Intrinsic motoneuron properties in typical human development

Abstract

Motoneuron properties and their firing patterns undergo significant changes throughout development and in response to neuromodulators such as serotonin. Here, we examined the age‐related development of self‐sustained firing and general excitability of tibialis anterior motoneurons in a young development (7–17 years), young adult (18–28 years) and adult (32–53 years) group, as well as in a separate group of participants taking selective serotonin reuptake inhibitors (SSRIs, aged 11–28 years). Self‐sustained firing, as measured by Δ F , was larger in the young development (∼5.8 Hz, n = 20) compared to the young adult (∼4.9 Hz, n = 13) and adult (∼4.8 Hz, n = 8) groups, consistent with a developmental decrease in self‐sustained firing mediated by persistent inward currents (PIC). Δ F was also larger in participants taking SSRIs (∼6.5 Hz, n = 9) compared to their age‐matched controls (∼5.3 Hz, n = 26), consistent with increased levels of spinal serotonin facilitating the motoneuron PIC. Participants in the young development and SSRI groups also had higher firing rates and a steeper acceleration in initial firing rates (secondary ranges), consistent with the PIC producing a steeper acceleration in membrane depolarization at the onset of motoneuron firing. In summary, both the young development and SSRI groups exhibited increased intrinsic motoneuron excitability compared to the adults, which, in the young development group, was also associated with a larger unsteadiness in the dorsiflexion torque profiles. We propose several intrinsic and extrinsic factors that affect both motoneuron PICs and cell discharge which vary during development, with a time course similar to the changes in motoneuron firing behaviour observed in the present study. image

Key points
Neurons in the spinal cord that activate muscles in the limbs (motoneurons) undergo increases in excitability shortly after birth to help animals stand and walk.
We examined whether the excitability of human ankle flexor motoneurons also continues to change from child to adulthood by recording the activity of the muscle fibres they innervate.
Motoneurons in children and adolescents aged 7–17 years (young development group) had higher signatures of excitability that included faster firing rates and more self‐sustained activity compared to adults aged ≥18 years.
Participants aged 11–28 years of age taking serotonin reuptake inhibitors had the highest measures of motoneuron excitability compared to their age‐matched controls.
The young development group also had more unstable contractions, which might partly be related to the high excitability of the motoneurons.

Full text

J Physiol 602.9 (2024) pp 2061–2087 2061
The Journal of Physiology
Intrinsic motoneuron properties in typical human
development
Ghazaleh Mohammadalinejad1,2,3 ,BabakAfsharipour
1,2,3 ,AlexYacyshyn
1,2 ,
Jennifer Duchcherer1,2,JackBashuk
1,2,ErinBennett
1,2 ,GregoryE.P.Pearcey
4, Francesco Negro5,
Katharina A. Quinlan6, David J. Bennett2,7 and Monica A. Gorassini1,2,3
1Department of Medicine, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB, Canada
2Neuroscience and Mental Health Institute, University of Alberta, Edmonton, AB, Canada
3Women and Children’s Health Research Institute, University of Alberta, Edmonton, AB, Canada
4School of Human Kinetics and Recreation, Memorial University of Newfoundland, St John’s Canada and Physical Therapy & Human Movement
Sciences, Northwestern University, Chicago, IL, USA
5Clinical and Experimental Sciences, Universita degli Studi di Brescia, Brescia, Italia
6George and Anne Ryan Institute forNeuroscience, Biomedical and PharmaceuticalS ciences, College of Pharmacy, University of Rhode Island, Kingston,
RI, USA
7Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, AB, Canada
Handling Editors: Richard Carson & Madeleine Lowery
The peer review history is available in the Supporting Information section of this article
(https://doi.org/10.1113/JP285756#support-information-section).
Abstract Motoneuron properties and their ring patterns undergo signicant changes throughout
development and in response to neuromodulators such as serotonin. Here, we examined the
age-relateddevelopmentofself-sustainedringandgeneralexcitabilityoftibialisanterior
motoneurons in a young development (7–17 years), young adult (18–28 years) and adult
(32–53 years) group, as well as in a separate group of participants taking selective serotonin reuptake
Ghazaleh Mohammdalinejad earned her Master’s degree in Neuroscience from the University of Alberta in 2023, having pre-
viously completed a Bachelor’s degree in Electrical Engineering from Sharif University of Technology in 2020. Throughout her
Master’s project, Ghazaleh explored the ring behaviour of human motoneurons via single motor unit recordings, specically
investigating how intrinsic motoneuron properties change during development from child to adulthood. Such data will be used
to examine dysfunction in motoneuron behaviour following early brain injury. In her current role as a Research Analyst at the
Centre for Addiction and Mental Health, she analyses functional magnetic resonance imaging data, providing valuable insights
to advance the eld of mental health and addiction research.
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society. DOI: 10.1113/JP285756
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits
use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial
purposes.

2062 G. Mohammadalinejad and others J Physiol 602.9
inhibitors(SSRIs,aged11–28years).Self-sustainedring,asmeasuredbyF, was larger in the
young development (∼5.8 Hz, n=20) compared to the young adult (∼4.9 Hz, n=13) and adult
(∼4.8 Hz, n=8) groups, consistent with a developmental decrease in self-sustained ring mediated
by persistent inward currents (PIC). Fwas also larger in participants taking SSRIs (∼6.5 Hz, n=9)
compared to their age-matched controls (∼5.3 Hz, n=26), consistent with increased levels of
spinal serotonin facilitating the motoneuron PIC. Participants in the young development and SSRI
groups also had higher ring rates and a steeper acceleration in initial ring rates (secondary ranges),
consistent with the PIC producing a steeper acceleration in membrane depolarization at the onset of
motoneuron ring. In summary, both the young development and SSRI groups exhibited increased
intrinsic motoneuron excitability compared to the adults, which, in the young development group,
was also associated with a larger unsteadiness in the dorsiexion torque proles. We propose several
intrinsic and extrinsic factors that aect both motoneuron PICs and cell discharge which vary during
development, with a time course similar to the changes in motoneuron ring behaviour observed in
the present study.
(Received 27 October 2023; accepted after revision 6 March 2024; rst published online 30 March 2024)
Corresponding author M. A. Gorassini: 5 005-A Katz Group Building, University of Alberta, Edmonton, AB, T6G 2E1,
Canada. Email: monica.gorassini@ualberta.ca
Abstract gure legend The ring behaviour of multiple, single motor units from the tibialis anterior muscle was
decomposed from high-density surface EMG to examine changes in motoneuron excitability across dierent phases
of development from child to adulthood in 50 participants aged 7–53 years. Compared to adults (≥18 years: young
adult and adult groups), motor unit discharge of children and adolescents aged 7–17 years (young development group)
accelerated faster at the onset of ring , reached higher pe ak rates and displayed longer self-sustained ring as represented
by the amplitude of the pink boxes. Participants aged 11–28 years (red bar), taking serotonin reuptake inhibitors (SSRI)
that are known to increase motoneuron excitability by facilitating persistent inward currents, had the highest measures
of motoneuron excitability. High motoneuron excitability may contribute to some of the force unsteadiness during the
execution of skilled ankle movements observed at the younger ages of development.
Key points
rNeurons in the spinal cord that activate muscles in the limbs (motoneurons) undergo increases
in excitability shortly after birth to help animals stand and walk.
rWe examined whether the excitability of human ankle exor motoneurons also continues to
change from child to adulthood by recording the activity of the muscle bres they innervate.
rMotoneurons in children and adolescents aged 7–17 years (young development group) had higher
signatures of excitability that included faster ring rates and more self-sustained activity compared
to adults aged ≥18 years.
rParticipantsaged11–28yearsofagetakingserotoninreuptakeinhibitorshadthehighestmeasures
of motoneuron excitability compared to their age-matched controls.
rThe young development group also had more unstable contractions, which might partly be related
to the high excitability of the motoneurons.
Introduction
Spinal motoneurons in rodents undergo marked changes
in intrinsic electrical properties during the rst 3 weeks
of postnatal development. These changes help to facilitate
motoneuron recruitment and repetitive ring that are
needed for sustained motor behaviours such as posture
and weight bearing locomotion (Carrascal et al., 2005;
Jean-Xavier et al., 2018). One such marked change is
the decrease in threshold and increase in amplitude of
persistent inward currents (PICs) in the rst 3 weeks
after birth (Harris-Warrick et al., 2023; Quinlan et al.,
2011; Revill et al., 2019; Sharples & Miles, 2021). PICs are
mediated by voltage-activated sodium (NaV)andcalcium
(CaV) channels (Li et al., 2004) and by a calcium-activated
sodium conductance (ICaN) (Bos et al., 2021). Because
PICs are activated near the recruitment threshold of the
motoneuron, they provide an additional depolarization
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2063
to amplify synaptic inputs, help secure the recruitment
of the motoneuron and accelerate their initial ring rates
(Bennett et al., 1998; Bennett, Li Siu et al., 2001; Gorassini
et al., 1998; Hounsgaard et al., 1988; Kiehn & Eken, 1997;
Lee & Heckman, 2000; Lee et al., 2003; Li et al., 2004).
Thus, the initial period of ring rate acceleration during a
slowly increasing input, termed secondary range (Bennett
et al., 1998; Granit et al., 1966; Kernell, 1965; Li et al.,
2004), is partly a reection of the onset activation of the
intrinsic PIC in motoneurons.
Following the secondary range, the PIC further shapes
the ring behaviour of motoneurons. After the PIC is
fully activated and produces a sustained plateau potential,
theringrateofthemotoneuronincreaseslinearlyin
response to an increasing intracellular current injection,
but with a much shallower slope (Bennett et al., 1998;
Lee et al., 2003). This lower-gain response, termed the
tertiary or saturation range, results partly from an increase
in membrane conductance produced by the sustained
opening of the NaVand CaVchannels mediating the
PIC (Binder et al., 2020). In adult decerebrate or intact
animals, the tertiary range also exhibits a pronounced
hysteresis where the sustained plateau potential mediated
by the PIC allows the motoneuron to re at inputs well
below that needed to initially recruit the motoneuron
(Bennett et al., 1998; Gorassini et al., 1999; Hounsgaard
et al., 1988). This hysteresis in ring reects how much
the PIC contributes to the self-sustained ring of the
motoneuron and can be measured by the reduction in
current injected into the soma at the oset of motoneuron
ring (de-recruitment) compared to the higher current
required to initiate ring at recruitment (I) (Bennett, Li,
Harvey et al., 2001; Gorassini et al., 2004; Li et al., 2004;
Powers et al., 2008). In wild-type mice, Iremains stable
from adolescence (4–9 weeks) to adulthood (13–17 weeks)
(Huh et al., 2021); however, these measures were made
under pentobarbital anaesthesia where PICs can be
suppressed.
Duringhumandevelopment,itisnotknownwhether
motoneuron PICs also increase in amplitude and decrease
in recruitment threshold in the rst few weeks and
months after birth or whether there are continual changes
from preadolescence to adulthood. To address the latter
question,weusedthesimultaneousringbehaviourof
multiple motor units from the tibialis anterior (TA)
muscle to indirectly examine the potential contribution of
motoneuron PICs to self-sustained ring in participants
between the ages of 7–17 years in the preadolescent
and adolescent stages (Brix et al., 2019; Wood et al.,
2019) and during young adulthood (18–28 years of age).
These two groups were also compared to an adult group
(32–53 years) before appreciable motor unit loss (McNeil
et al., 2005) or decreases in self-sustained ring of motor
units (Hassan et al., 2021; Orssatto et al., 2021). The
ringrateproleoflowthreshold(control)TAmotor
units activated during a triangular isometric dorsiexion
were used to estimate the synaptic input to higher
threshold (test) motor units (Afsharipour et al., 2020;
Gorassini et al., 2002a). Similar to the hysteretic current
measurements described above (I), we calculated the
dierence in the estimated synaptic input needed to
terminate ring of the test unit (control unit ring rate,
FT), compared with the synaptic input needed to recruit
ring (control unit ring rate, FR). Thus, the resulting
variable, F=FR–FT, was employed as a measure of
the self-sustained (hysteretic) ring of the motor unit,
potentially mediated by a PIC.
Interestingly, the ring rate proles of human motor
unitsactivatedduringaslowincreaseinvoluntaryeort,
andpresumablysynapticdrive,alsodisplayaninitial
acceleration followed by a shallower increase in ring
rate (Afsharipour et al., 2020; Beauchamp et al., 2023;
Binder et al., 2020), similar to the secondary and tertiary
ranges observed in animal motoneurons in response
to a slowly increasing somatic current injection. Thus,
we t two straight lines to the ascending portion of
theTAmotorunitringproletodelineatetheinitial
ringrateaccelerationduringthepresumedactivation
of the PIC (secondary range) and the more gradual
increase in ring rate following full PIC activation during
the tertiary range. We hypothesized that a steep and
brief secondary range is produced by a subthreshold
activation of a PIC that is accelerating rapidly at the
time of motoneuron recruitment compared to a shallower
and longer secondary range mediated by a more supra-
threshold PIC that is activating more gradually during
motoneuron recruitment (Afsharipour et al., 2020). Thus,
we measured both the slope and duration of the secondary
range to estimate whether there are any dierences in the
activation threshold of the PIC during development. The
slope and duration of the tertiary range was also compared
between age groups to estimate whether there were any
changes in the ring gain of the motoneuron.
