Coherent anti-Stokes Raman scattering imaging of lipids in cancer metastasis
© Le et al; licensee BioMed Central Ltd. 2009
Received: 16 October 2008
Accepted: 30 January 2009
Published: 30 January 2009
Lipid-rich tumours have been associated with increased cancer metastasis and aggressive clinical behaviours. Nonetheless, pathologists cannot classify lipid-rich tumours as a clinically distinctive form of carcinoma due to a lack of mechanistic understanding on the roles of lipids in cancer development.
Coherent anti-Stokes Raman scattering (CARS) microscopy is employed to study cancer cell behaviours in excess lipid environments in vivo and in vitro. The impacts of a high fat diet on cancer development are evaluated in a Balb/c mice cancer model. Intravital flow cytometry and histology are employed to enumerate cancer cell escape to the bloodstream and metastasis to lung tissues, respectively. Cancer cell motility and tissue invasion capability are also evaluated in excess lipid environments.
CARS imaging reveals intracellular lipid accumulation is induced by excess free fatty acids (FFAs). Excess FFAs incorporation onto cancer cell membrane induces membrane phase separation, reduces cell-cell contact, increases surface adhesion, and promotes tissue invasion. Increased plasma FFAs level and visceral adiposity are associated with early rise in circulating tumour cells and increased lung metastasis. Furthermore, CARS imaging reveals FFAs-induced lipid accumulation in primary, circulating, and metastasized cancer cells.
Lipid-rich tumours are linked to cancer metastasis through FFAs-induced physical perturbations on cancer cell membrane. Most importantly, the revelation of lipid-rich circulating tumour cells suggests possible development of CARS intravital flow cytometry for label-free detection of early-stage cancer metastasis.
Excess lipid in the body has been shown to aggravate cancer. Animal studies showed that high fat diets and obesity enhanced cancer metastasis . In human, a body mass index above 30 kg/m2 is strongly correlated with increased risk for various types of cancer . It is generally accepted that diet and obesity are accountable for 30% of preventable causes of cancer . Indeed, diet, nutrition, and physical activity are widely promoted as effective means for cancer prevention by the World Cancer Research Fund and the American Institute for Cancer Research . Nonetheless, the relationship between diet and cancer incidence and mortality in human remains controversial due to conflicting experimental results from clinical studies [4, 5]. Currently, a consensus is lacking on the benefits of a certain type of fatty acid or nutritional ingredient to cancer prevention . A contributing factor to the controversy is the lack of mechanistic understanding on how excess lipid or adiposity affects cancer development [2, 6].
Excess lipid at the cellular level has also been associated with cancer aggressiveness. Intracellular lipid body accumulation has been observed in many types of cancers including mammary, brain, adrenal, and others [7–9]. Early clinical studies in the 1970s of patients with lipid-rich carcinoma of mammary glands found high incidence of cancer mortality, metastatic tumours, and other aggressive clinical behaviours . Since then, lipid-rich carcinoma continued to be reported widely in human and animal [10, 11]. Nonetheless, the relationship between intracellular lipid accumulation and cancer behaviour has not been investigated. Without a mechanistic understanding of the role of lipid in cancer development, pathologists are refrained from classifying lipid-rich tumours as a morphologically and clinically distinctive form of carcinoma .
To investigate the roles of lipid in cancer development, we employ a lipid-sensitive imaging technique called coherent anti-Stokes Raman scattering (CARS) microscopy . CARS is a four wave mixing process where two synchronized lasers, pump and Stokes, are tightly focused into a diffraction-limited focal volume. The interaction of the pump field at frequency ωp and the Stokes field at frequency ωS with the medium generates an anti-Stokes field at frequency 2ωp-ωS. CARS signal is significantly enhanced when ωp-ωS matches a Raman-active vibrational band. Furthermore, the intrinsic coherent property allows CARS signal to increase quadratically with respect to the number of molecular vibrations in the focal volume. Such property renders CARS highly sensitive to lipid-rich structures when ωp-ωS matches the symmetric CH2 stretch vibration at 2840 cm-1. CARS microscopy has been widely applied as a label-free imaging technique to visualize lipid bilayers, cell membranes, adipocytes, myelin sheaths, foam cells of atheroma, and others [13, 14]. An additional unique advantage of CARS microscopy is its intrinsic capability for multimodal imaging. A typical CARS microscope with picosecond pulse excitation is capable of simultaneous CARS, sum frequency generation (SFG), and two-photon excitation fluorescence (TPEF) imaging. Such multimodal imaging capability has allowed characterization of the impact of obesity on the composition and architecture of mammary tumour stroma . Here, CARS microscopy is employed to elucidate the mechanistic link between lipid-rich tumours and aggressive tumour behaviours.
