Metabolite Imaging to Identify Drug Resistant Breast Cancer

Institution: Lawrence Berkeley National Laboratory
Investigator(s): Trent Northen, Ph.D. -
Award Cycle: 2009 (Cycle 15) Grant #: 15IB-0063 Award: $171,892
Award Type: IDEA
Research Priorities
Detection, Prognosis and Treatment>Innovative Treatment Modalities: search for a cure



Initial Award Abstract (2009)

A major challenge in the treatment of breast cancer is the effective application of chemotherapy. Selecting therapies based on the clinical and molecular characteristics of the tumor has the potential for an individualized, more effective, and less toxic treatment. Nearly 50% of breast cancer patients demonstrate primary and/or secondary resistance to doxorubicin, a drug used in four out of the six protocols defined by the American Cancer Society (CAF, AC, AC+, and A-CMF). The ability to subtype breast cancers based on doxorubicin sensitivity would therefore allow for improved treatment. Doxorubicin like many other chemotherapeutics activates enzymes that catalyze the production of ceramide—a metabolite known to induce apoptosis. Many studies link ceramide (sphingolipid) expression to drug resistance: (1) Elevated levels of glucosylceramide synthase (GCS) and the corresponding elevation in glucospingosines suggest that apoptosis can be avoided through altered ceramide metabolism, (2) Overexpression of the ceramide transport protein (CERT) in resistant cancers suggest a role of ceramide transport in drug resistance, (3) Recent reports also link GCS, glycolipids, and over expression of p-glycoprotein (ABC transporter of the MDR/TAP subfamily) which is widely associated with drug resistance. Therefore, sphingolipids represent a promising class of molecular markers of doxorubicin resistant breast cancer. Recent advances in mass spectrometry make it possible to detect, identify and quantify spingolipids from tissues. However, existing approaches require tissue homogenization/extraction averaging out locally altered and clinical important levels within breast tumors, often associated with clusters of single tumor cells.

Our novel approach is to utilize state-of-the-art nanotechnology (Nanostructure-Initiator Mass Spectrometry, NIMS) imaging to identify both the metabolic ‘fingerprint’ (i.e. metabolomics) of resistance as well as the heterogeneity within individual tumors. The aim is to utilize NIMS metabolite imaging technology to detect locally elevated levels of sphingolipids at the cellular level to define heterogeneity and identify drug resistant/sensitive breast cancers within individual tumors. For these studies we, first, plan to use drug resistant breast cancer cell lines to define a set of metabolic markers of resistance. The NIMS technique has a unique combination of high lateral resolution (10-75 γm), sensitivity (single cancer cells), and the ability to image metabolites at the cellular level within frozen tissue sections. The breast cancer samples include a diverse set of archival breast tumors (basal, luminal, ER+/-, PR+/-, and HER2+/-), and a 50+ cell line collection that has been well-characterized by our collaborator, Dr. Joe Gray at LBL and UCSF.

The successful development and implementation of the proposed metabolomics and imaging technologies for the detection and prognosis of drug resistant breast cancers will allow clinicians to distinguish between sensitive and drug sensitive cancers.




Final Report (2011)

One of the major challenges in treating breast cancer is selecting the most effective chemotherapy. Approaches that allow personalized therapies based on the clinical and molecular characteristics of the tumor have the potential for more effective and less toxic treatment. Nearly 50% of breast cancer patients demonstrate primary and/or secondary resistance to doxorubicin (aka adriamycin), a drug used in four out of the six protocols defined by the American Chemical Society (CAF, AC, AC+, and A-CMF). While tumors are stained and viewed by a pathologist to distinguish cancer cells based on morphology, there are molecular differences not currently ‘visible’ to the pathologist that have the potential to distinguish between drug resistant and sensitive breast cancers. Metabolites (sugars, amino acids, etc) are one class of molecular features that are altered in drug resistant breast cancers. Therefore, new technologies that can detect and ‘fingerprint’ drug resistant tumors using metabolites would enable individualized treatment based on predicted drug resistance of the tumor.