The general excitability of TA motoneurons was also
examined across development by measuring the discharge
rates of the motor units at dierent points along the
ringrateprole.Forexample,weexaminedwhether
higher start rates were associated with a steeper secondary
range slope where a fast accelerating, subthreshold PIC
should produce a larger jump in initial ring rates
compared to a more gradually activating PIC activated
near recruitment. Maximum ring rates were quantied to
determine whether they were higher in potentially smaller
motoneurons of children as a result of greater input
resistance (Jean-Xavier et al., 2018) and shorter after-
hyperpolarization potentials (AHP) (Piotrkiewicz et al.,
2007). We also examined whether ring rates decreased
with increasing recruitment threshold of the motor units,
a phenomenon known as an onion skin eect (De Luca
& Erim, 1994), given the conicting ndings in the
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2064 G. Mohammadalinejad and others J Physiol 602.9
literature, with some studies showing an onion skin eect
in the TA (De Luca & Hostage, 2010) and others not (Erim
et al., 1996; Jesunathadas et al., 2012). Lastly, we examined
the end ring rates at de-recruitment of the motor units
given that very low ring rates during periods of low
synaptic drive can be produced by strong, regenerative
activation of the NaPIC (Gorassini et al., 2004; Li et al.,
2004).
The increased amplitude and persistence of
motoneuron PICs during early development in animals
may be mediated by a maturation of descending neuro-
modulatory inputs, such as serotonin or noradrenaline,
to the ventral spinal cord (Bregman, 1987; Hounsgaard
et al., 1988; Perrier et al., 2013; Xia et al., 2017). Both these
neuromodulators work through Gq-protein coupled
receptors, such as 5-HT2B/C and α1 (Murray et al., 2010;
Murray et al., 2011), which activate protein kinase C
to subsequently phosphorylate and facilitate the NaV
and CaVchannels that contribute to the PIC (Mizuno &
Itoh, 2009; D’Amico et al., 2014; Perrier et al., 2013). In
adult humans, a single oral intake (20 mg) of selective
serotonin reuptake inhibitors (SSRIs), such as citalopram
or escitalopram, probably increases serotonin in the
spinal cord and is associated with increases in indirect
measures of PIC-mediated, self-sustained ring in spinal
motoneurons (D’Amico et al., 2013; D’Amico et al., 2014;
Murray et al., 2010; Thompson & Hornby, 2013; Wei
et al., 2014). Because some participants in the young
development and young adult groups were taking SSRIs
daily, such as escitalopram, sertraline or uoxetine (see
doses in Methods) to treat various aective and anxiety
disorders (Dwyer & Bloch, 2019), we treated these
participants as a separate group given the potential eect
increased spinal levels of serotonin may have on the
threshold and amplitude of the PIC (Harvey et al., 2006a;
Li et al., 2007).
Across the age span of 7–53 years, we hypothesized that
younger participants (<18 years) would have higher ring
rates and greater amounts of self-sustained ring (F),
potentially from smaller motoneurons and larger PICs,
respectively, compared to older participants (>18 years).
ThelargerPICsmaypartlybearesultoflowerlevelsof
spinal inhibition in the younger participants (Geertsen
et al., 2017; Willerslev-Olsen et al., 2014), which reduces
the amplitude and persistence of the PIC (Bennett et al.,
1998; Hyngstrom et al., 2007). Similarly, we hypothesized
that participants taking oral SSRIs would have higher
initial (start) rates, steeper secondary range slopes and
larger amounts of self-sustained ring given that serotonin
both reduces the threshold, and increases the amplitude,
of the PIC (Harvey et al., 2006a; Li et al., 2007). Lastly, we
estimatedtheskillinproducingthetriangulardorsiexion
contractions that were used to measure Fby quantifying
theamountofforcesteadiness(coecientofvariation
of the detrended torque trace) (Skinner et al., 2019) and
examined whether it changed during development along
withchangesinmotoneuronexcitability.
Methods
Ethical approval and participants
Experiments were approved by the Health Research
Ethics Board of the University of Alberta (Protocol
00 076790) and conformed to the Declaration of Helsinki.
All 50 participants provided their written informed
consent prior to the experiment. The young development
group had a mean ±SD age of 12.18 ±2.69 years
(range 7–17 years) with 12 males and eight females
(n=20), the young adult group had mean ±SD age of
22.61 ±3.60 years (range 18–28 years) with six males
and seven females (n=13)andtheadultgrouphada
mean ±SD age of 42.38 ±7.92 years (range 32–53 years)
with three males and ve females (n=8). A separate
group of participants were on selective serotonin reuptake
inhibitors (SSRI group) and were between the ages of 11
and 28 years (17.02 ±5.02 years), with four females and
ve males (n=9). Four of the SSRI participants were
taking a daily 10 mg oral dose of escitalopram, three were
taking a daily 25 mg (n=1) or 100 mg (n=2) oral
dose of sertraline and two were taking a daily 10 mg
oral dose of uoxetine for at least 2 months before the
experiment. This SSRI group was compared to their peers
(i.e. SSRI control group), also aged between 11 to 28 years
(18.17 ±5.32 years; P=0.584, Mann–Whitney rank sum
test) who were not taking SSRIs and comprised 12 females
and 14 males (n=26) from the young development
and young adult groups. All participants were excluded
for any history of central or peripheral nerve injury or
disease. Puberty, menstrual or menopausal status were
not recorded. In six of the adult participants, parts of the
data were taken from a previous study (Afsharipour et al.,
2020).
EMG recordings
To isolate single motor units, exible high-density surface
EMG (HDsEMG) electrodes with 64 recording sites
(GR08MM1305; OT Bioelecttronica Inc., Turin, Italy)
were placed over the TA muscle. The recordings sites,
arranged in a 5 (wide) by 13 (long) grid with 8 mm
spacing, were placed lengthwise ∼1cmlateraltothetibia
and 4–5 cm below the lower edge of the patella over
the belly of the TA muscle. Recordings were made in a
monopolar conguration with the reference and ground
electrode straps (WS2; OT Bioelecttronica Inc.) wrapped
around the lower leg just above the ankle, with the ground
strap most distal. To reduce impedance, the skin was
rst rubbed with an abrasive paste (Nuprep; Weaver and
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2065
Company, Denver, CO, USA) and any excess was removed
withasalineoralcohol-soakedgauze.HDsEMGsignals
were amplied (150 times) and ltered with a 10 Hz high
pass and 900 Hz low pass lter using a Quattrocento
amplier(16bit;OTBioelecttronica,Inc.).Apairof
Ag-AgCl surface EMG electrodes (3.2 cm by 2.2 cm;
Kendall; Chicopee, MA, USA) were placed in a bipolar
conguration over the soleus to measure antagonist
muscle activity during the isometric contractions. All
signals were digitized and sampled at a frequency of
∼2048 Hz with the Quattrocento A/D.
Experiment protocol
Participants sat in a chair with their knee extended
to ∼120° and their dominant foot tightly sandwiched
between two plates of an adjustable 3-D printed
foot holder, with the heel on the oor and ankle
at ∼90°. Dorsiexion torque was measured by an
S-shaped strain gauge (150-lb SSM; Interface Force
Measurement Solutions, Scottsdale, AZ, USA) attached
to the bottom of the foot holder directly underneath the
metatarsophalangeal joint. The maximum dorsiexion
torque was measured from the largest of two maximum
voluntary contractions (MVCs) separated by at least 30 s.
Participants were then tasked to perform an isometric
triangular dorsiexion at either 10%, 20% or 30% of
their MVC torque, and match their torque output to a
target triangle presented over a computer screen with the
exerted torque being displayed from left to right. Each
triangular contraction consisted of a 10 s ascending phase
anda10sdescendingphase,with20sbetweeneach
contraction to avoid frequency-dependent facilitation of
motor units (Gorassini et al., 2002b). At least eight to 10
contractionswereperformedateachlevelofMVCtorque
in two trial runs, and the four best trials were selected post
hoc for analysis. Participants started with a 20% or 30%
MVC contraction because these were easier to perform.
Data analysis
Decomposition of single motor units. The HDsEMG
signalswererstconvertedintoaMATLABleformat
(MATLAB, version R2021b/R2022a; Mathworks, Inc.,
Natick, MA, USA) with custom-built functions, and then
digitally ltered (fourth-order, zero lag Butterworth lter
with a bandpass of 10–500 Hz and an additional 60 Hz
notch). HDsEMG signals with a poor signal-to-noise ratio
were removed (typically two or three of the 64 recorded
signals). Blind source separation (Negro et al., 2016)
was used to decompose the HDsEMG signals into the
contributing single motor units. Only motor unit action
potentials with a mean silhouette (SIL) value of 0.85 or
higher were used (Negro et al., 2016). The trains of motor
unit action potentials (or pulses) were manually edited to
remove or add extra pulses to correct too high or too low
ringrates,respectively.Only10%orlessofthepulses
were modied in a single ring rate prole. A fth-order
polynomial curve was t to the ring rate prole of each
motor unit for subsequent analysis as described below.
Average number of motor units per contraction. In each
participant, the number of decomposed motor units in
each contraction that met the criteria for selection (see
above) was averaged across the four contractions and
plotted against the age of the participant. Two straight
lines of dierent slope were then iteratively t to the
data (see section ‘Piecewise linear t’ below) to determine
which age range had the largest change in the number of
decomposed motor units with age.
Distribution of motor unit recruitment thresholds. The
recruitment threshold of the motor units was measured
as the torque level when the motor unit began ring,
expressed as a percentage of the torque at MVC (%
MVC). To examine the distribution of motor units with
dierent recruitment thresholds for each group, the
total number of motor units (count) within a range of
recruitment thresholds (e.g. bins of 0–5% MVC, 5–10%
MVC, etc.) for all four contractions in each participant
was pooled within each group. Each bin count was
then divided by the number of participants in each
group and plotted as a histogram. Note that the dierent
number of decomposed motor units between the various
groups, or their distribution according to recruitment
threshold, may not represent physiological dierences
in motor unit numbers or types. Rather, dierences in
muscle geometry (e.g. size) and/or subcutaneous tissue
between the dierent groups may aect the isolation
of single motor units from HDsEMG and blind source
decomposition (Oliveira et al., 2022).
F. The contribution of the PIC to self-sustained ring
of motoneurons was estimated as per Afsharipour et al.
(2020). First, to estimate the synaptic input to the
TA motoneuron pool, the ring proles of the lowest
thresholdmotorunits(typicallytwotoveunitswith
recruitment thresholds <3% MVC), were superimposed
to produce a composite prole as an average measure of
the synaptic drive to the TA motoneuron pool. Abrupt
accelerations in ring rate at the onset of the composite
ringrateprole(i.e.secondaryrange)weremanually
removedsothatonlythetertiaryrangeremained,which
we propose provides a more linear estimate of the synaptic
drive to the TA motoneurons. A fth-order polynomial
line was then tted to the composite prole. In the
remaining higher-threshold (test) motor units that were
activated for at least 1 s before peak torque, the ring rate
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2066 G. Mohammadalinejad and others J Physiol 602.9
of the composite prole (measured from the polynomial
line) when the test unit was de-recruited or terminated
(FT)wassubtractedfromtheringratewhenthetest
unit was initially recruited (FR)toproduceaFvalue
(F=FR–FT). Thus, Fis an estimate of the reduction
in synaptic input present at recruitment that is needed
to terminate motor unit ring to counteract the added
depolarization provided by the motoneuron PIC. The
continued ring at synaptic levels below recruitment is
termed ‘self-sustained ring’. An average Facross all
test motor units was obtained from the four contra-
ctions in each participant and plotted across age in the
young development and young adult groups. Fwas
also averaged across participants within each group and
compared across the dierent groups.
Self-sustained ring duration (SSD). The SSD may
provide another measure of self-sustained ring of the
motoneuron.SSDisthepercentageoftimeamotorunit
res on the descending phase of the contraction that is
below the level of torque needed on the ascending phase to
initially recruit the motor unit (Afsharipour et al., 2020).
In this case, the torque prole is used as a rough measure
of synaptic input to the motoneuron. To calculate SSD,
the formula below is used where ‘duration of ascending
phase’ =the duration of time the test unit was active on
the ascending phase of the contraction and the ‘duration
of descending phase’ =duration of time the test unit was
activeonthedescendingphase:
SSD =duration o f descending phase −duration of ascending phase
duration of descending phase +duration o f ascending phase×100%
Thus, a motor unit that red for 10 s on the descending
phaseofthecontraction,butforonly5sontheascending
phase (i.e. 5 s longer than expected if the motor unit was
de-recruited at the same input as during recruitment),
would have a SSD of 33%.
Composite modulation depth (CMod). To measure the
excursion (modulation depth) in the ring rate of the
composite prole that was used to measure the F,
theminimumvalueofthetpolynomiallineonthe
descending phase of the contraction was subtracted from
the maximum value (CMod). This was carried out to
ensure that any dierences in Fbetween groups were not
the result of a small CMod constraining the Fvalues.