Multimodal nonlinear optical microscopy
A multimodal nonlinear optical (NLO) microscope capable of CARS, SFG, and TPEF imaging on a single platform has been previously described . For CARS imaging, the wave number difference ωp-ωS was tuned to 2840 cm-1 which matches the Raman shift of symmetric CH2 stretch vibration. The same CARS laser sources were used for SFG and TPEF imaging. CARS, SFG, and TPEF (green fluorescence/DIOC18/FITC/GFP) signals were collected through a 600/65 nm (Ealing Catalog, Rocklin, CA, Cat. No. 42-7336), a 375/50 nm (Chroma, Rockingham, Vermont, Cat. No. HQ375/50), and a 520/40 nm (Chroma, Rockingham, Vermont, Cat. No. HQ520/40) bandpass filters, respectively. To image TPEF signal for red fluorescence (Rh-DOPE or RFP), CARS laser sources were desynchronized such that there was no contribution from CARS signal to fluorescence signal. Red fluorescence TPEF signal was collected through a 600/65 nm (Ealing Catalog, Rocklin, CA, Cat. No. 42-7336) bandpass filter. For SFG and TPEF imaging, backward-reflected signals were collected. For CARS imaging of cell cultures, forward signals (F-CARS) were collected. For CARS imaging of tissue samples, backward-reflected signals (E-CARS) were collected. The combined laser power at the sample was kept constantly at 40 mW.
Tumour cell line and growth medium
Madison (M109) lung carcinoma cell line of Balb/c mice origin was a generous gift from P. Low (Purdue Cancer Center, Purdue University, West Lafayette, IN). M109 cells were grown in RPMI-1640 medium supplemented with 10% fetal bovine serum and antibiotics penicillin (50 U)/streptomycin (50 μg). A dual-color labelled M109 cell line was created by stable transfection of M109 cells with a dual-reporter plasmid which expresses both GFP and RFP.
To evaluate the impact of diet on tumour metastasis, Balb/c mice (6–8 weeks old) were subcutaneously injected with M109 cells (1 million cells in 0.1 ml PBS buffer per mouse) in the hind leg area. Then mice (week 0) were placed on two types of diet, a lean diet (20 mice) and a high fat diet (12 mice). Lean diet (Harlan Teklad, Indianapolis, IN, Cat. No. 7001) has 4.25% fat and 3.82 Kcal/g. High fat diet (Research Diets, New Brunswick, New Jersey, Cat. No. D12492) has 34.9% fat and 5.24 Kcal/g. On week 2, 4 mice from each diet group were sacrificed and visceral fat weights were measured. For the remaining mice, weight, tumour size, and number of circulating tumour cells were monitored for 4 weeks after tumour implantation. After 4 weeks, 8 mice from each diet group had terminal blood samples drawn to analyze for circulating tumour cells and fatty acid levels. Lung and primary tumour tissues were also collected and analyzed with nonlinear optical imaging or with histology. 8 mice on normal diet were kept until week 5, then lung tissues were collected for lung metastasis analysis. Furthermore, 4 mice on normal diet and 4 mice on high fat diet were injected with sterilized PBS buffer (0.1 ml) in the hind leg area for control experiments. All animal experiments were approved by Purdue Animal Care and Use Committee.
A dual-reporter plasmid to label M109 cells
A dual-reporter plasmid was constructed to label M109 cells with both GFP and RFP. A ~3.5 kb DNA fragment is generated by total synthesis (Genscript, Piscataway, NJ) and cloned into a pDSRed-Express-DR plasmid (Cat. No. 632423, Clontech, Mountain View, CA) between restriction sites XhoI and BamHI. This DNA fragment comprises a 1.4 kb fragment of vascular endothelial growth factor (VEGF) promoter (nucleotides -963 to +404) controlling the expression of a GFP (0.8 kb, Cat. No. 632428, Clontech, Mountain View, CA), a SV40 polyA transcription termination signal (0.25 kb), and a 1 kb fragment of collagen type I (col1a1) promoter (nucleotides -915 to +116). In the final plasmid construct, GFP expression is controlled by VEGF promoter and RFP expression is controlled by col1a1 promoter. The excitation and emission maxima for GFP are 496 nm and 506 nm, and for RFP are 557 nm and 579 nm, respectively.
