Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Journal Pre-proof
Quantitative determination of Neuronal Size and Density using Flow
Cytometry
Farrow L.F. (Conceptualization) (Methodology) (Validation) (Formal
analysis) (Investigation) (Writing - original draft) (Writing - review
and editing) (Visualization)<ce:contributor-role>Funding
acquisition), Andronicos N.M (Conceptualization) (Methodology)
(Validation) (Formal analysis) (Investigation) (Writing - review and
editing)<ce:contributor-role>Visualization), McDonald P.G.
(Supervision) (Funding acquisition) (Writing - review and editing),
Hamlin A.S. (Conceptualization) (Methodology) (Validation) (Formal
analysis) (Investigation) (Writing - review and editing) (Visualization)
(Resources) (Data curation)
PII: S0165-0270(21)00016-9
DOI: https://doi.org/10.1016/j.jneumeth.2021.109081
Reference: NSM 109081
To appear in: Journal of Neuroscience Methods
Received Date: 26 February 2020
Revised Date: 19 December 2020
Accepted Date: 14 January 2021
Please cite this article as: Farrow LF, Andronicos NM, McDonald PG, Hamlin AS, Quantitative
determination of Neuronal Size and Density using Flow Cytometry, Journal of Neuroscience
Methods (2021), doi: https://doi.org/10.1016/j.jneumeth.2021.109081
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.
Quantitative determination of Ne uronal Size and Density using Flow Cytometry.
Farrow, L.F.,1 Andronicos , N.M,2 McDonald, P.G.,1 Hamlin, A.S.2*
1Animal Behaviour and Ecology Laboratory, School of Environmental and Rural Science,
Faculty of Science, Agriculture, Business and Law, University of New England, Armidale,
NSW, Australia
2 Brain Behaviour Research Group, School of Science and Technology, Faculty of Science,
Agriculture, Business and Law, University of New England, Armidale, NSW, Australia
Corres pondence:
Dr Adam Hamlin
School of Science and Technology
Faculty of Science, Agriculture, Business & Law
University of New England
Armidale NSW 2351 AUSTRALIA
ahamlin@une.edu.au
Highlights:
Flow cytometric analysis represents a superior method for nuclear size and number
determination
Flow cytometry provides a methodological approach that allows for consistency in
research
Flow cytometry methods will allow comparable brain morphology analysis across
taxa
Journal Pre-proof
2
Abstract
Background: Recent anthropomorphic disturbances are occurring at an increasing
rate leading to organisms facing a variety of challenges. This change is testing the
information processing capacity (IPC) of all animals. Brain function is widely accepted to be
influenced by a variety of factors, including relative size, number of neurons and neuronal
densities. Therefore, in order to understand what drives an animals IPC, a methodological
approach to analyze these factors must be established.
New method: Here we created a protocol that allowed for high-throughput, non-biased
quantification of neuronal density and size across six regions of the brain. We used the
Isotropic Fractionator method in combination with flow cytometry to identify neuronal and
non-neuronal cells in the brains of adult rats.
Comparison with existing methods: The results obtained were comparable to those
identified using stereological counting methods.
Results: By employing this new method, the number of nuclei in a specific brain region can
be compared between replicate animals within an experiment. By calibrating the forward
scatter channel of the flow cytometer with size standard beads, neuronal and non-neuronal
nuclear sizes can be estimated simultaneously with nuclei enumeration. These techniques for
nuclear counting and size estimation are technically and biologically reproducible.
Conclusion: Use of flow cytometry provides a methodological approach that allows for
consistency in research, so that information on brain morphology, and subsequent function,
will become comparable across taxa.
Keywords: Isotropic fractionator; Flow Cytometry; High-throughput; Ne uronal Size;
Neuronal De nsity; Neuron Glia Ratios
Journal Pre-proof
3
1. Introduction.
Organisms face a variety of physical and cognitive challenges throughout their lives.
Recently, anthropomorphic disturbances, including climate change, are occurring at an
escalated rate (Thuiller, Albert et al. 2008). This increase in change and variability of
conditions is testing the information processing capacity (IPC) of all animals (Sih, Ferrari et
al. 2011), wherein working memory and mental manipulation will be key to survival at both
the species and community level (Berg and Ellers 2010). Consequently, an array of studies
aim to identify the impact IPC has on a species ability to adapt to rapid changes, yet
understanding why IPC varies across species remains controversial (Holekamp, Swanson et
al. 2013, Callaghan, Major et al. 2019).
