Viktor Szegedi
Abdennour Douida
Gábor Hutóczki3
László Novák
Karri Lamsa- 1Department of Physiology, Anatomy and Neuroscience, University of Szeged, Szeged, Hungary
- 2Hungarian Centre of Excellence for Molecular Medicine (HCEMM), Szeged, Szeged, Hungary
- 3Department of Neurosurgery, University of Debrecen Clinical Centre, Debrecen, Hungary
Homo sapiens has evolved a large and complex neocortex that underlies advanced cognitive capabilities. Neural computation, however, is inherently energy-intensive, and evolutionary pressures have shaped mechanisms that optimize both computational performance and energy efficiency in the human brain. Fast-spiking interneurons, particularly basket cells, are among the most active neuron types in the neocortex, where they play a key role in coordinating time and space in the activity of neuronal networks, but their high activity levels require high metabolic resources. Because the human neocortex is significantly larger than that of rodents—and contains a higher proportion of inhibitory interneurons relative to pyramidal cells—this expansion may have created evolutionary pressure to reduce the energetic cost of fast-spiking neurons. Compared with rodents, human fast-spiking neurons exhibit adaptations that appear to lower energy expenditure while preserving rapid and precise inhibition. One such adaptation is increased input resistance, which allows both excitation and inhibition to occur with reduced transmembrane ion currents, thereby decreasing the energy required to maintain ionic gradients across the plasma membrane. Since higher input resistance also slows down membrane potential changes, these cells show secondary adaptations that maintain rapid electrical signaling. Additional modifications—such as optimized ion channel composition in soma and axon initial segment, enhanced axon myelination, simplified structure of dendritic tree, and multivesicular synapses—further improve electrical signaling and are likely to reduce metabolic demand, collectively reducing ATP consumption in the neuronal network. By integrating cellular and synaptic perspectives, this review highlights how fast-spiking neurons in the human neocortex have evolved differently from those in rodents to balance energy efficiency while maintaining computational power, providing insight into the metabolic constraints of the human brain.
1 Introduction
The human brain, comprising approximately 90 billion neurons, possesses the most elaborated and complex cerebral cortex—the region primarily associated with cognition, reasoning, and consciousness—among all mammals (Pakkenberg and Gundersen, 1997). Despite its remarkable volumetric expansion relative to smaller mammals, human brain neurons adhere to the same fundamental cellular signaling and metabolic principles observed across species (Azevedo et al., 2009; Herculano-Houzel, 2009). Notably, humans have about 16 billion cortical neurons, the highest number recorded among mammals, a feature thought to underlie our exceptional cognitive and behavioral capacities. In contrast, even the largest-brained species, such as cetaceans and elephants, contain fewer cortical neurons despite their greater total brain mass (Azevedo et al., 2009; Herculano-Houzel et al., 2014a; Herculano-Houzel et al., 2014b). The remarkable increase in the number of neocortical neurons and the associated expansion of the neocortex that accompany it pose metabolic and energetic challenges for the human brain (Attwell and Laughlin, 2001; Herculano-Houzel, 2011). As brain metabolism scales with size and neuron number, energetic and metabolic limitations can critically influence brain evolution, particularly in large, neuron-rich cortices like the human neocortex (Karbowski, 2007).
Neural systems have evolved under persistent pressure to balance computational capacity with limited metabolic and spatial resources. Neurons are among the most energy-demanding cell types: although the brain constitutes only about 2% of body mass, it consumes nearly 20% of the body’s oxygen (Attwell and Laughlin, 2001; Herculano-Houzel, 2009). Across mammalian species, total brain energy consumption scales with neuron number (Laughlin et al., 1998; Herculano-Houzel et al., 2011; Hyder et al., 2011). In addition, larger neurons should, in theory, impose greater metabolic costs for maintaining membrane polarization—and because energy metabolism is tightly coupled to synaptic activity (Magistretti et al., 1999)—the relatively fixed energy budget per neuron suggests that large-brained apes, particularly humans, have evolved ‘primate-specific’ strategies to optimize their neuroeconomic efficiency.
Cortical organization is dominated by excitatory neurons, typically maintaining an excitatory-to-inhibitory cell ratio of around 4:1 (Douglas and Martin, 2007; Tremblay et al., 2016; Lim et al., 2018), although this ratio varies across regions and species. In the human neocortex, the relative abundance of inhibitory interneurons is higher than in rodents, and this is highest in higher-order associative areas such as the human prefrontal cortex, possibly reflecting enhanced demands for cognitive processing (Soltesz, 2006; Kooijmans et al., 2020; Hanson et al., 2025).
Among all cortical neuron types, parvalbumin-expressing fast-spiking (PVFS) interneurons are especially energy-intensive (Kann et al., 2014). Their ability to sustain high-frequency firing depends on rapid ion fluxes across the membrane and continuous ATP-dependent Na+/K+ pump activity. Basket cells—a major subclass of PVFS interneurons—is an electrically highly active neuron type important for structuring coordinated cortical network activities (Wang and Buzsaki, 1996; Traub et al., 2001; Hasenstaub et al., 2010; Melzer et al., 2012), but this precision comes at a high metabolic cost. In the expanded human neocortex, where inhibitory neuron number and density are both increased, these energetic constraints have likely exerted evolutionary pressure to preserve speed and reliability of electrical signaling while minimizing ATP consumption.
This review explores the cellular and synaptic adaptations that help human PVFS interneurons to sustain computational performance under tight energetic constraints. Cortical microcircuit operation depends on coordinated interactions among multiple neuronal and glial cell types, all of which have likely undergone evolutionary refinement. Here, however, we focus specifically on fast-spiking interneurons, particularly those in supragranular layer 2/3, which underwent pronounced expansion during human evolution (Balaram and Kaas, 2014).
2 Microstructural adaptations
A major challenge in scaling cortical networks of large-brained species is that larger neurons are intrinsically slower due to increased membrane capacitance and longer electrotonic path lengths (Gentet et al., 2000). Greater neurite length increases total membrane capacitance, thereby prolonging the time constant of membrane charging. It also extends the electrotonic distance over which signals must travel, leading to greater voltage attenuation and slower spread of voltage changes along neuronal processes (Das et al., 2025; Olah et al., 2025). These scaling-related factors, in theory, delay synaptic integration and action potential initiation. Yet, PVFS neurons show an evolutionarily conserved electrical phenotype with rapid and temporally precise action potential generation to synaptic input (Hu et al., 2014). This raises the question of how human PVFS neurons, which are larger and project over longer distances than those in smaller mammals, maintain the electrical rapidity within an expanded and energetically constrained cortical architecture (Benavides-Piccione et al., 2025; Kanari et al., 2025). To sustain rapid signaling despite these physical constraints, human interneurons appear to have evolved mechanisms—including specialized somatic, synaptic, and axonal properties—that maintain electrical signaling rapidity and temporal precision without substantially increasing ATP demand (Figure 1).