Stable transfection of M109 cells
M109 cells at density of 1 × 105 in 35 mm culture dishes were transfected with dual-reporter plasmid DNA using Fugene6 transfection reagent (Cat. No. 11815091001, Roche Diagnostics, Indianapolis, IN) at the ratio of 6 μl of Fugene6 reagent to 1 μg DNA. At 24 hours after transfection, cells were selected with 400 ng/ml Geneticin G418 Sulfate (Cat. No. 10131, Invitrogen, Carlsbad, CA) until individual colonies were observed under a microscope. Colonies were detached from culture dishes by incubation with non-enzymatic cell dissociation reagent (Cat. No. 1676949, MP Biomedicals, Solon, OH) for 5 minutes at 37°C. Individual colonies were carefully removed using pipette tips under a microscope. To ensure a single clone is used, a selected colony was diluted into multi-well plates at the concentration of one cell per well and re-grown in 100 ng/ml Geneticin G418 Sulfate (Cat. No. 10131, Invitrogen, Carlsbad, CA).
Intravital flow cytometry
To enumerate the number of circulating tumour cells in the blood, intravital flow cytometry in superficial veins of a mouse ear was employed. Using F-CARS or transmission microscopy, a vein of ~8 μm in diameter which has a steady flow rate of ~1000 cells/min was selected. This small diameter vein ensures that blood cells of ~5 μm in diameter flow through the vein in a single file. Additionally, fluctuation in detection signal due to varying cell position along the vertical axis of the detection volume is minimized because of restricted vessel volume. A laser scanning line (~8 μm) which spans the entire vein diameter was defined perpendicular to the direction of blood flow. The laser scanning speed of ~1.3 ms/line ensured sampling of all flow-through blood cells. In each mouse, two different veins were selected for measurement. Four sampling windows of ~1 minute per window were used to enumerate circulating tumour cells in each mouse. The average count of circulating tumour cells and the standard deviation between sampling windows were plotted for each mouse against time after tumour implantation. To visualize dual-labeled circulating tumour cells, a 488 nm Argon laser was used to excite GFP and a 543 nm Helium-Neon laser was used to excite RFP simultaneously. Emission from GFP and RFP were collected through a 520/40 nm and 600/65 nm filters, respectively. The power of each laser was kept constantly at 100 μW at the sample. Images were acquired using Fluoview software and processed using ImageJ software. High frequency noise was removed using ImageJ despeckle and Gaussian filter functions. The detection accuracy of circulating tumour cells is enhanced by the cross-correlation of signals arising from both GFP and RFP within the same cell.
Isolation and identification of circulating tumour cells
At 4 weeks after tumour implantation, mice were subjected to terminal blood collection. On average, approximately 1 ml of blood was collected from each mouse. Whole blood was centrifuged at 300 × g for 10 minutes at 4°C to separate plasma from blood cells. After plasma was removed for FFAs quantitation and GC-MS analysis, pellets were reconstituted in 10 ml red blood cell lysis buffer (eBioscience, San Diego, CA, Cat. No. 00-4333-57) and incubated for 5 minutes at room temperature. Then 20 ml of PBS buffer was added to stop lysis reaction. The sample was centrifuged at 300 × g for 10 minutes at 4°C to remove lysed red blood cells from other blood components. Pellets were reconstituted in supplemented RPMI medium and plated onto a glass-bottom chamber for NLO imaging. To detect circulating tumour cells (CTC), simultaneous F-CARS imaging for lipid detection and epi-reflected TPEF imaging for green fluorescent protein (GFP) detection were performed. Cells screened positive for lipid and GFP signals were further imaged for red fluorescent protein (RFP) using desynchronized laser sources for excitation and 600/65 nm filter for epi-reflected TPEF emission signals.
Histology analysis of lung metastasis
Lung tissues were collected at 4 or 5 weeks after tumour implantation and kept in 10% buffered formalin. Paraffin-embedded sections of ~5 μm were mounted on glass slides and stained with hematoxylin and eosin. Histology samples were analyzed using an Eclipse E400 microscope (Nikon, Tokyo, Japan) and a Spot Insight Camera (Diagnostic Instrument, Sterling Heights, Michigan). Tumour colonies were identified based on the density of stained nuclei.
