Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

© 2004 Wiley-Liss, Inc. Cytometry Part A 58A:13–20 (2004)

The Evolution of Cytometers

Howard M. Shapiro*
The Center for Microbial Cytometry, West Newton, Massachusetts

Microscopy is difficult when cells are on the fly;

It’s lucky the cytometer’s now quicker than the eye.
This hasn’t been the case throughout the gadget’s evolution,
But lasers and computers have provided a solution
To problems of illuminating cells, at least a myriad
Per second, and collecting ample data, in that period,
To pick out some to keep, and destine others for rejection.
For modern labs, the instrument’s a natural selection!

The first microscopes were as likely to be gentlemen’s son (3), working at the Karolinska Institute (Stockholm,
toys as scientists’ tools. By the beginning of the 20th Sweden), began to study cellular nucleic acids and their
century, the professionals had largely taken over from the relation to cell growth and function. He developed a series
amateurs, and it had become possible to draw at least of progressively more sophisticated microspectrophotom-
some quantitative conclusions from observations made eters, which made fairly precise measurements of nucleic
using microscopes. Hemocytometers allowed an observer acid and protein content based on the intrinsic ultraviolet
to derive a reasonably accurate count of the number of (UV) absorption of these substances near 260 and 280 nm.
cells or other particles in a unit volume of specimen, Caspersson’s early apparatus now seems hopelessly
although precision of counts was limited by both counting primitive. Cadmium spark sources were used for ultravi-
statistics and dilution errors. Using eyepiece reticles, olet illumination; photocurrent measurements were done
grids, etc., one could also measure the size of microscopic with string electrometers, unless the signal was strong
objects, at least in two dimensions. It was not, however, enough to permit use of a vacuum-tube amplifier. How-
until some time later that the tools developed by chemists ever, even this primitive apparatus got results, and at-
and physicists for spectroscopy and photometry or radi- tracted the attention of other researchers; many of the
ometry were adapted for use with the microscope, thus advances in analytical cytology from the 1940s on were
producing the first true optical cytometers. made by people who had made the pilgrimage to Stock-
holm. Ornstein (2) documents the influence of Caspers-
THE DAWN OF CYTOMETRY son’s work in establishing the role of DNA as the genetic
I have written at length about the development of material.
cytometers on numerous occasions, most recently in the It was during the 1950s that analytical cytology ac-
4th edition of Practical Flow Cytometry (1); in this briefer quired its name, coined by Francis O. Schmitt of the
exposition, influenced by the title suggested by the Editor Massachusetts Institute of Technology (MIT; Cambridge,
of this journal, I will emphasize the lineage of the instru- MA); the first and second editions of a book entitled
ments. I will apologize in advance to the many colleagues Analytical Cytology, edited by Robert Mellors, of the
whose contributions I cannot discuss in the allotted space, Memorial Sloan-Kettering Cancer Institute (New York,
and thus hope to avoid possible derogatory comments NY), appeared in 1955 and 1959 (4). The book included
about my own lineage. I am pleased that Leonard Orn- chapters on the fluorescent antibody method, on histo-
stein, a pioneer in the development of both static and flow chemistry, and on phase, interference, and polarizing mi-
cytometric apparatus and techniques, appears to share my croscopy; Arthur Pollister, Ornstein’s mentor at Columbia
perspective on the early history of cytometry, as evi- University (New York, NY), and Ornstein contributed
denced by an informative and entertaining reminiscence several chapters, including one on the theory and practice
he published in 1987 (2).
It is probably fair to say that the evolution of cytometers
from microscopes began in the 1930s in Stockholm. By *Correspondence to: H.M. Shapiro, The Center for Microbial Cytom-
etry, 283 Highland Avenue, West Newton, MA 02465-2513.
this time, conventional histologic staining techniques of E-mail: [email protected]
light microscopy had suggested that tumors might have Published online in Wiley InterScience (www.interscience.wiley.com).
abnormalities in DNA and RNA content. Torbjörn Caspers- DOI: 10.1002/cyto.a.10111
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