Figure 1. Cellular and metabolic adaptations of human parvalbumin fast-spiking (PVFS) interneurons supporting neuroeconomic efficiency. The figure illustrates some of the evolutionary specializations of human PVFS interneurons compared with their rodent counterparts. Human PVFS cells exhibit a denser perineuronal net (PNN), simplified dendritic architecture with fewer branch points, likely to have mitochondria with increased number and density with structural and metabolic adaptations, an altered molecular composition of the axon initial segment (AIS), and more extensive axonal myelination. Functionally, they show a more depolarized resting membrane potential (Vrest), higher input resistance (Ri), larger excitatory postsynaptic potentials (EPSPs), and a lower voltage threshold for action potential initiation. Together, these features enable temporally precise and energy-efficient signaling, allowing human PVFS interneurons to sustain high-frequency inhibition within the metabolically demanding human cortex.
2.1 Axonal myelination
Axonal myelination is a conserved feature of PVFS interneurons across mammals. Kole et al. (2022) showed that basket cells possess myelinated axons with mitochondria clustered within myelinated segments in rodent neocortex, supporting the energetic demands of rapid signaling. Myelination reduces effective axonal membrane capacitance, approximately 0.05 μF/cm2 for a 10-fold myelin wrap, thereby increasing conduction velocity and decreasing the energetic cost of action potential propagation (Castelfranco and Hartline, 2015).
PVFS interneuron axons are more frequently and more extensively myelinated in humans (Micheva et al., 2018; Stedehouder et al., 2019; Benamer et al., 2020). In the mouse neocortex, roughly 25–40% of PV interneurons have detectable myelination on parts of their axons. In the human neocortex, this proportion rises to 60–80%, depending on the cortical area and age. Both human and rodent PV interneurons display proximal axonal myelination, typically beginning near the soma, distal to the AIS. Human PVFSs exhibit greater overall myelination, with longer total myelinated lengths and more internodes (Stedehouder et al., 2017).
In addition, the myelin sheath tends to be thicker in humans relative to axon diameter. The g-ratio (axon diameter/fiber diameter) is smaller in humans (~0.6–0.7) than in rodents (~0.75–0.8) for PV interneurons, indicating thicker relative myelin. The g-ratio optimizes speed. Empirically, conduction velocity is maximal around g ≈ 0.6.
The primary benefit of thicker myelin is in active saltatory conduction of the action potential along the myelinated axon from one node to the next. A thicker myelin sheet reduces cell capacitance per surface area in the axon. In addition, a thicker myelin sheet increases transmembrane resistance, since it acts as a high-resistance electrical insulator by decreasing current leakage across the membrane. Each additional layer of myelin roughly doubles the effective membrane thickness, thereby reducing membrane capacitance and increasing membrane resistance exponentially. These changes increase the membrane length constant that determines how far along an axon a passive voltage change can spread before it decays significantly. Voltage changes spread farther along the axon before decaying, and there is less voltage attenuation between nodes of Ranvier, where action potentials are generated (Table 1).
Table 1. Summary of differences in axonal myelinization in PVFS interneurons in the human and rodent neocortex.
Increased myelination in human cells likely reflects an adaptation to larger neuronal size and longer conduction paths in the human cortex. By lowering axonal capacitance and leak conductance, myelination enhances propagation speed and reduces energy expenditure, contributing to the metabolic efficiency of fast and precise inhibition in large cortical networks (Hartline and Colman, 2007; Fields, 2015; Benamer et al., 2020). However, if myelin is too thick, the capacitance gain saturates, and metabolic cost increases (Table 2).
Table 2. The effects of enhanced myelinization on electrical signal propagation and energy efficiency in human PVFS interneuron axons.
2.2 Perineuronal nets
Perineuronal nets (PNNs) are specialized extracellular matrix structures that envelop the soma and proximal dendrites of PVFS interneurons in both rodents and humans. These structures mainly contain chondroitin sulfate proteoglycans (CSPGs), hyaluronan, tenascin-R, and link proteins (see Morphett et al., 2024 for review). Yet, the PNN is more than just an extracellular scaffold: it strongly influences metabolic efficiency, ion homeostasis, and electrical signaling (for a review of PNN role in metabolism see Zhang et al., 2024).
PNN can save metabolic energy indirectly by reducing ion leak and buffering extracellular ions. First, the dense extracellular matrix limits extracellular diffusion and movement of K+ and Na+ ions near the soma, helping maintain local ionic stability, which reduces the need for constant active pumping (via Na+/K+-ATPase) - a major ATP consumer - to restore gradients after each spike. By restricting ion and water diffusion near the membrane, PNNs can lower the effective membrane capacitance of fast-spiking neurons, thereby facilitating electrical rapidity by shortening the soma time constant for membrane potential changes. In addition, PNN can reduce energy requirements by restricting cell volume changes (Tewari et al., 2018). Second, dense PNN has strengthened the electrostatic stabilization effect where negatively charged glycosaminoglycan chains in the PNN create a fixed negative potential that may help stabilize extracellular charge distribution and buffer cations, indirectly reducing leak currents and ionic perturbations during fast firing. Third, by anchoring perisomatic inhibitory synapses PNN stabilizes synaptic connections (Bosiacki et al., 2019; Fawcett et al., 2019; Auer et al., 2025). Experimental degradation of PNNs in rodents reduces firing rates and electrical gain of PVFS cells, leading to instability and impaired rhythmic activity at the neuronal network level (Hanssen et al., 2023) (Table 3).
Table 3. The effects of thicker and denser PNNs on the electrical properties of the soma and cell metabolic costs in human PVFS interneurons.
Although the general morphology of PNNs is conserved across species, molecular composition and density appear to differ. Human and other primate PNNs show higher molecular density and more dense organization, whereas rodent PNNs are simpler and less dense (Mcgillis, 2018; Belliveau et al., 2024; da Silva et al., 2024). These molecular and structural differences may influence the biophysical properties of PNNs, and the more dense and molecularly rich PNN in humans may contribute to enhanced energy-saving function by enhanced ion buffering (Chaunsali et al., 2021) and protection against oxidate stress in human PVFS neurons (Morawski et al., 2004).
PNN also affects electrical signaling speed indirectly. Because the PNN is extracellular, it does not directly change the membrane’s capacitance or resistance. However, it can enhance the electrical signaling by stabilizing the extracellular ionic and electric field environment, which in turn affects how currents flow during synaptic and action potential activity. By reducing ion diffusion at the extracellular surface of the membrane, PNN helps maintain low extracellular K+ and high Na + during rapid firing. This supports consistent spike amplitude and timing, effectively increasing the temporal precision of firing.
2.3 K+ leak channels in the membrane
Human PVFS neurons exhibit higher input resistance than their rodent counterparts (Szegedi et al., 2016; Szegedi et al., 2023; Wilbers et al., 2023; Furdan et al., 2025), likely due to species-specific differences in leak channel density and transmembrane ionic conductances. Leak channels mediate passive ion fluxes that mainly determine the resting membrane potential (Gadsby, 2009) and generate an energetic cost, as ionic gradients must be continuously restored by Na+/K+-ATPases. In neurons with a larger membrane surface area, high leak conductance increases ATP demand. Thus, the higher input resistance of human PV interneurons likely reflects a reduction in passive ion leakage, lowering baseline ATP demand. The increased input resistance allows the resting membrane potential to be maintained and the generation of membrane potential changes, including synaptic excitation and inhibition and action potentials, with smaller ion currents, thereby reducing the energy cost of maintaining ionic gradients (Table 4).
