Analysis of Protein Turnover Rates in Skyline with the TurnoveR External Tool

Nathan Basisty, PhDTranslational Geroproteomics Unit, NIH

Skyline User MeetingOctober 27th, 2021

@nbasistyNathan.basisty@nih.gov

Basisty et al. Proteomics. 2018

Metabolic Labeling and Mass Spectrometry Enable Comprehensive Measurement of Protein Turnover Rates

Ad lib Calorie Restricted (CR)

Metabolic Labeling and Mass Spectrometry Enable Comprehensive Measurement of Protein Turnover Rates

TurnoveR

Basisty et al. Unpublished

Mike MacCoss(UW)

Brendan MacLean(UW)

Nicholas Shulman(UW)

Ali Marsh

Recommendations for Using the TurnoveR Tool in Skyline

iRT alignment

mProphet peak scoring

Skyline Document

Populate Skyline w/ all isotope permutations

Labeling experiment and MS

TurnoveR

1. Isotope analysis

2. Regressions3. Statistics4. Reports

Workflow for TurnoveR: A Skyline External Tool for the Analysis of Protein Turnover from Metabolic Labeling Studies

p < 0.0001p < 0.0001

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

Workflow for TurnoveR: A Skyline External Tool for the Analysis of Protein Turnover from Metabolic Labeling Studies

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

TurnoveR Calculates “True Isotope Distributions”

Unlabeled

0 LH 1 LH 2 LH

36% 48% 16%

M M + 3 M + 61 Leucine 2 Leucine

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

TurnoveR Calculates “True Isotope Distributions”

Unlabeled

0 LH 1 LH 2 LH

36% 48% 16%

M M + 3 M + 61 Leucine 2 Leucine

Brauman et al. 1966. Least Squares Analysis and Simplification of Multi-Isotope Mass Spectra

Unlabeled

0 LH 1 LH 2 LH

39% 47% 14%

M M + 3 M + 61 Leucine 2 Leucine

Natural isotope correction prevents systematic overestimation of heavy label enrichment

36% 48% 16%

Precursor pool correction is required for accurate calculation of turnoverLL = light leucineLH = heavy leucine

LL

LHLH

LL

LLProtein synthesis

40% heavy labeled

40% newly synthesized?

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

LH

LH

LH

LH

LH

LHLH

LH

LH

LH

LH

LH

100% of precursor leucines are heavy

LL = light leucineLH = heavy leucine

LL

LHLH

LL

LLProtein synthesis

40% heavy labeled

40% newly synthesized

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

Precursor pool correction is required for accurate calculation of turnover

New protein is synthesized from a mixture of heavy and light leucine

LL

LHLH

LL

LLProtein synthesis

40% heavy labeled

LL = light leucineLH = heavy leucine

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

LL

LH

LL

LL

LL

LLLL

LL

LL

LH

LH

LH

40% of precursor leucines are heavy

Precursor pool correction is required for accurate calculation of turnover

New protein is synthesized from a mixture of heavy and light leucine

LL

LHLH

LL

LLProtein synthesis

40% heavy labeled

100% newly synthesized

LL = light leucineLH = heavy leucine

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

LL

LH

LL

LL

LL

LLLL

LL

LL

LH

LH

LH

40% of precursor leucines are heavy

Precursor pool correction is required for accurate calculation of turnover

Always overestimated

Half-life without precursor RIA = 12.9 days

Half-life with precursor RIA = 6 days‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

Hsieh E, Shulman N … MacCoss. 2012. Molecular & Cellular Proteomics

Precursor pool correction is required for accurate calculation of turnover

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

Precursor pool correction is required for accurate calculation of turnover

CTL = ControlCR = Calorie Restricted

‘true’ isotope distributions

precursor pool enrichment

% newly synthesized

Turnover regressions (slopes, half-lives)

Statistical comparisons (Treated vs untreated)

TurnoveR Computational Pipeline

Precursor pool correction is required for accurate calculation of turnover

p = 0.0164 p < 0.0001

p < 0.0001 p = 0.0031

CTL = ControlCR = Calorie Restricted

TurnoveR Recapitulates Results Calculated by Topograph Software.

TurnoveR

Questions to Address w/ Protein Turnover

• Which proteins turnover during age-related muscle atrophy?

• How does acylation affect protein turnover?

• Is longer half-life a biomarker for interventions that extend lifespan?

• What is the turnover of insoluble proteins during aging?

Christopher Adams(Mayo Clinic) Scott EbertMatthew Miller

ATF4 is a Key Promoter of Age-Related Skeletal Muscle Atrophy

(Gadd45a, p21)

?

TA

Quad?

ATF4 May Modulate Atrophy by Reducing Protein Synthesis

(Gadd45a, p21)

?

TA

Quad?

ATF4 Alters Turnover of a Subset of Proteins

TurnoveR

Isocitrate Dehydrogenase II (IDH2)

ATF4 Reverses an Age-Associated Decrease in IDH2 Turnover

(Gadd45a, p21)

ATF4 May Modulate Atrophy by Altering Protein Turnover

Altered turnover of: Mitochondrial proteins (IDH2)Myofibrillar proteins

AcknowledgementsSchilling Lab (Buck):Birgit SchillingJacob RoseCameron WehrfritzSandip PatelSamah ShahFrancesco NeriAmy O’BroinJoanna BonsPierre-Yves DesprezChristina King

MacCoss Lab: (Univ. of Washington)Michael MacCossBrendan MacLeanNicholas ShulmanAlexandra Marsh

Funding:1K99AG065484-01A1, U01AG060906, Glenn Foundation

Rabinovitch Lab: (Univ. of Washington)Peter RabinovitchYing Ann ChiaoPabalu KarunadharmaDao-Fu DaiEllen QuarlesJeannie Fredrickson

Translational Geroproteomics Unit (TGU)Nathan BasistyNathan.basisty@nih.gov

Hiring Postdocs!

Adams Lab: (Mayo Clinic)Christopher AdamsScott EbertMatthew Miller Translational Geroproteomics Unit (NIA)

Anjana RamReema Banarjee

Thank you, Nat Bruce!