WO2011053435A1 - Cell detection using targeted nanoparticles and magnetic properties thereof - Google Patents

Abstract

The present invention provides methods and apparatuses for detecting or measuring cells or substances present in biological samples The method comprises (a) providing targeted nanoparticles compnsing a magnetic nanoparticle conjugated with a targeting agent that preferentially binds to cells of a first type, (b) introducing the targeted nanoparticles to the sample in a manner that allows bonding of the targeting agents to cells of the first type, (c) subjecting the bound targeting nanoparticles to the operation of a magnetic device to extract the sample cells to which are bound targeted nanoparticles The method can also compnse (a) prepanng a second sample by combining a first sample with a plurality of targeted nanoparticles, (b) subjecting the second sample to an applied magnetic field, and (c) measunng the relaxation of the magnetic field induced in the bound nanoparticles to determine the presence, concentration, or other characteristic of cells of the predetermined type

Description

CELL DETECTION USING TARGETED NANOPARTICLES AND MAGNETIC PROPERTIES THEREOF

[0001] TECHNICAL FIELD

This invention relates to the detection and measurement of cells using targeted nanoparticles and a magnetic needle, a magnetic sensor, or a combination thereof, and is particularly useful in determining minimum residual disease in leukemia patients and in finding rare cells such as circulating tumor cells.

[0002] BACKGROUND ART

There are various applications where the detection of cells that appear in very low quantities is important. For example, determination of minimum residual disease in leukemia patients in important in managing treatment. Acute leukemia is a hematopoietic malignancy for which the accurate measurement of minimal residual disease is critical to determining prognosis and treatment. Bone marrow aspiration and light microscopy comprise the current standard of care for detecting residual disease, but these approaches cannot reliably discriminate <5% lymphoblast cells. There are about 50,000 new leukemia cases each year in the U.S. Each new case typically has at least 4 bone marrow biopsies; cases in remission typically have one bone marrow biopsy per year. After initial treatment and while in remission, the bone marrow biopsy is expected to reveal zero leukemia cells. Current techniques, however, cannot reliably measure very low concentrations of leukemia cells. Consequently, harmful chemotherapy can be continued longer than necessary, and recurrence or incomplete treatment can be missed until the leukemia has progressed farther than desired (and produced enough cells to be measured by conventional means).

[0003] As another example, detection of circulating tumor cells has been proposed as a method to detect cancer, or metastasis or recurrence of cancer. However, the concentration of such cells can be very low, and conventional sample enrichment and light microscopy analysis has not been able to reliably and cost effectively detect such cells in such very low concentrations. As one example, about 400,000 prostatectomies are performed annually in the U.S. Post-prostatectomy, there should be no prostate-specific antigen (PSA) in the blood. PSA in the blood is likely to come from prostate cancer cells that have metastasized to other parts of the body, and consequently not removed in the prostatectomy. The ability to detect very low concentrations of PSA in blood can allow for early detection of prostate cancer metastasis. Other cancers can also "shed" cells prior to or as part of metastasis; those cells can be present in blood samples but at very low concentrations. Detecting such cells at such low cell concentrations can allow for early detection of cancer presence, cancer recurrence, or cancer metastasis.

[0004] DISCLOSURE OF INVENTION This application is related to the following applications, each of which is incorporated herein by reference: U.S. 11/957,988 filed Dec. 17, 2007; U.S. 11/069,361 filed Feb. 28, 2005; U.S. 60/549,501 filed Mar. 1, 2004; U.S. 12/337,554 filed Dec. 17, 2008; U.S. 61/248,775 filed Oct. 5, 2009; U.S. 61/329,076 filed Apr. 28, 2010; 61/377,854; U.S. 61/377,854 filed Aug. 27, 2010

[0005] The present invention provides methods and apparatuses for detecting, measuring, or both, cells or substances present in even very low concentrations in biological samples. The present invention provides a method of extracting cells of a first type from a sample, comprising: (a) providing targeted nanoparticles, each comprising a magnetic nanoparticle conjugated with a targeting agent that preferentially binds to cells of the first type; (b) introducing the targeted nanoparticles to the sample in a manner that allows bonding of the targeting agents to cells of the first type; (a) subjecting the bound targeting nanoparticles to the operation of a magnetic device to extract from the sample cells to which are bound targeted nanoparticles. In some embodiments, the targeted nanoparticles comprise magnetic nanoparticles conjugated an antibodies specific to cells of the first type. In some embodiments, the magnetic device comprises an elongated member having disposed thereon a plurality of magnetic regions. In some embodiments, the magnetic device further comprises a nonmagnetic sheath removably mounted over the elongated member. In some embodiments, the elongated member comprises a stainless steel rod, and the plurality of magnetic regions comprises a plurality of permanent magnets. In some embodiments, the permanent magnets comprise NdFeB magnets. In some embodiments, the sheath comprises a polyimide sheath.

[0006] The present invention provides an apparatus for the extraction of cells of a first type from a sample, comprising a plurality of targeted nanoparticles, each comprising a magnetic nanoparticle conjugated with a targeting agent that preferentially binds to cells of the first type; a sample holder configured to contain a sample which can include cells of the first type; and a magnetic device configured to be disposed in relation to the sample such that the magnetic device acts on targeted nanoparticles that have been bound to cells of the first type in the sample holder and extracts said cells from the sample. In some embodiments, the targeted nanoparticles comprise magnetic nanoparticles conjugated an antibodies specific to cells of the first type. In some embodiments, the magnetic device comprises an elongated member having disposed thereon a plurality of magnetic regions. In some embodiments, the magnetic device further comprises a nonmagnetic sheath removably mounted over the elongated member. In some embodiments, the elongated member comprises a stainless steel rod, and the plurality of magnetic regions comprises a plurality of permanent magnets. In some embodiments, the permanent magnets comprise NdFeB magnets. In some embodiments, the sheath comprises a polyimide sheath. [0007] The present invention provides a method of detecting, measuring, or a combination thereof, cells of a predetermined type in a first sample comprising: (a) preparing a second sample by combining the first sample with a plurality of targeted nanoparticles; (b) subjecting the second sample to an applied magnetic field; (c) measuring the relaxation of the magnetic field induced in the bound nanoparticles, and from the measurement determining the presence, concentration, prevalence, or other characteristic of cells of the predetermined type. In some embodiments, subjecting the second sample to an applied magnetic field comprises subjecting the second sample to a uniform magnetic field of strength and duration sufficient to impose a uniform magnetization on the nanoparticles in the second sample that are bound to cells of the determined type. In some embodiments, the targeted nanoparticles comprise magnetic nanoparticles bound to targeting agents, and the cells of the predetermined type have binding sites or other affinities for the targeting agents. In some embodiments, the targeting molecules comprise antibodies.

[0008] The present invention provides a method of detecting the presence of cancer in a patient, comprising: (a) obtaining a first sample of fluid, serum, tissue, or other substance collected from the patient, wherein the presence of a predetermined substance in the sample indicates the presence of cancer in the patient; (b) providing targeted nanoparticles, comprising magnetic nanoparticles conjugated with targeting agents, wherein a targeting agent comprises an agent that preferentially binds with the predetermined substance; (c) providing a second sample by subjecting the first sample to the targeted nanoparticles under conditions that allow binding of the targeting agents to the predetermined substance; (d) imposing a known magnetization on the nanoparticles in the second sample; (e) measuring the relaxation of the magnetization of the nanoparticles after imposition of the known magnetization; and (f) determining the presence of the predetermined substance in the second sample from the relaxation of the magnetization. In some embodiments, the predetermined substance is cells of a specific type of cancer, and the targeting agent comprises one or more antibodies that bind to those cells. In some embodiments, the predetermined substance is PSA, and the targeting agent comprises a PSA-specific antibody. In some embodiments, the predetermined substance and targeting agent comprise one or more or the pairs set forth in the specification.

[0009] The present invention provides a method of localizing a recurrence or metastasis of cancer in a patient, comprising: (a) detecting the presence of cancer according to the methods described herein; (b) injecting the patient with targeted nanoparticles, wherein the targeted nanoparticles preferentially bind with cells indicative of the recurrence or metastasis of cancer; (c) scanning the patient with a magnetic relaxation instrument, and identifying when the measured magnetic relaxation indicates bound nanoparticles; (d) Indicating the presence of recurrence or metastasis of cancer responsive to the identification of bound nanoparticles. In some embodiments, the cells indicative of recurrence or metastasis comprise cancer cells. In some embodiments, the targeted nanoparticles comprise magnetic nanoparticles conjugated with antibodies that preferentially bind with cancer cells.

[0010] BRIEF DESCRIPTION OF DRAWINGS

The accompanying drawings, which are incorporated in and form part of the specification, illustrate the present invention and, together with the description, describe the invention. In the drawings, like elements are referred to by like numbers.

[0011] Fig. 1. (a) Modified Langevin function in universal form where y=0 is the normal Langevin. As an example, a 25 nm diameter magnetite particle in a field of 4 mT with bulk properties at 300K would be have coordinates x = 3.7 and y = 26. (b) Ratio of modified Langevin divided by normal Langevin vs. nanoparticle diameter with same properties.

[0012] Fig. 2. Normalized probability distribution function,w(D) based on TEM measurements on 2602 particles. A cubic spline interpolates between the measured points for both the susceptibility and relaxometry. A lognormal fit is also shown.

[0013] Fig. 3. Measured susceptibilities on curves of constant frequency, from top down on left: 0.1, 0.3, 1.0, 3, 10, 30, 100, 300, 1000 Hz. The measurements are 5°K apart and joined by straight lines. The ordinates are volume susceptibility in rationalized MKS. (a) Real part, (b) Imaginary part.

[0014] Fig.4. Representative samples of fitted constant temperature curves for susceptibility. The 3 parameters in the model, Ms, K and T₀ , are adjusted for each temperature to give best fit to the 9 frequency points, (a) Real part for temperatures starting right bottom up: 260 to 350 K in 10 K increments, (b) Imaginary part for 290 K, diamonds, 300 K, circles, 310 K, squares. Note that the largest fluctuations in the measured points occur at the two highest frequencies. There is relatively more noise in the imaginary than the real measurements because the signals are an order of magnitude smaller. All 30 constant T data sets were fit in this manner; only this representative sample is shown.

[0015] Fig.5. Kernel functions, from eqs. 11 & 12, at T=300K, solid, and the size distribution function, dotted, arbitrary vertical scale. The highest frequency, 1000 Hz, envelope is on the left, and the lowest, 0.1 Hz, is on the right. Similar clusters for higher (lower) temperatures correspond to larger (smaller) particle sizes, (a) Real part, (b) Imaginary part; here each kernel probes only a narrow band in the size spectrum, in contrast with the much broader sensitivity for the real parts.

[0016] Fig.6. Best fitted values of the three parameters vs temperature. In each case the values from the real dramatically cease to agree with the imaginary below 300K. This discrepancy can be minimized by adding smaller particles to the size distribution (a Gaussian cluster, "lump", centered at 20 nm, sigma, 3nm, comprising 15% of particle population), (a) Saturated magnetization, Ms. (b) Power = -logT₀ . (c) Anisotropy energy density K .

[0017] Fig. 7. Correlations, (a) K vs logT₀ . Note that the data indicate an abrupt termination of - logT₀ at 10. The anomalous tail for greater than 25 disappears by inserting the arbitrary Gaussian cluster (see Fig. 6). (b) The effective diameter, D_eii, that depends on K over log T₀ , corresponds to a time constant of 1 second.

[0018] Fig. 8. SQUID relaxometry model (dotted line) compared with measurements for the three pulse durations, (a) Relaxation (b) Excitation. All three model curves were normalized to the data with one number determined by eq. 24, using the same values for the parameters as found in the susceptibility fits, with the addition of fitted value for alpha of 0.89. The normalization number for relaxation was different from that used to find the particle number.

