Induction of the Nicotinamide Riboside Kinase NAD+ Salvage Pathway in Skeletal Muscle of H6PDH KO Mice

Home , Cellular compartment

bioRxiv preprint doi: https://doi.org/10.1101/567297; this version posted March 4, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

1 Induction of the nicotinamide riboside kinase NAD+

2 salvage pathway in skeletal muscle of H6PDH KO mice

4 Craig L. Doig1,2, Agnieszka E. Zelinska1, Lucy A. Oakey1,2, Rachel S. Fletcher1,2, Yasir El

5 Hassan1,2, Antje Garten1, David Cartwright1,2, Silke Heising1,2, Daniel A Tennant1,2, David

6 Watson3, Jerzy Adamski4 & Gareth G. Lavery1,2.

8 1Institute of Metabolism and Systems Research, 2nd Floor IBR Tower, University of

9 Birmingham, Birmingham, B15 2TT, UK. 2Centre for Endocrinology, Diabetes and

10 Metabolism, Birmingham Health Partners, Birmingham, B15 2TH. 3Strathclyde Institute of

11 Pharmacy and Medical Sciences, Hamnett Wing John Arbuthnott Building, Glasgow G4

12 0RE. 4Helmholtz Zentrum Munchen GmbH, Ingolstadter Landstrasse 1, D-85764

13 Neuherberg. Germany.

15 Corresponding author:

16 Gareth G Lavery, PhD

17 Wellcome Trust Senior Research Fellow

18 Institute of Metabolism and Systems Research

19 University of Birmingham

20 Edgbaston

21 Birmingham, UK

22 B15 2TT

23 Tel: +44 121 414 3719 Fax: +44 121 415 8712

24 Email: [email protected]

26 Disclosure statement: The authors have nothing to disclose.

27 Abbreviated title: NAD+ metabolism and ER stress in skeletal muscle myopathy

28 Word Count: 5504; 7 Figures (Inclusive of 3 Supplemental)

1 Abstract

2 Background: Hexose-6-Phosphate Dehydrogenase (H6PDH) is the only known generator

3 of the pyridine nucleotide NADPH from NADP+ in the Endoplasmic/Sarcoplasmic Reticulum

4 (ER/SR). However, its ability to influence cell wide pyridine biosynthesis and metabolism is

5 unclear.

6 Methods: Skeletal muscle from H6PDH Knockout mice (H6KO) and Wild type controls (WT)

7 was subject to metabolomic assessment and pathway analysis. NAD boosting was

8 performed through intraperitoneal injection of Nicotinamide Riboside (NR) (400mg/kg) twice

9 daily for 4 days. Mice were culled and skeletal muscle tissue collected the next day. Double

10 knockout mice were generated through to breeding of H6KO with NRK2 knockout mice.

11 These mice were assessed for NAD levels and ER stress response.

12 Results: H6KO skeletal muscle shows significant changes in the pathway regulating NAD+

13 biosynthesis with the Nicotinamide Riboside Kinase 2 gene (NRK2) being the most elevated

14 gene. Assessment of the metabolic impacts of this dysregulation showed decreases in

15 mitochondrial respiration and Acylcarnitine metabolism. Pharmacologically boosting NAD+

16 levels with NR had no significant impact over ER stress or Acetyl-CoA metabolism. A duel

17 H6-NRK2 KO mouse exhibited changes in NAD+ levels but no overt change in metabolism

18 compared to the H6KO mouse.

19 Conclusions: This work shows a relationship between ER/SR status and wider muscle cell

20 metabolism. The utility of NAD+ boosting to skeletal muscle is demonstrated however, it is

21 also evident that the extent of cellular stress and/or REDOX status may overcome any

22 beneficial impact.

24 Keywords: Hexose-6-Phosphate Dehydrogenase, Endoplasmic/Sarcoplasmic Reticulum,

25 Skeletal muscle, Nicotinamide Riboside, Nicotinamide Adenine Dinucleotide

1 Background

2 A complete understanding of REDOX biology within the Sarcoplasmic/Endoplasmic

3 Reticulum (SR/ER) and any influence over cellular metabolism is lacking. Hexose-6-

4 phosphate dehydrogenase (H6PDH) is the only recorded generator of NADPH (from NADP+)

5 within this cellular compartment. This exclusivity ensures its importance to not only SR/ER

6 REDOX control and coenzyme generation for pyridine nucleotides, but also implicates it into

7 wider skeletal muscle homeostasis1,2.

8 The pyridine Nicotinamide Adenine Dinucleotide (NAD+) is a fundamental contributor to

9 cellular REDOX maintenance and energetic utilisation3,4. The rate limiting nature of NAD+ to

10 biochemical reactions make it fundamental to energetic metabolism including mitochondrial

11 fatty acid catabolism and ATP generation.

12 Pharmacologic augmentation of NAD+ through the use of precursors such as Nicotinamide

13 Riboside (NR) has shown restorative potential in NAD+ based therapies using a variety of

14 cell types and animal models with supportive human studies5-10. Recent studies back this,

15 showing a loss of NAD+ salvage function leads to progressive muscle degeneration and

16 boosting of the nicotinamide salvaging pathways liberates skeletal muscles from some of the

17 deleterious impacts of ageing, muscular dystrophy and high-fat diet phenotypes10-13. The

18 phosphorylation of NR into nicotinamide mononucelotide (NMN) occurs through two

19 nicotinamide riboside kinases (NRKs) NRK1 and NRK2. In many tissues NRK1 is the

20 dominant and rate limiting source of NR phosphorylation14. However, in skeletal muscle

21 NRK2 is highly expressed and recent data highlights its roles within myocyte physiology8.

22 Biosynthetic salvage of nicotinamide mediated by NRK2 has only recently been identified as

23 having relevance to skeletal muscle biology and its function within atrophic models, ER

24 stress and REDOX dysfunction are unstudied.

25 Furthermore, maintenance of the Sarcoplasmic Reticulum (skeletal muscle only) SR/ER

26 REDOX pyridine pools and its contribution to whole cell NAD+ levels is unclear. Hexose-6-

27 phosphate dehydrogenase (H6PD) converts glucose 6-phosphate (G6P) to 6-

28 Phosphoglucogonate (6-PG) and 6-Phosphglucogonatelactonase (6-PGL). Within the

1 SR/ER H6PD generates the rate limiting cofactor for intracellular glucocorticoid activation by

2 11beta-hydroxysteroid dehydrogenase type 1 (HSD11b1)15. The knockout of H6PDH

3 produces profound metabolic dysfunction, with glycogen storage shifts in liver accompanied

4 by improved glucose sensitivity16. Adipose analysis of whole body H6PD KO mice revealed

5 reduced adiposity and changes in lipolytic rates, suggesting that absence of H6PDH

6 switches the direction with which HSD11b1 acts upon glucocorticoids17. Thus the action of

7 H6PDH upon metabolism are multifaceted and tissue-specific. Skeletal muscle from H6PDH

8 KO mice exhibits changes in REDOX balance of the ER/SR and inducing Unfolded Protein

9 Response. This in-turn induces myofibre loss, vacuole formation and fibre type switching.

10 Permutations extend to the cytosolic and nuclear environment shifting metabolic activity of

11 purine biosyntheisis18. Given the importance of the SR/ER to skeletal muscle and H6PD

12 being the only source of NADPH within the SR/ER there is potential for SR/ER pyridine

13 metabolism may be shifted in ways that are not yet attributed to H6PD function. In this study,

14 we examined H6PD as a regulator skeletal muscle metabolism through the conversion of

15 NADP+ to NADP(H), and how this impacts upon cellular processes extending beyond the

16 SR/ER confined environment. These data show that H6PDH not only has implications for the

17 biosynthetic processes governing NAD+ generation but also impacts wider cellular pathways

18 such as lipid homeostasis and mitochondrial fatty acid oxidation.

19 Here we explore the relationship between H6PDH and NRK2, attempting to boost skeletal

20 muscle NAD+ bioavailability with acute NR supplementation. In order to further understand

21 dependence of SR/ER actions upon NRK2 mediated NAD+ biosynthesis we generated a

22 whole body double-knockout of H6PD and NRK2, exploring how ablation of NRK2 within

23 H6PDH defunct skeletal muscle alters metabolic function and mitochondrial activity. Our

24 work reveals previously unconsidered aspects of NAD+ metabolism showing an ER sensing

25 mechanism that up-regulates a skeletal muscle NAD+ biosynthesis pathway in response to

26 altered NADPH availability.