14 SHAPIRO

of absorption measurements that is well worth reading middle of the decade, there were at least a few groups of
even today. analytical cytologists ready to use them. The TICAS sys-
tem, assembled in the late 1960s by George Wied of the
Early Microspectrophotometry University of Chicago, Gunter Bahr of the Armed Forces
and Image Cytometry Institute of Pathology, and Peter Bartels of the University
Microspectrophotometers were first made by putting a of Arizona (Tucson, AZ), interfaced Zeiss’s commercial
small “pinhole” aperture, technically known as a field version of the Caspersson microspectrophotometer to a
stop, in the image plane of a microscope, restricting the minicomputer, with the aim of automating interpretation
field of view to the area of a single cell, and placing a of Pap smears (6).
photodetector behind the field stop. If a 40⫻ objective The use of stage motion for scanning made operation
lens is used, measuring the transmission through, or the extremely slow; it could take many minutes to produce a
absorption of, a cell 10 ␮m in diameter requires a 400 ␮m high-resolution scanned image of a single cell, and there
diameter field stop; with a smaller field stop, it becomes were no computers available to capture the data. Some-
possible to measure the transmission through a corre- what higher speed could be achieved by using Nipkow
spondingly smaller area of the specimen. A 40 ␮m field discs or galvanometer-driven moving mirrors for image
stop allows measurement of a 1 ␮m diameter area of the scanning, and limiting the tasks of the motorized stage to
specimen, and. By moving the specimen in precise incre- bringing a new field of the specimen into view and into
mental steps in the x and y directions (i.e., in the plane of focus; this required some electronic storage capability,
the slide) in a raster pattern, and recording the informa- and made measurements susceptible to errors due to un-
tion, it becomes possible to measure the integrated ab- even illumination across the field, although this could be
sorption of a cell, and/or to make an image of the cell with compensated for. My colleagues and I at National Insti-
each pixel corresponding in intensity to the transmission tutes of Health (NIH) built “Spectre II” (7), which incor-
or absorption value. This was the first, and, until the porated a galvanometer mirror scanning system (8) devel-
1950s, the only approach to scanning cytometry. By the oped by Kendall Preston, an Airborne Instruments
1960s, Zeiss (Oberkochen, Germany) had commercialized alumnus then working at Perkin-Elmer (Norwalk, CT), and
a current version of Caspersson’s apparatus, and others a Digital Equipment Corporation (Maynard, MA) LINC-8
had begun to build high-resolution scanning microscopes computer. While this system had sufficient computer
incorporating a variety of technologies. power to capture high-resolution cell images (0.2 ␮m
pixels), data were recorded on 9-track tape and trans-
Motivation for Progress: Cancer Cytology ported (by “sneakernet”) to a mainframe elsewhere on the
and Hematology Automation NIH campus for analysis (9).
By the mid-1950s, it had become apparent that malig- Since the late 1940s and early 1950s had already given
nant cells were likely to contain more nucleic acid than us Howdy Doody, Milton Berle, and the Ricardos, it might
normal cells, and Mellors proposed construction of an be expected that, somewhere around that time, someone
automatic scanning instrument for screening cervical cy- would have tried to automate the process of looking down
tology (Papanicolaou or “Pap”) smears. Tolles et al. (5), at the microscope and counting cells using video technol-
Airborne Instruments Laboratory (Mineola, NY), de- ogy. Most of the imaging cytometers developed then were
scribed the “Cytoanalyzer” built for this purpose. A “Nip- not based on video cameras, for a number of reasons, not
kow disc” containing a series of apertures rotated in the the least of which was the variable light sensitivity of
image plane of a microscope, producing a raster scan of a different regions of a camera tube, which made quantita-
specimen with approximately 5 ␮m resolution. A hard- tive measurements difficult. However, it was recognized
wired analyzer extracted nuclear size and density informa- that the raster scan mechanism of a cathode ray tube
tion; cells were then classified as normal or malignant could be used on the illumination side of an image analysis
using these parameters. The Cytoanalyzer was, to make a system, with the “flying spot” illuminating only a small
long story short, right more of the time than it was wrong, segment of the specimen plane at any given time (10). The
but its false-positive and false-negative rates were too high CYDAC system, a flying spot scanner built at Airborne
for it to be suitable for clinical use. The results were Instruments Laboratory, was used in studies of the auto-
encouraging enough for the American Cancer Society and mation of differential leukocyte counting (11) and chro-
the National Cancer Institute to continue funding research mosome analysis (12) by Mortimer Mendelsohn (later the
on cytology automation. first president of the Society for Analytical Cytology),
Recording and storing cell images was a nontrivial task Brian Mayall (later the founding editor of this journal),
in the 1960s, when mainframe computers occupied entire Judith Prewitt, and their colleagues, then at the University
rooms, required kilowatts of power and heavy-duty air of Pennsylvania.
conditioning, and cost millions of dollars, which bought a
processor with 160 kb of memory and a 6.4 ␮s instruction FLOW CYTOMETRY AND SORTING:
cycle (equivalent clock speed 160 kHz). Input and output WHY AND HOW
and data storage typically used tape drives; large random Somewhat simpler tasks of cell or particle identification,
access storage media had not yet arrived on the scene. characterization, and counting than those involved in Pap
However, when minicomputers became available in the smear analysis and differential white cell counting had
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

