HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only All Versions of this Article: 2223010812v1 222/3/814    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bremer, C. Articles by Weissleder, R. Search for Related Content PubMed Articles by Bremer, C. Articles by Weissleder, R.

Imaging of Differential Protease Expression in Breast Cancers for Detection of Aggressive Tumor Phenotypes¹

¹ From the Center for Molecular Imaging Research, Massachusetts General Hospital and Harvard Medical School, Boston. Received April 19, 2001; revision requested May 25; revision received June 22; accepted July 16. Supported in part by NIH NOI-C0M065, P50-CA86355-01, and a Dana Farber grant. C.B. supported by the Deutsche Forschungsgemeinschaft. Address correspondence to C.B., Department of Clinical Radiology, University of Muenster, Albert-Schweitzer-Strasse 33, D-48129 Muenster, Germany (e-mail: bremerc@umi-muenster.de).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To determine if different expression levels of tumor cathepsin-B activity in well differentiated and undifferentiated breast cancers could be revealed in vivo with optical imaging.

MATERIALS AND METHODS: A well differentiated human breast cancer (BT20, n = 8) and a highly invasive metastatic human breast cancer (DU4475, n = 8) were implanted orthotopically in athymic nude mice. Tumor-bearing animals were examined in vivo with near-infrared fluorescence (NIRF) imaging 24 hours after intravenous injection of an enzyme-sensing imaging probe. Immunohistochemistry, Western blotting (on cells and whole tumor samples), and correlative fluorescence microscopy were performed.

RESULTS: Both types of breast cancers activated the NIRF probe so that tumors became readily detectable. However, in tumors of equal size, there was a 1.5-fold higher fluorescence signal in the highly invasive breast cancer (861 arbitrary units ± 88) compared with the well differentiated lesion (566 arbitrary units ± 36, P < .01). Western blotting confirmed a higher cathepsin-B protein content in the highly invasive breast cancer (DU4475) of about 1.4-fold (whole tumor samples) to 1.7-fold (cells). Immunohistochemistry and fluorescence microscopy findings confirmed the imaging findings.

CONCLUSION: Cathepsin-B enzyme activity can be determined in vivo with NIRF optical imaging, while differences in tumoral expression may correlate with tumor aggressiveness.

© RSNA, 2002

Index terms: Breast neoplasms, experimental studies • Infrared and near-infrared spectroscopy • Animals

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
A variety of proteases are involved in the degradation of the extracellular matrix and basement membranes and, therefore, are key factors in cancer progression (1). Thiol proteases such as cathepsin-B correlate with invasiveness and metastatic capabilities in many tumors (2). In breast cancer specifically, high expression levels of cathepsin-B have been linked to highly aggressive tumors and poor clinical outcome (3–10). A series of new fluorescent imaging agents, which facilitate measurement of specific protease enzyme activity in vivo (11,12), have recently been developed. These agents are converted from their nonfluorescent native state to a highly fluorescent activated state by enzymatic cleavage of fluorochromes from a macromolecular carrier molecule (11). In past studies (11,13), a fluorescent probe with high selectivity for cathepsin-B enzyme activity has been used to depict tumors in the submillimeter range in murine models.

This study was conducted to determine if different expression levels of tumoral cathepsin-B activity in well differentiated and undifferentiated breast cancers could be revealed in vivo with optical imaging.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Imaging Probe
A cathepsin-B–sensitive probe was synthesized as follows. Multiple fluorochrome residues (Cy5.5; Amersham, Piscataway, NJ) were bound to a long circulating graft copolymer consisting of a poly-lysine backbone sterically shielded through methoxy polyethylene glycol, or MPEG, side chains (11). Approximately 44 unmodified lysines that served as cleavage sites for k-k recognizing proteases including cathepsin-B were present in each probe. Overall, the graft copolymer contained an average of 92 methoxy polyethylene glycol side chains and 12 molecules of the near-infrared fluorochrome Cy5.5 (excitation and emission maxima were 675 nm and 694 nm, respectively). In this nonactivated state, fluorochromes are positioned in close proximity to each other, which results in mutual energy transfer and thus quenching of fluorochromes. After enzymatic cleavage of the backbone, the fluorochromes are released, resulting in a bright fluorescence signal that can be detected in vivo. Thus, baseline fluorescence in vitro ranged around only 100 arbitrary units (at 0.24 fmol of Cy5.5) and increased up to 3,000 arbitrary units 24 hours after enzymatic activation. Fluorescence measurements were performed with a fluorometer (U4500; Hitachi, Tokyo, Japan), by using an excitation wavelength of 675 nm and an emission wavelength of 694 nm.