1 Materials and Methods

2 Animal care, Mouse strains, storage

3 C57/BL5J mice were group housed in a standard temperature (22oC) and humidity-

4 controlled environment with 12:12-hour light:dark cycle. Nesting material was provided and

5 mice had ad libitum access to water and standard chow. Mice were sacrificed using

6 schedule one cervical dislocation and tissues were immediately. Collections were all

7 performed at 10-11am. Intraperitoneal injections of Nicotinamide Riboside (400mg/kg) were

8 given twice daily. (10am and 3pm) for 4 days before cervical dislocation and skeletal muscle

9 bed collection on Day 5 (10am), tissues were flash frozen, fixed or for high resolution

10 respirometry stored in BIOPS buffer. Heterozygotes of H6PD KO and NRK2 KO were breed

11 to generate H6-NRK2 KO mice. Mice were utilised between 8-12 weeks of age. Experiments

12 were consistent with current UK Home Office regulations in accordance with the UK Animals

13 (Scientific procedures) Act 1986.

15 RNA extraction and qRT-PCR

16 Total RNA was extracted from skeletal muscle tissue using TRI-reagent (Invitrogen). RNA

17 quality was determined by visualisation on a 1.5% agarose gel and quantity was measured

18 by nanodrop absorbance at 260nm. Reverse transcription was conducted using 1µg RNA

19 that was incubated with 250uM random hexamers, 5.5mM MgCl2, 500uM dNTPs, 20units

20 RNase inhibitor 63units multi-scribe reverse transcriptase and 1x reaction buffer. Reverse

21 transcription was performed on a thermocycler set at the following conditions: 25oC

22 10minutes, 48oC for 30 minutes before the reaction was terminated by heating to 98oC for 5

23 minutes. cDNA levels were determined using an ABI7500 system (Applied Biosystems),

24 reactions were conducted in a 384-well plate in single-plex format. Primers and probes were

25 purchased as Assay on Demand (FAM) products (Applied Biosystems). Total reaction

26 volumes used were 10ul containing Taqman Universal PCR mix (Applied Biosystems). All

27 reactions were normalised to 18s rRNA (VIC) (Applied Biosystems). The real-time PCR

28 reaction was performed at the following conditions: 95oC for 10 minutes then 40 cycles of

1 95oC for 15 seconds and 60oC for 1 minute. Data were collected as Ct values and used to

2 obtain deltaCt (dCt) values and expressed as fold change ±standard error of the

3 mean(SEM).

5 mtDNA Copy number

6 DNA was isolated using phenol-chloroform and washed in Ethanol before quantification

7 using nanodrop. DNA from all samples was diluted to 50ng/µl and subject to real-time PCR

8 using SYBRgreen master mix (Applied Biosystems) using the following primers specific for

9 nuclear and mitochondrial DNA.

10 Nuclear DNA: TERT Forward CTAGCTCATGTGTCAAGACCCTCTT Reverse

11 GCCAGCACGTTTCTCTCGTT, Mitochondrial DNA: DLOOP Forward

12 AATCTACCATCCTCCGTGAAACC Reverse TCAGTTTAGCTACCCCCAAGTTTAA, ND4

13 Forward AACGGATCCACAGCCGTA Reverse AGTCCTCGGGCCATGATT.

15 Western Immunoblotting

16 Protein lysates were collected in RIPA buffer (50mmol/l Tris pH 7.4, 1% NP40, 0.25%

17 sodium deoxycholate, 150mmol/l NaCl, 1mmol/l EDTA), protease & phosphatase inhibitor

18 cocktail (Roche, Lewes, U.K.), stored at -80°C (>30 minutes), defrosted on ice and

19 centrifuged at 4°C (10mins, 12000 rpm). The supernatant was recovered and total protein

20 concentration was assessed by Bio-Rad assay. Total proteins were resolved on a 12% SDS-

21 PAGE gel and transferred onto a nitrocellulose membrane. Primary antibodies; Anti-Goat

22 H6PDH (in house), Anti-Mouse NAMPT (Bethyl A300-372A), Anti-Rabbit NRK1 (in house),

23 Anti-Rabbit NRK2 (in house), Anti-Rabbit CHOP (Cell Signaling D46F1 5554), Total Lysine

24 Ac (Cell Signaling 9441), Anti-Mouse Rodent OXPHOS Mitoprofile (Abcam ab110413), IDH2

25 K413ac (GeneTel AC0004), IDH2 Total (Abcam ab55271), H3K9ac (Cell Signaling 9649),

26 H3K18ac (Cell Signaling 9675), ERO1a (Cell Signaling 3264), PDI (Cell Signaling 3501),

27 alpha-Tubulin (Santa Cruz B-7 sc-5286). Secondary antibodies (Dako) anti-mouse, anti-goat

28 and anti-rabbit conjugated with HRP added at a dilution of 1/10,000. Equal loading of protein

1 content was verified using alpha-tubulin as a housekeeping protein and bands visualised

2 using ECL detection system (GE Healthcare, UK). Bands were measured using Image J

3 densitometry, and normalised to those of loading control (alpha tubulin).

5 High Resolution Respirometry of permeabilised muscle fibres by Oroboros

6 Respirometry studies in fast (Extensor Digitorum Longus) and slow twitch (Soleus) skeletal

7 muscle myofibres were performed using high-resolution respirometry (Oroboros Instruments,

8 Innsbruck, Austria). Fresh skeletal muscle (5mg) was stored in ice-cold 2mls BIOPS buffer.

9 Any connective tissue and adipose was discarded and 20µl of Saponin stock solution for

10 permeabilisation of muscle fibres (5mg Saponin per ml BIOPS buffer) added. The samples

11 were then incubated for 30 minutes at 4°C. Muscles were washed in Mir05 for 10 minutes at

12 4°C before drying on filter paper. Approximately 3mgs of muscle tissue was introduced into

13 the chamber and left for >10 minutes to give a stable baseline signal, re-oxygenation

14 occurred once O2 concentration reached 250µM. Sequential addition of inhibitors and

15 substrates occurred as per Oroboros Instruments specified protocol for fatty acid oxidation.

17 NAD+ Quantification by Colorimetric Assay

18 NAD+ was extracted from flash frozen skeletal muscle and NAD+/NADH quantified using the

19 NAD+/NADH Assay kit (BioAssay Systems) according to the manufacturers instructions.

21 Unbiased Metabolomic measurement

22 20mg of skeletal muscle previously flash frozen tissue from Tibialis Anterior (TA) or Soleus

23 (SOL) was placed in dry ice. Metabolites were extracted into 0.5ml of ice-cold LCMS grade

24 methanol:acetonitrile:water (50:30:20). Samples were pulverised and stored at 4°C for 10

25 minutes before centrifugation at 0°C 1300rpm for 15 minutes. Supernatant was then

26 transferred to HPLC vials ready for LC/MS analysis. To preclude variables introduces from

27 preparing samples separately all samples were prepared simultaneously in a randomised

28 order and re-randomised before injection in the LC/MS.

1 Biocrates Targeted Metabolite kit

2 Skeletal muscle (Quadriceps), liver and plasma were collected from WT and H6KO and WT

3 littermates. Frozen tissue was homogenised in tubes containing ceramic beads of 1.4mm

4 diameter using extraction solvents. The AbsoluteIDQ p150 kit (Biocrates Life Science AG,

5 Austria) was prepared according to the manufacturers instructions. Briefly, 10µl of tissue

6 extract being used for PITC-derivatisation of amino acids, extraction with organic solvent and

7 liquid handling steps. Metabolites were quantified with a reference to appropriate internal

8 standards using FIA-MS/MS on a 400 QTRAP instrument (AB Sciex, Germany) coupled to a

9 HPLC Prominence (Shimadzu Deutschland GmbH, Germany). A total of 163 metabolites

10 were measured and divided into 6 groups (Supplementary Table 1).

12 Metabolomic set enrichment analysis

13 We conducted metabolite set enrichment analysis using MetaboAnalyst19,20. Significantly

14 dysregulated metabolites from the unbiased measurement in H6KO compared to WT muscle

15 were uploaded to Metaboanalyst. These were processed with the over-representation

16 analysis (ORA) algorithm and the metabolic pathway associated metabolite library set. This

17 examines the group of metabolites using Fisher’s exact test, calculating the probability of

18 seeing a particular number of metabolites containing the biological pathway of interest in the

19 list of metabolites.

21 Statistical Analysis

22 Students T-test or ANOVA statistical comparisons were used with the Graphpad Software

23 Inc. Prism version 5. Data presented as mean ±SEM with statistical significance determined

24 as *. p<0.05, **. p<0.01, ***. p<0.001. Unpaired T-Test compared treatments or genotypes.

25 Statistical analysis derived from real-time PCR data was determined using dCt values

26 throughout.