CYTOMETER EVOLUTION 15
attracted the attention of other groups of researchers at device, which would have represented a new level of
least since the 1930s. During World War II, the U.S. Army sophistication in flow cytometry, was ever actually built.
became interested in developing devices for rapid detec- Around the same time, the promising results obtained
tion bacterial biowarfare agents in aerosols (a continuing with the Cytoanalyzer in attempts to automate reading of
preoccupation); this would require processing a relatively Pap smears (5) encouraged executives at the International
large volume of sample in substantially less time than Business Machines Corporation (IBM) (Armonk, NY) to
would have been possible using even a low-resolution look into producing an improved instrument. Assuming
scanning system. The apparatus built in the Chemistry this would be some kind of image analyzer, IBM gave
Department of Northwestern University (Evanston, IL) by technical responsibility for the program to Louis Ka-
Gucker et al. (13) in support of this project achieved the mentsky, who had developed a successful optical charac-
necessary rapid specimen transport by injecting the air ter reader. He did some calculations of what would be
stream containing the sample into the center of a larger required in the way of light sources, scanning rates, and
sheath stream of flowing air that passed through the focal computer storage and processing speeds to solve the
point of a dark-field microscope. Particles passing through problem using image analysis, and concluded that a dif-
the system scattered light into a collection lens, eventually ferent approach would be required.
producing electrical signals from a photodetector. The Having learned from pathologists in New York that cell
instrument could detect objects on the order of 0.5 ␮m in size and nucleic acid content could provide a good indi-
diameter, and is generally recognized as having been the cator of whether cervical cells were normal or abnormal,
first flow cytometer used for observation of biological Kamentsky traveled to Caspersson’s laboratory in Stock-
cells. holm and learned microspectrophotometry. He then built
By the late 1940s and early 1950s, the same principles, a microscope-based flow cytometer that used a transmis-
including the use of sheath flow, were applied to the sion measurement at visible wavelengths to estimate cell
detection and counting of red blood cells in saline solu- size and a 260 nm UV absorption measurement to estimate
tions (14), providing effective automation for a diagnostic nucleic acid content (19,20). Subsequent versions of this
test notorious for its imprecision when performed by a instrument, which incorporated a dedicated computer
human observer using a hemocytometer and a micro- system, could measure as many as four cellular parameters
scope. Neither the bacterial counter nor the early red cell (21). A brief trial on cervical cytology specimens indicated
counters had any significant capacity either for discrimi- the system had some ability to discriminate normal from
nating different types of cells or for making quantitative abnormal cells (22); it could also produce distinguishable
measurements. Both types of instrument were measuring signals from different types of cells in blood samples
what we would now recognize as side scatter signals; stained with a combination of acidic and basic dyes, sug-
although larger particles, in general, produced larger sig- gesting that flow cytometry might be usable for differen-
nals than smaller ones, correlations between particle sizes tial leukocyte counting.
and signal amplitudes were not particularly strong. The first commercial flow cytometric differential
An alternative flow-based method for cell counting was counter, introduced in the early 1970s, was Technicon’s
developed in the 1950s by Wallace Coulter (15). Recog- (Tarrytown, NY) Hemalog D (23,24); Ornstein was a
nizing that cells, which are surrounded by a lipid mem- prime mover in its development, having interacted with
brane, are relatively poor conductors of electricity as com- Kamentsky’s group along the way (2). The Hemalog D
pared to saline, he devised an apparatus in which cells used light scattering and absorption measurements made
passed one by one through a small (⬍100 ␮m) orifice at different wavelengths in three different flow cytometers
between two chambers filled with saline. When a cell to classify leukocytes. Chromogenic enzyme substrates
passed through, the electrical impedance of the orifice were used to identify neutrophils and eosinophils by the
increased in proportion to the volume of the cell, produc- presence of moderate to high and very high concentra-
ing a voltage pulse. The Coulter counter was widely tions of peroxidase, while another channel identified
adopted in clinical laboratories for blood cell counting; it monocytes by their esterase content. Basophil identifica-
was soon established that it could provide more accurate tion was based on detection of glycosaminoglycans in
measurements of cell size than had previously been avail- basophil granules using Alcian blue. A single tungsten-
able (16,17). halogen lamp served as light source for all three flow
In the early 1960s, investigators working with Leitz systems.
(Wetzlar, Germany) (18) conceived a hematology counter Although the Hemalog D employed cytochemical stain-
that added a fluorescence measurement to the light scat- ing procedures that were well regarded by hematologists
tering measurement used in red cell counting. If a fluores- for such purposes as determination of lineage of leukemic
cent dye such as acridine orange were added to the blood cells, the apparatus, which worked pretty well, was ini-
sample, white cells would be stained much more brightly tially regarded with a great deal of suspicion, at least in
than red cells; the white cell count could then be derived part due to the novelty of flow cytometry. The developers
from the fluorescence signal, and the red cell count from and manufacturers of image analyzing differential
the scatter signal. It was also noted that acridine orange counters, which certainly didn’t perform much better
fluorescence could be used to discriminate mononuclear than did the Hemalog D, did what they could to keep
cells from granulocytes. However, it is not clear that the potential users suspicious of flow cytometry for as long as
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

16 SHAPIRO

possible; the technology would eventually be legitimized incorporated the orthogonal “body plan” now standard in
by its dramatic impact on immunology, which was facili- laser-source instruments, with the optical axes of illumi-
tated by the introduction of cell sorting and immunofluo- nation and light collection at right angles to each other
rescence measurements. and to the direction of sample flow. Kamentsky, who had
Although impedance (Coulter) counters and optical left IBM to found Bio/Physics Systems (Mahopac, NY),
flow cytometers could analyze hundreds of cells per sec- produced the Cytofluorograf, an orthogonal geometry flu-
ond, providing a high enough data acquisition rate to be orescence flow cytometer that was the first commercial
useful for clinical use, microscope-based static cytometers product to incorporate an argon ion laser; Göhde’s Partec
offered a significant advantage. A system with computer- (Münster, Germany) Impulscytophotometer (ICP) instru-
controlled stage motion could be programmed to reposi- ment, built around a fluorescence microscope with arc
tion a cell on a slide within the field of view of the lamp illumination, was distributed commercially by
objective (7), allowing the cell to be identified or other- Phywe (Göttingen, Germany).
wise characterized by visual observation; it was, initially, Leonard Herzenberg and his colleagues (34), at Stanford
not possible to extract cells with known measured char- University, realizing that fluorescence flow cytometry and
acteristics from a flow cytometer. Until this could be subsequent cell sorting could provide a useful and novel
done, it would be difficult to verify any cell classification method for purifying living cells for further study, devel-
arrived at using a flow cytometer, especially where the oped a series of instruments after exposure to a Ka-
diagnosis of cervical cancer or leukemia might be in- mentsky prototype lent to them by IBM (35). Although
volved. their original apparatus (36), with arc lamp illumination,
This problem was solved in the mid-1960s, when both was not sufficiently sensitive to permit them to achieve
Mack Fulwyler (25), working at the Los Alamos National their objective of sorting cells from the immune system,
Laboratory (Los Alamos, NM), and Kamentsky, at IBM based on the presence and intensity of staining by fluo-
(26), demonstrated cell sorters built as adjuncts to their rescently labeled antibodies, the second version (37),
flow cytometers. Kamentsky’s system used a syringe which used a water-cooled argon laser, was more than
pump to extract selected cells from its relatively slow- adequate. This was commercialized as the Fluorescence-
flowing sample stream. Fulwyler’s was based on ink jet Activated Cell Sorter (FACS) in 1974 by a group at Becton-
printer technology then recently developed by Richard Dickinson (B-D, now BD Biosciences, San Jose, CA), led by
Sweet (27) at Stanford University (Stanford, CA); following Bernard Shoor.
passage through the cytometer’s measurement system Coulter Electronics (now Beckman Coulter, Fullerton,
(originally a Coulter orifice), the saline sample stream was CA), which by 1970 had become a very large and success-
broken into droplets, and those droplets that contained ful manufacturer of laboratory hematology counters, pur-
cells with selected measurement values were electrically sued the development of fluorescence flow cytometers
charged at the droplet break-off point. The selected through a subsidiary, Particle Technology, under Mack
charged droplets were then deflected into a collection Fulwyler’s direction in Los Alamos. The Two Parameter
vessel by an electric field; uncharged droplets went, as it Sorter (TPS-1), Coulter’s first product in this area, reached
were, down the drain. the market in 1975. It used an air-cooled 35 mW argon ion
In the early 1970s, the group at Los Alamos led the way laser source and could measure forward scatter and fluo-
in implementation of practical multiparameter flow cy- rescence.
tometers; their larger instruments, with droplet sorting Multiple wavelength fluorescence excitation was intro-
capability, combined two-color fluorescence measure- duced to flow cytometry in apparatus built at Block Engi-
ments with measurements of Coulter volume and (thanks neering (Cambridge, MA) during an abortive attempt to
to the contributions of Paul Mullaney, Gary Salzman, and develop a hematology instrument. The first instrument
others) light scattering at several angles (28 –31). The (38) derived five illuminating beams from a single arc
cytometers were interfaced to Digital Equipment Corpo- lamp; the second (39) used three laser beams; both could
ration minicomputers. Several instruments made at Los analyze over 30,000 cells per second and, using hardwired
Alamos were delivered to the NIH; other institutions cop- preprocessors and integral minicomputers, identify cells
ied most or all of the Los Alamos design in their own comprising less than 1/100,000 of the total sample. The
laboratory-built apparatus. laser source system incorporated forward and side scatter
measurements, which permitted lymphocyte gating (40),
FLUORESCENCE AND FLOW: MADE influenced by work done at Los Alamos (31). Block also
FOR EACH OTHER built a slow flow system intended for detection of hepa-
Fluorescence measurement was introduced to flow cy- titis B virus and antigen in serum; it could discriminate
tometry in the late 1960s as a means of improving both scatter signals from large viruses (41) and could theoreti-
quantitative and qualitative analyses. By that time, Van cally detect a few dozen fluorescein molecules above
Dilla et al. (32) at Los Alamos and Dittrich and Göhde (33) background. The Block cytometers were never sold com-
in Germany had built fluorescence flow cytometers to mercially, but influenced the optical, electronic, and sys-
measure cellular DNA content, facilitating analysis of ab- tems design of later instruments.
normalities in tumor cells and of cell cycle kinetics in both By the time the Society for Analytical Cytology (now
neoplastic and normal cells. The Los Alamos instrument ISAC) came into being in 1978, B-D, Coulter, and Ortho (a
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

