389
Preparative cell sorting: Sorting large numbers of cells for biochemical and other studies
remains a tantalizing prospect: obtaining pure samples of chosen subsets of the particles in the
ocean would enable us to answer many questions about plankton communities. For example,
direct measurements of C:N or C:Chlorophyll ratios in pure samples of phytoplankton
uncontaminated by detritus could tell us whether N is limiting growth in the oligotrophic
oceans, and stable isotope ratios of selected classes of particles could tell us much about
marine food chains. Flow cytometry is a tool with potential for providing these samples, but
the sheer numbers of particles required for most analytical techniques is daunting for a sorter
that operates on individuals. Consequently, the number of studies in which preparative cell
sorting has been utilized is still rather small.
In examples described above, samples were sorted for pigment analyses by fluorescence
techniques (Olson et al. 1988; Chisholm et al. 1988) which can be very sensitive. Similarly,
sorting of relatively small numbers of cells can be useful in physiological studies utilizing
radioisotopes, though it has been shown that the relatively harsh sorting step must take place
at the end of the experiment (Rivkin et al. 1986). Using this approach, relative uptake rates
of 14C-bicarbonate were demonstrated to vary considerably between different constituents of
the phytoplankton in a coastal seawater sample (Li 1986). In addition, the first evidence that
prochlorophyte picoplankton were actually phototrophic came from 14C sorting experiments
(Fig. 24). In another application, preparative cell sorting has been used to "synchronize"
phytoplankton cultures growing in continuous conditions by selecting large cells ready to
divide (Campbell and Yentsch 1989; Campbell et al. 1989; Armbrust et al. 1990). Druehl et
al. (1989) detected differences in the DNA stain fluorescence of male and female brown algal
meiospores (previously indistinguishable) by sorting and subsequent cultivation. Subba Rao
et al. (1991), while investigating the life history and domoic acid production of the diatom
Nitzchia pungens, went a step further and mixed sorted male and female gametes
(distinguished based on light scatter properties) to produce new vegetative cells.
However, for many of the most exciting applications, such as measuring the chemical
composition of individual species in the field, the number of cells required is too large to
allow sorting to be a practical method, given existing analytical constraints. For example, a
typical measurement of particulate N requires a sample of at least 10 J.l.g N (e.g. Lee and
Preparative cell sorting: Sorting large numbers of cells for biochemical and other studies
remains a tantalizing prospect: obtaining pure samples of chosen subsets of the particles in the
ocean would enable us to answer many questions about plankton communities. For example,
direct measurements of C:N or C:Chlorophyll ratios in pure samples of phytoplankton
uncontaminated by detritus could tell us whether N is limiting growth in the oligotrophic
oceans, and stable isotope ratios of selected classes of particles could tell us much about
marine food chains. Flow cytometry is a tool with potential for providing these samples, but
the sheer numbers of particles required for most analytical techniques is daunting for a sorter
that operates on individuals. Consequently, the number of studies in which preparative cell
sorting has been utilized is still rather small.
In examples described above, samples were sorted for pigment analyses by fluorescence
techniques (Olson et al. 1988; Chisholm et al. 1988) which can be very sensitive. Similarly,
sorting of relatively small numbers of cells can be useful in physiological studies utilizing
radioisotopes, though it has been shown that the relatively harsh sorting step must take place
at the end of the experiment (Rivkin et al. 1986). Using this approach, relative uptake rates
of 14C-bicarbonate were demonstrated to vary considerably between different constituents of
the phytoplankton in a coastal seawater sample (Li 1986). In addition, the first evidence that
prochlorophyte picoplankton were actually phototrophic came from 14C sorting experiments
(Fig. 24). In another application, preparative cell sorting has been used to "synchronize"
phytoplankton cultures growing in continuous conditions by selecting large cells ready to
divide (Campbell and Yentsch 1989; Campbell et al. 1989; Armbrust et al. 1990). Druehl et
al. (1989) detected differences in the DNA stain fluorescence of male and female brown algal
meiospores (previously indistinguishable) by sorting and subsequent cultivation. Subba Rao
et al. (1991), while investigating the life history and domoic acid production of the diatom
Nitzchia pungens, went a step further and mixed sorted male and female gametes
(distinguished based on light scatter properties) to produce new vegetative cells.
However, for many of the most exciting applications, such as measuring the chemical
composition of individual species in the field, the number of cells required is too large to
allow sorting to be a practical method, given existing analytical constraints. For example, a
typical measurement of particulate N requires a sample of at least 10 J.l.g N (e.g. Lee and