Cell Culture and Characterization of Cell Lines
Two human breast adenocarcinoma cell lines were obtained from the American Tissue Culture Collection (ATCC, Manassas, Va). The first cell line was harvested from a primary lesion of a well differentiated adenocarcinoma (BT20) (14), and the second cell line was derived from a metastatic nodule of a highly aggressive breast adenocarcinoma (DU4475) (15). These tumors were chosen because of their known biologic behavior (aggressive vs nonaggressive). Moreover, a previous study with a macromolecular carrier similar to the one used in this study, but labeled with gadopentetate dimeglumine, revealed similar vascularity and probe delivery in both tumors. Both cell lines were cultured in minimum essential medium (Cellgro; Mediatech, Washington, DC) supplemented with 2 mM L-glutamine and Earle salt solution, 1.5 g/L sodium bicarbonate, 0.1 mM nonessential amino acids, 1.0 mM sodium pyruvate, and 10% heat-inactivated fetal bovine serum at 37°C in a humidified 5% CO₂ atmosphere.

For Western blot analysis, Petri dishes with approximately 80% confluent cells were lysed (20 mM Tris pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% Triton X-100) for 20 minutes at 4°C. The cell lysate was aspirated and centrifuged for 10 minutes at 16,000 rpm. Western blotting of whole tumors was carried out by first homogenizing tissues in a 50 mM Tris buffer (pH 7.5) containing 0.25% Triton X-100 and complete protease inhibitor (Boehringer-Mannheim, Indianapolis, Ind). Minced tissue was then repeatedly sonicated and centrifuged at 10,000 rpm for 10 minutes.

The protein content of the samples was measured with a bicinchoninic acid assay (BCA; Pierce, Rockford, Ill). Equal amounts of protein were applied to a 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis, or SDS-PAGE. Following separation, proteins were blotted onto a polyvinylidene difluoride membrane (Bio-Rad, Hercules, Calif). After blocking, immunoblots were incubated with an antihuman cathepsin-B polyclonal primary antibody (Santa Cruz Biotechnology, Santa Cruz, Calif), followed by incubation with a biotinylated secondary antibody (Vector Laboratories, Burlingame, Calif). Binding of the antibodies was revealed by using an avidin-peroxidase conjugate and diaminobenzidine as a peroxidase substrate (Vecstastain; Vector Laboratories). To reveal murine rather than tumor-produced human cathepsin-B, additional Western blotting was performed by using an antimouse cathepsin-B polyclonal primary antibody (Santa Cruz Biotechnology) and a similar protocol as described. All samples for Western blotting were tested in triplicate.

In Vivo Studies
To assess the in vivo biologic behavior (tumor take and tumor growth) of both tumors, 3 x 10⁶ cells of either cell line were implanted separately in a subset of animals (female athymic nude mice, n = 4 for DU4475; n = 6 for BT20). Tumors were grown for 3 weeks, excised, weighed, and processed for histologic study.

For imaging studies, tumors were coimplanted in a second set of animals with a 7-day offset to yield similar tumor sizes at the time of imaging experiments. Three million cells of BT20 (eight animals) or DU4475 (eight animals) breast adenocarcinomas were injected subcutaneously into the mammary fat pad of female athymic nude mice. Tumors measured about 2–3 mm in diameter at the time of imaging. Animals were anesthetized by using intraperitoneal injection of ketamine (90 mg per kilogram of body weight) and xylazine (10 mg/kg) and the cathepsin-B–sensitive probe containing 2 nmol of fluorochrome was administered by using tail vein injection 24 hours prior to the imaging study. A subset of animals (n = 3) was imaged prior to injection of the fluorochrome to assess near-infrared autofluorescence of the tumors. All animal studies were approved by the animal care committee at our institution.

