Article

ISSN 0006-2979, Biochemistry (Moscow), 2025, Vol. 90, No. 11, pp. 1678-1697 © Pleiades Publishing, Ltd., 2025.

Published in Russian in Biokhimiya, 2025, Vol. 90, No. 11, pp. 1794-1815.

1678

Relationships of GDAP1 Mutations to Disease

Phenotype and Mechanisms of Therapeutic Action

of Oxidative Metabolism Activators in a Patient

with Charcot–Marie–Tooth Neuropathy Type2K

Nadejda R. Borisova

, Alina A. Emelyanova

, Olga N. Solovjeva

Natalia V. Balashova

4,5

, Olga P. Sidorova

and Victoria I. Bunik

1,2,3,a

Sechenov University, 119048 Moscow, Russia

Faculty of Bioengineering and Bioinformatics, Lomonosov Moscow State University,

119991, Moscow, Russia

Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University,

119991 Moscow, Russia

Faculty of Advanced Medicine, Vladimirsky Moscow Regional Research and Clinical Institute,

129110 Moscow, Russia

Faculty of Continuing Medical Education, RUDN Medical Institute,

117198 Moscow, Russia

e-mail: bunik@belozersky.msu.ru

Received July 6, 2025

Revised October 5, 2025

Accepted October 18, 2025

Abstract— The development of personalized medicine, including the treatment of hereditary diseases, re-

quires translation of advances in biochemistry into medical practice. Our work is dedicated to solving this

problem in a clinical case of hereditary Charcot–Marie–Tooth neuropathy type 2K (CMT2K), induced by the

compound heterozygous mutations in the GDAP1 gene leading to the protein variants with the most common

in Europe substitution L239F (inherited from the father) and previously uncharacterized substitution A175P

(inherited from the mother). The ganglioside-induced differentiation-associated protein  1 (GDAP1) encoded by

the GDAP1 gene is located in the outer mitochondrial membrane and belongs to the glutathione S-transferase

superfamily. Our structure-function analysis of GDAP1 shows that dimerization of its monomers with either

L239F or A175P substitutions, along with the half-of-the-sites reactivity of GDAP1 to hydrophobic ligands, may

synergistically impair the binding due to the double amino acid substitution in one of the active sites. This

mechanism explains the early disease onset and progress in the child, whose parents heterozygous by each

of the mutations are asymptomatic. Published phenotypes of amino acid substitutions in the GDAP1 region

comprising the binding site for hydrophobic compounds are analyzed, including phenotypes of the homozy-

gous L239F substitution and its compound heterozygous combinations with other substitutions in this region.

Based on the found association of these substitutions with the axonal form of Charcot–Marie–Tooth disease

(CMT) and disturbances in the NAD

- and thiamine diphosphate (ThDP)-dependent mitochondrial metabolism,

the therapeutic effect of nicotinamide riboside (NR) and thiamine (precursors of NAD

and ThDP, respective-

ly) in the patient is studied. Oral administration of thiamine and NR increases levels of ThDP and NAD

the patient’s blood, improves the hand grip strength, and, after a long-term administration, normalizes the

ThDP-dependent metabolism. After the therapy, the diseased-altered activities of transketolase (TKT) and its

apo-form, as well as the relationship between the activity of the TKT holoenzyme and ThDP and NAD

levels

in the patient’s blood, approach those of healthy women. Our results demonstrate the therapeutic potential

of thiamine and NR in correcting metabolic dysregulation in CMT caused by mutations in GDAP1, suggesting

* To whom correspondence should be addressed.

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1679

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

the underlying molecular mechanisms. Genetic diagnostics and biochemical characterization of mechanisms

involved in the pathogenicity of mutations in clinically asymptomatic patients or patients at the early CMT

stages may increase the efficacy of therapy, as it is easier to protect from the accumulating metabolic damage

than to reverse it.

DOI: 10.1134/S0006297925601911

Keywords: GDAP1, pyruvate dehydrogenase kinase, compound heterozygous mutations, muscle strength,

Charcot–Marie–Tooth neuropathy, nicotinamide riboside, thiamine, transketolase, tricarboxylic acid cycle

INTRODUCTION

Deciphering molecular mechanisms of patholo-

gies caused by missense mutations is a challenge in

personalized medicine [1]. Solving this problem is

essential for developing patient-specific therapeutic

interventions. In some cases, such as Charcot–Marie–

Tooth disease (CMT), the pathological effects of mis-

sense mutations can develop asymptomatically over a

long period of ontogenesis [2]. CMT is a very hetero-

geneous group of inherited motor and sensory poly-

neuropathies with a high incidence (1  :  2500) and au-

tosomal dominant, autosomal recessive, and X-linked

inheritance patterns [2]. Based on electrophysiological

criteria, two major forms of CMT are considered, i.e.,

the demyelinating and axonal ones. X-linked subtypes

are traditionally classified as a separate group CMTX.

Early onset and severe progression might be associ-

ated with a mixed-type CMT, which is initially diag-

nosed as the axonal form [3]. In such cases, demyelin-

ation in the course of disease progression presumably

results from continuous mitochondrial dysfunction,

which usually induces axonal damage, representing

a common biochemical feature of various neuromus-

cular diseases [4]. The clinical symptoms common for

all CMT forms include progressive muscle weakness

and muscle atrophy in lower limbs (especially, in

distal regions), which later affect arms and proximal

regions, as well as sensory loss, diminished tendon

reflexes, and impaired nerve impulse generation and

conduction [2].

Mutations in the ganglioside-induced differenti-

ation-associated protein  1 (GDAP1), encoded by the

GDAP1 gene, are the fifth most frequent cause of the

CMT in certain populations, leading to the develop-

ment of its demyelinating, axonal, or mixed forms [2].

GDAP1 is highly expressed in neurons of the periph-

eral nervous system and Schwann cells (neuroglial

cells involved in peripheral myelinogenesis) [5]. The

clinical phenotypes of homozygous and compound

heterozygous GDAP1 mutations are characterized by

early and significant impairments of predominantly

motor functions that worsen with age. Depending on

a particular GDAP1 mutation (the number of known

GDAP1 mutations exceeds 80 [6]), the associated phe-

notypes range from very severe, diagnosed during the

first year of life, to moderate, which develop over 2-3

decades [2, 3].

The effects of GDAP1 knockouts and knockdowns

could vary depending on the experimental model.

For example, disturbances in the mitochondrial mor-

phology and calcium homeostasis detected in the

GDAP1-deficient models correlate with a reduced

number of contact sites between the mitochondrial

and other cellular membranes (plasma membrane and

membranes of the endoplasmic reticulum, lysosomes,

and peroxisomes) [7-9]. Pleiotropic clinical and cellu-

lar phenotypes may be observed even for the same

GDAP1 mutations or their heterozygous combinations

[10-16], suggesting that the cellular and organismal

consequences of GDAP1 dysfunction depend on the

context in which its molecular functions are realized.

Unfortunately, these functions are characterized insuf-

ficiently.

According to the analysis of the amino acid se-

quence of GDAP1, this protein belongs to the gluta-

thione S-transferase (GST) superfamily [17, 18]. Thus,

GDAP1 contains domains homologous to the N- and

C-terminal domains of GST, which include the glu-

tathione-binding site and the hydrophobic-co-sub-

strate-binding site, respectively (Fig. 1). Although the

data on the glutathione binding and catalytic activ-

ity of GDAP1 in the GST reaction are contradictory

[19-21], the ability of GDAP1 to bind hydrophobic com-

pounds, such as etacrynic and hexadecanedioic acids,

is demonstrated in independent studies [19, 20]. Fur-

thermore, in patients with the CMT neuropathy type

2K (CMT2K), missense mutations in the GST domains

of GDAP1 result in a faster disease progression com-

pared to the missense mutations located outside of

these domains, indicating the key role of the GST-

homologous domains of GDAP1 in the CMT2K pathol-

ogy [22].

In addition to the GST-homologous domains,

GDAP1 contains the so-called α-loop within the C-ter-

minal GST domain and two additional domains: the

hydrophobic domain (HD) and the transmembrane

(TM) domain (Fig.  1). The presence of extended α-loop,

BORISOVA et al.1680

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 1. Domain structure of GDAP1: pink, N-terminal GST domain containing glutathione-binding site; gray, C-terminal GST

domain containing hydrophobic-co-substrate-binding site. Compared to the GST structure, the C-terminal GST domain in

GDAP1 contains an insertion named the α-loop, most of which is not visualized in X-ray structures. Following the C-terminal

GST domain, GDAP1 contains the hydrophobic domain (HD, blue) and the transmembrane domain (TM, black). Positions

of mutations in GDAP1 of the studied patient are shown; the boundaries of the GDAP1 structural fragments are numbered

according to [19].

which is largely unresolved in the GDAP1 crystal

structures, is essential for GDAP1 dimerization, bind-

ing of hydrophobic compounds in the C-terminal GST

domain, and formation of heterologous complexes

with β-tubulin, Ras-related protein Rab-6B, and cay-

taxin (proteins involved in the vesicular transport)

[19, 20, 23]. Another characterized property of GDAP1

is its membrane binding mediated by the C-terminal

TM domain, which is important for the formation of

contact sites between the mitochondria and other or-

ganelles, such as endoplasmic reticulum and lyosomes

[9, 24]. Due to the influence of GDAP1 on the mito-

chondrial morphology, suggestions on its GTPase activ-

ity, similar to that of dynamin, have been considered,

but neither the GTPase activity, nor the structural

similarity of GDAP1 to GTPases have been found [19].

The aim of this study is to analyze the molecular

basis of the pathogenicity of GDAP1 mutations, and

of the therapeutic action of metabolic activators, in a

female CMT2K patient with a compound heterozygous

combination of two GDAP1 missense mutations result-

ing in the amino acid substitutions L239F (with pre-

viously characterized clinical phenotype) and A175P

(of unknown pathogenicity). The mutations are in-

herited from apparently healthy heterozygous father

and mother, respectively. L239F is the most frequent

GDAP1 mutation in the Central and Eastern Europe,

whose pathogenicity is manifested in the homozygous

state  [15], while the pathogenicity of A175P and its

combination with L239F remains unknown. To eval-

uate it, we integrate the results of our analysis of the

structure-function relationships in the GDAP1 mole-

cule with data on the clinical and cellular phenotypes

of mutations in the GDAP1 gene. Our findings suggest

a negative impact of compound heterozygous L239F/

A175P substitutions on the binding of hydrophobic

compounds by GDAP1, which functions as a dimer

with a half-of-the-sites reactivity for such compounds.