[0019] Fig. 9. (A) Transmission electron microscopy of Ocean nanoparticles with nominal diameters of 20, 25, 30, and 35 nanometers. The SHP-20 particles show a generally uniform intensity in TEM. In the images of the larger particles, an increasing number of particles exhibit intensity variations (light and dark bands), which we attribute to polycrystallinity. (B) Particle size distributions determined by user-guided image analysis of 2500 particles from each sample.

[0020] Fig. 10 (A) The amplitude of the out-of-phase component of the AC susceptibility χ" vs. frequency for the SHP particles. (B) The moment/kg observed by SQUID relaxometry (gray bars) correlates well with χ" at 1.0 Hz (black bars).

[0021] Fig. 11. Normalized magnetization vs. field obtained by DC susceptometry (solid symbols) compared to the Langevin function computed for 25 nm (dashed line) and 8 nm (dotted line) particles. The fits to the data (solid lines) are described in the text.

[0022] Fig. 12. High-resolution TEM of the SHP-30 particles shows that some particles are polycrystalline.

[0023] Fig. 13. Normalized magnetization vs. field data (solid symbols) obtained by relaxometry. The solid line was computed using the Langevin function, assuming 27.4 nm particles. The dashed lines were computed using the Moment Superposition Model, as described in the text.

[0024] Fig. 14. Volume magnetization vs. magnetic field for the SHP-25 and SHP-30 particles. The magnetic susceptibilities (slope of Mv vs. H at small H) differ by a factor of 5.

[0025] Fig. 15. (A) Observed magnetic moment as a function of time for Ocean-SHP 30

nanoparticles, conjugated to CD3 antibody and incubated with either normal Jurkat cells (blue diamonds) or Jurkat cells treated with trypsin (red squares) to remove cell surface antigens. Error bars represent the standard error of the mean (N=3). (B) In photomicrographs of slides prepared after the 60 min. incubation, the Jurkat cells are stained pink, while the nanoparticles are stained blue. Untreated cells (left) express high levels of the CD3 receptor and show greater nanoparticle binding, while the trypsin-treated cells (right) are stripped of cell surface antigens and show significantly reduced binding. The observed SQUID relaxometry signal appears to be attributable to binding of nanoparticles to cells and nanoparticle aggregation.

[0026] Fig. 16 is an illustration of magnetic relaxometry for in-vivo detection of disease.

[0027] Fig. 17 is an illustration of an example Superconducting Quantum Interference Device (SQUID) sensor system for relaxometry.

[0028] Fig. 18 is an illustration of results with particular cells.

[0029] Fig. 19 is an illustration of biomagnetic sensitivity compared to other methods. Fig. 20 is an illustration of results with animal models.

[0030] Fig. 21 is an illustration of the application to a magnetic biopsy needle.

[0031] Fig. 22 is an illustration of issues associated with production and characterization of nanoparticles.

[0032] Fig. 23 is an illustration of susceptometry measurements of nanoparticle properties.

[0033] Fig. 24 is an illustration f properties before and after addition of PEG.

[0034] Fig. 25 is a depiction of an analysis of nanoparticle binding using light microscopy and SQUID. Human leukemic cell lines U937, GA-10, and Jurkat. Photomicrographs show 20x imaging of Prussian blue staining for the presence of iron-oxide nanoparticles on cells incubated with either CD34 antibody-conjugated or unconjugated nanoparticles. Scale bar, 20 μιη (A). Nanoparticle (NP) binding was confirmed by SQUID magnetometry (B). Nanoparticle conjugated with CD34 (+) or unconjugated nanoparticles (-) are indicated. Experiments were performed thrice; columns, mean; bars, SD.

[0035] Fig. 26 is an illustration of SQUID detection of specific Nanoparticle Binding. U937 cells (1.0 x 107) were incubated with varying numbers of nanoparticles (A and B) or varying number of cells were incubated with nanoparticles (C and D). The curves shown are nonlinear fits of the data. Data represents results from three separate experiments.

[0036] Fig. 27 is a schematic representation of magnetic needle and sampling of spiked sample. Shown is the schematic of the magnetic needle covered with a polyimide sheath prepared for insertion into a bone marrow sample (A), as well as the sheath-covered needle itself (C) and the magnetic needle inserted into an experimental sample (B). Sample tubes are placed into a sample holder on a platform under the SQUID sensors (D).

[0037] Fig. 28 is an illustration of the percentage of lymphoblasts in pre-needle and needle harvest samples of three leukemia patients. The patient samples were diluted 1:1024 in donor blood, and the lymphoblast percentage was calculated both before and after needle harvest. Since neutrophils were present primarily due to dilution of sample with blood so that the percentage of blasts was low, the percentage of lymphoblasts was calculated with neutrophils included in the WBC count (A) and excluded from the WBC count (B).

[0038] Fig. 29 is a schematic illustration of the production and operation of targeting agents according to the present invention.

[0039] Fig. 30 is a schematic illustration of a sample 85 such as that in Fig. 29 after introduction of a magnetic device 86.

[0040] Fig. 31 is a schematic illustration of an example preparation of a sample for measurement according to the present invention.

[0041] Fig. 32(a,b,c,d) provide a schematic illustration of an example measurement in accord with the present invention.

[0042] Fig. 33 is a schematic illustration of measurements from the process described in connection with Fig. 32.

[0043] Fig. 34 is a schematic illustration of an apparatus suitable for use in the present invention.

[0044] MODES FOR CARRYING OUT THE INVENTION AND INDUSTRIAL APPLICABILITY

The present invention provides methods and apparatuses using magnetic nanoparticles that have been conjugated with targeting agents so that they are preferentially bound to cells of one or predetermined types. After the targeted nanoparticles have bound to such cells, those cells can be separated from the rest of a sample with a magnetic device. Also, the magnetic properties of the targeted nanoparticles can be used to measure the number of target cells in the sample using magnetic relaxation, as an alternative to or in connection with separation using a magnetic device.

[0045] Fig. 29 is a schematic illustration of the production and operation of targeting agents according to the present invention. The illustrations in the figure are highly simplified and intended for ease of explanation only, and are not intended to represent the actual shapes, sizes, proportions, or complexities of the actual materials involved. A sample 81, e.g., a cell culture, a blood or tissue sample, or in vivo blood or tissue, comprises some cells of the type of interest (shown in the figure as circles with "V" shaped structures around the periphery) and some cells of other types (shown in the figure as ovals with rectangular structures around the periphery). A plurality of magnetic nanoparticles 82 is provided, shown in the figure as small circles. A plurality of targeting agents 83 is also provided, shown in the figure as small triangles. The nanoparticles and targeting agent are combined (or conjugated), forming targeted nanoparticles 84.

[0046] The targeted nanoparticles can then be introduced into the sample 85. Cells of the type of interest have binding sites or other affinities for the targeting molecule, illustrated in the figure by "V" shaped structures around the periphery of such cells. The number of binding sites can vary, but some types of cancer cells have been found to have many thousands of binding sites for specific antibodies, so the 8 sites shown on each cell in the figure is a very significant underrepresentation of the density of nanoparticles that can attach to such cells. The targeting agents attach to the cells of the type of interest, illustrated in the figure by the triangular targeting molecules situated within the "V" shaped structures. Generally, each cell will have a large number of such binding or affinity sites. Cells of other types do not have such binding sites or affinities, illustrated in the figure by ovals with no targeted nanoparticles attached. Targeted nanoparticles that do not bind to cells are left free in the body, illustrated in the figure by small circles with attached triangles that are not connected with any specific cell (both inside and outside the organ or other structure represented by the irregular curve).

[0047] Fig. 30 is a schematic illustration of a sample 85 such as that in Fig. 29 after introduction of a magnetic device 86. The magnetic device 86 has attracted the nanoparticles in the sample 85, with the strongest attraction for those cells 88 that have many nanoparticles attached by the targeting agent. Cells without nanoparticles 87, i.e., those that did not bind to the targeting agent, are not attracted to the magnetic device 86. The magnetic device 86 can be removed from the sample 85 and examined to determine whether any cells of the targeted type are present. Since substantially all cells of the targeted type present in the sample will have nanoparticles bound to them by the targeting agent, if there are any cells of the targeted type present in the sample originally there will be cells of the targeted type on the magnetic device. The figure shows roughly equal number of targeted cells and non-targeted cells; in practice, there can be many times as many non-targeted cells as targeted cells, and the operation of the targeted nanoparticles and magnetic device will serve to pull only targeted cells from the sample. Accordingly, the presence of targeted cells in the sample can be determined even in the presence of very low targeted cell concentrations.

[0048] The ability to find even very low concentrations of targeted cells can be significant in many applications. As an example, some cancers "shed" cells that are then present in minute

concentrations in the blood or other tissue. While the blood is easy to sample, finding such rare cells is nearly impossible with current techniques. The ability of the present invention to find such cells can enable much earlier detection and treatment of cancer, cancer metastasis, and other diseases. As another example, determination of minimum residual disease in leukemia is important in managing and assessing treatment. Current techniques, however, rely on optical microscopy examination of small numbers of cells taken at random from a bone marrow biopsy sample.

Leukemia cells can be present in the sample, but if they are not part of the small sample taken for microscopic examination, the determination will be "no disease", when actually the disease is still present. An embodiment of the present invention has been tested in connection with leukemia cell detection in bone marrow biopsy samples.

[0049] EXAMPLE MAGNETIC DEVICE EMBODIMENT AND APPLICATION. Acute leukemias are bone marrow-derived malignancies for which the development of sensitive detection methods is crucial to improving clinical detection and outcomes. Currently, technologies used to examine bone marrow samples may fail to detect the presence of leukemia cells below 1% to 5% of total leukocytes, i.e., minimal residual disease. A major difficulty in detecting minimal residual disease using bone marrow aspiration is that random sampling can neglect areas of focal disease. As a result, opportunities opportunities to intensify therapy may be overlooked, leading to relapsed disease. In these cases, the ability to reliably detect residual leukemia cells, when present below 5%, to monitor the efficacy of therapy is critical to improving care.

[0050] Magnetic nanoparticles have become an increasingly important tool in both targeting and detecting cancer cells. Cellular targeting may be achieved through attachment of receptor-specific ligands, including antibodies. As a result, cell surface receptors expressed on tumor cells, such as CD34 on acute leukemia cells, could allow for targeting nanoparticles labeled with antibodies. In addition, receptor-specific binding of nanoparticles leads to internalization in vitro and in vivo, thus increasing the potential number of nanoparticles associated with each cell target. By using superparamagnetic nanoparticles composed of iron oxide (SPIONs), conjugated to anti-CD34 antibodies, we hypothesized that we could create magnetically charged leukemia cells that could be preferentially collected using a magnetic source during standard bone marrow sampling procedures.

[0051] Once magnetically charged leukemia cells are collected, nanoparticle binding and lymphoblast collection efficiency of the magnetic needle needed to be assessed. In addition to using standard techniques, such as light microscopy, we used a highly sensitive magnetometer called a superconducting quantum interference device (SQUID) to allow assessment of very small numbers of nanoparticle-coated cells. SQUID magnetometry has been used for clinically detecting magnetic fields under a variety of conditions because of its acute sensitivity. One such method uses a SQUID biosusceptometer, which can detect small aberrations in iron seen in iron-based pathologies such as hemochromatosis and thalassemia- induced iron storage disease. Our method uses

magnetorelaxometry, whereby nanoparticles are briefly magnetized by a pulsed field, and the SQUIDs detect the nanoparticle magnetization as it relaxes back to equilibrium. Pertinent to our study, SPIONs have three specific properties that make them highly compatible for SQUID relaxometry detection: (a) they are superparamagnetic, (b) the individual magnetic moments of these particles align with a magnetic field, so that cells labeled with sufficient numbers of bound single particles with magnetic moments of ~4 x 10-18 A-m2 are detectable by SQUIDs, and (c) unbound single particles, even when present in large numbers, do not generate detectable SQUID signals. Magnetic moments measured by SQUID relaxometry provide additional information regarding cellular binding and a secondary confirmation of microscopy results from magnetic needle collections.

[0052] The present invention provides for the enhancement of leukemia cell sampling using a magnetic device and nanoparticles. The sensitivity and ability of the SQUID to quantify cell sampling is also described. The invention can provide enhanced technologies for marrow sampling and for extraction of leukemia cells from bone marrow biopsy samples, which can improve clinical decision making and patient outcomes.

[0053] Materials and Methods. The following materials and methods were demonstrated in an connection with example embodiment of the invention.

[0054] Cell culture. U937, Jurkat, and GA-10 cells were purchased commercially from American Type Culture Collection and cultured in PMI supplemented with 10% fetal bovine serum (FBS; v/v; HyClone), 1% penicillin streptomycin (v/v; Life Technologies-Bethesda Research Laboratories), and 4 μg/mL ciprofloxacin (Bayer). Cells were cultured in an incubator at 37°C with 5% C02 and maintained at a cell concentration between 1 x 10^s and 1 x 10⁶ viable cells/mL. U937, GA-10, and Jurkat represent myeloid, B-cell, and T-cell lineage leukemia cell lines. Each cell line expresses CD34.

[0055] Peripheral blood and bone marrow collection. Peripheral whole blood was obtained from donors through venous puncture and was anticoagulated in 10 U/mL of heparin (Becton Dickinson). Bone marrow aspirations were performed in patients with acute leukemia who required a bone marrow evaluation as a part of their routine clinical care. Human subjects provided consent in accordance with local and federal guidelines. The patients were placed in the supine position, and the sacral area was draped in a sterile fashion. Local anesthesia was achieved with 1% Xylocaine (Abraxis Pharmaceutical) administered s.c, and periosteally. A Jamshidi needle (Baxter Healthcare Corporation) was inserted into the cortex of the posterior superior iliac spine.

[0056] Determination of cell surface antigen expression on live cells. A Quantum Simply Cellular kit (Bangs) was used for quantitation of cellular antigen expression in antibody binding capacity units as per manufacturer's instructions and described briefly here. The Quantum Simply Cellular bead populations provided a means for constructing a QuickCal calibration curve (antibody binding capacity values versus fluorescence intensity). Cells were compared with antibody-labeled beads, and cell surface antigen expression was quantified in antibody binding capacity units. Approximately 1 x 10^s cells and Quantum Simply Cellular beads were incubated with mouse FITC-labeled anti- human anti-CD34 (Caltag). Labeled cells were analyzed for cellular antigen expression using FACScan (Becton Dickinson) flow cytometry as described previously. [0057] Production of ligand-bearing nanopartlcles. The nanopartlcles were provided with an available surface carboxyl group to which an amino group on the antibody was attached using the carbodiimide method modified from a commercial protocol (Ocean Nanotec7) through

experimentation and is described briefly here. Ten milligrams of SiMAG-TCL 100-nm nanopartlcles (Chemicell) were aliquoted into a 15-mL conical tube (Greiner Bio-One) and brought to a total volume of 10 mL with double distilled water. N-hydroxysulfosuccinimide (Piercel) and l-ethyl-3-[3- dimethylaminopropyl] carbodiimide hydrochloride (EDC; Pierce) were prepared fresh at a concentration of 25 mg/mL each in separate tubes with double distilled water. One hundred microliters each of the EDC and N-hydroxysulfosuccinimide were added to the nanopartlcles and incubated at room temperature on a LabQuake shaker (Lablndustries, Inc.) for 20 min. Nanopartlcles were brought to pH 8.0 with 50 mmol/L NaHC03 (Sigma-Aldrich), 50 μg of anti-CD34 antibody (BD Biosciences) was added, and the mixture was incubated at room temperature on a Lab- Quake shaker for 2 h. The antibody-nanoparticle mixture was centrifuged at 7,500 relative centrifugal force for 30 min at 4°C. The supernatant was removed and 10 mL of double distilled water was added to the pelleted nanopartlcles. The centrifugation parameters were repeated once more and the supernatant was removed. The remaining pellet was resuspended in a total volume of 240 μΐ PBS (Life Technologies-Bethesda Research Laboratories)/0.5% FBS (HyClone). CD34-conjugated nanopartlcles were stored at 4°C before use.

[0058] Cell labeling and sampling. Cultured U937, Jurkat, or GA-10 cells were harvested and washed using sterile PBS. Harvested cells were counted using a hemocytometer (Hausser Scientific). Each sample, unless otherwise specified, contained 1 x 10⁷ cells suspended in 200 μΐ of cold PBS/0.5% FBS solution and 20 μΐ of CD34-coupled SiMAG-TCL nanopartlcles, at a concentration of 41.67 μg [solids]_/^L, corresponding to 22.8 μg [FeJ/μί. Cells and CD34 nanopartlcles were incubated on ice for 1 h. The magnetic needle used for cell collection is similar in size to a standard bone marrow biopsy needle. However, the magnetic needle is composed of a 17-cm-long, 1-mm-diameter stainless steel rod, at the end of which are two cylindrical NdFeB magnets 2 mm in length and separated by a stainless steel spacer 2 mm in length. The components of the magnetic needle are contained within a tight-fitting polyimide tube and the needle assembly is then inserted into a removable polyimide sheath. Magnetically labeled cells are attracted to the both poles on both magnets and stick to the outside of the sheath. After extracting magnetic material from the marrow, the sheath is removed from the needle and inserted into media where the cells are liberated from the sheath. Additional physical characterizations of the magnetic needle used in this study have been described previously by Bryant and colleagues and more recently by Adolphi and colleagues. [0059] For leukemia cell dilution experiments, U937 cells were each diluted in donor whole blood. Cells were incubated with CD34-conjugated nanoparticles for 60 min at 4°C on a nutator (Becton Dickinson). SQUID measurements were taken after incubation and pre-needle draw, post-needle draw, and of the needle draw sample. The needle draws were performed following the nanoparticle incubation by placing the magnetic portion of the needle in the center of the samples in 1.5-mL microcentrifuge tubes for 1 min.

[0060] Glass slide preparations were made using a Cyto-centrifuge (Shandon). Slides were stained with either Diff-Quik stain (Dade Behring), which is similar to a Wright-Giemsa stain, or Prussian blue stain, which reveals the presence of iron. Prussian blue staining was performed by TriCore Reference Laboratories. Stained samples were qualitatively assessed for nanoparticle attachment using light microscopy. Light microscopy was performed on an Axiovert 200 MAT (Zeiss) and captured images were taken with Moticam 2300 using provided Motic Images Plus software (Motic). Percent lymphoblasts in bone marrow diluted in blood were assessed microscopically by the number of lymphoblast cells per 200 leukocytes. The collection enhancement factor was determined by comparison of the lymphoblast percent in the needle and pre-needle draw samples.

[0061] Complete leukocyte counts of bone marrow, donor blood, post-needle draw, and needle draw samples were performed by lysis of the erythrocytes using ammonium chloride followed by collection of the leukocytes by centrifugation. Leukocytes were counted using a Beckman Coulter Counter. Absolute lymphoblast counts were then calculated from the total leukocyte counts using the percent lymphoblast counts determined microscopically.

[0062] Cell detection using SQUID relaxometry. Cell nanoparticle attachment was quantified using SQUID relaxometry, which involves briefly magnetizing the nanoparticles using a pulsed magnetic field and then detecting the decaying magnetization of bound particles over time using the SQUID sensors. In the case of cell-bound nanoparticles (which are not free to rotate), the decay is due to the Neel mechanism (20), involving internal reorientations of the nanoparticle magnetic moment and gives a detectable signal in the time window required by SQUID relaxometry for nanoparticles with the appropriate core diameter (~24 ± 4 nm iron oxide). The magnetization of unbound nanoparticles in fluid decays by Brownian rotation of the particle, and this relaxation is generally too fast to be detected by our technique. Samples in 1.5 mL microcentrifuge tubes were centered under the seven channel SQUID sensor array (BTi 2004, 4D-Neuroimaging) and placed at a distance of 3.4 cm below the center sensor. A uniform magnetic field of 38 Gauss was produced at the location of the sample for 0.3 s using a square Helmholtz array. After a 50-ms delay immediately following the magnetic field pulse, the decaying magnetization at each sensor was then sampled at a rate of 1 kHz and digitized using a NI-PXI8336 16-channel digitizer and LabVIEW 8.5.1 acquisition software (National Instruments). The pulsing sequence was repeated 10 times and the results averaged to increase the signal to noise ratio.

[0063] SQUID analysis. Analysis of the SQUID data was performed using the Multi-Source Analysis program, written in our laboratory using MATLAB (The MathWorks, Inc.). The SQUID system consists of seven sensors that measure the distribution of field lines from sources placed below. The field magnitudes from multiple sensors give the ability to localize a source and measure its magnitude moment. To determine the magnetic field amplitude immediately after the magnetizing pulse at each of the seven sensors, curve fitting was performed to extrapolate the relaxation curves back to the time at which the magnetizing pulse was switched off. The location (x, y, z) and magnetic moment (mz) of the sample were then determined by fitting the spatial dependence of the magnetic field amplitudes with a single magnetic dipole model.

[0064] Analysis of unlabeled and antibody-conjugated nanoparticle binding to human cell lines in vitro. To evaluate ligand-specific nanoparticle binding behavior, three leukemic cell lines were identified with varied expression of CD34. The number of CD34 receptors on the cell surface was quantified using a bead-based technique flow cytometry. This flow cytometry method is extremely sensitive and a leukemia cell line expressing no CD34, and therefore truly CD34 "negative", was not experimentally observed by us since very low numbers of receptors can be detected (21, 22). U937 which expressed 4.6 x 10⁴ sites, and two relatively low CD34 expressing cell lines, GA-10 and Jurkat, expressing 6.2 x 10³ and 2.3 x 10³ sites, respectively, were incubated with anti-CD34 antibody- tagged nanoparticles and unlabeled nanoparticles. Following incubation, cytospin slides were prepared and stained for cells and for the presence of iron-containing nanoparticles. The presence of nanoparticles bound to the cell was assessed microscopically. By light microscopy, qualitative differences in the number of anti-CD34 antibody- tagged nanoparticles bound to the three cell lines was observed; U937, the high CD34-expressing cell line (Fig. 25A), shows a high level of CD34 antibody-conjugated nanoparticle association in the cytospin preparations. Lower levels of cell associationwith CD34 antibody-conjugated nanoparticles were found in the low CD34-expressing cell lines GA-10 and Jurkat (Fig. 25A). Very little nanoparticle association was found between any of the cell lines and the unconjugated nanoparticles (Fig. 25A). To confirm the light microscopy results, cells were assessed for nanoparticle binding using SQUID magnetometry, which showed increasingly higher magnetic moment in samples with higher CD34 expression (Fig. 25B). These data suggest increased nanoparticle binding as the number of CD34 receptors increases on the cell surface and indicate that the antibody attached to the nanoparticles has sufficient orientation to provide functionality. [0065] Quantifying binding specificity using SQUID. To understand the capability of SQUID to quantitate nanoparticle binding, we performed two sets of experiments. First, we measured the SQUID signal obtained by adding varying amounts of CD34-labeled nanoparticles to U937 cells. When using 1.0 x 10⁷ U937 cells, -9.2 x 10 CD34 receptors were available in the sample. An increasing SQUID signal was observed up to ~7.1 x 10 nanoparticles, which then plateaued, indicating that additional sites for nanoparticle binding were not available once 7 x 10 sites were bound, although up to 3.5 x 10¹² nanoparticles were added (Fig. 26A and B). Interestingly, unconjugated nanoparticles also produced a detectable signal, indicating that low amounts of cellular nanoparticle uptake can be detected by SQUID, but is not easily visualized by light microscopy and special staining, or that the nanoparticles are agglomerated so they intrinsically produce a small magnetic moment.

