Glucose intolerance in aging is mediated by the Gpcpd1-GPC metabolic axis

Abstract

Skeletal muscle plays a central role in the regulation of systemic metabolism during lifespan. With aging, muscle mediated metabolic homeostasis is perturbed, contributing to the onset of multiple chronic diseases. Our knowledge on the mechanisms responsible for this age-related perturbation is limited, as it is difficult to distinguish between correlation and causality of molecular changes in muscle aging. Glycerophosphocholine phosphodiesterase 1 (GPCPD1) is a highly abundant muscle enzyme responsible for the hydrolysis of the lipid glycerophosphocholine (GPC). The physiological function of GPCPD1 remained largely unknown. Here, we report that the GPCPD1-GPC metabolic pathway is dramatically perturbed in the aged muscle. Muscle-specific inactivation of Gpcpd1 resulted in severely affected glucose metabolism, without affecting muscle development. This pathology was muscle specific and did not occur in white fat-, brown fat- and liver-specific Gpcpd1 deficient mice. Moreover, in the muscle specific mutant mice, glucose intolerance was markedly accelerated under high sugar and high fat diet. Mechanistically, Gpcpd1 deficiency results in accumulation of GPC, without any other significant changes in the global lipidome. This causes an aged-like transcriptomic signature in young Gpcpd1 deficient muscles, changes in myofiber osmolarity, and impaired insulin signaling. Finally, we report that GPC levels are markedly perturbed in muscles from both aged humans and patients with Type 2 diabetes. These results identify the GPCPD1-GPC metabolic pathway as critical to muscle aging and age-associated glucose intolerance.

Full text

Glucose intolerance in aging is
mediated by the Gpcpd1-GPC metabolic axis
Domagoj Cikes1*, Michael Leutner2, Shane J.F. Cronin1, Maria Novatchkova1, Lorenz Pfleger2,
Radka Klepochová2, Benjamin Lair3, Marlène Lac3, Camille Bergoglio3, Gerhard Dürnberger4,
Elisabeth Roitinger4, Eric Rullman5,6, Thomas Gustafsson5, Astrid Hagelkruys1, Geneviève
Tavernier3, Virginie Bourlier3, Claude Knauf7, Cedric Moro3, Michael Krebs2, Alexandra
Kautzky-Willer2, Martin Krssak2, Michael Orthofer1,9, Josef M. Penninger1,10*
1 IMBA, Institute of Molecular Biotechnology of the Austrian Academy of Sciences, Vienna,
1030, Austria
2 Division of Endocrinology and Metabolism Department of Internal Medicine III, Medical
University of Vienna, Vienna, 1090, Austria
3Institute of Metabolic and Cardiovascular Diseases, Team MetaDiab, Inserm UMR1297,
Toulouse, France
4 VBCF, Vienna Biocenter Core Facilities, Vienna Biocenter, Vienna 1030; Austria
5 Division of Clinical Physiology, Department of Laboratory Medicine, Karolinska Institutet,
and Unit of Clinical Physiology, Karolinska University Hospital, Stockholm; Sweden
6 Cardiovascular Theme, Karolinska Institutet, Karolinska University Hospital Huddinge,
Stockholm; Sweden
7 INSERM U1220 Institut de Recherche en Santé Digestive, CHU Purpan, Université Toulouse
III Paul Sabatier Toulouse; France
8 Institute of Metabolic and Cardiovascular Diseases, Inserm/Paul Sabatier University UMR
1297, Toulouse, France
9JLP health, Vienna; Austria
10 Department of Medical Genetics, Life Sciences Institute, University of British Columbia,
Vancouver; Canada
*Correspondence
domagoj.cikes@imba.oeaw.ac.at (D.C.),
josef.penninger@ubc.ca (J.M.P)
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Abstract
Skeletal muscle plays a central role in the regulation of systemic metabolism during lifespan.
With aging, muscle mediated metabolic homeostasis is perturbed, contributing to the onset
of multiple chronic diseases. Our knowledge on the mechanisms responsible for this age-
related perturbation is limited, as it is difficult to distinguish between correlation and causality
of molecular changes in muscle aging. Glycerophosphocholine phosphodiesterase 1
(GPCPD1) is a highly abundant muscle enzyme responsible for the hydrolysis of the lipid
glycerophosphocholine (GPC). The physiological function of GPCPD1 remained largely
unknown. Here, we report that the GPCPD1-GPC metabolic pathway is dramatically perturbed
in the aged muscle. Muscle-specific inactivation of Gpcpd1 resulted in severely affected
glucose metabolism, without affecting muscle development. This pathology was muscle
specific and did not occur in white fat-, brown fat- and liver-deficient Gpcpd1 deficient mice.
Moreover, in the muscle specific mutant mice, glucose intolerance was markedly accelerated
under high sugar and high fat diet. Mechanistically, Gpcpd1 deficiency results in accumulation
of GPC, without any other significant changes in the global lipidome. This causes an “aged-
like” transcriptomic signature in young Gpcpd1 deficient muscles and impaired insulin
signaling. Finally, we report that GPC levels are markedly perturbed in muscles from both
aged humans and patients with Type 2 diabetes, with highly significant positive correlation of
GPC levels with advanced age. These results identify the GPCPD1-GPC metabolic pathway as
critical to muscle aging and age-associated glucose intolerance.
Keywords: Aging; Muscle; Insulin signaling; Glycerophosphocholine phosphodiesterase 1;
Glycerophosphocholine; Glucose intolerance; Metabolic dysfunction.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Introduction
Skeletal muscle is the biggest organ in the body, essential for mobility, health-span and
lifespan (Cruz-Jentoft and Sayer, 2019; Yanagi et al., 2021; Yang et al., 2019). Apart from
mobility, muscle plays a crucial role in systemic metabolism (Baskin et al., 2015). Critical to
this regulatory role is the control of glucose metabolism, with muscle metabolizing up to 80%
of ingested glucose (DeFronzo et al., 1981; Jue et al., 1989). Aged muscles exhibit multiple
molecular and metabolic perturbations, that can potentially affect muscle function and
whole-body metabolism during the lifespan (Demontis et al., 2013a). With muscle ageing,
glucose utilization is significantly impaired, leading to systemic perturbations of glucose
metabolism (Chia et al., 2018; Kern et al., 1992). This systemic metabolic dysfunction
contributes to the onset and progression of multiple chronic diseases (Cowie et al., 2009).
Glycerophosphodiester phosphodiesterase 1 (Gpcpd1; Gdpd6; Edi3) is a member of a large
family of glycerophosphodiester phosphodiesterases, highly conserved enzymes found in
bacteria, protozoa, and mammals (Corda et al., 2014). Mammalian
glycerophosphodiesterases exist in multiple isoforms (seven in humans) with a high degree
of tissue and substrate specificity (Corda et al., 2014). Gpcpd1 is a 76.6 kD protein that
hydrolyzes lipid glycerophosphocholine (GPC), yielding choline and glycerol-3-phosphate.
Unlike other phosphodiesterases, Gpcpd1 does not contain a transmembrane region and is
localized in the cytoplasm. An early study reported that Gpcpd1 expression is enriched in
heart and skeletal muscle (Okazaki et al., 2010). The same study also reported that over-
expression of a truncated version of Gpcpd1 in the muscles resulted in muscle atrophy,
suggesting that Gpcpd1 regulates muscle differentiation (Okazaki et al., 2010). High
expression of Gpcpd1 was found in metastatic endometrial cancers, while inhibition of
Gpcpd1 in breast cancer cells inhibited migration and invasion (Stewart et al., 2012). Apart
from these studies, the physiological function of Gpcpd1 remains largely unknown, and an in
vivo loss-of-function analysis has never been reported.
A recent untargeted metabolomic profiling of skeletal muscle from old mice indicated
glycerophosphocholine (GPC), the Gpcpd1 substrate, as a significantly elevated metabolite in
aged mouse muscles (Houtkooper et al., 2011). Moreover, a single nucleotide polymorphism
in proximity to the human GPCPD1 locus was found to be associated with longevity (Pilling et
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

al., 2016). Therefore, we set out to investigate if there is a direct link between the Gpcpd1-
GPC metabolic pathway and aging-related health decline.
MATERIALS AND METHODS
Animal care
All experimental protocols were approved by the institutional Animal Ethics Committee in
accordance with the Austrian legal guidelines on Animal Care. All mice were on a C57BL/6J
background (in-house colony). MckCre, Ap2Cre, Ucp1Cre, AlbCre mice were purchased from
the Jackson Laboratory (Bar Harbor, US, stock number 006405, 005069, 024670, 003574). The
mice were maintained on a 12:12 hour light-dark cycle (lights on 08:00-20:00) while housed
in ventilated cages at an ambient temperature of 25oC. Mice were fed ad libitum standard
chow diet (17% kcal from fat; Envigo, GmbH), high-fat diet (HFD; 60% kcal from fat; Envigo,
GmbH), or high fructose diet (30% fructose in drinking water; #F0127 Sigma Aldrich Gmbh)
starting from 8 weeks of age.
Human biopsies
All human experiments were approved by the regional ethical review board in Stockholm
(2014/516-31/2 and 2010/786-31/3) and complied with the Declaration of Helsinki. Oral and
written informed consent were obtained from all subjects prior to participation in the study.
8 healthy young adults (age 21-29) and 8 middle-aged (age 45-62) subjects were recruited.
The subjects did not use any medications and were nonsmokers. Biopsies of the quadriceps
skeletal muscle (vastus lateralis) were obtained under local anesthesia using the Bergström
percutaneous needle biopsy technique (Bergström and Hultman, 1967). The biopsies were
immediately frozen in isopentane, cooled in liquid nitrogen, and stored at −80°C until further
analysis.
Determination of muscle glycerophosphocholine in Type 2 diabetes patients
This prospective clinical study was performed at the Department of Internal Medicine 3,
Division of Endocrinology and Metabolism at the Medical University of Vienna between June
2021 and September 2021. Patients were recruited at the Diabetes Outpatient Department
of the Medical University of Vienna and enrolled for the study after giving written informed
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