Analysis of the patient’s clinical phenotype and other

known phenotypes associated with GDAP1 missense

mutations affecting the hydrophobic ligand-binding

region, reveals a common clinical picture of the axo-

nal CMT neuropathy with the onset in the first decade

of life. Based on the studies conducted in cells ob-

tained from patients with GDAP1 mutations altering

the binding region for hydrophobic compounds, we

predict the therapeutic effect of thiamine and nico-

tinamide riboside (NR) as precursors of the metabol-

ic activators thiamine diphosphate (ThDP) and NAD

respectively. Consistent with the molecular mecha-

nisms inferred from our analysis, administration of

thiamine and NR improves and/or stabilizes the hand

grip strength (typically decreased in CMT) and asso-

ciated metabolic parameters, as demonstrated using

a minimally invasive blood test. The use of dietary

supplements therapy, similar to our administration

of metabolic activators, attracts an increasing interest

in the treatment of CMT [25].

MATERIALS AND METHODS

Reagents. Commercial reagents for biochemical

assays were obtained from Helicon, Russia and were

of the highest purity available. A mixture of pentose

phosphates for the transketolase (TKT) activity as-

say was synthesized from ribose 5-phosphate using

ribose 5-phosphate isomerase and xylulose 5-phos-

phate epimerase from bovine spleen acetone pow-

der [26, 27]. Yeast TKT apoenzyme was isolated by

immunoaffinity chromatography with polyclonal an-

tibodies obtained from the serum of immunized rab-

bit according to the published protocol [28,  29]. The

storage of TKT and its use in reactions have been

described previously [30, 31]. Formate dehydrogenase

for the NAD

assay was from Innotech MSU (Moscow,

Russia).

Patient’s clinical characteristics and medical

history at the enrollment. The study was not pre-

registered. Prior to the study, written informed con-

sent for participation in the study and publication

of its results had been obtained from the mother

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1681

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

of the patient (a minor). The patient was diagnosed

with CMT2K (an axonal form of hereditary motor and

sensory CMT neuropathy) at the age of 7, based on

clinical findings and genetic testing, which revealed

two missense mutations in the GDAP1 gene in a com-

pound heterozygous state: Ala175Pro (inherited from

the mother) and Leu239Phe (inherited from the fa-

ther). Both parents considered themselves healthy,

while the absence of native GDAP1 protein in the

child led to the early disease manifestation.

At the time of enrollment into the study at the

age of 16, the patient presented with complaints of

weakness in the legs and arms, slow gait, and in-

ability to climb the stairs without support, rise from

a squatting position, and stand on tiptoes or heels.

Physical examination revealed muscle atrophy in the

feet, lower legs, and hands (with interphalangeal joint

contractures), as well as joint hypermobility with the

Beighton score of 5. Hand grip strength was severely

reduced to 5.1  kg in the right hand and 7.5 kg in the

left hand, compared to the normal grip strength of

~26  kg for girls of the same age [32-34]. Tendon and

periosteal reflexes in the upper extremities were mod-

erately active; knee and ankle reflexes were absent.

The patient experienced painful hypesthesia in hands

and feet and maintained the Romberg position with

difficulty due to weak legs, but accurately performed

the finger-to-nose test. Vibration sensitivity in the toes

was preserved.

Electroneuromyography at the time of enrollment

showed a significant reduction in the amplitude of

the compound muscle action potential and a less pro-

nounced decrease in the conduction velocity of effer-

ent (motor) fibers, particularly in the nerves of the

lower limbs. The conduction velocity of afferent (sen-

sory) fibers was at the lower limit of normal. The con-

duction velocity in the median nerve was 47-50  m/sec

in both arms, which is above the threshold of 38  m/sec

for diagnosing the demyelinating form of CMT, thus

supporting the classification of this case as the axonal

type [33]. However, ultrasound examination of nerves

in the upper and lower limb revealed signs of altered

nerve structure, suggesting demyelination.

Analysis of pathogenicity of the patient’s mu-

tations. For analysis, a structure of the GDAP1 dimer

complex with a hydrophobic ligand (PDB  ID: 7AIA

[20]) was selected. To model the 3D structures of the

α-loop, AlphaFold [35] as a general approach and PEP-

Fold as a specialized approach to predict the spatial

structures of short peptides up to 50 amino acid res-

idues in length [36], were used. All structures were

visualized with Pymol [37].

The 3D structures of proteins were shown as rib-

bon diagrams, and amino acid substitutions resulting

from missense mutations were represented with ball-

and-stick models. The regions present in both mod-

eled and X-ray structures were superimposed using

the Pymol align method [37]. The different conforma-

tions of the crystallographic structures of the dimer

subunits in this region caused the differences in the

superimposed regions and RMSD values of the su-

perposition. For the subunit with the α-loop directed

towards the hexadecanedioic acid molecule present

in the crystal, these were the superimposed regions

T147-G165 (N-terminal) and H199-D210 (C-terminal),

with RMSD 4.578Å. Similar superimposed regions for

the α-loop of the other subunit of the dimer were

T147-S162 and A184-N201, RMSD 6.162  Å.

Assessment of the effects of thiamine and NR

on the clinical parameters. Thiamine (as thiamine

hydrochloride) and NR were administered orally be-

tween the prescribed courses of neurometabolic ther-

apy, which the patient received 2-3 times per year.

These courses lasted 1.5 month each and included

α-lipoic acid (Berlition 600  mg, intravenous infusion,

or Thiogamma 600  mg, tablets), vitamin  B12 (cyano-

cobalamin, 0.5  mg, intramuscular injection, or meth-

ylcobalamin, 1  mg, tablets), anticholinesterase agent

ipidacrine (Axamon 5  mg, intramuscular injection or

Axamon 20  mg, tablets), combined enzyme prepara-

tions (Wobenzym, 8 tablets per day), omega-3 poly-

unsaturated fatty acids, lecithin, and potassium iodide

(Iodomarin, tablets).

During the course of thiamine and NR adminis-

tration, the strength of the finger flexor muscles was

measured regularly as a highly reliable and valid

indicator of the CMT progression in adult patients

[38, 39]. Standardized hand dynamometry with a

DK-25 hand dynamometer (Russia) was performed as

previously described [40].

Biochemical assessment of thiamine and NR ef-

fects. Whole blood samples were collected from the

median cubital vein in the morning after an over-

night fasting into heparin-containing vacuum tubes.

The samples were frozen at −20°C and transported on

ice to the laboratory, where the samples were stored

at −70°C before being used in the experiments.

TKT activity in sonicated whole blood was

measured spectrophotometrically at 340 nm from

the absorbance of NADH produced in a coupled re-

action using a CLARIOstar Plus microplate reader

(Helicon, Russia) as described in [40]. The activi-

ty of TKT in the absence of ThDP corresponded to

the activity of endogenous holo-TKT, which contains

tightly bound, non-dissociating ThDP. The activity

determined in the presence of ThDP was taken as

the total TKT activity. The activity of endogenous

apo-TKT lacking ThDP was determined as the dif-

ference between the total TKT activity and endoge-

nous holo-TKT activity. Accordingly, the activation of

TKT by ThDP added to the medium was calculated

as [1  −  (endogenous holo-TKT/total TKT)]  ×  100%.

BORISOVA et al.1682

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

ThDP content was measured in the extracts ob-

tained from sonicated blood using a CLARIOstar Plus

microplate reader in the spectrophotometric mode

as described in [40]. The concentration of ThDP was

determined enzymatically from the ThDP activation

of yeast apo-TKT.

Blood NAD

content was determined in the

methanol-acetic acid extracts prepared as described

in [41] with modifications. After thawing, the blood

was thoroughly mixed, and 30-µl aliquots were trans-

ferred to new microtubes. Next, 240 µl of ice-cold

methanol was added to each tube. The blood-meth-

anol mixture was sonicated with a Bioruptor (Diag-

enode; Liège, Belgium) in 3 cycles of 30  s each at the

maximum power using ice cooling. After sonication,

40.5  µl of 2% acetic acid was added to each tube, and

the samples were mixed on ice for 30  min in a New

Brunswick Excella E24R incubator (Eppendorf, Rus-

sia) at 180  rpm. To remove denatured proteins, the

extracts were centrifuged for 20  min at 21,500g at

4°C in a Hitachi CT15RE centrifuge (Helicon, Russia).

The supernatants were transferred to clean tubes.

The concentration of NAD

was measured enzymat-

ically on the day of extract preparation based on the

fluorescence of NADH formed in the reaction with

formate dehydrogenase using a CLARIOstar Plus mi-

croplate reader according to the previously described

protocol [41].

Female CMT patients and healthy women used

in the study. For the correlation analysis, female par-

ticipants (Table  S1 in the Online Resource  1) were se-

lected from a previously established database [40].

Parameters from young women of the age closest to

that of the patient, were used as reference values.

Statistical analysis was performed using Graph-

Pad Prism 8.0 (GraphPad Software, USA). The normal

distribution of data was confirmed by the Shapiro–

Wilk and Kolmogorov–Smirnov tests, and the correla-

tion analysis for the cohorts was performed using the

Pearson correlation coefficient. Multiple comparisons

between the baseline values and NAD

content after

thiamine and NR administration were performed us-

ing one-way ANOVA with the Tukey’s post hoc test.

RESULTS

Prediction of effects of A175P and L239F sub-

stitutions on the structure and function of GDAP1.

Deciphering molecular mechanisms of hereditary dis-

eases caused by amino acid substitutions requires

understanding the structure-function relationships in

the protein molecule. Unfortunately, for many pro-

teins, including GDAP1, such relationships are poorly

characterized due to insufficient information on their

molecular functions. This, in turn, is linked to diffi-

culties in obtaining the full-length GDAP1 protein in a

functionally active state regulated by natural mecha-

nisms. Furthermore, the properties of various GDAP1

constructs studied in independent works differ signifi-

cantly, depending on the structural parts removed to

obtain a soluble protein and on the used expression

system [19-21]. Nevertheless, the binding of hydropho-

bic compounds is a molecular function of GDAP1, es-

tablished in independent studies using functional tests

and crystallization. Essential structural prerequisites

for the binding of hydrophobic ligands and strength

of this binding are: the GDAP1 domain homologous

to the C-terminal domain of GST, the GDAP1-specific

α-loop within the C-terminal domain of GST, and the

α-loop-facilitated dimerization of GDAP1 [19, 20]. Both

patient’s heterozygous GDAP1 mutations cause amino

acid substitutions in these structural elements (Figs.  1

and 2a). Moreover, the substituted amino acids are

strictly conserved in animals, from fish and amphibi-

ans to humans (Fig.S1 in the Online Resource1), tes-

tifying to the functional importance of these residues.

In the structures of GDAP1 constructs resolved

by X-ray analysis [monomer (PDB ID:6UIH) and di-

mers (PDB ID: 7AIA, 7YWD and 8EXZ) of the T157P

mutant], L239 is identified in a loop region. The con-

formational mobility of such regions, as well as the

location of the L239-containing loop in the vicinity of

the hexadecanedioic-acid-binding site (Fig. 2a), sup-

port a contribution of L239 to the formation of the

hydrophobic-ligand-binding site of GDAP1. The substi-

tution of L239 with a bulkier phenylalanine residue

as a result of mutation may impair such binding by

altering both the binding pocket geometry and the

hydrophobic interactions with surrounding residues,

in particular, with neighboring C240, which plays an

important role in maintaining GDAP1 stability [17].