Imaging was performed by using a NIRF reflectance imaging system, which has been described in detail (13). Briefly, the system consisted of a light-tight chamber equipped with a halogen white light source and an excitation bandpass filter (Omega Optical, Brattleboro, Vt) providing excitation wavelengths of 610–650 nm (13). Fluorescence was depicted by using a 12-bit monochrome charged-coupled device camera (Kodak, Rochester, NY) equipped with a 12.5–75-mm zoom lens and an emission bandpass filter at 680–720 nm (Omega Optical). Exposure time was 30 seconds per image, with maximum input photon flux delivered by the light source. Images were analyzed by using commercially available software (Digital Science 1D software; Kodak). Regions of interest (200–500 pixels for each location) were placed by one of the authors (C.B.) within the visible tumor margins, the adjacent skin, and a reference standard containing 10 nM free Cy5.5 fluorescent dye.

Histologic Studies
Tumors were excised, fixed in phosphate-buffered formalin for 24 hours, paraffin embedded, and cut into 8-µm sections. Sections were incubated with a primary anti-cathepsin-B polyclonal antibody (Santa Cruz Biotechnology) and an alkaline phosphatase–labeled secondary antigoat antibody (Sigma, St Louis, Mo). Endogenous alkaline phosphatase activity was eliminated by heating (65°C for 30 minutes), and specific alkaline phosphatase activity was visualized by using a substrate (NBT/BCIP; Boehringer-Mannheim). Sections were counterstained with nuclear fast red dye. Control sections were processed identically but without incubation of the primary antibody.

To determine the spatial location and extent of the imaging probe, NIRF microscopy was performed on frozen tumor sections by using an inverted epifluorescence microscope (Zeiss Axiovert, Thornwood, NY). Excitation and emission wavelength were 650 and 700 nm, respectively. A cooled CCD camera (Sensys; Photometrics, Tucson, Ariz) adapted with a bandpass filter was used for image capture.

Statistical Analysis
Data are presented as means plus or minus the standard error of the mean. Statistical analysis of in vivo tumor fluorescence was conducted by using a two-tailed paired Student t test (n = 8 tumors each). For comparison of tumor growth, a two-tailed unpaired Student t test was performed (n = 4 tumors for DU4475, n = 5 tumors for BT20). A P value of .05 or less was considered to indicate a statistically significant difference.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In Vivo Biologic Behavior of Tumor Cell Lines
The tumor cell lines showed different growth characteristics. The overall tumor take was approximately 84% for the BT20 cells and 100% for the DU4475 tumor cells. Three weeks after implantation (3 x 10⁶ cells each) the BT20 tumors had grown to a mean size of 29.5 mg ± 3.3, while the mean weight of DU4475 tumors was 593.8 mg ± 171.0 (P < .01). Animals with DU4475 tumors did not survive more than 5–6 weeks, while animals with BT20 tumors survived several months after tumor cell inoculation (data from earlier survival studies).

Molecular Characterization of Tumors
Western blot analysis showed human cathepsin-B expression in both tumor cell lines (Fig 1). However, the DU4475 cells revealed an approximately 1.7-fold higher level in culture when compared with BT20 (Fig 1). In whole tumors, cathepsin-B levels were about 1.4-fold higher compared with BT20 tumors, respectively.

View larger version (37K): [in this window] [in a new window]   Figure 1. Cathepsin-B expression revealed by Western blot analysis of cell lysates (top: gel, bottom: density profiles) is 1.7-fold higher in highly invasive DU4475 cells compared with BT20 cells (A = 1.7 x B).