CYTOMETER EVOLUTION 17
division of Johnson & Johnson (Raritan, NJ) that bought specimen is/are illuminated at any given time. Although a
Bio/Physics Systems) were producing flow cytometers field stop can restrict the area being measured in a mi-
that could measure small- (forward scatter) and large- (side crospectrofluorometer or scanning fluorescence micro-
scatter) angle light scattering and fluorescence in at least scope, the image resolution is adversely affected by the
two wavelength regions, analyzing several thousand cells collection of fluorescence emission from planes above
per second, and with droplet deflection cell sorting capa- and below the focal plane. Various means of eliminating
bility. Ortho was also distributing the ICP, which, by this light are employed in different types of confocal
virtue of its optical design, could make higher precision microscopes, with the end result that it has become pos-
measurements of DNA content than could laser-based sible to create high-resolution images of very thin sections
flow cytometers. DNA content analysis was receiving con- of a specimen, and to reconstruct three-dimensional detail
siderable attention as a means of characterizing the aggres- from a succession of such images. Although confocal
siveness of breast cancer and other malignancies, and, at microscopes are far outnumbered by flow cytometers at
least in part due to the results of a Herzenberg sabbatical present, they are becoming essential components of the
in Cesar Milstein’s lab at the University of Cambridge cytometric armamentarium.
(UK), monoclonal antibodies had begun to emerge as
practical reagents for dissecting the stages of development FLOW CYTOMETRY IN AND OUT
of cells of the blood and immune system. Loken, Parks, OF THE MAINSTREAM
and Herzenberg had successfully performed a two-color From the early 1970s on, commercial production of
immunofluorescence experiment, introducing fluores- instruments allowed researchers who had not developed
cence compensation in the process (42), although it was and built their own apparatus to pursue applications of
clear that a great deal needed to be done in the area of fluorescence flow cytometry and sorting. Advances in the
fluorescent label development to realize the potential of technology itself continued to occur primarily in a rela-
monoclonal antibodies. tively small community of academic, government, and
industrial labs. What got done in any one lab was deter-
STILL IN THE PICTURE: STATIC CYTOMETERS mined by the biological problems and/or clinical applica-
The late 1960s and early 1970s saw the development of tions under investigation, and also by the migration of
a number of image analyzing automated differential leu- instruments and/or investigators from one place to an-
kocyte counters. Corning (Corning, NY) produced the other. This process has gotten some attention from real
LARC, based on work by Bacus (43); Geometric Data historians of science, resulting in publications by Peter
(Wayne, PA) introduced the Hematrak (44), and Coulter Keating and Alberto Cambrosio (52) and in a video history
offered a system based on research by Young (45). Other by Ramunas Kondratas, which was funded and otherwise
manufacturers subsequently entered the market. The in- heavily influenced by B-D; a summary is available online
struments all incorporated bright field microscopes and from the Smithsonian Institution Archives (www.si.edu/
examined cells stained with Wright’s, Giemsa’s, or similar archives/ihd/videocatalog/9554.htm).
stains in smears on slides. Bacus later founded a company At Stanford University, the emphasis remained on sort-
celled Cell Analysis Systems, (Lombard, IL) which sold ing on the basis of immunofluorescence signals with the
bright field image analyzing systems that were applied to aim of isolating morphologically indistinguishable viable
DNA content measurement in tumor cells, using Feulgen lymphocytes with differences in functional characteris-
staining, and detection of hormone receptors, using (im- tics. Placing the observation point in a jet in air, rather
muno)cytochemistry. than in a flow chamber, shortened the distance between
Image cytometers incorporating both bright field and this point and the droplet break-off point, making faster
fluorescence microscopes existed in research laborato- sorting possible. One notable descendant of the Stanford
ries, but were much slower, usually somewhat less precise instrument was the computer-controlled, multiparameter
and much less sensitive (not easily adapted for immuno- cytometer/sorter built by the Jovins at the Max Planck
fluorescence analysis), and almost always even less user- Institute for Biophysical Chemistry in Göttingen, Germany
friendly than the early flow cytometers. In general, these (53). This apparatus could operate at short ultraviolet
instruments couldn’t sort, although Meridian Instruments’ wavelengths; it was used to measure such parameters as
(Okemos, MI) ACAS system, which incorporated both intrinsic protein fluorescence, membrane fluidity (using
lamp illumination and laser illumination for fluorescence fluorescence polarization), and receptor proximity (using
measurements, could destroy unwanted cells with high- energy transfer), and to establish the utility of Hoechst
intensity laser light (46). This technique of photodamage (Frankfurt-am-Main, Germany) 33342 as a vital DNA stain
cell selection, or “cell zapping,” was also envisioned as an and thioflavin T as an RNA stain.
adjunct to flow cytometry (47– 49). The multiparameter flow cytometers developed at Los
However, by far the most exciting development in Alamos were copied by investigators at other institutions,
static cytometry in the 1980s was the adaptation of the e.g., the Salk Institute (La Jolla, CA), Colorado State Uni-
technique of confocal microscopy to fluorescence imag- versity (Ft. Collins, CO), the University of California at Los
ing. Historical details can be found in books by Inoué and Angeles (Los Angeles, CA), and the University of Houston
Spring (50) and Pawley (51). In a conventional fluores- (Houston, TX), where an instrument was applied to mul-
cence microscope, a fairly large area (and volume) of the tiparameter flow cytometric analysis of bacteria (54). Los
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