consent. Routine laboratory measurements were analyzed at the certified Department of
Medical and Chemical Laboratory Diagnostics (http://www.kimcl.at/) of the Medical
University of Vienna. In the present study, we included patients with diagnosed diabetes
mellitus type 2 and patients without diabetes mellitus aged between 50 and 65 years.
Exclusion criteria were infectious diseases such as hepatitis B or C or human
immunodeficiency virus (HIV) We performed a detailed metabolic characterization, including
measurements of lipid parameters, glucose metabolism (e.g. HbA1c, fasting glucose levels),
routine laboratory analyses and anthropometric measurements. In vivo magnetic resonance
spectroscopy (MRS) of Vastus lateralis muscle was performed on individuals that were
selected based on their age and without the use of any medications or a record of
musculoskeletal or cardiovascular disease (Krumpolec et al., 2020; Valkovič et al., 2013).
Laboratory analyses and MRS measurements were performed after a 12-hour fasting period.
The study was approved by the ethics board of the Medical University of Vienna.
Cross tissue Gpcpd1 expression analysis
The data on Gpcpd1 mRNA expression analysis in several tissues were derived from the gene
expression profiling atlas GSE10246 (Lattin et al., 2008) and visualized using the BioGPS portal
(Wu et al., 2016).
Generation of Gpcpd1 conditional and tissue specific mutant mice
Gpcpd1 conditional knockout mice were derived from targeted ES cells obtained from
EUCOMM (European Conditional Mouse Mutagenesis Program). Exons 9 and exon 10 of the
Gpcpd1 gene were flanked by loxP sites. Cre mediated deletion of the floxed region creates a
frameshift and truncated protein. Upon confirmation of correct targeting by Southern
blotting, the ES cell clone G8 (C57Bl/6N) was injected into C57BL/6J-Tyrc-2J/J blastocysts and
offspring chimeric mice were crossed to C57BL/6J mice. Following germline transmission,
targeted mice were crossed to transgenic mice expressing FLPe recombinase leading to
excision of the NEO cassette. Offspring mice were backcrossed onto a C57Bl/6J genetic
background. Following primers were used to identify the floxed allele:
Forward - GTGCAGGGAACTCAACAACG
Reverse - AGTGATGACAAAGAGGCCAAAAAG
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Subsequently, MckCre, Ap2Cre, Ucp1Cre, and AlbCre mutant animals were crossed to the
Gpcpd1flox/flox mice to generate MckCre-Gpcpd1, Ap2Cre-Gpcpd1, Ucp1Cre-Gpcpd1, and
AlbCre-Gpcpd1 mutant mice.
Mass spectrometry
25µl of each sample containing 100µg protein in RIPA buffer were mixed with 25 µl of 2x lysis
buffer (iST, PO 00065, PreOmics) and sample preparation and tryptic digest were performed
using the iST 96x kit (PO 00027, PreOmics) according to the manufacturer’s description.
750ng of each generated peptide sample was analysed by nanoLC-MS/MS. The nano HPLC
system (UltiMate 3000 RSLC nano system, Thermo Fisher Scientific) was coupled to an Exploris
480 mass spectrometer equipped with a FAIMS pro interfaces and a Nanospray Flex ion
source (all parts Thermo Fisher Scientific). Peptides were loaded onto a trap column (PepMap
Acclaim C18, 5 mm × 300 μm ID, 5 μm particles, 100 Å pore size, Thermo Fisher Scientific) at
a flow rate of 25 μl/min using 0.1% TFA as mobile phase. After 10 minutes, the trap column
was switched in line with the analytical column (a prototype 110 cm µPAC™ Neo HPLC column,
Thermo Fisher Scientific) operated at 30°C. Peptides were eluted using a flow rate of 300
nl/min, starting with the mobile phases 98% A (0.1% formic acid in water) and 2% B (80%
acetonitrile, 0.1% formic acid) and linearly increasing to 35% B over the next 180 minutes.
The Exploris mass spectrometer was operated in data-dependent mode, performing a full
scan (m/z range 350-1200, resolution 60,000, target value 1E6) at 3 different compensation
voltages (CV -45, -60, -75), each followed by MS/MS scans of the most abundant ions for a
cycle time of 0.8 sec per CV. MS/MS spectra were acquired using a collision energy of 30,
isolation width of 1.0 m/z, resolution of 30.000, target value of 1E5 and intensity threshold of
2.5E4, maximum injection time was set to 30 ms. Precursor ions selected for fragmentation
(include charge state 2-6) were excluded for 40 s. The monoisotopic precursor selection filter
and exclude isotopes feature were enabled. For peptide identification, the RAW-files were
loaded into Proteome Discoverer (version 2.5.0.400, Thermo Scientific). All MS/MS spectra
were searched using MSAmanda v2.0.0.16129 (Dorfer V. et al., J. Proteome Res. 2014 Aug
1;13(8):3679-84). The peptide and fragment mass tolerance was set to ±10 ppm, the
maximum number of missed cleavages was set to 2, using tryptic enzymatic specificity
without proline restriction. Peptide and protein identification was performed in two steps.
For an initial search the RAW-files were searched against the database
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

uniprot_reference_mouse_2022-03-04.fasta (21,962 sequences; 11,728,099 residues),
supplemented with common contaminants and sequences of tagged proteins of interest,
using the following search parameters: Iodoacetamide derivative on cysteine was set as a
fixed modification, oxidation of methionine as variable modification. The result was filtered
to 1 % FDR on protein using the Percolator algorithm (Käll L. et al., Nat. Methods. 2007 Nov;
4(11):923-5) integrated in Proteome Discoverer. A sub-database of proteins identified in this
search was generated for further processing. For the second search, the RAW-files were
searched against the created sub-database using the same settings as above plus considering
additional variable modifications: Phosphorylation on serine, threonine and tyrosine,
deamidation on asparagine and glutamine, and glutamine to pyro-glutamate conversion at
peptide N-terminal glutamine, acetylation on protein N-terminus were set as variable
modifications. The localization of the post-translational modification sites within the peptides
was performed with the tool ptmRS, based on the tool phosphoRS (Taus T. et al., J. Proteome
Res. 2011, 10, 5354-62). Identifications were filtered again to 1 % FDR on protein and PSM
level, additionally an Amanda score cut-off of at least 150 was applied. Peptides were
subjected to label-free quantification using IMP-apQuant (Doblmann J. et al., J. Proteome Res.
2019, 18(1):535-541). Proteins were quantified by summing unique and razor peptides and
applying intensity-based absolute quantification (iBAQ, Schwanhäusser B. et al., Nature 2011,
473(7347):337−42) with subsequent normalisation based on the MaxLFQ algorithm (Cox J. et
al., Mol Cell Proteomics. 2014, 13(9):2513-26). Identified proteins were filtered to contain at
least 3 quantified peptide groups. Protein-abundances were normalized using sum
normalization. Ion chromatograms extracted by apQuant were plotted using R.
Quantitative real-time PCR (RT-qPCR)
After animal sacrifice, tissues were extracted, separated into regions and flash frozen in liquid
nitrogen. Subsequently, mRNA was isolated using the RNeasy Lipid Tissue Mini Kit (Qiagen,
GmbH). The RNA concentrations were estimated by measuring the absorbance at 260 nm
using Nanodrop (Thermofisher, GmbH). cDNA synthesis was performed using the iScript
Advanced cDNA Synthesis Kit for RT-qPCR (Bio-Rad, GmbH) following manufacturer’s
recommendations. cDNA was diluted in DNase-free water (1:10) before quantification by real-
time PCR. mRNA transcript levels were measured in triplicate samples per animal using CFX96
touch real-time PCR (Bio-Rad GmbH). Detection of the PCR products was achieved with SYBR
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Green (Bio-Rad, GmbH). At the end of each run, melting curve analyses were performed, and
representative samples of each experimental group were run on agarose gels to ensure the
specificity of amplification. Gene expression was normalized to the expression level of 18S
ribosomal rRNA as the reference gene. The following primers were used:
Murine Gpcpd1:
Forward 5’- GTGGTGCAGGGAACTCAACAACG - 3’
Reverse 5’- TGAGGTCATGATACACCACGGGC - 3’
Murine 18SrRNA:
Forward 5’- GGCCGTTCTTAGTTGGTGGAGCG -3’
Reverse 5’- CTGAACGCCACTTGTCCCTC - 3’
Human Gpcpd1:
Forward 5’- GCATCTGTGGTGCTAGGTGA - 3’
Reverse 5’- TGCCTTGTGAAAAACATGCAG - 3’
Human 18SrRNA:
Forward 5’- GGCCCTGTAATTGGAATGAGTC -3’
Reverse 5’- CCAAGATCCAACTACGAGCTT - 3’
Grip strength analysis
20 month old mice (control and Pcyt2 Myf5 KO mice) were subjected to grip strength tests
using a grip strength meter (Bioseb, USA), following standardized operating procedures. Prior
to tests, mice were single caged for two weeks, in order to avoid any littermate influence on
their performance. Each mouse was tested three times, with a 15-minute inter-trial interval,
and values were averaged among the three trials. Clasping index was evaluated as described
previously 84. Each mouse was scored three times, and an average of scores was calculated.
Glucose tolerance test (GTT), insulin tolerance test (ITT) and insulin measurements
For glucose tolerance tests (GTT), mice were fasted for 16 h and a D-glucose solution was
administered by oral gavage at a dose of 2 g/kg (chow diet), or 1 g/kg (high sugar, high fat
diet). Blood samples were collected from the tail vein and glucose was measured using a
glucometer (Roche, Accu-Chek Performa). For analysis of insulin levels, tail vein blood samples
were added to a NaCl/EDTA solution to avoid blood clotting and plasma insulin concentrations
were determined using the Alpco Mouse Ultrasensitive Insulin ELISA (80-INSMSU-E10). For
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