Brain function, and subsequently IPC, is widely accepted to be influenced by a variety
of factors, including relative size (Sol 2005, Deaner 2007), number of neurons and neuronal
densities (Dicke 2016), glial to neuronal ratios (Herculano‐Houzel 2014) and potentially
size of neuronal and non-neuronal cells (Herculano-Houzel and Lent 2005). Therefore, in
order to understand what drives an animals IPC, a methodologica l approach to analyze these
factors must be established.
Originally, neuronal counts and densities were estimated using stereological methods,
such as the optical dissector (Sterio 1984), that were restricted to well-defined structures and
measurable volumes (West 1999). To account for these limitations Herculano-Houzel and
Lent, (2005) developed the Isotropic Fractionator (IF) method that allowed for semi-
quantitative counts of neuronal and non-neuronal cells regardless of density. This method
provided a reliable and reproducible means of measuring total cell numbers across
neuroanatomical regions. While successful, this approach is labor intensive and requires
Journal Pre-proof
4
stereological counting. As such, Young et al. (2012) elaborated upon the IF method and
employed flow cytometry cell counting technology to analyze the proportion of neurons in
cortical regions of baboons (Papio hamadryas Anubis) that was analogous to the IF method
employed by Herculano-Houzel and Lent (2005).
Flow cytometric technology enables the rapid analysis of single cells or particles (e.g.
cell nuclei) as they flow past lasers while suspended in a buffered salt-based solution
(McKinnon 2018). Semi-quantitative parameters such as particle sizes (Forward Scatter,
FSC) and granularity (Side Scatter, SSC) can be easily determined. The nuclei of cells are
identifiable by staining the DNA with propidium iodide (PI), a red fluorescent dye (Deitch,
Law et al. 1982). Labeling of cells or particles with fluorescent antibodies and/or dyes enable
identification of cells or particles in a complex solution. This method results in high-
throughput cell population estimates allowing the creation of neuronal distribution maps.
While these methods allow for semi-quantitative analysis of neuronal and non-neuronal cell
counts, there is a need to understand the size and overall composition of these cells across
brain regions and the brain overall.
Calibration of flow cytometers with beads of different sizes converts semi-
quantitative forward scatter parameter into an absolute physical dimension. Thus, it is
hypothesized that the number and size of cell nuclei within a tissue homogenate, spiked with
enumeration beads can be efficiently quantified using a flow cytometer that has been
calibrated with size beads. Therefore, the aim of this study was to define a reproducible,
quantitative flow cytometric method for neuron and non-neuronal cell enumeration and
characterization that minimizes errors associated with traditional semi-quantitative
stereological techniques thereby increasing the throughput and reliability of acquired data
Journal Pre-proof
5
obtained within a given sample. To achieve this, quantitative flow cytometric analysis of
NeuN and PI labelled nuclei in homogenates isolated from the six anatomical regions of the
brain were performed to determine the neuronal (NeuN+) and non-neuronal (NeuN-) nuclear
(PI+) densities and sizes in each region.
This method allows for intra-sample sampling to ensure repeatability of results, as
well as an ability to compare regions within a given individual and across various specimens.
It is anticipated that this methodological approach will thus make it possible to compare
factors of the brain postulated to be responsible for variation in function (particularly IPC)
across individuals of various ages and/or species.
2. Mate rials and Methods.
2.1 Animals
Experimentally naive male (n=6) and female (n=3) adult Wistar rats (10-12 weeks) (250–350
g) were obtained from The Centre for Research and Teaching, University of New England.
The procedures were approved by the University of New England Animal Ethics Committee
(AEC18-132) and conducted in accordance with the National Institutes of Health Guide for
the Care and Use of Laboratory Animals (NIH Publications No. 8023) revised 1996. The
procedures were designed to minimize the number of animals used and their suffering.
2.2 Tissue preparation and flow cytometric analysis
Rats were deeply anesthetized with sodium pentobarbital (100 mg/kg i.p.) and perfused
transcardially with 100 ml of 0.9% saline, containing 1.25ml 1% sodium nitrite and 0.036ml
heparin sodium (5000 i.u./ml), followed by 500 ml of 4% paraformaldehyde in 0.1 M
phosphate buffer (PB), pH 7.4. Brains were dissected and post-fixed for 1 hour in the same
fixative before being placed in 1% NaN3 in 0.1 M PBS for a minimum of three days. Each
Journal Pre-proof
6
brain was weighed (whole brain post-fixative weight) and mid-sagittally hemisected. The
brain was then micro dissected into the regions identified by (Olkowicz 2016); tectum,
cerebellum, cortex, forebrain, brain stem and the diencephalon (Figure 1).