Piecewise linear t to the ascending ring rate prole
(three-slope analysis). The ring rate proles during
the ascending portion of the contraction exhibited a
non-linear shape with roughly three distinct ranges
(slopes) that may arise from the activation of motoneuron
intrinsic conductances. We approximated this non-linear
response by three connected straight lines of dierent
slopes using the unconstrained MATLAB simplex search
method called fminsearch. The rst range is a rapid
increase in motor unit ring at the onset of the contraction
(termed secondary range), probabaly as a result of
the activation of the motoneuron PIC at recruitment
(Afsharipour et al., 2020; Binder et al., 2020) and we
approximated this with a xed line. After full PIC
activation, the ring rate of the motor unit increases more
slowly (termed the tertiary range), which we model with
a second joined line of lower slope. During the tertiary
range,wherewepresumesynapticdriveisstillincreasing,
motorunitringcandecreasebeforepeaktorqueis
reached. We refer to this as the tertiary sag range and these
data will be presented in a future paper. To t a piecewise
linear model to the secondary, tertiary and tertiary sag
ranges, we employed a non-linear search for parameters
m1–m3(slopes) and b1–b3(osets) of three linear
equations:
F(t)=m1t+b1if t<tBP1 Secondary range
F(t)=m2t+b2if tBP1 <t<tBP2 Te rt i a r y r a ng e 
F(t)=m3t+b3if t>tBP2 Te rt i a r y s a g r a n ge 
where F(t)isthemodelringrateandtis time. These
lines are constrained to be joined at two breakpoints
in time, tBP1 and tBP2, because we do not see discontinuities
in ring at the transitions between ranges. Thus,
F(tBP1)=m1tBP1 +b1=m2tBP1 +b2and thus, by
rearrangingwecompute:
tBP1 =(b1−b2)/(m1−m2)and similarly for the
second breakpoint
tBP2 =(b2−b3)/(m2−m3)
The six unknown parameters (m1,m2,m3,b1,b2and
b3)inthisfunctionF(t) were determined iteratively by
minimizing the least squared error between the model and
actual ring rates, starting with initial guesses (values) for
the breakpoints, and using the unconstrained MATLAB
simplexsearchmethodcalledfminsearch.Initialguesses
for the mand bparameters were obtained by separate
linear regressions on the segments of data between the
breakpoint values. Because of the non-linear nature of
this model (with variable breakpoints), there are multiple
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2067
possible solutions (at minima in criteria) depending on
thestartinginitialvaluesgiventothebreakpoints.We
systematically searched the solution space by testing all
possible initial values for tBP1 and tBP2 pairings at each
frequency point. Many nearby initial values converged
onto a single solution with a common minimum error, and
we manually chose the nal solution with both the near
minimum error and the largest number of initial value
pairs that led to that error. Sometimes two or more such
solutions arose, and we chose the one that best t the bin
averaged (10 ms bins) ring rate prole, as determined
visually. Final solutions were also constrained to have a
positive secondary range slope that was at least twice as
steepasthetertiaryrangeslopeandatertiarysagrange
with a slope <0Hzs
–1. If there was no secondary range
(i.e. b1<2x’s b2), the ring rate prole was t with just
one or two straight lines to delineate the tertiary range
only or the tertiary and tertiary sag ranges, respectively.
In many of the higher threshold units when there was no
period of low slope ring that was similar to the tertiary
range observed in the lower threshold units, we also t
oneortwolinestotheringrateproletodelineatethe
secondary range only or the secondary and tertiary sag
ranges, respectively. Finally, in cases where there was no
tertiary sag range, the straight line through the tertiary
range was extended to the peak torque to measure its slope
and duration.
Theproportionofmotorunitsexhibitingasecondary
and/or tertiary range was calculated in each participant
by rst binning the motor units according to their
recruitment threshold (see below) and pooling all of the
motor units from the four contractions together. The
percentage of total units in each bin with a secondary or
tertiary range was then calculated (e.g. number of units
with a secondary or tertiary range/number of total units
within the bin ×100%) and this percentage was then
averaged across participants within each group for each
bin. The slope and duration of the secondary and tertiary
ranges were also analysed like Fwith respect to age (see
above) and the recruitment threshold of the motor units
(described below).
Start, maximum and end rates. The start, maximum
or end ring rate for all units in each contraction were
measured from the t polynomial line using a custom
MATLAB program. The average value measured from
all units in a single contraction was calculated and this
value was then averaged across the four contractions in
each participant and then across participants within a
group. The mean start, maximum and end ring rates
were also plotted against the age of the participant in the
young development and young adult groups (7–28 years)
to visualize trends across age and across the recruitment
threshold of the motor units.
Antagonist muscle activation. Because the ring rate
proles of the TA motor units could be aected by a
counter load produced from antagonist muscle activation,
we used MATLAB to measure the mean rectied and
smoothed (fourth-order, zero lag Butterworth lter with
30 Hz low pass) soleus EMG during the entire triangular
dorsiexion. The mean noise of the signal, measured
before or after the contraction with no EMG activity, was
subtracted to better compare soleus EMG values across the
dierent participants. The mean amplitude of the soleus
EMG during the 30% MVC triangular contraction was
compared with the mean soleus EMG produced during
the less controlled, dorsiexion MVC where, in the latter,
we expect a greater amount of co-contraction.
Onionskineffect. As noted in the Introduction, the
onionskineectcanpresentwhenthemean(andby
extension the maximum) ring rate decreases with the
recruitment threshold of the motor units (De Luca
& Hostage, 2010; Erim et al., 1996). To determine
whether this eect occurred in the TA motor units
for our recording and decomposition protocol, in each
participant, the maximum ring rate of each motor
unit was plotted against its recruitment threshold,
grouping data from all four contractions together and
calculating repeated measures correlation coecients
(rmcorr). Briey, rmcorr avoids violating the assumption
of independence of observations by using analysis of
covariance to statistically adjust for inter-individual
variability and generating the best linear t for each
participant using parallel regression lines (the same slope)
with a unique intercept (Bakdash & Marusich, 2017). The
rmcorr can range from −1 to 1 and indicates the strength
of the linear relationship between two variables, similar
to a Pearson correlation coecient, with rmcorr of |0.10|,
|0.30| and |0.50| considered to have small, medium and
large eect sizes respectively (Bakdash & Marusich, 2017).
Analysis according to recruitment threshold of the
motor units. To examine whether the F,secondaryand
tertiary range or ring rate values varied according to
therecruitmentthresholdofthemotorunits,valuesfrom
motor units having a recruitment threshold within a given
range (e.g. bins of 0–3%, 3–5%, 5–7%, … 25–27% MVC)
from the four contractions were averaged together in each
participant using a custom MATLAB program. These
binned averages were then averaged across participants
within each group to compare between dierent group
pairings (e.g. young development vs.youngadult,SSRIvs.
SSRI control, etc.).
Torque steadiness and motor unit interspike inter-
val. Torque steadiness was quantied by measuring the
coecient of variation (CoV) of the detrended torque
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2068 G. Mohammadalinejad and others J Physiol 602.9
trace. In each participant, the four triangular torque ramps
were subjectively split into ascending and descending
components, determined visually as the point where the
torque transitioned from an ascending to a descending
prole near its peak. The MATLAB detrend function
was applied to the separated ascending and descending
components where the best straight-line t to the data was
removed. The SD of the detrended torque was calculated
across the entire ascending or descending phase. The SD
of the ascending or descending detrended torque was
then divided by the mean of the non-detrended torque
to produce the CoV (SD/mean) (Skinner et al., 2019).
Note that the homoscedasticity of the SD across the
ascending or descending phase was not measured and
was probably greater near peak torque compared to the
onsetofthecontraction.However,becauseweareonly
using CoV as a measure of torque variability, this was not
a large concern. The CoV of the interspike interval for
the motor unit discharge times was also measured in the
young development and young adult groups to determine
whether it followed a similar trend across age similar to
the CoV of the detrended torque. The SD and mean inter-
spike interval across the entire ascending and descending
discharge prole for each motor unit was measured to
calculate the CoV (SD/mean). In each participant, the
CoV of the interspike interval for all motor units from all
four contractions was pooled and then averaged together
to obtain a single CoV value and plotted against age.
Statistical analysis
Sigma Plot, version 11.0 (Grati LLC, Palo Alto, CA,
USA), was used for all statistics except for the repeated
measures correlation coecients described below.
Because the group Fandmeanringratesdidnotdier
between males and females during the 30% MVC contra-
ctions,aswillbedetailedelsewhere,datafromthetwo
sexes were collapsed within each group. A Shapiro–Wilk
test was used to assess the normality of the data. Group
comparisons consisted of young development vs.young
adult, young development vs.adult,youngadultvs.adult
and SSRI vs.SSRIcontrols.Unpaired,between-group
comparisons for normally distributed data were assessed
with Student’s ttest and non-parametric data were
assessed with the Mann–Whitney rank sum test. The
relationship between the dierent dependent variables
(i.e. F, CMod, SSD, secondary and tertiary range
slopes and durations, ring rates and torque steadiness)
and the age of the participant was evaluated with a
Pearson product correlation moment (r). Within each
group, a one-way analysis of variance (ANOVA) was
performedonnormallydistributeddataandaone-way
ANOVAonrankswasusedonnon-normallydistributed
data to compare the eect of motor unit recruitment
threshold on the F, secondary and tertiary range slope
and duration, and the various ring rates. A two-way
ANOVA was also performed to analyse the eect of
group and recruitment threshold and their interaction
on these dierent variables. To examine maximum ring
rates across the recruitment threshold of the motor
units, repeated measures correlation coecients were
calculated independently for each group using the rmcorr
package (Bakdash & Marusich, 2017) in R (R studio,
version 1.4.1106; R Foundation for Statistical Computing,
Vienna, Austria). P<0.05 was considered statistically
signicant. Data are presented as the mean ±SD or
median (range) when appropriate.
Results
Number and ring rate proles of decomposed motor
units
Using the blind-source separation algorithm (Negro et al.,
2016), 2164 single motor units from the four contractions
were decomposed from the TA HDsEMG activated during
the 30% MVC dorsiexion, with 575 units decomposed in
the young developmental group (aged 7–17 years, n=20),
669 units in the young adult group (aged 18–28 years,
n=13), 384 units in the SSRI group (aged 11–28 years,
n=9) and 536 units in the adult group (32–53 years,
n=8). In all groups, the average pulse-to-noise ratio,
or silhouette (SIL) value, was greater than the cuto
of 0.85 (young development =0.89 ±0.023; young
adult =0.90 ±0.030, SSRI =0.900 ±0.018 and adult
group =0.94 ±0.025). Example motor unit ring proles
with the t polynomial lines (black) from a single 30%
MVC contraction are shown in Fig. 1 for an 11-year-old
in the young development group, a 19-year-old in the
young adult group and from a 16-year-old taking an
SSRI; for similar data in adults, see Afsharipour et al.
(2020). All ten decomposed motor units are displayed
for the 11-year-old and are plotted in order of increasing
recruitment threshold, whereas 10 representative motor
units with similar increments in recruitment threshold
are displayed for the other two participants. The motor
unit ring rate proles of the 11-year-old participant
were similar to the participant taking SSRIs and distinct
from the older 19-year-old participant. For example, ring
rates were more variable in the young development and
SSRI participants and reached higher maximum rates
compared to the young adult participant, with the latter
reected by a greater number of polynomial lines crossing
above the coloured vertical lines. Moreover, motor units
recruited at torques of ∼10% MVC or higher continued
to re for longer as the dorsiexion torque (Fig. 1,
top trace) decreased towards zero at the end of the
contraction(Fig.1,verticaldottedline),especiallyinthe
SSRI participant compared to the young adult, suggesting
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2069
that these motor units had greater amounts of prolonged,
self-sustained ring.
When rst comparing across all participants, the
average number of motor units decomposed per
contractionincreasedmoresteeplywithageupuntil
∼30.25 years as measured by tting two straight lines to
the data (Fig. 2A,SSRIparticipantsinred).Onaverage,
thereweremoreunitsdecomposedpercontractionin
the young adult (12.86 ±7.65) and adult (16.75 ±6.40)
groups compared to the young development (7.03 ±3.67)
group (P=0.041 and 0.021, respectively, Mann–Whitney
rank sum test), with no dierence in the young adult and
adult groups compared to the SSRI (10.66 ±5.96) group
(P=0.479 Student’s ttest and P=0.295, Mann–Whitney
rank sum test, respectively). Moreover, the young adult
and adult participants had a greater proportion of
lower-threshold units (i.e. peak of gaussian curve at
2.5% MVC) (Fig. 2B,middle)comparedtotheyoung
Figure 1. Dorsiexion torque (top trace) and ring rate proles of 10 decomposed TA motor units
during a triangular 30% maximum voluntary contraction (MVC) plotted from top to bottom in order
of increasing recruitment threshold
Left: 11-year-old female (10 out of 10 units decomposed). Middle: 19-year-old male (10 out of 21 units). Right:
16-year-old male taking an SSRI (10 out of 19 units). The vertical long dashed lines mark the time of peak torque
and the dotted lines mark the time of zero torque at the end of the contraction. The black lines show the fth-order
polynomial t to the ring rate proles. The 0 Hz baseline (x-axis) is displayed in the corresponding colour for each
motor unit, except for the yellow units where grey is used to improve visibility. [Colour gure can be viewed at
wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2070 G. Mohammadalinejad and others J Physiol 602.9
developmentgroup(peakat7.5%MVC)(Fig.2B,top
left) and the participants taking SSRIs (peak at 12.5%
MVC) (Fig. 2B,bottomright).Becausethedistribution
of recruitment thresholds in the decomposed motor units
was dierent across groups, in addition to age we also
plotted Fand other measures across the recruitment
threshold of the motor units to determine whether it was
acontributingfactor.
Figure 2. Mean number and recruitment threshold
distribution of motor units
A, mean number of TA motor units per contraction plotted against
the age of the participant. Participants taking SSRIs are marked with
red circles. Two lines of different slope (green) with the smallest error
were iteratively t to the data. B, histogram distribution of number
of motor units binned according to the recruitment threshold of the
motor unit (in increments of 5% MVC) and divided by the number
of participants in each group. The combined young development
and young adult group (Combo: 7–28 years) is also displayed. N,
number of participants in each group. [Colour gure can be viewed
at wileyonlinelibrary.com]
Estimation of self-sustained ring
Fwas measured to estimate whether there were
changes in self-sustained ring of motoneurons across
development and whether it was larger in participants
taking SSRIs. To measure F, the prole of the
synaptic input to the TA motoneurons was estimated
byaveragingthemeanringrateproleofthetwoor
three lowest-threshold control units (e.g. bottom proles
in Fig. 3Awith grey, light blue and green dots) to produce
a composite control motor unit prole as per Afsharipour
et al. (2020). This estimated synaptic input was typically
lower when the test motor units (orange dots, top proles)
were de-recruited compared to when they were rst
recruited (Fmarked by length of downward arrows
in Fig. 3A), potentially from the added depolarization
of the PIC assisting self-sustained ring at low synaptic
inputs. Fwas larger in both the young development
(7.1 Hz) and SSRI (9.3 Hz) participants compared to
the young adult participant (4.6 Hz), consistent with the
motoneuron PIC in the former two groups producing
greater amounts of self-sustained ring.