Evaluating the impacts of VF conditioned medium on M109 cells
Visceral fat tissues (0.3 g) were collected from Balb/c mice at 8 to 12 weeks old, added to a culture dish containing 2 ml RPMI medium, and incubated at 37°C with 5% CO2 for 4 days. Conditioned medium (CM) was removed and added to approximately 1 million M109 cells in a 35 mm glass-bottom culture dish (MatTek, Ashland, MA, Cat. No. P35G-0-10-C). M109 cells in CM were kept in an incubator at 37°C with 5% CO2. The impacts of CM on M109 were monitored over time using CARS imaging and biochemical assays.
Isolation of FFAs from CM or blood plasma
4 ml of hexane (Sigma-Aldrich, St Louis, MO, Cat. No. 296090) was added to 1 ml of CM or blood plasma and vortex-mixed. The mixture was centrifuge at 300 × g for 5 minutes to separate organic phase (top) from aqueous phase (bottom). Organic phase containing hexane and fatty acids was removed into a new container. Steady stream of nitrogen gas was used to evaporate hexane for 2 hours. FFAs were reconstituted into supplemented RPMI medium and used immediately or stored at -80°C. Aqueous phase which contains cytokines was used immediately or stored at -80°C.
Probing fatty acids induced phase separation on cell membrane
To evaluate how different type of FFAs perturb tumour cell membrane, two membrane dyes were employed, DIOC18 (3,3'-dioctadecyloxacarbocyanine perchlorate, Invitrogen, Carlsbad, CA, Cat. No. D275) to probe liquid-ordered (Lo) phase and Rh-DOPE (rhodamine B sulfonyl dioleoyl phosphatidyl ethanolamine, Avanti Polar Lipids, Alabaster, AL, Cat. No. 810150) to probe liquid-disordered (Ld) phase on cell membrane. Final concentration of both DIOC18 and Rh-DOPE dyes used was at 25 μg/ml. For visualization of dye-labeled membrane, a 488 nm Argon laser was used to excite DIOC18 and a 543 nm Helium-Neon laser was used to excite Rh-DOPE. Emission signal was collected through a 520/40 nm filter for DIOC18 and 600/65 nm filter for Rh-DOPE. To avoid signal cross-over or possible FRET activity, the membrane was imaged sequentially using one laser source at a time. First, only the 488 nm Argon laser was used to visualize DIOC18 signal. Then, the Argon laser was blocked and the 543 nm Helium-Neon laser was used to visualize Rh-DOPE signal. The power of each laser was kept constantly at 80 μW at the sample.
Three-dimensional M109 cells migration assay
In these assays, migration of M109 cells into explanted visceral fat tissues (VF) was examined. VF (0.3 g) was placed directly above the image area such that VF came into direct contact with M109 cells. After 24 hours of incubation, VF was imaged for invading M109 cells.
Two-dimensional M109 cells migration assay
In these assays, membrane morphology and lipid content of migrating M109 cells along the bottom surface of a culture dish were examined. A polydimethylsiloxane (PDMS) well of 10 mm in diameter was used to seed ~10,000 M109 cells in 0.2 ml supplemented RPMI medium onto a culture dish for 24 hours. Seeded M109 cells were washed several times with fresh medium to remove unattached cells, then culture medium was removed with aspiration. The PDMS well holding seeded M109 cells was removed and the culture dish was filled with 2 ml of supplemented RPMI medium. VF (0.3 g) was placed in the culture dish away from the seeded M109 cells. M109 cells were allowed to migrate toward VF for 24 hours, then imaged with transmission and CARS microscopy.
Extracellular matrix (ECM) M109 cell invasion assay
The ability of M109 cells to invade reconstituted basement membrane was evaluated using a QCM 96-Well Cell Invasion Assay kit (Chemicon, Temecula, CA, Cat. No. ECM 555). Experiments were performed according to manufacturer's protocol. We evaluated ECM invasion capability of four different pre-treated M109 cells toward four different chemoattractants. Pre-treated M109 cells are: 1) M109: M109 cells starved 24 hours by incubating in serum-free RPMI medium, 2) M109/CM: M109 cells in CM for 4 days, 3) M109/FFAs: M109 cells incubated for 4 days with extracted FFAs from CM, 4) M109/cytokines: M109 cells incubated for 4 days with extracted cytokines from CM. All pre-conditioned M109 cells were adjusted to 1 million cells per ml prior to ECM invasion assay. Chemoattractants are: 1) RPMI: complete RPMI medium with 10% supplemented fetal bovine serum, 2) CM: conditioned medium obtained by incubating 0.3 g VF in 2 ml of RPMI medium for 4 days, 3) FFAs: extracted FFAs from CM, 4) cytokines: extracted cytokines from CM. M109 cells were allowed to migrate across ECM membrane for 24 hours, then collected and assayed with CyQuant GR dye. Fluorescence reading was recorded with a multi-well fluorescence plate reader (Gemini XPS, Molecular Devices, Sunnyvale, CA) using a 480/520 nm filter set.
Results and discussion
Primary tumour growth was unaffected by a high fat diet
High fat diet induced early increase of circulating tumour cells
High fat diet increased metastasis to lung tissues
FFAs induced membrane phase separation and intracellular lipid accumulation
While all four free fatty acids induced intracellular lipid accumulation, only polyunsaturated linoleic acid induced cell polarity observable with CARS imaging (Fig. S5, see Additional file 2). Because cellular contents concentrated toward Lo phase of the membrane, an increase in Ld phase due to incorporation of polyunsaturated linoleic acid separated a single cell into two distinctive poles with one rich and the other poor in cellular contents. CARS imaging of content-rich cell poles which include cell membrane and lipid droplets and of content-poor cell poles which consist mainly of cell membrane yielded strong and weak CARS contrast, respectively (Fig. 6E). On the other hand, saturated FFAs-induced membrane separation was observable only with fluorescent imaging of phase probing dyes, but not with CARS imaging of lipid-rich structures (Fig. 6F). Hence, cell polarity observable with CARS imaging of CTCs isolated from HD mice strongly indicated incorporation of polyunsaturated FFAs into CTCs membranes (Fig. 3D). Furthermore, the presence of intracellular lipid in primary, circulating, and metastasized cells clearly indicated the exposure of cells to excess FFAs.
FFAs induced cell polarity and increased surface adhesion capability
FFAs reduced cell-cell contact leading to increased chemotactic motility
Visceral adipose tissues are generally accepted as endocrine organs which regulate body lipid and energy homeostasis [22, 23]. Increased visceral adipose mass is strongly associated with hyperlipidemia, elevated organ adiposity, altered free fatty acids metabolism, and abnormal adipokines secretion . Here, we showed that secreted cytokines and FFAs from visceral adipose tissues exerted profound impacts on cancer cell motility and tissue invasion. Secreted cytokines attracted cancer cells; whereas, secreted FFAs perturbed cancer cell membrane leading to reduced cell-cell contact. Direct evidence of FFAs exposure was indicated by intracellular lipid accumulation in migrating cancer cells. While the negative impacts of increased visceral adipose tissues on cancer development had been focused mainly on altered chemical signaling induced by secreted cytokines, we showed that physical perturbations on cancer cell membrane induced by secreted FFAs also contributed to increased cancer aggressiveness. Taken together, our in vitro data supported a correlation between increased visceral adiposity or increased blood plasma FFAs level and increased cancer aggressiveness through the action of FFAs on cancer cells.
FFAs impacts on M109 cells were observable in mammary cancer
In this study, we present a mechanistic dissection into the relationship between excess lipid and cancer aggressiveness. Using an animal cancer model, we observe mice with excess visceral adipose tissue or plasma FFAs due to a high fat diet experience early bodyweight loss, early appearance of high number of CTCs, and increased lung metastasis. CARS imaging reveals significant lipid accumulation in primary, circulating, and metastasized tumour cells. Furthermore, CTCs isolated from HD mice exhibit polarized distribution of lipid bodies and cytoplasmic proteins. Using M109 cell cultures, we show that FFAs induce intracellular lipid accumulation. FFAs incorporation into cell membrane causes membrane rounding, which leads to reduced cell-cell adhesion and increased tissue invasion. Moreover, polyunsaturated FFAs induce membrane phase separation and polarized distribution of surface, cytoplasmic, and cytoskeletal proteins. Cell polarity induced by polyunsaturated FFAs, and observed in CTCs of HD mice, increases surface binding capability. This observation could explain increased extravasation and metastasis of CTCs of HD mice. Taken together, both in vitro and in vivo studies strongly associate the effects of excess FFAs on cancer cell membrane with increased risk of cancer aggressive behaviours.