[0066] In the second set of experiments, we fixed the number of nanoparticles and varied the number of CD34-expressing cells. Using 7.1 x 10 nanoparticles, a peak specific signal was seen with 1.0 x 10⁷ cells that then plateaued (Fig. 26C and D). Again, background uptake of unconjugated nanoparticles was also detected. Taken together, the findings of these two experiments are consistent and indicate that ~7.1 x 10⁴ nanoparticles bound per cell, although 9.2 x 10⁴ CD34 receptors are present, suggesting some level of steric hindrance. Nonetheless, SQUID relaxometry provided a highly sensitive method to quantitatively measure the specific binding of anti-CD34 antibody-bound SPIONs to leukemia cells.

[0067] Isolation of CD34-positive cells from peripheral blood using anti-CD34-conjugated nanoparticles and the magnetic needle. Specific binding of anti-CD34 SPIONs to the CD34-positive cells suggests that it may be possible to specifically sample these cells from blood or bone marrow. To achieve this, a magnetic needle based on a standard bone marrow sampling needle was developed (Fig. 27). The magnetic needle consists of very strong rare earth magnets attached at the bottom of a stainless steel rod (Fig. 27A). To allow collection of cells from the needle after sampling, the needle is covered by a removable polyimide sheath (Fig. 27C). Once covered by the sheath, the needle is lowered into a bone marrow or blood sample for lminute to collect nanoparticle-bound cells (Fig. 27B). After sampling, the needle sheath is removed and placed in media to allow removal of collected cells from the sheath and assessment of these cells by preparation of cytospin slides. To confirm binding, samples were placed under the SQUID sensor system and assessed by SQUID magnetometry (Fig. 27D).

[0068] To assess the capacity to isolate CD34-positive cells using anti- CD34-labeled nanoparticles and the magnetic needle, we "spiked" human blood with a leukemia cell line, U937 (Table 4). U937 cells were serially diluted in whole blood from a healthy volunteer, incubated with ligand-bearing nanoparticles, and collected using the magnetic needle. The presence of nanoparticle-bound cells in the magnetic needle samples was verified using Diff-Quik and Prussian blue stained cytology. Notably, the collection enhancement of CD34-expressing cells increased at lower absolute numbers and percentages. When these slides were examined by light microscopy, only U937 cells and neutrophils were seen. Other types of leukocytes were not present. Prussian blue staining indicated that the neutrophils in the peripheral blood had phagocytosed nanoparticles; otherwise, the enhanced collection of U937 cells would have been markedly increased.

[0069] Table 4

[00114] Sampling of bone marrow in patients with CD34-positive leukemia. Next, we assessed our ability to harvest CD34-expressing leukemia cells from patient samples. Bone marrow aspirate from three patients with known acute leukemia were serially diluted in donor whole blood and the lymphoblasts were extracted using the magnetic needle. There was a significant increase in the number of lymphoblasts harvested from each patient bone marrow sample and each dilution (Table 5; Fig. 28). Examination of the percent lymphoblasts present in the pre-needle draw samples to the post-needle draw samples shows a consistent increase in the percent lymphoblasts sample harvested with the magnetic needle. The increase in percent lymphoblast was greatest in the more dilute samples, similar to the previous experiments with a leukemia cell line and showing the utility of the magnetic needle for isolating the rare lymphoblasts.

[00115] Table 5 bone

marrow percentage of percentage of dilution in pre-needle percentage of lymphoblasts after lymphoblasts without blood lymphoblasts harvest neutrophils after harvest

1 2 3 1 2 3 1 2 3

1:64 32.0% 6.0% 3.5% 49.5% 15.5% 11.0% 100.0% 100.0% 99.0%

1:128 27.0% 3.5% 2.5% 41.0% 12.5% 8.5% 99.0% 100.0% 100.0%

1:256 19.5% 3.5% 1.5% 39.5% 8.5% 7.5% 100.0% 100.0% 100.0%

1:512 13.0% 1.0% 1.0% 34.5% 7.5% 7.0% 100.0% 99.0% 100.0%

1:1024 5.5% 1.0% 0.0% 29.5% 5.5% 6.5% 100.0% 99.0% 98.0%

1:2058 0.5% 0.0% 4.5% 5.0% 100.0% 100.0% 100.0%

[00116] Once again, only the neutrophils from the blood used for dilution were seen in addition to the lymphoblasts. Neutrophils were only present in these samples due to the use of donor blood to dilute the bone marrow and should not present a significant factor in collection of lymphoblasts pure bone marrow specimens. To address the collection of blasts in the absence of blood dilution, a second lymphoblast percentage without the neutrophil component was calculated. As neutrophils have a very different morphology than lymphoblasts, neutrophils were easily indentified visually and excluded when examining needle collection slides microscopically (Fig. 28). Excluding the neutrophils, virtually 100% of the cells identified were lymphoblasts. Since neutrophils are primarily present in these samples because peripheral blood was used to dilute the number of blasts to determine the detection limits of this technique, we would not anticipate the same number of neutrophils in undiluted specimens.

[00117] The present example provides improvements to cellular targeting using superparamagnetic particles that provide new opportunities for its use as a clinical tool. Within the area of

nanoparticles, several techniques, including antibody-bound, small-molecule modification, and aptamer-based technologies have been studied. Since the accurate detection of residual disease has previously been shown to lead to response-based improvements in leukemia-free survival, we investigated the binding of ligand-bearing nanoparticles to CD34+ cells to enhance collection of lymphoblast cells using a magnetic needle as an initial step toward the development of a tool to increase the sensitivity of detection of residual disease in acute leukemia patients. We found the binding of nanoparticles to CD34+ cell lines was dependant on cell line receptor expression and the presence CD34 antibody ligand on the nanoparticle surface. We also showed the capacity for ligand- bearing particles to bind to CD34+ cell lines and patient leukemic cells and for these cells to be collected using a magnetic needle at levels that improve upon standard bone marrow aspirate technique. Finally, these results suggest that ligand-bearing nanoparticles in combination with magnetic needle extraction present a novel option for oncologists to use targeted leukemic sampling to improve detection of low levels of disease.

[00118] In our studies, attachment of nanoparticles to cells was ligand and receptor expression dependant whether assessed qualitatively, by light microscopy, or quantitatively, by SQUID magnetometry. Increased nanoparticle attachment was observed on cells that expressed higher levels of CD34 when compared with cell lines that had lower expression. Interestingly, our titration studies indicated that there was a limit to the number of nanoparticles that bound to a cell.

Nanoparticles (7.1 x 10⁴) bound to each cell, although ~9.2 x 10⁴ receptors were present. The limit of nanoparticle binding to a cell may be a result of steric considerations, either based on cell size or access to the receptor, or incomplete coupling of the nanoparticles with antibodies directed against CD34. Since U937 has a reported diameter of between 10 and 20 μιη, and the nanoparticles used in this study have a diameter of 140 nm, and a rough calculation indicates that ~1.6 x 10⁴ to 6.5 x 10⁴ nanoparticles can bind to the surface of a cell. Our observed limit seems to be in close agreement with this theoretical calculation.

[00119] Magnetic needle extraction enhanced the number of CD34- expressing cells visible by microscopy in both spiked U937 cells as well as freshly collected patient bone marrow samples diluted into blood. The enhancement increased as the percentage of CD34-expressing cells was lowered in the sample. Notably, because we were required to dilute samples, we were required to use human peripheral blood, which, unlike bone marrow, had a high percentage of neutrophils (50- 70%). Our enhancement was observed in spite of the nonspecific uptake of nanoparticles by neutrophils and their resultant harvesting. As we move to a formal clinical trial, we anticipate testing bone marrow samples with a low percentage of leukemic blasts and few neutrophils, as the rule rather than the exception, because neutrophils are not as prevalent in marrow as blood.

Accordingly, the future clinical use of this device may show even higher collection enhancement than we observed. Nonetheless, the number of lymphoblasts collected on the needle remained fairly constant regardless of dilution factor, and at the lowest dilutions, the number of lymphoblasts in the pre-needle sample approaches the number of lymphoblasts recovered on the needle (data not shown).

[00120] Nanoparticles, in the absence or presence of cells but without a targeting ligand, also showed a measurable background magnetic relaxometry signal. This background is derived from two sources: nanoparticle agglomeration visible by light microscopy in sample without cells as well as nonspecific uptake and binding of naked nanoparticles to cells. Agglomeration of the nanoparticles causes a signal similar to cell-nanoparticle binding and was visible by SQUID relaxation

magnetometry. Agglomeration has not been described previous studies using nanoparticles as agents in magnetic resonance imaging studies. However, our experiments indicate that low levels of agglomeration is present in virtually all nanoparticle preparations and is reflective of the high sensitivity of the SQUID magnetometry relative to other methods. Additionally, nanoparticle aggregation was easily distinguishable from cell-nanoparticle binding by microscopy. Nonspecific cellular attachment and uptake of nanoparticles may also occur but was not easily visualized by light microscopy. The nonspecific uptake of nanoparticles has been reported by previous investigators, which showed that cells lines preferentially take up nanoparticles regardless of ligand. Nonetheless, our background SQUID signals were lower than the specific nanoparticle-cell signal and could easily be subtracted.

[00121] The magnetic needle, when combined with microscopy, proved to enhance the collection and identification of CD34-expressing cells. The present invention can provide for ex vivo extraction of leukemia cells from patient bone marrow samples. In addition, this sampling modality has the capacity to identify minimal residual disease earlier, which may improve survival and reduce therapy-related patient toxicity. An advantage of the nanoparticle used in this study is its potential for in vivo use for the detection and possible treatment of leukemias. The present invention can also provide an ability to directly inject the antibody-labeled nanoparticles into the bone marrow and then specifically harvest or kill CD34-expressing cells.

[001221 EXAMPLE MAGNETIC ELAXOMET Y EMBODIMENT AND APPLICATION. Special magnetic nanoparticles that are targeted towards specific cells can be used to sensitively detect metastatic cells in serum by adding the nanoparticles to the extracted fluid and placing the combined fluid under very sensitive magnetic sensors. Only cells that are specific to the antibodies or other markers on the magnetic nanoparticles will be detected by the sensors using a magnetic relaxometry method. Detection of the presence of these cells in quantities smaller than existing methods is a desirable goal. For example, in the case of prostate cancer metastasis that might have occurred before prostate removal, a level of 0.2 ng/ml of PSA in the blood would indicate a tumor containing ten to one hundred million cells which would be shedding some cells into the blood. A more sensitive determination of metastatic cells is important to determine that treatment for this reoccurrence of cancer should begin at an earlier stage. Indications of increased numbers of markers or detection of metastatic cells in the blood can be followed by searches throughout the body for the tumors using magnetic relaxometry sensor arrays that are targeted towards the type of cancer identified by the markers or detected cells in the serum. These sensor arrays use the same type of ultra-sensitive sensors, such as SQUIDs, that are used to detect the cells in the blood but are arranged in an array of sensors to permit localization of the metastatic tumor. The patient can be placed under this array which then scans the body, after injection of magnetic nanoparticles with antibodies specific to the markers or cell cancer type identified in the serum assay. This method increases the specificity for finding cancer while also increasing the detection sensitivity since the cancer type has been identified.

[00123] This method of metastatic cell detection using magnetic relaxometry or sensitive marker detection can be followed by localization of the cancer site responsible for shedding these cells into the serum and can be used following any type of cancer treatment whether it is surgery, chemotherapy or ablation. Detection of presence of these cells or markers can then indicate further therapy is needed such as continuation of the chemotherapy or radiation or targeted therapy when the source of the cells or markers is located.