Table 4. Effect of increased input resistance on electrical properties and energy efficiency in soma of human PVFS interneurons.
On the other hand, higher resistance also slows the time course of membrane potential kinetics, potentially compromising temporal precision (Szegedi et al., 2023). To compensate, human PVFS neurons exhibit additional adaptations. One is the somatic expression of hyperpolarization-activated cyclic nucleotide–gated (HCN) channels, which facilitate kinetics of membrane potential changes in soma to shorten the action potential generation delay (Szegedi et al., 2023). In contrast, rodent PVFS neurons express HCN channels primarily on their axons (Roth and Hu, 2020), showing that while HCN channel expression is evolutionarily conserved, its subcellular localization has shifted in humans. This somatic localization likely enhances the rapidity to transform excitatory synaptic inputs to an AP (Kalmbach et al., 2018; Szegedi et al., 2023). HCN channels are strongly modulated by intracellular factors such as pH or Ca2+ (Combe and Gasparini, 2021), which make the electrical properties of the human PVFS cell soma a substrate for plasticity regulations.
In addition to HCN channels, inwardly rectifying potassium (Kir) channels also help regulate PVFS membrane resistance and excitability (Furdan et al., 2025). Human PVFSs achieve comparable input resistance regulation with lower Kir conductance than mouse cells, owing to their higher baseline membrane resistance. This reduced Kir activity for equal excitability regulation represents another energy-conserving adaptation, minimizing unnecessary ion flow and ATP requirements through voltage-activated ion channels (Furdan et al., 2025).
2.4 Dendritic tree
The electrotonic transduction of membrane potential changes is less efficient in human neurons with longer dendritic elongations. Human PVFSs in the supragranular layer, which is a particularly proliferated neocortical layer in humans compared to other mammals, have structural specializations to compensate for this (Wilbers et al., 2023). A recent study shows that human PVFSs have a simplified structure of large dendrites compared to those in rats, with a smaller number of branching points. This structural specialization improves the electrotonic propagation of membrane potential changes over longer dendritic distances, since the simplified dendritic structure with fewer branching points results in less signal attenuation. Long dendrites suffer greater voltage attenuation and temporal dispersion of signals reaching the soma. In addition, from the point of view of the main dendrite, each branch behaves as a current sink where the axial current divides, causing attenuation and slowdown of the forward propagation of voltage along the main dendrite (Wilbers et al., 2023). Signals decay faster and arrive with smaller amplitude at the soma. Therefore, PVFS interneurons in the supragranular layer of human cortex may have evolved with a simplified dendritic tree, because less division of current enables less voltage attenuation along the main shaft. The shorter total electrotonic path allows faster and stronger signal transmission from distal to proximal sites. This is particularly advantageous for fast-spiking interneurons that rely on precise temporal signaling. However, the reduced dendritic tree comes at a cost of reduced synaptic integration capacity but fewer total synapses. However, the human neocortex compensates this functional trade-off between the electrical efficiency of PVFS neurons and the network computational capacity by an increased proportional number of interneurons compared to pyramidal cells (see, e.g., Kanari et al., 2025).
In addition, fewer branches reduce metabolic cost. There is less maintenance of ion gradients across the total membrane surface area (via Na+/K+ ATPase), less metabolic maintenance of cytoplasmic and membrane-bound processes in distal locations from the soma for structural (cytoskeleton) and transport support, and reduced biosynthetic costs for ion channels and receptors (Attwell and Laughlin, 2001; Karbowski, 2007) (Table 5).
Table 5. The effect of dendritic length and branching on electrotonic attenuation, propagation, and metabolic cost.
2.5 Axon initial segment
Compared with rodents, human parvalbumin-expressing fast-spiking interneurons (PVFSs) exhibit a lower action potential (AP) threshold at the axon initial segment (AIS) in response to somatic excitatory inputs. Together with the higher input resistance of human interneurons, this reduced AP threshold decreases the amount of depolarizing current—and therefore transmembrane ion flux—required to generate an output spike. Both features are neuroeconomically advantageous, as they minimize energetic cost per signaling event.
The reduced AP threshold in human PVFSs arises from a marked depletion of Kv1-type potassium channels in the AIS (Bakos et al., 2025). Human PVFS neurons lack Kv1.1 channels entirely, show no detectable expression of the corresponding gene KCNA1, and express Kv1.2 channels at substantially lower levels in the AIS compared with rodents (Bakos et al., 2025). In the absence of this Kv1-mediated outward current, smaller depolarizations are sufficient to trigger regenerative Na+ channel activation and initiate an AP.
Consequently, baseline K+ efflux at rest and near threshold is reduced during spike initiation. The accompanying reduction in both Na+ and K+ transmembrane currents lowers the workload of the Na+/K+-ATPase required to restore ionic gradients after spiking. In addition, the absence of Kv1 channels reduces the biosynthetic and maintenance costs associated with producing and trafficking these ion channels, further contributing to energetic efficiency in human PVFS interneurons.
Beyond energetic considerations, a lowered AP threshold also compensates for the slower membrane potential dynamics associated with the higher input resistance and larger somatic size of human interneurons. At the AIS, the critical transformation from excitatory postsynaptic potential (EPSP) to AP depends on how rapidly the local membrane potential reaches the Na+ channel activation threshold. In an AIS with low Kv1 channel density, depolarizing current originating in the soma spreads more efficiently into the AIS, while the reduced outward K+ conductance allows the local membrane potential to rise more steeply.
By diminishing subthreshold K+ “drag,” the AIS shortens the latency between EPSP onset and spike initiation. Thus, the lowered AP threshold in human PVFSs compensates for slower membrane charging, ensuring rapid and reliable action potential generation in response to excitatory synaptic input (Wilbers et al., 2023; Bakos et al., 2025) (Table 6).
Table 6. Reduced Kv1 channel density in the axon initial segment both lowers the energy cost of maintaining and generating action potentials and accelerates the input–output conversion process.
3 Metabolic adaptations
The majority of the cell’s energy currency, ATP, is consumed by the process of restoring ionic homeostasis, primarily through Na+/K + pumps (i.e., Na/K exchangers). ATP is produced in the mitochondria, specialized organelles whose collective composition is highly dynamic and sensitive to the intracellular milieu. Human PVFS interneurons exhibit a reduced resting membrane potential, which lowers metabolic costs. Additionally, human neurons have specialized mitochondria.