insulin tolerance tests (ITT), mice were fasted for 6 h and insulin solution (0.375 IU insulin/kg)
administered i.p., followed by blood glucose measurements.
Radioactive 2-deoxyglucose-6-phosphate tissue uptake measurements
To determine 2-DG uptake in various tissues, a bolus injection of 2-[1,2-3H(N)]deoxy-D-
glucose (PerkinElmer, Boston, Massachusetts) (0.4 µCi/g body weight) and insulin (3mU/g
body weight) was injected intraperitoneally in mice (Laurens et al., 2016a, 2016b). Mice were
killed 30 min after injection and tissues were snap-frozen and stored at -80°C until further
processing. Tissue-specific accumulation of 2-deoxyglucose-6-phosphate was determined as
previously described with minor modifications (Kraegen et al., 1985). The total of the [3H]-
radioactivity found in 2-deoxyglucose-6-phosphate was divided by the mean specific activity
of glucose at 30 minutes to obtain the tissue-specific metabolic clearance index (Rg) (µmol
per 100 g of wet tissue per minute).
Determination of muscle glycogen levels
Skeletal muscle tissue (quadriceps) was surgically removed, and flash frozen in liquid nitrogen.
The tissue was crushed on dry ice, weighed, and further processed according to the
manufacturers’ instructions (Sigma Aldrich; #MAK016). Tissue glycogen levels were
determined based on the formulation derived from the standard curve and normalized to the
tissue weight.
Muscle RNA purification
Total RNA was extracted from skeletal muscles using the Qiagen miRNeasy Mini kit (cat#
217004, Qiagen, GmbH) per the manufacturer’s instructions. In brief, samples were
homogenized in QIAzol lysis reagent using a rotor stator homogenizer, and then centrifuged
after chloroform addition at 12,000 g to separate the organic and aqueous phases. Total RNA
was purified from the aqueous phase using the spin column provided in the kit. DNA was
digested on-column per the manufacturer’s instructions. RNA concentration was measured
using the Nanodrop and RNA integrity was measured with an Agilent 2200 Tapestation
instrument (cat#5067–5576 and cat#5067–5577, Agilent, Santa Clara, CA). All samples had
RIN values greater than 8.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Quantseq analysis and sequencing
Libraries were prepared using the QuantSeq 3’mRNA-Seq Library Prep Kit-FWD (cat #15,
Lexogen, Vienna, Austria) using 1 μg of RNA per library, following manufacturer’s instructions.
11 cycles of library amplification were performed with indices from the first two columns of
the i7 Index Plate for QuantSeq/SENSE with Illumina adapters 7001–7096 (cat #044, Lexogen).
Libraries were eluted in 22 μL of the Elution Buffer, and double stranded DNA concentration
was quantified by the KAPA Library Quantification Kit. The molar concentration of cDNA
molecules was calculated from the double stranded DNA concentration and the region
average size (determined by analyzing each sample on an Agilent 2200 Tapestation
instrument (cat#5067–5584 and cat#5067–5585, Agilent, Santa Clara, CA)). Aliquots
containing an equal number of nmoles of cDNA molecules from each library were pooled to
a concentration of 10 nM of cDNA. The final pool was purified once more (to remove any free
primers to prevent index-hopping) by adding 0.9x volumes of PB and proceeding from Step
30 onwards in the QuantSeq User Guide protocol. The library was eluted in 22 μL of the kit’s
Elution Buffer. The pooled libraries were sequenced using an Illumina HiSeq4000 instrument
(Illumina, San Diego, CA).
Transcript coverage
RNA-seq reads were trimmed using BBDuk v38.06 (ref=polyA.fa.gz,truseq.fa.gz k=13 ktrim=r
useshortkmers=t mink=5 qtrim=r trimq=10 minlength=20) and reads mapping to abundant
sequences included in the iGenomes Ensembl GRCm38 bundle (mouse rDNA, mouse
mitochondrial chromosome, phiX174 genome, adapter) were removed using bowtie2
v2.3.4.1 alignment. Remaining reads were analyzed using genome and gene annotation for
the GRCm38/mm10 assembly obtained from Mus musculus Ensembl release 94. Reads were
aligned to the genome using star v2.6.0c and reads in genes were counted with featureCounts
(subread v1.6.2) using strand-specific read counting for QuantSeq experiments (-s 1).
Differential gene expression analysis on raw counts and variance-stabilized transformation of
count data for heatmap visualization were performed using DESeq2 v1.18.1. Functional
annotation overrepresentation analysis of differentially expressed genes was conducted
using clusterprofiler v3.6.0 in R v3.4.1.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Western blot analysis and immunofluorescence
A D-glucose solution was administered by oral gavage and mice were sacrificed after 15
minutes. Muscle tissue (quadriceps) was immediately surgically removed, and flash frozen in
liquid nitrogen. Tissues were further homogenized in RIPA buffer (Sigma; R0278) containing
Halt protease and phosphatase inhibitor cocktail (Thermo Scientific; 78440). Protein levels
were determined using the Bradford assay kit (Pierce, GmbH) and lysates containing equal
amounts of protein were subjected to SDS-PAGE, further transferred to nitrocellulose
membranes. Western blotting was carried out using standard protocols. Blots were blocked
for 1 hour with 5% BSA (Sigma Aldrich; #820204) in TBST (1× TBS and 0.1% Tween-20) and
were then incubated overnight at 4°C with primary antibodies diluted in 5% BSA in TBST
(1:250 dilution). Blots were washed three times in TBST for 15 min, then incubated with HRP-
conjugated secondary antibodies diluted in 5% BSA in TBST (1:5000 dilution) for 45 min at
room temperature, washed three times in TBST for 15 min and visualized using enhanced
chemiluminescence (ECL Plus, Pierce, 1896327). The following primary antibodies were used:
anti-phospho-Insulin Receptor β (Tyr1150/1151) (# 3021 CST, DE, 1:250), anti-total Insulin
Receptor β (# 3025 CST, DE, 1:250), anti-phospho-Akt (Ser473) (# 4060 CST, DE, 1:250), anti-
total Akt (#4685, CS, DE 1:250). Secondary antibodies were anti-rabbit IgG HRP (CST, DE,
#7074). For immunocytochemistry, quadriceps muscles were harvested and fixed in PFA (4%)
for 72h and cryoprotected by further immersing in 30% sucrose solution for another 72 hr.
After embedding in OCT, sections were cut and stained using standard immunohistochemistry
using the following antibodies: anti-Dystrophin (#ab15277, Abcam, UK1:150). For the analysis
of myofiber types, quadriceps muscles were harvested and frozen by immersing the unfixed
quadriceps muscle in OCT and slowly freezing the block in pre-cooled isopentane solution in
liquid nitrogen. Sections were cut and stained using standard immunohistochemistry using
the following antibodies: BA-D5, SC-71 and BF-F3 (DSHB, Iowa, US 1:150) and counterstained
with DAPI.
For analysis of mean fiber area, quadriceps cross-sectional area and mean muscular fiber
number, dystrophin-stained muscle sections were drawn, segmented, and analyzed using Fiji
software.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Electron microcopy:
Six month old animals were sacrificed, muscles were surgically excised, trimmed to small
pieces and immediately immersed in cold 2.5% glutaraldehyde. Muscles were processed for
electron microscopy as described previously (Takeshima et al., 2000). Briefly, tissue was cut
into small pieces and fixed in 2.5% glutaraldehyde in 0.1 mol/l sodium phosphate buffer, pH
7.4. for 1 hour. Subsequently samples were rinsed with the same buffer and post-fixed in 2%
osmium tetroxide in 0.1 mol/l sodium phosphate buffer on ice for 40 min. After 3 rinsing steps
in ddH2O, the samples were dehydrated in a graded series of acetone on ice and embedded
in Agar 100 resin. 70-nm sections were post-stained with 2% uranyl acetate and Reynolds lead
citrate. Sections were examined with a FEI Morgagni 268D (FEI, Eindhoven, The Netherlands)
operated at 80 kV. Images were acquired using an 11 megapixel Morada CCD camera
(Olympus-SIS).
Lipidomics
Quadriceps muscles were isolated from 8-week-old mice and snap frozen in liquid nitrogen.
Muscle tissue was homogenized using a Precellys 24 tissue homogenizer (Precellys CK14
lysing kit, Bertin). Per mg tissue, 3µL of methanol were added. 20 µL of the homogenized
tissue sample was transferred into a glass vial, into which 10 µL internal standard solution
(SPLASH® Lipidomix®, Avanti Polar Lipids) and 120 µL methanol were added. After vortexing,
500 µL Methyl-tert-butyl ether (MTBE) were added and incubated in a shaker for 10 min at
room temperature. Phase separation was performed by adding 145 µL MS-grade water. After
10 min of incubation at room temperature, samples were centrifuged at 1000xg for 10min.
An aliquot of 450 µL of the upper phase (organic) was collected and dried in a vacuum
concentrator. The samples were reconstituted in 200 µL methanol and used for LC-MS
analysis. The LC-MS analysis was performed using a Vanquish UHPLC system (Thermo Fisher
Scientific) combined with an Orbitrap Fusion™ Lumos™ Tribrid™ mass spectrometer (Thermo
Fisher Scientific). Lipid separation was performed by reversed phase chromatography
employing an Accucore C18, 2.6 µm, 150 x 2 mm (Thermo Fisher Scientific) analytical column
at a column temperature of 35oC. As mobile phase A we used an acetonitrile/water (50/50,
v/v) solution containing 10 mM ammonium formate and 0.1 % formic acid. Mobile phase B
consisted of acetonitrile/isopropanol/water (10/88/2, v/v/v) containing 10 mM ammonium
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

formate and 0.1% formic acid. The flow rate was set to 400 µL/min. A gradient of mobile
phase B was applied to ensure optimal separation of the analyzed lipid species. The mass
spectrometer was operated in ESI-positive and -negative mode, capillary voltage 3500 V
(positive) and 3000 V (negative), vaporize temperature 320oC, ion transfer tube temperature
285oC, sheath gas 60 arbitrary units, aux gas 20 arbitrary units and sweep gas 1 arbitrary unit.
The Orbitrap MS scan mode at 120000 mass resolution was employed for lipid detection. The
scan range was set to 250-1200 m/z for both positive and negative ionization mode. The AGC
target was set to 2.0e5 and the intensity threshold to 5.0e3. Data analysis was performed
using the TraceFinder software (ThermoFisher Scientific). Lipidomics results from five
biological replicates per group were analyzed. The amount of each lipid was calculated as
concentration per mg of tissue for all lipid species, measured in a single biological replicate.
Values were next averaged over the five biological replicates for control and MckCre-Gpcpd1
muscle samples, log2 transformed, and compared between the groups using lipidR software.
Glycerophosphocholine and choline targeted metabolomics
Choline and choline glycerophosphate were quantified by reversed phase chromatography
on-line coupled to mass spectrometry, injecting 1 l of the methanolic serum extracts onto a
Kinetex (Phenomenex) C18 column (100 Å, 150 x 2.1 mm) connected with the respective
guard column. Metabolites were separated by employing a 7-minute-long linear gradient
from 96% A (1 % acetonytrile, 0.1 % formic acid in water) to 80% B (0.1 % formic acid in
acetonytrile) at a flow rate of 80 µl/min. On-line tandem mass spectrometry (LC-MS/MS) was
performed using the selected reaction monitoring (SRM) mode of a TSQ Altis mass
spectrometer (Thermo Fisher Scientific), with the transitions m/z 104.1 → m/z 60.1, CE=20
(choline) and m/z 258.1 → m/z 104.1, CE=20 in the positive ion mode. Additionally, several
amino acids (serine, isoleucine, leucine, tryptophan and phenylalanine) were recorded by
respective Selected Reaction Monitoring (SRM) as internal standards. Data interpretation was
performed using TraceFinder (Thermo Fisher Scientific). Authentic metabolite standards were
used for validating experimental retention times by standard addition.
Osmolarity measurements
Quadriceps muscles were isolated from 8-week-old mice, snap frozen in liquid nitrogen,
ground to a powder and weighed. Tissue lysates or cell lysates were prepared using
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

metabolomics lysis buffer (20% water/40% Acetonytril/40% Methanol) to reduce the
interference of proteins on the measurements. The final solution was mixed with deionized
water (1:3 ratio) and measured using K-7400S Semi-Micro Osmometer (Knauer, DE). Obtained
measurements were normalized to tissue weight.
Myofiber glucose uptake
C2C12 were seeded in 24-well dishes (Corning, Sigma-Aldrich, GmbH) and grown in
DMEM/F12 with 10% FCS and penicillin–streptomycin (Life Technologies, final conc., 50 U/mL
of penicillin). Differentiation was achieved by culturing the cells in differentiation media
(DMEM/F12 with 2% horse serum and penicillin–streptomycin), after achieving 80%
confluency. 48 hours after differentiation, in the experiment group differentiation media was
supplemented with 10mM glycerophosphocholine (#20736, Cayman Chemicals, USA). Cells
were incubated further in differentiation media with or without glycerophosphocholine for
additional 7 days, with daily medium changes and intracellular glycerophosphocholine levels
were determined using mass spectrometry as stated above. On the day of the glucose uptake
measurements, differentiated myotubes were washed twice with Hank's Balanced Salt
Solution (Gibco) (HBSS) and further starved in HBSS with or without glycerophosphocholine
(10mM) for 6h. Afterwards, cells were washed twice with HBSS, and incubated with HBSS
containing only D-glucose (15mM) and 200nM insulin (#I0516, Sigma Aldrich, GmbH) for 20
minutes. The solution was washed after indicated time with HBSS, the reaction was stopped
by adding 0.6N HCl, neutralized with 1M Tris solution and lysates were further processed for
measurement using the Glucose-Glo Assay (Promega, GmbH) according to the manufacturer’s
instructions.
Statistical analysis
All data are expressed as mean +/- standard error of the mean (SEM). Statistical significance
was tested by Student’s two tailed, unpaired t-test or two-way ANOVA followed by Sidak’s
multiple comparison test. All figures and statistical analyses were generated using Prism 8
(GraphPad) or R. Details of the statistical tests used are stated in each figure legend. In all
figures, statistical significance is represented as *P <0.05, **P <0.01, ***P <0.001, ****P
<0.0001.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

RESULTS
Aging affects the metabolic Gpcpd1-GPC pathway in the muscles
We first set out to investigate if Gpcpd1 mRNA expression and GPC metabolite levels change
as a result of muscle ageing. We isolated skeletal muscles from adult (6 months) and old (24
months) mice. Gpcpd1 mRNA levels were markedly downregulated, while there was a highly
significant accumulation of GPC in the aged muscles (Figure 1A, B). Further, we mined a
recently reported multi-tissue age-dependent gene expression analysis in rats for potential
changes in Gpcpd1 expression associated with ageing (Shavlakadze et al., 2019). Indeed, we
found Gpcpd1 to be among the topmost significantly reduced genes in muscles with advanced
age, displaying a high level of inverse correlation between Gpcpd1 mRNA levels and age (R2 =
0.5953; P<0.0001) (Figure 1C). This reduction was only evident in the skeletal muscle, while
other tissues tested did not show such age-related Gpcpd1 mRNA decline (Supplementary
Figure 1). These data indicate that Gpcpd1 expression is downregulated in the aged muscle
of rodents, with a significant increase in its catalytic substrate GPC.
Muscle specific Gpcpd1 deficiency does not affect muscle development but causes
hyperglycemia in aged mice
To assess the physiological function of Gpcpd1 in vivo, we generated Gpcpd1flox/flox mice to
specifically delete Gpcpd1 in defined metabolically important tissues that display a high
Gpcpd1 expression (Figure 1D; Supplementary Figure 2). Successful gene targeting was
confirmed by Southern Blot analysis (Figure 1E). As we found that Gpcpd1 expression was
strongly reduced in aged muscles, we first crossed Gpcpd1flox/flox and MckCre transgenic mice
to generate muscle-specific Gpcpd1 deficient offspring, here termed MckCre-Gpcpd1flox/flox
mice. The efficiency of the Gpcpd1 deletion in the muscle of MckCre-Gpcpd1flox/flox animals
was evaluated by RT-PCR. Gpcpd1 mRNA levels were significantly reduced, with some residual
expression most likely resulting from non-myogenic cells (Figure 1F). Subsequent loss of
Gpcpd1 protein in muscles from MckCre-Gpcpd1flox/flox mice was confirmed by mass
spectrometry (Supplementary Figure 3). Contrary to a previous report suggesting that over-
expression of a truncated version of Gpcpd1 results in muscle atrophy (Okazaki et al., 2010),
MckCre-Gpcpd1flox/flox mice exhibited apparently normal muscle development, muscle
morphology, and muscle sizes at 6 month old and 20 months of age (Figure 1G,H,
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 4A). Moreover, the distribution of fiber type, myofiber size, muscle and
mitochondrial ultrastructure, and finally muscle strength of 20 months old MckCre-
Gpcpd1flox/flox mice appeared comparable to control mice (Supplementary Figure 4B-G).
Intriguingly, aged (20 months old) developed fed and fasting hyperglycemia (Figure 1I). At the
same time, there was a significant reduction in muscle glycogen levels (Figure 1J). Blood
insulin levels remained unchanged (Figure 1K). These results indicate that muscle specific loss
of Gpcpd1 has no apparent effect on muscle development or muscle mass in aging, but results
in a muscle-mediated perturbation in glucose metabolism.
Muscle-specific Gpcpd1 mutant mice display glucose intolerance
To directly address muscle glycemic control, we performed a glucose tolerance test in young
(12-week-old) littermate control and MckCre-Gpcpd1flox/flox mice. In contrast to older mice
(Figure 1I) young MckCre-Gpcpd1flox/flox mice did not yet display fasting hyperglycemia
(Supplementary Figure 5A). However, 12-week-old mutant mice already displayed a severe
glucose intolerance (Figure 2A,B) and overall increased insulin release (Figure 2C). The glucose
intolerance was apparent also in the aged (20 months old) MckCre-Gpcpd1flox/flox mice
(Supplementary Figure 5B,C). To trace which organ is responsible for this defect, we bolus-
fed (12-week-old) control and MckCre-Gpcpd1flox/flox mice with radioactive 2-deoxy-D-glucose
(2DG) and measured the uptake in several tissues. 2DG uptake was significantly reduced in
the skeletal muscle (Figure 2D,E) and in the heart muscle (Supplementary Figure 5D,E) By
contrast, 2DG uptake in white fat, subcutaneous fat, brown fat and liver was unaffected in
MckCre-Gpcpd1flox/flox mice (Supplementary Figure 5F). To test whether Gpcpd1 might also
play a role in other metabolically relevant tissues that show high Gpcpd1 expression
(Supplementary Figure 2), we crossed the Gpcpd1flox/flox mice to mouse lines that mediate
deletion in the liver, both white and brown fat, or the brown fat only. However, 12-week-old
liver Gpcpd1-deficient (Figure 2F,G), white and brown fat Gpcpd1-deficient (Figure 2H,I), and
brown fat Gpcpd1-deficient (Figure 2J,K) mice did not exhibit any apparent effects either on
weight gain nor glucose metabolism, as addressed by glucose tolerance tests. These data
show that muscle-specific Gpcpd1 deficiency results in glucose intolerance.
Muscle-specific Gpcpd1 deficiency exacerbates high sugar and high fat diet induced
metabolic syndrome
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

To further examine the metabolic dysfunction that develops in MckCre-Gpcpd1flox/flox mice,
adult (3 months old) mice were first fed a high sugar diet (30% fructose in drinking water).
There were no detectable differences in weight gain between controls and MckCre-
Gpcpd1flox/flox mice over the 10 weeks observation period (Figure 3A). However, MckCre-
Gpcpd1flox/flox mice displayed a severe defect in glucose clearance, as evaluated by a glucose
tolerance test (Figure 3B,C). We next fed mice a high fat diet; again we observed no effects
on weight gain but a severe glucose intolerance of the MckCre-Gpcpd1flox/flox mice as
compared to their control littermates (Figure 3D,F). Insulin release appeared unaffected
under both high sugar and high fat diet conditions (Figure 3G,H). Compared to the standard
diet, high sugar and high fat diet tended to worsen the ability to regulate glucose metabolism
of adult MckCre-Gpcpd1flox/flox mice (Figure 3I). These data show that inactivation of Gpcpd1
in muscle does not affect weight gain under high sugar and fat diets but exacerbates glucose
intolerance in adult MckCre-Gpcpd1flox/flox mice.
Muscle-specific Gpcpd1 deficiency results in accumulation of glycerophosphocholine
Since Gpcpd1 is responsible for hydrolysis of glyceroposphocholine (GPC), we reasoned that
muscle Gpcpd1 deficiency would result in accumulation of GPC. To address this, we
performed targeted lipidomic analysis on quadriceps muscle isolated from control and
MckCre-Gpcpd1flox/flox mice. As expected, there was a significant accumulation of GPC in the
muscle of MckCre-Gpcpd1flox/flox mice (Figure 4A). Besides the marked increase in GPC, there
were no significant changes in the global lipidome in the muscles from MckCre-Gpcpd1flox/flox
mice (Supplementary Figure 6A), indicating that loss of Gpcpd1 specifically impairs
degradation of GPC. Choline levels remained unchanged (Supplementary Figure 6B), which
could be explained because choline is mainly provided in the food (Fagone and Jackowski,
2013). As GPC acts as an osmolyte (Gallazzini et al., 2008), we addressed whether the aberrant
accumulation of GPC in muscles from MckCre-Gpcpd1flox/flox mice changes tissue osmolarity.
Indeed, the osmolarity was significantly increased in muscles from MckCre-Gpcpd1flox/flox mice
compared to control mice (Figure 4B). Similar changes were observed in muscles isolated
from aged (24 months old) mice when compared to adult (6 months old) wild type mice
(Figure 4C). To address whether the increased intracellular accumulation of
glycerophosphocholine can directly affect glucose uptake of muscle cells, we exposed C2C12-
derived myotubes for several days with glycerophosphocholine (GPC) supplemented media.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Using mass spectrometry, we confirmed that this treatment increased intracellular GPC levels
in C2C12-derived myotubes (Supplementary Figure 6C), with a trend towards increased
osmolarity in GPC treated myotubes (P=0.074) (Supplementary Figure 6D). Importantly,
intracellular accumulation of GPC resulted in significantly reduced glucose uptake of C2C12-
derived myotubes after 16 h starvation (Figure 4D). Taken together, inactivation of Gpcpd1 in
the muscle results in marked and specific accumulation of the lipid GPC and a change in
muscle osmolarity and glucose uptake.
Muscle-specific Gpcpd1 deficiency impairs insulin signaling
To obtain an overview of transcriptional changes occurring in muscles upon Gpcpd1
deficiency, we performed genome wide mRNA-sequencing analysis (Quant-Seq) to compare
the skeletal muscle gene expression profiles from young (12 weeks old) MckCre-Gpcpd1 and
control littermate mice (Figure 4E). Overall, 894 genes were significantly changed in skeletal
muscles of MckCre-Gpcpd1flox/flox mice (401 down-regulated, 493 up-regulated; adjusted p-
value < 0.01). Interestingly, a large set of significantly dysregulated genes (Figure 4E and
Supplementary Figure 7C) were also the top-most dysregulated genes previously associated
with muscle ageing (Shavlakadze et al., 2019). This data indicates that, at least partially, young
MckCre-Gpcpd1flox/flox mice display an” aged like” muscle transcriptional profile. Functional
enrichment analysis of down-regulated genes in muscles of MckCre-Gpcpd1flox/flox mice
revealed an enrichment for pathways linked to glucose and carbohydrate metabolism, and
insulin signaling (adjusted p-value < 0.05) (Figure 4F, Supplementary Figure 8A). Indeed,
MckCre-Gpcpd1flox/flox mice displayed reduced insulin sensitivity as determined by an insulin
tolerance test (ITT) (Figure 4G,H). Moreover, upon glucose challenge in vivo, we observed
impaired insulin signaling as addressed by decreased phosphorylation of the insulin receptor
beta (IR) and Akt (Figure 4I,J). Thus, inactivation of Gpcpd1 in the muscle results in altered
gene expression profiles associated with metabolic syndrome and aging and impaired Insulin
receptor signaling.
Changes of GPCPD1 and GPC levels in aging and type 2 diabetes in humans
We finally investigated how our findings relate to human physiology and aging. We found that
in the skeletal muscles of otherwise healthy, middle aged/old (49-62 years of age) individuals,
there was a significant reduction in Gpcpd1 mRNA expression (Figure 5A). In parallel, there
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