18 SHAPIRO

Alamos also provided the inoculum for the subsequent an ergonomically designed, computer-controlled “knob-
growth of another major center for flow cytometer devel- less” instrument that included sorting capability. By the
opment, that at Lawrence Livermore National Laboratory mid-1980s, with the emergence of AIDS, increasing de-
(Livermore, CA), where high-speed flow sorting was per- mand for clinical instruments, B-D had brought out the
fected as a means for separating human chromosomes FACScan, a three-color bench-top analyzer using a rectan-
(55,56). The MoFlo high-speed sorter developed by Ger gular cuvette with a gel-coupled lens for highly efficient
van den Engh and others at Livermore was subsequently light collection, which could make more sensitive immu-
refined by Cytomation (now DakoCytomation, Ft. Collins, nofluorescence measurements using a 15 mW air-cooled
CO), and has been produced commercially by them since argon laser source than were possible using ten times
1994. more laser power in stream-in-air sorters. The FACScan
The digital pulse processing recently introduced into was followed by the FACSort, which included a relatively
commercial instruments from Luminex Corporation (Aus- slow fluidic sorter; both were succeeded by the FACSCali-
tin, TX) and BD Biosciences mirrors work done beginning bur, which offered both a fluidic sorting option and a
in the 1970s by Leon Wheeless et al. (57) at the University fourth fluorescence channel with excitation from a red
of Rochester (Rochester, NY) on slit-scanning flow sys- (635– 640 nm) diode laser.
tems, leading to the development of progressively more From the mid-1990s on, there has been a proliferation
elaborate apparatus for processing pulse waveforms and of diode and solid-state lasers, and these small, energy-
for imaging cells in flow, intended for use in cancer efficient, and (usually) relatively inexpensive sources have
cytology. Digital pulse processing was used in bench-top increasingly been incorporated into flow cytometers. The
instruments produced in the early 1990s by RATCOM use of violet (395– 415 nm) diode lasers in cytometry was
(Miami, FL) (58), but, because analog-to-digital converters first described at the 2000 ISAC meeting (63); by the 2002
(ADCs) with sufficient speed and resolution were not meeting, at least three manufacturers had incorporated
available, neither these systems nor those built at Roches- such sources into instruments. Frequency-doubled diode-
ter could achieve a large enough dynamic range to elimi- pumped YAG lasers, emitting green light at 532 nm, have
nate the need for the logarithmic amplifiers that became also come into use, as have doubled semiconductor lasers
commonplace in flow cytometers during the 1980s. emitting at 488 – 492 nm, now available as replacements
An arc source instrument first described by Lindmo and for both air- and water-cooled argon lasers. Both diode and
Steen in 1979 (59,60) observed cells in sheath flow after a solid-state lasers are now available as ultraviolet (350 –375
jet in air intersected the flat surface of a cover slip, making nm) sources.
multiangle scatter and fluorescence measurements with Although the instruments Kamentsky built at IBM were
sufficient sensitivity to characterize bacteria (61). An early computer controlled, computers were expensive options
commercial version of this apparatus was produced by for most flow cytometers until the early 1980s, by which
Leitz; a later version was made by Skatron (Lier, Norway), time microprocessor-based systems could do the work of
and an even later one, formerly available from Bio-Rad an earlier generation of minicomputers. The FACScan was
(Hercules, CA) as the Bryte HS™, is now being produced equipped with a desktop computer system built by
by Apogee (Northwood, Middlesex) in the U.K. Hewlett-Packard (Palo Alto, CA); within a few years, per-
sonal computer systems were integrated into almost all
Flow Cytometer Domestication: From commercial flow cytometers, with B-D switching to Apple
Behemoths to Bench-tops (Cupertino, CA) Macintosh systems for its product line
Relatively small arc source systems such as the Im- and Coulter and Cytomation favoring IBM PC-compatible
pulscytophotometer and the Bryte HS managed to achieve products and Microsoft (Redmond, WA) software, initially
reasonably high measurement precision and sensitivity, running under DOS and later under Windows. An increas-
due at least in part to their use of high–numerical aperture ing amount of the internal electronics of flow cytometers
(NA) microscope optics, which provided more efficient has become computer-based, with the latest systems in-
light collection than was available in laser source flow corporating special-purpose large-scale integrated circuits,
cytometers. In the mid and late 1970s, Kamentsky’s Bio/ microprocessors, microcontrollers, and digital signal pro-
Physics Systems and its successor, Ortho Diagnostics Sys- cessing chips.
tems (Westwood, MA), introduced flow cytometers and The development of digital audio, telephony, and video
sorters in which measurements were made in flat-sided has resulted in large increases in the performance, and
quartz flow cuvettes, and in which “high-dry” microscope decreases in the price, of ADCs, which are critical ele-
objectives were used to increase light collection. This ments in data acquisition systems for any type of instru-
made it possible to use air-cooled rather than water-cooled mentation, flow cytometers included. The ADCs originally
lasers for immunofluorescence measurements, decreasing used with flow cytometers had only 8- or 10-bit resolution,
the size, cost, and power consumption of instruments and making it necessary to use logarithmic amplifiers to pro-
making them easier to introduce into an emerging clinical cess signals with a large dynamic range. This necessitated
market. Ortho’s Spectrum III, a highly automated system, the use of hardware for fluorescence compensation.
without sorting, was aimed at clinical users, as were B-D’s While this approach is feasible when three or four colors
FACS analyzer, a small but sensitive analytical apparatus are measured, it is essentially impossible to implement for
employing an arc lamp source, and Coulter’s EPICS C (62), modern multibeam instruments in which measurements
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