In contrast to L239, A175 is located in a part of

the α-loop that is not resolved in any of the published

crystal structures of GDAP1, indicating multiple con-

formations and a high mobility of this region. More-

over, the visualized parts of α-loops in the subunits

of the GDAP1 dimer (PDBID: 7AIA) differ, depending

on the hydrophobic ligand binding. That is, in the

GDAP1dimer presented in Fig.2a, the unresolved part

of the α-loop of the upper subunit with the bound

hexadecanedioic acid, comprises 22 residues (resi-

dues up to 162 and from 184 are visible), whereas

in the ligand-free lower subunit, the unresolved part

comprises 34 residues (residues up to 165 and from

199 are visible). The distances from the H′199 resi-

due of the lower subunit to M116 (35  Å) and L239

(39  Å) in the ligand-bound site of the upper subunit

are smaller than the distances from H199 of the up-

per subunit to the same residues in the lower sub-

unit, i.e., M’116 (38  Å) and L’239 (46  Å). Remarkably,

the GDAP1 dimer (PDB ID: 7AIA) with structurally

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1683

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 2. Location of GDAP1 substitutions L239F and A175P in the hydrophobic-ligand-binding region (PDB ID: 7AIA). The

terminal residues of the α-loop visualized in the dimeric X-ray structure are colored green for the upper subunit and

light blue for the lower subunit (also labeled with a prime). a) Location of the GDAP1 α-loop and A175 and L239 residues

substituted in the patient (framed), relative to the hexadecanedioic-acid-binding site in the GDAP1 dimer (PDB ID: 7AIA).

b) Modeling of the GDAP1 α-loop structure and its flanking regions (residues 140-210) using AlphaFold. c)Modeling of the

GDAP1 α-loop structure (residues 150-199) using PEP-Fold. Panels b and c show changes in the in silico modeled α-loop

conformation caused by the A175P substitution (on the right) vs. the native peptide structure (on the left). The color code

for the structural parts of GDAP1 is the same as in Fig. 1; in panels b and c, residue 175 (which is substituted in the pa-

tient’s GDAP1) is shown in orange.

different α-loops binds the hydrophobic ligand in only

one of the subunits, pointing to the half-of-the-sites

reactivity of GDAP1 toward the hydrophobic ligand,

which is associated with different conformations of

its α-loops. The latter is confirmed by independent

data on the significance of α-loop for the binding of

hydrophobic compounds [19, 20]. These findings are

consistent with the structure of the GDAP1 dimer pre-

sented in Fig.2a, which implies dimer stabilization not

only through the dimerization interface represented

by the N-terminal GST domain (shown in pink), but

also through the α-loop participation in the formation

of the binding site for hydrophobic compounds, con-

tributed by different subunits of the dimer. The high

mobility allows one of the α-loops (α-loop of the low-

er subunit in Fig.  2a, whose unresolved structure is

shown with yellow dotted line) to “close” the hydro-

phobic-ligand-binding site formed by the C-terminal

GST domain of the other subunit (Fig.  2a). The length

of the α-loop with (44  Å, Fig.  2b) or without (30  Å,

Fig.  2c) its flanking regions, is comparable to the dis-

tance between H′199 at the bottom of the α-loop and

L239 located near the hexadecanedioic acid (39  Å).

Thus, the proposed involvement of the α-loop of one

of the GDAP1 subunits in blocking the hydrophobic-

compound-binding site of the other subunit, inferred

from the GDAP1 dimeric structure (PDB ID: 7AIA;

Fig.  2a), is confirmed both by the known role of the

α-loop in enhancing the intersubunit contacts upon

the hydrophobic ligand binding [19, 20] and by the

BORISOVA et al.1684

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

sensitivity of the α-loop conformation (residues 145-

200, Fig.  1) to the binding of hydrophobic ligands

(Fig.  2). The structure of the GDAP1 dimer presented

in Fig.  2a suggests that the higher flexibility of the

lower subunit α-loop (37 residues not visualized in

the structure) compared to the more structured α-loop

of the upper subunit (22 residues not visualized in

the structure) is a manifestation of multiple α-loop

conformations during its rearrangements required for

closing the hexadecanedioic-acid-binding site, and a

higher α-loop spiralization in the open conformation

of the binding site.

Analysis of the GDAP1 X-ray structure (PDD ID:

7AIA; Fig.  2a) also reveals the proximity of the mobile

α-loop (shown in yellow) to residues of the hydropho-

bic domain (shown in blue), which may ensure reg-

ulation of the α-loop conformation through its inter-

action with the hydrophobic domain participating in

the autoinhibition of the GDAP1 GST activity [21]. The

existence of structural premises for the α-loop interac-

tion with the hydrophobic substrate required for the

GST reaction, on one hand, and regulatory hydropho-

bic domain, on the other hand (Fig. 2a), is consistent

with the results of independent functional studies on

the α-loop involvement in the binding of hydrophobic

ligands [19, 20] and autoinhibition of GST activity by

the hydrophobic domain of GDAP1 [21].

The A175P substitution should significantly alter

the spectrum of loop conformations due to (i) the

greater stability of the cis-conformation of the peptide

bond preceding the proline residue compared to the

bonds formed by other amino acids, and (ii) proxim-

ity of the proline residues P175 (mutant) and P179.

The differences in the conformations of the A175- and

P175-containing loops are also confirmed by modeling

the structures of GDAP1 peptides containing the native

and substituted α-loops. Regardless of the program

used and loop-flanking residues, the A175P substitu-

tion changes the peptide fold (Fig. 2, b, c). Although

only the most stable structures with the trans-confor-

mation of peptide bonds are modeled by default, in

native proteins, peptide bonds between prolines and

their preceding residues often have the cis-conforma-

tion [42]. Conformational changes caused by the cis-

trans isomerization of such bonds are widely used in

biological systems [43]. This factor may additionally

influence the α-loop conformation induced by the

A175P substitution and, accordingly, position of the

loop relative to the hydrophobic-ligand-binding site.

Thus, the A175P substitution may alter the con-

formational spectrum of the α-loop involved in di-

merization and binding hydrophobic compound in

one subunit, while the L239F substitution may affect

the same binding pocket for hydrophobic compound

in its part formed by the other subunit (Fig. 2). As a

result of the combination of the monomers with either

L239F or A175P substitutions into the GDAP1 dimer,

one of the binding sites of the dimer will get the dou-

ble substitution. Hence, a greater dysfunction of this

double-substituted binding site is expected, compared

to the wild type and to the binding sites with the

single substitutions L239F or A175P. Moreover, given

the half-of-the-sites reactivity of GDAP1 towards hexa-

decanedioic acid considered above (Fig.2a), impaired

binding due to the amino acid substitutions in one

binding site of the dimer will negatively affect the li-

gand binding in the other of the interacting sites. This

mechanism can explain the dysfunction of GDAP1 in

the patient with two heterozygous mutations in the

absence of noticeable defects in the parents carrying

single substitutions in GDAP1.

Common features of GDAP1 phenotypes in-

duced by amino acid substitutions in the hydro-

phobic-compound-binding site. The structure–func-

tion analysis of GDAP1 conducted above shows that

the A175P and L239F substitutions, which are not

pathogenic in the heterozygous state (as in the par-

ents at the time of study), may cause an early onset

of pathology in the child due to a synergistic impair-

ment in the GDAP1 binding of hydrophobic ligands.

To get further evidence for this mechanism, available

clinical and cellular phenotypes of GDAP1 mutations

leading to substitutions in the protein regions within

or near the hexadecanedioic-acid-binding site, have

been analyzed (Table  1, Fig.  3). Because of the com-

pound heterozygous state of GDAP1 mutations in the

patient, Table  1 also includes the known phenotypes

of compound heterozygotes of the widespread L239F

substitution located in the studied protein region.

As can be seen from Table  1, all presented sub-

stitutions lead to the axonal type of CMT, i.e., affect

generation of the nerve signal rather than the con-

duction velocity. The L239F substitution in the het-

erozygous state produces no clinical phenotype, while

phenotypes resulting from the homozygous and com-

pound heterozygous states of L239F are mainly char-

acterized as moderate, with the onset during the first

decade of life [15, 44-47]. The facts that the pathology

caused by the compound heterozygous state of L239F

was exacerbated to the level typical of homozygous

replacements of residues in the region binding hydro-

phobic compounds, confirms the assumption about

the critical role of hydrophobic ligand binding and

subunit interactions in the dimer for the physiological

function of GDAP1.

Given the progressive nature of CMT, its earlier

onset maybe considered as a sign of a more severe

phenotype. This occurs in patients carrying L239F in

combination with mutations leading to the loss of ar-

ginine residues either near the hydrophobic- ligand-

binding site (R282C) or in the region where α-loop

interacts with the hydrophobic domain (R273G).

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1685

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 3. Location of the GDAP1 amino acid residues substituted due to the mutations considered in Table 1, in the 3D

structure of the GDAP1 dimer in complex with hexadecanedioic acid (PDB ID: 7AIA). The discussed residues are shown

as ball-and-stick models in both subunits, but labeled only in the area of one hexadecanedioic-acid-binding site for better

visualization. Residues of the lower subunit are marked with prime next to the residue abbreviation; residues substituted

in the patient’s GDAP1 are framed. The boundaries of the α-loop regions not visualized in the X-ray structure are indicated

in light blue for the lower subunit (G165′ and H199′) and green for the upper subunit (S162 and A184). α-Loops modeled

by AlphaFold (Fig.2b) are superimposed by the Pymol align method, using the regions present in both X-ray and modeled

structures. After that, the regions of the modelled α-loop, which correspond to those visualized in the X-ray structure of

the dimer, are trimmed off.

In these cases, the disease onset shifts from the first

decade to the first years of life (Table  1, Fig.  3). More-

over, the double substitution L239F/R273G, which po-

tentially affects the regulatory interactions with the

hydrophobic domain (Fig. 3), is associated with spe-

cific features of the clinical phenotype, such as hand

tremor and hoarseness (Table 1).

According to ClinVar, the A175P substitution in

GDAP1 is identified in germline cells, but neither

its clinical phenotype is described, nor its pathoge-

nicity and inheritance type are established (https://

www.ncbi.nlm.nih.gov/clinvar/variation/657278/). The

absence of pronounced symptoms of CMT in the het-

erozygous carrier of this mutation (patient’s mother)

indicates the absence of toxic action or significant

dysfunction of GDAP1A175P, if the normal protein is

synthesized from the other chromosome. The same

applies to the synthesis of GDAP1 and GDAP1 L239F

in patient’s father. In contrast, the early onset of clin-

ical symptoms in the compound heterozygous patient

with both A175P and L239F substitutions in GDAP1

means strongly perturbed GDAP1 function, similar

to that of the L239F homozygous state (Table 1).