In addition to cancer cells, many other cell types also accumulate lipid in elevated serum FFAs condition . Indeed, excess intracellular lipid accumulation in non-adipose tissues is associated with insulin resistance, cell death, and heart failure . Currently, it is hypothesized that lipid accumulation is a general defence mechanism against lipotoxicity . By channelling excess fatty acids toward lipid metabolism pathway and away from apoptosis pathway, cells improve their survival . Many non-adipocyte cells, which have limited capacity to convert FFAs into neutral lipid, have been shown to alternatively incorporate FFAs into cell membrane, leading to increased cell motility , increased surface adhesion , and altered transmembrane signaling . It is reasonable to believe that cancer cells also accumulate lipid to protect themselves against lipotoxicity (Fig. S9, see Additional file 2). However, unintended physical perturbations induced by FFAs on cancer cell membrane cause cancer cells to lose adhesion to neighbouring cells, become more susceptible to migration, and have increased opportunity to extravasate from bloodstream. Lipid-rich tumours could be a consequence of excess FFAs in tumour microenvironments. Therefore, treatment or classification of lipid-rich carcinoma should account for the risks induced by excess FFAs.
In our animal model studies, diet-induced increase of plasma FFAs level accelerates cancer metastasis. Nonetheless, the contribution of FFAs to cancer metastasis should be viewed in the context of a complex in vivo environment where multiple risk factors interact. For instance, our co-culture experiments show a strong influence of CM cytokines on cancer cell migration (Fig. 8D). Previous studies have also identified a number of other metastasis risk factors in tumour microenvironment including stem cells , fibroblasts , macrophages , collagen fibers , and others. Furthermore, FFAs have been shown to exert immunosuppressive effect by perturbing T cell membrane and inhibiting T cell signal transduction . Given many cell types can be induced to accumulate intracellular lipid, it is unlikely that FFAs exert their impact specifically on cancer cells. To fully understand the impact of FFAs on cancer metastasis, the effects of FFAs on other cell types, particularly tumour stromal cells, should be comprehensively investigated.
Finally, the revelation of lipid-rich CTCs by CARS imaging could enable future development of label-free intravital CARS flow cytometry for early-stage diagnosis of cancer metastasis. Using our current nonlinear optical microscopy set-up, intravital CARS flow cytometry is yet able to discriminate lipid-rich CTCs from lipid-poor blood cells in a microvessel. Possible explanations could be attributed to CARS low sensitivity to flowing cells , interference of lipid-rich components of mouse skins  which descramble CARS signals arising from flowing cells, or un-optimized laser scanning speed per blood flow speed ratio . However, recent advances in optics suggest that challenges for intravital CARS flow cytometry can be overcame. Such advances include adaptive optics which minimizes optical distortion in thick tissues , CARS signal generation from a single picosecond synchronously pumped optical parametric oscillator which increases penetration depth , and video-rate CARS microscopy which improves laser scanning speed . Furthermore, advances in CARS endoscopy  also suggest possible clinical application of CARS imaging in the near future. The discovery of lipid-rich CTCs presented in this paper should facilitate translational development of CARS-based imaging tools for clinical cancer diagnosis.
coherent anti-Stokes Raman scattering
circulating tumour cells
conditioned medium of 0.3 g visceral fat tissue
free fatty acids
green fluorescent protein
high fat diet
red fluorescent protein
sum frequency generation
two-photon excitation fluorescence
visceral fat tissue.
This work is supported by a NIH postdoctoral fellowship F32HL089074 to TTL and a NIH R21 grant EB004966 to JXC. The authors thank IG Camarillo, K Buhman and RT Phan for insightful discussion.
- Rose DP, Connolly JM, Meschter CL: Effect of Dietary-Fat on Human Breast-Cancer Growth and Lung Metastasis in Nude-Mice. J Natl Cancer Inst. 1991, 83 (20): 1491-1495. 10.1093/jnci/83.20.1491.View ArticlePubMedGoogle Scholar
- Calle EE, Kaaks R: Overweight, obesity and cancer: Epidemiological evidence and proposed mechanisms. Nat Rev Cancer. 2004, 4 (8): 579-591. 10.1038/nrc1408.View ArticlePubMedGoogle Scholar
- World Cancer Research Fund/American Institute for Cancer Research: Food, nutrition, physical activity, and the prevention of cancer: a global perspective. 2007, Washington DC: AICRGoogle Scholar
- Thiebaut ACM, Schatzkin A, Ballard-Barbash R, Kipnis V: Dietary fat and breast cancer: Contributions from a survival trial. J Natl Cancer Inst. 2006, 98 (24): 1753-1755.View ArticlePubMedGoogle Scholar
- Fontham ETH: Diet and lung cancer. Cancer Causes & Control. 1997, 8 (6): 819-820. 10.1023/A:1018432916540.View ArticleGoogle Scholar
- MacLean CH, Newberry SJ, Mojica WA, Khanna P, Issa AM, Suttorp MJ, Lim YW, Traina SB, Hilton L, Garland R, Morton Sc: Effects of omega-3 fatty acids on cancer risk – A systematic review. JAMA. 2006, 295 (4): 403-415. 10.1001/jama.295.4.403.View ArticlePubMedGoogle Scholar
- Ramos CV, Taylor HB: Lipid-rich carcinoma of the breast: A clinicopathologic analysis of 13 examples. Cancer. 1973, 33 (3): 812-819. 10.1002/1097-0142(197403)33:3<812::AID-CNCR2820330328>3.0.CO;2-4.View ArticleGoogle Scholar
- Metser U, Miller E, Lerman H, Lievshitz G, Avital S, Even-Sapir E: F-18-FDG PET/CT in the evaluation of adrenal masses. J Nucl Med. 2006, 47 (1): 32-37.PubMedGoogle Scholar
- Sijens PE, Levendag PC, Vecht CJ, vanDijk P, Oudkerk M: H-1 MR spectroscopy detection of lipids and lactate in metastatic brain tumors. NMR Biomed. 1996, 9 (2): 65-71. 10.1002/(SICI)1099-1492(199604)9:2<65::AID-NBM397>3.0.CO;2-N.View ArticlePubMedGoogle Scholar
- Rosen PP: Rosen's Breast Pathology. 2008, New York: Lippincott Williams & Wilkins, 556-556. 3Google Scholar
- de los Monteros AE, Hellmen E, Ramirez GA, Herraez P, Rodriguez F, Ordas J, Millan Y, Lara A, de las Mulas JM: Lipid-rich carcinomas of the mammary gland in seven dogs: Clinicopathologic and immunohistochemical features. Vet Pathol. 2003, 40 (6): 718-723. 10.1354/vp.40-6-718.View ArticleGoogle Scholar
- Cheng JX, Xie XS: Coherent anti-Stokes Raman scattering microscopy: Instrumentation, theory, and applications. J Phys Chem B. 2004, 108 (3): 827-840. 10.1021/jp035693v.View ArticleGoogle Scholar
- Cheng JX: Coherent anti-Stokes Raman Scatteriong microscopy. Appl Spectrosc. 2007, 61 (9): 197-208.View ArticlePubMedPubMed CentralGoogle Scholar
- Hellerer T, Axang C, Brackmann C, Hillertz P, Pilon M, Enejder A: Monitoring of lipid storage in Caenorhabditis elegans using coherent anti-Stokes Raman scattering (CARS) microscopy. Proc Natl Acad Sci USA. 2007, 104 (37): 14658-14663. 10.1073/pnas.0703594104.View ArticlePubMedPubMed CentralGoogle Scholar
- Le TT, Rehrer CW, Huff TB, Nichols MB, Camarillo IG, Cheng JX: Nonlinear optical imaging to evaluate the impact of obesity on mammary gland and tumor stroma. Mol Imaging. 2007, 6 (3): 205-211.PubMedPubMed CentralGoogle Scholar
- He W, Wang HF, Hartmann LC, Cheng JX, Low PS: In vivo quantitation of rare circulating tumor cells by multiphoton intravital flow cytometry. Proc Natl Acad Sci USA. 2007, 104 (28): 11760-11765. 10.1073/pnas.0703875104.View ArticlePubMedPubMed CentralGoogle Scholar
- Hanahan D, Weinberg RA: The hallmarks of cancer. Cell. 2000, 100 (1): 57-70. 10.1016/S0092-8674(00)81683-9.View ArticlePubMedGoogle Scholar
- Baumgart T, Hammond AT, Sengupta P, Hess ST, Holowka DA, Baird BA, Webb WW: Large-scale fluid/fluid phase separation of proteins and lipids in giant plasma membrane vesicles. Proc Natl Acad Sci USA. 2007, 104 (9): 3165-3170. 10.1073/pnas.0611357104.View ArticlePubMedPubMed CentralGoogle Scholar