[00124] Detection of cells in the body's fluids can also be used to indicate that further imaging for cancer is desirable. Magnetic relaxometry is a very sensitive method for determining the presence of cancer clusters resulting from metastasis and can be used to examine the entire body for growths of cancer. In this method, the treatment can then be focused on the specific areas where cancer cell clusters have been detected. This can result in a considerable decrease in side effects as compared with a general therapy applied to the entire individual. Therapy can also be more effective when there is minimal spread of the disease, made possible by detection before significant spread. The result is more effective treatment, a general cost savings in medical care, and increased quality of life because the body's normal cells are not subjected to therapeutical agents that harm normal cells.

[00125] Fig. 31 is a schematic illustration of an example preparation of a sample for measurement according to the present invention. The illustrations in the figure are highly simplified and intended for ease of explanation only, and are not intended to represent the actual shapes, sizes, proportions, or complexities of the actual materials involved. A sample 11, e.g., blood or serum or tissue, comprises some cells of the type of interest (shown in the figure as circles with "V" shaped structures around the periphery) and some cells of other types (shown in the figure as ovals with rectangular structures around the periphery). A plurality of magnetic nanoparticles 12 is provided, shown in the figure as small circles. A plurality of targeting molecules 13 is also provided, shown in the figure as small triangles. The nanoparticles and targeting molecules are combined (or conjugated), forming targeted nanoparticles 14.

[00126] The targeted nanoparticles can then be combined with the sample 15. Cells of the type of interest have binding sites or other affinities for the targeting molecule, illustrated in the figure by "V" shaped structures around the periphery of such cells. The targeting molecules attach to the cells of the type of interest, illustrated in the figure by the triangular targeting molecules situated within the "V" shaped structures. Generally, each cell will have a large number of such binding or affinity sites. Cells of other types do not have such binding sites or affinities, illustrated in the figure by ovals with no targeted nanoparticles attached. Targeted nanoparticles that do not bind to cells are left free in the prepared sample, illustrated in the figure by small circles with attached triangles that are not connected with any specific cell.

[00127] Fig. 32(a,b,c,d) provide a schematic illustration of an example measurement in accord with the present invention. In Fig. 32a, the prepared sample is as in Fig. 31, with the addition of arrows near each nanoparticle. The arrows are representative of the magnetization of each nanoparticle, and indicate that the magnetization of the nanoparticles in the prepared sample is random (in the figure, the arrows are shown in one of four directions for ease of illustration only; in practice the magnetization can have any direction).

[00128] In Fig. 32b, an external magnetic field (represented by the outlined arrow at the lower right of the figure) is applied. The magnetization of the nanoparticles in response to the applied magnetic field is now uniform, represented in the figure by all the magnetization arrows pointing in the same direction.

[00129] Fig. 32c illustrates the sample a short time after the magnetic field is removed. The nanoparticles not bound to cells are free to move by Brownian motion, and their magnetization rapidly returns to random, represented in the figure by the magnetization arrows of the unbound nanoparticles pointing in various directions. The nanoparticles bound to cells, however, are inhibited from such physical motion and hence their magnetization remains substantially the same as when in the presence of the applied magnetic field.

[00130] Fig. 32d illustrates the prepared sample a longer time after removal of the applied magnetic field. The magnetization of the bound nanoparticles has by now also returned to random.

[00131] Fig. 33 is a schematic illustration of measurements from the process described in connection with Fig. 32. Magnetic field is shown as a function of time in a simplified presentation for ease of illustration; in actual practice the units, scales, and shapes of the signals can be different and more complex. At the beginning of the process, corresponding to the state of Fig. 32a, the nanoparticle magnetization is random and the external magnetic field is applied. After that time, the magnetization of the nanoparticles is uniform, corresponding to the state of Fig. 32b. The magnetic field can be ignored for a short time while the unbound nanoparticles return to random

magnetization, corresponding to the state of Fig. 32c. The magnetization can then be measured as the bound nanoparticles transition from uniform to random magnetization, corresponding to the state of Fig. 32d. The characteristics of the measurement magnetization from the state of Fig. 32c to that of Fig. 32d are related to the number of bound nanparticles in the sample, and hence to the number of cells of the type of interest in the sample.

[00132] Fig. 34 is a schematic illustration of an apparatus suitable for use in the present invention. A sample holder 41 is configured to contain a sample such as those described elsewhere herein. A magnetizing system 42, for example Helmholtz coils, mounts relative to the sample holder so that the magnetizing system can apply a magnetic field to the sample. A magnetic sensor system 43 counts relative to the sample so that it can sense the small magnetic fields associated with the magnetized nanoparticles. The system is controlled and the sensor data analyzed by a control and analysis system 44; for example by a computer with appropriate programming.

[00133] The present invention provides methods and apparatuses that can detect metastatic cancer cells in the blood, bone marrow, urine or other body fluids and that can identify the tumor source of these cells. The present invention also provides methods and apparatuses that can identify the tumor source of cancer-indicating markers that can be found in the body serum. The present invention uses biomagnetic sensors and targeted superparamagnetic nanoparticles. The biomagnetic sensors detect magnetic nanoparticles bound by antibodies or other specific agents to cells of interest such as cancer cells.

[00134] In one application of this invention, body serum, such as blood or bone marrow is extracted and has magnetic nanoparticles with specific targeting molecules attached, such as antibodies, mixed into it. Any cells targeted by the targeting molecules (e.g., cancer cells targeted by specific antibodies) will be coated with the nanoparticles specific to that type of cell. The sample is placed under a magnetic field sensor system using magnetic relaxometry to magnetize any nanoparticles in the sample and observe their magnetic field decay. The magnetic relaxometry system allows the specific detection of nanoparticles bound to targeted cells.

[00135] One example of a suitable sensor system is Superconducting Quantum Interference Devices (SQUIDs) which are very sensitive magnetometers that can be used to measure extremely weak magnetic fields based on superconducting loops containing Josephson junctions. In particular, SQUIDs are capable of detecting a few cells targeted by these nanoparticles when the sample is placed in close proximity to the sensor and in a shielded environment. Other types of

magnetometers including atomic magnetometers can also be used for this application.

[00136] When cells are detected and identified, a sensor array consisting of these same or similar sensors can be used to search for tumors in the body of the same type of cells identified in the serum, again using magnetic relaxometry. The particular type of cell identified by the body fluid survey identifies the specific targeting molecule for these cells. Magnetic nanoparticles conjugated with these targeting molecules are injected into the body, providing a sensitive and specific test for finding the tumor or metastasis that has produced the metastatic cancer cells in the serum. This method is capable of detecting several hundred cells that might be localized in a growing tumor.

[00137] In another application of this invention, the presence of certain markers in the serum indicating that there is a metastatic tumor growing somewhere in the body, can be used to perform a search for this tumor using a magnetic relaxometry sensor array. Examples of this are detection of PSA in an individual who has undergone a prostatectomy (where there should be no observable PSA in the blood). Another example is for women who have had ovarian surgery and a level of CA125 is observed in the blood, indicative of an ovarian tumor growing somewhere. In both, and similar, cases, observation of markers in the serum indicating metastasis can be followed by a search for the metastatic tumor of the type identified by the marker. Knowledge of the tumor type indicates the type of antibody or other targeting molecule that will be attached to the magnetic nanoparticle injected into the body to target the responsible tumor. This substantially increases the specificity and sensitivity of the sensor search for the metastatic tumor as compared with previous methods.

[00138] Magnetic relaxation detection of suitable magnetic nanoparticles in serum can be performed in 5 minutes or less. The source contrast of the magnetic relaxation superparamagnetic nanoparticle method of the present invention is many orders of magnitude greater than other detection methods because only bound or hindered particles are observed; the measurement is not impaired by the presence of unbound nanoparticles. The magnetic relaxation sensor method can compete with all prior methods for detecting metastatic cancer cells and offers a new method for then locating the tumor source of these cells in the body for therapeutical applications.

[00139] Examples of suitable antibodies and cells lines, and associated diseases, are listed below. Those skilled in the art will appreciate many more, and the invention is applicable with those and others not yet discovered.

Disease Cell Line Antibody

Leukemia U937 CD34

Leukemia Jurkat CD34

Leukemia GA-10 CD34

Transplant ejection T-cells CD4

Ovarian NIH :OVCAR3 CA125

Ovarian CAOV-3 CA125

Ovarian SKJDV-3 CA125

Ovarian TOV-12D CA125

Breast MCF7-HER2 HER2

Breast MDA-MB-231 HER2 Breast BT474 HE 2

Prostate LNCaP PSMA

Prostate PC-3 PSMA

[00140] More general antibodies can also be used to detect broader ranges of cells. For example, an epithelial marker such as EMA can be used to bind with many types of cells. For example, a blood sample can be combined with targeted nanoparticles having EMA attached, and the nanoparticles will then indicate the presence of some substance to which EMA binds (many of which will be unexpected in blood). Additional examination, such as examination by optical microscopy, can then determine the specific type of cells found. More specific targeting agents can then be used to localize the source of those cells in the patient.

[00141] An example embodiment of the present invention using SQUID sensors and conventional shielding can detect roughly 200 cells or more in a typical blood or tissue sample. An example embodiment using SQUID sensors and conventional shielding can detect about lOng or more of targeted cells in a typical blood or tissue sample.

[001421 NANOPARTICLE PRODUCTION AND MAGNETIC PROPERTIES.. In order to label biological structures magnetically for detection by magnetic relaxometry, the nanoparticles with their coatings of specific antibodies should exhibit different behavior when attached to targets than when not attached to target objects. As an example, when the particles of interest are free to reorient in their suspending fluid, they have relaxation times much shorter than the observation time; however, if rotation of the nanocrystal is hindered by its attachment through antibody-antigen interactions with the target structure, the relaxation time can be comparable to the magnetic relaxometry observation time, typically of the order of a second. In the former case, the relaxation time for rotation in a fluid is governed by the well-known Brownian formula, linearly dependent on the viscosity and the particle's hydrodynamic volume and inversely on the temperature of the suspending fluid. In the latter case, when rotation of the crystal lattice of the particle is hindered, the magnetic moment of the single domain nanocrystal, typically of the order of half a million Bohr magnetons in magnetite, reorients with characteristic relaxation times first discussed by Neel. The Neel time is important because its exponential dependence on particle diameter can limit the sizes of the particles that are suitable for SQUID relaxometry. As discussed below, the " magnetic relaxometry window", the material-dependent size range for which magnetic relaxometry can readily sense the decaying magnetism from relaxing nanoparticles, can be about 2 nm wide centered on a diameter of 25 nm at room temperature, for the magnetite particles discussed as examples here. [001431 Exam le nanoparticles. The nanoparticles used as examples in some of the description herein were obtained from Ocean NanoTech (Springdale, A , USA), designated Ocean SHP 30 lot DE4G. These magnetite particles were suspended in water with an iron content measured to be 28.8 mg [Fe] per ml. Ten μΙ of suspension were pipetted onto a cotton swab "QTip" (Unilever, Trumbull, CT, USA), and were allowed to dry. The description herein assumes the nanoparticles become firmly attached to the cotton fibers of the swab. This was a primary sample for the measurements reported for the methods and results described herein. Since Fe304 is 0.7235 Fe by weight, the mass of magnetite nanoparticles in the sample is 398 μg. From TEM images (Tecnai G2 F30 at 300kV, FEI Corporation, Hillsboro, OR, USA), a distribution of Feret diameters of the particles was determined using the program, ImageJ (produced through an NIH grant and made available on the internet). The Feret diameter is the largest caliper measurement that could be made on the TEM image of the particle. A caliper measurement is the distance between two parallel planes each just touching the surface of the object. The distribution peaked at 25 nm with a full width at half maximum (FWHM) of 4 nm.