Cathepsin-B expression was also confirmed by immunohistochemistry results in both tumor types (Fig 2). The majority of enzyme was localized pericellularly or was cell membrane bound in both tumor types. However, DU4475 tumors revealed a markedly stronger cathepsin-B expression compared with BT20 tumors.

View larger version (154K): [in this window] [in a new window]   Figure 2. Photomicrograph of cathepsin-B expression in DU4475 (A) and BT20 (B) tumors. (Original magnification, x40.) Inserts show controls without primary antibody. Cathepsin-B expression is revealed as a blue precipitate of the alkaline phosphatase-labeled secondary antibody. Note the strong cathepsin-B expression in the highly invasive DU4475 tumor (A) compared with the well differentiated adenocarcinoma (B).

View larger version (53K): [in this window] [in a new window]   Figure 3. NIRF imaging 24 hours after intravenous injection of the cathepsin-B-sensitive autoquenched probe in a representative animal. A, light image; B, raw NIRF image; and C, color encoded NIRF signal (arbitrary units of NIRF intensity) superimposed on light image. The highly invasive breast adenocarcinoma (DU4475) was implanted on the right of the chest and the well differentiated adenocarcinoma (BT20) on the left. Note the higher fluorescent signal depicted on the highly invasive breast lesion (B, C).

View larger version (13K): [in this window] [in a new window]   Figure 4. Bar graph shows in vivo NIRF signal in the tumors. Note the significantly higher signal intensity (SI) for the highly invasive breast cancer (DU4475) compared with the less aggressive lesion (BT20) (*, P < .01). AU = arbitrary units.

View larger version (58K): [in this window] [in a new window]   Figure 5. NIRF microscopy of unstained frozen tumor sections. (Original magnification, x40.) The cathepsin-B-generated fluorescence signal is higher in the highly invasive DU4475 tumor (A) compared with the well differentiated BT20 adenocarcinoma (B).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Imaging Protease Activity by Using Smart Optical Probes
Proteases have been shown to play a critical role in tumor formation in local and metastastic tumor infiltration (1,5,8,10,16). For breast cancer specifically, there is considerable clinical and experimental data that strongly suggest cathepsin-B expression to be associated with higher grades of invasiveness and poorer outcome of disease (3–7,9,17). Cathepsins are thought to be part of a protease cascade involving other enzymes such as matrix metalloproteases that finally results in digestion of the extracellular matrix (18). Degradation of the extracellular matrix is an integral part of tumor cell infiltration and angiogenesis (1). Targeting and imaging protease activity with activatable fluorescent probes (and potentially other probes for MR) thus offers an opportunity to visualize key molecules in carcinogenesis and metastagenesis in vivo.

Diagnostic Imaging with Light
Optical mammography has for many years been the object of intense experimental and clinical investigations. However, investigators in early studies (19–22) have failed to show sufficient sensitivity and specificity of this method, primarily because contrast was based on absorption by hemoglobin and/or scattering. In this study, we used a fundamentally different approach, the use of specific fluorochromes to generate molecular contrast. The agents are nondetectable in their native state (ie, zero background) but became highly fluorescent at target interaction (ie, the fluorescence activation in the infrared is the sole source of contrast). Tumors of submillimeter size have been depicted by using these agents and the NIRF reflectance imager used in this study (11,13).

In general, NIRF reflectance imaging, as used in this study, offers only limited depth penetration due to the highly scattering properties of the tissue (11,13). However, the main focus of this work was to demonstrate molecular specificity of the fluorescent probes as opposed to developing new optical imaging devices.

Practical application: More efficient illumination and reconstruction techniques have been used to apply tomographic principles for optical imaging (diffuse optical tomography) (23–25). Recently, the feasibility of detecting breast lesions with diffuse optical tomography has been shown in a clinical scenario (26). Combining new imaging methods such as diffuse optical tomography with smart optical probes is a next logical step for clinical imaging with light.