was a highly significant increase of GPC and GPC/PDE levels in skeletal muscles of sedentary
aged individuals (Figure 5B). A significant positive correlation of higher GPC and GPC-PDE
levels was only evident with aging (Figure 5C,D), while there was no correlation of higher GPC
and GPC/PDE levels with increased fat mass, muscle mass and body mass index
(Supplementary Figure 9). Thus, the muscle GPC and GPC-PDE levels gradually increase with
advancing age. As we observed that deficiency of Gpcpd1 in mice results in impaired glucose
metabolism, we also addressed if hyperglycemia in type 2 diabetic patients is associated with
perturbed GPC metabolism. Indeed, there was a significant accumulation of GPC and GPC-
PDE in skeletal muscles of diabetic individuals compared to age, gender and BMI matched
controls (Figure 5E), while several other metabolites remained unchanged (Figure 5F). Taken
together, our data show that the GPCPD1-GPC metabolic axis is dysregulated with aging in
humans, and that GPC accumulates in muscles of type 2 diabetes patients.
Discussion
Muscle aging is accompanied by a myriad of molecular and metabolic changes. Yet, it still
remains elusive which of these perturbations are causative to the aging process, affecting
both the muscle and systemic health (Demontis et al., 2013b). Here, we report that the
Gpcpd1-GPC metabolic axis is perturbed with aging in muscles of rodents and humans. Based
on these observations, we generated conditional and subsequently tissue specific Gpcpd1
deficient mice to investigate if the observed age-related changes in this metabolic pathway
affect systemic health. We found that muscle specific Gpcpd1 deficiency results in a severe
glucose intolerance, a disorder that is commonly seen in the elderly (Chia et al., 2018). These
data indicate that age-related perturbations of the Gpcpd1-GPC metabolic pathway is
involved in altered glucose metabolism in aging.
Gpcpd1 deficiency has no apparent effect on body weight, muscle development, nor muscle
mass even in old age. The apparently normal muscle mass and functionality in our model
contrasts with a previous report where overexpression of a truncated Gpcpd1 protein in vivo
resulted in muscle atrophy (Okazaki et al., 2010). This could be attributed to high levels of
transgene expression, which can result in cell and tissue toxicity (Bolognesi and Lehner, 2018;
Kulak et al., 2014; Moriya, 2015), not excluding other reasons for this previously observed
phenotype such as dominant-negative effects of the overexpressed protein in defined
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

pathways. We also deleted Gpcpd1 specifically in several other metabolically important
mouse tissues that exhibit Gpcpd1 expression, namely the liver and brown and white fat.
However, we never observed any gross-developmental defects nor impairment in glucose
metabolism. Thus, our results show that muscle-regulated glucose homeostasis is critically
dependent on Gpcpd1 which could be, in part, explained by the relatively high Gpcpd1
expression in the muscles.
Both skeletal and heart muscle displayed impaired glucose uptake in the MckCre-
Gpcpd1flox/flox mice. Given the relative organ to body weight, it is conceivable that the impaired
glucose uptake in the skeletal muscle is mainly responsible for the systemic glucose
metabolism perturbation in MckCre-Gpcpd1flox/flox mice. Future studies could reveal how
Gpcpd1 loss only in the heart muscle affects heart function. Intriguingly, once we further
generated Gpcpd1 deficient mice in several tissues with a relatively high Gpcpd1 expression
and that are also essential to whole-body metabolic homeostasis, we did not observe any
gross-developmental defects or impairment in glucose metabolism. Since Gpcpd1 has the
highest expression levels in the striated muscles, we infer that the muscle tissue is specifically
dependent on Gpcpd1 and its hydrolysis activity on GPC. A previous study has shown
beneficial effects of GPC supplementation on the brain as well as cartilage tissue though the
effect on the metabolic homeostasis was never tested (Matsubara et al., 2018). This could
further indicate the muscle specific effects of the perturbation of the Gpcpd1-GPC metabolic
pathway.
Mechanistically, our data show that Gpcpd1 deficiency and increased GPC impaired muscle
glucose uptake by interfering with glucose metabolism and insulin signaling. How would high
levels of GPC, the specifically increased substrate of Gpcpd1, affect these pathways? It has
been shown that the excess accumulation of other lipid species in the skeletal muscle, such
as diacylglycerol, triacylglycerol, and ceramides, impairs insulin sensitivity and glucose uptake
by perturbing mitochondrial beta oxidation and triggering an inflammatory program (Park
and Seo, 2020; Wu and Ballantyne, 2017). It is possible that increased GPC level exerts similar
effects, although we did not observe any signs of muscle inflammation in our MckCre-
Gpcpd1flox/flox mice. At the transcriptional level, one of the most affected genes in the muscles
from MckCre-Gpcpd1flox/flox mice in the insulin signaling network was Mss51 (Mitochondrial
Translational Activator; also known as Zymnd17). Interestingly besides Gpcpd1, Mss51 is also
one of the top-most downregulated genes in skeletal muscles with aging (Supplemental
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 7) (Shavlakadze et al., 2019). Disruption of Mss51 in C2C12 using Cas9 system resulted
in increased cellular ATP levels, upregulated glycolysis and oxidative phosphorylation (Rovira
Gonzalez et al., 2019). However, in vivo disruption of Mss51 resulted in opposing effects in
regard to glucose metabolism and insulin responses (Fujita et al., 2018; Rovira Gonzalez et al.,
2019). MckCre-Gpcpd1flox/flox mice exhibit phenotype of severe glucose intolerance and
impaired insulin signaling, without displaying any signs of mitochondrial dysfunction both at
the transcriptional as well as ultrastructural levels. Further studies will be needed to elucidate
which transcriptional or epigenetic mechanism is modulated due to Gpcpd1 loss and
subsequent GPC accumulation. Interestingly, it has been previously reported that prolonged
exposure to osmotic shock severely inhibits insulin stimulated glucose uptake and insulin
signaling in adipose cells (Chen et al., 1999; Gual et al., 2003; Stookey et al., 2004). GPC acts
as an osmolyte in kidney cells (Gallazzini et al., 2008). We detected increased osmolarity both
in the muscle tissue of Gpcpd1 mutant mice and in older wild type animals, while exposure of
myofibers to GPC indeed impaired glucose transport. Therefore it is plausible that a chronic
exposure to GPC changes this physicochemical property of muscle cells, contributing to the
aging-induced impaired glucose homeostasis (Chia et al., 2018).
Importantly, muscles from aged humans and from patients with type 2 diabetes, displayed
persistently increased levels of GPC, with highly significant positive correlation between high
muscle GPC levels and aging. Our findings highlight the key role for the GPCPD1-GPC
metabolic pathway in aging and the crosstalk between lipid metabolism in the muscle and
systemic glucose tolerance. Whether this pathway can be therapeutically exploited to
ameliorate the muscle-mediated dysregulation of metabolic glucose homeostasis in the
elderly needs to be further explored.
Conclusions. Our data provide evidence that ageing results in a perturbed Gpcpd1-GPC
metabolic axis in the muscle. Muscle specific Gpcpd1 deficient mice display systemic glucose
intolerance, without apparently affecting muscle development and size. Assessing tissue
specific deletion of Gpcpd1 in several organs that are of key importance to whole-body
metabolic glucose control, we further show that this effect on glucose intolerance and
hyperglycemia appears to be muscle specific. Mechanistically, we find that specific
accumulation of GPC due to loss of Gpcpd1 in the muscle perturbs muscle carbohydrate
metabolism, myofiber osmolarity, and insulin signaling and is associated with an “aged like”
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

transcriptional profile. Finally, we report that in the muscles of aged humans as well as of type
2 diabetes patients, the GPCPD1-GPC axis is markedly changed. Thus, in mice and men we
have identified a critical pathway in muscle that regulates systemic glucose homeostasis in
aging.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Author contributions
D.C. and J.M.P. conceived, coordinated, and designed the study. D.C. and M.O. performed
experiments and analyzed the data with contributions from M.L. and S.J.F. M.N. performed
bioinformatic analysis. A.H. assisted in tissue sampling. E.R. and T.G. collected human muscle
biopsies. L.P., R.K. and M.Krs. performed and analyzed in vivo MRS measurements. B.L., M.L.,
C.B., C.K., G.T., V.B., and C.M. performed glucose uptake experiments. G.D. and E.R.
performed mass spectrometry experiments. M.L., M.K., M.Krs. and A.K.-W. designed,
coordinated and oversaw the human T2DM experiments. D.C. and J.M.P. wrote the
manuscript. All authors edited the manuscript and approved the final manuscript.
Acknowledgements
We would like to thank all members of our laboratories for helpful discussions. We are
grateful to Transgenic unit, Comparative medicine and Metabolomics unit from Vienna
Biocenter Core Facilities for their service. We would also like to thank M.G. from the
Histopathology unit in Vienna Biocenter Core Facilities for histological processing. We also
than the lipidomic service of Center for Molecular Medicine (CEMM). We would like to thank
Dr. Patrik Krumpolec from the Biomedical Research Center, Institute of Experimental
Endocrinology, Slovak Academy of Sciences, Bratislava, Slovakia, for the processing of MRS
data from young and senior participants. We would like to thank Prof. S. Trattnig, High Field
MR Centre, Department of Biomedical Imaging and Image guided Therapy for logistic support.
We would like to thank Prof. James D Johnson from University of British Columbia, Vancouver,
Canada for critical reading of the manuscript.
J.M.P. is supported by IMBA, a Wittgenstein award, the T. von Zastrow foundation, and a
Canada 150 Research Chair in functional genetics. D.C. is supported by the Austrian Academy
of Sciences and the T. von Zastrow foundation. M. Krs., R. K. and L. P. are supported by
Austrian Science Foundation (KLI-904-B).
Conflict of interest
The authors have declared that no conflict of interest exists.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