[001441 Susceptometry. The induced magnetic moment of the Q Tip sample was determined at 9 frequencies and 30 temperatures, using an MPMS-7 SQUID magnetometer (Quantum Design, San Diego, CA, USA), with a palladium reference sample for calibration. The software associated with the apparatus presents the magnetic moment results in emu (erg/Gauss). These numbers convert to volume susceptibility in the emu system by dividing the magnetic moment by the amplitude of the driving field (2 Oersteds) and by the total volume of the magnetite, 7.56 x 10-5 cc (398 μg at a density of 5.24 g/cc) in the sample. The resulting (dimensionless emu) volume susceptibilities is converted into (dimensionless rationalized MKS) volume susceptibilities by multiplying the former by 4π.

[001451 Magnetic relaxometry. The magnetic relaxometry measurement technique and apparatus used for the measurements reported here has been described in other documents, including the following, each of which is incorporated herein by reference: N. L. Adolphi, D. L. Huber, J. E. Jaetao, et al., J. Magn. Magn. Mater. 321 (2009)1459; E. R. Flynn and H.C. Bryant, Phys. Med. Biol. 50 (2005)1273.

[001461 Neel relaxation, anisotropy and field effects. Neel relaxation is principally governed by the coupling energy of the collective electron spin with special directions in the crystal lattice, the so- called preferred "easy axes", and conversely the "hard axes". This anisotropy energy can also have contributions from the morphology of the nanoparticle such as shape and surface anisotropy.

Configurational anisotropy has been studied in single and pseudo-single domain grains of magnetite and the effects are comparable to magnetocrystalline anisotropy. The treatment below includes, as is commonly done in this field, the anisotropy energy density in a single parameter K (J/m3), and treat magnetite as a uniaxial crystal, ignoring the fact that its magnetocrystalline anisotropy is triaxial (cubic). This approach to understanding the particle is therefore semiempirical in an attempt to restrict parameters to as few as possible for a reasonable description of the magnetic behavior for biomedical applications.

[00147] The Neel relaxation time is also affected by the strength of the applied magnetic field. For purposes of the description below we shall represent the relaxation time in the presence of an applied field by

where β = (1— B I B_k )^a and B_K = 2K / M _s is the anisotropy field. Here a, of order 1, is taken as a fitting parameter; Ms, B, k and Tare the spontaneous magnetization (J/Tm²), applied field (Tesla), Boltzmann's constant (1.38xl0^~23 J/K) and absolute temperature (K) respectively. Note that the presence of a magnetic field decreases the relaxation time, and its effect is gauged by B_K , which for bulk magnetite at room temperature is about 57 mT. For susceptibility measurements the reduction is negligible, but it should be taken into account for relaxometry measurements that require much higher fields (4 mT and higher); in this case we use cs as a fitting parameter. In the modeling described herein, we leave T₀ to be fit with the susceptibility data. We treat T₀ as solely a function of temperature, and not dependent on the particle size.

[001481 The Langevin function. The equilibrium polar angle θ alignment of a classical dipole with an applied field is determined by a balance between the torque exerted on the dipole by the field and the disorienting effect of thermal fluctuations that increases with T. When these two effects are the only agents present, the equilibrium average value of the cosine of the polar angle is determined by the well-known Langevin function L, the classical limit of the quantum-mechanical Brillouin function. However, when the dipole is a single domain nanoparticle, fixed so the crystal cannot rotate, the anisotropy should not be neglected; the anisotropy can hinder alignment with the external field and we can modify the Langevin function as described below.

[001491 Modified Langevin function. Suppose we have a fixed single domain particle whose easy axis is aligned at an angle φ to the z axis. In the absence of a field the collective magnetic moment of the particle will be expected to lie along an easy axis, where the potential energy is minimized, to the extent allowed by thermal fluctuations. In the presence of a field along the z axis the moment-/! will be pulled toward it as well, remaining in the plane determined by the z and easy axes. Here we recognize that the weighting factor in the determination of the average value of cos9 should include, in addition to the term coupling the moment to the field, a term due to the anisotropy energy that also depends on To compute this function we apply the Gibbs distribution: j _e-^{uw, kT} cos 0sin 6H0

< cos 0 >= ^ , (2)

j e-^uw/kT sm Me

0

where υ(θ) =—μΒοο?,θ— Κνοο&² ( —θ) . The easy axis, the direction of the applied field, and the magnetic moment are taken here to lie in the same plane, in that this configuration would be the lowest energy for the moment, and there is nothing to hinder this arrangement. The second term is the contribution from uniaxial anisotropy. Since < cos# > is now a function of φ, the angle between the applied field and the easy axis, we must integrate over it. We assume an isotropic distribution.

« cos Θ »=— < cosd > άφ . (3)

π ^Jo

[00150] Defining dimensionless quantities x = μΒ/kT and y = vK/kT, we find

1 3

« cos # »=— dt— In dds^' #exp(xcos Θ + y cos (θ - φ))

dx ^Jo (4)

[00151] This "Modified Langevin function", L(x, y) , is odd in x and even in y. The result of a numerical computation of eq. 4 is shown Fig. la for various values of y. As an example of the application of this universal function to the specific example case herein, Fig. lb displays the ratio of the Modified Langevin to the Langevin as a function of particle diameter, taking the particles to be uniform spheres, and using bulk values of the parameters for magnetite at 300 K.

[00152] Particles in a pulsed field. To illustrate the application of these results we describe the response of a nanoparticle distribution to a pulsed magnetic field using the so-called Moment

Superposition Model (MSM). The particles are assumed to be spherical and homogeneous. The observed induced magnetic moment and its subsequent decay arises from the mechanisms represented below. A pulse of strength B is applied for a time tpulse, and the magnetic moment M(t) of the sample is given at a time t after the pulse is turned off.

M(t) = n Lw(D)neel(tpulse) exp(-t I )dD

(5)

where n is the number of particles in the sample, μ is the magnetic moment of a particle of diameter D . Here μ = νΜ and L is the modified Langevin function discussed above. w(D) is the diameter probability distribution of the sample. The "Neel factor" is given by

neel{tpulse) = (1 - exp(-tpulse I _N )) , and τ is just τ_Ν with the field off ( β = 1) .

[00153] Susceptibility model. We start with representations of the ac response as described by osensweig and Shliomis , who use the definition:

where M (t) is the magnetization (dipole moment per unit volume as a function of time) in an alternating magnetic field, H(t) = H₀ cos(O)t) , and M₀(t) = %₀H₀ cos(OX) is the "equilibrium magnetization" in the applied field. That is, at any given instant the magnetization is relaxing toward a value it would have were the relaxation time zero. Ignoring frequency dependence of the equilibrium volume susceptibility χ_ϋ , and using the dc expression :

χ₀ = N// < cos 0 > / H_o , (7)

where the number density N is the number of monodisperse magnetic dipoles with moment μ divided by the total volume, however we wish to define it. This is the same as volume susceptibility of a single nanoparticle as described by Worm when we replace N by 1 / . Thus χ₀ is the dc volume susceptibility of an ensemble of noninteracting monodisperse nanoparticles immobilized in a magnetically inert medium. We should be careful with a polydisperse sample, however, such as described below, where we can use the reciprocal of the average volume per particle.

[00154] For small argument x, L{x) ~ , we can replace < cos9 > by BvM_sL(x, y)/3L(x). Calling r = L( , y)/L( ), we see from Fig. la that r does not depend on B when B is small. Thus χ₀ = ₀NM ²r(D)/3fcr, (8)

where μ₀ is the permeability of free space.

[00155] The real and imaginary components of the susceptibility are respectively,

γ> = _J-° (9)

1+(ωτ)² '

and ^λ 1+(ωτ)² * '

[00156] In order to apply these formulae to model the measurements we can integrate over the size distribution of the nanoparticles, which we characterize by the diameter D. The ratio r(D) is shown in Fig. la. Ignoring any possible dependence of K on D, the formulae above become:

y' = _i£2_ Λ · f °° fi²r(D)w(D)dP . .

X 3kT ^J0 [1 + (ωτ)²] * '

and

£o_ ²r(D)w{D)dD

X 3kT ^J0 [1+(ωτ)²] ¹ '

[00157] The concept of susceptibility makes use of the assumption that the induced moment is linear with the applied field in the "small field limit". The small field limit implicit in the susceptibility measurements described herein is below 0.2 mT. One can show numerically that r is essentially independent of B in such small fields for values of D of interest in our measurements.

[001581 Susceptibility measurements of size distribution. The size distribution of the nanoparticles was determined by TEM measurements on a representative sample, about 2500 particles. Fig. 2 displays the distribution of Feret diameters of the sample measured. Although we assume spherical particles, the circularity (4π area/perimeter² ), for example, for a Feret diameter of 25 nm in our sample is around 0.77. A circle is 1.

[001591 Susceptibility measurements of the number density. Using the measured probability distribution w(D) of Feret diameters (Fig. 2) and ignoring any departure from sphericity, we find the average particle volume in our sample,

v = J~ w(D)vdD = 8.1 x 10^"24 m³ . (13)

[00160] From the total volume of magnetite in the sample, we estimate the number of nanoparticles on the Q Tip as n = 9.38 X l0¹² . We take the volume V of the sample to be just the total volume of the particles themselves, so that V = nv . Consequently, the number density N in eqs. 11 and 12 reduces to 1 / v , and n does not enter into our model even though it determines the strength of the total dipole moment measured by the magnetometer.

[00161] A study of a sub sample of the particles whose size distribution determined by TEM is shown in Fig. 2 was analysed by the program ImageJ. There is evidence that the particles in this sample tend to be less round as the Feret diameter increases between 20 and 30 nm, indicating that some of the larger particles may have multiple domains. As such they would contribute to the smaller size particles, for which the measured distribution function, Fig. 2, may be deficient.

[001621 Comparison with model . The model for the susceptibility (MKS) was fitted to the measurements, Fig. 3a, b, by adjusting the values of the three temperature-dependent parameters to minimize the rms differences between the modeled and measured quantities, for each of the 30 temperatures for which the real and imaginary susceptibilities were measured. In the fitting the possibility that these three parameters could depend upon the particle size is ignored. The real and imaginary parts were treated separately and the outcomes can be compared for consistency. Each 3- paramenter curve fit contained 9 data points corresponding to the 9 frequencies. The numerical work was done using Matlab. See Fig. 4a, b.

[001631 The anatomy of susceptibility. The susceptibility at each frequency and temperature is modeled by equations 11 and 12. Representative kernels contributing to the model computation are examined to see where each value arises in the nanoparticle distribution. By "kernel" we refer to the integrand of eqs. 11 or 12 for the real or imaginary part, respectively. To compute these functions we use the parameters obtained by making a best fit to the data at each temperature, except we have fixed Ms at a nominal 2.68x105 J/Tm3. The fitted values of Ms are given below. See Fig. 5.

[00164] As T decreases, the cluster of kernels for each T moves progressively toward the small diameter side of the distribution. For the real parts each kernel extends down to the smallest diameters, whereas for the imaginary parts each kernel is confined to a rather narrow range of sizes. Thus, for the real part the T=400 measurement is sensitive to the entire distribution, and the T=255 K cluster is influenced only by the small diameter particles. See Fig. 5a. We would therefore expect results in the low temperature regions for the real part to be less robust than for the high temperature regions since the modeling is patterned by only the lower wing of the distribution where the statistics are relatively poor. As we shall see, the fitted values of the parameters for the real part for temperatures below 300 K appear to be anomalous and differ significantly from those obtained from the imaginary susceptibility fits.

[00165] In contrast, the kernels for the imaginary susceptibility at each temperature are clustered well within the well-populated part of the size distribution, as shown for T=300K in Fig. 5b. Although the imaginary signal is smaller, with consequently larger relative noise, the parameters do not do not appear anomalous at low temperatures.

[00166] Fitted parameters. The susceptibility model discussed above is compared with the measured values at the nine frequencies for each temperature and the parameters Ms, K and T₀ are adjusted to give the best fit to the these data points. The three parameters are assumed to be temperature dependent, so that the fit at each temperature is made independent of any other. Size dependence of these parameters has been ignored.