CYTOMETER EVOLUTION 19
of 12 or more colors may be made. The alternative is (Cambridge, MA) (70 –72), to simple fluorescence micro-
software compensation (63), which is best applied to scope cytometers using inexpensive charge-coupled de-
linear data digitized to a resolution of at least 16 bits. In vice (CCD) detectors and even less expensive LED light
the early 1990s, Auer et al. (64) implemented software sources.
compensation in the Beckman Coulter EPICS XL analyzer, The Society for Analytical Cytology was founded in
which captures 20-bit linear data, eliminating the need for 1978, and became the International Society for Analytical
logarithmic amplifiers. Other manufacturers, e.g., BD Bio- Cytology in 1991. In the interval, AIDS provided flow
sciences, DakoCytomation, and Partec, have developed cytometry with what the computer people call a “killer
their own approaches to high-resolution digital data anal- application,” in both figurative and literal senses. As ISAC
ysis, and we can expect that, as has been the case for passes its 25th Anniversary, it is gratifying to see at least
audio and video, digital techniques in flow cytometry will some of its members working toward simple, inexpensive
drive their predecessors to extinction. cytometric apparatus to aid in the fight against AIDS and
other health problems in both developed and developing
FUTURE DIRECTIONS IN CYTOMETRY countries. May we thus provide “a light unto the nations,”
Speaking of extinction, we may see renewed interest in even if it is only at the milliwatt level.
this measurement parameter in the near future. Extinction
measurements, which were favored for cell sizing in older
Ortho flow cytometers, require low-noise light sources; LITERATURE CITED
air-cooled argon and helium-neon lasers are too noisy, but 1. Shapiro HM. Practical flow cytometry. 4th edition. Hoboken, NJ:
Wiley-Liss; 2003. 681 p.
diode and solid-state lasers (and light emitting diodes 2. Ornstein L. Tenuous but contingent connections. Electrophoresis
[LEDs]) can readily meet the requirements. 1987;8:3–13.
With the introduction of the FACSAria sorter in late 3. Caspersson TO. Cell growth and cell function. New York: Norton;
1950, 185 pp.
2002, BD Biosciences successfully hybridized the behe- 4. Mellors RC, editor. Analytical cytology. 2nd edition. New York:
moth high-speed cell sorter and the bench-top analyzer; McGraw-Hill; 1959, 534 pp.
other manufacturers will almost certainly be demonstrat- 5. Tolles WE. The cytoanalyzer: an example of physics in medical
research. Trans N Y Acad Sci 1955;17:250 –256.
ing similar progeny at the 2004 ISAC meeting. Most hy- 6. Wied GL, Bahr GF, editors. Automated cell identification and cell
brids are sterile, but this should be an advantage as far as sorting. New York: Academic Press; 1970, 403 pp.
7. Stein PG, Lipkin LE, Shapiro HM. Spectre II: general-purpose micro-
cell sorters are concerned. scope input for a computer. Science 1969;166:328 –333.
We can expect flow cytometers to continue to decrease 8. Ingram M, Preston K, Jr. Automatic analysis of blood cells. Sci Am
in size and energy consumption, and hope that they will 1970;223:72–78.
9. Shapiro HM, Bryan SD, Lipkin LE, Stein PG, Lemkin PF. Computer-
eventually decrease in cost. It is widely known that flow aided microspectrophotometry of biological specimens. Exp Cell Res
cytometry makes it possible to analyze and sort over 1971;67:81– 85.
100,000 cells per second, to identify rare cells that repre- 10. Young JZ. A flying-spot microscope. Nature 1951;167:231.
11. Prewitt JMS, Mendelsohn ML. The analysis of cell images. Ann NY
sent only one of every 10 million cells in mixed popula- Acad Sci 1966;128:1035–1053.
tions, to simultaneously measure light scattering at two or 12. Mendelsohn ML, editor. Automation of cytogenetics. Asilomar Work-
shop, November 30 –December 2, 1975. CONF-751158. Livermore,
three angles and fluorescence in 12 or more spectral CA: Lawrence Livermore Laboratory; 1976, 187 p.
regions, to measure fluorescence with a precision better 13. Gucker FT, Jr, O’Konski CT, Pickard HB, Pitts JN, Jr. A photoelec-
than one percent, and to detect and quantify a few hun- tronic counter for colloidal particles. J Am Chem Soc 1947;69:2422–
2431.
dred molecules of fluorescent antibody bound to a cell 14. Crosland-Taylor PJ. A device for counting small particles suspended in
surface. It is less widely appreciated that it may be difficult fluid through a tube. Nature 1953;171:37–38.
or impossible to accomplish two or three of these amazing 15. Coulter WH. High speed automatic blood cell counter and cell size
analyzer. Proc Natl Electronics Conf 1956;12:1034 –1042.
feats at once, but it is anticipated that newer systems will 16. Brecher G, Schneiderman M, Williams GZ. Evaluation of electronic
be taught new tricks. red blood cell counter. Am J Clin Pathol 1956;26:1439 –1449.
Slow flow systems, with millisecond rather than micro- 17. Mattern CFT, Brackett FS, Olson BJ. Determination of number and
size of particles by electrical gating: blood cells. J Appl Physiol
second observation times, have already been used to de- 1957;10:56 –70.
tect single fluorescent molecules (65), and, if they are ever 18. Hallermann L, Thom R, Gerhartz H. Elektronische Differentialzählung
von Granulocyten und Lymphocyten nach intravitaler Fluochrom-
made available to the right users, may prove useful for ierung mit Acridinorange. Verh Deutsch Ges Inn Med 1964;70:217–
high-precision analysis of bacteria (66), viruses, and mac- 219.
romolecules (67,68). The microfluidic technology incor- 19. Kamentsky LA. Cytology automation. Adv Biol Med Phys 1973;14:93–
161.
porated in some of these systems (66,68) makes it possible 20. Kamentsky LA, Melamed MR, Derman H. Spectrophotometer: new
to stop and reverse flow. Large-scale microfluidic integra- instrument for ultrarapid cell analysis. Science 1965;150:630 – 631.
tion (69) should allow thousands of individual cells to be 21. Kamentsky LA, Melamed MR. Instrumentation for automated exami-
nations of cellular specimens. Proc IEEE 1969;57:2007–2016.
placed on an individual “chip,” subjected to various ma- 22. Koenig SH, Brown RD, Kamentsky LA, Sedlis A, Melamed MR. Efficacy
nipulations, and studied over time. Looking at the same of a rapid cell spectrophotometer in screening for cervical cancer.
Cancer 1968;21:1019 –1026.
cells over time is a tall order for a flow cytometer, but is 23. Ornstein L, Ansley HR. Spectral matching of classical cytochemistry
easily accomplished using instruments ranging down in to automated cytology. J Histochem Cytochem 1974;22:453– 469.
complexity and cost from multiphoton confocal micro- 24. Mansberg HP, Saunders AM, Groner W. The Hemalog D white cell
differential system. J Histochem Cytochem 1974;22:711–724.
scopes through laser scanning cytometers, such as those 25. Fulwyler MJ. Electronic separation of biological cells by volume.
developed by Kamentsky’s CompuCyte Corporation Science 1965;150:910 –911.
 Reprinted with permission of Cytometry Part A, John Wiley and Sons, Inc.