The similarity in the pathophysiological action of

the A175P and L239F substitutions agrees with the

results of our structure-function analysis, suggesting

that both mutations affect the same molecular func-

tion of GDAP1, namely, the binding of hydrophobic

compounds (Figs. 2 and 3).

Thus, amino acid substitutions in the hydropho-

bic-compound-binding site of GDAP1 lead to a com-

mon clinical phenotype of CMT, namely, its axonal

form with the onset during the first decade of life.

A combination of L239F with the substitutions affect-

ing positive charges in the binding site or region ad-

jacent to the regulatory hydrophobic domain, shifts

the disease onset to the first year of life and may

present additional clinical symptoms, as in the case

of L239F/R273G substitutions (Table 1).

The most relevant models for elucidating the

mechanisms underlying pathologies caused by pro-

tein variants are cells obtained from the patients

BORISOVA et al.1686

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Table 1. Phenotypes of GDAP1 missense mutations within or near the hydrophobic-compound-binding site

Mutation Clinical phenotype Cellular models References

A175P No symptoms of CMT

in an adult carrier of mutation

in the heterozygous state

No data This work

A175P/

L239F

Axonal type

The disease begins at age 7. By age 16,

weakness in the legs is accompanied

bysevere weakness in the arms

and painful hypesthesia in hands and feet

No data This work

P231L Axonal type in homozygotes

The disease onset in homozygotes

is during the first years of life, with

gradual progression. With age, the upper

extremities become affected, and sensory

impairment in the distal extremities

develops [10, 11, 13]. Vocal cord paralysis

and dysphonia may also occur [12]

No data [10-13]

L239F Axonal type in homozygotes

The disease onset in homozygotes occurs

during the first decade of life, primarily

affecting the legs. Later, weakness and

atrophy in the arms and mild sensory

disturbances are observed. By the age

of 20-30, some patients are forced

to use a wheelchair [15, 45, 68].

Heterozygotes are asymptomatic

Expression of L239F GDAP1 in HeLa

cells with low GDAP1 expression affects

morphology of the Golgi apparatus

and reduces the level of endogenous

GDAP1expression [69].

Structural data suggests decreased

stability of GDAP1with this

substitution [17]

[15, 17, 45,

68, 69]

L239F/

M116T

Axonal type

Onset at 5-10 years of age [15]

M116R GDAP1 does not differ from the

wild-type protein in its ability to protect

against glutamate toxicity in HT22R

cells [16] or to induce mitochondrial

fragmentation in COS7 cells [5]

[5, 15, 16]

L239F/

N227D

Axonal type

Onset at 4-10 years of age [15]

No data [15]

L239F/

R273G

Axonal type

Onset at 2 years of age.

Hand tremors, hoarseness [15]

Comparison of motor neurons

obtained from pluripotent stem cells

of healthy control and a patient with

compound heterozygous L239F/R273G

mutation shows that the substitutions

in GDAP1 decreased mitochondrial Ca

increased PDH phosphorylation, increased

glutamate dehydrogenase activity,

and decreased glutamate level [14].

In the patient’s fibroblasts, expression

of GDAP1 mRNA and protein is

significantly reduced, the glutathione

level is reduced by 40%, and the

mitochondrial potential decreases by 30%,

compared to the control fibroblasts [16]

[14-16]

L239F/

R282C

Axonal type

Onset at 2-3 years of age [15]

No data [15]

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1687

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Table 1 (cont.)

Mutation Clinical phenotype Cellular models References

C240Y Axonal type in heterozygotes

The ages of three studied patients,

all carrying mutation in the heterozygous

state, ranged from 48 to 78 years.

Sensory nerve potentials were not

recorded, while motor nerve potentials

and conduction velocity were normal.

Reflexes in the lower extremities

were absent, and tactile hypesthesia

of the sock-glove type was observed.

Chronic denervation in the upper

extremities was present without

histological abnormalities in muscle

biopsy specimens. In contrast

to the predominant motor impairment

of the distal segments’ characteristic

of CMT, weakness of the proximal

segments of the legs was observed;

foot deformity characteristic of CMT

was absent [22, 48]

The lactate to pyruvate ratio

in the culture medium of fibroblasts

from patients heterozygous

for the C240Y mutation in GDAP1

increases to 12-22, while the control

values are 4-12. In the patients’

fibroblasts, the rate of mitochondrial

respiration and oxidative phosphorylation

on pyruvate/malate, activity of complexI

of the electron respiratory chain,

and activity of aconitase are reduced

by 40-50%. Expression and composition

of complex I, activities of citrate synthase

and fumarase, and the ratio of isocitrate

dehydrogenase to citrate synthase

are not significantly changed by the

substitution. The level of reactive oxygen

species is increased; the expression

of sirtuin 1 is reduced, while the

expression of sirtuin 3 remains

unchanged in cells with the C240Y

substitution (vs. C240 in wild-type

GDAP1). The size and mass of tubular

mitochondria are increased by 33

and 20%, respectively [22, 48]

[22, 48]

Note. Mutations identified in the patient are shown in bold italic. Common symptoms of the developed CMT typically in-

clude weakness, decreased tendon reflexes, and muscle atrophy (predominantly in the lower limbs), deformity of the lower

extremities (bilateral hollow foot), and gait impairment.

possessing these protein variants, since in such cells

the native and/or substituted proteins function with-

in the metabolic networks, which are not externally

manipulated. Due to the limited availability of such

cells for scientific research, in particular, because

of the rare occurrence of specific GDAP1 mutations,

inference from available published data is particu-

larly important for the development of personalized

medicine. The results of our analysis presented in

Table  1 show that despite different sets of indicators

used in independent studies, a common feature of

patients’ cells expressing GDAP1 with substitutions

affecting the hydrophobic-compound-binding site, is

altered mitochondrial metabolism. In particular, the

oxidation of pyruvate in the tricarboxylic acid (TCA)

cycle is changed. That is, motor neurons differenti-

ated from pluripotent stem cells of a patient with

the compound heterozygous L239F/R273G mutations

demonstrate a number of interrelated changes, such

as decreased mitochondrial Ca

, increased phos-

phorylation of α-subunit of pyruvate dehydrogenase

(PDHA), increased level of glutamate dehydrogenase

and decreased glutamate content [14]. Patient’s fi-

broblasts with the compound heterozygous L239F/

R273G mutation show decreased mitochondrial po-

tential and reduced glutathione level, along with a

significant downregulation of the endogenous GDAP1

expression [16]. Cultured fibroblasts from a patient

with the substitution of a residue next to L239, i.e.,

C240Y, exhibit an increased lactate/pyruvate ratio in

the culture medium, 40% reduction in the mitochon-

drial respiration on pyruvate/malate, 40% reduction

in the mitochondrial complex I activity, and 50% de-

crease in oxidative phosphorylation [22, 48]. In con-

trast to the C240Y substitution, the R120W substitu-

tion outside the hydrophobic-compound-binding site

causes no significant increase in the lactate/pyruvate

ratio [22]. According to an independent classification

[49], the lactate/pyruvate ratio <25 observed in the

studied cells (Table1) usually indicates deficient PDH

function, whereas the lactate/pyruvate ratio >25 indi-

cates insufficient activities of pyruvate carboxylase,

TCA cycle, and respiratory chain.

Thus, the studies in cells from patients with

mutations leading to the amino acid substitutions in

the hydrophobic-compound-binding region of GDAP1

show that such substitutions are associated with in-

creased PDH phosphorylation and reduced mitochon-

drial oxidation of pyruvate and NADH. ThDP is an

inhibitor of PDH phosphorylation and, simultaneously,

BORISOVA et al.1688

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 4. The thiamine status of the CMT2K patient compared to young healthy women and young female patients with other

forms of CMT (CMT1A and CMTX1). All parameters are presented as mean ±standard error of mean (SEM).

a coenzyme activating dephosphorylated PDH [50,

51]. Hence, the ThDP precursor thiamine may have

a therapeutic effect in the studied patient having the

hydrophobic-compound-binding region of GDAP1 im-

paired by the L239F/A175P substitutions. On the other

hand, association of the GDAP1 substitutions in the

hydrophobic-compound-binding region with the re-

duced PDH function indicates that the substrate flux

through the TCA cycle is decreased in the patient.

In a number of pathologies, an NAD

precursor NR

not only activates the TCA cycle, but also improves

the exercise tolerance [52-54]. Based on these obser-

vations, the ability of NR along with thiamine to cor-

rect the metabolic dysfunction in the patient with the

L239F/A175P substitutions has been tested.

Patient’s thiamine and NAD

status before the

therapy. The assessment of the thiamine status of an

individual is based on the determination of blood lev-

els of the major intracellular form of thiamine, the

coenzyme ThDP, as well as the activity and coenzyme

saturation of the ThDP-dependent enzyme TKT [40].

Based on these parameters, we found no thiamine in-

sufficiency in the patient’s blood compared to other

young women, either without or with CMT (Fig.  4).

Before the thiamine administration, the ThDP level

in the patient’s blood (200  nM) corresponds to the

upper limit in the control group. In contrast to ThDP,

the TKT activity in the patient’s blood is at the lower

limit of the normal range, i.e., approximately two-fold

lower than the average TKT activity in other young

women, either without or with CMT (Fig.  4). Further-

more, no ThDP activation of TKT in the patient’s blood

is observed upon the ThDP addition to the reaction

medim, while in the control group, TKT is activated

by ThDP (Fig. 4). A relatively low TKT activity and

the absence of ThDP-induced TKT activation in the

patient’s blood are the features previously established

in the CMT patients [40]. In particular, both the TKT

activity in the blood and its activation by ThDP are

decreased at the disease advanced stages [40], which

in the case of GDAP1 mutations are observed in the

first decades of life. As a result, the thiamine status

of the studied patient is consistent with the published

results, indicating the absence of thiamine deficien-

cy in CMT patients and the marker role of the blood

TKT activity and TKT regulation by ThDP in the CMT

pathology [40].

The concentration of NAD

in the patient’s blood

(6-7  µM) is significantly lower than in the control (15-

23  µM) and neurological group (13-16  µM) [41]. Such

low NAD

level may be associated with a much earlier

disease onset in the case of GDAP1 mutations, sug-

gesting a rapid development of pathology, compared

to the previously studied patients with other forms

of CMT starting in the 4th-5th decades of life. Indeed,

the blood NAD

content shows a trend to decrease

in cardiology patients with stage 3 of heart failure,

compared to those with stage 2 [41].

Effects of mitochondrial metabolism activators

on the hand grip strength, ThDP and NAD

levels,

and TKT activity in patient’s blood. The therapeu-

tic effects of thiamine and NR were studied between

the courses of neurometabolic therapy, alongside dai-

ly administration of L-carnitine at a dose of 500  mg.

Thepatient received thiamine hydrochloride orally at

a daily dose of 100  mg for the first 2 months, followed

by 100  mg every other day for the next 4  months.

Starting from the 3rd month, NR was added at a dai-

ly dose of 100 mg and from the 6th month – at the

same dose every other day.

Daily oral administration of thiamine hydrochlo-

ride at a dose of 100 mg increases the ThDP blood

level, while reducing the dose to 100  mg every other

day decreases it (Fig.  5a). Similar effects are observed

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1689

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 5. Effect of oral administration of thiamine and NR on the blood levels of ThDP, NAD

, TKT and on the hand grip

strength. a)Changes in the hand grip strength and the blood ThDP and NAD

concentrations. b)Changes in the blood TKT

activity and ThDP content. c)Changes in the hand grip strength and blood TKT activity and NAD

content. All biochemical

parameters are presented as mean ±SEM (error bars are not seen if within the symbol size).

for the hand grip strength: daily thiamine admin-

istration increases it, while administration of the

same dose every other day leads to a gradual de-

crease in the grip strength compared to the maxi-

mum value observed during the daily administra-

tion (Fig. 5a).

The dynamics of changes in the blood ThDP

levels upon thiamine administration (Fig.  5a) has a

non-monotonous pattern and demonstrates a delayed

effect of thiamine administration on the hand grip

strength, compared to changes in the blood ThDP

concentration. That is, the peak of the blood ThDP

concentration is observed after one month of daily

thiamine intake. After that, the level of ThDP de-

creases even with the same vitamin  B1 dosage, and

remains stable after the dose has been reduced from

100  mg per day to 100  mg every two days. Through-

out this period, the hand grip strength of both right

and left hands gradually increases. Compared to the

blood ThDP levels, the changes in the grip strength

are more monotonous, which is likely related to the

buffering capacity of tissues for thiamine and in-

crease in this capacity in response to a higher thi-

amine availability, for instance, due to the altered

tissue-specific expression of proteins of thiamine me-

tabolism. Presumably, the blood ThDP concentration

is influenced not only by adaptations of blood cells to

a high thiamine availability, which stabilizes its level

and resulting blood metabolism, but also by the over-

all saturation of tissues with thiamine delivered by

the blood. The influence of these factors is consistent

with the observed non-monotonous changes in the

BORISOVA et al.1690

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

ThDP content and metabolism in the blood, as well

as by the delayed effect of thiamine on the hand grip

strength, compared to the blood ThDP levels (Fig.  5a).

The regulatory significance of thiamine for cen-

tral metabolism is evident from changes in the blood

levels of another coenzyme – NAD

. Even with the

thiamine administration alone, i.e., without addition-

al supplementation with the NAD

precursor NR, the

content of NAD

increases simultaneously with the

increase in the ThDP levels (Fig.  5a). After one and

two months of daily thiamine intake, the blood levels

of NAD

become significantly higher (p  <  0.04) than

those observed initially. Reducing the thiamine dose

by switching from a daily regimen to every other day

administration decreases the NAD

levels, but not the

ThDP content (Fig.  5a). However, the following admin-

istration of the biosynthetic NAD

precursor NR along-

side thiamine changes the positive relation between

the ThDP and NAD

levels observed upon taking thi-

amine alone, to a negative one. Specifically, combined

administration of thiamine (100  mg every other day)

and NR (100mg daily) expectedly increases the NAD

level compared to the initial one (p  <  0.03). However,

under these conditions, the increase in the NAD

con-

tent is accompanied by the decrease in the ThDP level

(Fig.5). Reducing the dose of NR to 100  mg every oth-

er day while maintaining the same thiamine dosage,

stabilizes the NAD

level and increases the level of

ThDP (Fig. 5). Both thiamine and NR are imported by

the cells via specific transporters [55,  56]. However,

at high doses, the uptake of these organic cations or

their metabolites could also involve other proteins,

such as organic cation transporters (OCTs) [56,  57].

The competition for transporters may contribute to

the observed negative correlation between the ThDP

and NAD

levels during combined administration of

thiamine and NR. On the other hand, the positive

correlation between the ThDP and NAD

levels ob-

served when thiamine is administered alone, may

reflect an improvement in the body’s biosynthetic ca-

pacity due to the thiamine-dependent normalization

of metabolism.

Comparison of changes in the hand grip strength

and blood ThDP or NAD

levels (Fig.  5a), shows a

greater correlation of the grip strength with the levels

of ThDP than those of NAD

. When the ThDP content

is increased by thiamine administration, the muscle

strength of both right and left hands increases, while

subsequent decrease in the ThDP content upon the

dosage reduction is paralleled by a decrease in the

hand grip strength. In contrast, the decrease in the

blood NAD

levels after reducing the thiamine dose

does not affect the increase in the hand grip strength,

while increase in the blood NAD

levels caused by

NR administration does not prevent decreases in the

hand grip strength (Fig. 5a).

Given a previously demonstrated role of blood

TKT as an indicator of pathological metabolic chang-

es in CMT patients [40], the association of the blood

ThDP-dependent TKT activity with the intake of thi-

amine and NR has been investigated in the patient.

Figure5b shows that the blood TKT activity decreases

with the increase in the blood ThDP content and, vice

versa, increases with the ThDP decrease. At the same

time, thiamine administration alters the activities of

both total and holo-TKT to a similar extent, without

causing significant changes in the ThDP-induced TKT

activation (Fig.  5b). However, a combined intake of thi-

amine and NR, which decreases the blood ThDP level,

increases the ThDP-induced TKT activation (Fig.  5b).

Importantly, blood ThDP itself is not the only deter-

minant of the TKT activation, i.e., even if the blood

ThDP levels in the initial state and after the combined

intake of thiamine and NR are the same, the activa-

tion of TKT by the coenzyme is observed only in the

latter case (Fig. 5b). Reducing the NR dose upon con-

tinuous thiamine administration increases the blood

ThDP levels, that results in the disappearance of the

ThDP-induced TKT activation. The appearance of the

ThDP-induced activation of the blood TKT (Fig.  5b),

which is a sign of normal metabolism and is almost

absent in CMT (Fig. 4) [40], indicates normalization

of the patient’s metabolism as a result of combined

administration of thiamine and NR. However, chang-

es in the blood TKT and hand grip strength do not

correlate, similar to those in the grip strength and

blood NAD

(Fig. 5c).

Therefore, the patient’s TKT activity and blood

ThDP levels show an inverse relationship, and admin-

istration of 100 mg/day NR upon the thiamine intake

of 100 mg every other day promotes metabolic nor-

malization, assessed by the activation of blood TKT

by ThDP (Fig. 5).

Analysis of correlation between the blood

NAD

levels and parameters of thiamine status in

healthy women and female patients with CMT. Our

results on the separate and combined intake of thia-

mine and NR by a patient with CMT2K has revealed

novel patterns of interaction between the blood lev-

els of NAD

and ThDP or TKT activity (Fig.  5). Taking

into account previously established sex differences in

the thiamine metabolism [40], correlations between

the blood levels of NAD

and thiamine status param-

eters (ThDP, TKT activity) have been examined, using

available samples of women without and with CMT.

Comparison of correlations between the blood levels

of NAD

, ThDP, and holo-TKT activity in these samples

(Fig. 6) shows that healthy women exhibit a negative

correlation between the blood levels of NAD

and

ThDP (Rp  =  −0.58, p  =  0.06). The positive character of

this correlation in the female CMT patients (Rp  =  0.75,

p  =  0.03) points to metabolic changes associated

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1691

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 6. Correlations between the NAD

levels and thiamine status parameters in the blood of healthy women (control,

n  =  11) and female CMT patients before (n  =  8) and after 1-3 months of thiamine supplementation (n  =  6). Pearson co-

efficients (R

) and significance values (p) of the correlations are shown on the graphs; significant correlations (p  < 0.05)

are marked in bold. The data for different CMT forms are indicated by different symbols according to the legend.

Abnormally low NAD

values of the CMT2K patient investigated in this work (with the disease onset during the first de-

cade of life), compared to other patients (CMT1A and CMTX1, with the disease onset during the 4th-5th decade of life),

are not included in the correlation analysis.

with the disease. During the initial stage of thiamine

intake by the CMT2K patient (Fig.  5a, first month of

thiamine administration), the relation between the

blood levels of NAD

and ThDP is also positive. Hence,

the direction of correlation between the blood levels

of NAD

and ThDP is different in the affected (pos-

itive correlation) and healthy (negative correlation)

women.

As a result of thiamine administration, the cor-

relation coefficient between the NAD

and ThDP lev-

els in female CMT patients decreases from 0.75 to 0.03

(Fig.  6). The disappearance of the positive correlation

BORISOVA et al.1692

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

suggests a therapy-induced shift from the positive

relationship between the NAD

and ThDP levels,

characteristic of CMT patients (R

  =  0.75), towards

the negative correlation typical for healthy women

  =  −0.58). This transition is presumably facilitat-

ed by the combined administration of thiamine and

NR, since under these conditions, the blood levels of

NAD

and ThDP in the CMT2K patient show the op-

posite changes (Fig. 5), corresponding to the negative

correlation of these parameters in healthy females

(Fig. 6, control). Similarly, a significant negative cor-

relation between the blood levels of NAD

and endog-

enous holo-TKT activity, which is observed in healthy

women and disappears in the CMT female patients,

is restored after the thiamine therapy (Fig. 6). In the

case of CMT2K patient, the increase in the NAD

con-

tent upon combined administration of thiamine and

NR decreases the level of endogenous holo-TKT activ-

ity from 94 to 74% of the total TKT activity (Fig.  5b),

which corresponds to the negative correlation be-

tween the blood levels of NAD

and TKT activity ob-

served in healthy women (Fig. 6).

Thus, the relationships between the blood lev-

els of NAD

and thiamine status parameters are sig-

nificantly altered in female CMT patients, compared

tohealthy women. The therapy with thiamine and NR

shifts these relationships towards the control ones,

indicating metabolic normalization.

DISCUSSION

The molecular mechanisms of CMT development

induced by mutations in GDAP1 have been extensive-

ly studied in a number of independent laboratories,

because in certain populations mutations in this pro-

tein are the fifth most frequent cause of this common

hereditary motor and sensory neuropathy [2]. In our

work, the mechanisms of GDAP1 dysfunction upon its

L239F/A175P substitutions in a compound heterozy-

gous carrier with the early development of CMT2K

pathology are analyzed. Using the structure-function

analysis of GDAP1, we show that the amino acid sub-

stitutions caused by these mutations, may affect the

same hydrophobic-compound-binding region (Fig.  3).

As the heterozygous substitutions in parents jointly

influence the same GDAP1 binding function in the

compound heterozygous child, the function becomes

strongly perturbed, leading to early onset of the dis-

ease in the child in the absence of symptoms in the

parents.

Cellular phenotypes associated with GDAP1 mu-

tations leading to the amino acid substitutions in

the hydrophobic-compound-binding region are ana-

lyzed in the most relevant models, such as patients’

motor neurons and fibroblasts (Table  1). According

to the analysis of these data, a key common conse-

quence of amino acid substitutions in this region is

a dysfunction of the mitochondrial TCA cycle. Inde-

pendent studies investigating the effects of the L239

and C240 substitutions in GDAP1 report increased

PDH phosphorylation (and reduction of NADH pro-

duction in the TCA cycle) and decreased activity of

the NADH-oxidizing complex I, respectively (Table  1).

We have therefore suggested that thiamine and NR,

which are biosynthetic precursors of ThDP and NAD

both activating PDH and TCA cycle, should be useful

in the treatment of the CMT2K patient with the com-

pound heterozygous L239F/A175P substitution. ThDP

(the coenzyme form of thiamine) activates PDH not

only by saturating the enzyme active sites, but also

by inhibiting PDH phosphorylation, since ThDP acts as

an inhibitor of pyruvate dehydrogenase kinases (PDK)

[50, 51]. It should be mentioned that increased PDH

phosphorylation is associated with CMT not only upon

substitutions in the hydrophobic-compound-binding

region of GDAP1 (Table  1), but also in the case of up-

regulation of PDH phosphorylation by the activating

mutation in PDK3 [58]. Moreover, inactivation of PDH

due to a heterozygous mutation leading to the L138F

substitution in the enzyme α-subunit causes demye-

lination, which is one of the characteristic features of

CMT [59, 60]. It is worth noting that both GDAP1 [61,

62] and PDK isoenzyme 4 (PDK4) are involved in the

dynamin-related protein 1 (DRP1)-mediated mitochon-

drial fragmentation [63]. That is, DRP1 interaction

with the mitochondria is controlled by phosphoryla-

tion of the non-canonical PDK4 substrate septin2 [63].

ThDP not only inhibits PDK [50, 51], but also reduces

the activating phosphorylation of DRP1 at S616 [64],

which promotes mitochondrial fragmentation. Hence,

mitochondrial fragmentation depends on the action of

PDK and ThDP on the phosphorylation of septin2 and

DRP1. These data provide independent evidence link-

ing PDH phosphorylation to the observed disturbanc-

es in the mitochondrial dynamics in GDAP1 mutants.

It should be noted that PDKs exhibit tissue-specific

expression. Expression of PDK4 in the nervous tissue

is negligible under normal conditions. However, PDK4

may significantly increase in metabolic disorders [65].

Probably, the energy imbalance in GDAP1 mutants ac-

tivates PDK4 expression in the nervous tissue and/

or muscles. Hence, the therapeutic action of thiamine

as a precursor of ThDP in L239F/A175P-substituted

GDAP1 and in other forms of CMT [40] may be an

indicator of common disturbances in the PDH phos-

phorylation associated with CMT of different etiology.

As discussed above, mitochondrial dysfunction and

fragmentation, often observed in GDAP1 mutations,

may be a coupled effect of PDK upregulation.

When characterizing the therapeutic action of

thiamine and NR, we have found that the increase

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1693

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

in the hand grip strength correlates more with the

blood levels of ThDP than those of NAD

or TKT activ-

ity. However, a combined administration of thiamine

and NR promotes metabolic normalization, expressed

in the activating effect of ThDP on TKT and in the

correlations between the blood levels of the thia-

mine status parameters (ThDP, holo-TKT activity) and

NAD

. At the same time, it should be kept in mind

that the metabolism of NR and other NAD

precursors

involves complex interactions between body tissues

[66, 67]. Furthermore, available low-invasive blood

tests used in clinical studies is not optimal for de-

tecting an increase in the NAD

biosynthesis, since

most NAD

is located in the mitochondria, which are

absent in erythrocytes representing the major frac-

tion of blood cells. For the same reason, whole-blood

analysis is not appropriate for characterization of

mitochondrial metabolism. The small changes in the

blood levels of NAD

, observed upon the intake of NR

(Fig. 5), may be a consequence of the low NAD

con-

tent in this tissue. Nevertheless, the measured blood

parameters, such as the TKT regulation by ThDP and

the correlations between NAD

, ThDP, and holo-TKT

activity can be used as markers of metabolic changes

in the entire body. In particular, comparison of these

indicators and their relationships in the control and

CMT group points to the metabolic normalization in

the CMT2K patient treated with a combination of thi-

amine and NR. It is possible that an earlier adminis-

tration of thiamine and NR could be more efficient,

as slowing down the development of a pathology is

more feasible than repairing already damaged sys-

tem. This consideration justifies the metabolic cor-

rection therapy based on the early identification of

CMT-causing mutations and biochemical mechanisms

of their pathogenicity before the appearance of pro-

nounced clinical symptoms and significant disability

of patients.

Supplementary information

The online version contains supplementary material

available at https://doi.org/10.1134/S0006297925601911.

Contributions

N.R.B. performed NAD

assays, analyzed published re-

ports on phenotypes of GDAP1 mutations and regula-

tory role of kinases, analyzed and visualized dynamics

of the clinical parameters; A.A.E. performed struc-

ture-function analysis and visualization of GDAP1,

analyzed published data of invitro studies of GDAP1;

O.N.S. performed ThDP and TKT activity assays; N.V.B.

collected blood, performed NAD

assays, and curated

the data; O.P.S. performed neurologic tests and clini-

cal follow-up of the patient; V.I.B. developed the study

concept, supervised the study, analyzed the results

and literature, and wrote the manuscript. All authors

discussed the results, read the paper, and agreed on

the presented version of the paper.

Funding

This work was supported by the State Program

AAAA-A19-119042590056-2.

Ethics approval and consent to participate

The study was approved by the Ethics Committee of

Vladimirsky Moscow Regional Research and Clinical

Institute (protocol no.17, December10, 2020). All the

study participants provided voluntary informed con-

sent. For the minor patient, the consent was provided

by the mother.

Conflict of interest

The authors of this work declare that they have

noconflicts of interest.

REFERENCES

1. Bunik, V. I. (2024) A challenging interplay between

basic research, technologies and medical educa-

tion to provide therapies based on disease mecha-

nisms, Front. Med. (Lausanne), 11, 1464672, https://

doi.org/10.3389/fmed.2024.1464672.

2. Dai, S., Zheng,J., Chen, Y., Zhu, J., Wang, X., Peng,Y.,

Luo,Y., Lin,T., Li,Y., Ma,M., Shi,Z., Meng,X., Sun,L.,

and Zhou, J.C. (2025) A cross-sectional survey on the

health status of patients with Charcot-Marie-Tooth

disease in a Chinese national patient group, J.Neurol.,

272, 322, https://doi.org/10.1007/s00415-025-13063-7.

3. Cassereau, J., Chevrollier, A., Bonneau, D., Verny, C.,

Procaccio,V., Reynier,P., and Ferré,M. (2011) A locus-

specific database for mutations in GDAP1 allows ana-

lysis of genotype-phenotype correlations in Charcot-

Marie-Tooth diseases type 4A and 2K, Orphanet J.

Rare Dis., 6, 87, https://doi.org/10.1186/1750-1172-6-87.

4. Kotov,S., Sidorova,O., and Borodataya,E. (2019) Mi-

tochondrial disorders in neuromuscular pathology,

Neuromusc. Dis., 9, 22-31, https://doi.org/10.17650/

2222-8721-2019-9-3-22-31.

5. Niemann, A., Ruegg, M., La Padula, V., Schenone, A.,

and Suter, U. (2005) Ganglioside-induced differenti-

ation associated protein 1 is a regulator of the mi-

tochondrial network: new implications for Charcot-

Marie-Tooth disease, J. Cell Biol., 170, 1067-1078,

https://doi.org/10.1083/jcb.200507087.

6. Pakhrin, P.S., Xie,Y., Hu,Z., Li,X., Liu,L., Huang,S.,

Wang,B., Yang,Z., Zhang,J., Liu,X., Xia,K., Tang,B.,

and Zhang,R. (2018) Genotype-phenotype correlation

and frequency of distribution in a cohort of Chi-

nese Charcot-Marie-Tooth patients associated with

GDAP1 mutations, J. Neurol., 265, 637-646, https://

doi.org/10.1007/s00415-018-8743-9.

BORISOVA et al.1694

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

7. González-Sánchez,P., Satrústegui,J., Palau,F., and Del

Arco, A. (2019) Calcium deregulation and mitochon-

drial bioenergetics in GDAP1-related CMT, Int. J. Mol.

Sci., 20, 403, https://doi.org/10.3390/ijms20020403.

8. Cantarero, L., García-Vargas, G., Hoenicka, J., and

Palau, F. (2023) Differential effects of Mendelian

GDAP1 clinical variants on mitochondria-lysosome

membrane contacts sites, Biol. Open, 12, bio059707,

https://doi.org/10.1242/bio.059707.

9. Xia, M., Zhang, Y., Jin, K., Lu, Z., Zeng, Z., and

Xiong, W. (2019) Communication between mitochon-

dria and other organelles: a brand-new perspective

on mitochondria in cancer, Cell Biosci., 9, 27, https://

doi.org/10.1186/s13578-019-0289-8.

10. Xin, B., Puffenberger, E., Nye, L., Wiznitzer, M., and

Wang, H. (2008) A novel mutation in the GDAP1

gene is associated with autosomal recessive Char-

cot-Marie-Tooth disease in an Amish family, Clin.

Genet., 74, 274-278, https://doi.org/10.1111/j.1399-

0004.2008.01018.x.

11. Ortiz-Santiago, A., and Ramos, E. (2021) Childhood

onset homozygous recessive GDAP1 (p.Pro231Leu)

mutation in a 9-year-old puerto rican pediatric fe-

male with axonal Charcot-Marie-Tooth disease: a case

report, J. Pediatr. Rehabil. Med., 14, 533-537, https://

doi.org/10.3233/prm-200695.

12. Manzoor, U., Ali, A., Ali, S. L., Abdelkarem, O.,

Kanwal, S., Alotaibi, S. S., Baazeem, A., Baiduisse-

nova, A., Yktiyarov, A., Hajar, A., and Olzhabay, A.

(2023) Mutational screening of GDAP1 in dysphonia

associated with Charcot-Marie-Tooth disease: clinical

insights and phenotypic effects, J. Genet. Eng. Bio-

technol., 21, 119, https://doi.org/10.1186/s43141-023-

00568-9.

13. Vivar, C., and Avila, J. D. (2019) GDAP1-related

Charcot- Marie-Tooth disease: additional evidence for

the c.692C>T variant as a pathogenic mutation (P3.4-

045), Neurology, 92, P3.4-045, https://doi.org/10.1212/

WNL.92.15_supplement.P3.4-045.