References
Baskin, K.K., Winders, B.R., and Olson, E.N. (2015). Muscle as a “mediator” of systemic metabolism.
Cell Metab. 21.
Bergström, J., and Hultman, E. (1967). A study of the glycogen metabolism during exercise in man.
Scand. J. Clin. Lab. Invest.
Bolognesi, B., and Lehner, B. (2018). Protein Overexpression: Reaching the limit. Elife 7.
Chen, D., Fucini, R. V., Olson, A.L., Hemmings, B.A., and Pessin, J.E. (1999). Osmotic Shock Inhibits
Insulin Signaling by Maintaining Akt/Protein Kinase B in an Inactive Dephosphorylated State. Mol.
Cell. Biol. 19.
Chia, C.W., Egan, J.M., and Ferrucci, L. (2018). Age-related changes in glucose metabolism,
hyperglycemia, and cardiovascular risk. Circ. Res. 123.
Corda, D., Mosca, M.G., Ohshima, N., Grauso, L., Yanaka, N., and Mariggiò, S. (2014). The emerging
physiological roles of the glycerophosphodiesterase family. FEBS J. 281.
Cowie, C.C., Rust, K.F., Ford, E.S., Eberhardt, M.S., Byrd-Holt, D.D., Li, C., Williams, D.E., Gregg, E.W.,
Bainbridge, K.E., Saydah, S.H., et al. (2009). Full accounting of diabetes and pre-diabetes in the U.S.
population in 1988-1994 and 2005-2006. Diabetes Care 32.
Cruz-Jentoft, A.J., and Sayer, A.A. (2019). Sarcopenia. Lancet 393.
DeFronzo, R.A., Jacot, E., Jequier, E., Maeder, E., Wahren, J., and Felber, J.P. (1981). The effect of
insulin on the disposal of intravenous glucose. Results from indirect calorimetry and hepatic and
femoral venous catheterization. Diabetes 30.
Demontis, F., Piccirillo, R., Goldberg, A.L., and Perrimon, N. (2013a). Mechanisms of skeletal muscle
aging: Insights from Drosophila and mammalian models. DMM Dis. Model. Mech. 6.
Demontis, F., Piccirillo, R., Goldberg, A.L., and Perrimon, N. (2013b). The influence of skeletal muscle
on systemic aging and lifespan. Aging Cell 12.
Fagone, P., and Jackowski, S. (2013). Phosphatidylcholine and the CDP-choline cycle. Biochim.
Biophys. Acta - Mol. Cell Biol. Lipids 1831.
Fujita, R., Yoshioka, K., Seko, D., Suematsu, T., Mitsuhashi, S., Senoo, N., Miura, S., Nishino, I., and
Ono, Y. (2018). Zmynd17 controls muscle mitochondrial quality and whole-body metabolism. FASEB
J. 32.
Gallazzini, M., Ferraris, J.D., and Burg, M.B. (2008). GDPD5 is a glycerophosphocholine
phosphodiesterase that osmotically regulates the osmoprotective organic osmolyte GPC. Proc. Natl.
Acad. Sci. U. S. A. 105.
Gual, P., Gonzalez, T., Grémeaux, T., Barrés, R., Le Marchand-Brustel, Y., and Tanti, J.F. (2003).
Hyperosmotic stress inhibits insulin receptor substrate-1 function by distinct mechanisms in 3T3-L1
adipocytes. J. Biol. Chem. 278.
Houtkooper, R.H., Argmann, C., Houten, S.M., Canto, C., Jeninga, E.H., Andreux, Ṕeńelope A.,
Thomas, C., Doenlen, R., Schoonjans, K., and Auwerx, J. (2011). The metabolic footprint of aging in
mice. Sci. Rep. 1.
Jue, T., Rothman, D.L., Shulman, G.I., Tavitian, B.A., DeFronzo, R.A., and Shulman, R.G. (1989). Direct
observation of glycogen synthesis in human muscle with 13C NMR. Proc. Natl. Acad. Sci. U. S. A. 86.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Kern, M., Dolan, P.L., Mazzeo, R.S., Wells, J.A., and Dohm, G.L. (1992). Effect of aging and exercise on
GLUT-4 glucose transporters in muscle. Am. J. Physiol. - Endocrinol. Metab. 263.
Kraegen, E.W., James, D.E., Jenkins, A.B., and Chisholm, D.J. (1985). Dose-response curves for in vivo
insulin sensitivity in individual tissues in rats. Am. J. Physiol. - Endocrinol. Metab. 11.
Krumpolec, P., Klepochová, R., Just, I., Tušek Jelenc, M., Frollo, I., Ukropec, J., Ukropcová, B.,
Trattnig, S., Krššák, M., and Valkovič, L. (2020). Multinuclear MRS at 7T Uncovers Exercise Driven
Differences in Skeletal Muscle Energy Metabolism Between Young and Seniors. Front. Physiol. 11.
Kulak, N.A., Pichler, G., Paron, I., Nagaraj, N., and Mann, M. (2014). Minimal, encapsulated
proteomic-sample processing applied to copy-number estimation in eukaryotic cells. Nat. Methods
11.
Lattin, J.E., Schroder, K., Su, A.I., Walker, J.R., Zhang, J., Wiltshire, T., Saijo, K., Glass, C.K., Hume, D.A.,
Kellie, S., et al. (2008). Expression analysis of G Protein-Coupled Receptors in mouse macrophages.
Immunome Res. 4.
Laurens, C., Badin, P.M., Louche, K., Mairal, A., Tavernier, G., Marette, A., Tremblay, A., Weisnagel,
S.J., Joanisse, D.R., Langin, D., et al. (2016a). G0/G1 Switch Gene 2 controls adipose triglyceride
lipase activity and lipid metabolism in skeletal muscle. Mol. Metab. 5.
Laurens, C., Bourlier, V., Mairal, A., Louche, K., Badin, P.M., Mouisel, E., Montagner, A., Marette, A.,
Tremblay, A., Weisnagel, J.S., et al. (2016b). Perilipin 5 fine-tunes lipid oxidation to metabolic
demand and protects against lipotoxicity in skeletal muscle. Sci. Rep. 6.
Matsubara, K., Okuda, M., Shibata, S., Miyaki, S., Ohkubo, T., Izu, H., and Fujii, T. (2018). The delaying
effect of alpha-glycerophosphocholine on senescence, transthyretin deposition, and osteoarthritis in
senescence-accelerated mouse prone 8 mice. Biosci. Biotechnol. Biochem. 82.
Moriya, H. (2015). Quantitative nature of overexpression experiments. Mol. Biol. Cell 26.
Okazaki, Y., Ohshima, N., Yoshizawa, I., Kamei, Y., Mariggiò, S., Okamoto, K., Maeda, M., Nogusa, Y.,
Fujioka, Y., Izumi, T., et al. (2010). A novel glycerophosphodiester phosphodiesterase, GDE5, controls
skeletal muscle development via a non-enzymatic mechanism. J. Biol. Chem. 285.
Park, S.S., and Seo, Y.K. (2020). Excess accumulation of lipid impairs insulin sensitivity in skeletal
muscle. Int. J. Mol. Sci. 21.
Pilling, L.C., Atkins, J.L., Bowman, K., Jones, S.E., Tyrrell, J., Beaumont, R.N., Ruth, K.S., Tuke, M.A.,
Yaghootkar, H., Wood, A.R., et al. (2016). Human longevity is influenced by many genetic variants:
Evidence from 75,000 UK Biobank participants. Aging (Albany. NY). 8.
Rovira Gonzalez, Y.I., Moyer, A.L., LeTexier, N.J., Bratti, A.D., Feng, S., Sun, C., Liu, T., Mula, J., Jha, P.,
Iyer, S.R., et al. (2019). Mss51 deletion enhances muscle metabolism and glucose homeostasis in
mice. JCI Insight 4.
Shavlakadze, T., Morris, M., Fang, J., Wang, S.X., Zhu, J., Zhou, W., Tse, H.W., Mondragon-Gonzalez,
R., Roma, G., and Glass, D.J. (2019). Age-Related Gene Expression Signature in Rats Demonstrate
Early, Late, and Linear Transcriptional Changes from Multiple Tissues. Cell Rep.
Stewart, J.D., Marchan, R., Lesjak, M.S., Lambert, J., Hergenroeder, R., Ellis, J.K., Lau, C.H., Keun, H.C.,
Schmitz, G., Schiller, J., et al. (2012). Choline-releasing glycerophosphodiesterase EDI3 drives tumor
cell migration and metastasis. Proc. Natl. Acad. Sci. U. S. A. 109.
Stookey, J., Pieper, C., and Cohen, H. (2004). Hypertonic hyperglycemia progresses to diabetes faster
than normotonic. Eur. J. Epidemiol. 19.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Takeshima, H., Komazaki, S., Nishi, M., Iino, M., and Kangawa, K. (2000). Junctophilins: A novel family
of junctional membrane complex proteins. Mol. Cell.
Valkovič, L., Ukropcová, B., Chmelík, M., Baláž, M., Bogner, W., Schmid, A.I., Frollo, I., Zemková, E.,
Klimeš, I., Ukropec, J., et al. (2013). Interrelation of 31P-MRS metabolism measurements in resting
and exercised quadriceps muscle of overweight-to-obese sedentary individuals. NMR Biomed. 26.
Wu, H., and Ballantyne, C.M. (2017). Skeletal muscle inflammation and insulin resistance in obesity.
J. Clin. Invest. 127.
Wu, C., Jin, X., Tsueng, G., Afrasiabi, C., and Su, A.I. (2016). BioGPS: Building your own mash-up of
gene annotations and expression profiles. Nucleic Acids Res. 44.
Yanagi, N., Koike, T., Kamiya, K., Hamazaki, N., Nozaki, K., Ichikawa, T., Matsunaga, A., Kuroiwa, M.,
and Arai, M. (2021). Assessment of sarcopenia in the intensive care unit and 1-year mortality in
survivors of critical illness. Nutrients 13.
Yang, M., Jiang, J., Zeng, Y., and Tang, H. (2019). Sarcopenia for predicting mortality among elderly
nursing home residents: SARC-F versus SARC-CalF. Med. (United States) 98.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure legends
Figure 1. Muscle-specific Gpcpd1 deficiency causes hyperglycemia in old mice.
A) mRNA levels in quadriceps muscles of adult (6 months old) and old (24 months old) mice.
N=5 per group. B) Glycerophosphocholine (GPC) levels in quadriceps muscles of young/adult
(6 months old) and old (24 months old) mice. N=5-10 per group. C) Correlation of mRNA
expression of Gpcpd1 in rat muscles with age. D) Schematic outline and E) Southern blot
validation of successful generation of a conditional Gpcpd1 allele in mice. F) Validation of
Gpcpd1 mRNA deletion in muscles from muscle-specific Gpcpd1 mutant mice (MckCre-
Gpcpd1flox/flox). N=6 per group. G) Bodyweight of 20 months old control and littermate MckCre-
Gpcpd1flox/flox mice. N=7-8 per group. H) Weights of skeletal muscle (quadriceps, QA;
gastrocnemicus, GC; and Tibialis anterior, TA) isolated from 20 months old control and Mck-
CreGpcpdflox/flox mice. N=7 per group. I) Blood glucose levels of 20 months old littermate
control and MckCre-Gpcpd1flox/flox mice under fasted and re-fed states (2h after re-feeding;
standard diet). N=6-7 per group. J) Skeletal muscle (quadriceps) glycogen levels in 20 months
old control and Mck-CreGpcpd1flox/flox mice. N=6. K) Blood insulin levels of standard diet fed 20
months old control and MckCre-Gpcpd1flox/flox mice. N=6-7 per group. Unless otherwise stated,
each dot represents an individual mouse. Data are shown as means ± SEM. Student’s two
tailed, unpaired t-test was used for statistical analysis; ns, not significant; *p < 0.05; **p <
0.01; ***p < 0.001, ****p<0.0001.
Figure 2. Muscle Gpcpd1 deficiency causes glucose intolerance in young mice.
A) Blood glucose levels and B) Area under curve (AUC) after an oral glucose tolerance test
(OGTT) in 12 weeks old months old control and MckCre-Gpcpd1flox/flox mice. N=16-17 per
group. Student’s two tailed un-paired t test with Welch correction was used for AUC statistical
analysis. C) Blood insulin levels during OGTT in 12 weeks old control and MckCre-Gpcpd1flox/flox
mice. N=12 per group. D) and E) Radioactive 2DG glucose uptake in skeletal muscles (soleus)
of 12 weeks old control and Mck-CreGpcpd1flox/flox mice after bolus glucose feeding. F) Body
weights of 3 months old control and hepatocyte-specific Gpcpd1 deficient (AlbCre-
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Gpcpd1flox/flox) littermates. N=17-20 per group. F) Blood glucose levels after OGTT test in 2-3
months old control and AlbCre-Gpcpd1flox/flox mice. N=17-20 per group. G) Body weights of 3
months old control and white and brown fat-specific Gpcpd1 deficient (Ap2Cre-Gpcpd1flox/flox)
littermates. N=4-6 per group. H) Blood glucose levels after OGTT in 2-3 months old control
and Ap2Cre-Gpcpd1flox/flox mice. N=4-6 per group. I) Body weights of 3 months old control and
brown fat-specific Gpcpd1 deficient (Ucp1Cre-Gpcpd1flox/flox) littermates. N=4-6 per group. J)
Blood glucose levels after OGTT test in 2-3 months old control and Ucp1Cre-Gpcpd1flox/flox
mice. N=4-6 per group. Unless otherwise stated, each dot represents an individual mouse.
Data are shown as means ± SEM. Repeated measures Two-way ANOVA followed by Sidak’s
multiple comparison test was used for statistical analysis unless otherwise stated; ns, not
significant; *p < 0.05; **p < 0.01; ****p < 0.0001.
Figure 3. Muscle specific Gpcpd1 deficiency exacerbates high sugar and high fat
diet induced metabolic syndrome.
A) Body weight gain of 3 months old control and MckCre-Gpcpd1flox/flox mice on high
carbohydrate diet (HCD; 30% fructose). N=7 per group. B) Blood glucose levels after OGTT in
and C) Area under curve (AUC) after 3 months old control and MckCre-Gpcpd1flox/flox mice fed
a HCD diet. N=7 per group. Student’s two tailed un-paired t test with Welch correction was
used for AUC statistical analysis. D) Body weight gain of 3 months old control and MckCre-
Gpcpd1flox/flox mice on high fat diet (HFD; 60% fat). N=9 per group. E) Blood glucose levels after
OGTT and F) Area under curve (AUC) after in 3 months old control and MckCre-Gpcpd1flox/flox
littermate mice fed a HFD diet. N=9 per group. Student’s two tailed un-paired t test with Welch
correction was used for AUC statistical analysis. F) Blood insulin levels during OGTT in 3
months old control and MckCre-Gpcpdflox/flox mice fed HCD. N=7 per group. G) Blood insulin
levels during OGTT in 3 months old control and MckCre-Gpcpd1 mice fed HFD. N=6 per group.
H) Blood insulin levels during OGTT in 3 months old control and MckCre-Gpcpd1 mice fed HFD.
N=11-12 per group. I) Area under the curve (AUC) comparison during oral glucose tolerance
test (OGTT) of 12 weeks old control and Mck-CreGpcpd1flox/flox mice that were fed either
standard (chow diet), high sugar diet (HCD), and high fat diet (HFD). Multiple comparisons
One-way ANOVA was used for statistical analysis. The effect of diets was compared to the AUC
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