20 SHAPIRO

26. Kamentsky LA, Melamed MR. Spectrophotometric cell sorter. Science selection using bromodeoxyuridine/Hoechst 33342 photosensitized
1967;156:1364 –1365. cell killing. Cytometry 1988;9:143–149.
27. Sweet RG. High frequency recording with electrostatically deflected 50. Inoué S, Spring KR. Video microscopy. The fundamentals. 2nd edi-
ink jets. Rev Sci Instrum 1965;36:131–136. tion. New York: Plenum Press; 1997. 742 p.
28. Mullaney PF, Van Dilla MA, Coulter JR, Dean PN. Cell sizing: a light 51. Pawley JB, editor. Handbook of biological confocal microscopy. 2nd
scattering photometer for rapid volume determination. Rev Sci In- edition. New York: Plenum Press; 1995. 632 p.
strum 1969;40:1029 –1032. 52. Keating P, Cambrosio A. Biomedical platforms. realigning the normal
29. Steinkamp JA, Fulwyler MJ, Coulter JR, Hiebert RD, Horney JL, Mul- and the pathological in late twentieth-century medicine. Cambridge,
laney PF. A new multiparameter separator for microscopic particles MA: MIT Press; 2003. 544 p.
and biological cells. Rev Sci Instrum 1973;44:1301–1310. 53. Arndt-Jovin DJ, Jovin TM. Computer-controlled multiparameter anal-
30. Salzman GC, Crowell JM, Goad CA, Hansen KM, Hiebert RD, LaBauve ysis and sorting of cells and particles. J Histochem Cytochem 1974;
PM, Martin JC, Ingram M, Mullaney PF. A flow-system multiangle 22:622– 625.
light-scattering instrument for cell characterization. Clin Chem 1975; 54. Bailey JE, Fazel-Madjlessi J, McQuitty DN, Lee YN, Allred JC, Oro JA.
21:1297–1304. Characterization of bacterial growth by means of flow microfluorom-
31. Salzman GC, Crowell JM, Martin JC, Trujillo TT, Romero A, Mullaney etry. Science 1977;198:1175–1176.
PF, LaBauve PM. Cell identification by laser light scattering: identifi- 55. Peters D, Branscomb E, Dean P, Merrill T, Pinkel D, Van Dilla M, Gray
cation and separation of unstained leukocytes. Acta Cytol 1975;19: JW. The LLNL high-speed sorter: design features, operational charac-
374 –377. teristics, and biological utility. Cytometry 1985;6:290 –301.
32. Van Dilla MA, Trujillo TT, Mullaney PF, Coulter JR. Cell microfluorom- 56. Gray JW, Dean PN, Fuscoe JC, Peters DC, Trask BJ, van den Engh GJ,
etry: a method for rapid fluorescence measurement. Science 1969; Van Dilla MA. High-speed chromosome sorting. Science 1987;238:
163:1213–1214. 323–329.
33. Dittrich W, Göhde W. Impulsfluorometrie bei Einzelzellen in Suspen- 57. Wheeless LL, Jr. Slit-Scanning. In: Melamed MR, Lindmo T, Mendel-
sionen. Z Naturforsch 1969;24b:360 –361. sohn ML, editors. Flow cytometry and sorting. 2nd edition. New
34. Herzenberg LA, Sweet RG, Herzenberg LA. Fluorescence activated York: John Wiley and Sons, Inc.; 1990. p 109 –125.
cell sorting. Sci Am 1976;234:108 –117. 58. Cram LS. Spin-offs from the NASA space program for tumor diagnosis.
35. Saunders AM, Hulett HR. Microfluorometry: comparison of single Cytometry 2001;43:1.
measurements to a rapid flow system. J Histochem Cytochem 1969; 59. Steen HB, Lindmo T. Flow cytometry: a high resolution instrument for
17:188. everyone. Science 1979;204:403– 404.
36. Hulett HR, Bonner WA, Barrett J, Herzenberg LA. Automated separa- 60. Steen HB. Further developments of a microscope-based flow cytom-
tion of mammalian cells as a function of intracellular fluorescence. eter: light scatter detection and excitation intensity compensation.
Science 1969;166:747–749. Cytometry 1980;1:26 –31.
37. Bonner WA, Hulett HR, Sweet RG, Herzenberg LA. Fluorescence 61. Steen HB, Boye E, Skarstad K, Bloom B, Godal T, Mustafa S. Applica-
activated cell sorting. Rev Sci Instrum 1972;43:404 – 409. tions of flow cytometry on bacteria: cell cycle kinetics, drug effects,
38. Curbelo R, Schildkraut ER, Hirschfeld T, Webb RH, Block MJ, Shapiro and quantitation of antibody binding. Cytometry 1982;2:249 –257.
HM. A generalized machine for automated flow cytology system 62. Leif SB, Leif RC, Auer R. The EPICS C analyzer: an ergometrically
design. J Histochem Cytochem 1976;24:388 –395. designed flow cytometer computer system. Anal Quant Cytol Histol
39. Shapiro HM, Schildkraut ER, Curbelo R, Turner RB, Webb RH, Brown 1985;7:187–191.
DC, Block MJ. Cytomat-R: a computer-controlled multiple laser 63. Shapiro HM, Perlmutter NG. Violet laser diodes as light sources for
source multiparameter flow cytophotometer system. J Histochem cytometry. Cytometry 2001;44:133–136.
Cytochem 1977;25:836 – 844. 64. Bagwell CB, Adams EG. Fluorescence spectral overlap compensation
40. Shapiro HM. Fluorescent dyes for differential counts by flow cytom- for any number of flow cytometry parameters. Ann NY Acad Sci
etry: does histochemistry tell us much more than cell geometry? 1993;677:167–184.
J Histochem Cytochem 1977;25:976 –989. 65. Nguyen DC, Keller RA, Jett JH, Martin JC. Detection of single mole-
41. Hercher M, Mueller W, Shapiro HM. Detection and discrimination of cules of phycoerythrin in hydrodynamically focused flows by laser-
individual viruses by flow cytometry. J Histochem Cytochem 1979; induced fluorescence. Anal Chem 1987;59:2158 –2161.
27:350 –352. 66. Fu AY, Chou HP, Spence C, Arnold FH, Quake SR. An integrated
42. Loken MR, Parks DR, Herzenberg LA. Two-color immunofluorescence microfabricated cell sorter. Anal Chem 2002;74:2451–2457.
using a fluorescence-activated cell sorter. J Histochem Cytochem 67. Goodwin PM, Johnson ME, Martin JC, Ambrose WP, Marrone BL, Jett
1977;25:899 –907. JH, Keller RA. Rapid sizing of individual fluorescently stained DNA
43. Bacus JW, Gose EE. Leukocyte pattern recognition. IEEE Trans Sys fragments by flow cytometry. Nucleic Acids Res 1993;21:803– 806.
Man Cybernet 1972;SMC-2:513–525. 68. Chou H-P, Spence C, Scherer A, Quake S. A microfabricated device
44. Miller MN. Design and clinical results of HEMATRAK: an automated for sizing and sorting DNA molecules. Proc Natl Acad Sci USA 1999;
differential counter. IEEE Trans Biomed Eng 1976;BME-23:400 – 405. 96:11–13.
45. Young IT. The classification of white blood cells. IEEE Trans Biomed 69. Thorsen T, Maerkl SJ, Quake SR. Microfluidic large scale integration.
Eng 1972;BME-19:291–298. Science 2002;298:580 –584.
46. Schindler ML, Olinger MR, Holland JF. Automated analysis and sur- 70. Kamentsky LA, Kamentsky LD. Microscope-based multiparameter la-
vival selection of anchorage-dependent cells under normal growth ser scanning cytometer yielding data comparable to flow cytometry
conditions. Cytometry 1985;6:368 –374. data. Cytometry 1991;12:381–387.
47. Martin JC, Jett JH. Photodamage, a basis for super high speed cell 71. Kamentsky LA, Burger DE, Gershman RJ, Kamentsky LD, Luther E.
selection. Cytometry 1981;2:114. Slide-based laser scanning cytometry. Acta Cytol 1997;41:123–143.
48. Shapiro HM. Apparatus and method for killing unwanted cells. United 72. Darzynkiewicz Z, Bedner E, Li X, Gorczyca W, Melamed MR. Laser-
States Patent 4,395,397, 1983. scanning cytometry: a new instrumentation with many applications.
49. Herweijer H, Stokdijk W, Visser JWM. High speed photodamage cell Exp Cell Res 1999;249:1–12.