14. Wolf,C., Pouya,A., Bitar,S., Pfeiffer,A., Bueno,D., Rojas-

Charry,L., Arndt,S., Gomez-Zepeda,D., Tenzer,S., Bello,

F. D., Vianello, C., Ritz, S., Schwirz, J., Dobrindt, K.,

Peitz,M., Hanschmann, E.M., Mencke,P., Boussaad,I.,

Silies, M., Brüstle, O., et al. (2022) GDAP1 loss of

function inhibits the mitochondrial pyruvate dehy-

drogenase complex by altering the actin cytoskele-

ton, Commun. Biol., 5, 541, https://doi.org/10.1038/

s42003-022-03487-6.

15. Kabzińska, D., Strugalska-Cynowska, H., Kostera-

Pruszczyk, A., Ryniewicz, B., Posmyk, R., Midro, A.,

Seeman, P., Báranková, L., Zimoń, M., Baets, J., Tim-

merman,V., Guergueltcheva,V., Tournev,I., Sarafov,S.,

De Jonghe, P., Jordanova, A., Hausmanowa-Petruse-

wicz,I., and Kochański,A. (2010) L239F founder mu-

tation in GDAP1 is associated with a mild Charcot-

Marie-Tooth type 4C4 (CMT4C4) phenotype, Neuroge-

netics, 11, 357-366, https://doi.org/10.1007/s10048-010-

0237-6.

16. Noack,R., Frede,S., Albrecht,P., Henke,N., Pfeiffer,A.,

Knoll,K., Dehmel,T., Meyer Zu Hörste,G., Stettner,M.,

Kieseier, B. C., Summer, H., Golz, S., Kochanski, A.,

Wiedau-Pazos, M., Arnold, S., Lewerenz, J., and

Methner, A. (2012) Charcot-Marie-Tooth disease CM-

T4A: GDAP1 increases cellular glutathione and the mi-

tochondrial membrane potential, Hum. Mol. Genet.,

21, 150-162, https://doi.org/10.1093/hmg/ddr450.

17. Sutinen, A., Paffenholz, D., Nguyen, G. T. T., Ruska-

mo,S., Torda, A.E., and Kursula,P. (2023) Conserved

intramolecular networks in GDAP1 are closely con-

nected to CMT-linked mutations and protein stabil-

ity, PLoS One, 18, e0284532, https://doi.org/10.1371/

journal.pone.0284532.

18. Marco, A., Cuesta, A., Pedrola, L., Palau, F., and

Marín,I. (2004) Evolutionary and structural analyses

of GDAP1, involved in Charcot-Marie-Tooth disease,

characterize a novel class of glutathione transfer-

ase-related genes, Mol. Biol. Evol., 21, 176-187, https://

doi.org/10.1093/molbev/msh013.

19. Googins, M.R., Woghiren-Afegbua, A.O., Calderon,M.,

St Croix, C. M., Kiselyov, K. I., and VanDemark, A. P.

(2020) Structural and functional divergence of

GDAP1 from the glutathione S-transferase superfam-

ily, FASEB J., 34, 7192-7207, https://doi.org/10.1096/

fj.202000110R.

20. Nguyen, G. T. T., Sutinen, A., Raasakka, A., Muruga-

nandam,G., Loris,R., and Kursula,P. (2020) Structure

of the complete dimeric human GDAP1 core domain

provides insights into ligand binding and clustering

of disease mutations, Front. Mol. Biosci., 7, 631232,

https://doi.org/10.3389/fmolb.2020.631232.

21. Huber, N., Bieniossek, C., Wagner, K. M., Elsässer,

H. P., Suter, U., Berger, I., and Niemann, A. (2016)

Glutathione-conjugating and membrane-remodeling

activity of GDAP1 relies on amphipathic C-terminal

domain, Sci. Rep., 6, 36930, https://doi.org/10.1038/

srep36930.

22. Cassereau, J., Chevrollier, A., Codron, P., Goizet, C.,

Gueguen, N., Verny, C., Reynier, P., Bonneau, D.,

Lenaers,G., and Procaccio,V. (2020) Oxidative stress

contributes differentially to the pathophysiology

of Charcot-Marie-Tooth disease type 2K, Exp. Neu-

rol., 323, 113069, https://doi.org/10.1016/j.expneurol.

2019.113069.

23. Pla-Martín,D., Rueda, C.B., Estela,A., Sánchez- Piris,M.,

González-Sánchez, P., Traba, J., de la Fuente, S.,

Scorrano,L., Renau-Piqueras, J., Alvarez,J., Satrúste-

gui, J., and Palau, F. (2013) Silencing of the Charcot-

Marie-Tooth disease-associated gene GDAP1 induces

abnormal mitochondrial distribution and affects Ca

homeostasis by reducing store-operated Ca

entry,

Neurobiol. Dis., 55, 140-151, https://doi.org/10.1016/

j.nbd.2013.03.010.

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1695

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

24. Sutinen,A., Jones, N.C., Hoffmann, S.V., Ruskamo,S.,

and Kursula, P. (2023) Conformational analysis of

membrane-proximal segments of GDAP1 in a lipid-

ic environment using synchrotron radiation sug-

gests a mode of assembly at the mitochondrial out-

er membrane, Biophys. Chem., 303, 107113, https://

doi.org/10.1016/j.bpc.2023.107113.

25. Chen, Y., Zhu, J., Wang, M., Zhao, Q., Huang, C.,

Tarjibayeva, S., Wang, L., Sun, L., and Zhou, J. C.

(2025) Advances and clues of nutritional adju-

vant therapy for Charcot-Marie-Tooth disease, J.

Nutr., 155, 3642-3653, https://doi.org/10.1016/j.tjnut.

2025.09.026.

26. Gubler, C. J., Johnson, L. R., and Wittorf, J. H.

(1970) Yeast transketolase (sedoheptulose-7-phos-

phate:d-glyceraldehyde-3-phosphate dihydroxyac-

etonetransferase, EC 2.2.1.1) assay of thiamine di-

phosphate, Methods Enzymol., 18, 120-125, https://

doi.org/10.1016/0076-6879(71)18290-0.

27. Kochetov, G. A. (1980) Practice Guidelines on Bio-

chemistry [in Russian], Vysshaya Shkola, Moscow.

28. Solovjeva, O. N. (2002) Isolation and properties of

noncovalent complex of transketolase with RNA,

Biochemistry (Moscow), 67, 667-671, https://doi.org/

10.1023/a:1016198321838.

29. Tikhomirova, N.K., and Kochetov, G.A. (1990) Purifi-

cation of transketolase from baker’s yeast by an im-

munoadsorbent, Biochem. Int., 22, 31-36.

30. Solovjeva, O. N., Selivanov, V. A., Orlov, V. N., and

Kochetov, G. A. (2019) Stages of the formation of

nonequivalence of active centers of transketolase

from baker’s yeast, Mol. Catal., 466, 122-129, https://

doi.org/10.1016/j.mcat.2019.01.00.

31. Heinrich, C.P., Noack,K., and Wiss,O. (1972) Chemi-

cal modification of tryptophan at the binding site of

thiamine-pyrophosphate in transketolase from Bak-

er’s yeast, Biochem. Biophys. Res. Commun., 49, 1427-

1432, https://doi.org/10.1016/0006-291x(72)90498-6.

32. Kim,M., Won, C.W., and Kim,M. (2018) Muscular grip

strength normative values for a Korean population

from the Korea National Health and Nutrition Exam-

ination Survey, 2014-2015, PLoS One, 13, e0201275,

https://doi.org/10.1371/journal.pone.0201275.

33. Harding, A. E., and Thomas, P. K. (1980) The clin-

ical features of hereditary motor and sensory neu-

ropathy types I and II, Brain, 103, 259-280, https://

doi.org/10.1093/brain/103.2.259.

34. Wong, S.L. (2016) Grip strength reference values for

Canadians aged 6 to 79: Canadian Health Measures

Survey, 2007 to 2013, Health Rep., 27, 3-10.

35. Jumper,J., Evans,R., Pritzel,A., Green,T., Figurnov,M.,

Ronneberger, O., Tunyasuvunakool, K., Bates, R.,

Žídek,A., Potapenko,A., Bridgland,A., Meyer,C., Kohl,

S.A.A., Ballard, A.J., Cowie,A., Romera-Paredes,B., Ni-

kolov,S., Jain,R., Adler,J., Back,T., etal. (2021) Highly

accurate protein structure prediction with AlphaFold,

Nature, 596, 583-589, https://doi.org/10.1038/s41586-

021-03819-2.

36. Maupetit, J., Derreumaux, P., and Tuffery, P. (2009)

PEP-FOLD: an online resource for de novo pep-

tide structure prediction, Nucleic Acids Res., 37,

W498-W503, https://doi.org/10.1093/nar/gkp323.

37. DeLano, W. L., and Bromberg, S. (2004) PyMOL user’s

guide, DeLano Scientific LLC.

38. Svensson,E., and Häger-Ross,C. (2006) Hand function

in Charcot Marie Tooth: test retest reliability of some

measurements, Clin. Rehabil., 20, 896-908, https://

doi.org/10.1177/0269215506072184.

39. Bohannon, R.W. (2017) Test-retest reliability of mea-

surements of hand-grip strength obtained by dyna-

mometry from older adults: a systematic review of

research in the PubMed database, J.Frailty Aging, 6,

83-87, https://doi.org/10.14283/jfa.2017.8.

40. Artiukhov, A. V., Solovjeva, O. N., Balashova, N. V.,

Sidorova, O. P., Graf, A. V., and Bunik, V. I. (2024)

Pharmacological doses of thiamine benefit patients

with the Charcot-Marie-Tooth neuropathy by chang-

ing thiamine diphosphate levels and affecting regu-

lation of thiamine-dependent enzymes, Biochemis-

try (Moscow), 89, 1161-1182, https://doi.org/10.1134/

S0006297924070010.

41. Balashova, N. V., Zavileyskiy, L. G., Artiukhov, A. V.,

Shaposhnikov, L. A., Sidorova, O. P., Tishkov, V. I.,

Tramonti, A., Pometun, A. A., and Bunik, V. I. (2022)

Efficient assay and marker significance of NAD(+)

in human blood, Front. Med. (Lausanne), 9, 886485,

https://doi.org/10.3389/fmed.2022.886485.

42. Reimer,U., Scherer,G., Drewello,M., Kruber,S., Schut-

kowski, M., and Fischer, G. (1998) Side-chain effects

on peptidyl-prolyl cis/trans isomerisation, J. Mol. Biol.,

279, 449-460, https://doi.org/10.1006/jmbi.1998.1770.

43. Schmidpeter, P.A., Koch, J.R., and Schmid, F.X. (2015)

Control of protein function by prolyl isomerization,

Biochim. Biophys. Acta, 1850, 1973-1982, https://

doi.org/10.1016/j.bbagen.2014.12.019.

44. Ammar,N., Nelis,E., Merlini,L., Barisić,N., Amouri,R.,

Ceuterick,C., Martin, J.J., Timmerman,V., Hentati,F.,

and De Jonghe,P. (2003) Identification of novel GDAP1

mutations causing autosomal recessive Charcot-

Marie-Tooth disease, Neuromuscul. Disord., 13, 720-

728, https://doi.org/10.1016/s0960-8966(03)00093-2.