Figure 1: Dissected regions of rat brain for flow cytometric analysis. Midsagittal section of the rat brain
showing regions dissected for flow cytometric analysis. Tectum (T), cerebellum (CB), cortex (CTX), forebrain
(FB), brain stem (BS), diencephalon (D) and the optic bulb (OB).
Once dissected, the regions were weighed before being placed into 1.5 ml microfuge tubes
containing 1 ml homogenization buffer (40mM sodium citrate buffer (pH 6.0) with 1% Triton
X-100 (Herculano-Houzel and Lent 2005) The tissue was then placed in a glass dounce
homogenizer (Kontes glass company tube rod = C35, USA) and homogenized for 60 seconds
to form an isotropic suspension of cell debris including nuclei. To further reduce between
sample artefacts in the sample, the isotropic suspension was then filtered through a 30 μm
pre-separation filter (Miltenyi Biotec, USA) placed directly over a 15 mL centrifuge tube.
Journal Pre-proof
7
The 30 µm filter was washed with an additional 1 ml of homogenization buffer to
quantitatively collect the nuclei in the filtrate and 100 µl samples of the isotropic suspension
were placed into microfuge tubes and centrifuged at 300g for 3 minutes at room temperature.
An additional 100 µl sample of the isotropic suspension for each brain region was collected
(n=4) for biological replication. The supernatant was discarded, and the pellet resuspended in
100 µl of homogenization buffer containing a 1:100 dilution of mouse anti-NeuN monoclonal
antibody (EMD Millipore, USA) or isotype control antibody (EMD Millipore, USA) for 30
minutes on ice.
The brain tissue lysate was washed by adding 1 ml of ice-cold PBS (pH 7.4) and centrifuged
at 300g for 3 minutes and the supernatant discarded. The pellets were resuspended in 100 µl
of homogenization buffer containing a 1:200 dilution of goat anti-mouse IgG-FITC antibody
(EMD Millipore, USA) for 30 minutes on ice in the dark. Brain tissue lysate was washed by
adding 1 ml of PBS (pH7.4) and centrifuged at 300g for 3 minutes. The supernatant was
discarded and the cells resuspended in 100 µl of PBS (pH 7.4) and were stored on ice in the
dark.
Just prior to flow cytometric acquisition, 1 µl of 1 mg/ml propidium iodide (PI) was added to
each sample, along with 50 μL of precision count beads (Biolegend, USA). The sample tubes
were inverted six times to homogenously mix the beads and PI stained nuclei. Flow
cytometric data were acquired using a FlowSight imaging flow cytometer (Merck, USA).
Data acquisition was completed for each sample when the number of events in the single
bead counting gate reached 500. This process was repeated twice for four individual samples
for each brain region for technical replication. To quantify nuclear size the forward scatter
Journal Pre-proof
8
(FSC) scale of the flow cytometer was converted into microns via the acquisition of size
calibration beads (1 to 15 µm; Thermoscientific, USA) using the same cytometer settings.
2.3 Statistical analysis
Statistical analyses of flow cytometric data were performed using the SPSS statistical
analysis package, version 25 (IBM, USA). For flow cytometric data analysis only single
particle events were analysed. Neuronal nuclei (PI+NeuN+) and non-neuronal nuclei
(PI+NeuN-) were gated and the size distribution determined relative to the FSC (mean ± SD)
of the size calibration beads for each brain region. For the multivariate comparison of
neuronal and non-neuronal nuclei in the various brain regions the Box's Test of Equality of
Covariance Matrices was not significant (p = 0.214), therefore linear regression ANOVA
with Tukey’s post-hoc tests were used as the tests of significance. The significance of
neuronal nuclear size distribution was determined by Kruskal-Wallis test for each brain
region as these data were not normally distributed. The distribution of non-neuronal cell
nuclei and neuronal nuclei had a skewed distribution, therefore, to determine significant
differences of the neuronal nuclei to non-neuronal nuclei for the various brain regions Mann-
Whitney U tests were performed. For all statistical tests of significance, a p-value < 0.05 was
deemed significant.