3.1 Depolarized resting membrane potential
Human PVFSs have lower (more depolarized) resting membrane potential (Em) than their rodent counterparts (Szegedi et al., 2023), which, together with the increased input resistance and the lowered AP threshold in the cells, synergistically works to reduce the amount of excitatory current and transmembrane ion current to generate AP by excitatory synaptic inputs. This is neuroeconomical because neuronal energy consumption primarily stems from restoring electrochemical gradients through ATP-dependent ion pumps following electrical activity (Harris et al., 2012; Muangkram et al., 2023). The Na+/K+ pump accounts for up to 50% of neuronal ATP use (Attwell and Laughlin, 2001; Meyer et al., 2022).
Pump activity increases with more negative resting membrane potential, as its generation and maintenance require greater electrical work to expel Na+ and intake K+ (Gadsby et al., 1985; Hansen et al., 2002). A more depolarized resting membrane potential, as reported in human FSPV neurons (Szegedi et al., 2023; Wilbers et al., 2023), is likely to reduce this energetic burden, thus being neuroeconomical by lowering ATP demand while maintaining high excitability in human PVFSs. The depolarized Em is likely to stem from fewer open K+ channels, causing reduced passive K+ efflux, in resting conditions (for a review of two-pore-domain K + channels, or leak channels, see Lesage and Lazdunski, 2000). This also means less biosynthetic burden (Table 7).
Table 7. A more depolarized resting membrane potential (Em) is metabolically and biosynthetically economical because it reduces passive ionic leaks, lowers the workload of the Na+/K+-ATPase, and allows smaller K+ channel densities to sustain excitability.
3.2 Mitochondrial specialization
Mitochondria provide the ATP required for neuronal signaling via oxidative phosphorylation (Mattson et al., 2008). There is some evidence indicating species differences in mitochondrial structure, metabolism, and distribution between human and rodent neurons, which will be discussed in more detail below.
Comparative analyses indicate that mitochondria in the human cerebral cortex differ molecularly and functionally from those in rodents, reflecting species-specific adaptations in bioenergetics (Westi et al., 2022). These differences suggest a distinct “mitotype” in humans that supports the energetic demands of large, high-performance cortical networks (Pontzer et al., 2016). Although differences in overall brain metabolism—such as substrate usage, coupling efficiency, and mitochondrial proteome—have been reported between species, no cell-type-specific studies have yet examined these features in parvalbumin-expressing interneurons. To date, there are no direct comparative data on the number, density, or subcellular distribution of mitochondria in human versus rodent PV interneurons (Table 8).
Table 8. Comparative evidence for mitochondrial specializations addressing mitochondrial structure, density, and localization in human and rodent PVFS interneurons.
In the human cortex, mitochondria display a high degree of regional and cellular specialization that presumably aligns with local computational and metabolic demands. The distribution of mitotypes follows a phylogenetic gradient, meaning that mitochondrial adaptations co-evolved with the expansion of higher-order association cortices in primate evolution (Mosharov et al., 2025). In isolated preparations, human cortical mitochondria exhibit lower oxygen consumption but equal or greater ATP output compared with mouse mitochondria, indicating more efficient oxidative phosphorylation coupling (Jeffreys and Craig, 1974). In parallel, transcriptomic and proteomic analyses reveal divergent gene expression and protein synthesis patterns between human and rodent mitochondria (for a review on cardiac cells, see Alibrandi and Lionetti, 2025), further emphasizing functional specialization at the molecular level.
Human and mouse neurons differ in mitochondrial development and metabolic activity, thereby affecting neuronal maturation. In humans, cortical neurons exhibit slower mitochondrial growth and lower oxidative metabolism, leading to a longer developmental timeline. In contrast, mouse neurons undergo faster mitochondrial maturation and generate energy more rapidly through oxidative phosphorylation (Iwata et al., 2023).
Mitotypes vary not only between species but also across neuronal classes (see Kuznetsov and Margreiter, 2009 for a review) (Fecher et al., 2019) indicating cell–type–specific energetic specializations. Even within a single neuron, the mitochondrial population is heterogeneous: axonal and dendritic mitochondria display divergent genetic and functional characteristics (Hirabayashi et al., 2024), as shown in pyramidal cells. It is therefore reasonable to assume that PVFS interneurons harbor mitochondria adapted to their exceptionally high energetic and temporal demands.
Parvalbumin basket cell myelination accumulates axonal mitochondria in internodes. A recent study in a mouse model shows that in PV-positive basket cell axons, myelinated segments show a higher mitochondrial density than unmyelinated segments within the same axon. Regarding that human PVFS cell axons are more extensively myelinated; they may also be richer in mitochondria. Given the larger size and longer axonal/dendritic distances of human cortical neurons, one might predict that human PV interneurons would have increased mitochondrial content or different localization (e.g., more mitochondria in distal processes) or enhanced coupling efficiency — but this remains to be empirically verified.
However, across many mammalian species, the density of mitochondria per unit area of neuropil in the cortex appears conserved, and mitochondria density does not scale with brain size, thus large- vs. small-brained mammals have similar mitochondria densities (Karl et al., 2024). This suggests an evolutionary constraint on mitochondrial number and density in the cortical neuropil, and also indicates that brain enlargement in mammals is accompanied by an increase in the number of mitochondria per neuron. To date, there is no direct comparison of mitochondrial number, density, or subcellular localization specifically in human PV interneurons compared to rodent PV interneurons.
4 Synaptic adaptations
Human cortical signaling appears optimized for energy efficiency through high-probability neurotransmitter release and sparse coding strategies (Hunt et al., 2023). In sparse coding, only a subset of neurons is active at any given moment, reducing overall metabolic load while maintaining high information capacity (Attwell and Laughlin, 2001; Willmore et al., 2011).
4.1 Excitatory synaptic input
Excitatory synapses from pyramidal neurons onto PV basket cells are more efficient in humans than in rodents, in terms of generating action potential discharge (Molnar et al., 2008; Komlosi et al., 2012; Szegedi et al., 2016; Szegedi et al., 2017). Human connections contain approximately four times more functional release sites—around 6.2 per active zone compared with 1.6 in rats (Molnar et al., 2016). Electron microscopy further shows that the synapses’ active zones have approximately four docked vesicles in humans, compared to one in rodents. This enables a single presynaptic spike to release more neurotransmitter and elicit a larger excitatory postsynaptic current, sufficient to recruit local interneurons synchronously (Molnar et al., 2008; Szegedi et al., 2016).
In rodents, excitatory inputs onto basket cells are more numerous (Veres et al., 2017). The higher frequency of miniature EPSPs reported in rats compared with monkeys corresponds to denser excitatory innervation (Povysheva et al., 2008). In contrast, primates achieve comparable or greater postsynaptic efficacy with fewer but stronger synapses—a distinct energy-efficient strategy.
4.2 Inhibitory synapses of PVFSs
Cellular reconstruction analysis reveals that human basket cells receive more GABAergic synapses on their somata and proximal dendrites than those of mice (Rebollo et al., 2025). The high number of inhibitory synapses is likely to reflect the increased number of interneurons and cell density in the human cortex compared to the rodent cortex, as evidenced by the similarity in the number of autaptic GABAergic synapses to human and mouse PV immunopositive basket cells (Szegedi et al., 2020). However, in smaller GABAA-receptor activation is needed in human PVFSs for an equal level of inhibition, measured as a relative drop of the cell input resistance by GABAAR-mediated synaptic conductance (Szegedi et al., 2020)-because of the higher starting resting input resistance of the human cells (Szegedi et al., 2023).
