1 of 26

Systems Insights into Molecular Variability and Divergent Cancer Phenotypes

B. Bishal Paudel, PhD

IMAG/MSM working group

Aug 8, 2024

@paudelbb

2 of 26

Biological systems span several scales and are heterogeneous

Part 1

Part 2

3 of 26

Part 1: Identifying the drivers and mechanisms of tumorigenesis

4 of 26

Different network topologies drive cell-to-cell variability

Single-cell

phenotype

Gardner et al. (2000)

(PMID: 10659857)

Angeli et al. (2004)

(PMID: 14766974)

Lu et al. (2013) (PMID: 24154725)

5 of 26

Different network topologies drive cell-to-cell variability

Single-cell

phenotype

Multi-cell

phenotype

Gardner et al. (2000)

(PMID: 10659857)

Angeli et al. (2004)

(PMID: 14766974)

Lu et al. (2013) (PMID: 24154725)

-AP

+AP

Muthuswamy et al. (2001) (PMID: 11533657)

6 of 26

ErbB activation leads to divergent accumulation of cargo types

Wang*, Paudel* et al. Nat Commun. (2023) (PMID: 37055441)

7 of 26

ErbB activation leads to divergent accumulation of cargo types

Model reconstructed from: Riddick & Macara (2005) (PMID: 15795315)

Wang*, Paudel* et al. Nat Commun. (2023) (PMID: 37055441)

8 of 26

ErbB activation leads to divergent accumulation of cargo types

Model reconstructed from: Riddick & Macara (2005) (PMID: 15795315)

Wang*, Paudel* et al. Nat Commun. (2023) (PMID: 37055441)

9 of 26

ErbB activation leads to divergent accumulation of cargo types

Model reconstructed from: Riddick & Macara (2005) (PMID: 15795315)

Wang*, Paudel* et al. Nat Commun. (2023) (PMID: 37055441)

10 of 26

ErbB receptors internalize and repress the frequency of multicellular outgrowth

Wang*, Paudel* et al. Nat Commun. (2023) (PMID: 37055441)

Knockdown 1

Knockdown 2

11 of 26

Cross-inhibitory feedbacks give rise to switch-like cargo localization

Outgrowth

Wang*, Paudel* et al. Nat Commun. (2023) (PMID: 37055441)

K_D

Switch

12 of 26

Transport signatures and its relevance in racial disparity

Log2 [NC ratio]

NLS Cargo type1

NLS Cargo type2

PMID: 24570268

Collaboration with Dr. Clayton Yates (Johns Hopkins)

Gene expression

Transport signature

13 of 26

High nuclear localization of classical cargoes indicates worse prognosis in AA group NOT in EA

{

European American

High

Low

Survival Probability

Time (months)

African American

Survival Probability

Time (months)

High

Low

14 of 26

Part 2: Characterizing molecular variability and its clinical implications

DISCLOSURES: Penn State Research Foundation has licensed CNL to Keystone Nano, Inc (PA). MK is CTO and co-founder of Keystone Nano. TPL is a member of the SAB of Keystone Nano, Bioniz Therapeutics, Kymera Therapeutics and Dren Bio.

15 of 26

Role of sphingolipid metabolism in Acute Myeloid Leukemia (AML)

Acute myeloid leukemia (AML) is a highly heterogeneous disease.

Several genomic classifications have been proposed for AML, however a majority of patients lack known genomic features.

Therapies have not substantially improved over 30 years.

Patel et al. (2012) (PMID: 22417203)

Ceramide

Sphingomyelin

SMS

Sphingosine

kinase

C1P

Hexosylceramide

GCS

S1P

Ceramidase

Increasing evidence suggest a key role of dysfunctional sphingolipid metabolism in AML.

Sphingolipid regulate a wide range of cellular processes, and depending on the species, they are either pro- or anti-apoptotic.

Dysregulation of the pathway and its balance contributes to many diseases including cancer.

Sphingolipid Pathway

16 of 26

Sphingolipidomic based AML patient stratification?

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

17 of 26

AML separates into two distinct sphingolipidomic subtypes

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

18 of 26

Sphingolipidome integrates patient cases and cell lines better than global gene expression

Lipidomics-based clustering

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

19 of 26

Sphingolipidomic subtypes differ in treatment and survival outcomes

Primary Response

Survival Outcome

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

20 of 26

Sphingolipidomic subtypes differ in gene expression and abundances of distinct cell types

Van Galen et al. (2019) (PMID: 30827681)

Zeng et al. (2022) (PMID: 35618837)

Log₂(Score)

Undifferentiated cell types

Myeloid cell types

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

21 of 26

Sphingolipidomic subtypes differ gene expression and abundances of distinct cell types

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

SM^hi

SM^lo

Stemness

Differentiation

22 of 26

Machine-learning classifier can infer sphingolipidomic subtypes from gene expression

Machine learning approach

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

23 of 26

Sphingolipidomic subtype, SM^hi is a high-risk subtype with poor clinical outcome

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

24 of 26

Sphingolipidomic subtype, SM^hi stratifies clinically ambiguous intermediate risk group

Intermediate Risk Group:

~ 25-30% patients

‘Basket category’

Favorable and Adverse risk groups not significant.

Paudel et al. Blood Adv. (2024) (PMID: 37131653)

25 of 26

Part 2: Summary and Future Work

Sphingolipid-focused metabolic profiles identify two distinct AML subtypes.

High sphingomyelin subtype is a previously unrecognized high-risk subtype with poor clinical outcome.

Sphingolipid is an independent predictor of outcomes —could lead to new patient stratification.

Sphingolipid subtypes exhibit differential drug sensitivity offering new therapeutic options for high-risk AML subtype.

26 of 26

Acknowledgements

Kevin A. Janes, PhD

Thomas P. Loughran Jr., MD

David J. Feith, PhD

UVA Cancer Center

AML P01 Team: P01CA171983

Janes Lab @ UVA

Collaborators

Francine Garrett-Bakelman, MD, PhD (UVA)

David Claxton, MD (Penn State)

Clayton Yates, PhD (Johns Hopkins)

Charles Chalfant, PhD (UVA)

Todd Fox, PhD (UVA)

Jeff Smith, PhD (UVA)

Kelsey Fisher-Wellman, PhD (ECU)

Walter Coulter Translational Research Grant