1 Results

2 Sarco/Endoplasmic confined H6PD knockout produces muscle wide metabolic

3 alterations in Nicotinamide metabolism. The Sarcoplasmic/Endoplasmic Reticulum

4 serves as a major calcium store, driving the contractile response and contributing towards

5 REDOX homeostasis through the generation of NADPH by H6PD. Previously, studies show

6 the ablation of H6PDH in conditional skeletal muscle and whole body models provoking a

7 muscle myopathy driven by SR/ER stress and the unfolded protein response. Additionally,

8 the physiological role of H6PD upon glucocorticoid metabolism have been characterised,

9 further highlighting the importance of SR/ER REDOX to NADPH generation15,18. We

10 subjected skeletal muscle extracted from WT and H6KO mice to an unbiased metabolomic

11 screen to better understand the role H6PD plays in wider muscle cell metabolism and SR/ER

12 REDOX maintenance. This unbiased metabolite profile shows significant differential

13 abundance of metabolites in the H6KO muscle (Supplementary Table 1). The N6,N6,N6.-

14 Trimethyl-L-Lysine metabolite is a methylated derivative of the amino acid lysine, a

15 constituent of nuclear histone proteins and a precursor to carnitine and fatty acid oxidation in

16 the mitochondria. 3-Hydroxysebasicacid was also significantly elevated in the H6KO muscle,

17 a metabolite elevated in the urine of humans with peroxisomal disorders and glycogen

18 storage disease21,22, signalling a defect in fatty acids synthesis and metabolic defects such

19 as medium chain acyl-CoA dehydrogenase deficiency. Cis-5-Tetradecenoylcarnitine is a

20 marker of long chain acyl-dehydrogenase (VLCAD) deficiency, alterations in metabolites

21 regulating beta oxidation are contributory to skeletal muscle myopathy23. There are

22 significant elevations in both the metabolite Nicotinamide Adenine Dinucleotide (NAD+) and

23 the methylated form of NAD+, 1-methylnicotinamide. This dataset was subject to pathway

24 analysis to reveal overrepresentation of metabolites that sharing a metabolic pathway. The

25 most highly overrepresented pathways were nicotinamide and pyridine biosynthesis (Figure

26 1A-B). Changes in NAD+ availability to skeletal muscle profoundly impacts downstream

27 reactions intrinsic to metabolic function and defects in NAD+ homeostasis are contributory to

28 disease development24-26. To gain further insight NAD+ generation in relation to the H6KO

1 model we conducted quantification of NAD+, NAD(H) and NAD(H)total abundance in WT and

2 H6KO skeletal muscle. These data show no significant NAD+ changes of H6KO in

3 comparison to WT. NADH is significantly decreased in H6KO, however, measurement of

4 total NAD(H) levels were indifferent to WT (Figure 1E-H). Critical to cellular NAD+

5 maintenance are the multiple salvage and biosynthetic enzymes culminating with NAD+

6 formation. The schematic showing these routes indicates that combinatorial sources leading

7 to NRK2 generates NMN from NR contributing towards cellular NAD+, which is consumed by

8 cellular reactions and the by-product NAM is salvaged and regenerated into usable NAD+

9 (Figure 1G). Real-time PCR analysis of these genes responsible for muscle NAD+

10 biosynthesis and salvage demonstrates changes in both biosynthetic and NAD+

11 phosphorylating genes (Figure 1H). There is significant increase in the skeletal muscle

12 restricted nicotinamide riboside kinase 2 gene (NRK2), whilst the constitutively expressed

13 NRK1 and NAMPT are unchanged. Responsible for the phosphorylation of the NAD+

14 precursor nicotinamide riboside into nicotinamide mononucleotide NRK2 has been shown to

15 be elevated in a model of cardiomyopathy7. Significantly down regulated genes in the H6KO

16 implicated in NAD+ biosynthesis show reductions in levels of nmnat2 a key enzyme skeletal

17 muscle conversion of NMN into NAD+ (but no change in nmnat1), suggesting restriction of a

18 rate limiting response to the upregulated NRK2, moderating NAD+ levels. Downregulation of

19 NAD kinase limits generation of NADP+, (the rate limiting coenzyme for H6PD within the

20 SR/ER). This indicates an adaptive response serving to prevent NADP+ build up in response

21 to H6PDH ablation. Purine Nucleoside Phosphorylase (PNP) (converts NR to NAM) and the

22 NAD+ utilising ADP-Ribosyltransferase (ART1) are both downregulated underlying the

23 disturbance in NAD+ metabolism driving the wider metabolic disturbances showed in the

24 H6KO muscle. Protein measurement of both NRK1 and NRK2 revealed significant

25 elevations in H6KO samples. The rate limiting enzyme for NAD+ salvage NAMPT remained

26 unchanged indicating that H6KO mouse muscle may have elevated capacity to generate

27 NMN from NR, whilst the contribution from salvaged NAM remained unchanged (Figure 1 I-

1 K). This presents the opportunity to examine the biosynthetic contribution as part of an

2 adaptive metabolic response to cellular stress.

3 H6KO skeletal muscle has reduced mitochondrial fatty acid oxidative capacity and

4 widespread changes in acylcarnitines.

5 Elevations in NAD+ and NMN concentrations within skeletal muscle have been identified as

6 promoting mitochondrial function27-29. We therefore, challenged the potential of upregulated

7 NRK2 in the H6KO model to impact mitochondrial respiration rates. Despite a lack of NAD+

8 increase within the H6KO muscle we reasoned that flux of NAD+ turnover may still be

9 changed, and this may have implications for energetic utilisation of mitochondria. Therefore,

10 to assess mitochondrial metabolic function we subjected permeabilised skeletal muscle

11 fibres from both the Tibialis anterior (TA) and Soleus (SOL) muscle beds to high-resolution

12 respirometery, exploiting mitochondria’s fatty acid oxidative capacity. Measurement of the

13 fast twitch TA muscle from H6KO mice shows a decreased ability to consume oxygen for

14 energetic metabolism in the H6KO model (Figure 2A). This defect was also apparent in slow

15 twitch Soleus muscle from H6KO (Figure 2B). This suggests H6KO ablation has a

16 detrimental impact upon mitochondria and the elevations in NAD+ biosynthesis measured in

17 NRK2 increased abundance are ineffectual and their impact upon mitochondrial

18 metabolism. These results show H6KO (fast and slow twitch) muscle has significantly

19 decreased ability to utilise substrates for fatty acid beta-oxidation and overall respiratory

20 capacity. To understand if these measurements are a measure of functionality and not

21 mitochondrial abundance we examined mtDNA content in both WT and H6KO muscle and

22 found them to be unchanged (Figure 2C). Together these data demonstrate in addition to

23 the SR/ER disturbances H6KO skeletal muscle there is a widespread inability to utilise

24 sources of energy for cellular respiration rising from occlusion in the generation of pyridine

25 nucleotide NADPH within the SR/ER. This leads to widespread physiological deficits in

26 mitochondrial function without altering mitochondrial protein levels. Total protein

27 measurement and quantification of mitochondrial respiratory complex subunits indicated no

28 change in overall abundance of ATP generating complexes (Figure 2D). Using a targeted

1 metabolomic analysis of acylcarnitines and other metabolites (Supplementary Figure 2) we

2 find short-chain acylcarnitines (C2, C3, C4, C5, C6 & C10:1) were differentially upregulated

3 in H6KO muscle tissue, the ratio of acetylcarnitine to free carnitine (C2:C0) was increased

4 signifying elevations in even numbered fatty acids within the H6KO. Free-carnitine was not

5 significantly different in H6KO versus WT tissue (Figure 2E). To gain further insight and

6 reconcile observed changes in myofibre carnitine composition and shifts in phospholipid

7 species we used real-time qPCR to examine expression of genes integral to carnitine

8 acylation, lipid mobilisation and fatty acid metabolism. H6KO tissue shows significant

9 elevations in acss2, far1, acsl2, acaa2 and scd1 (Figure 2F). A schematic of disrupted

10 genes in H6KO critical to the generation of acylcarnitine and fatty acid esters, contributing

11 towards mitochondrial B-oxidation (Figure 2G). Protein lysates collected from H6KO muscle

12 show increases in Isocytrate Dehydrogenase 2 (IDH2) acetylation and histone 3 lysine 9 and

13 lysine 18 acetylation, reflecting fatty acid oxidation and acyl-CoA changes observed in the

14 unbiased metabolome measurements and the targeted metabolic screen (Figure 2H).

15 Boosting of NAD+ with the precursor Nicotinamide Riboside has no influence

16 metabolic composition of H6KO mouse skeletal muscle.