Human PVFSs may also produce faster and stronger inhibitory outputs compared with rodents (Wilbers et al., 2023). Individual human basket cells form more inhibitory boutons per postsynaptic pyramidal neuron than those of rodents (Karube et al., 2004; Kubota et al., 2016). The increased number of boutons per target improves transmission reliability, allowing efficient control of pyramidal neuron firing with fewer presynaptic spikes (Table 9).
Table 9. Comparative organization of excitatory and inhibitory synapses in human and rodent PVFS cells, with their electrophysiological and metabolic implications.
5 Summary
Human parvalbumin-expressing fast-spiking interneurons combine a repertoire of structural and metabolic adaptations that sustain rapid, precise inhibition in the large human cortex. Despite their larger size and higher input resistance, they maintain fast signaling through specialized AIS and HCN channel expression, strong axonal myelination, simplified dendritic trees, and stabilizing perineuronal nets. A more depolarized resting potential and efficient mitochondrial ATP production reduce energy costs. The human PVFS interneurons are well adapted to operate within a sparse cortical coding regime, where only a limited subset of neurons is active at any given time, and which is therefore energy efficient. Their strong excitatory postsynaptic potentials, elevated intrinsic excitability, and lowered input–output transformation threshold allow them to respond rapidly to minimal excitatory drive. At the same time, their powerful and temporally precise inhibitory output ensures accurate control over local network activity, preventing runaway excitation.
Author contributions
VS: Conceptualization, Supervision, Visualization, Writing – original draft, Writing – review & editing. AD: Writing – review & editing. GH: Resources, Writing – review & editing. LN: Resources, Writing – review & editing. KL: Conceptualization, Funding acquisition, Project administration, Writing – review & editing.
Funding
The author(s) declared that financial support was received for this work and/or its publication. KL declares support for the research of this work from the Ministry of Culture and Innovation of Hungary from the National Research (Project no. TKP-2021-EGA-05) provided by, Development and Innovation Fund financed under the TKP2021-EGA funding scheme, the Ministry of Culture and Innovation of Hungary from the National Research, Development and Innovation Fund, financed under the 2022-2.1.1-NL funding scheme (Project no. 2022-2.1.1-NL-2022-00005), EU’s Horizon 2020 research and innovation program under grant (No. 739593). Nemzeti Kutatási, Fejlesztési és Innovaciós Alap, Magyar Tudományos Akadémia - the National Brain Research Program Hungary OTKA K 134279 and ADVANCED 150382
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was used in the creation of this manuscript. The figure was prepared with the help of AI.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Alibrandi, L., and Lionetti, V. (2025). Interspecies differences in mitochondria: implications for cardiac and vascular translational research. Vasc. Pharmacol. 159:107476. doi: 10.1016/j.vph.2025.107476,
Attwell, D., and Laughlin, S. B. (2001). An energy budget for signaling in the grey matter of the brain. J. Cereb. Blood Flow Metab. 21, 1133–1145. doi: 10.1097/00004647-200110000-00001,
Auer, S., Schicht, M., Hoffmann, L., Budday, S., Frischknecht, R., Blumcke, I., et al. (2025). The role of perineuronal nets in physiology and disease: insights from recent studies. Cells 14:321. doi: 10.3390/cells14050321,
Azevedo, F. A., Carvalho, L. R., Grinberg, L. T., Farfel, J. M., Ferretti, R. E., Leite, R. E., et al. (2009). Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically scaled-up primate brain. J. Comp. Neurol. 513, 532–541. doi: 10.1002/cne.21974,
Bakos, E., Tiszlavicz, Á., Szegedi, V., Douida, A., Furdan, S., Welter, D. K., et al. (2025). Adaptations of the axon initial segment in fast-spiking interneurons of the human neocortex support low action potential thresholds. PLoS Biol. 23:e3003549. doi: 10.1371/journal.pbio.3003549
Balaram, P., and Kaas, J. H. (2014). Towards a unified scheme of cortical lamination for primary visual cortex across primates: insights from NeuN and VGLUT2 immunoreactivity. Front. Neuroanat. 8:81. doi: 10.3389/fnana.2014.00081,
Belliveau, C., Theberge, S., Netto, S., Rahimian, R., Fakhfouri, G., Hosdey, C., et al. (2024). Chondroitin sulfate glycan sulfation patterns influence histochemical labeling of perineuronal nets: a comparative study of interregional distribution in human and mouse brain. Glycobiology 34:cwae049. doi: 10.1093/glycob/cwae049,
Benamer, N., Vidal, M., Balia, M., and Angulo, M. C. (2020). Myelination of parvalbumin interneurons shapes the function of cortical sensory inhibitory circuits. Nat. Commun. 11:5151. doi: 10.1038/s41467-020-18984-7,
Benavides-Piccione, R., Fernaud-Espinosa, I., Kastanauskaite, A., and DeFelipe, J. (2025). Principles for dendritic spine size and density in human and mouse cortical pyramidal neurons. J. Comp. Neurol. 533:e70060. doi: 10.1002/cne.70060,
Bosiacki, M., Gassowska-Dobrowolska, M., Kojder, K., Fabianska, M., Jezewski, D., Gutowska, I., et al. 2019). Perineuronal nets and their role in synaptic homeostasis. Int. J. Mol. Sci. 20:4108. doi: 10.3390/ijms20174108, 31443560.