3. Results
3.1 Flow cytometric gating strategy and reproducibility.
Figure 2 describes the flow cytometric gating strategy used to quantify the size and number
of neuronal (PI+NeuN+) and non-neuronal (PI+NeuN-) nuclei in various regions of the rat
brain. Only single particle signals were used to generate flow cytometric data (Figure 2A). A
scatter plot of FSC against PI fluorescence intensity of the single particle cell debris was
Journal Pre-proof
9
plotted to identify (gate) the population of single nuclei (PI+ particles; Figure 2B). By only
analyzing single particles there was a clear demarcation between the spiked enumeration
beads and particles in the homogenate. Histograms of PI+ nuclei stained with either NeuN
antibody to identify neuronal nuclei or an irrelevant isotype control antibody were used to
define the NeuN+ nuclei gate relative to the isotype control (NeuN-) signal (Figure 2C). The
FSC against NeuN fluorescence intensity scatter plot of PI+ particles (nuclei; Figure 2D) were
used to differentiate between neuronal and non-neuronal nuclei and demonstrated that
PI+NeuN+ neuronal nuclei had a greater size distribution than PI+NeuN- non-neuronal nuclei.
Once neuronal and non-nuclei populations were defined, histograms of the non-neuronal
nuclei (PI+NeuN-) and neuronal nuclei (PI+NeuN+; Figure 2F) were plotted and superimposed
with bead size (mean FSC ± 1 SD; from Figure 2E) calibration gates to determine the number
of nuclei within each size gate, thereby quantifying the non-neuronal (PI+NeuN-) and
neuronal (PI+NeuN+) nuclei sizes in the respective homogenates from the various anatomical
locations of the brain. Thus, the gating strategy used, facilitated the identification of nuclei
within the cell debris which enabled the quantitation of neuronal and non-neuronal nuclear
sizes and numbers in the different regions of the rat brain by two methods. First, overlaying
bead size FSC gates (mean FSC ± 1 SD; Figure 2E) on to figure 2F histograms or second, by
using the equation from the bead diameter against FSC standard curve (Figure 2E) to
calculate the physical size from the FSC of each particle from figure 2F.
Journal Pre-proof
10
Figure 2: Imaging flow cytometric gating strategy. (A) A scatter plot of Area against Aspect Ratio to define
single particles within the brain tissue homogenates. (B) A scatter plot of single particles of forward scatter
(FSC) against propidium iodide (PI) fluorescent intensity used to define the PI+ nuclei in the cell debris. (C)
NeuN channel intensity histograms of brain homogenates stained with either NeuN antibody or an irrelevant
isotype control antibody. (D) FSC against NeuN scatter plot of the PI+ nuclei population from (B) used to define
neuronal (PI+NeuN+) and non-neuronal (PI+NeuN-) nuclei populations using NeuN+ and NeuN- gates defined in
(C) as well as flow cytometric micrographs defining the staining pattern of neuronal and non-neuronal nuclei.
(E) Bead dimensions against FSC for the size calibration beads (1 to 15 µm). (F). Histograms of PI+NeuN- non-
neuronal nuclei and PI+NeuN+ neuronal nuclei with superimposed bead size (median ± SD from (E)) calibration
gates.
The reproducibility of the nuclei sampling technique employed was determined (Figure 3).
Homogenates from different brain regions were stained with PI and NeuN, spiked with
enumeration beads and acquired on the flow cytometer. The gating strategy described in
figure 2 was used for all technical replicate comparisons. Overall, there was a significant (p =
Journal Pre-proof
11
0.001) difference between technical replicates. As expected, the variation between different
anatomical regions of the brain was responsible for the significant (p = 0.002) difference
observed for the total technical replicate variation. In contrast, there were no significant
differences between technical replicates when nuclei type or antibody stain parameters.
However, nuclear sizes between 4 and 6 µm had greater, but non-significant variability
between technical replicates. Collectively these data suggested that the reproducibility of this
flow cytometric technique for nuclei size and enumeration was adequate.
Figure 3: Reproducibility of sizes of neuronal and non-neuronal nuclei for the various regions of the rat brain.