[00167] Fig. 6 presents the resulting estimates for the three parameters used in our model. In all three cases the parameters determined from the real and imaginary susceptibilities are much more consistent with each other for temperatures above 300 K than for those below. The likelihood this discrepancy arises from a deficiency in the lower wing of the measured size distribution is demonstrated in these figures, by showing the consequences of adding a small cluster of particles to the lower portion of the size distribution.

[00168] The correlation between K and logT_¾ . Although the variability of the fitted parameters K and logT₀ shown in Fig. 6b and 6c is large, the correlation between the variations is strong. Fig. 7a shows that there is an abrupt threshold just above 10 for -log T₀ . For T greater than about 300K the ratio of these two parameters agrees well for the real and imaginary contributions. For

temperatures below 300 K, the real-based ratio deviates strongly from the imaginary. We have fitted a straight line to the complete set of points from the imaginary susceptibility and those from the real susceptibility above 300 K: = a - bT , (14)

- log T₀

where a = 1636/ I m³ and b = 1.40/ / Km³ . We see the physical significance of this ratio in Fig. 7b, when it inserted into Eq. 15, below.

[00169] In magnetic relaxometry applications the size range of nanoparticles should correspond to a relaxation time of about 1 second. In that case log i ~ 0 , allowing us via Eq.l to find the effective particle diameter,

D_eff « (-6kT\og t₀ \nmi πΚ) . (15)

Consequently, it is the ratio Eq. 14 that determines the required diameter for SQUID relaxometry, an example of which we discuss below.

[001701 A closer look at the fits at 300K. We examine more closely the fitting process for real part data at 300 K, the temperature region of interest in biomedical applications. At this temperature our best fit corresponds to a K of 2.66xl0⁴ J/m³, which is about twice that of the bulk value of magnetite, about 1.35xl0⁴ J/m³. The accompanying fit for the log of T₀ , usually taken as -10, is a -23.4, giving a ratio of 1.14xl0³ J/m³ that when multiplied by 10 would be well within the usually accepted range for K. In spite of the strong correlation between power and K, we cannot simply "renormalize" logT₀ back to 10, however, since in this case the fit to the data is much worse. Below we shall use these same 300 K parameters successfully to fit relaxometry data taken on the same sample.

[001711 The value of T_¾ . Conventionally one takes τ₀ 0.1 or 1 ns for fitting purposes principally because small changes in K can compensate for rather large uncertainties in τ₀. According to Aharoni quoting from Brown (see also Worm) for uniaxial particles:

fo = 2K_Yo {a/n†/2/M_s , (16)

where τ₀ = l/f₀ and a = Taking γ₀ as the gyromagnetic ratio of the electron, 1.76xl0¹¹T ¹s ^Λ , and with K and M_s the bulk values for magnetite near 300 K (1.35xl0⁴ J/m³ & 4.72x10^s J/Tm³, respectively), we obtain— log(r₀) = 10.23, in line with conventional usage and our experimental lower limit, Eq. 13. Notice that τ₀ depends on the particle volume, v, that we take as the average for our distribution, 8.1xl0^"24 m³.

[001721 Summary of susceptibility study. We are able to fit the measured ac susceptibility of a sample of Fe₂0₃ Ocean nanoparticles, and extract values for the properties τ₀, K and M_s as a function of temperature, making use of the MSM model and a modification of the Langevin function to reflect the restraint that anisotropy places on alignment of the dipole moment with external field. Discrepancies from the usually accepted values appear in the preference for large negative values of logj₀ although the ratio -K/logj₀ appears to be remarkably robust for temperatures above 300 K and close to constant for a particular relaxation time. The relative sensitivities of the real and imaginary parts of the ac susceptibility at different frequencies depend on the size distribution and temperature. In our case there is indication from the real part of the susceptibility, which at low temperatures is very sensitive to the smallest sizes, that the measured size distribution is deficient in smaller particles. A possible, but not unique, explanation for this apparent deficit could be that a fraction of the largest particles have broken up into multiple domains. A small change in the number density of large particles could result in a large change in the number density of small particles. This possibility is supported by the observation that the larger particles tend to appear less round than the smaller ones.

[001731 Magnetic elaxometry Measurements. Methods. Here we apply the model discussed above, as well as the susceptibility results, to the understanding of measurements done on the same Ocean 30 DH4G nanoparticles in magnetic relaxometry. As in the susceptibility measurements, the sample consists of nanoparticles in suspension that were allowed to dry on a cotton tip. We assume the particles are firmly attached to the cotton fibers so that the crystal lattices themselves cannot rotate in response to an applied field; the magnetic moment must reorient with respect to its lattice by the Neel mechanism. For these observations the sample is subjected to pulsed flat-topped magnetic fields ranging in strength up to 4 mT. The pulsed fields rise abruptly to a constant amplitude and are terminated abruptly after a fixed duration, with a decay time of a few ms, such that the lingering effects of the applied pulsed field and associated transients are essentially undetectable, beyond 50 ms past the switch-off, by a SQUID array above the sample. The sample and the SQUID array lie along the axis of a Helmholtz pair, with the sample centered between the two coils. The apparatus is described in more detail elsewhere herein.

[00174] We examine the responses of the sample of tethered nanoparticles to fields of three pulse durations, 0.3, 0.75 and 1.5 seconds, comparing to the model the measured decay trajectories of the induced moment of the sample as a function of the time after cutoff. We also examine the excitation curves for the three cases, namely, the induced moment 50 ms after the pulses are switched off as a function of the pulsed field strength.

[00175] The two sets of data, consisting of three curves each, can be modeled with Eq. 6 above. In the decay case we have the magnetic field produced by the decaying remanence moment from the particles as measured by the SQUID array, Bs(t), vs. time, ranging up to two seconds after the pulse is switched off, and, in the excitation case, we have the magnetic moment of the sample M(t) at t = 50 ms (after the field is switched off) vs. the strength of the pulsed field. [00176] Models. To model these data we use the cubic spline fit to the size distribution w(D) for the Ocean SHP 30 DE4G sample (Fig.3) discussed above. We define the "magnetic relaxometry window" as the size range of a particular lot of nanoparticles for which the time constants will allow the SQUI Ds to pick up a measurable signal from the relaxation of the particles after the alignment pulse is switched off (Fig. 7b). The amplitude for the maximum signal we can expect with the above- described apparatus is proportional to

fsw(D) = v - w - neel■ L■ exp(-50ms I τ) , (18)

where L is the modified Langevin function, since the equilibrium alignment with an applied field of the tethered particles is affected by the anisotropy.

[00177] The initial induced moment from a sample of n particles is

M(t = 50ms) = n - Ms - ^ fsw(D) dD . ( 19)

The total magnetic moment of the sample is

M_sample = n - Ms - v w dD . (20)

[00178] The fraction of the magnetic material that will be SQUID-visible is then just the ratio of these two quantities. Integrands of these functions are given in Table 1. The size of the average particle dipole moments excited for the three pulse durations may be calculated:

= M_s - [ fsw dD/[ (fsw/v) - dD

[00179] For the examples shown, with Ms = 2.75e5 J/Tm³, K = 2.06e4 J/m³, logr₀ = -19.4, a = 0.89, we have Table 1.

Table 1.

[00180] Results. Fig. 8 presents the SQUID measurements of a) relaxation and b) excitation of the sample of magnetite nanoparticles for three different pulse durations compared with the model results, using parameters determined for 300 K. The model predictions were normalized for Fig. 8a by one fitted number. In Fig. 8b, in addition to an overall normalization, the exponent was optimized to 0.89. From the Fig. 8b normalization constant the number of particles involved may be extracted and compared with the number expected, as explained elsewhere herein. [001811 Determination of the number of particles. The model moment prediction for the single nanoparticle, y = Ms^■ [ fsw(D)^■ dD , produces a model data set, y_;, for a set of parameters. The measured data set corresponding to the model set is d_;. The predicted value for d_; is ny_t , where n is the number of particles. A least squares fit yields

[00182] For the data and the models to which they are compared in Fig. 8b, we find n is 1.5xl0¹³. Earlier, in the susceptibility measurements described above, this sample, a cotton tip, on which fluid containing the suspended particles had been allowed to evaporate, was estimated to contain 9.38xl0¹² particles.

[00183] This discrepancy may lie with the role of K in the modified Langevin function, which in a limited range of D alters only the normalization of the modeled signal. As we saw in the results in 3.5, the data indicate that K and — 10^ /logT₀ should be larger than conventionally expected, although their ratio seems well-behaved. However, since log T₀ does not play a role in the modified Langevin, K, as appearing in this function, it should perhaps be "renormalized" to be the expected value if — log T₀ = 10. Thus, if we replace K m the modified Langevin by— 10^ / log T₀ , we find the fits are essentially the same, except that the number of particles is now 1.05xl0¹³,in much better agreement.

[001841 Characterization of nanoparticles. The study described below allows us to determine to what extent single-core magnetite nanoparticles exhibiting relatively low size polydispersity will enable us to improve the sensitivity (detectable moment/mg[Fe] ) of the associated techniques. This description focuses on commercially-available single-core magnetite nanoparticles (Ocean

Nanotech, SHP series). In addition to SQUID-relaxometry, we performed SQUID-susceptometry, transmission electron microscopy (TEM), zeta-potential analysis, and dynamic light scattering (DLS). While the measurement timescale of SQUID-relaxometry (50 ms - 2 s) is sensitive to a narrow distribution of particles, SQUID-susceptometry is sensitive to all particles with relaxation times up to ~100 s (i.e., all unblocked particles) enabling a more complete magnetic characterization of the nanoparticle ensemble. Similarly, TEM, DLS, and zeta-potential measurements are sensitive to properties (size and surface charge) of the entire ensemble. Following the methods of Chantrell, we interpret the magnetization curves (M vs. H) as obtained by relaxometry by using a moment superposition model (MSM), which explicitly includes the distribution of nanoparticle sizes. The results of antibody conjugation and incubation of conjugated nanoparticles with target cells are also presented. [001851 Nanoparticles. SHP nanoparticles (Ocean Nanotech, Springdale A ) are single-core magnetite particles, coated with a thin (~4 nm thick) layer of polymer and functionalized with carboxyl groups to enable conjugation to antibodies. We characterized SHP-20 (lot OCK8), SHP-25 (lot SEP8-0), SHP-30 (lots DE4), and SHP-35 (lot SA07) with nominal core diameters of 20, 25, 30, and 35 nm, respectively. The iron concentration (mg[Fe]/mL) of the stock solution was determined destructively by dissolving in acid, forming the phenanthroline/Fe2+ complex, and then quantifying the concentration of a known dilution spectrophotometrically [14] . For comparison, SiMAG nanoparticles (Chemicell, Berlin) consist of multiple magnetite cores embedded in a silica matrix with an overall diameter of ~100 nm.

[00186] SQUID relaxometry. Detection of nanoparticles by relaxometry was performed using a seven-channel low-temperature SQUID array (BTi 2004, 4D-Neuroimaging, San Diego, CA) originally designed for magnetoencephalography. Second order gradiometers with a baseline of 4 cm are used to reject background magnetic fields due to distant sources, allowing the measurements to be performed in an unshielded environment. Due to RF interference, the sensitivity of the system is currently limited to ~10 ¹² T/VHZ. The seven gradiometer coils are located at the bottom of the liquid He dewar, 1.9 cm from the outer dewar surface, arranged with six in a circle of 2.15 cm radius and one at the center. For in vitro measurements on small samples, the sample is located at a distance z ~ 2.8-3.5 cm below the bottom of the center coil. The samples are uniformly magnetized (parallel to the center gradiometer axis) using a 60 cm-square Helmholtz array (L = 16 mH, B/l = 0.107 mT/A) powered by a 3 kW current-regulated supply. The decaying magnetization is sampled at a rate of 1 kHz (beginning 50 ms after switching off the magnetizing pulse) and digitized using a National Instruments PXI8336 16-channel digitizer and LabVIEW 8.5.1 acquisition software (National Instruments, Austin, TX). Our standard measurement protocol is to apply a 3.8 mT field for 0.3 s and then acquire data for 2 s (see Fig. 1), with 10 repetitions, subsequently averaged, to improve signal- to-noise. Magnetization curves (M vs. H) are obtained, using the same apparatus and timing, by varying the applied field between 0 and 4.0 mT.