of control mice (dashed line) on standard diet after OGTT. Unless otherwise stated, each dot
represents an individual mouse. Data are shown as means ± SEM. Repeated measures Two-
way ANOVA followed by Sidak’s multiple comparison test was used for statistical analysis
unless otherwise stated; ns, not significant; *p < 0.05, **p < 0.01.
Figure 4. Muscle aging gene expression signatures and impaired Insulin
receptor signaling upon Gpcpd1 loss in the skeletal muscle.
A) Glycerophosphocholine (GPC) levels in skeletal muscle (quadriceps) of 3 months old control
and MckCre-Gpcpd1flox/flox mice. N=10-12 per group. B) Skeletal muscle tissue osmolarity
(quadriceps) of 3 months old control and MckCre-Gpcpd1flox/flox mice. N=6-7 per group. C)
Skeletal muscle tissue osmolarity (quadriceps) of 6 months old and 24 months old wild type
mice. N=5-10 per group. Each dot represents one individual animal. D) Glucose uptake in
C2C12 differentiated myofibers without or with GPC pretreatment. Each dot represents one
myofiber culture. E) Quantseq transriptomic analysis in skeletal muscle (gastrocnemicus) of 3
months old control and MckCre-Gpcpd1flox/flox littermate mice. N=4 per group. F) Gene
ontology analysis of down-regulated genes in skeletal muscle from 3 months old MckCre-
Gpcpd1flox/flox mice, compared to control littermates. G) Insulin tolerance test and H) area
under the curve (AUC) of 3 months old control and MckCre-Gpcpd1flox/flox mice. N=7-8 per
group. Mice were fasted 6h before the test. Student’s two tailed un-paired t test with Welch
correction was used for AUC statistical analysis. I) and J) Phsopho-Insulin receptor beta (IR)
and phospho-Akt levels in skeletal muscles (quadriceps) from 3 months old control and
MckCre-Gpcpd1flox/flox mice after overnight fasting and 15 minutes after oral glucose delivery
(bolus 1g glucose/kg of body weight). Total IR levels, phosphorylated IR
(phosphoTyr1150/Tyr1151 IR total Akt and phosphorylated Akt are shown. GAPDH is shown
as a loading control. N=3 mice per group, per treatment. Each dot represents one myofiber
culture. Unless otherwise stated, each dot in A-C represents an individual mouse. Data are
shown as means ± SEM. Repeated measures Two-way ANOVA followed by Sidak’s multiple
comparison test and Student’s two tailed, unpaired t-test with Welch correction was used for
statistical analysis; ns, not significant. *p < 0.05, ****p<0.0001.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 5. The metabolic GPCPD1-GPC axis is perturbed with aging and type 2
diabetes in humans.
A) Gpcpd1 mRNA levels in quadriceps muscles in otherwise healthy young (20-30yrs) and old
(49-65) humans. Relative mRNA expression normalized to 18S RNA is shown. B)
Representative 31P-magnetic resonance (MR) spectra, and quantification of GPC and GPC-PDE
levels acquired from the skeletal muscle (gastrocnemius medialis) from young (20-30yrs) and
senior (49-65) participants. C) GPC and GPC-PDE levels plotted against age for each individual.
Linear-regression analysis was used for R2 and P values calculation. E) Representative 31P-MR
spectra, and quantification of GPC and GPC-PDE levels acquired from the skeletal muscle
(gastrocnemius medialis) from healthy volunteers and type 2 diabetes patients (see Methods
for cohort description). F) Levels of phosphocreatine (PCr), inorganic phosphate (Pi),
phosphodiester (PDE), phosphocholine (PC), phosphomonoesthers (PME) acquired from the
skeletal muscle from healthy volunteers and type 2 diabetes patients. Unless otherwise
stated, each dot represents an individual human. Data are shown as means ± SEM. Student’s
two tailed, unpaired t-test with Welch correction was used for statistical analysis unless
otherwise stated; ns, not significant; *p < 0.05, ***p < 0.001, ****p < 0.0001.
Supplementary Figure legends
Supplementary Figure 1. Assessment of age-related Gpcpd1 levels in rat
tissues.
Correlation of mRNA expression of Gpcpd1 in the indicated rat tissues with age. Data are
shown as means ± SEM. Student’s two tailed, unpaired t-test was used for statistical analysis
unless otherwise stated. There were no statistically significant differences.
Supplementary Figure 2. Gpcpd1 mRNA expression across different cell and
tissue type.
Gpcpd1 mRNA transcript per million in different tissues and cell types (http//biogps.org).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 3. Validation of Gpcpd1 protein disruption in MckCre-
Gpcpd1 mice by mass spectrometry
Representative extracted ion chromatograms (XICs) of Gpcpd1 peptides from skeletal muscle
proteome of control and MckCre-Gpcpd1 mice. Integrated signal used for quantification is
indicated in grey. Peptide identifications are represented by red dashed lines. XICs from N=3
mice per group are shown.
Supplementary Figure 4. Grip strength, muscle weights, fiber type, and muscle
ultrastructure assessment of Mck-CreGpcpd1flox/flox mice.
A) Skeletal muscle weights (quadriceps, QA; gastrocnemicus, GC; and Tibialis anterior, TA) of
20 months old control and Mck-CreGpcpd1flox/flox mice N=6 per group, respectively. B)
Representative images and quantification of MyHC!, MyhCIIA and MyHCIIB fibers in skeletal
muscle (quadriceps) from 20 months old control and Mck-CreGpcpd1flox/flox mice. Images were
taken under 5x magnification, and ≥100 myofibers were counted at 3 different matching
histological areas. N=4 animals per group. Scale bar 500µm. C) Representative cross sections
of skeletal muscle (quadriceps) and myofiber diameter size from 20 months old control and
Mck-CreGpcpd1flox/flox mice. Myofibers were imaged using 10X magnification with ≥ 100
myofibers analyzed per mouse. n=4 animals per group. Scale bar 100µm. E) Ultrastructure of
skeletal muscle (quadriceps) and F) intermyofibrillar mitochondria from 20 months old control
and Mck-CreGpcpd1flox/flox mice. No apparent structural defects were observed in the muscles
of Mck-CreGpcpd1flox/flox mice. n=3 animals per group. Scale bar=1µm and scale bar=500nm
respectively. G,H) Muscle strength evaluation of 20 months old Control and MckCre-Gpcpd1
mice. Unless otherwise stated, each dot represents one individual animal. Data are shown as
means ± SEM. Student’s two tailed, unpaired t-test with Welch correction was used for
statistical analysis unless otherwise stated, ns, not significant.
Supplementary Figure 5. Fasting glucose levels and tissue glucose uptake in
Mck-CreGpcpd1flox/flox mice.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

