Discussion
We assessed the linearity between two common measurements for the size
of yeast populations in laboratory cultures, cell density (OD, obtained
from spectrophotometry) and cell count (obtained from flow cytometry),
across all eight species of Saccharomyces (Figure 1). The
steepness of the slope of this relationship is informative for
researchers conducting studies across species and strain backgrounds,
comparing the competitive fitness of stains, or specific aspects of
their growth kinetics. We found that the average slope of the
relationship between cell density and cell count significantly differs
between species. Especially S. cerevisiae has a significantly
steeper slope than most other species, i.e. they produce higher cell
counts at lower OD values. Interestingly, the patterns in slope
variation that we observed across the genus, closely follow the
phylogenetic relationship of the species, with more closely related
pairs (S. cerevisiae /S paradoxus , S.
mikatae /S. jurei , and S. eubayanus /S. uvarum )
having more similar slopes (Figure 2). This suggests that the species-
and strain- specificity of this relationship should be taken into
account when setting up and interpreting results of competitive fitness
assays, or when fine-tuning cell titers and inocula of non-commercial
strains for industrial applications. It also shows the limitations of
using only a single laboratory S. cerevisiae strain as a
reference point when exploring the population biology of wild strains.
We then speculated that differences in cell size may explain variation
in the steepness of the slopes between species. Variation in cell size
is expected, given the large diversity of ecological niches these
species inhabit (Table S1), and has been described in S.
cerevisiae as a result of temperature and nutrient availability
(Kellogg & Levin, 2022), affecting basic physiological functions such
as protein synthesis and cell division rate . The intensity and the
radius of light scattering in the spectrophotometer depends on cell
size, thus affecting the absorbance of the microbial culture . Indeed,
we found that species significantly differ in average cell size, withS. eubayanus having the smallest (1.16 ± 0.12 log
µm2) and S. cerevisiae having the largest cells
(1.4 ± 0.11 log µm2, Figure 3). But against our
expectation, that strains with larger average cells would give lower
cell counts at similar OD readings, increasing cell size did not predict
a decrease in slopes from cell count – cell density regressions.
Instead, we found the opposite to be true. Slopes increased slightly
with increasing cell size (Figure 4), suggesting that at a given
OD-value, strains with larger cells also produce higher cell counts.
Besides cell size, other species-specific cellular features may affect
the cell count – cell density relationship. The species investigated
here are ecologically and genetically vastly divergent and likely differ
in the composition of their cell wall, determining its strength and
rigidity . For instance, chitin, chitosan, and cell wall proteins have
been shown to have different refractive indices that can affect OD
readings . Wall-resident proteins have diversified rapidly over
evolutionary time in Saccharomyces as a result of gene silencing
through epigenetic mechanisms and environmentally induced expression
regulation, providing adaptability to different habitats and lifestyles
. Chitin and chitosan are also known to vary in structure between fungal
species . Besides cell wall composition, the number and structure of bud
scars may also affect the cell’s refractive index . Mother cells
accumulate chitinous scar tissue from cytokinesis over their lifetime
and different growth conditions (poor vs . rich media) can lead to
variation in bud scar number . If species vary in the average number or
structure of bud scars a mother cell carries (e.g. due to heritable
differences in cell longevity or species-specific responses to nutrient
availability), this may change their refractive indices and affect the
translation of optical density readings into cell counts.