The mechanism of the imaging approach presented in this study can theoretically be applied to a variety of enzymes in vivo, for example, cathepsin-D and matrix metalloprotease-2 (12,27). Imaging protease activity patterns in vivo might in the near future have important implications not only for earlier tumor detection (ie, before major phenotypical changes occur) but also for therapeutic management of breast cancers. More aggressive treatment protocols, for example, could be more appropriate for tumors with higher protease levels, while tumors with less protease burden might be treated more conservatively. Moreover, protease inhibitors as a novel class of anticancer drugs are currently undergoing clinical evaluation (6,28–30). Experimental data suggest that smart optical probes may be helpful to monitor these novel therapeutics (31). In summary, the results of this study demonstrate the feasibility of imaging expression levels of a key protease related to tumor aggressiveness in vivo.

	ACKNOWLEDGMENTS

 
The authors thank Nikolai Sergeyev, PhD for technical assistance in Western blotting.

	FOOTNOTES

 
Abbreviation: NIRF = near-infrared fluorescence

Author contributions: Guarantors of integrity of entire study, all authors; study concepts and design, all authors; literature research, C.B.; experimental studies, C.B., C.H.T., A.B.; data acquisition, C.B., C.H.T.; data analysis/interpretation, C.B., R.W.; statistical analysis, C.B.; manuscript preparation, C.B., R.W.; manuscript definition of intellectual content, editing, revision/review, and final version approval, all authors.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Edwards DR, Murphy G. Cancer: proteases—invasion and more. Nature 1998; 394:527-528.
2. Frosch BA, Berquin I, Emmert-Buck MR, Moin K, Sloane BF. Molecular regulation, membrane association and secretion of tumor cathepsin B. APMIS 1999; 107:28-37.
3. Krepela E, Bartek J, Skalkova D, et al. Cytochemical and biochemical evidence of cathepsin B in malignant, transformed and normal breast epithelial cells. J Cell Sci 1987; 87:145-154.
4. Barbi GP, Margallo E, Margiocco M, et al. Evaluation of cathepsin D as prognostic predictor in breast cancer. Oncology 1994; 51:329-333.
5. Abecassis J, Collard R, Eber M, Pusel J, Fricker JP, Methlin G. Proteinases and sialyltransferase in human breast tumors. Int J Cancer 1984; 33:821-824.
6. Xing R, Wu F, Mason RW. Control of breast tumor cell growth using a targeted cysteine protease inhibitor. Cancer Res 1998; 58:904-909.
7. Maguire TM, Shering SG, Duggan CM, McDermott EW, O’Higgins NJ, Duffy MJ. High levels of cathepsin B predict poor outcome in patients with breast cancer. Int J Biol Markers 1998; 13:139-144.
8. Benitez-Bribiesca L, Martinez G, Ruiz MT, Gutierrez-Delgado F, Utrera D. Proteinase activity in invasive cancer of the breast: correlation with tumor progression. Arch Med Res 1995; 26:S163-S168.
9. Schultz DC, Bazel S, Wright LM, et al. Western blotting and enzymatic activity analysis of cathepsin D in breast tissue and sera of patients with breast cancer and benign breast disease and of normal controls. Cancer Res 1994; 54:48-54.
10. Recklies AD, Tiltman KJ, Stoker TA, Poole AR. Secretion of proteinases from malignant and nonmalignant human breast tissue. Cancer Res 1980; 40:550-556.
11. Weissleder R, Tung CH, Mahmood U, Bogdanov A, Jr. In vivo imaging of tumors with protease-activated near-infrared fluorescent probes. Nat Biotechnol 1999; 17:375-378.
12. Tung CH, Bredow S, Mahmood U, Weissleder R. Preparation of a cathepsin D sensitive near-infrared fluorescence probe for imaging. Bioconjug Chem 1999; 10:892-896.