17 To establish the utilisation of NRK2 and their importance to skeletal muscle metabolic

18 function we supplemented both WT and H6KO animals with the NAD+ precursor

19 Nicotinamide Riboside (400mg/kg twice daily over 4 consecutive days). With the hypothesis

20 being intraperitoneal injection of NR would provide additional substrate for utilisation by

21 elevated NRK2 in H6KO muscle. Thus exploiting the upregulated biosynthetic route to NAD+

22 ultimately rescuing the defects in these mice. Quantification of NAD levels in the skeletal

23 muscle of mice receiving i.p NR shows it was successful at enhancing NAD+ levels recorded

24 in both WT and H6KO samples when compared to their sham treated controls (Figure 3 A-

25 D). However, NADH was only successfully elevated by i.pNR in H6KO, WT trended upwards

26 but failed to achieve significance. This potentially signifies that although H6PD is the sole

27 source of NADPH in the SR/ER and despite the high abundance of SR/ER in skeletal

28 muscle it may not be a significant contributor towards total myocellular NAD(H). These data

1 demonstrate i.p NR over an acute schedule (4 days) is successful and elevating cellular

2 NAD+ in both WT and H6KO metabolically challenged mice. However, NADH is only

3 augmented in the H6KO model, reflecting the enhanced NAD+ biosynthetic and salvage

4 network identified in the unbiased metabolomic screen. Transcript analysis of genes integral

5 to NAD biosynthesis and salvage showed a decrease in NRK2 in WT, not shown in H6KO.

6 I.P NR was successful at increasing the expression of NMNAT1, NAMPT and NRK1 in

7 H6KO muscle, responses not measured in WT cells and may reflect shifts in the biosynthetic

8 generation of the H6KO model. Additionally, a lack of response from NRK2 in the H6KO may

9 suggest NRK2 is transcript is functioning towards physiological capacity (Figure 3E).

10 Measurement of transcripts regulating Acyl-CoA and lipid metabolism enzymes contributing

11 towards beta-oxidation reveals that NAD+ boosting has no impact over their expression

12 (Figure 3F). This suggests that increased NAD+ concentrations produced by i.p.NR have

13 limited/no bearing upon the mitochondrial energy utilisation of NAD+ flux within skeletal

14 muscle of WT animals. Protein quantification of NAMPT and NRK2 showed no change in

15 abundance in response to i.p.NR (Figure 3G+H). In order to further asses the effect of NR in

16 skeletal muscle of WT and H6KO muscle and challenge the elevated NRK2 levels in H6KO

17 muscle we subjected samples to high resolution respirometry fatty acid oxidation protocols

18 and found H6KO muscle +i.p NR is significantly reduced in its O2 flux measurements when

19 compared to WT i.p NR. The i.p NR data sets were not significantly different from their sham

20 treated genotype controls suggesting that although acute i.p NR does increase NAD+

21 concentrations within muscle there is no measurable bearing upon mitochondrial O2 flux.

22 Furthermore, in H6KO mice attempts to exploit increased NRK biosynthetic route through i.p

23 NR, and NAD+ concentration elevation does not overcome the metabolic defects induced by

24 H6PDH ablation (Figure 3 I-J). The H6KO model exhibits extensive ER stress response and

25 unfolded stress response. To understand any impact upon this protein lysates from muscle

26 were probed for the ER stress response protein CHOP and the protein folding regulators

27 ERO1a and PDI (Figure 3K). From these data it is evident that the acute i.p NR protocol has

28 no significant impact upon the protein expression, suggesting that benefits of NAD+ boosting

1 in other cellular functions may not extend into the ER/SR of skeletal muscle. NAD+ boosting

2 as a strategy to increase availability of the rate-limiting factor for the sirtuin enzyme family

3 has shown some promising outcomes. Therefore, to assess acetylation levels changes in

4 response to acute i.p NR in both WT and H6KO muscle protein lysates from these mice

5 were examined for global acetylation and acetylation of the mitochondrial enzyme IDH2.

6 These data show there was no change in acetylation status shown when compared to sham

7 controls (Figure 3L).

8 Ablation of NRK2 within the H6KO muscle decreases NAD+ levels and is independent

9 to the H6KO skeletal muscle myopathy.

10 As boosting of NAD+ did not change the mitochondrial function or ER/SR stress status of the

11 H6KO muscle we investigated the potential to perturb the H6KO induced NRK2 response.

12 With our hypothesis stating NRK2 as being a dominant muscle NAD+ contributing pathway

13 crucial to whole cell pyridine levels. To challenge this we produced a double knockout

14 model, deficient in both H6PD and NRK2. As NRK2 is skeletal muscle restricted, we

15 reasoned that its ablation in the H6KO would negatively affect cellular NAD+ content,

16 exacerbating the H6KO muscle myopathy phenotype. Quantification of NAD+ levels revealed

17 that the Double knockout TA has significantly decreased NAD+ and total NAD(H) total,

18 indicating that the ablation of NRK2 in the H6 null muscle (Figure 4A-D). H6KO skeletal

19 muscle beds exhibit myopathy, this is reflected in the decreased muscle bed weights (Figure

20 4 E-G). Assessment of NAD biosynthesis and salvage gene expression revealed no

21 significant change in the DKO muscle compared to H6KO (NRK2 KO has no significant

22 change compared to WT mice), suggesting that ablation of skeletal muscle restricted NRK2

23 can reduce the levels of NAD(H) generated but this may not affect cellular metabolic function

24 (Figure 4H). Quantitative real-time PCR of the fatty acids and Acyl-CoA genes upregulated

25 in the H6 model have equivalent expression to DKO indicating no altered metabolic

26 phenotype in the DKO muscle tissue. (Figure 4I). Protein levels of the unfolded protein

27 response is upregulated in the H6KO muscle and protein measurement of the DKO reflects

28 an unchanged level of expression of the protein unfolded responses factors ERO1a and PDI.

1 These upregulated sensors of ER stress are not significantly altered in expression compared

2 to H6KO suggesting NRK2 has little/no impact over the ability of the SR/ER UPR pathway

3 (Figure 4J+K).

5 Discussion

6 The SR/ER located H6PD is the only know source of NADPH within this largely impermeable

7 compartment. Currently there is a gap in our understanding of how nucleotide pools

8 independent to the cytosol (a high consumer of cellular NADPH with its own maintenance

9 stores) relate to cellular pyridine homeostasis. We demonstrate that ER REDOX dysfunction

10 perturbs mitochondrial lipid metabolism and cytosolic NRK2 mediated NAD+ biosynthesis

11 within skeletal muscle. The reduced bioenergetic capacity of H6KO mitochondria

12 demonstrates that SR/ER stress signals confer extensive metabolic impairment. However,

13 these direct signalling mechanism remains unresolved.

14 In contrast to H6KO conditional and global murine models that show a profound upregulation

15 of the ER stress response human cancer cell lines of breast origin transient knockdown of

16 H6PD produced a down regulation in ER stress30. Also reported was a reduction in migration

17 and enhanced cell adhesion, these data suggest not only that cancer cells show reliance

18 upon H6PD mediated NADPH generation for their proliferation but also that H6PD maybe a

19 novel target. Importantly, supportive of our work this study in human cancer cells also

20 demonstrated an increased mitochondrial superoxide levels, indicative of mitochondrial

21 impairment. We demonstrate the SR/ER capability to create mitochondrial dysfunction,

22 further quantifying the level of mitochondrial respirometry in H6PD defunct myofibres.

23 Supportive of this notion is the identification of H6PD as a target of PARP inhibitor

24 Rucaparib, (PARPs being NAD+ dependent enzymes) indicating current anti-cancer

25 therapeutic strategies may already be exploiting the ability of the SR/ER to impact wider cell

26 metabolism31.

27 In an attempt to exploit enhanced NAD+ biosynthetic enzyme expression in the H6KO

28 skeletal muscle we delivered a saturated supply of high dose precursor, nicotinamide

1 riboside. This was successful at enhancing NAD+ levels within the tissue, also it elevated

2 NADH within H6KO samples. However, there was no change over the recorded metabolic

3 features of the H6KO model with ER stress, acylation and mitochondrial fatty acid oxidation

4 unchanged. Thus physiologically the elevation in NAD+ was unable to overcome the

5 negative attributes of H6PD ablation. The acute nature of the supply of NR (4 days) maybe

6 considered a limitation of the study, with this duration it is unlikely that structural aberrations

7 induced by the H6KO might be rescued. However, the acute delivery of NR did induce NAD

8 sifts recorded within skeletal muscle in addition to regulatory changes in genes important to

9 NAD+ utilisation and salvage.

10 As an alternative strategy to challenge the relationship between NAD+ biosynthesis and the

11 SR/ER REDOX enzyme H6PD we produced a duel-knockout model with ablated H6PD-

12 NRK2. This model exhibited a level of skeletal muscle myopathy comparable to the H6KO

13 that was not exacerbated or rescued by the lack of NRK2 gene product. These data suggest

14 generation of NADPH within the SR/ER by H6PD is essential to cellular metabolic functions

15 external to the SR/ER environment. Differential NAD+ homeostasis has been demonstrated

16 to be present in muscular pathologies such as muscular dystrophy and infantile onset

17 Pompe disease32,33. The latter being a glycogen storage disease and reflective of the H6PD

18 glycogen storage defects34. Also in cardiac muscle it has been demonstrated that

19 upregulation of NRK2 occurs as a response to congestive heart failure7. Herein we detail a

20 similar mechanism for skeletal muscle with enhanced expression of NRK2 mRNA and

21 protein indicating that this is an adaptive response and occurs independently of NAMPT

22 activity. Attempting to delineate the NAD+ biosynthetic changes introduced by H6PD showed

23 although there was significant elevations in NRK2 mediated expression its NAD+ generating

24 contribution is presumably attenuated by the unchanged expression levels of the rate limiting

25 NMNAT enzymes. Furthermore, the increase of NRKs (in particular NRK2) maybe an

26 attempt to rectify the REDOX alteration within the H6PD KO SR/ER. Should this notion be

27 confirmed would support the concept of organelle sensing and exchange of nucleotides, an

28 attribute whose existence remains unclear partly due to a lack of identified transporters for

1 pyridine nucleotides. However, recent published data support the existence of mitochondrial

2 membrane transporters able to distribute pyridine nucleotides between cytosolic and

3 mitochondria, indeed the compartmentalisation of NAD+ as a cellular mechanism capable of

4 regulating metabolism and their transcriptional programmes is now apparent35. Movement of

5 NAD+ has been more specifically evaluated and the existence of a transporter able to flux

6 NAD+ to and from the mitochondria has been shown, although the identification of the

7 responsible protein(s) remains unknown36.

9 Conclusion

10 Our work shows although NAD+ boosting has pharmacologic utility in maintaining metabolic

11 health of skeletal muscle, the underlying REDOX status of the cell can be determinant of its

12 impact. Here we find that adaptive upregulation of the NRK pathway is unproductive, yielding

13 no increase in NAD+ or NADH. Furthermore, attempts to exacerbate the H6KO model

14 defective phenotype by ablating the NRK2 gene and producing a double knockout had no

15 impact upon the skeletal muscle. This is supportive of the recent findings regarding the

16 redundant nature of the NRKs, indicating NRK1 and NAMPT adaptive actions. Additionally,

17 attempts to increase NAD+ levels with acute intraperitoneal NR in the H6KO models were

18 successful but had no significant impact over the physiology with many pathways failing to

19 show a response differential from WT mice. Therefore, our work shows for the first time that

20 NAD+ biochemistry and the value of NAD+ boosting as a strategy is dependent upon the

21 metabolic context and REDOX stress environment within the target tissue. These being

22 previously unidentified key determinants in the potential use of NAD+ precursor compounds.

23 Production of a double knockout of H6PD and NRK2 was successful at reducing total levels

24 of skeletal muscle NAD(H) suggesting the existence REDOX influence over the cytosol and

25 mitochondria that may impact metabolic functionality beyond the confined SR/ER

26 environment.

1 List of Abbreviations

2 ER/SR: Endoplasmic/Sarcoplasmic Reticulum.

3 H6PD: Hexose-6-Phosphate Dehydrogenase.

4 NAD+: Nicotinamide Adenine Dinucleotide.

5 NMN: Nicotinamide Mononucleotide.

6 NR: Nicotinamide Riboside.

7 NRK: Nicotinamide Riboside Kinase.

8 TA: Tibialis Anterior

9 SOL: Soleus

11 Author contributions

12 CLD and GGL conceived the study. CLD, GGL and DT wrote the manuscript. Experimental

13 work was performed by CLD, AEZ, LOB, RSF, YH, DC, AG,DW, JA.

15 Conflicts of interest

16 None

18 Acknowledgements

19 We thank Chromadex (Irvine, California) for providing nicotinamide riboside.

21 Funding

22 This work was supported by a Wellcome Trust Senior Research Fellowship awarded to

23 G.G.L. (GGL-104612/Z/14/Z).

25 Availability of data and materials

26 Datasets used in this study are available from the author upon request.

1 References

2 1 Semjonous, N. M. et al. Hexose-6-phosphate dehydrogenase contributes to skeletal muscle 3 homeostasis independent of 11beta-hydroxysteroid dehydrogenase type 1. Endocrinology 4 152, 93-102, doi:10.1210/en.2010-0957 (2011). 5 2 Marbet, P., Klusonova, P., Birk, J., Kratschmar, D. V. & Odermatt, A. Absence of hexose-6- 6 phosphate dehydrogenase results in reduced overall glucose consumption but does not 7 prevent 11beta-hydroxysteroid dehydrogenase-1-dependent glucocorticoid activation. FEBS 8 J 285, 3993-4004, doi:10.1111/febs.14642 (2018). 9 3 Belenky, P., Bogan, K. L. & Brenner, C. NAD+ metabolism in health and disease. Trends 10 Biochem Sci 32, 12-19, doi:10.1016/j.tibs.2006.11.006 (2007). 11 4 Bogan, K. L. & Brenner, C. Nicotinic acid, nicotinamide, and nicotinamide riboside: a 12 molecular evaluation of NAD+ precursor vitamins in human nutrition. Annu Rev Nutr 28, 13 115-130, doi:10.1146/annurev.nutr.28.061807.155443 (2008). 14 5 Schondorf, D. C. et al. The NAD+ Precursor Nicotinamide Riboside Rescues Mitochondrial 15 Defects and Neuronal Loss in iPSC and Fly Models of Parkinson's Disease. Cell Rep 23, 2976- 16 2988, doi:10.1016/j.celrep.2018.05.009 (2018). 17 6 Hou, Y. et al. NAD(+) supplementation normalizes key Alzheimer's features and DNA damage 18 responses in a new AD mouse model with introduced DNA repair deficiency. Proceedings of 19 the National Academy of Sciences of the United States of America 115, E1876-E1885, 20 doi:10.1073/pnas.1718819115 (2018). 21 7 Diguet, N. et al. Nicotinamide Riboside Preserves Cardiac Function in a Mouse Model of 22 Dilated Cardiomyopathy. Circulation 137, 2256-2273, 23 doi:10.1161/CIRCULATIONAHA.116.026099 (2018). 24 8 Fletcher, R. S. et al. Nicotinamide riboside kinases display redundancy in mediating 25 nicotinamide mononucleotide and nicotinamide riboside metabolism in skeletal muscle cells. 26 Mol Metab 6, 819-832, doi:10.1016/j.molmet.2017.05.011 (2017). 27 9 Zhang, H. et al. NAD(+) repletion improves mitochondrial and stem cell function and 28 enhances life span in mice. Science 352, 1436-1443, doi:10.1126/science.aaf2693 (2016). 29 10 Ryu, D. et al. NAD+ repletion improves muscle function in muscular dystrophy and counters 30 global PARylation. Sci Transl Med 8, 361ra139, doi:10.1126/scitranslmed.aaf5504 (2016). 31 11 Canto, C., Menzies, K. J. & Auwerx, J. NAD(+) Metabolism and the Control of Energy 32 Homeostasis: A Balancing Act between Mitochondria and the Nucleus. Cell Metab 22, 31-53, 33 doi:10.1016/j.cmet.2015.05.023 (2015). 34 12 Mouchiroud, L. et al. The NAD(+)/Sirtuin Pathway Modulates Longevity through Activation of 35 Mitochondrial UPR and FOXO Signaling. Cell 154, 430-441, doi:10.1016/j.cell.2013.06.016 36 (2013). 37 13 Canto, C. et al. The NAD(+) precursor nicotinamide riboside enhances oxidative metabolism 38 and protects against high-fat diet-induced obesity. Cell Metab 15, 838-847, 39 doi:10.1016/j.cmet.2012.04.022 (2012). 40 14 Ratajczak, J. et al. NRK1 controls nicotinamide mononucleotide and nicotinamide riboside 41 metabolism in mammalian cells. Nat Commun 7, 13103, doi:10.1038/ncomms13103 (2016). 42 15 Lavery, G. G. et al. Hexose-6-phosphate dehydrogenase knock-out mice lack 11 beta- 43 hydroxysteroid dehydrogenase type 1-mediated glucocorticoid generation. The Journal of 44 biological chemistry 281, 6546-6551, doi:10.1074/jbc.M512635200 (2006). 45 16 Lavery, G. G. et al. Hypoglycemia with enhanced hepatic glycogen synthesis in recombinant 46 mice lacking hexose-6-phosphate dehydrogenase. Endocrinology 148, 6100-6106, 47 doi:10.1210/en.2007-0963 (2007). 48 17 Bujalska, I. J. et al. Lack of hexose-6-phosphate dehydrogenase impairs lipid mobilization 49 from mouse adipose tissue. Endocrinology 149, 2584-2591, doi:10.1210/en.2007-1705 50 (2008).

1 18 Lavery, G. G. et al. Deletion of hexose-6-phosphate dehydrogenase activates the unfolded 2 protein response pathway and induces skeletal myopathy. The Journal of biological 3 chemistry 283, 8453-8461, doi:10.1074/jbc.M710067200 (2008). 4 19 Xia, J., Sinelnikov, I. V., Han, B. & Wishart, D. S. MetaboAnalyst 3.0--making metabolomics 5 more meaningful. Nucleic Acids Res 43, W251-257, doi:10.1093/nar/gkv380 (2015). 6 20 Xia, J. & Wishart, D. S. Web-based inference of biological patterns, functions and pathways 7 from metabolomic data using MetaboAnalyst. Nat Protoc 6, 743-760, 8 doi:10.1038/nprot.2011.319 (2011). 9 21 Tserng, K. Y., Jin, S. J., Kerr, D. S. & Hoppel, C. L. Urinary 3-hydroxydicarboxylic acids in 10 pathophysiology of metabolic disorders with dicarboxylic aciduria. Metabolism 40, 676-682 11 (1991). 12 22 Bergoffen, J., Kaplan, P., Hale, D. E., Bennett, M. J. & Berry, G. T. Marked elevation of urinary 13 3-hydroxydecanedioic acid in a malnourished infant with glycogen storage disease, 14 mimicking long-chain L-3-hydroxyacyl-CoA dehydrogenase deficiency. J Inherit Metab Dis 16, 15 851-856 (1993). 16 23 Strauss, A. W. et al. Molecular basis of human mitochondrial very-long-chain acyl-CoA 17 dehydrogenase deficiency causing cardiomyopathy and sudden death in childhood. 18 Proceedings of the National Academy of Sciences of the United States of America 92, 10496- 19 10500 (1995). 20 24 Oakey, L. A. et al. Metabolic tracing reveals novel adaptations to skeletal muscle cell energy 21 production pathways in response to NAD (+) depletion. Wellcome Open Res 3, 147, 22 doi:10.12688/wellcomeopenres.14898.1 (2018). 23 25 Deloux, R., Tannous, C., Ferry, A., Li, Z. & Mericskay, M. Aged Nicotinamide Riboside Kinase 2 24 Deficient Mice Present an Altered Response to Endurance Exercise Training. Front Physiol 9, 25 1290, doi:10.3389/fphys.2018.01290 (2018). 26 26 Tarrago, M. G. et al. A Potent and Specific CD38 Inhibitor Ameliorates Age-Related Metabolic 27 Dysfunction by Reversing Tissue NAD(+) Decline. Cell metabolism 27, 1081-1095 e1010, 28 doi:10.1016/j.cmet.2018.03.016 (2018). 29 27 Agerholm, M. et al. Perturbations of NAD(+) salvage systems impact mitochondrial function 30 and energy homeostasis in mouse myoblasts and intact skeletal muscle. Am J Physiol 31 Endocrinol Metab 314, E377-E395, doi:10.1152/ajpendo.00213.2017 (2018). 32 28 Uddin, G. M., Youngson, N. A., Sinclair, D. A. & Morris, M. J. Head to Head Comparison of 33 Short-Term Treatment with the NAD(+) Precursor Nicotinamide Mononucleotide (NMN) and 34 6 Weeks of Exercise in Obese Female Mice. Front Pharmacol 7, 258, 35 doi:10.3389/fphar.2016.00258 (2016). 36 29 Frederick, D. W. et al. Loss of NAD Homeostasis Leads to Progressive and Reversible 37 Degeneration of Skeletal Muscle. Cell metabolism 24, 269-282, 38 doi:10.1016/j.cmet.2016.07.005 (2016). 39 30 Tsachaki, M., Mladenovic, N., Stambergova, H., Birk, J. & Odermatt, A. Hexose-6-phosphate 40 dehydrogenase controls cancer cell proliferation and migration through pleiotropic effects 41 on the unfolded-protein response, calcium homeostasis, and redox balance. FASEB J 32, 42 2690-2705, doi:10.1096/fj.201700870RR (2018). 43 31 Knezevic, C. E. et al. Proteome-wide Profiling of Clinical PARP Inhibitors Reveals Compound- 44 Specific Secondary Targets. Cell Chem Biol 23, 1490-1503, 45 doi:10.1016/j.chembiol.2016.10.011 (2016). 46 32 Palermo, A. T. et al. Transcriptional response to GAA deficiency (Pompe disease) in infantile- 47 onset patients. Mol Genet Metab 106, 287-300, doi:10.1016/j.ymgme.2012.05.004 (2012). 48 33 Goody, M. F. et al. NAD+ biosynthesis ameliorates a zebrafish model of muscular dystrophy. 49 PLoS Biol 10, e1001409, doi:10.1371/journal.pbio.1001409 (2012). 50 34 Ferreira, C. R. & Gahl, W. A. Lysosomal storage diseases. Transl Sci Rare Dis 2, 1-71, 51 doi:10.3233/TRD-160005 (2017).

1 35 Ryu, K. W. et al. Metabolic regulation of transcription through compartmentalized NAD(+) 2 biosynthesis. Science 360, doi:10.1126/science.aan5780 (2018). 3 36 Davila, A. et al. Nicotinamide adenine dinucleotide is transported into mammalian 4 mitochondria. Elife 7, doi:10.7554/eLife.33246 (2018). 5

1 Figure legends

2 Figure 1. Unbiased identification of metabolic shifts supported by transcriptional

3 changes. (A) Pathways analysis of unbiased metabolomics from WT (n=3) and H6KO (n=3)

4 Tibialis Anterior Skeletal Muscle. (B) Numerical output from pathway analysis of WT and

5 H6PDH KO Tibialis Anterior Skeletal Muscle. Values shown are original value from pathway

6 analysis. (C-F) Quantification of NAD+, NADH and total NAD(H) in WT (n=7) and H6KO

7 (n=6) TA muscle. (G) Schematic representation of the biosynthetic generation of NAD from

8 nicotinamide Riboside (NR) and nicotinamide (NAM) salvage, enzymes shown in red. (H)

9 qRT-PCR of NAD+ synthesis and salvage genes in WT (n=7) and H6KO (n=7) TA. (I+J)

10 Immunoblots and quantification of WT (n=9) and H6KO (n=9) TA lysates. *P<0.05, **P<0.01

11 and ***P<0.001.

13 Figure 2. Impaired mitochondrial fatty acid oxidation in skeletal muscle due to H6 KO.

14 (A) High resolution respirometry of fatty acid oxidation in permeabilised Tibialis Anterior WT

15 (n=3) in and H6KO (n=3). (B) High resolution respirometry of fatty acid oxidation using WT

16 (n=3) and H6KO (n=3) permeabilised Soleus. (C) Mitochondrial DNA (mtDNA) quantification

17 of WT (n=7) and H6KO (n=7) muscle, measured using qRT-PCR. (D) Immunoblots WT and

18 H6KO protein lysates (n=9) probed for oxidative phosphorylation enzyme subunit

19 abundance. (E) Acylcarnitine species levels in WT (n=9) and H6KO (n=11) muscle

20 measured using GC-MS/MS. Data expressed as heat-maps with z-score values representing

21 metabolite abundance in WT and H6KO. Box and Whisker plots showing significantly altered

22 short acylcarnitines. (F) qRT-PCR measurement of genes critical carnitine and fatty acids

23 metabolism in WT (n=7) and H6KO (n=7) TA. (G) Schematic showing carnitine and fatty acid

24 metabolism between cytosol and mitochondria. (H) Immunoblots of acetylated proteins

25 within WT (n=6) and H6KO (n=6) skeletal muscle. *P<0.05, **P<0.01 and ***P<0.001.

1 Figure 3. H6 KO upregulates the muscle specific metabolic pathway regulating

2 NAD(H) salvage and biosynthesis. (A-D) NAD+, NAD(H) and total NAD(H) quantification

3 in skeletal muscle of WT and H6KO +/- intraperitoneal (i.p.) Nicotinamide Riboside (NR)

4 (n=6-9). (E) qRT-PCR data of genes involved in the biosynthesis of NAD+ and salvage of

5 NAM in WT and H6KO muscle +/- i.p. NR (n=6-9). (F) qRT-PCR data of genes critical

6 carnitine and fatty acids metabolism in WT and H6KO muscle +/- i.p. NR (n=6-9). (G-H)

7 Immunoblots and quantification of the NAM salvage protein NAMPT and the skeletal muscle

8 specific protein NRK2 (n=6). (I) High resolution respirometry of fatty acid oxidation in

9 permeabilised Tibialis Anterior WT in and H6KO i.p NR (n=3). (I) High resolution respiration

10 for fatty acid oxidation using WT and H6KO after i.p NR permeabilised Soleus (n=3). (J)

11 Immunoblots showing ER stress regulator CHOP and protein folding factors PDI and ERO1a

12 in WT and H6KO +/- i.p NR. (K) Immunoblots showing total lysine acetylation and IDH

13 acetylation in WT and H6KO muscle +/- i.p NR. (n=6-9). *P<0.05, **P<0.01 and ***P<0.001.

15 Figure 4. Double Knockout of H6PDH and NRK2 reduces NAD+ in skeletal muscle

16 independently of mitochondrial FAO and REDOX status. (A-D) NAD+, NADH and total

17 NAD(H) levels in WT, H6KO, NRK2KO and DKO (H6KO-NRK2KO) TA muscle (n=3-6). (E-

18 G) Skeletal muscle tissue weights from WT, H6KO, NRK2KO and H6-NRK2 Double

19 Knockout (DKO) (n=3-6). (H) qRT-PCR of NAD+ biosynthetic gene expression in WT, H6KO,

20 NRK2 KO and DKO TA (n=3-6). (I) qRT-PCR of mitochondrial and Acyl-CoA genes in TA of

21 WT, H6KO, NRK2 KO and DKO (n=3-6). (J) qRT-PCR of ER/SR stress response genes in

22 WT, H6KO, NRK2KO and DKO TA (n=3-6). (F-K) Immunoblots of CHOP and protein folding

23 factors PDI and ERO1a in WT, H6KO, NRK2 KO and DKO muscle lysates (n=3-6). (L)

24 Immunoblots of total lysine acetylation, IDH2 acetylation in WT, H6KO, NRK2KO and DKO

25 muscle protein lysates (n=3-6). *P<0.05, **P<0.01 and ***P<0.001.

1 Supplementary Figure 1. Targeted metabolomics of WT and H6KO Muscle, Serum and

2 Liver (n=9-11).

4 Supplementary Figure 2. Metabolic impact of H6PD deficiency upon the skeletal

5 muscle. (A-D) Metabolomic assessment of WT and H6KO Tibialis Anterior measured by LC-

6 MS. (A) Metabolomics analysis of WT (n=9) and H6PDH KO (n=11) Tiabialis Anterior

7 Skeletal Muscle. Data expressed as boxplots representing abundance. (A-D) Box and

8 Whisker plot showing selected (A) Lysophospholipids, (B) Phosphotidylcholines, (C) Amino

9 acids and (D) Sphingomyelins quantification from WT and H6KO TA. Error bars represent

10 minimum and maximum values. *p<0.05, **p<0.01, ***p<0.001. two-tailed unpaired Students

11 t-test. (C)

13 Supplementary Table 1. Unbiased metabolite abundance in H6KO and WT skeletal

14 muscle. (A) Metabolomic assessment of WT (n=3) and H6KO (n=3) Tibialis Anterior

15 measured by LC-MS. Shown in table 1 are the most differentially abundant metabolites in

16 H6KO muscle compared to WT.

bioRxiv preprint doi: https://doi.org/10.1101/567297; this version posted March 4, 2019. The copyright holder for this preprint (which was Fig 1. A. not certified by peer review) is the author/funder, who has granted bioRxivB. a license to display the preprint in perpetuity. It is made 5 available under aCC-BY-ND 4.0 International license. Nicotinamide Metabolism Metabolic Pathway Total Expected Hits Raw p Nicotinamide Metabolism 13 0.12 2 5.73E-03 4 Pyrimidine Metabolism 41 0.38 2 5.21E-02 Nitrogen Metabolism 9 0.08 1 7.98E-02 Pyrimidine Metabolism Histidine Metabolism 15 0.14 1 1.30E-01 3 Pentose Phosphate Pathway 19 0.17 1 1.62E-01

log(p) Alanine, Aspartate and Glutamate Metabolism 24 0.22 1 2.00E-01 - Histidine Metabolism Inositol Phosphate Metabolism 28 0.26 1 2.29E-01 2 Amino Sugar and Nucleotide Sugar Metabolism 37 0.34 1 2.92E-01 Purine Metabolism 68 0.62 1 4.74E-01 Aminoacyl-tRNA Biosynthesis 69 0.63 1 4.79E-01 1

0.00 0.05 0.10 0.15 0.20 0.25 0.30 Pathway Impact C. D. E. F. T T 1 2 0 T 1 2 0 1 2 0 1 2 W W

W *

s s

1 0 0 s 1 0 0 1 0 0 u u

u 9 s s s r 8 0 8 0 r 8 0 r e e e v o v

i v

t 6 0 6 0 6 0 6 e e a e g g R g n

4 0 4 0 n 4 0 n a a

a 3 h h 2 0 h 2 0 2 0 C C

0 0 % 0 0 % N A D + N A D H N A D (H ) T o ta l N A D + /N A D H G. H. NR T * * *

nrk1/2 W 2 5

u 2 0 *

NMN s r

nmnat1/2 e 1 5 v NAAD * * * * * * * * * nampt pnp e 2 art1 nadsyn g ADPr NAD n a

h 1

nadk C

NADP d l 0 o

NAM F 2 n 1 t 1 2 k p 1 k t p k t t r y a r a d n r s m a p a n d n a n n n a m n m I. J. n n n o t H6PDH 1 .5 e * * * v i n t

i 1 .2 l a NAMPT l * * u e b R u 0 .9 e T -

NRK1 g a n

h 0 .6 a p h l C NRK2 A 0 .3 d l o

Tubulin F 0 .0

2 1 T D K K P P WT H6KO R R M 6 N N A H N bioRxiv preprint doi: https://doi.org/10.1101/567297; this version posted March 4, 2019. The copyright holder for this preprint (which was Fig 2. A. not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made C. Tibialis Anterior available under aCC-BY-NDB. 4.0 International license. Soleus

3 0 0 3 0 0 * * * * * 8 0 0 0 s t s

2 2 5 i

s 2 2 5 s n s

* * * ) ) a

a * * * * * * U 6 0 0 0 g

g * * * * m m 1 5 0 1 5 0 y

* * * * * m r r * r * a e

s * e s t ( i ( p / p /

l 7 5 7 5 b 4 0 0 0 r x x o o u A u l

l m m F F p

* * * A p

( ( 2 2 2 0 2 0 N O O D 2 0 0 0 t

1 0 1 0 m

0 0 0 t t k S S P S S X k S S P S S X y a y T a O O X T T O O C O X T T O O e C E e O E E L H r H O E R L H r H R W K I II P P I II 6 P io P I io I O X r X C C O X r X C C H A O P O + A O P O + F I I F I I O C C O C C A A F F

D. E. 3 1 CV-ATP5A 0 -1

CIII-UQCRC2 Score Z -3 C0 C2 CIV-MTCO1 300 150 *** 250 120 CII-SDHB /mg) 200 90 CI-NDUFB8 150

pmol 60 ( 100 Tubulin Acylcarnitines 50 30 WT H6KO 0 0 WT H6KO WT H6KO F. T

W 8 WT H6KO s

u * * * C0:C2 C4 C6 s r 6 * * * 0.8 5 0.8 *** e *** *** v * * * 4 e 4 * * * * * * 0.6 0.6 g /mg) o n i 3 t a

a 0.4 0.4 h

2 R 2 C pmol

(

d 0.2 0.2 l 1

o 0 F 0 0.0 2 1 2 2 1 0.0 r l WT H6KO WT H6KO WT H6KO s s a d s fa a c c c c s a a a H. G. IDH2ac IDH2total H3K9ac H3K18ac Tubulin WT H6KO A. bioRxiv preprint doi: https://doi.org/10.1101/567297B. ; this version posted MarchC. 4, 2019. The copyright holder for this preprintD. (which was Fig 3. not certified by peer** review) is the author/funder, who has granted bioRxiv a license to display the preprint in* perpetuity.** It is made 1 8 0 available1 8 0 under aCC-BY-ND** 4.0 International license1 8 0. 1 2 s s s l l l u u 1 5 0 u 1 5 0 1 5 0 o o o s s * s r r r r r r t t t 9 e e e n n n o

v 1 2 0 v 1 2 0 v 1 2 0 i

o o o t e e e a C C C

g g g

9 0 9 0 9 0 R 6 e e e n n n l l l a a a c c c i i i h h 6 0 h 6 0 6 0 h h h C C C e e e

3 V V V % % 3 0 % 3 0 3 0

0 0 0 0 N A D + N A D H N A D (H ) T o ta l N A D + /N A D H E. F. 8

5 0 e e l l c c 2 5 i i e h e h

g 6 e g e

* * * n V n * V

a a

2 T h T h C

W 4 C

d s d s l l u o u o

1 s s F r F 2 r e e v v

0 0

2 n 1 t 1 2 k p 2 1 2 2 1 k y t p k t d n s r l a r s a r a a a s d n n m n n p s f a c d a n c c c s a m n m a a a n n n n n

i 1 .5

H. o l G. i u s b s 1 .2 e NAMPT u

t * * r

p 0 .9 o x t

NRK2 e 0 .6 v n i i t e a t 0 .3 l

Tubulin o e r r

P 0 .0 WT WT H6KO H6KO N R K 2 N A M P T NR - + - + I. J. Tibialis Anterior Soleus 3 0 0 3 0 0 *** ** s s 2 2 5 2 2 5 s s ** ** ) ) a a ** *** g

g **

m ** m ***

m 1 5 0

m 1 5 0 * r * r * e s e s ( ( p / p /

l 7 5

l 7 5 x o x o *** u u l m l m F p F p

(

2 2 0 2 2 0 O O 1 0 1 0 0 0 k S t S X k S t S P S X y P S S a y S a O O X T T O O O X T T O e C e C L H H O E E R L H r H O E E R r I I I P o P II I P o P I I O X i X O X i X r C C r C C A O P O + A O P O + F I F I I I O C C O C C A A F F K. L. Tubulin Tubulin CHOP Acetylated Lysine ERO1a IDH2ac PDI IDH2Total WT WT H6KO H6KO NR - + - + WT WT H6KO H6KO NR - + - + bioRxiv preprint doi: https://doi.org/10.1101/567297B. ; this version posted March 4,C. 2019. The copyright holder for this preprintD. (which was Fig 4. A. not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is6 made 1 2 5 available1 2 5 under aCC-BY-ND 4.0 International1 2 license5 . s s s * ** ** u u u ** 1 0 0 1 0 0 1 0 0 s s s e e e r r r l l l e e e c c c 4 i i i v v v o

7 5 7 5 7 5 i h h h t e e e e e e a g g g V V V R

n n n 5 0 5 0 5 0 a a a T T T 2 h h h W W W C C C

2 5 2 5 2 5 % % %

0 0 0 0 N A D + N A D H N A D (H ) T o ta l N A D + /N A D H E. F. G. 0 .2 5 6 0 1 0

** ** ) 0 .2 0 ** ) 8 ) g g g

( 4 0 m m

( ( t

0 .1 5 6 t t h h h g i g g i i e 0 .1 0 4 e e

W 2 0 W W 0 .0 5 2

0 .0 0 0 0 Q U A D T A S O L E U S H. I. l * * * l o o 8 r 2 1 r t t

n *** *** 1 8 n o o e e *** *** C C

g 6

5 g *** *** n e n e l l a a c 4 c i h i h *** *** *** *** h h C * * * * * * 4 C

e 3 e d d V l V l

o o s 2 s F F u

u 2 s s r 1 r e e V V

0 0 n rk 2 n rk 1 n a m p t n a d s y n a c s s 2 fa r 1 a c s l2 a c a a 2 s c d 1

J. K. Tubulin Tubulin CHOP Total Lysine Ac ERO1a PDI IDH2ac WT H6KO NRK2KO DKO IDH2Total H3K5ac WT H6KO NRK2KO DKO bioRxiv preprint doi: https://doi.org/10.1101/567297; this version posted March 4, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made Muscleavailable under aCC-BY-ND 4.0 InternationalSerum license. Liver

3 4 5 1 0 0 0 Log2 Log2

Log2 -1 -4 -5 -3

Acylcarnitines

3 4 2 1 0 0

0 Log2 Log2 Log2 -1 -4 -2 -3

Phospholipids

3 4 5 1 0 0 3 Log2 Log2 Log2 -1 -4 Amino acids -3 1

3 30 6 1 20 0 0 Log2 Log2

Log2 10 -1 -6 -3 0 Sphingolipids

WT H6KO WT H6KO WT H6KO bioRxiv preprint doi: https://doi.org/10.1101/567297; this version posted March 4, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Lysophospholipids Phosphatidylcholines

LPC (C16:0) LPC (C17:0) LPC (C18:0) LPC (C18:1) PC (30:0) PC (30:2) PC (32:0) PC (32:1) 80 *** 1.5 * 40 *** 15 *** 50 *** 1.0 * 150 *** 250 *** 40 0.8 120 200 g 60 30 g 1.0 10 m m / / 30 0.6 90 150 l l

o 40 20 o 20 0.4 60 100 m 0.5 5 m p p

l 20 10 l 10 0.2 30 50 e e v v e 0 0.0 0 0 e 0 0.0 0 0 L L

e e

t LPC (C18:2) LPC (C20:3) LPC (C20:4) LPC (C24:1)

t PC (32:2) PC (32:3) PC (36:2) PC (36:3) i i l 25 2.0 6 2.0 l 50 5 300 400

o ** *** ** ** o *** ** ** ** b b a 20 a 40 4 t 1.5 1.5 t 300 e 4 e

o 200 o i i M 15 M 30 3 t t a 1.0 1.0 a 200 R 10 20 2 R 2 100 0.5 0.5 5 10 1 100 0 0.0 0 0.0 0 0 0 0

Sphingomyelins Amino Acids

SM C16:0 SM C16:1 Arginine Glutamine Glycine Ornithine 80 5 600 *** 4000 *** 5000 *** 50 *** *** *** 4 3000 4000 40 60 g g 400

m 3000 30

m 3 / / l

2000 l

o 40

2000 20 o

m 200 2 m p 1000

l 1000 10 p 20

l 1 v 0 0 0 0 e e v L

0 Proline Tryptophan Leucine e 0 e t L i

l 600 45 *** 800 SM C18:1 SM (OH) C22:2

*** ** e o t i b 40 l 12 2.0 a

t 600 *** * o

e 400 o b

35 i t M a t a 400 10 1.5 e 30 R 200 M 25 200 8 1.0 0 20 0 6 0.5

4 0.0

80 80

60 60WT H6KO

40 40

20 20

0 0 bioRxiv preprint doi: https://doi.org/10.1101/567297; this version posted March 4, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license. Compound Differentially Abundant in H6KO Compared to WT muscle p N6,N6,N6-Trimethyl-L-lysine 4.41E-04 L-Methionine S-oxide 3.94E-03 3-Hydroxysebacicacid 7.25E-03 4-Aminobutanoate 7.71E-03 2,3,4-trihydroxy-butanoic acid 9.10E-03 Leu-Gln-Pro 1.08E-02 1H-Imidazole-4-ethanamine 1.13E-02 (R)-1-Aminopropan-2-ol 1.19E-02 2',6'-Dihydroxy-4'-methoxy-3'-prenyldihydrochalcone 1.20E-02 Leu-Pro 1.33E-02 NAD+ 1.67E-02 L-rhamnitol 1.92E-02 L-Proline 1.93E-02 N-Undecanoylglycine 1.96E-02 L-Valine 2.01E-02 cis-5-Tetradecenoylcarnitine 2.02E-02 2,3,4-trihydroxy-butanoic acid 2.12E-02 Creatinine 2.46E-02 Citrate 3.03E-02 4-Aminobutanoate 3.11E-02 PE (18:0/20:4 (5Z,8Z,11Z,14Z)) 3.25E-02 4-Aminobutanoate 3.28E-02 N,N,N-Trimethylserotonin 3.36E-02 Dethiobiotin 3.68E-02 1-Methylnicotinamide 3.97E-02 D-Glucosamine 6-phosphate 4.24E-02 Tetradecanoylcarnitine 4.38E-02 GMP 4.43E-02 O-Butanoylcarnitine 4.43E-02 Choline phosphate 4.54E-02 D-Glucose 6-phosphate 4.79E-02