- Hoover RL, Fujiwara K, Klausner RD, Bhalla DK, Tucker R, Karnovsky MJ: Effects of Free Fatty-Acids on the Organization of Cytoskeletal Elements in Lymphocytes. Mol Cell Biol. 1981, 1 (10): 939-948.View ArticlePubMedPubMed CentralGoogle Scholar
- Palmantier R, George MD, Akiyama SK, Wolber FM, Olden K, Roberts JD: cis-polyunsaturated fatty acids stimulate β(1) integrin-mediated adhesion of human breast carcinoma cells to type IV collagen by activating protein kinases C-e and -μ. Cancer Res. 2001, 61 (6): 2445-2452.PubMedGoogle Scholar
- Ghosh PK, Vasanji A, Murugesan G, Eppell SJ, Graham LM, Fox PL: Membrane microviscosity regulates endothelial cell motility. Nat Cell Biol. 2002, 4 (11): 894-900. 10.1038/ncb873.View ArticlePubMedGoogle Scholar
- Rosen ED, Spiegelman BM: Adipocytes as regulators of energy balance and glucose homeostasis. Nature. 2006, 444 (7121): 847-853. 10.1038/nature05483.View ArticlePubMedPubMed CentralGoogle Scholar
- Despres JP, Lemieux I: Abdominal obesity and metabolic syndrome. Nature. 2006, 444 (7121): 881-887. 10.1038/nature05488.View ArticlePubMedGoogle Scholar
- Listenberger LL, Han XL, Lewis SE, Cases S, Farese RV, Ory DS, Schaffer JE: Triglyceride accumulation protects against fatty acid-induced lipotoxicity. Proc Natl Acad Sci USA. 2003, 100 (6): 3077-3082. 10.1073/pnas.0630588100.View ArticlePubMedPubMed CentralGoogle Scholar
- Kopelman PG: Obesity as a medical problem. Nature. 2000, 404 (6778): 635-643.PubMedGoogle Scholar
- Stulnig TM, Berger M, Sigmund T, Raederstorff D, Stockinger H, Waldhausl W: Polyunsaturated fatty acids inhibit T cell signal transduction by modification of detergent-insoluble membrane domains. J Cell Biol. 1998, 143 (3): 637-644. 10.1083/jcb.143.3.637.View ArticlePubMedPubMed CentralGoogle Scholar
- Karnoub AE, Dash AB, Vo AP, Sullivan A, Brooks MW, Bell GW, Richardson AL, Polyak K, Tubo R, Weinberg RA: Mesenchymal stem cells within tumour stroma promote breast cancer metastasis. Nature. 2007, 449 (7162): 557-563. 10.1038/nature06188.View ArticlePubMedGoogle Scholar
- Kalluri R, Zeisberg M: Fibroblasts in cancer. Nat Rev Cancer. 2006, 6 (5): 392-401. 10.1038/nrc1877.View ArticlePubMedGoogle Scholar
- Wyckoff JB, Wang Y, Lin EY, Li JF, Goswami S, Stanley ER, Segall JE, Pollard JW, Condeelis J: Direct visualization of macrophage-assisted tumor cell intravasation in mammary tumors. Cancer Res. 2007, 67 (6): 2649-2656. 10.1158/0008-5472.CAN-06-1823.View ArticlePubMedGoogle Scholar
- Provenzano PP, Eliceiri KW, Campbell JM, Inman DR, White JG, Keely PJ: Collagen reorganization at the tumor-stromal interface facilitates local invasion. BMC Med. 2006, 4: 38-10.1186/1741-7015-4-38.View ArticlePubMedPubMed CentralGoogle Scholar
- Wang HW, Bao N, Le TT, Lu C, Cheng JX: Microfluidic CARS cytometry. Opt Express. 2008, 16 (8): 5782-5789. 10.1364/OE.16.005782.View ArticlePubMedPubMed CentralGoogle Scholar
- Evans CL, Potma EO, Puoris'haag M, Cote D, Lin CP, Xie XS: Chemical imaging of tissue in vivo with video-rate coherent anti-Stokes Raman scattering microscopy. Proc Natl Acad Sci USA. 2005, 102 (46): 16807-16812. 10.1073/pnas.0508282102.View ArticlePubMedPubMed CentralGoogle Scholar
- Wright AJ, Poland SP, Girkin JM, Freudiger CW, Evans CL, Xie XS: Adaptive optics for enhanced signal in CARS microscopy. Opt Express. 2007, 15 (26): 18209-18219. 10.1364/OE.15.018209.View ArticlePubMedGoogle Scholar
- Ganikhanov F, Carrasco S, Xie XS, Katz M, Seitz W, Kopf D: Broadly tunable dual-wavelength light source for coherent anti-Stokes Raman scattering microscopy. Opt Lett. 2006, 31 (9): 1292-1294. 10.1364/OL.31.001292.View ArticlePubMedGoogle Scholar
- Legare F, Evans CL, Ganikhanov F, Xie XS: Towards CARS endoscopy. Opt Express. 2006, 14 (10): 4427-4432. 10.1364/OE.14.004427.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2407/9/42/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.