13. Mahmood U, Tung CH, Bogdanov A, Weissleder R. Near-infrared optical imaging of protease activity for tumor detection. Radiology 1999; 213:866-870.
14. Lasfargues EY, Ozzello L. BT20. J Natl Cancer Inst 1958; 21:1131-1147.
15. Langlois AJ, Holder WD, Iglehart JD, Nelson-Rees WA, Wells SA, Bolognesi DP. Morphological and biochemical properties of a new human breast cancer cell line. Cancer Res 1979; 39:2604-2613.
16. Poole AR, Tiltman KJ, Recklies AD, Stoker TA. Differences in secretion of the proteinase cathepsin B at the edges of human breast carcinomas and fibroadenomas. Nature 1978; 273:545-547.
17. Schwartz MK. Tissue cathepsins as tumor markers. Clin Chim Acta 1995; 237:67-78.
18. Sameni M, Moin K, Sloane BF. Imaging proteolysis by living human breast cancer cells. Neoplasia 2000; 2:496-504.
19. Monsees B, Destouet JM, Totty WG. Light scanning versus mammography in breast cancer detection. Radiology 1987; 163:463-465.
20. Monsees B, Destouet JM, Gersell D. Light scan evaluation of nonpalpable breast lesions. Radiology 1987; 163:467-470.
21. Drexler B, Davis JL, Schofield G. Diaphanography in the diagnosis of breast cancer. Radiology 1985; 157:41-44.
22. Geslien GE, Fisher JR, DeLaney C. Transillumination in breast cancer detection: screening failures and potential. AJR Am J Roentgenol 1985; 144:619-622.
23. Ntziachristos V, Chance B. Probing physiology and molecular function using optical imaging: applications to breast cancer. Breast Cancer Res 2001; 3:41-46.
24. Ntziachristos V, Weissleder R. Experimental three-dimensional fluorescence reconstruction of diffuse media using a normalized Born approximation. Opt Lett; (in press).
25. Pogue BW, Poplack SP, McBride TO, et al. Quantitative hemoglobin tomography with diffuse near-infrared spectroscopy: pilot results in the breast. Radiology 2001; 218:261-266.
26. Ntziachristos V, Yodh AG, Schnall M, Chance B. Concurrent MRI and diffuse optical tomography of breast after indocyanine green enhancement. Proc Natl Acad Sci U S A 2000; 97:2767-2772.
27. Tung C, Mahmood U, Bredow S, Weissleder R. In vivo imaging of proteolytic activity using a novel molecular reporter. Cancer Res 2000; 60:4953-4958.
28. Jones JL, Walker RA. Control of matrix metalloproteinase activity in cancer. J Pathol 1997; 183:377-379.
29. Rooprai HK, McCormick D. Proteases and their inhibitors in human brain tumours: a review. Anticancer Res 1997; 17:4151-4162.
30. Brown PD. Matrix metalloproteinase inhibitors. Breast Cancer Res Treat 1998; 52:125-136.
31. Bremer C, Tung CH, Weissleder R. In vivo molecular target assessment of matrix metalloproteinase inhibition. Nat Med 2001; 7:743-748.

This article has been cited by other articles:

				L. Matuszewski, T. Persigehl, A. Wall, W. Schwindt, B. Tombach, M. Fobker, C. Poremba, W. Ebert, W. Heindel, and C. Bremer Cell Tagging with Clinically Approved Iron Oxides: Feasibility and Effect of Lipofection, Particle Size, and Surface Coating on Labeling Efficiency Radiology, April 1, 2005; 235(1): 155 - 161. [Abstract] [Full Text] [PDF]

				U. Mahmood and R. Weissleder Near-Infrared Optical Imaging of Proteases in Cancer Mol. Cancer Ther., May 1, 2003; 2(5): 489 - 496. [Abstract] [Full Text] [PDF]

				T. F. Massoud and S. S. Gambhir Molecular imaging in living subjects: seeing fundamental biological processes in a new light Genes & Dev., March 1, 2003; 17(5): 545 - 580. [Full Text] [PDF]

This Article Abstract Figures Only All Versions of this Article: 2223010812v1 222/3/814    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bremer, C. Articles by Weissleder, R. Search for Related Content PubMed Articles by Bremer, C. Articles by Weissleder, R.