To explore the relationship of age to Ffurther,
we plotted the mean Ffor each of the young
development and adult participants across age and
notedanegativecorrelationasmarkedbytheblack
regression line (r=−0.496, P=0.006, Pearson’s
product), showing that Fdecreases with age from 7
to 28 years (Fig. 3Ba). Correspondingly, the average F
in the young development group (5.83 ±1.15 Hz)
(Fig. 3Ca)waslargerthantheyoungadultgroup
(4.88 ±0.80 Hz, P=0.015, Student’s ttest) and the
adult group (4.78 ±1.05 Hz, P=0.035, Student’s ttest)
(for tvalues, see Supporting information, Table S1b).
Moreover, the average Fwas higher in the SSRI group
(6.45 ±0.56 Hz) (Fig. 3Ca) compared to the age-matched
controls (5.27 ±1.05 Hz, P=0.003, Student’s ttest; see
also Supporting information, Table S1b).
The larger Fin the young development and
SSRI groups could have been mediated by a greater
excursion in ring rate modulation of the composite
control motor units (CMod, lower proles in Fig. 3A).
However, unlike F,CModintheyoungdevelopment
(11.22 ±2.09 Hz) (Fig. 3Cb)wasnotlargercompared
to the young adult (10.17 ±1.66 Hz, P=0.139) or
adult groups (10.94 ±3.28 Hz, P=0.787, Student
ttests), nor were there dierences between the SSRI
group (11.88 ±0.99 Hz) and the age-matched controls
(10.62 ±1.97 Hz, P=0.076, Student’s ttest), with CMod
being almost twice that of the Fvalues. Moreover, there
was no relationship between CMod and age across the
young development and young adult groups as marked
by the grey regression line in Fig. 3Bb (r=−0.166,
P=0.357, Pearson’s product). A similar CMod across
groups suggests that the lower Fin the young adult
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2071
and adult groups were not constrained by a lower CMod.
The smaller number of decomposed TA motor units in
the young development group (Fig. 2) also could have
biased sampling to produce a higher F. However, when
comparing Fto a group of young adults and adults
(n=10) with a similar number of decomposed motor
units per contraction (median number of units =7.13)
compared to the young development group (median: 7.50,
P=0.704 Mann–Whitney rank sum test), Fwas still
larger in the young development group (5.83 ±1.15 Hz)
compared to this combined adult group (4.95 ±0.99 Hz,
P=0.035, Student’s ttest).
The dierences in estimated PIC activation between
the three participants from Fig. 3Awere also reected
in the measure of SSD, which is the percentage of
time a motor unit res below the torque needed to
initially recruit the motor unit (Fig. 3A, horizontal double
arrow) (Afsharipour et al., 2020). The SSD was 54%
in the 11-year-old, 21% in the 19-year-old and 47% in
the 16-year-old taking an SSRI. However, in the group
data (Fig. 3Cc), SSD in the young development group
(19.44 ±18.51%) was not larger compared to the young
adult (14.74 ±12.29%, P=0.543) or the adult group
(10.52 ±7.71%, P=0.525, Mann–Whitney rank sum
Figure 3. F, Control Unit Modulation Depth (CMod) and Self-sustained ring Duration
A, representative ring rate proles of composite TA control units (bottom proles, different control units indicated
with different colours) and test units (top proles, orange) used to measure F, CMod and SSD in the same
three participants from Fig. 1. Solid black lines indicate tted fth-order polynomial. B,F(Ba), CMod (Bb)and
self-sustained ring duration (SSD) (Bc) plotted against age across the young development (n=20) and young
adult groups (n=13). Black regression lines indicate there is a signicant correlation with age; grey regression
lines indicate no correlation. Participants taking SSRIs (n=9, not included in regression analysis) are marked with
red circles. C, height of bar indicates mean F(Ca), CMod (Cb) and SSD (Cc) for the young development (green),
young adult (blue), adult (black), SSRI (red) and SSRI control (grey) participants. ∗Statistical difference between
groups with P<0.05; the results of unpaired comparisons are provided in the text and Supporting information
(Table S1b). [Colour gure can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2072 G. Mohammadalinejad and others J Physiol 602.9
test), nor was there a correlation between SSD and age
across the young development and young adult groups
(Fig. 3Bc, grey regression line, r=−0.149, P=0.407,
Pearson’s product). By contrast, SSD in the participants
taking SSRIs (25.27 ±10.12%) was larger compared to
the age-matched controls (15.02 ±12.57%, P=0.035,
Student’s ttest), similar to F(Fig. 3Cc).
We also examined Fas a function of recruitment
threshold of the test motor units (Fig. 4) given the dierent
distributions of motor unit recruitment thresholds
between the groups (Fig. 2B) and the potential for it
to aect F(Harris-Warrick et al., 2023; Sharples &
Miles, 2021). When tested with a one-way ANOVA, there
was no signicant main eect of recruitment threshold
on Fwithin any of the groups (see one-way ANOVA
results in the Supporting information, Table S2a), in
agreement with Afsharipour et al. (2020). As with the
unpaired comparisons in Fig. 3Ca, a two-way ANOVA
showed a main eect of group on Fbetween the young
development and young adult groups (F1,12 =24.781,
P<0.001), the young development and adult groups
(F1,12 =21.690, P<0.001) and the SSRI and SSRI control
groups (F1,12 =24.558, P<0.001) (see all two-way
ANOVA results in the Supporting information, Table
S2b).
Proportion of motor units having a secondary and/or
tertiary range
Three joined, straight lines of dierent slopes were t to
theascendingportionoftheringrateprolestoquantify
the non-linearities in motor unit ring in response to
the presumed linear increase in synaptic drive to the
TA motoneurons (see ‘Piecewise linear t’ in Methods).
The rst slope, termed the secondary range (Fig. 5A,
light blue), is considered to reect the acceleration in
ring rate when the PIC is activated during recruitment
of the motoneuron and thus may potentially provide an
estimate of the upswing in membrane potential during
PIC activation. As shown for the top ring rate proles in
Fig. 5A,∼85–90% of all motor units activated between 0%
and 15% MVC had a secondary range (Fig. 5B,topgraphs
plotted for each group), with the remaining units having
ring rate proles that jumped directly onto the tertiary
range (e.g. unit 2 from 11-year-old in Fig. 1). As the
recruitment threshold of the motor units increased above
18% MVC (e.g. bottom units in Fig. 5A), almost 100% of
all motor units had a secondary range. The next slope,
termed the tertiary range (Fig. 5A, green), is considered
to reect the lower gain ring of the motoneuron after
thefullactivationofthePICincreasestheconductance
of the cell (Bennett et al., 1998). In every group, all
motorunitsrecruitedbetween0%and5%MVChada
tertiary range (Fig. 5B,bottomrow),andthisproportion
rapidly dropped o for motor units with recruitment
thresholds >13% MVC when the secondary range started
todominatemostoftheascendingringrateprole(e.g.
Fig. 5A, bottom proles). There were no dierences in
the proportion of units with a secondary or tertiary range
between the various groups (see two-way ANOVA results
in the Supporting information, Table S3).
Secondary and tertiary range slopes and durations
Similar to FandSSD,werstplottedtheslopeand
duration of the secondary and tertiary ranges across
the age of the participants. Similar to F,theslope of
the secondary range decreased with age (r=−0.387,
P=0.026, Pearson’s product) (Fig. 6Aleft), with the
young development group having a steeper secondary
range slope (5.73 ±1.93 Hz s–1)(Fig.6A,right)compared
to the young adult (4.34 ±0.90 Hz s–1)andadult
Figure 4. Fand recruitment threshold of the test motor unit
F plotted against recruitment threshold of the test motor units (binned every 2% MVC) for A, young development
(n=20, green) and young adult (n=13, blue) groups; B, young development and adult (n=8, black) groups; C,
SSRI (n=9, red) and aged-match control (n=26, grey) groups. Solid symbols denote a signicant group effect
from the two-way ANOVA analysis (all results of two-way ANOVA are displayed in the Supporting information,
Table S2b). Small circles indicate individual participant data. [Colour gure can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2073
(4.35 ±1.20 Hz s–1)groups(P=0.026 and P=0.050,
respectively, Mann–Whitney rank sum). By contrast, the
duration of the secondary range did not vary with age
across the young development and young adult groups
(r=0.167, P=0.354, Pearson’s product) (Fig. 6B,left),nor
were there dierences between the young development
and young adult or adult groups (Fig. 6B,right;seeresults
of unpaired comparisons in the Supporting information,
Table S4c). The secondary range slope was steeper in the
SSRI group (7.20 ±2.38 Hz s–1)(Fig.6A,right)andwitha
shorter duration (1.39 ±0.25 s) (Fig. 6B,right)compared
to the age-matched controls (slope: 4.96 ±1.69 Hz s–1,
duration: 1.68 ±0.32 s, P=0.004 Mann–Whitney rank
sum test and P=0.019 Student’s ttest, respectively).
Interestingly, the slope (and sometimes duration) of the
secondary range followed a similar trend across the groups
as the Festimate of self-sustained ring (as detailed
below in Fig. 8). By contrast, there were no associations in
Figure 5. Proportion of motor units with secondary and tertiary ranges
A, examples of secondary (blue), tertiary (green) and tertiary sag (red) ranges t to the ascending prole of both
a lower threshold (middle plot) and higher threshold (bottom plot) motor unit in the same three participants as in
Fig. 1. The vertical dashed line marks peak torque and the black horizontal line represents 0 Hz for the top ring
rate proles. B, top row: large circles mark mean proportion of units within each 2% MVC bin with a secondary
range in the young development (green) and young adult (blue) groups (left); young development and adult (black)
groups (middle); and SSRI (red) and age-matched controls (grey) (right). Small circles mark individual participant
values with overlapping data points (e.g. at 100% proportion) skewed slightly to the left or right of the bin centre
(e.g. 1%, 3% and 5% MVC, etc.). Bottom row: Same as for the top row but for the tertiary range. [Colour gure
can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2074 G. Mohammadalinejad and others J Physiol 602.9
the slope (∼0.6 Hz s–1) or duration (∼5s)ofthetertiary
range with age where motoneuron ring is asssumed to
occur during full PIC activation (grey regression lines
in Fig. 6Cand D, left; for rvalues, see Supporting
information, Table S4b), nor were there any dierences
betweenthevariousgroups(Fig.6Cand Dright graphs;
see unpaired comparisons in the Supporting information,
Table S4c).
We also examined whether the slope and duration
of the secondary and tertiary ranges varied across
the recruitment threshold of the motor units within
each group (Fig. 7). In general, the slope of the
secondary range decreased, and the duration became
longer as the recruitment threshold of the motor units
increased (marked with an asterisk in Fig. 7A,top
and bottom, respectively; see one-way ANOVA results
in the Supporting information, Table S5), consistent
with the threshold of the PIC increasing with the
recruitment threshold of the motoneuron as described
in the Introduction. By contrast, there was no eect of
recruitment threshold on the slope of the tertiary range
for any group (Fig. 7B,top;seeone-wayANOVAresultsin
theSupportinginformation,TableS5).Asexpected,there
was an eect of recruitment threshold for the duration of
thetertiaryrangebecausemoreoftheringrateprolein
the higher threshold units was comprised of the secondary
range (marked with an asterisk in Fig. 7B,bottom;see
one-way ANOVA results in the Supporting information,
TableS5),withmanyofthemotorunitsrecruitedator
above 23% MVC having no tertiary range and only a
secondary range as shown in Fig. 5B. The lled in circles
in Fig. 7 indicate a main eect of group from a two-way
ANOVAanalysis(seeSupportinginformation,TableS6),
with similar results to the unpaired comparisons in Fig. 6.
Correlation between Fand slope of the secondary
range
The slope of the secondary range exhibited similar trends
across age and between the dierent groups as the F.
For example, both the Fand the slope of the secondary
range decreased with age across the young development
and young adult groups and both were greater in the
young development group compared to the young adult
and adult groups and between the SSRI group and the
age-matched controls (compare Figs 3BaCa and 6A).
Thus,weplottedtheaverageslopeofthesecondary
range for each participant against their average Fto
determine whether they were indeed correlated (Fig. 8).
Across the entire group, there was a correlation between
Fand the slope of the secondary range (r=0.536,
P<0.001, Pearson’s product n=50), showing that these
two estimates of motoneuron PIC activation co-vary with
one another.
Figure 6. Slope and duration of the secondary and tertiary ranges
Left: mean slope of the secondary (A) and tertiary (C) ranges plotted against the age of the participants in the young
development and young adult groups (black circles, n=33) and participants taking SSRIs (red circles, n=9). Solid
black line represents signicant regression with age. SSRI participants are not included in the regression. Right:
mean slope of secondary (A) and tertiary (C) range (height of bar) for young development (green), young adult
(blue) and adult (black) groups and for SSRI (red) and SSRI-control (grey) groups. Small circles indicate individual
participant values. Band D,asin(A)and(C) but for secondary and tertiary range duration. ∗Statistical difference
between groups (see values in text and bolded numbers in the Supporting information, Table S4c). [Colour gure
can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2075
Torque steadiness
Similar to measures of motoneuron excitability, the
steadiness of the dorsiexion torque when trying to track
the target triangle also changed with development as
assessed by the coecient of variation (CoV =SD/mean)
of the detrended torque during both the ascending or
descending phases of the contraction. Similar to Fand
theslopeofthesecondaryrange,theCoVofthedetrended
torque, for both the ascending (r=−0.587, P<0.001)
(Fig. 9A, left) or descending (r=−0.493, P=0.004)
(Fig. 9B, left, Pearson’s product) phase of the contraction,
decreased across age between 7 and 28 years, signifying
an improvement in dorsiexion skill during development.