We have developed a new-to-the-world nanotechnology (Nanostructure-Initiator Mass Spectrometry, NIMS) which allows us to ‘image’ metabolites associated with drug resistance within tumors. This approach has a unique combination of attributes: it can identify metabolites from complex mixtures, is very sensitive (highest reported, including single breast cancer cells), has been show to detect metabolites associated with drug resistance in breast cancer tumors, and can image tissues at the cellular level. We hypothesized that there is heterogeneity within tumors with respect to their drug sensitivity and that our NIMS metabolite imaging technology will detect locally elevated levels of lipids at the cellular level to define heterogeneity and identify drug resistant/sensitive breast cancers within individual tumors.

To test this hypothesis we used a combination of analysis of drug resistant breast cancer cell lines using a comprehensive ‘metabolomics’ approach [Bowen and Northen 2009, Curr Opin Microbiol. 12(5):547-52)] to establish the unique molecular signatures of resistance and gain new insights into resistance. These signatures were studied by imaging primary breast tumors to define the heterogeneity within tumors, the range of these metabolites between tumors, and association with existing subtypes and markers. This work revealed dramatic heterogeneity even between sections in individual tumors and established a need for a comprehensive 3D analysis of tumors to further examine heterogeneity. To analyze the heterogeneity within the tumors we developed technical and multivariate statistical approaches resulting in the first 3D metabolite imaging of a tumor (Integr Biol (Camb). 3(4):460-7, 2011). The heterogeneity in the tumor was examined in detail and linked to existing histopathological markers. Therefore metabolite imaging has the ability to resolve metabolic heterogeneity within tumors and has potential to inform application of drug therapies for more effective breast cancer treatment.




Symposium Abstract (2010)

Wolfgang Reindl, Ben P. Bowen, and Trent R. Northen (PI)

The goal of this project is to use our recently developed tissue imaging technique, Nanostructure-Initiator Mass Spectrometry (NIMS), for the identification of metabolites as biomarkers which can be used to discriminate between different breast cancer subtypes. Functional tissue imaging provides a tremendous opportunity to gain insights into the pathological processes of breast cancer and it allows to distinguish the effects of different tumor tissue microenvironments. Current approaches are focused on genomic and proteomic imaging. However, the importance of the cellular metabolism in cancer pathology coupled with the utility of small molecule biomarkers make it critical to develop complementary metabolite imaging approaches. Unfortunately, technical limitations of existing mass spectrometry approaches have largely limited this possibility. We have developed NIMS as a new surface based mass spectrometry approach that is well suited for metabolite profiling and imaging from frozen tissue sections combining high lateral resolution (10-75 μ m), sensitivity (highest reported including single cancer cells), and lack of matrix. Preliminary imaging results reveal dramatic metabolic differences between normal and tumorous breast tissue. Breast cancer is a very heterogeneous disease for which several subtypes (e.g. basal and luminal) can be distinguished and for which a broad range of drug sensitivities or resistance can be observed, making it difficult to select the most efficient treatment. The availability of metabolic biomarkers for particular drug responses would be an important resource for the selection of the appropriate medication.



Conserved features of cancer cells define their sensitivity to HAMLET-induced death
Periodical:Oncogene
Index Medicus: Oncogene
Authors: Storm, P; Aits, S; Puthia, MK; Urbano, A; Northen, T; Powers, S; Bowen, B; Chao, Y; Reind
Yr: 2011 Vol: 30 Nbr: June 6 Abs: Pg:4765-4779

Dealing with the Unknown: Metabolomics and Metabolite Atlases.
Periodical:Journal of the American Society for Mass Spectrometry
Index Medicus: J Am Soc Mass Spectrom
Authors: Bowen BP, Northen TR
Yr: 2010 Vol: 21 Nbr: 9 Abs: Pg:1471-76, ePub

Multivariate analysis of a 3D mass spectral image for examining tissue heterogeneity.
Periodical:Journal of Intergative Biology
Index Medicus: J Intergrative Biol
Authors: Reindl W, Bowen BP, Balamotis MA, Green JE, Northen TR
Yr: 2011 Vol: 3 Nbr: 4 Abs: Pg:460-7

Conserved features of cancer cells define their sensitivity to HAMLET-induced death; c-Myc and glycolysis.
Periodical:Oncogene
Index Medicus: Oncogene
Authors: Storm P, Aits S, Puthia MK, Urbano A, Northen T, et al.
Yr: 2011 Vol: ePub Nbr: June Abs: Pg:1-15