Castelfranco, A. M., and Hartline, D. K. (2015). The evolution of vertebrate and invertebrate myelin: a theoretical computational study. J. Comput. Neurosci. 38, 521–538. doi: 10.1007/s10827-015-0552-x
Chaunsali, L., Tewari, B. P., and Sontheimer, H. (2021). Perineuronal net dynamics in the pathophysiology of epilepsy. Epilepsy Curr 21, 273–281. doi: 10.1177/15357597211018688,
Combe, C. L., and Gasparini, S. (2021). I(h) from synapses to networks: HCN channel functions and modulation in neurons. Prog. Biophys. Mol. Biol. 166, 119–132. doi: 10.1016/j.pbiomolbio.2021.06.002,
da Silva, M. D. V., Bacarin, C. C., Machado, C. C. A., Franciosi, A., Mendes, J. D. L., da Silva Watanabe, P., et al. (2024). Descriptive study of perineuronal net in enteric nervous system of humans and mice. J. Neurochem. 168, 1956–1972. doi: 10.1111/jnc.16159,
Das, B., Baruah, S. M. B., Roy, S., and Bhattacharyya, D. K. (2025). An effective framework to study signal transmission due to non-homogeneous extracellular space in neuron. J. Biol. Phys. 51:24. doi: 10.1007/s10867-025-09689-3,
Douglas, R. J., and Martin, K. A. (2007). Recurrent neuronal circuits in the neocortex. Curr. Biol. 17, R496–R500. doi: 10.1016/j.cub.2007.04.024,
Faitg, J., Lacefield, C., Davey, T., White, K., Laws, R., Kosmidis, S., et al. (2021). 3D neuronal mitochondrial morphology in axons, dendrites, and somata of the aging mouse hippocampus. Cell Rep. 36:109509. doi: 10.1016/j.celrep.2021.109509,
Fawcett, J. W., Oohashi, T., and Pizzorusso, T. (2019). The roles of perineuronal nets and the perinodal extracellular matrix in neuronal function. Nat. Rev. Neurosci. 20, 451–465. doi: 10.1038/s41583-019-0196-3,
Fecher, C., Trovo, L., Muller, S. A., Snaidero, N., Wettmarshausen, J., Heink, S., et al. (2019). Cell-type-specific profiling of brain mitochondria reveals functional and molecular diversity. Nat. Neurosci. 22, 1731–1742. doi: 10.1038/s41593-019-0479-z,
Fields, R. D. (2015). A new mechanism of nervous system plasticity: activity-dependent myelination. Nat. Rev. Neurosci. 16, 756–767. doi: 10.1038/nrn4023,
Furdan, S., Douida, A., Bakos, E., Tiszlavicz, A., Molnar, G., Tamas, G., et al. (2025). Regulation of input excitability in human and mouse parvalbumin interneurons by kir potassium channels. bioRxiv. doi: 10.1101/2025.04.11.648314,
Gadsby, D. C. (2009). Ion channels versus ion pumps: the principal difference, in principle. Nat. Rev. Mol. Cell Biol. 10, 344–352. doi: 10.1038/nrm2668,
Gadsby, D. C., Kimura, J., and Noma, A. (1985). Voltage dependence of Na/K pump current in isolated heart cells. Nature 315, 63–65. doi: 10.1038/315063a0,
Gentet, L. J., Stuart, G. J., and Clements, J. D. (2000). Direct measurement of specific membrane capacitance in neurons. Biophys. J. 79, 314–320. doi: 10.1016/S0006-3495(00)76293-X,
Hansen, P. S., Buhagiar, K. A., Kong, B. Y., Clarke, R. J., Gray, D. F., and Rasmussen, H. H. (2002). Dependence of Na+-K+ pump current-voltage relationship on intracellular Na+, K+, and cs+ in rabbit cardiac myocytes. Am. J. Physiol. Cell Physiol. 283, C1511–C1521. doi: 10.1152/ajpcell.01343.2000
Hanson, K. L., Greiner, D. M. Z., Schumann, C. M., and Semendeferi, K. (2025). Inhibitory systems in brain evolution: pathways of vulnerability in neurodevelopmental disorders. Brain Behav. Evol. 100, 29–48. doi: 10.1159/000540865,
Hanssen, K. O., Grodem, S., Fyhn, M., Hafting, T., Einevoll, G. T., Ness, T. V., et al. (2023). Responses in fast-spiking interneuron firing rates to parameter variations associated with degradation of perineuronal nets. J. Comput. Neurosci. 51, 283–298. doi: 10.1007/s10827-023-00849-9,
Harris, J. J., Jolivet, R., and Attwell, D. (2012). Synaptic energy use and supply. Neuron 75, 762–777. doi: 10.1016/j.neuron.2012.08.019,
Hartline, D. K., and Colman, D. R. (2007). Rapid conduction and the evolution of giant axons and myelinated fibers. Curr. Biol. 17, R29–R35. doi: 10.1016/j.cub.2006.11.042,
Hasenstaub, A., Otte, S., Callaway, E., and Sejnowski, T. J. (2010). Metabolic cost as a unifying principle governing neuronal biophysics. Proc. Natl. Acad. Sci. USA 107, 12329–12334. doi: 10.1073/pnas.0914886107,
Herculano-Houzel, S. (2009). The human brain in numbers: a linearly scaled-up primate brain. Front. Hum. Neurosci. 3:31. doi: 10.3389/neuro.09.031.2009,
Herculano-Houzel, S. (2011). Scaling of brain metabolism with a fixed energy budget per neuron: implications for neuronal activity, plasticity and evolution. PLoS One 6:e17514. doi: 10.1371/journal.pone.0017514,
Herculano-Houzel, S., Avelino-de-Souza, K., Neves, K., Porfirio, J., Messeder, D., Mattos Feijo, L., et al. (2014a). The elephant brain in numbers. Front. Neuroanat. 8:46. doi: 10.3389/fnana.2014.00046,
Herculano-Houzel, S., Manger, P. R., and Kaas, J. H. (2014b). Brain scaling in mammalian evolution as a consequence of concerted and mosaic changes in numbers of neurons and average neuronal cell size. Front. Neuroanat. 8:77. doi: 10.3389/fnana.2014.00077,
Herculano-Houzel, S., Ribeiro, P., Campos, L., Valotta da Silva, A., Torres, L. B., Catania, K. C., et al. (2011). Updated neuronal scaling rules for the brains of Glires (rodents/lagomorphs). Brain Behav. Evol. 78, 302–314. doi: 10.1159/000330825,
Hirabayashi, Y., Lewis, T. L. Jr., Du, Y., Virga, D. M., Decker, A. M., Coceano, G., et al. (2024). Most axonal mitochondria in cortical pyramidal neurons lack mitochondrial DNA and consume ATP. bioRxiv. doi: 10.1101/2024.02.12.579972
Hu, H., Gan, J., and Jonas, P. (2014). Interneurons. Fast-spiking, parvalbumin(+) GABAergic interneurons: from cellular design to microcircuit function. Science 345:1255263. doi: 10.1126/science.1255263,
Hunt, S., Leibner, Y., Mertens, E. J., Barros-Zulaica, N., Kanari, L., Heistek, T. S., et al. (2023). Strong and reliable synaptic communication between pyramidal neurons in adult human cerebral cortex. Cereb. Cortex 33, 2857–2878. doi: 10.1093/cercor/bhac246,
Hyder, F., Herman, P., Sanganahalli, B. G., Coman, D., Blumenfeld, H., and Rothman, D. L. (2011). Role of ongoing, intrinsic activity of neuronal populations for quantitative neuroimaging of functional magnetic resonance imaging-based networks. Brain Connect. 1, 185–193. doi: 10.1089/brain.2011.0032,
Iwata, R., Casimir, P., Erkol, E., Boubakar, L., Planque, M., Gallego Lopez, I. M., et al. (2023). Mitochondria metabolism sets the species-specific tempo of neuronal development. Science 379:eabn4705. doi: 10.1126/science.abn4705,
Jeffreys, A., and Craig, I. (1974). Differences in the products of mitochondrial protein synthesis in vivo in human and mouse cells and their potential use as markers for the mitochondrial genome in human--mouse somatic cell hybrids. Biochem. J. 144, 161–164. doi: 10.1042/bj1440161,
Kalmbach, B. E., Buchin, A., Long, B., Close, J., Nandi, A., Miller, J. A., et al. (2018). H-channels contribute to divergent intrinsic membrane properties of Supragranular pyramidal neurons in human versus mouse cerebral cortex. Neuron 100, 1194–1208.e5. doi: 10.1016/j.neuron.2018.10.012,
Kanari, L., Shi, Y., Arnaudon, A., Barros-Zulaica, N., Benavides-Piccione, R., Coggan, J. S., et al. (2025). Of mice and men: dendritic architecture differentiates human from mouse neuronal networks. iScience 28:112928. doi: 10.1016/j.isci.2025.112928,
Kann, O., Papageorgiou, I. E., and Draguhn, A. (2014). Highly energized inhibitory interneurons are a central element for information processing in cortical networks. J. Cereb. Blood Flow Metab. 34, 1270–1282. doi: 10.1038/jcbfm.2014.104,
Karbowski, J. (2007). Global and regional brain metabolic scaling and its functional consequences. BMC Biol. 5:18. doi: 10.1186/1741-7007-5-18,
Karl, M. T., Kim, Y. D., Rajendran, K., Manger, P. R., and Sherwood, C. C. (2024). Invariance of mitochondria and synapses in the primary visual cortex of mammals provides insight into energetics and function. J. Comp. Neurol. 532:e25669. doi: 10.1002/cne.25669,
Karube, F., Kubota, Y., and Kawaguchi, Y. (2004). Axon branching and synaptic Bouton phenotypes in GABAergic nonpyramidal cell subtypes. J. Neurosci. 24, 2853–2865. doi: 10.1523/JNEUROSCI.4814-03.2004,
Kole, K., Voesenek, B. J. B., Brinia, M. E., Petersen, N., and Kole, M. H. P. (2022). Parvalbumin basket cell myelination accumulates axonal mitochondria to internodes. Nat. Commun. 13:7598. doi: 10.1038/s41467-022-35350-x,
Komlosi, G., Molnar, G., Rozsa, M., Olah, S., Barzo, P., and Tamas, G. (2012). Fluoxetine (prozac) and serotonin act on excitatory synaptic transmission to suppress single layer 2/3 pyramidal neuron-triggered cell assemblies in the human prefrontal cortex. J. Neurosci. 32, 16369–16378. doi: 10.1523/JNEUROSCI.2618-12.2012,
Kooijmans, R. N., Sierhuis, W., Self, M. W., and Roelfsema, P. R. (2020). A quantitative comparison of inhibitory interneuron size and distribution between mouse and macaque V1, using calcium-binding proteins. Cereb Cortex Commun 1:tgaa068. doi: 10.1093/texcom/tgaa068,
Kubota, Y., Karube, F., Nomura, M., and Kawaguchi, Y. (2016). The diversity of cortical inhibitory synapses. Front Neural Circuits 10:27. doi: 10.3389/fncir.2016.00027,
Kuznetsov, A. V., and Margreiter, R. (2009). Heterogeneity of mitochondria and mitochondrial function within cells as another level of mitochondrial complexity. Int. J. Mol. Sci. 10, 1911–1929. doi: 10.3390/ijms10041911,
Laughlin, S.B. van de Ruyter Steveninck, R.R., and Anderson, J.C. 1998 The metabolic cost of neural information Nat. Neurosci. 1 36–41 doi: 10.1038/236
Lesage, F., and Lazdunski, M. (2000). Molecular and functional properties of two-pore-domain potassium channels. Am. J. Physiol. Renal Physiol. 279, F793–F801. doi: 10.1152/ajprenal.2000.279.5.F793
Lim, L., Mi, D., Llorca, A., and Marin, O. (2018). Development and functional diversification of cortical interneurons. Neuron 100, 294–313. doi: 10.1016/j.neuron.2018.10.009,
Magistretti, P. J., Pellerin, L., Rothman, D. L., and Shulman, R. G. (1999). Energy on demand. Science 283, 496–497. doi: 10.1126/science.283.5401.496,
Mattson, M. P., Gleichmann, M., and Cheng, A. (2008). Mitochondria in neuroplasticity and neurological disorders. Neuron 60, 748–766. doi: 10.1016/j.neuron.2008.10.010,
Mcgillis, K. (2018). Developmental postnatal neuroplasticity of the primate amygdala: A quantitative analysis of parvalbumin neurons and perineuronal nets. Master’s Thesis, Harvard Extension School.
Melzer, S., Michael, M., Caputi, A., Eliava, M., Fuchs, E. C., Whittington, M. A., et al. (2012). Long-range-projecting GABAergic neurons modulate inhibition in hippocampus and entorhinal cortex. Science 335, 1506–1510. doi: 10.1126/science.1217139,
Meyer, D. J., Diaz-Garcia, C. M., Nathwani, N., Rahman, M., and Yellen, G. (2022). The Na(+)/K(+) pump dominates control of glycolysis in hippocampal dentate granule cells. eLife 11:e81645. doi: 10.7554/eLife.81645,
Micheva, K. D., Chang, E. F., Nana, A. L., Seeley, W. W., Ting, J. T., Cobbs, C., et al. (2018). Distinctive structural and molecular features of myelinated inhibitory axons in human neocortex. eNeuro 5:ENEURO.0297-18.2018. doi: 10.1523/ENEURO.0297-18.2018,
Molnar, G., Olah, S., Komlosi, G., Fule, M., Szabadics, J., Varga, C., et al. (2008). Complex events initiated by individual spikes in the human cerebral cortex. PLoS Biol. 6:e222. doi: 10.1371/journal.pbio.0060222,
Molnar, G., Rozsa, M., Baka, J., Holderith, N., Barzo, P., Nusser, Z., et al. (2016). Human pyramidal to interneuron synapses are mediated by multi-vesicular release and multiple docked vesicles. eLife 5:e18167. doi: 10.7554/eLife.18167,
Morawski, M., Bruckner, M. K., Riederer, P., Bruckner, G., and Arendt, T. (2004). Perineuronal nets potentially protect against oxidative stress. Exp. Neurol. 188, 309–315. doi: 10.1016/j.expneurol.2004.04.017
Morphett, J. C., Whittaker, A. L., Reichelt, A. C., and Hutchinson, M. R. (2024). Perineuronal net structure as a non-cellular mechanism contributing to affective state: a scoping review. Neurosci. Biobehav. Rev. 158:105568. doi: 10.1016/j.neubiorev.2024.105568
Mosharov, E. V., Rosenberg, A. M., Monzel, A. S., Osto, C. A., Stiles, L., Rosoklija, G. B., et al. (2025). A human brain map of mitochondrial respiratory capacity and diversity. Nature 641, 749–758. doi: 10.1038/s41586-025-08740-6,
Muangkram, Y., Himeno, Y., and Amano, A. (2023). Clarifying the composition of the ATP consumption factors required for maintaining ion homeostasis in mouse rod photoreceptors. Sci. Rep. 13:14161. doi: 10.1038/s41598-023-40663-y,
Olah, G., Lakovics, R., Shapira, S., Leibner, Y., Szucs, A., Csajbok, E. A., et al. (2025). Accelerated signal propagation speed in human neocortical dendrites. eLife 13:RP93781. doi: 10.7554/eLife.93781,
Pakkenberg, B., and Gundersen, H. J. (1997). Neocortical neuron number in humans: effect of sex and age. J. Comp. Neurol. 384, 312–320.
