Micro-NMR for rapid molecular analysis of human tumor samples - PubMed (original) (raw)

Micro-NMR for rapid molecular analysis of human tumor samples

Jered B Haun et al. Sci Transl Med. 2011.

Abstract

Although tumor cells obtained from human patients by image-guided intervention are a valuable source for diagnosing cancer, conventional means of analysis are limited. Here, we report the development of a quantitative micro-NMR (nuclear magnetic resonance) system for rapid, multiplexed analysis of human tumors. We implemented the technology in a clinical setting to analyze cells obtained by fine-needle aspirates from suspected lesions in 50 patients and validated the results in an independent cohort of another 20 patients. Single fine-needle aspirates yielded sufficient numbers of cells to enable quantification of multiple protein markers in all patients within 60 min. Moreover, using a four-protein signature, we report a 96% accuracy for establishing a cancer diagnosis, surpassing conventional clinical analyses by immunohistochemistry. Our results also show that protein expression patterns decay with time, underscoring the need for rapid sampling and diagnosis close to the patient bedside. We also observed a surprising degree of heterogeneity in protein expression both across the different patient samples and even within the same tumor, which has important implications for molecular diagnostics and therapeutic drug targeting. Our quantitative point-of-care micro-NMR technique shows potential for cancer diagnosis in the clinic.

PubMed Disclaimer

Figures

Fig. 1

Fig. 1

The μNMR clinical analysis system (DMR-3) and bioconjugation strategy (BOND-2). (Left, top) Complete μNMR system for use at the patient's bedside. The bottom component contains all of the circuitry for NMR measurements, whereas the top enclosure holds a permanent magnet and chip-sized microliter-volume sensors. (Left, bottom) State-of-the-art μNMR probe used for sensing within the mini magnet. (Right, top) Smart phone interface for operating the μNMR system. (Right, bottom) Fine-needle aspirates from each patient sample were processed with a bio-orthogonal amplification strategy adapted for clinical samples. The bio-orthogonal amplification allows ultrasensitive detection of cellular proteins.

Fig. 2

Fig. 2

Validation of NMR measurements. (Top) Plots showing the correlation of EGFR measurements obtained by μNMR versus ELISA, FACS, or immunohistochemistry (IHC) in clinical samples where sufficient cells were available for conventional proteomic techniques (typically 105 to 106 for ELISA and FACS versus 102 for μNMR). Note the excellent correlation coefficients for the different methods. (Bottom) Representative immunofluorescence stains of a representative human sample. The primary antibody (green) was labeled with Alexa Fluor 488 and TCO. The magnetic nanoparticle (MNP) (red) was labeled with VT680 and Tz. Note colocalization between nanoparticles (conferring NMR properties) and antibody (indicating protein content). For additional calibration data, see figs. S1 and S2. AU, arbitrary unit.

Fig. 3

Fig. 3

Distribution of protein expression markers. Waterfall plots showing the expression levels of each of the different biomarkers sorted from high (left) to low (right). Each column represents a different patient sample (green, malignant; blue, benign).

Fig. 4

Fig. 4

Expression of different protein biomarkers arranged by patient number. Patients 5, 12, 17, 18, 21, and 42 had benign lesions, and the remainder had various epithelial cancers (breast, gastrointestinal, genitourinary, gynecological, lung, pancreatic, or undifferentiated).

Fig. 5

Fig. 5

A graphical representation of the Spearman correlation coefficients (0, no correlation; 1, perfect correlation) between protein markers created to examine their interrelationship. For example, EGFR and HER2 have a good correlation (coefficient = 0.6), whereas EpCAM and HER2 have a poor correlation (coefficient = 0.1).

Fig. 6

Fig. 6

Variability of protein marker expression stratified by diagnosis and by global leukocyte versus nonleukocyte comparisons. (Left) Individual marker expression for both malignant and benign samples. (Right) Overall leukocyte and nonleukocyte cell counts.

Fig. 7

Fig. 7

ROC curves. ROC curves were calculated for single protein markers, for a dual marker set, as well as for triple and quadruple marker combinations to determine optimum μNMR threshold values. Az, area under the curve; 95% CL, 95% confidence limits.

Fig. 8

Fig. 8

A representative clinical case illustrating the potential use of μNMR in cancer diagnostics. Patient 3 underwent CT-guided biopsy for an enlarging (2.5 cm by 6.8 cm) presacral lesion in the setting of active metastatic rectal adenocarcinoma. Both cytology and core biopsy assessed the lesion as benign (inflammatory tissue). The lesion was thus treated with a drainage catheter. μNMR analysis, using the quadruple marker combination (MUC-1 + HER2 + EGFR + EpCAM), unequivocally classified the lesion as malignant (aggregate μNMR value was 11.25 and well above the malignancy threshold of ≥1.6). Repeat chest and abdomen CT after 2 months revealed a significant interval enlargement of the biopsied lesion as well as new metastases.

Fig. 9

Fig. 9

Analysis of sample heterogeneity. (A) Repeat measurement of the same samples (note the different scale compared to other graphs). (B) Measurement of repeat fine-needle aspirate samples obtained with the same coaxial needle (see table S3 for variance component estimates for intrasubject variability). (C) Measurement of repeat fine-needle aspirates from different tumor sites. (D and E) Effect of prospective preservation treatments on extracellular and intracellular protein measurements. Live, living cells; FA, 2% formaldehyde; Meth, 100% methanol; TX, 0.05% Triton X-100 in PBS; FB1, Fix Buffer 1; Sap, saponin. Asterisk, optimized conditions chosen for subsequent experiments. (F) Effect of time at 4°C before fixation (for example, during transport to a central laboratory facility) on protein measurements. There was a rapid change in protein expression in unfixed samples.

Similar articles

Cited by

References

    1. Ramaswamy S, Ross KN, Lander ES, Golub TR. A molecular signature of metastasis in primary solid tumors. Nat. Genet. 2003;33:49–54. - PubMed
    1. Taylor BS, Varambally S, Chinnaiyan AM. A systems approach to model metastatic progression. Cancer Res. 2006;66:5537–5539. - PubMed
    1. Maheswaran S, Sequist LV, Nagrath S, Ulkus L, Brannigan B, Collura CV, Inserra E, Diederichs S, Iafrate AJ, Bell DW, Digumarthy S, Muzikansky A, Irimia D, Settleman J, Tompkins RG, Lynch TJ, Toner M, Haber DA. Detection of mutations in EGFR in circulating lung-cancer cells. N. Engl.J.Med. 2008;359:366–377. - PMC - PubMed
    1. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100:57–70. - PubMed
    1. Spencer SL, Gaudet S, Albeck JG, Burke JM, Sorger PK. Non-genetic origins of cell-to-cell variability in TRAIL-induced apoptosis. Nature. 2009;459:428–432. - PMC - PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources