells could then be subjected to detailed microscopic analysis. In 1969, Wolfgang Dittrich and Wolfgang Göhde in Münster, Germany, described a flow chamber for a microscope-based system with which fluorescence intensity histograms could be generated based on the ethidium bromide fluorescence from the DNA of alcohol-fixed cells.

During this period of advances in flow microscopy, so-called Coulter technology had been developed by Wallace Coulter for analysis of blood cells. In the 1950s, instruments were produced that counted cells as they flowed in a liquid stream; analysis was based on the amount by which the cells increased electrical resistance as they displaced an isotonic saline solution while flowing through an orifice. Cells were thereby classified more or less on the basis of their volume because larger cells have greater electrical resistance. These Coulter counters soon became essential equipment in hospital hematology laboratories, allowing the rapid and automated counting of white and red blood cells. They actually incorporated many of the features of analysis that we now think of as being typical of flow cytometry: the rapid flow of single cells in file through an orifice, the electronic detection of signals from those cells, and the automated analysis of those signals.

Fig. 1.2. A diagram of Kamentsky's original flow sorter.

From Kamentsky and Melamed (1967)/American Association for the Advancement of Science - AAAS.

At the same time as Kamentsky was working on cervical screening, Mack Fulwyler at the Los Alamos Laboratory in New Mexico was in need of a new project. He had been studying the contamination of food by radioactive fallout from nuclear bombs. This project had become less urgent because, in 1963, the United States, the Soviet Union, and Great Britain signed the Nuclear Test Ban Treaty; this Treaty limited the testing of nuclear weapons under water, in the atmosphere, and in outer space. For a new project, Fulwyler decided to investigate an issue well known to everyone looking at red blood cells in Coulter counters: Red cells show a bimodal distribution of their electrical resistance ("Coulter volume"). Anyone looking at erythrocytes under the microscope cannot help but be impressed by the remarkable structural uniformity of these cells. Some biologists had hypothesized that the bimodal distribution of the electrical resistance of erythrocytes resulted from the existence of two classes of these cells. Fulwyler doubted that the bimodal Coulter volume distribution represented two classes of these microscopically very uniform cells. He felt that the bimodal profile might simply be an artifact based on some quirky aspect of the electronic resistance measurements. The most direct way of testing these two alternatives was to separate red blood cells according to their electronic resistance signals and then to determine whether the separated classes remained distinct (i.e., unimodal) when they were reanalyzed.

The technique that Fulwyler developed for sorting the erythrocytes combined Coulter methodology with the inkjet technology being developed at Stanford University by Richard G. Sweet for running computer printers. Inkjet technology involves the vibration of a nozzle so as to generate a stream that breaks up into discrete drops-followed by the charging and grounding of that stream at appropriate times so as to leave indicated drops, as they break off, carrying an electrical charge. For purposes of printing, those charged drops of ink can then be deflected to positions on the paper as required by computer print messages. Fulwyler took the intellectual leap of combining this methodology with Coulter flow technology; he developed an instrument that would charge drops containing suspended cells, thereby allowing deflection of the cells (within the drops) as dictated by signals based on the cell's measured Coulter volume.

The data from this limited but pioneering experiment led to a conclusion that with hindsight seems obvious: Erythrocytes are indeed uniform. When they are sorted according to their electrical resistance, the resulting cells from one class or the other still show a bimodal distribution when reanalyzed for their electrical resistance profile. The bimodal "volume" signal from erythrocytes was therefore artifactual-resulting in part from the discoid (nonspherical) shape of the cells. The technology developed for this landmark experiment is the essence of all the technology required for flow sorting as we now know it. That experiment also, unwittingly, emphasized an aspect of flow cytometry that has remained with us to this day: Flow cytometrists still need to be continually vigilant (and know how to use a microscope) because signals from cells (particularly signals that are assumed to be related to cell volume) are subject to artifactual influences and may not be what they seem. (Fulwyler's 1965 paper actually describes the separation of mouse from human erythrocytes and the separation of a...