3.2 Quantitative flow cytome tric nuclei count analysis
The first step in defining the neuronal density of various anatomical regions of the rat brain
was to quantify total neuronal (PI+NeuN+) and non-neuronal (PI+NeuN-) nuclei in the
homogenates using quantitative flow cytometry (Figure 4). The quantitative acquisition of
Journal Pre-proof
12
nuclei in brain homogenates were standardized via the addition of a known quantity of
enumeration beads and creating a stopping gate when a set number of beads had been
acquired. In the brain stem, cortex, diencephalon, forebrain and tectum regions of the rat
brain there were significantly (p = 1.2×10-28) more (6-22 fold) non-neuronal nuclei
(PI+NeuN-) compared to neuronal nuclei (PI+NeuN+; Figure 4). In contrast, the numbers of
non-neuronal (PI+NeuN-) and neuronal (PI+NeuN+) nuclei in the cerebellum were not
significantly different (Figure 4).
Figure 4: Total number of neuronal (PI+NeuN+;) and non-neuronal (PI+NeuN-;) nuclei in the different
anatomical regions of the rat brain as determined by quantitative flow cytometric analysis. * p < 0.05 non-
neuronal nuclei compared to neuronal nuclei numbers. # p < 0.05 for neuronal nuclei comparisons across brain
regions.
3.3 Quantitative flow cytometric nuclei count analysis
Journal Pre-proof
13
Two methods for determining nuclear sizes from the flow cytometric FSC data were
evaluated. First, neuronal and non-neuronal nuclear sizes were stratified by overlaying the
FSC parameter for each bead size plus/minus one standard deviation as gates on the
PI+NeuN+ (neuronal) and PI+NeuN- (non-neuronal) nuclei FSC histograms (Figure 2F). For
this analysis the size by nuclear type varied significantly (p = 4.7×10-51) and accounted for
18.2% of the total variation. Figure 5 demonstrated that quantitative flow cytometric analysis
of neuronal nuclear size indicated that there were significant (p < 0.05) enrichments of
neuronal nuclear sizes between 6-10 µm in the brain stem, cortex and the diencephalon
regions of the brain when compared to 1 µm neuronal nuclei (Figure 5). Similarly, there was
a significant enrichment of 4-10 µm neuronal nuclei in the forebrain and tectum regions of
the rat brain compared to 1 µm neuronal nuclei. Finally, the cerebellum contained a more
homogenous neuronal nuclei composition with 6 µm nuclei being significantly enriched
relative to 1 µm nuclei.
In contrast, non-neuronal nuclei had smaller nuclear sizes than neuronal nuclei (Figure 5).
Specifically, there were significantly more 1 and 2 µm non-neuronal nuclei in all the brain
regions compared to neuronal nuclei.
Journal Pre-proof
14
Figure 5: Nuclear size distribution of neuronal (PI+NeuN+) and non-neuronal (PI+NeuN-) nuclei defined using
the FSC ± 1 SD for each size calibration bead overlaid as gates onto each of the PI+NeuN+ and PI+NeuN-
histograms. p<0.05 for neuronal nuclear sizes relative 1 µm neuronal nuclei and p<0.05 for non-neuronal
nuclear sizes relative 10 µm non-neuronal nuclei.
The second nuclear size analysis method involved using the equation defined by the bead size
against FSC standard curve (Figure 2E) to convert FSC data into micron dimensions for each
PI+ particle (Figure 6). The bead size calibration standard curve enabled the estimation of
nuclei numbers per micron from 1 to 14 µm and demonstrated that neuronal nuclear sizes
were normally distributed for the cerebellum, cortex, diencephalon and brain stem having 6.6,
7.1, 7.0 and 7.3 µm as their average nuclear sizes, respectively, whereas the forebrain and
tectum had smaller neuronal nuclear sizes with mean nuclear size of 6.0 and 5.8 µm,
respectively. In contrast, smaller non-neuronal nuclei were present in these same anatomical
regions but demonstrated a skewed size distribution and demonstrated a clear reduction in the
number of non-neuronal nuclear sizes greater than at 4 µm for each region other than the
Journal Pre-proof
15
tectum (data not shown). However, the number of non-neuronal nuclei estimated using the
standard curve method were approximately 2 orders of magnitude above the estimates using
the histogram gating method (data not shown).
Journal Pre-proof
16
Figure 6: Estimates of neuronal nuclear size distributions in various brain anatomical regions using the standard
curve estimation method.