Overall, the CoV of the ascending torque prole in the
young development group was larger (0.093 ±0.036)
compared to the young adult (0.059 ±0.016) and adult
(0.057 ±0.011) groups (P=0.002 and P=0.004,
Figure 7. Slope and duration of the secondary and tertiary ranges according to the recruitment
threshold of the motor units
A, slope (top) and duration (bottom) of the secondary range plotted against motor unit recruitment threshold (2%
MVC bin widths): left: young development vs. young adult; middle panels: young development vs. adult; right:
SSRI vs. SSRI controls. B,asin(A) but for tertiary range. ∗Statistically signicant effect of recruitment threshold
on secondary and tertiary range slope or duration from one-way ANOVA colour coded to the respective group
(see bold numbers in the Supporting information, Table S5). Solid circles indicate a main effect of group from a
two-way ANOVA (see bold numbers in the Supporting information, Table S6). Small circles represent individual
participant values. [Colour gure can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2076 G. Mohammadalinejad and others J Physiol 602.9
respectively,Mann–Whitneyranksumtest)(Fig.9A,
right). The CoV of the detrended torque during the
descending phase of the contraction was also larger in
the young development group (0.089 ±0.034) but only
compared to the young adult group (0.067 ±0.023,
P=0.041, Mann–Whitney rank sum test) (Fig. 9Bright;
results of all unpaired comparisons in the Supporting
information, Table S7c). By contrast, torque steadiness in
the participants taking SSRIs (Fig. 9Aand B,openred
circles) was not dierent compared to the age-matched
controls (see values in the Supporting information, Table
S7a,c), consistent with a similar level of dorsiexion skill
inthetwogroups.SimilartotheCoVofthedetrended
torque, the CoV of the interspike interval of the motor unit
ringtimesfortheentireringrateprolealsodecreased
across age (r=−0.524, P=0.002, Pearson’s product)
(Fig. 9C, left), with the CoV of the interspike interval in the
young development group (0.51 ±0.09) higher compared
to the young adult group (0.43 ±0.07, P=0.018 Student’s t
test) (Fig. 9C, right), signifying that motor unit ring rates
were also more variable at ages less than 18 years.
Start, maximum and end ring rates across age
The mean ring rates of the TA motor units activated
during a 30% MVC isometric dorsiexion were often
higher in the young development and SSRI groups
compared to the adult and SSRI control participants,
respectively. As shown for a young development (10 years
old) and 21-year-old SSRI participant (Fig. 10A, left
and right), the maximum and end ring rates of the
various motor units were often higher compared to
young adults (e.g. 24-year-old) (Fig. 10A, middle) and
Figure 8. Fversus slope of secondary range
The average secondary range slope plotted against the
corresponding average Fin each participant (n= 50): young
development (green), young adult (blue), adult (black) and SSRI (red).
The black regression line was t through all data points. [Colour
gure can be viewed at wileyonlinelibrary.com]
adults (not shown). Across the young development and
young adult groups from 7 to 28 years of age, there
was a negative correlation between the maximum or
end ring rates and age (r=−0.363, P=0.038 and
r=−0.484, P=0.004 respectively, Pearson’s product)
(Figs 10Bb and Bc). Moreover, maximum ring rates
in the young development group (18.32 ±2.44) were
higher compared to the young adult group (16.51 ±1.75,
P=0.028, Student’s ttest) (Fig. 10Cb)and,similarly,
the end ring rates in the young development group
(7.29 ±1.64) were higher compared to both the young
adult (6.13 ±1.03) and adult (6.04 ±0.85) groups
(P=0.029 and P=0.050 respectively, Student’s ttest)
(Fig. 10Cc). By contrast, the start rates in the young
development group (10.38 ±1.95 Hz) were not dierent
compared to the young adult (9.36 ±1.54 Hz) and
adult (9.84 ±1.98 Hz) groups (Pall >0.05, Student’s t
test) (Fig. 10Ca;seestatisticalresultsintheSupporting
information, Table S8c), nor was there a correlation to age
(r=−0.261, P=0.142, Pearson’s product) (Fig. 10Ba).
However, both the start (11.18 ±1.57 Hz) and maximum
(19.02 ±1.98 Hz) ring rates were higher in the SSRI
group compared to the age-matched controls: 9.64 ±1.82,
P<0.001 Mann–Whitney rank sum test) (Fig. 10Ca)and
17.23 ±2.26 Hz, P=0.042, Student’s ttest) (Fig. 10Cb),
respectively.
We examined whether the higher ring rates of the TA
motor units in the young development group occurred
because of a larger activation of the antagonist soleus
muscle that produced a larger counter load to the ankle
dorsiexion. The mean amplitude of the rectied and
smoothed soleus EMG during the triangular dorsiexion
contraction was very low in all groups (∼2μV) compared
to the soleus activity generated during the less controlled,
maximum dorsiexion contraction (∼50 μV). The
median rectied and ltered soleus EMG in the young
developmentgroupwas2.156μV (range 1.219–2.863μV)
and was not dierent to the young adult group at 1.773
μV (range 1.342–2.657 μV, Mann–Whitney sum rank
test, P=0.401, T=2304.000). Thus, the coactivation
of the antagonist soleus muscle probably had little to no
eect on the ring rates of the TA motor units or on the
other measures of unit ring such as F,secondaryand
tertiary range values.
Start, maximum and end ring rates across motor
unit recruitment threshold
The start, maximum and end ring rates were further
analysed based on the recruitment threshold of the motor
units (Fig. 11). When tested with a one-way ANOVA,
there was no main eect of recruitment threshold of the
motor units on the start or end ring rates in any in any
ofthegroups(seenon-boldnumbersintheSupporting
information, Table S9a). By contrast, a one-way ANOVA
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2077
on ranks showed a signicant main eect of recruitment
thresholdonthemaximumringratesintheyoung
adults (marked with an asterisk in Fig. 11; H1,12 =22.850,
P=0.029), increasing as the recruitment threshold
increased. When taking the recruitment threshold of the
motor units into account, the higher ring rates in the
young development and SSRI groups compared to the
adults and age-matched controls, respectively, were more
apparent. The two-way ANOVA showed a signicant
main eect of group on the start, maximum and end
rates between the young development and young adult
groups and for the maximum and end ring rates between
theyoungdevelopmentandadultgroups(solidcirclesin
Fig. 11; see also bold values in the Supporting information,
Table S9b). Similarly, the two-way ANOVA showed a
signicant main eect of group for the start and maximum
ring rates between the SSRI and SSRI control groups.
Maximum ring rate vs. recruitment threshold of
motor unit
We further investigated the relationship between maximal
ring rates and recruitment threshold of the motor units
(e.g.Fig.12A) by calculating the repeated measures
correlation coecients (rmcc) for each participant within
each group. Across all groups, the relationship between
the maximum ring rates and recruitment threshold
of the motor units was signicantly positive during
contractions performed up to 30% MVC (Fig. 12B–F),
indicating that the recruitment threshold and ring rate
patterns did not follow an onion skin pattern as in De
Luca & Hostage (2010). Although this relationship was
signicantly positive in all groups, it varied from very
weak in the adults [r533 =0.12 (0.037, 0.2), P=0.005]
to moderate in the young development [r610 =0.27 (0.2,
0.34), P<0.0001] and SSRI groups [r420 =0.28 (0.19,
0.37), P<0.0001], whereas the young adult [r667 =0.21
(0.14, 0.28), P<0.0001] and SSRI control [r1133 =0.21
(0.15, 0.26), P<0.0001] groups were in between. The
stronger relationship in the young development and
SSRI groups occurred because the higher threshold units
(blue) typically had higher ring rates compared to the
lower threshold units, as shown for the superimposed
polynomial ring rate proles in a representative SSRI
participant (Fig. 12G). By contrast, some (but not all) of
the highest threshold units in the adults had lower ring
Figure 9. Variability of torque steadiness and interspike interval
Left: average coefcient of variation (CoV) of the detrended torque plotted by age across the young development
and young adult groups for the ascending (A) and descending (B) phase of the contraction. The black line indicates
a signicant correlation. Participants taking SSRIs are marked by red open circles and were not included in the
regression. Right: height of bar indicates mean value for each group with individual participant data marked by
small circles. C, left: average CoV of the interspike interval for all motor units in each participant plotted against
age for the young development and young adult groups. Right: as in (A)and(B) but for young development and
young adult groups only. ∗P<0.05 and ∗∗P<0.01, denotes signicant differences between groups (see bold
numbers in the Supporting information, Table S7c). [Colour gure can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2078 G. Mohammadalinejad and others J Physiol 602.9
rates compared to the lowest-threshold units (Fig. 12H),
resulting in a shallower slope and lower rmcc.
Discussion
In rodents, during the rst 3 weeks after birth,
motoneuron PICs decrease in threshold and increase
in amplitude to help secure motoneuron recruitment
and amplify synaptic inputs (Quinlan et al., 2011;
Sharples & Miles, 2021). At early developmental stages in
vitro, PICs produce little self-sustained activation of the
motoneuronthatwouldprolongtheeectsofsynaptic
inputs, especially in small motoneurons (Harris-Warrick
et al., 2023). Using indirect methods, in the present study,
weshowinhumansthat,by7yearsofage,motoneurons
have appreciable self-sustained activity, with participants
Figure 10. Start, maximum and end ring rates
A, representative ring rate proles of motor units used to measure the start, maximum, and end ring rates
from the t fth-order polynomial line (black). Left: 10-year-old (yr) female. Middle: 24-year-old female. Right:
24-year-old male taking an SSRI. The horizontal lines matched to the colour of the corresponding motor unit
indicate the baseline of 0 Hz (except for the last yellow unit). B, similar format to Fig. 3 with start (Ba), maximum
(Bb)andend(Bc) ring rates plotted against age for the young development and young adult groups. C, similar
format to Fig. 3 for start (Ca), maximum (Cb)andend(Cc) ring rates. ∗Statistically signicant difference between
groups with P<0.05; results of unpaired comparisons are presented in the text and Supporting information (Table
S8c). [Colour gure can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2079
between the ages of 7 and 17 years having larger F
values compared to participants between the ages of
18 to 28 years or 32 to 53 years of age. We also directly
quantied the slope and duration of the initial acceleration
in motor unit ring (secondary range) during a slowly
increasing synaptic drive as an indirect measure of the
size of the PIC eect during its initial, abrupt activation.
Because PIC activation is initiated over a xed time in an
all-or-none manner (Bennett et al., 1998), faster increases
in ring during the secondary range should occur for
larger PICs (Afsharipour et al., 2020), as detailed further
below. Furthermore, PICs are activated subthreshold to
ring and continue to activate rapidly at the onset of
ring, and thus, lower threshold PICs that are in the
steep part of their activation at recruitment should lead
to a steeper secondary range. Taken together, steeper
secondary range slopes suggest larger, lower threshold
PICsandisconsistentwithourndingthatmotoneurons
with larger estimated PICs by the Fmethod have
steeper secondary range slopes. For example, the young
development group had both larger Fvalues and
secondary range slopes compared to adults, suggesting
the motoneurons had larger and lower threshold PICs
before and during adolescence compared to adulthood.
Similarly, in participants taking SSRIs, both measures of
PIC activation, Fand the slope of the secondary range,
were also larger and steeper respectively compared to
their age-matched controls, consistent with serotonin
increasing the amplitude and decreasing the threshold
of motoneuron PICs (Harvey et al., 2006a; Li et al.,
2007). The increased excitability of motoneuron PICs
was also associated with higher ring rates in the young
development and SSRI groups. Below, we discuss some
of the intrinsic and extrinsic factors that may produce
the developmental changes in motoneuron excitability
that could combine to lower the torque steadiness and
Figure 11. Start, maximum and end ring rates according to recruitment threshold of the motor units
The start (top), maximum (middle), and end (bottom) ring rates plotted against the recruitment threshold of
the motor units for the young development (green), young adult (blue), adults (black), SSRI (red), and SSRI-age
matched controls (grey). ∗Statistically signicant effect of recruitment threshold on maximum ring rate in the
young adult group (bold numbers for one-way ANOVA in the Supporting information, Table S9a). Solid symbols
indicate signicant effect of group on the start, maximum or end rate from a two-way ANOVA (bold numbers in
the Supporting information, Table S9b). Small circles indicate individual participant values. [Colour gure can be
viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2080 G. Mohammadalinejad and others J Physiol 602.9
dorsiexion skill observed in the young development
group.
Relation of the secondary range and start rates to PIC
activation
Theinitialaccelerationinringrateattheonsetofmotor
unit recruitment, as quantied by the slope and duration
of the secondary range, is assumed to be produced by
the low-voltage activation of the PIC near the onset
of motoneuron recruitment (Afsharipour et al., 2020;
Bennett et al., 1998; Binder et al., 2020; Lee & Heckman,
1998). In addition, a slow inactivation of Kv1.2 channels
in the axon initial segment may also contribute to the
ringrateaccelerationduringthesecondaryrange(Bos
et al., 2018). This could occur if the Kv1.2 channels were
activated quickly at the onset of ring to resist recruitment
and then slowly inactivated over the next few seconds
to facilitate the initial ring rate acceleration mediated
bytheincreasingsynapticinputandPIC.Aswehave
previously predicted (Afsharipour et al., 2020), motor
units with low recruitment thresholds have a steeper
secondary range slope compared to higher threshold units
(Fig. 7). A steep secondary range in low threshold units
could be produced if a substantial portion of the PIC is
activated subthreshold to ring, with initial motor unit
ringoccurringduringthelatter,steeprisingportionof
the PIC. By contrast, the shallower secondary range slope
in the higher threshold motor units could result from the
PIC being activated closer to motoneuron recruitment
where the initial acceleration of the PIC is slower (Bennett
et al., 1998; Binder et al., 2020; Svirskis & Hounsgaard,
1997).