A) Fasting blood glucose levels in 12 weeks old control and Mck-CreGpcpd1flox/flox mice. N=6
and N=7 per group. B) Blood glucose levels and C) Area under curve (AUC) after an oral glucose
tolerance test (OGTT) in 20 month old months old control and MckCre-Gpcpd1flox/flox mice. N=6
per group. Student’s two tailed un-paired t test with Welch correction was used for AUC
statistical analysis. D) and E) Radioactive 2DG glucose uptake in heart muscle of 12 weeks old
control and Mck-CreGpcpd1flox/flox mice after bolus glucose feeding. F) Radioactive 2DG
glucose uptake in white fat (eWAT), beige fat (iWAT), brown fat (BAT), and liver of 12 weeks
old control and Mck-CreGpcpd1flox/flox mice after bolus glucose feeding. Each dot represents
one individual animal. Data are shown as means ± SEM. Student’s two tailed, unpaired t-test
with Welch correction was used for statistical analysis unless otherwise stated; ns, not
significant;*p < 0.05, ***p < 0.001.
Supplementary Figure 6. Muscle lipidome evaluation upon Gpcpd1 loss and
C2C12 derived myotubes GPC treatment
A) Mass spectrometry based un-targeted lipidomic analysis in quadriceps muscles isolated
from 3 months old control and Mck-Cre-Gpcpd1 flox/flox littermate mice. Abbreviations denote
triacylglycerols (TAG), sphingomyelins (SM), phosphatidylinositols (PI),
phosphatydilethanolamines (PE), phosphatydilcholines (PC), lysophosphatydilethanolamines
(LPE), lysophosphatydilcholines (LPC), diacylglycerols (DAG), ceramides (Cer), and
cholesterolesther (CE). Numbers denote the carbon numbers, heatmap denotes higher and
lower abundant lipid species in the muscles. No significant differences were found in the
indicated lipid species abundance between muscles isolated from control and Mck-Cre-
Gpcpd1 flox/flox mice. N=5 per group. B) Choline levels in skeletal muscle (quadriceps) of 3
months old control and MckCre-Gpcpd1flox/flox mice. N=10-12 per group. C) GPC levels in C2C12
differentiated myofibers without or with GPC pretreatment. Each dot represents one cell
culture. D) Cellular osmolarity of C2C12 derived myofibers treated with vehicle or GPC for 7
days. Each dot represents one cell culture. Data are shown as means ± SEM. Student’s two
tailed, unpaired t-test was used for statistical analysis unless otherwise stated, ns, not
significant, ****p < 0.0001.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 7. Insulin signaling pathway s is affected in muscles from
young Mck-CreGpcpd1flox/flox mice
Analysis of insulin signaling network on transcriptional level in quadriceps from young 3
months old control and MckCre-Gpcpd1flox/flox mice. The significantly downregulated genes
(P<0,05) are labelled in green.
Supplementary Figure 8. Muscles from young Mck-CreGpcpd1flox/flox mice
display an “aged-like” transcriptomic signature.
Comparison of upregulated (red) and downregulated (blue) gene expression changes in aged
skeletal muscles in rats (27 month old versus 6 month old) (Shavlakadze et al., 2019) with
gene expression changes of skeletal muscles from 3 month old control and Mck-
CreGpcpd1flox/flox mice. Overlapping dysregulated genes are highlighted.
Supplementary Figure 9. Correlation of GPC and GPC-PDE levels in humans
with fat mass, muscle mass and body mass index
Correlation of skeletal muscle Glycerophosphocholine (GPC) and Glycerophosphocholine
phosphodiester (GPC-PDE) levels with fat mass, muscle mass and body mass index in humans,
irrespective of age. No significant correlation was found for any of the parameters.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 1
A
**
0.0
0.5
1.0
1.5
***
GPC levels
(per mg of tissue)
Gpcpd1 expression levels
(normalized to 18S)
BC
Gpcpd1 expression levels
(arbitrary units)
Age (months)
Blood Insulin (µg/L)
DE
0.0
0.5
1.0
1.5
0.0
0.2
0.4
0.6
0.8
Gpcpd1 expression levels
(arbitrary units)
FG
H
0.0
0.1
0.2
0.3
0.4
Organ weight (g)
QA GC TA
I
J
****
Blood glucose (mg/dl)
0
50
100
150
200
250
*
*
0
20
40
60
Muscle glycogen (µg/g)
K
*
Fasted Fed
0
50000
100000
150000
6 9 12 15 18 21 24 27
0
20
40
60
80
100
R2=0.5953
P<0.0001
ns
ns
ns
ns
6 months old
24 months old
6 months old
24 months old
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Body weight (g)
0
10
20
30
40
ns Control
MckCre-Gpcpd1
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 2
0
5000
10000
15000
20000
0 15 30 45 60
50
100
150
200
250
300
350
400
0 15 30 45 60
0.0
0.2
0.4
0.6
Blood Glucose (mg/dL)
AUC
Blood Insulin (ug/L)
B
AC
Pgenotype<0.0001
****
*
Time (min)
Liver
(Alb Cre)
0 15 30 45 60
0
50
100
150
200
250
300
350
400
Body weight (g)
Blood Glucose (mg/dL)
Time (min)
FG
ns
ns
white and
brown fat
(Ap2 Cre)
0 15 30 45 60
0
50
100
150
200
250
300
350
0
10
20
30
40
Body weight (g)
Blood Glucose (mg/dL)
HI
ns
ns
Time (min)
0
10
20
30
40
0 15 30 45 60
0
50
100
150
200
250
300
350
400
Time (min)
Body weight (g)
JK
Blood Glucose (mg/dL)
ns
brown fat
(Ucp1 Cre)
ns
Control
AlbCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
Ap2Cre-Gpcpd1
Control
Ucp1Cre-Gpcpd1
Control
AlbCre-Gpcpd1
Control
Ap2Cre-Gpcpd1
Control
Ucp1Cre-Gpcpd1
0
10
20
30
D
*Control
MckCre-Gpcpd1
0
20
40
60
80
100
Radioactivity
(μmol/100g tissue/min)
Time (min)
*
Control
MckCre-Gpcpd1
0
50
100
150
2DG uptake (%)
E
Pgenotype=0.026
*
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 3
0 15 30 45 60
100
150
200
250
300
350
400
0 15 30 45 60
1
2
3
4
5
6
Body weight (g)
Weeks on HFD Time (min)
Blood Insulin (ug/L)
Time (min)
Blood Glucose (mg/dL)
D
GH
Blood Insulin (ug/L)
Time (min)
*
0 15 30 45 60
0
1
2
3
4
5
0 2 4 6 8 10
25
30
35
40
45
AC
Body weight (g)
Weeks on HCD
Blood Glucose (mg/dL)
0 15 30 45 60
50
100
150
200
250
300
350
400
Time (min)
*
B
0
5000
10000
15000
20000
0
5000
10000
15000
20000
*
*
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
AUC
AUC
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
0 2 4 6 8 10
25
30
35
40
45
50
55
EF
Pgenotype=0.028
Pgenotype=0.01
*
B
I
0
5000
10000
15000
20000
Control (HS)
***
P=0.093
AUC
P=0.75
Control (HFD)
MckCre-Gpcpd1 (CD)
MckCre-Gpcpd1 (HS)
MckCre-Gpcpd1 (HFD)
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 4
Muscle GPC
(per mg of tissue)
A
EMck-Gpcpd1 Control
*
0
2×105
4×105
6×105
8×105
1×106
1×107
2×107
3×107
F
**
Blood Glucose (mg/dL)
Fasted Glucose
Mck-Gpcpd1
Fasted Glucose
pIRẞ
pAkt
GAPDH
Akt
IRẞ
Control
Time (min)
*
0 15 30 45 60 75 90
0
50
100
150
200
0
5000
10000
15000
20000
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
AUC
GControl
MckCre-Gpcpd1
Pgenotype=0.0085
IJ
Normalized pIR levels
0
2
4
6
8
0
5
10
15
Normalized pAkt levels
Control
MckCre-Gpcpd1
** *
** *
H
0
20000
40000
60000
80000
100000
D********
Intracellular glucose
levels (RLU)
HBSS/-GPC/-glucose
HBSS/+GPC/+glucose
HBSS/-GPC/+glucose
B
50
100
150
200
250
300
*
Tissue osmolarity
(Osm per mg of tissue)
Control
MckCre-Gpcpd1
50
100
150
200
250
300
*
Tissue osmolarity
(Osm per mg of tissue)
C6 months old
24 months old
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Figure 5
A
0
1
2
3
4
5
*
Gpcpd1 expression levels
(normalized to 18S)
C
3
4
5
6
****
Metabolite levels
(arbitrary units)
D
0
5
10
15
20
25
****
E
GPC levels
(arbitrary units)
GPC levels
(arbitrary units)
Age (years)
20 30 40 50 60 70 80
4
6
8
10
12
R2=0.493
P<0.0001
0
2
4
6
20
25
30
35
40
PCr Pi PDE PC PME
F
Metabolite levels
(arbitrary units)
ns
ns ns ns
Young
Old
****
GPC GPC/PDE
GPC GPC/PDE
Young
Old
Young
Old
Control
Type 2 diabetes
Control
Type 2 diabetes
B
20 30 40 50 60 70 80
4
8
12
16
20
GPC/PDE levels
(arbitrary units)
Age (years)
R2=0.5212
P<0.0001
Control
T2DM
ns
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 1
0.0
0.5
1.0
1.5
9 12 17 21 24 27
Liver
0.0
0.5
1.0
1.5
9 12 17 21 24 27
Hippocampus
0.0
0.5
1.0
1.5
9 12 17 21 24 27
Kidney
Gpcpd1 expression levels
(arbitrary units)
Gpcpd1 expression levels
(arbitrary units)
Gpcpd1 expression levels
(arbitrary units)
Age (months) Age (months)
Age (months)
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 2
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 3
VQENTIASLR
YEADPVELFEIPVK
VDFIIIK
IYDWMPEQPNIFQVEQLER
LSHVTALK
SAGILTLPIMSR
YPILFLTQGK
IYDWMPEQPNIPFQVEQLER
Control MckCre-Gpcpd1
RT (min)
Intensity
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

0
100
200
300
0
200
400
600
800
1000
0
50
100
150
200
Organ weight (mg)
QA GC TA
C
Dystrophin
DAPI
Control MckCre-Gpcpd1
ns ns
ns
Supplementary Figure 4
DE
F
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Grip strength (g)
MckCre-Gpcpd1
Control
B
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
A
0
20
40
60
80
100
Fiber type (%)
Control
MckCre-Gpcpd1
MyHCI MyHCIIA MyHCIIB
Myofiber size (μm2)
Control
MckCre-Gpcpd1
0
2
4
6
8
10
Control
MckCre-Gpcpd1
G
Grip strength
(g/g of body weight)
MyHC I
MyHC IIA
MyHC IIB
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

60
70
80
90
100
110
A
ns
Blood glucose (mg/dl)
Supplementary Figure 5
C
DE
eWAT iWAT BAT Liver
0
50
100
150
200
FControl
MckCre-Gpcpd1
** **
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
Control
MckCre-Gpcpd1
ns
ns
ns
ns
0
50
100
150
200
Radioactivity
(μmol/100g tissue/min)
0
50
100
150
200
2DG uptake (%)
Rg
(μmol/100g tissue/min)
0 15 30 45 60
100
200
300
400
500
Time (min)
Blood Glucose (mg/dL)
B
MckCre-Gpcpd1
Control
Pgenotype<0.0001
****
0
5000
10000
15000
20000
25000
AUC
**
Control
MckCre-Gpcpd1
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 6
Muscle choline
(per mg of tissue)
0
5×104
1×105
1.5×105
2×105
ns
Control
MckCre-Gpcpd1
B
C
0.0
0.5
1.0
1.5
2.0
2.5
***
*
Intracellular GPC
levels (AU)
-GPC
+GPC
D
0.00
0.02
0.04
0.06
0.08
0.10
Cellular Osmolarity
(Osm per DNA content)
-GPC
+GPC
P=0.074
A
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 7
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

Supplementary Figure 8
Old muscle (27 month old) versus young muscle (6 month old) (Log2Fold change)
MckCre-Gpcpd1 muscle versus Control muscle (Log2Fold change)
upregulated
downregulated
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint

15 20 25 30 35 40
0
2
4
6
8
10
0 20 40 60
0
2
4
6
8
10
10 20 30 40 50
0
2
4
6
8
10
GPC levels
(arbitrary units)
GPC levels
(arbitrary units)
GPC levels
(arbitrary units)
Body mass index (BMI)
Body fat (%) Muscle mass (%)
R2=0.0728
R2=0.00313 R2=0.01675
Supplementary Figure 9
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission.
The copyright holder for thisthis version posted September 25, 2022. ; https://doi.org/10.1101/2021.10.26.465828doi: bioRxiv preprint