3.4 Neuronal to non-neuronal ratios
The non-neuronal to neuron densities for each brain region were calculated by combining
flow cytometric nuclear estimates with the tissue weight of anatomical brain regions (Table
1). Non-neuronal : neuronal ratios for the anatomical regions of the brain ranged from 0.8:1
in the cerebellum up to 13.5:1 in the diencephalon. The average non-neuronal : neuronal ratio
across all the studied brain anatomical regions was estimated at 7.5:1 ± 4.10.
Journal Pre-proof
17
Table 1: Average weights (g) of brain regions and the mean non-neuronal and neuronal densities expressed as a
ratio of flow cytometric nuclei counts per brain tissue weight (cells/g).
aThe flow cytometric nuclei count data used to calculate the PI+NeuN- (non-neuronal) and PI+NeuN+ (neuronal)
nuclei densities were derived from the total nuclei count data of the PI+NeuN- and PI+NeuN+ histograms (Figure
2F), respectively for each anatomical brain region. n = 8; means ± SEM. *p < 0.05 relative to the cerebellum
neuronal nuclei density. # p < 0.05 relative to the forebrain.
4. Discussion
Quantitative flow cytometric determination of cell nuclei numbers and sizes has several
advantages over microscopy-based nuclei enumeration methods, including higher sample
throughput and reproducibility. However, before flow cytometry can be used for determining
cell nuclei number and sizes two criteria must be satisfied: First, only single particles should
be analyzed. Second, semi-quantitative flow cytometric data must be converted into
quantitative data by calibration of the flow cytometer with counting and size beads. These
beads are readily available from numerous biotech supply companies. To identify nuclear
material in the brain region homogenates the DNA were labelled with PI. Neuronal nuclei in
these brain homogenates also expressed NeuN antigen (i.e. PI+ NeuN+). These labelled
samples were spiked with a set volume of cell counting beads and acquisition of data from
the homogenates stopped when a pre-defined target number of beads (500) in each sample
was detected by the bead gate of the flow cytometer. By employing this method, the number
of nuclei in a specific brain region could be compared between replicate animals within an
Brain Region
Tissue Weight
(g)
aNon-Neuronal Density
(104 cells/g)
aNeuronal Density
(104 cells/g)
Non-
Neuronal:Neuronal
Cortex
0.44 ± 0.009
8.2# ± 0.9870
0.94* ± 0.1976
8.7:1
Forebrain
0.050 ± 0.01
71.4 ± 24.644
8.74 ± 1.858
8.1:1
Diencephalon
0.08 ± 0.012
44.9 ± 13.211
3.33 ± 0.4096
13.5:1
Tectum
0.03 ± 0.002
55.0 ± 12.918
8.74 ± 1.773
6.3:1
Cerebellum
0.14 ± 0.006
87.1#± 2.040
10.84 ± 3.923
0.8:1
Brain Stem
0.13 ± 0.006
19.3 ± 3.965
2.53 ± 0.6598
7.6:1
Journal Pre-proof
18
experiment. In addition, by calibrating the FSC channel of the flow cytometer with size
standard beads, neuronal and non-neuronal nuclear sizes could be estimated simultaneously
with nuclei enumeration. These techniques for nuclear counting and size estimation were
technically and biologically reproducible.
For nuclear size determination there were two possible methods for estimating nuclear size
stratification. First, overlaying the neuronal and non-neuronal histograms with the FSC of
bead sizes plus/minus one standard deviation. This analysis method provided data similar to
previous studies (Sterio 1984, Herculano-Houzel and Lent 2005, Young, Flaherty et al. 2012)
because only nuclear sizes that were within a single standard deviation were considered
thereby reducing overestimation errors. An alternative analytical method to stratify nuclear
sizes is to define the equation for the FSC bead calibration curve and use this equation to
convert FSC of PI+NeuN+ (neuronal) and PI+NeuN- (non-neuronal) nuclei into physical
sizes. The equation-based stratification method was comparable to the nuclear size overlay
method for neuronal nuclei. However, for non-neuronal nuclei the equation-based method
was too sensitive and resulted in a massive overestimation of non-nuclear material that were
PI+. Possible reasons for this discrepancy in nuclear estimates between the two calculation
techniques may be DNA cross contamination. For example, neuronal nuclei were identified
by the presence of PI and the nuclear antigen NeuN. Thus, NeuN antigen differentiated
neuronal nuclei from PI+ non-neuronal signals, which may be composed of non-neuronal
nuclei, organelles with extra-nuclear DNA such as mitochondria or free DNA in the
homogenate non-specifically binding to non-nuclear material. To improve the estimation of
the non-neuronal content of specific brain regions nuclear markers for glial cells and vascular
cells for example are required. Digestion of homogenates with DNase to reduce extranuclear
DNA contamination may provide short term improvements to the accuracy of non-neuronal
Journal Pre-proof
19
counts. Therefore, the current histogram overlay size gate method was preferred over the size
standard equation estimation method because the PI+ particles included in the estimation were
within 1 standard deviation of the FCS of the bead, thereby reducing error. In contrast, sizes
estimated using the equation-based method were rounded to the nearest whole micron. The
limitation with the gating overlay method is that nuclear sizes may be under-estimated. For
example, 3, 5, 7, 8, 9, 11, 12, 13 and 14 µm sizes were not estimated using this method.