ThendingthattheyoungdevelopmentandSSRI
participants had a shorter duration secondary range
Figure 12. Relation of maximum ring rate and motor unit recruitment threshold
A, straight line t through data points of maximum ring rate (obtained from the t polynomial line) plotted
against recruitment threshold for all motor units from four contractions in an 18-year-old female participant. B–F,
best linear t from the analysis of covariance for each participant using parallel regression lines but with unique
y-intercepts for the young development (B), young adult (C), adult (D), SSRI (E) and SSRI age-matched controls
(F). Each participant is identied with a different colour. G, overlay of polynomial ts to the ring rate proles for
one contraction from a 16-year-old female participant in the SSRI group, as well as from a 52-year-old female
participant in the adult group (H). [Colour gure can be viewed at wileyonlinelibrary.com]
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2081
would also suggest that the PICs in these two groups have
a reduced threshold compared to the adult groups. In
supportofthis,oneofthemajoreectsofserotoninisto
reduce the threshold of the PIC (Li et al., 2007) and this
might contribute to more of the lowest threshold units in
theSSRIgrouphavingabrief,orevenlackofasecondary
range with ring starting directly on the tertiary range
(Fig. 5). Similarly, when taking recruitment threshold into
account, the start rates of the motor units in the young
development and SSRI groups were also higher compared
to the adults and age-matched controls, consistent with
larger and more rapidly activating PICs at recruitment.As
discussed below (see ‘Functional consequences of lower
threshold and larger PICs in the young development
group’), a steep secondary range and high initial ring
rate produced by a large subthreshold PIC may not only
provide a more secure recruitment of motor units, but
also may make gradually increasing torque at the onset
of a contraction more dicult if several motor units
are abruptly recruited together at high ring rates (Del
Vecchio, Sylos-Labini et al., 2020).
Larger SSD and/or Fin young development and
SSRI participants
Interestingly, our results demonstrate that the slope
of the secondary range may be a very eective
and simpler estimate of PIC activation given that
participants with steep secondary range slopes also
had longer self-sustained ring as estimated by the
F(see correlation in Fig. 8), suggesting that the two
measures could be used interchangeably even though
they represent dierent aspects of PIC activation. The
larger self-sustained ring, as measured by the F,may
be produced by a larger (or more sustained) motoneuron
PIC in the young development and SSRI groups compared
to the adults and age-matched controls, respectively. The
SSD, which uses the torque prole as an estimate of the
synaptic input to the motoneurons, may also reect the
contribution of the PIC to self-sustained ring but could
be less sensitive because larger SSD values were only
seen for the SSRI group compared to the age-matched
controls. The larger self-sustained ring (F)intheyoung
development group may result, in part, from reduced
spinal inhibition. Excitatory reexes, such as the stretch
reex, show an age-related decline during childhood that
reaches adult levels around 12–14 years, potentially from
a maturation of inhibitory spinal circuitry (Geertsen
et al., 2017; Willerslev-Olsen et al., 2014), the latter
which reduces PICs (Bennett et al., 1998; Hyngstrom
et al., 2007; Hyngstrom et al., 2008). Hormones, such
as oestradiol and testosterone, facilitate the maturation
of inhibitory GABAergic signalling (Gilfarb & Leuner,
2022). Because these hormones increase during the
young development period (Barrientos et al., 2019; Wood
et al., 2019), increases in the excitability of inhibitory
circuitry may contribute to the age-related decrease in
self-sustained ring we observed in this study. We did not
measure whether participants in the young development
group were before, within or after puberty, to determine
whether motoneuron PICs changed in direct relation to
predicted changes in sex hormones. There may also be a
developmental decrease in the activation of the thermo-
sensitive, transient receptor potential melastatin 5 (i.e.
Trpm5) channel, which is the main sodium ion carrier
for ICaN that also contributes to the self-sustained ring
of spinal motoneurons in young (postnatal days 5–12)
mice (Bos et al., 2021). Lastly, the larger self-sustained
ring (F) in the SSRI group may result from the direct
facilitation of the NaVand CaVchannels that mediate the
PIC via activation of serotonin (i.e. 5-HT2B/C)receptors
on the motoneuron (D’Amico et al., 2013; Goodlich et al.,
2023; Murray et al., 2010; Murray et al., 2011; Wei et al.,
2014) or microglia (El Oussini et al., 2016).
Confounds of different motor unit numbers and
thresholds between groups in measuring F
The smaller number of decomposed motor units in
the young development group, potentially as a result of
muscles with smaller cross-sectional area (Del Vecchio,
Holobar et al., 2020; Oliveira et al., 2022; Taylor et al.,
2022), may have biased the sampling of test motor units
with higher Fvalues compared to adults. However, when
participants in the young adult and adult groups with a
similarnumberofdecomposedmotorunitswereselected,
the mean Fwas still lower compared to the young
development group. The young development and SSRI
groups also had a smaller proportion of low threshold
motor units, potentially because smaller action potentials
that were less readily decomposed from the HDsEMG)
that could have lower Fs. For example, measuring F
in the lowest threshold test units could have a oor eect
wheretheringrateofthecompositecontrolmotorunits
when the test motor units are de-recruited cannot reach
rates that are lower than the rate when the test motor
units are recruited early in the contraction. However, there
were no eects of motor unit recruitment threshold on
the dierences in Finanyofthegroupcomparisons,in
agreement with earlier studies (Afsharipour et al., 2020).
Moreover, the Fvalues of the lowest threshold units
activated between 0% and 3% MVC, which probably
includes small slow motoneurons, were just as large as
those of the higher threshold units. This contrasts with
mice motoneurons recorded in vitro up to adolescence
(3 weeks) where small motoneurons lack self-sustained
(bistable) activity (Harris-Warrick et al., 2023), raising
the question of whether dierences in species (mice vs.
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2082 G. Mohammadalinejad and others J Physiol 602.9
human) or recording conditions (in vitro vs. in vivo)are
driving these dierent ndings.
Estimation of input/output gain of motoneurons
from the tertiary range
There appears to be a slightly higher gain in transducing
ring frequency from synaptic inputs in motoneurons of
the young development group compared to young adults.
That is, the slope of the tertiary range when compared
across motor units of dierent recruitment threshold was
slightly larger in the young development group compared
totheyoungadultgroup,althoughthiswasaweak
eect not seen in the overall group averages. This would
suggest that the transduction of synaptic inputs into
motoneuronringfollowingfullPICactivationoccursat
ahighergainintheyoungdevelopmentgroupandagrees
with the overall observed higher ring rates; however,
dierences in synaptic inputs between groups cannot be
ruled out. Further experiments with better quantiable
and controlled synaptic inputs, such as those generated
during muscle stretch (Gorassini et al., 1999; Powers et al.,
2008), are needed to examine the true input/output gain
of the tertiary range and determine whether it varies
across development (Smith & Brownstone, 2020), after
brain or spinal cord injury (Harvey et al., 2006a) or from
motoneuron disease such as amyotrophic lateral sclerosis
(Huh et al., 2021; Jensen et al., 2020) as shown in animal
models.
Direct measurement of the secondary and tertiary
ranges
We have demonstrated that it is important to measure
the secondary and tertiary range slopes directly using
optimization techniques, which demonstrates a strong
relationship between the secondary range slope and the
F(PIC) and subtle changes in tertiary slope that
reect motoneuron gain as detailed above. Other indirect
methods of estimating these slopes were suggested
(Beauchamp et al., 2023), including our work where only
a single slope was t to the ascending or descending ring
rate prole (Afsharipour et al., 2020), and have produced
dierent ndings. For example, the average slope of
the tertiary range using our optimization technique is
∼0.6 Hz s–1, which was around 10 times shallower than
theaverageslopeofthesecondaryrangeat∼6.0 Hz s–1.
These slope values are higher compared to the slopes
measured by drawing a straight line from the discharge
rate at motor unit recruitment to the ring rate at the end
oftheascendingphaseofthecontractionandmeasuring
the location of the longest orthogonal line to determine
the transition point between the secondary and tertiary
ranges; secondary range slope: 2.44 Hz s–1, tertiary range
slope: 0.4 Hz s–1 in Beauchamp et al. (2023). The lower
slopes from this method probably stem from including
moreofthetertiaryrangeintothesecondaryrangeand
not excluding the tertiary sag range in contrast to when
these slopes are directly tted with three joined straight
lines from our linear piecewise method.
Higher motor unit ring rates in young development
and SSRI participants
Previous reports have suggested that in prepubescent
children, ring rates in rst dorsal interosseous motor
units are higher compared to adults at matched levels
of MVC force to compensate for the inability to activate
motor units with high recruitment thresholds (Dotan
et al., 2012; Miller et al., 2019). In agreement with this
and other studies (Okudaira et al., 2023; Piotrkiewicz
et al., 2007), we also show that the maximum and end
ring rates of TA motor units decrease with age from 7 to
28 years, with the young development group having higher
ringratescomparedtotheyoungadultand/oradult
groups. Although superimposition of motor unit action
potentials at peak contractions may reduce the accuracy
of decomposition and measures of maximum ring rates,
the nding that the end ring rates were also higher, where
there is less superimposition, supports the conclusion of
higher maximum ring rates in the young development
group. The higher ring rates in the young development
group may be produced by a higher synaptic drive needed
to reach a comparable level of 30% MVC torque compared
totheadultgroupsifnotasmanylarge,fastmotor
units with higher force generation were recruited (Dotan
et al., 2012). Alternatively, or in addition, the higher
ring rates in the young development group may result
from motoneurons with shorter AHPs, as demonstrated
for biceps motoneurons in 5.5–19-year-old compared to
37.5–79-year-old participants (Piotrkiewicz et al., 2007).
Whatever the mechanism, the need to contract against
a higher antagonist (e.g., soleus) load likely did not
contribute to the higher ring rates in the TA motor units.
The higher ring rates in the young participants might be
matched to their faster twitch contraction times, where, in
the TA muscle, it progressively slows from 20 to 100 years
of age (Vandervoort & McComas, 1986) and are probably
even shorter at earlier ages.
Unexpectedly, the end ring rates were not lower in
the young development or SSRI groups even though the
PICs may have been larger compared to the adults and
age-matched controls, respectively, where NaPICs could
regeneratively drive very slow ring rates at low synaptic
drive near de-recruitment (Li et al., 2004). Perhaps the
synaptic drive at the end of the contraction decreased
too rapidly to allow for appreciable NaPIC-mediated
discharge that is more readily seen during a steady,
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2083
weak contraction (Gorassini et al., 2004). The start
and maximum ring rates were also higher in the
SSRI group compared to the age-matched controls,
consistent with 5-HT increasing the input resistance of the
motoneuron (Harvey et al., 2006a), facilitating the NaPIC
and shortening the AHP (Harvey et al., 2006b; Perrier
et al., 2013) to possibly produce higher rates of initial and
steady ring.
Increasing maximum ring rates with recruitment
threshold of the motor units
Dierent methods of isolating single motor units
from EMG signals have led to dierent conclusions
between ring rates and recruitment threshold, which
we reinvestigate here. Studies utilizing intramuscular
EMG and visual motor unit identication have shown
that low threshold motor units peak at lower rates and
forces compared to higher threshold motor units where
ring rates continue to increase up to 100% MVC (Barry
et al., 2007; Gydikov & Kosarov, 1974; Jesunathadas et al.,
2012; Moritz et al., 2005; Oya et al., 2009). This pattern
of ring rate discharge is logical for ecient muscle force
production because low threshold motor units achieve
peak tetanic forces at lower ring rates compared to higher
threshold units given the slower twitch contraction times
in the former (Piotrkiewicz & Turker, 2017). Thus, the
low ring rates in lower threshold motor units keep the
maximum ring rates tuned to their slower contraction
times, potentially preventing unnecessary high ring rates
and energy expenditure (Burke, 1968). Similarly, we also
observe a positive relationship between maximum ring
rates and recruitment threshold of the motor units isolated
with HDsEMG and blind source decomposition where
several motor units can be simultaneously identied
during a single contraction. However, in some cases, the
ring rates of the highest threshold motor units recruited
near peak torque were less than the lower threshold motor
units (e.g. dark blue proles in Fig. 12H). This may be
produced by a reduced availability of increasing supra-
threshold current near the 30% MVC peak torque to drive
the ring rates higher (Powers & Heckman, 2017). Future
experiments will examine whether the ring rates of these
motorunitscanbedrivenhigherbylargersynapticinputs
during contractions that are >30% MVC.
How are these results reconciled with other studies
which show a decrease in maximum ring rates as the
recruitment threshold of the motor unit increases to
producetheonionskineect?Previousstudiesshowing
an onion skin eect in the TA at comparable contraction
intensities (20% to 50% MVC) were decomposed using
ve-pin surface arrays and a precision decomposition
technique (De Luca & Hostage, 2010; Nawab et al., 2010).
This technique is prone to a greater probability of missing
spikes from the superimposition of motor units during
stronger contraction intensities (Farina & Enoka, 2011).
Thus, missed spikes in the higher threshold units at higher
forces would articially produce lower ring rates. Pre-
cision decomposition also selects motor unit activation
timesbasedontheunit’slocalpredictedringrate,
potentially producing articially higher ring rates in
the lower threshold units as well. By contrast, motor
units visually isolated from intramuscular EMG or from
64-electrode HDsEMG and blind source decomposition
may be less susceptible to missed spikes and forced
ring rate proles (Caillet et al., 2023; Farina & Enoka,
2011) and thus, show higher ring rates in progressively
activated motor units. Therefore, we suggest that in the
TA, decreases in maximum ring rates with recruitment in
theonionskineectisanartifactofmissedspikesand/or
forced ring rate proles when using 5-pin surface arrays
and precision decomposition. However, further studies
using HDsEMG and blind source decomposition at higher
contractions are needed to determine whether increases in
maximum ring rates with recruitment threshold is still
maintained, as occurs with contractions up to 90% MVC
for TA motor units visually identied from intramuscular
EMG (Jesunathadas et al., 2012).
