Numbers of neurons as biological correlates of cognitive capability
Introduction
If the brain is the organ that directly organizes behavior, how does the enormous diversity in brain size, spanning a range of over one billion-fold across vertebrates and invertebrates, impact behavioral complexity and flexibility? The difficulty in answering this apparently simple question underscores how non-trivial it has been to quantify the two variables involved: behavior (should we measure repertoire size? Flexibility? Complexity? Speed of learning? Memory capacity? Self-control? And how should they be measured?) and brain diversity (is the relevant variable simply size? Should it be two-dimensional size, that is, surface area, or three-dimensional size, that is, volume? Number of neurons? Synapses? Glial cells? Fibers?).
Brain tissue is made of neurons, glial cells and vasculature, and neurons are the functional units that integrate synaptic activity and pass it on. Across species spanning orders of magnitude differences in brain size, and given that brain organization is remarkably conserved within clades, the ultimate causes for diversity in behavioral capability (however it is measured) should thus lie on the numbers of neurons in well-defined circuits in the brain: the more the neurons that compose a circuit, the more the possibilities that the circuit in principle admits, just like duplicating an entire genome opens new venues for complexity in life. However, until recently, the numbers of cells that compose different brains were not available for analysis. The best proxies were morphometric variables such as brain or structure mass, volume, or surface area—besides whole body mass.
One problem with using brain mass as a proxy for whatever underlying brain features do correlate with behavioral capabilities is that brain mass (and the mass of its components) is very obviously correlated with body mass across species (Figure 1a). The problem is that if a larger body requires a larger brain to operate it, then increases in brain mass may not necessarily contribute to behavioral complexity beyond simple body control.
The usual solution to this problem has been to mathematically factor body mass out of the equation by assuming that there is a predictable component of brain mass, required to operate the body depending on the size of the latter, that can be estimated from body mass, then discounting it from total brain mass, which should ‘neutralize’ effects of body size. This is done by calculating quotients between actual values and those predicted allometrically from body mass (the basis of the encephalization quotient proposed by Jerison [1]), or by the related expedient of calculating the residuals for correlations between the variable of interest and body mass, in the hope of looking specifically at “what was left” after accounting for body mass.
The whole rationale behind the encephalization quotient and later analyses of residuals against body mass is that Jerison not only explicitly expected larger bodies to require more neurons to operate them, but also, predicting that the number of neurons in the brain is related to brain mass in a universal manner across species, proposed that the amount of extra tissue measured by his encephalization quotient could be converted into a measure of ‘extra neurons’ contained in that tissue [1]. Jerison acknowledged that numbers of neurons and glial cells would have been “more meaningful biological parameters” than brain mass, but in their absence, he argued that brain size could be used instead.
We now have enough data from 75 species (47 mammals and 28 birds) that allow us to examine numbers of neurons directly and how they relate to body mass or cognitive capability. All data available on total numbers of neurons in the brain or its main structures have been generated so far by our own group and collaborators using the isotropic fractionator [2], a non-stereological method based on dissolving fixed, dissected brain tissue and counting free cell nuclei that gives results comparable to those obtained with stereology but in much less time [3, 4, 5]. Using this method, we have directly estimated numbers of neurons in the brains of rodents [6, 7], primates [8, 9, 10], eulipotyphlans [11], afrotherians [12, 13], artiodactyls [14], marsupials [15], as well as different orders of birds [16]. As reviewed recently [17••, 18], the data indicate that not only brain structure mass can be a misleading proxy for number of brain neurons (because of clade-specific relationships between numbers of neurons and neuronal density, which is inversely proportional to average neuronal cell size [19]), but there is also no universal relationship across mammalian or bird species between body mass and the number of brain neurons directly involved with operating the body.
Section snippets
Numbers of neurons in the brainstem should reflect body-related needs
However body mass influences or regulates numbers of neurons that directly or indirectly innervate and control body targets and sources of information, the number of neurons situated in structures most directly in contact with the body should offer a more direct proxy for the amount of neural processing involved with the body. Ideally, one would analyze numbers of neurons in the spinal cord; however, dissecting this organ in its integrity is a difficult task, and therefore many more brains than
Variable scaling of numbers of neurons in the cerebral cortex and rest of brain
Although the ‘rest of brain’ in our dataset includes the diencephalon and striatum, it provides the best opportunity so far for examining how numbers of neurons in the main structure implicated in cognition, the cerebral cortex, increase in relation to the number of neurons involved in operating the body. Although organized in nuclei and not layers, the bird pallium shares a similar functional organization as the mammalian cerebral cortex [21•], fulfilling equivalent complex cognitive
The need for comparable, quantitative behavioral data
The lack of systematic, comparative studies of cognitive capability across species has recently begun to be addressed. Comparative cognition is a difficult issue to tackle because of obvious hurdles to overcome, including making sure that tasks accommodate species-specific body form (paws, toes, claws, beak?), motivation, social anxiety, and non-trivial reasons why an animal does not perform (uncontrolled factors such as context, previous exposure, willingness to use different materials).
Conclusions
The history of the search for neural correlates of variation in cognitive capabilities across species has involved a long tradition of using brain and body mass and making assumptions about how they reflect underlying numbers of neurons. Now that numbers of neurons can be estimated rapidly and reliably, one side of the equation has been taken care of; what the field badly needs are more systematic quantitative studies of behavior and cognition that can be crossed to those data in search of an
Conflict of interest statement
I hereby declare having no conflicts of interest regarding the manuscript.
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:
• of special interest
•• of outstanding interest
Acknowledgements
Thanks to the many collaborators involved in generating the data analyzed here. This work was supported by the James McDonnell Foundation [grant 220020232].
References (25)
- et al.
Validation of the isotropic fractionator: comparison with unbiased stereology and DNA extraction for quantification of glial cells
J Neurosci Methods
(2014) Evolution of the Brain and Intelligence
(1973)- et al.
Isotropic fractionator: a simple, rapid method for the quantification of total cell and neuron numbers in the brain
J Neurosci
(2005) - et al.
Three counting methods agree on cell and neuron number in chimpanzee primary visual cortex
Front Neuroanat
(2014) - et al.
How to count cells: the advantages and disadvantages of the isotropic fractionator compared with stereology
Cell Tissue Res
(2015) - et al.
Cellular scaling rules for rodent brains
Proc Natl Acad Sci U S A
(2006) - et al.
Updated neuronal scaling rules for the brains of Glires (rodents/lagomorphs)
Brain Behav Evol
(2011) - et al.
Cellular scaling rules for primate brains
Proc Natl Acad Sci U S A
(2007) - et al.
Equal numbers of neuronal and non-neuronal cells make the human brain an isometrically scaled-up primate brain
J Comp Neurol
(2009) - et al.
Cellular scaling rules for the brain of an extended number of primate species
Brain Behav Evol
(2010)