45. Baránková,L., Vyhnálková,E., Züchner,S., Mazanec,R.,

Sakmaryová, I., Vondrácek, P., Merlini, L., Bojar, M.,

Nelis,E., De Jonghe, P., and Seeman,P. (2007) GDAP1

mutations in Czech families with early-onset CMT,

Neuromuscul. Disord., 17, 482-489, https://doi.org/

10.1016/j.nmd.2007.02.010.

46. Auer-Grumbach, M., Fischer, C., Papić, L., John, E.,

Plecko, B., Bittner, R. E., Bernert, G., Pieber, T. R.,

Miltenberger, G., Schwarz, R., Windpassinger, C.,

Grill,F., Timmerman,V., Speicher, M.R., and Janecke,

A. R. (2008) Two novel mutations in the GDAP1

BORISOVA et al.1696

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

and PRX genes in early onset Charcot-Marie-Tooth

syndrome, Neuropediatrics, 39, 33-38, https://

doi.org/10.1055/s-2008-1077085.

47. Moroni, I., Morbin, M., Milani, M., Ciano, C.,

Bugiani, M., Pagliano, E., Cavallaro, T., Pareyson, D.,

and Taroni, F. (2009) Novel mutations in the GDAP1

gene in patients affected with early-onset axonal

Charcot-Marie-Tooth type 4A, Neuromuscul. Disord.,

19, 476-480, https://doi.org/10.1016/j.nmd.2009.04.014.

48. Rougeot,C., Chabrier,S., Camdessanche, J.P., Prieur,F.,

d’Anjou, M. C., and Latour, P. (2008) Clinical, electro-

physiological and genetic studies of two families with

mutations in the GDAP1 gene, Neuropediatrics, 39,

184-187, https://doi.org/10.1055/s-0028-1085467.

49. Binięda, K., Rzepnikowska, W., Kolakowski, D.,

Kaminska,J., Szczepankiewicz, A. A., Nieznańska,H.,

Kochański, A., and Kabzińska, D. (2021) Mutations

in GDAP1 influence structure and function of the

trans-Golgi network, Int. J. Mol. Sci., 22, 914, https://

doi.org/10.3390/ijms22020914.

50. Cassereau, J., Chevrollier, A., Gueguen, N., Malinge,

M.C., Letournel,F., Nicolas,G., Richard,L., Ferre,M.,

Verny, C., Dubas, F., Procaccio,V., Amati-Bonneau, P.,

Bonneau, D., and Reynier, P. (2009) Mitochondrial

complex I deficiency in GDAP1-related autosomal

dominant Charcot-Marie-Tooth disease (CMT2K),

Neurogenetics, 10, 145-150, https://doi.org/10.1007/

s10048-008-0166-9.

51. Van den Bossche, D., Schiettecatte, J., Vekens, E.,

De Smet, D., Gorus, F. K., and Martens, G. A. (2012)

Enzymatic pyruvate measurement by Cobas 6000

open channel assay, Clin. Lab., 58, 1091-1095.

52. Bunik,V. (2023) The therapeutic potential of vitamins

B1, B3 and B6 in Charcot-Marie-Tooth disease with

the compromised status of vitamin-dependent pro-

cesses, Biology (Basel), 12, 897, https://doi.org/10.3390/

biology12070897.

53. Jonus, H.C., Byrnes, C.C., Kim,J., Valle, M.L., Bartlett,

M. G., Said, H. M., and Zastre, J. A. (2020) Thiamine

mimetics sulbutiamine and benfotiamine as a nu-

traceutical approach to anticancer therapy, Biomed.

Pharmacother., 121, 109648, https://doi.org/10.1016/

j.biopha.2019.109648.

54. Pawlosky,R., Demarest, T.G., King, M.T., Estrada,D.,

Veech, R. L., and Bohr, V. A. (2025) Effect of dietary

ketosis and nicotinamide riboside on hippocampal

krebs cycle Intermediates and mitochondrial energet-

ics in a DNA repair-deficient 3xTg/POLβ(+/–) Alzhei-

mer disease mouse model, J. Neurochem., 169, e16295,

https://doi.org/10.1111/jnc.16295.

55. Schaefer, P. M., Huang, J., Butic, A., Perry, C.,

Yardeni, T., Tan, W., Morrow, R., Baur, J. A., and

Wallace, D.C. (2022) Nicotinamide riboside alleviates

exercise intolerance in ANT1-deficient mice, Mol.

Metab., 64, 101560, https://doi.org/10.1016/j.molmet.

2022.101560.

56. Ahmadi, A., Begue, G., Valencia, A. P., Norman, J. E.,

Lidgard,B., Bennett, B.J., Van Doren, M.P., Marcinek,

D.J., Fan,S., Prince, D.K., Gamboa, J., Himmelfarb,J.,

de Boer, I.H., Kestenbaum, B.R., and Roshanravan,B.

(2023) Randomized crossover clinical trial of coen-

zyme Q10 and nicotinamide riboside in chronic kid-

ney disease, JCI Insight, 8, e167274, https://doi.org/

10.1172/jci.insight.167274.

57. Kropotov,A., Kulikova,V., Nerinovski,K., Yakimov,A.,

Svetlova, M., Solovjeva, L., Sudnitsyna, J., Migaud,

M.E., Khodorkovskiy,M., Ziegler,M., and Nikiforov,A.

(2021) Equilibrative nucleoside transporters mediate

the import of nicotinamide riboside and nicotinic

acid riboside into human cells, Int. J. Mol. Sci., 22,

1391, https://doi.org/10.3390/ijms22031391.

58. Overton, E., Emelyanova, A., and Bunik, V. I. (2025)

Thiamine, gastrointestinal beriberi and acetylcho-

line signaling, Front. Nutr., 12, 1541054, https://

doi.org/10.3389/fnut.2025.1541054.

59. Ailabouni, A.S., Vijaywargi,G., Subash,S., Singh, D.K.,

Gaborik,Z., and Prasad,B. (2025) Is N1-methylnicotin-

amide a good organic cation transporter 2 (OCT2) bio-

marker? Metabolites, 15, 80, https://doi.org/10.3390/

metabo15020080.

60. Perez-Siles, G., Cutrupi, A., Ellis, M., Screnci, R.,

Mao, D., Uesugi, M., Yiu, E. M., Ryan, M. M., Choi,

B.O., Nicholson,G., and Kennerson, M.L. (2020) Ener-

gy metabolism and mitochondrial defects in X-linked

Charcot-Marie-Tooth (CMTX6) iPSC-derived motor

neurons with the p.R158H PDK3 mutation, Sci. Rep.,

10, 9262, https://doi.org/10.1038/s41598-020-66266-5.

61. Singhi, P., De Meirleir, L., Lissens, W., Singhi, S., and

Saini, A. G. (2013) Pyruvate dehydrogenase-e1α defi-

ciency presenting as recurrent demyelination: an un-

usual presentation and a novel mutation, JIMD Rep.,

10, 107-111, https://doi.org/10.1007/8904_2012_211.

62. Bonne, G., Benelli, C., De Meirleir, L., Lissens, W.,

Chaussain, M., Diry, M., Clot, J. P., Ponsot, G., Geof-

froy, V., Leroux, J. P., and Marsac, C. (1993) E1 py-

ruvate dehydrogenase deficiency in a child with

motor neuropathy, Pediatr. Res., 33, 284-288, https://

doi.org/10.1203/00006450-199303000-00016.

63. Huber, N., Guimaraes, S., Schrader, M., Suter,U., and

Niemann,A. (2013) Charcot-Marie-Tooth disease-asso-

ciated mutants of GDAP1 dissociate its roles in per-

oxisomal and mitochondrial fission, EMBO Rep., 14,

545-552, https://doi.org/10.1038/embor.2013.56.

64. Niemann, A., Wagner, K. M., Ruegg, M., and

Suter, U. (2009) GDAP1 mutations differ in their ef-

fects on mitochondrial dynamics and apoptosis

depending on the mode of inheritance, Neurobi-

ol. Dis., 36, 509-520, https://doi.org/10.1016/j.nbd.

2009.09.011.

65. Thoudam,T., Chanda,D., Sinam, I.S., Kim, B.G., Kim,

M. J., Oh, C. J., Lee, J. Y., Kim, M. J., Park, S. Y., Lee,

S.Y., Jung, M.K., Mun, J.Y., Harris, R.A., Ishihara,N.,

MECHANISMS OF PATHOLOGY AND THERAPY IN GDAP1 MUTATIONS 1697

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Jeon, J. H., and Lee, I. K. (2022) Noncanonical PDK4

action alters mitochondrial dynamics to affect the

cellular respiratory status, Proc. Natl. Acad. Sci.

USA, 119, e2120157119, https://doi.org/10.1073/pnas.

2120157119.

66. Yamada, Y., Kusakari, Y., Akaoka, M., Watanabe, M.,

Tanihata, J., Nishioka, N., Bochimoto, H., Akaike, T.,

Tachibana, T., and Minamisawa, S. (2021) Thiamine

treatment preserves cardiac function against isch-

emia injury via maintaining mitochondrial size

and ATP levels, J. Appl. Physiol., 130, 26-35, https://

doi.org/10.1152/japplphysiol.00578.2020.

67. Klyuyeva,A., Tuganova,A., Kedishvili,N., and Popov,

K. M. (2019) Tissue-specific kinase expression and

activity regulate flux through the pyruvate dehydro-

genase complex, J.Biol. Chem., 294, 838-851, https://

doi.org/10.1074/jbc.RA118.006433.

68. Kropotov, A., Kulikova,V., Solovjeva, L., Yakimov,A.,

Nerinovski,K., Svetlova,M., Sudnitsyna,J., Plusnina,A.,

Antipova, M., Khodorkovskiy, M., Migaud, M. E.,

Gambaryan, S., Ziegler, M., and Nikiforov, A. (2022)

Purine nucleoside phosphorylase controls nicotin-

amide riboside metabolism in mammalian cells,

J.Biol. Chem., 298, 102615, https://doi.org/10.1016/

j.jbc.2022.102615.

69. Song, W. S., Shen, X., Du, K., Ramirez, C. B., Park,

S.H., Cao,Y., Le,J., Bae,H., Kim, J., Chun, Y., Khong,

N.J., Kim,M., Jung,S., Choi,W., Lopez, M.L., Said,Z.,

Song,Z., Lee, S.G., Nicholas,D., Sasaki,Y., etal. (2025)

Nicotinic acid riboside maintains NAD(+) homeosta-

sis and ameliorates aging-associated NAD(+) decline,

Cell Metab., 37, 1616-1618, https://doi.org/10.1016/

j.cmet.2025.05.004.

Publisher’s Note. Pleiades Publishing remains

neutral with regard to jurisdictional claims in published

maps and institutional affiliations. AI tools may have

been used in the translation or editing of this article.