Pontzer, H., Brown, M. H., Raichlen, D. A., Dunsworth, H., Hare, B., Walker, K., et al. (2016). Metabolic acceleration and the evolution of human brain size and life history. Nature 533, 390–392. doi: 10.1038/nature17654,
Povysheva, N. V., Zaitsev, A. V., Rotaru, D. C., Gonzalez-Burgos, G., Lewis, D. A., and Krimer, L. S. (2008). Parvalbumin-positive basket interneurons in monkey and rat prefrontal cortex. J. Neurophysiol. 100, 2348–2360. doi: 10.1152/jn.90396.2008,
Rebollo, B., Abrahamyan, A., Thomale, U. W., Kaindl, A. M., Herman, M. A., and Rosenmund, C. (2025). GABAergic synaptic components are largely preserved across human and mouse neuronal models. Front. Cell. Neurosci. 19:1588894. doi: 10.3389/fncel.2025.1588894,
Roth, F. C., and Hu, H. (2020). An axon-specific expression of HCN channels catalyzes fast action potential signaling in GABAergic interneurons. Nat. Commun. 11:2248. doi: 10.1038/s41467-020-15791-y,
Santuy, A., Turegano-Lopez, M., Rodriguez, J. R., Alonso-Nanclares, L., DeFelipe, J., and Merchan-Perez, A. (2018). A quantitative study on the distribution of mitochondria in the neuropil of the juvenile rat somatosensory cortex. Cereb. Cortex 28, 3673–3684. doi: 10.1093/cercor/bhy159,
Soltesz, I. (2006). Diversity in the neuronal machine: Order and variability in interneuronal microcircuits. Oxford; New York: Oxford University Press.
Stedehouder, J., Brizee, D., Slotman, J. A., Pascual-Garcia, M., Leyrer, M. L., Bouwen, B. L., et al. (2019). Local axonal morphology guides the topography of interneuron myelination in mouse and human neocortex. eLife 8. doi: 10.7554/eLife.48615,
Stedehouder, J., Couey, J. J., Brizee, D., Hosseini, B., Slotman, J. A., Dirven, C. M. F., et al. (2017). Fast-spiking parvalbumin interneurons are frequently myelinated in the cerebral cortex of mice and humans. Cereb. Cortex 27, 5001–5013. doi: 10.1093/cercor/bhx203,
Szegedi, V., Bakos, E., Furdan, S., Kovacs, B. H., Varga, D., Erdelyi, M., et al. (2023). HCN channels at the cell soma ensure the rapid electrical reactivity of fast-spiking interneurons in human neocortex. PLoS Biol. 21:e3002001. doi: 10.1371/journal.pbio.3002001,
Szegedi, V., Molnar, G., Paizs, M., Csakvari, E., Barzo, P., Tamas, G., et al. (2017). High-precision fast-spiking basket cell discharges during complex events in the human neocortex. eNeuro 4:ENEURO.0260-17. doi: 10.1523/ENEURO.0260-17.2017,
Szegedi, V., Paizs, M., Baka, J., Barzo, P., Molnar, G., Tamas, G., et al. (2020). Robust perisomatic GABAergic self-innervation inhibits basket cells in the human and mouse supragranular neocortex. eLife 9:e51691. doi: 10.7554/eLife.51691,
Szegedi, V., Paizs, M., Csakvari, E., Molnar, G., Barzo, P., Tamas, G., et al. (2016). Plasticity in single axon glutamatergic connection to GABAergic interneurons regulates complex events in the human neocortex. PLoS Biol. 14:e2000237. doi: 10.1371/journal.pbio.2000237,
Tewari, B. P., Chaunsali, L., Campbell, S. L., Patel, D. C., Goode, A. E., and Sontheimer, H. (2018). Perineuronal nets decrease membrane capacitance of peritumoral fast spiking interneurons in a model of epilepsy. Nat. Commun. 9:4724. doi: 10.1038/s41467-018-07113-0,
Traub, R. D., Kopell, N., Bibbig, A., Buhl, E. H., LeBeau, F. E., and Whittington, M. A. (2001). Gap junctions between interneuron dendrites can enhance synchrony of gamma oscillations in distributed networks. J. Neurosci. 21, 9478–9486. doi: 10.1523/JNEUROSCI.21-23-09478.2001,
Tremblay, R., Lee, S., and Rudy, B. (2016). GABAergic interneurons in the neocortex: from cellular properties to circuits. Neuron 91, 260–292. doi: 10.1016/j.neuron.2016.06.033,
Veres, J. M., Nagy, G. A., and Hajos, N. (2017). Perisomatic GABAergic synapses of basket cells effectively control principal neuron activity in amygdala networks. eLife 6:e20721. doi: 10.7554/eLife.20721,
Wang, X. J., and Buzsaki, G. (1996). Gamma oscillation by synaptic inhibition in a hippocampal interneuronal network model. J. Neurosci. 16, 6402–6413. doi: 10.1523/JNEUROSCI.16-20-06402.1996,
Westi, E. W., Jakobsen, E., Voss, C. M., Bak, L. K., Pinborg, L. H., Aldana, B. I., et al. (2022). Divergent cellular energetics, glutamate metabolism, and mitochondrial function between human and mouse cerebral cortex. Mol. Neurobiol. 59, 7495–7512. doi: 10.1007/s12035-022-03053-5,
Wilbers, R., Galakhova, A. A., Driessens, S. L. W., Heistek, T. S., Metodieva, V. D., Hagemann, J., et al. (2023). Structural and functional specializations of human fast-spiking neurons support fast cortical signaling. Sci. Adv. 9:eadf0708. doi: 10.1126/sciadv.adf0708,
Willmore, B. D., Mazer, J. A., and Gallant, J. L. (2011). Sparse coding in striate and extrastriate visual cortex. J. Neurophysiol. 105, 2907–2919. doi: 10.1152/jn.00594.2010,
Keywords: ATP - adenosine triphosphate, basket cell, fast spiking cell, Homo sapiens , neocortex, neuroeconomic, parvalbumin (PV)
Citation: Szegedi V, Douida A, Hutóczki G, Novák L and Lamsa K (2026) Evolutionary neuroeconomic adaptations of fast-spiking neurons in the human neocortex. Front. Synaptic Neurosci. 18:1741452. doi: 10.3389/fnsyn.2026.1741452
Edited by:
Dmitri A. Rusakov, University College London, United KingdomReviewed by:
Stefano Masoli, University of Pavia, ItalyMaximiliano Jose Nigro, Norwegian University of Science and Technology, Norway
Copyright © 2026 Szegedi, Douida, Hutóczki, Novák and Lamsa. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Viktor Szegedi, dmlrdG9yLnN6ZWdlZGlAaGNlbW0uZXU=