Functional consequences of lower threshold and
larger PICs in the young development group
A lower-threshold and potentially larger motoneuron PIC
in the young development group may help descending
synaptic inputs recruit and prolong motoneuron ring
during voluntary contractions. However, these facilitated
PICs may produce a more abrupt and synchronous
recruitment of motor units at higher rates of discharge,
contributingtothegreaterringratevariability
and torque unsteadiness (i.e. higher CoV) observed
during the ascending phase of the contraction in the
young development group (Fig. 9). Similarly, greater
self-sustainedringofmotoneuronsmaycontributeto
the greater torque unsteadiness during the descending
phase of the contraction, making the descending control
of motoneuron de-recruitment and muscle relaxation
more challenging. Although PICs are instrumental in
amplifying synaptic inputs and facilitating repetitive
discharge, the membrane potential of motoneurons
canvarymorewhenonaplateaupotential(Bennett
et al., 1998), although we do not know why this is the
case. Thus, the lower threshold and larger PIC in the
younger participants may introduce more variability in
the motoneuron membrane potential, making the task
of matching the triangular torque prole more dicult.
This may be especially problematic in participants aged
less than 12–14 years in whom the corticospinal tract
and cortical grey matter are still undergoing substantial
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2084 G. Mohammadalinejad and others J Physiol 602.9
anatomical and functional development (Christova &
Georgopoulos, 2023; Nezu et al., 1997; Petersen et al.,
2010; Yeo et al., 2014), along with a slower processing of
visual information and motor execution to visual targets
(Davies et al., 2015). By contrast, participants taking
SSRIsalsohadsteeperandshortersecondaryrangesand
more self-sustained motoneuron ring but without a
larger torque unsteadiness. Because this group was older
(17.02 ±5.02 years) compared to the young development
group (12.18 ±2.69 years), the SSRI group may have
a more developed visual feedforward/feedback system,
corticospinal tract and spinal inhibition to help control
the lower threshold and larger amplitude motoneuron
PICs.
Relation to previous animal and human work and
clinical implications
As mentioned in the Introduction, in rodents, the
amplitudeofthePICincreasesintherst3weeksafter
birth as the animal gains weight bearing locomotion. We
observed evidence consistent with large PIC activation
producing longer self-sustained ring in motoneurons
of children as young as 7 years of age where Fdeclines
from 7 to 28 years and remains stable throughout
adulthood (Huh et al., 2021) until it starts to decline
around 65–70 years of age (Hassan et al., 2021; Orssatto
et al., 2021). Given the importance of sex hormones
such as oestrogens, progestogens and androgens on
metabolic function in spinal motoneurons (Vegeto et al.,
2020), it would be important to also look at changes in
motoneuron properties according to hormonal prole
across development, rather than just across chronological
age. Finally, it would be interesting to determine in
newborns whether motoneuron PICs, as in rodents, also
start out small and non-hysteretic and become larger
and more persistent to provide synaptic amplication
and self-sustained ring as bipedal locomotion is
achieved. None-the-less, data from this study can be
used to compare to childhood neurological disorders
such as cerebral palsy, Down’s syndrome and spinal
muscular atrophy to determine whether alterations in the
threshold and persistence of PICs and transduction of
synaptic inputs into ring may contribute to the motor
abnormalities, such as hyper- and hypotonia, present in
these conditions.
References
B., Manzur N., Duchcherer J., Fenrich K. F., Thompson C.
K., Negro F., Quinlan K. A., Bennett D. J., & Gorassini M.
A. (2020). Estimation of self-sustained activity produced
by persistent inward currents using ring rate proles of
multiple motor units in humans. Journal of Neurophysiology,
124(1), 63–85.
Bakdash J. Z., & Marusich L. R. (2017). Repeated measures
correlation. Frontiers in Psychology,8, 456.
Barrientos R. M., Brunton P. J., Lenz K. M., Pyter L., &
Spencer S. J. (2019). Neuroimmunology of the female brain
across the lifespan: Plasticity to psychopathology. Brain,
Behavior, and Immunity,79, 39–55.
Barry B. K., Pascoe M. A., Jesunathadas M., & Enoka R.
M. (2007). Rate coding is compressed but variability is
unaltered for motor units in a hand muscle of old adults.
Journal of Neurophysiology,97(5), 3206–3218.
Beauchamp,J.A.,Pearcey,G.E.P.,Khurram,O.U.,Chardon,
M., Wang, Y. C., Powers, R. K., Dewald, J. P. A., &
Heckman, C. J. (2023). A geometric approach to quantifying
the neuromodulatory eects of persistent inward currents
on individual motor unit discharge patterns. Journal of
Neural Engineering 20(1), 016034.
Bennett D. J., Hultborn H., Fedirchuk B., & Gorassini
M. (1998). Synaptic activation of plateaus in hindlimb
motoneurons of decerebrate cats. Journal of Neuro-
physiology,80(4), 2023–2037.
Bennett D. J., Li Y., Harvey P. J., & Gorassini M. (2001).
Evidence for plateau potentials in tail motoneurons of
awake chronic spinal rats with spasticity. Journal of Neuro-
physiology,86(4), 1972–1982.
Bennett D. J., Li Y., & Siu M. (2001). Plateau potentials in
sacrocaudal motoneurons of chronic spinal rats, recorded
in vitro. Journal of Neurophysiology,86(4), 1955–1971.
Binder, M. D., Powers, R. K., & Heckman, C. J. (2020). Non-
linear input-output functions of motoneurons. Physiology
(Bethesda, Md.),35, 31–39.
Bos R., Drouillas B., Bouhadfane M., Pecchi E., Trouplin
V., Korogod S. M., & Brocard F. (2021). Trpm5 channels
encode bistability of spinal motoneurons and ensure motor
control of hindlimbs in mice. Nature Communications,
12(1), 6815.
Bos R., Harris-Warrick R. M., Brocard C., Demianenko L.
E.,ManuelM.,ZytnickiD.,KorogodS.M.,&Brocard
F. (2018). Kv1.2 channels promote nonlinear spiking
motoneurons for powering up locomotion. Cell reports,
22(12), 3315–3327.
Bregman B. S. (1987). Development of serotonin immuno-
reactivity in the rat spinal cord and its plasticity after
neonatal spinal cord lesions. Brain Research,34(2), 245–263.
Brix N., Ernst A., Lauridsen L. L. B., Parner E., Støvring H.,
Olsen J., Henriksen T. B., & Ramlau-Hansen C. H. (2019).
Timing of puberty in boys and girls: A population-based
study. Paediatric and Perinatal Epidemiology,33(1), 70–78.
Burke R. E. (1968). Firing patterns of gastrocnemius motor
unitsinthedecerebratecat.The Journal of Physiology,
196(3), 631–654.
Caillet A. H., Avrillon S., Kundu A., Yu T., Phillips A. T. M.,
Modenese L., & Farina D. (2023). Larger and denser: An
optimal design for surface grids of EMG electrodes to
identify greater and more representative samples of motor
units. eNeuro,10(9), ENEURO.0064-23.2023.
Carrascal L., Nieto-Gonzalez J. L., Cameron W. E., Torres B.,
& Nunez-Abades P. A. (2005). Changes during the postnatal
development in physiological and anatomical characteristics
of rat motoneurons studied in vitro. Brain Research Brain
Research Reviews,49(2), 377–387.
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2085
Christova P., & Georgopoulos A. P. (2023). Changes of
cortical gray matter volume during development: A human
connectome project study. Journal of Neurophysiology,
130(1), 117–122.
D’Amico, J. M., Condlie, E. G., Martins, K. J., Bennett,
D. J., & Gorassini, M. A. (2014). Recovery of neuro-
nal and network excitability after spinal cord injury and
implications for spasticity. Frontiers in Integrative Neuro-
science,8, 36.
D’amico J. M., Murray K. C., Li Y., Chan K. M., Finlay M.
G., Bennett D. J., & Gorassini M. A. (2013). Constitutively
active 5-HT2/α1 receptors facilitate muscle spasms after
human spinal cord injury. Journal of Neurophysiology,
109(6), 1473–1484.
Davies B. L., Gehringer J. E., & Kurz M. J. (2015). Age-related
dierences in the motor planning of a lower leg target
matching task. Human Movement Science,44, 299–306.
Deluca C., & Erim Z. (1994). Common drive of motor units
in regulation of muscle force. Trends in Neuroscience (Tins),
17(7), 299–305.
De Luca C. J., & Hostage E. C. (2010). Relationship between
ring rate and recruitment threshold of motoneurons
in voluntary isometric contractions. Journal of Neuro-
physiology,104(2), 1034–1046.
DelVecchioA.,HolobarA.,FallaD.,FeliciF.,EnokaR.
M., & Farina D. (2020). Tutorial: Analysis of motor unit
discharge characteristics from high-density surface EMG
signals. Journal of Electromyography and Kinesiology,53,
102426.
Del Vecchio A., Sylos-Labini F., Mondì V., Paolillo P., Ivanenko
Y., Lacquaniti F., & Farina D. (2020). Spinal motoneurons
of the human newborn are highly synchronized during leg
movements. Science Advances,6(47), eabc3916.
Dotan R., Mitchell C., Cohen R., Klentrou P., Gabriel D.,
& Falk B. (2012). Child-adult dierences in muscle
activation–a review. Pediatric Exercise Science,24(1), 2–21.
Dwyer, J. B., & Bloch, M. H. (2019). Antidepressants for
pediatric patients. Current Psychiatry,18, 26–42F.
El Oussini H., Bayer H., Scekic-Zahirovic J., Vercruysse P.,
Sinniger J., Dirrig-Grosch S., Dieterlé S., Echaniz-Laguna
A., Larmet Y., Müller K., Weishaupt J. H., Thal D. R.,
Van Rheenen W., Van Eijk K., Lawson R., Monassier L.,
Maroteaux L., Roumier A., Wong P. C., Van Den Berg L.
H.,LudolphA.C.,VeldinkJ.H.,WittingA.,&DupuisL.
(2016). Serotonin 2B receptor slows disease progression
and prevents degeneration of spinal cord mononuclear
phagocytes in amyotrophic lateral sclerosis. Acta Neuro-
pathologica,131(3), 465–480.
Erim Z., De Luca C. J., Mineo K., & Aoki T. (1996).
Rank-ordered regulation of motor units. Muscle & Nerve,
19(5), 563–573.
Farina D., & Enoka R. M. (2011). Surface EMG decomposition
requires an appropriate validation. Journal of Neuro-
physiology,105(2), 981–982.
Geertsen S. S., Willerslev-Olsen M., Lorentzen J., & Nielsen
J. B.o (2017). Development and aging of human spinal
cord circuitries. Journal of Neurophysiology,118(2),
1133–1140.
Gilfarb R. A., & Leuner B. (2022). GABA system modications
during periods of hormonal ux across the female lifespan.
Frontiers in Behavioral Neuroscience,16, 802530.
Goodlich B. I., Del Vecchio A., Horan S. A., & Kavanagh J. J.
(2023). Blockade of 5-HT(2) receptors suppress motor unit
ring and estimates of persistent inward currents during
voluntary muscle contraction in humans. The Journal of
Physiology,601(6), 1121–1138.
Gorassini M., Bennett D. J., Kiehn O., Eken T., & Hultborn
H. (1999). Activation patterns of hindlimb motor units in
the awake rat and their relation to motoneuron intrinsic
properties. Journal of Neurophysiology,82(2), 709–717.
Gorassini M., Yang J. F., Siu M., & Bennett D. J. (2002a).
Intrinsic activation of human motoneurons: Possible
contribution to motor unit excitation. Journal of Neuro-
physiology,87(4), 1850–1858.
Gorassini M., Yang J. F., Siu M., & Bennett D. J. (2002b).
Intrinsic activation of human motoneurons: reduction of
motor unit recruitment thresholds by repeated contractions.
Journal of Neurophysiology,87(4), 1859–1866.
Gorassini M. A., Bennett D. J., & Yang J. F. (1998).
Self-sustained ring of human motor units. Neuroscience
Letters,247(1), 13–16.
Gorassini M. A. (2004). Role of motoneurons in the
generation of muscle spasms after spinal cord injury. Brain,
127(10), 2247–2258.
Granit R., Kernell D., & Lamarre Y. (1966). Synaptic
stimulation superimposed on motoneurones ring in the
‘secondary range’ to injected current. The Journal of Physio-
logy,187(2), 401–415.
Gydikov A., & Kosarov D. (1974). Some features of dierent
motor units in human biceps brachii. Pugers Archiv:
European Journal of Physiology,347(1), 75–88.
Harris-Warrick, R. M., Pecchi, E., Drouillas, B., Brocard, F.,
& Bos, R. (2023). A size principle for bistability in mouse
spinal motoneurons. bioRxiv. https://doi.org/10.1101/2023.
09.29.559784.
Harvey P. J., Li X., Li Y., & Bennett D. J. (2006a). 5-HT2
receptor activation facilitates a persistent sodium current
and repetitive ring in spinal motoneurons of rats with
and without chronic spinal cord injury. Journal of Neuro-
physiology,96(3), 1158–1170.
Harvey P. J., Li Y., Li X., & Bennett D. J. (2006b). Persistent
sodium currents and repetitive ring in motoneurons of
the sacrocaudal spinal cord of adult rats. Journal of Neuro-
physiology,96(3), 1141–1157.