The methodology described herein allowed for identification of a significant difference in the
number of neurons present across the six regions dissected. Primarily, the cerebellum
contained significantly more neurons than the forebrain, cortex, brain stem, diencephalon and
tectum. These results are comparable to previous studies, in particular Herculano-Houzel and
Lent (2005) that developed the isotropic fractionator (IF) using stereological counting
methods that are more time-consuming and capture less events.
Previous research has stated that glial cells are more prominent than neuronal cells in the
brain (Allen and Barres 2009). This is the result of glia playing a crucial role in metabolic
support for neurons (Sherwood, Stimpson et al. 2006) and control of synaptic formation
(Ullian, Sapperstein et al. 2001). While this was seen in the majority of brain regions
analyzed in this study the cerebellum was an exception in that it offered a near 1:1 non-
neuronal to neural nuclei ratio, possibly as a result of the neuronal density of granule cells in
the cerebellum. However further research is required to investigate this potential link.
The original method of IF (Herculano-Houzel and Lent 2005) identified the main
disadvantage of IF as the inability to identify cell composition as the very nature of
homogenization is to destroy the brain tissue. However, here, through overlay of size
Journal Pre-proof
20
calibration bead gates onto PI+NeuN+ and PI+NeuN- FSC histograms, we were able to
identify nuclear sizes (and density of these various sizes) across the six brain regions
analyzed.
In conclusion, we have demonstrated a novel method of using IF to obtain accurate estimates
of neuron sizes and density, across major regions of the brain, rapidly and with minimum
opportunity for human error that is potentially applicable across a range of taxa. Further, it is
anticipated that in creating a methodology that is concise and obtainable for many
laboratories that have access to a flow cytometer, future results regarding brain morphology
across taxa will be comparable. This in turn, would help shed some forward scatter light onto
what makes a brain intelligent.
Credit Author Statement
Lucy F. Farrow: Conceptualization, Methodology, Validation, Formal analysis,
Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization, Funding
acquisition
Nicholas M. Andronicos: Conceptualization, Methodology, Validation, Formal
analysis, Investigation, Writing - Review & Editing, Visualization
Paul, G. McDonald: Supervision, Funding acquisition, Writing - Review & Editing
Adam, S. Hamlin: Conceptualization, Methodology, Validation, Formal analysis,
Investigation, Review & Editing, Visualization, Resources, Data Curation
Journal Pre-proof
21
Ethical Standards:
The authors certify that these experiments were carried out in accordance with the
National Institute of Health Guide for the Care and Use of Laboratory Animals (NIH
Publications No. 80-23) revised 1996 or the UK Animals (Scientific Procedures) Act 1986
and associated guidelines, or the European Communities Council Directive of 24 November
1986 (86/609/EEC). The authors also certify that formal approval to conduct the experiments
described has been obtained from the animal subjects review board of their institution and
could be provided upon request. The authors further attest that all efforts were made to
minimize the number of animals used and their suffering.
Declaration of interests: None
Acknowledgements:
This project was funded partially by the Holsworth Wildlife Research Endowment – Equity
Trustees Charitable Foundation & the Ecological Society of Australia (Awarded to LF) and
internal PhD research funding from the University of New England (to LF).
References:
Allen, N. J. and B. A. Barres (2009). "Glia — more than just brain glue." Nature 457(7230):
675-677.
Berg, M. P. and J. Ellers (2010). "Trait plasticity in species interactions: a driving force of
community dynamics." Evolutionary Ecology 24(3): 617-629.