HassanA.S.,FajardoM.E.,CummingsM.,McphersonL.M.,
NegroF.,DewaldJ.P.A.,HeckmanC.J.,&PearceyG.E.P.
(2021). Estimates of persistent inward currents are reduced
inupperlimbmotorunitsofolderadults.The Journal of
Physiology,599(21), 4865–4882.
Hounsgaard J., Hultborn H., Jespersen B., & Kiehn O.
(1988). Bistability of alpha-motoneurones in the
decerebrate cat and in the acute spinal cat after intravenous
5-hydroxytryptophan. The Journal of Physiology,405(1),
345–367.
Huh S., Heckman C. J., & Manuel M. (2021). Time course
of alterations in adult spinal motoneuron properties in
the SOD1(G93A) mouse model of ALS. eNeuro,8(2),
ENEURO.0378-20.2021.
Hyngstrom A. S., Johnson M. D., & Heckman C. J. (2008).
Summation of excitatory and inhibitory synaptic inputs
by motoneurons with highly active dendrites. Journal of
Neurophysiology,99(4), 1643–1652.
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

2086 G. Mohammadalinejad and others J Physiol 602.9
HyngstromA.S.,JohnsonM.D.,MillerJ.F.,&HeckmanC.J.
(2007). Intrinsic electrical properties of spinal motoneurons
vary with joint angle. Nature Neuroscience,10(3), 363–369.
Jean-Xavier C., Sharples S. A., Mayr K. A., Lognon A.
P., & Whelan P. J. (2018). Retracing your footsteps:
Developmental insights to spinal network plasticity
following injury. Journal of Neurophysiology,119(2),
521–536.
Jensen D. B., Kadlecova M., Allodi I., & Meehan C. F. (2020).
Spinal motoneurones are intrinsically more responsive in
the adult G93A SOD1 mouse model of amyotrophic lateral
sclerosis. The Journal of Physiology,598(19), 4385–4403.
JesunathadasM.,KlassM.,DuchateauJ.,&EnokaR.M.
(2012). Discharge properties of motor units during steady
isometric contractions performed with the dorsiexor
muscles. Journal of Applied Physiology (1985),112(11),
1897–1905.
Kernell D. (1965). High-frequency repetitive ring of cat
lumbosacral motoneurones stimulated by long-lasting
injected currents. Acta Physiologica Scandinavica,65(1–2),
74–86.
Kiehn O., & Eken T. (1997). Prolonged ring in motor units:
Evidence of plateau potentials in human motoneurons?
Journal of Neurophysiology,78(6), 3061–3068.
Lee R. H., & Heckman C. J. (1998). Bistability in spinal
motoneurons in vivo: Systematic variations in persistent
inward currents. Journal of Neurophysiology,80(2), 583–593.
Lee R. H., & Heckman C. J. (2000). Adjustable amplication
of synaptic input in the dendrites of spinal motoneurons in
vivo. Journal of Neuroscience,20(17), 6734–6740.
Lee R. H., Kuo J. J., Jiang M. C., & Heckman C. J. (2003).
Inuence of active dendritic currents on input-output
processing in spinal motoneurons in vivo. Journal of Neuro-
physiology,89(1), 27–39.
Li X., Murray K., Harvey P. J., Ballou E. W., & Bennett D. J.
(2007). Serotonin facilitates a persistent calcium current in
motoneurons of rats with and without chronic spinal cord
injury. Journal of Neurophysiology,97(2), 1236–1246.
Li Y., Gorassini M. A., & Bennett D. J. (2004). Role of
persistent sodium and calcium currents in motoneuron
ring and spasticity in chronic spinal rats. Journal of Neuro-
physiology,91(2), 767–783.
Mcneil C. J., Doherty T. J., Stashuk D. W., & Rice C. L. (2005).
Motor unit number estimates in the tibialis anterior muscle
of young, old, and very old men. Muscle & Nerve,31(4),
461–467.
MillerJ.D.,SterczalaA.J.,TrevinoM.A.,WrayM.E.,
Dimmick H. L., & Herda T. J. (2019). Motor unit action
potential amplitudes and ring rates during repetitive
muscle actions of the rst dorsal interosseous in children
and adults. European Journal of Applied Physiology,119(4),
1007–1018.
Mizuno N., & Itoh H. (2009). Functions and regulatory
mechanisms of Gq-signaling pathways. Neuro-Signals,
17(1), 42–54.
Moritz C. T., Barry B. K., Pascoe M. A., & Enoka R. M. (2005).
Discharge rate variability inuences the variation in force
uctuations across the working range of a hand muscle.
Journal of Neurophysiology,93(5), 2449–2459.
MurrayK.C.,NakaeA.,StephensM.J.,RankM.,D’amico
J.,HarveyP.J.,LiX.,HarrisR.L.W.,BallouE.W.,Anelli
R., Heckman C. J., Mashimo T., Vavrek R., Sanelli L.,
Gorassini M. A., Bennett D. J., & Fouad K. (2010). Recovery
of motoneuron and locomotor function after spinal
cord injury depends on constitutive activity in 5-HT2C
receptors. Nature Medicine,16(6), 694–700.
Murray K. C., Stephens M. J., Ballou E. W., Heckman C. J., &
Bennett D. J. (2011). Motoneuron excitability and muscle
spasms are regulated by 5-HT2B and 5-HT2C receptor
activity. Journal of Neurophysiology,105(2), 731–748.
Nawab S. H., Chang S.-S., & De Luca C. J. (2010). High-yield
decomposition of surface EMG signals. Clinical Neuro-
physiology,121(10), 1602–1615.
Negro F., Muceli S., Castronovo A. M., Holobar A., & Farina
D. (2016). Multi-channel intramuscular and surface EMG
decomposition by convolutive blind source separation.
JournalofNeuralEngineering,13(2), 026027.
Nezua A., Kimura S., Uehara S., Kobayashia T., Tanaka M., &
Saito K. (1997). Magnetic stimulation of motor cortex in
children: Maturity of corticospinal pathway and problem of
clinical application. Brain & Development,19(3), 176–180.
Okudaira M., Takeda R., Hirono T., Nishikawa T., Kunugi S.,
& Watanabe K. (2023). Motor unit ring properties during
force contrl task and associations with neurological tests in
children. Pediatric Exercise Science,36(1), 23–29.
Oliveira D. S. D.e, Casolo A., Balshaw T. G., Maeo S., Lanza
M.B.,MartinN.R.W.,MaulliN.,KinfeT.M.,Eskoer
B. M., Folland J. P., Farina D., & Del Vecchio A. (2022).
Neural decoding from surface high-density EMG signals:
Inuence of anatomy and synchronization on the number
of identied motor units. JournalofNeuralEngineering,
19(4), 046029.
Orssatto L. B. R., Mesquita R. N. O., & Phillips K. M. (2021).
Looking at the bigger PICture: Understanding and counter-
acting the decline of persistent inward currents in older
adults. The Journal of Physiology,599(23), 5137–5139.
Oya T., Riek S., & Cresswell A. G. (2009). Recruitment and
rate coding organisation for soleus motor units across entire
range of voluntary isometric plantar exions. The Journal of
Physiology,587(19), 4737–4748.
Perrier J.-F., Rasmussen H., Christensen R., & Petersen
A. (2013). Modulation of the intrinsic properties of
motoneurons by serotonin. Current Pharmaceutical Design,
19(24), 4371–4384.
PetersenT.H.,Kliim-DueM.,FarmerS.F.,&NielsenJ.B.o
(2010). Childhood development of common drive to a
human leg muscle during ankle dorsiexion and gait. The
Journal of Physiology,588(22), 4387–4400.
Piotrkiewicz M., Kudina L., Mierzejewska J., Jakubiec M., &
Hausmanowa-Petrusewicz I. (2007). Age-related change in
duration of afterhyperpolarization of human motoneurones.
The Journal of Physiology,585(2), 483–490.
Piotrkiewicz M., & Türker K. S. (2017). Onion skin or
common drive? Frontiers in Cellular Neuroscience,11,2.
Powers R. K., & Heckman C. J. (2017). Synaptic control of
the shape of the motoneuron pool input-output function.
Journal of Neurophysiology,117(3), 1171–1184.
Powers R. K., Nardelli P., & Cope T. C. (2008). Estimation
of the contribution of intrinsic currents to motoneuron
ring based on paired motoneuron discharge records in
the decerebrate cat. Journal of Neurophysiolog y,100(1),
292–303.
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.

J Physiol 602.9 Intrinsic motoneuron properties in typical human development 2087
Quinlan K. A., Schuster J. E., Fu R., Siddique T., & Heckman
C. J. (2011). Altered postnatal maturation of electrical
properties in spinal motoneurons in a mouse model of
amyotrophic lateral sclerosis. The Journal of Physiology,
589(9), 2245–2260.
Revill A. L., Chu N. Y., Ma L.i, Leblancq M. J., Dickson
C. T., & Funk G. D. (2019). Postnatal development of
persistent inward currents in rat XII motoneurons and their
modulation by serotonin, muscarine and noradrenaline.
The Journal of Physiology,597(12), 3183–3201.
Sharples S. A., & Miles G. B. (2021). Maturation of persistent
and hyperpolarization-activated inward currents shapes
the dierential activation of motoneuron subtypes during
postnatal development. Elife,10, e71385.
Skinner J. W., Christou E. A., & Hass C. J. (2019). Lower
extremity muscle strength and force variability in persons
with Parkinson disease. Journal of Neurologic Physical
Therapy,43(1), 56–62.
Smith C. C., & Brownstone R. M. (2020). Spinal motoneuron
ring properties mature from rostral to caudal during post-
natal development of the mouse. The Journal of Physiology,
598(23), 5467–5485.
Svirskis G., & Hounsgaard J. (1997). Depolarization-induced
facilitation of a plateau-generating current in ventral
horn neurons in the turtle spinal cord. Journal of Neuro-
physiology,78(3), 1740–1742.
Taylor C. A., Kopicko B. H., Negro F., & Thompson
C. K. (2022). Sex dierences in the detection
of motor unit action potentials identied
using high-density surface electromyography.
Journal of Electromyography and Kinesiology,65,
102675.
Thompson C. K., & Hornby T. G. (2013). Divergent
modulation of clinical measures of volitional and reexive
motor behaviors following serotonergic medications in
human incomplete spinal cord injury. Journal of Neuro-
trauma,30(6), 498–502.
Vandervoort A. A., & Mccomas A. J. (1986). Contractile
changes in opposing muscles of the human ankle joint with
aging. Journal of applied physiology (1985),61(1), 361–367.
VegetoE.,VillaA.,DellaTorreS.,CrippaV.,RusminiP.,
Cristofani R., Galbiati M., Maggi A., & Poletti A. (2020).
The role of sex and sex hormones in neurodegenerative
diseases. Endocrine Reviews,41(2), 273–319.
Wei K., Glaser J. I., Deng L., Thompson C. K., Stevenson I.
H.,WangQ.,HornbyT.G.,HeckmanC.J.,&KordingK.
P. (2014). Serotonin aects movement gain control in the
spinal cord. Journal of Neuroscience,34(38), 12690–12700.
Willerslev-Olsen M., Andersen J. B., Sinkjaer T., & Nielsen J.
B.o (2014). Sensory feedback to ankle plantar exors is not
exaggerated during gait in spastic hemiplegic children with
cerebral palsy. Journal of Neurophysiology,111(4), 746–754.
Wood C. L., Lane L. C., & Cheetham T. (2019). Puberty:
Normal physiology (brief overview). Best Practice &
Research-Clinical Endocrinology & Metabolism,33(3),
101265.
Xia Y., Chen D., Xia H., Liao Z., Tang W., & Yan Y.i (2017).
Serotonergic projections to lumbar levels and its plasticity
following spinal cord injury. Neuroscience Letters,649,
70–77.
Yeo S. S., Jang S. H.o, & Son S.u M. (2014). The dierent
maturation of the corticospinal tract and corticoreticular
pathway in normal brain development: Diusion tensor
imaging study. Frontiers in Human Neuroscience,8, 573.
Additional information
Data availability statement
Thedatathatsupportthendingsofthisstudyareavailable
from the corresponding author upon reasonable request. The
statistical results are summarized in the Supporting information.
Competing interests
Theauthorsdeclaretheyhavenocompetinginterests.
Author contributions
G.M., B.A., A.Y., F.N., K.A., DJB and MAG conceived and
designed research. G.M., B.A., A.Y., J.D. and MAG performed
experiments. G.M., B.A., A.Y., J.D., J.B., E.B., GEP and MAG
analysed the data. G.M., B.A., Y.A., GEP, K.A., DJB and MAG
interpreted the results of experiments. G.M., J.N., A.Y., GEP
and MAG prepared gures. G.M., GEP and MAG drafted and
revisedthemanuscriptandallauthorseditedthemanuscript.All
authors approved the nal version of the manuscriupt submitted
for pub;ication. All authors agre e to be accountable for all aspects
of the work. The authors conrm that all persons designated as
authors are qualied.
Funding
This work was supported by a Natural Sciences and Engineering
Research Council of Canada Grant 0 5205 to MAG, an NIH
RO1 NS104436 grant to KAQ and MAG, a Canadian Institute
of Health Research Grant PS 180 430 to MAG, and a Women
and Children’s Health Research Grant 3475 to MAG.
Acknowledgements
We thank Dr Kelvin Jones for helping with the rst draft of the
manuscript and for advice on statistics.
Keywords
motor units, onion skin, persistent inward currents, serotonin,
tibialis anterior, torque steadiness
Supporting information
Additional supporting information can be found online in the
Supporting Information section at the end of the HTML view of
the article. Supporting information les available:
Peer Review History
Supplemental Tables 1–10
© 2024 The Authors. The Journal of Physiology published by John Wiley & Sons Ltd on behalf of The Physiological Society.