Callaghan, C. T., R. E. Major, J. H. Wilshire, J. M. Martin, R. T. Kingsford and W. K.
Cornwell (2019). "Generalists are the most urban-tolerant of birds: a phylogenetically
controlled analysis of ecological and life history traits using a novel continuous measure of
bird responses to urbanization." Oikos 128(6): 845-858.
Journal Pre-proof
22
Deaner, R. O., Isler, K., Burkart, J., & Van Schaik, C. (2007). "Overall brain size, and not
encephalization quotient, best predicts cognitive ability across non-human primates." Brain,
behavior and evolution 70(2): 115-124.
Deitch, A. D., H. Law and R. deVere White (1982). "A stable propidium iodide staining
procedure for flow cytometry." Journal of Histochemistry & Cytochemistry 30(9): 967-972.
Dicke, U., & Roth, G. (2016). "Neuronal factors determining high intelligence."
Philosophical Transactions of the Royal Society B: Biological Sciences 371(1685):
20150180.
Herculano-Houzel, L. (2005). "Isotropic fractionator: a simple, rapid method for the
quantification of total cell and neuron numbers in the brain." Journal of Neuroscience 25(10):
2518-2521.
Herculano-Houzel, S. and R. Lent (2005). "Isotropic fractionator: a simple, rapid method for
the quantification of total cell and neuron numbers in the brain." The Journal of
neuroscience : the official journal of the Society for Neuroscience 25(10): 2518-2521.
Herculano-Houzel, S. and R. Lent (2005). "Isotropic fractionator: a simple, rapid method for
the quantification of total cell and neuron numbers in the brain." Journal of Neuroscience
25(10): 2518-2521.
Herculano‐Houzel, S. (2014). "The glia/neuron ratio: how it varies uniformly across brain
structures and species and what that means for brain physiology and evolution." Glia 62(9):
1377-1391.
Holekamp, K. E., E. M. Swanson and P. E. Van Meter (2013). "Developmental constraints on
behavioural flexibility. " Philos Trans R Soc Lond B Biol Sci 368(1618): 20120350.
McKinnon, K. M. (2018). "Flow Cytometry: An Overview." Current protocols in
immunology 120: 5.1.1-5.1.11.
Journal Pre-proof
23
Olkowicz, S., Kocourek, M., Lučan, R. K., Porteš, M., Fitch, W. T., Herculano-Houzel, S., &
Němec, P. (2016). "Birds have primate-like numbers of neurons in the forebrain."
Proceedings of the National Academy of Sciences 113(26): 7255-7260.
Sherwood, C., C. Stimpson, M. Raghanti and D. Wildman (2006). "Evolution of increased
glia-neuron ratios in the human frontal cortex." Proceedings of the National Academy of
Sciences of the United States of America 103(37): 13606.
Sih, A., M. C. Ferrari and D. J. Harris (2011). "Evolution and behavioural responses to
human-induced rapid environmental change." Evol Appl 4(2): 367-387.
Sol, D., Duncan, R. P., Blackburn, T. M., Cassey, P., & Lefebvre, L. (2005). "Big brains,
enhanced cognition, and response of birds to novel environments." Proceedings of the
National Academy of Sciences 102(15): 5460-5465.
Sterio, D. C. (1984). "The unbiased estimation of number and sizes of arbitrary particles
using the disector." J Microsc 134(Pt 2): 127-136.
Thuiller, W., C. Albert, M. B. Araujo, P. M. Berry, M. Cabeza, A. Guisan, T. Hickler, G. F.
Midgely, J. Paterson, F. M. Schurr, M. T. Sykes and N. E. Zimmermann (2008). "Predicting
global change impacts on plant species' distributions: Future challenges." Perspectives in
Plant Ecology Evolution and Systematics 9(3-4): 137-152.
Ullian, E., S. Sapperstein, K. Christopherson and B. Barres (2001). "Control of synapse
number by glia." Science 291(5504): 657-661.
West, M. J. (1999). "Stereological methods for estimating the total number of neurons and
synapses: issues of precision and bias." Trends Neurosci 22(2): 51-61.
Young, N. A., D. K. Flaherty, D. C. Airey, P. Varlan, F. Aworunse, J. H. Kaas and C. E.
Collins (2012). "Use of flow cytometry for high-throughput cell population estimates in brain
tissue." Front Neuroanat 6: 27.
Journal Pre-proof