[00187] Data analysis is performed using the Multi-Source Analysis program, written in our lab using MATLAB (The MathWorks Inc., Natick, MA). After signal averaging and removal of 60Hz line frequency contamination, background data (acquired with no sample) is subtracted from the sample data. The seven relaxation curves are fit by a logarithmic function, to determine the DC offset, and an exponential function to the first 200 ms of the decay to determine the magnetic field amplitude at each sensor position 50 ms after the magnetizing pulse is switched off. In order to solve the inverse problem, we fit the spatial dependence of the magnetic field by modeling the sample as a single magnetic dipole, which allows us to determine the location (x, y, z) and magnetic moment (m_z) of the source. It is assumed that the magnetic vector points along the z axis, the direction of the magnetizing field. The least-squares fit is performed using the Levenberg-Marquardt algorithm. For medical imaging, a multiple dipole model may be used to determine the spatial coordinates and moments for multiple discrete sources using n different sample positions - equivalent to a sensor array with 7n elements.

[00188] To obtain the desired Neel relaxation of the nanoparticles, the particles should be immobilized. In the case of cell samples, the antibody-conjugated nanoparticles are immobilized by the binding of the antibodies to receptors on the cell surface. In the case of unconjugated nanoparticles, 10-20 μΙ of stock nanoparticle solution is applied to a Q-tips cotton swab (Unilever, Trumball, CT) and allowed to dry in air.

[001891 SQUID-susceptometry. Magnetic characterizations were performed using an MPMS-7 SQUID magnetometer system (Quantum Design, San Diego, CA, USA). DC Magnetization curves were acquired by equilibrating the sample at the measurement temperature in zero field, then incrementally increasing the field and pausing 100 seconds at each field before measurement. Five sequential measurements were taken at each field, a mean of those measurements calculated, and the three values with the lowest deviation from the mean retained. The mean of these three values was then reported as the moment. Zero Field Cooled (ZFC) curves were determined by cooling the sample in the absence of a magnetic field to 5 K, then slowly warming in a 1 mT field. After thermally equilibrating at a target temperature, a series of five measurements was taken and the values were processed as described above for magnetization curves to obtain a reported magnetization value. AC susceptibility curves were performed by applying a series of 0.2 mT AC magnetic fields (f = 0.1 - 1000 Hz) to a thermally equilibrated sample. The in-phase (χ') and out-of-phase (χ") AC susceptibilities of the sample were digitized and recorded.

[001901 TEM and Image Analysis. The nanoparticles were imaged by transmission electron microscopy (Tecnai G² F30 at 300kV, FEI Corporation, Hillsboro, Oregon, USA).

[001911 Nanoparticle-antibody conjugation. Nanoparticles were conjugated by the following method. Ten mg of nanoparticles were aliquoted into 15 ml conical tubes (Greiner Bio-One, San Diego, CA) and brought to a total volume of 10 ml with double distilled water. N- hydroxysulfosuccinimide (Sulfo-NHS) (Pierce, Rockford, II) and l-ethyl-3-[3- dimethylaminopropyl]carbodiimide hydrochloride (EDC) (Pierce, Rockford, II) were prepared fresh at a concentration of 25 mg/ml each in separate tubes with double distilled water. One hundred microliters of the EDC and Sulfo-NHS were each added to the nanoparticles and incubated at room temperature on a LabQuake shaker (Lablndustries, Inc., Berkeley, CA) for 20 minutes. Nanoparticles were brought to pH 8.0 with 50 mM NaHC0₃ (Sigma-Aldrich, St. Louis, MO), 50 μg of antibody (BD Biosciences, San Jose, CA) was added, and the mixture was incubated at room temperature on a LabQuake shaker for 2 hours. The antibody-nanoparticle mixture was centrifuged at 7,500 RCF for 30 minutes at 4°C. The supernatant was removed and 10 ml of double distilled water was added to the pelleted NPs. The centrifugation parameters were repeated once more and the supernatant was removed. The remaining pellet was resuspended in a total volume of 240 μΙ PBS/ (Gibco-BRL, Rockville, MD) 0.5% fetal bovine serum (FBS) (HyClone, Logan, UT). Conjugated nanoparticles were stored at 4°C for up to 2 days.

[001921 Incubation of cells with nanoparticles. Jurkat cells were harvested and washed using sterile PBS (Gibco-BRL). Trypsinized Jurkat cells were treated with 3mls Trypsin/EDTA (Gibco/BRL) at 37C for 5 minutes. After tyrpsinization, cells were washed with sterile PBS. Harvested cells were counted using 0.4% Trypan Blue Solution (Sigma, St Louis, MO) and a hemocytometer (Hausser Scientific, Horsham, PA). Each sample contained lxlO⁷ cells suspended in 200 μΙ of cold PBS/0.5% FBS solution to which 0.8mg of CD3-coupled Ocean nanoparticles were added. Cells and CD3- nanoparticles were incubated on ice with SQUID measurements taken at 1, 15, 30 and 60 minutes. For light microscopy, glass slide preparations were prepared using a Cyto-centrifuge (Shandon, Pittsburgh, PA) and stained with Prussian blue by TriCore Reference Laboratories (Albuquerque, NM).

[00193] Characterization bv TEM, Relaxometry, and Susceptometry. Fig. 9A shows TEM images of each nanoparticle sample. The SHP-20 particles show the most uniform intensity in TEM, whereas the number of particles exhibiting dark and light banding becomes increasingly prevalent in the SHP- 25, -30, and -35 particles. The particle size distributions (feret diameter), determined by using the program Image J by analyzing 2500 particles from each sample using multiple TEM fields, are shown in Fig. 9B. The average particle diameter determined for each set of particles is in good agreement with the nominal diameters specified by the manufacturer (see Table 2). The narrowest size distribution is obtained for the SHP-20 particles, and the standard deviation in particle diameter is observed to increase with average diameter.

[00194] Fig. 10A shows the imaginary component of the AC susceptibility at room temperature as a function of frequency. While the SHP-20 particles show a clear peak at ~300 Hz and the SHP-25 sample appears to be peaking at ~0.1 Hz, there is only a weak maximum in the AC loss for the SHP- 30 sample, and no peak is evident for the SHP-35 sample. The measurement timescale of the SQUID- relaxometry technique (50 ms to ~2 s) corresponds to frequencies of approximately 0.1-3 Hz.

Indeed, the amplitude of χ" (per kg) at 1.0 Hz correlates well with the magnetic moment (per kg) detected by SQUID relaxometry at room temperature (using a 0.3 s, 3.8 mT pulse) as shown in Fig. 10B, demonstrating that AC susceptibility measurements may be used to predict the SQUID relaxometry response.

[00195] Table 2. Physical and electrical properties

[00196 Table 3. Magnetic properties

[00197 I *measured by SQUID relaxometry using a 3.8 mT, 0.30 s magnetizing pulse

[00198] ** measured by DC susceptometry

[00199] The SHP-25 particles give rise to the largest detectable moment/kg, roughly 4 times greater than the moment/kg of the SHP-30 particles, and nearly an order of magnitude greater than the moment/kg of multi-core particles we have characterized previously. Therefore, we have achieved a significant improvement in detection sensitivity by switching from multi-core particles to single-core particles of relatively uniform diameter.

[00200] However, the improvement is not as great as anticipated based on the narrow size distributions of the SHP particles. Given that a relaxometry signal is detected from all four sets of particles, an examination of the size distributions in Fig. 9B suggests that the optimal particle diameter is roughly 26 nm (the diameter at which all of the distributions overlap). If this is the ideal diameter, then theoretically, particles in the 26 ± 1 nm size range should have the appropriate relaxation time to contribute to the detected moment/kg. In that case, we can estimate from the size distributions in Fig. 9B that the detected moment/kg should be approximately 30 J/T/kg for the SHP-25 particles and 40 J/T/kg for the SHP-30 particles. These estimates further assume that all particles exhibit bulk values of K and Ms, and are magnetized according to the Langevin function with B=3.8 mT and T=300 K. However, the observed moments/kg are an order of magnitude lower than these estimates, which demonstrates that particle diameter is not the only significant factor determining the detectable moment/kg. In particular, the fact that the observed moment/kg from the SHP-25 particles is ~4 times greater than that of the SHP-30 particles, although the diameters are nearly the same, clearly indicates that other nanoparticle properties must be considered.

[00201] In order to determine why the moment/kg is not as large as predicted based on the size distributions, further investigations of the nanoparticle magnetic properties were undertaken. Fig. 11 shows magnetization curves measured by DC susceptometry near the blocking temperature (see Table 3) of each sample. The data is plotted as M/M_s (dimensionless) vs. H to compare the shapes of the curves. For all samples, the magnetization does not rise nearly as sharply with field as expected for magnetite particles with diameters in the 20-40 nm range. The theoretical curves for 25 nm and 8 nm diameter particles, calculated using the Langevin function, are plotted in Fig. 11 (dashed and dotted lines, respectively) for comparison. For the theoretical curves, we assumed a volume saturation magnetization of 3.3 x 10^s J/T/m³ (70% of the bulk value), which is typical for nanoparticles, and T = 335 K. The observed slopes of M/M_s vs. H at small H are therefore much smaller than expected, and are theoretically consistent with particles in the 7-10 nm diameter range. Furthermore, the Langevin formalism predicts that the slope of M/M_s vs. H at small H should decrease with decreasing particle diameter; but surprisingly, this set of samples shows the opposite trend, with the highest slope obtained from the smallest (SHP-20) particles.

[00202] Examination of the SHP-30 particles by high resolution TEM suggests that the reduced slopes are caused by polycrystallinity, which may also explain the non-uniform TEM intensities of some particles observed in Fig. 9A. A representative image showing a particle with multiple crystallite orientations is shown in Fig. 12. We suggest that the observed decrease in the slope of M/Ms vs. H (Fig. 11) with increasing particle size may therefore be explained as a consequence of a decrease in average crystallite size with increasing particle diameter. Even assuming a small effective particle size, the magnetization curves in Fig. 11 are be adequately fit using the Langevin function, if a single value of μ is assumed. In order to estimate the saturation magnetization for each sample, satisfactory fits of the data (solid lines in Fig. 11) were obtained using the sum of Langevin functions corresponding to two distinct μ values. The saturation magnetization values so-obtained (summarized in Fig. 10) fall between 55 and 92% of the bulk value (92 J/T/kg) for magnetite. We note that the lower saturation magnetization of the SHP-30 sample, relative to the SHP-25 sample, partially explains the lower observed moment/kg.

[00203] Without a direct measurement of K, it is not possible to determine the particle diameter that is giving rise to the SQUID-relaxometry signal. However, as noted above, a comparison of the size distributions (Fig. 9B) and the observed moments/kg (Fig. 10A) suggests that the diameter of the nanoparticles detected by relaxometry is approximately 26 nm, which is close to the predicted value of 23.8 nm assuming K_bu|_k (=1.35 x 10⁴ J/m³). If we assume that the diameter of the relaxometry- detected particles is 26 nm, this implies that K = 1 x 10⁴ J/m³, about 75% of the bulk value. We note that K would have to be extremely large (10-40 times the bulk value) to yield relaxation times of order 1 s from crystallites with diameters in the 7 - 10 nm range, suggesting that the small crystallites within the polycrystalline particles in the ensemble do not contribute to the observed SQUID relaxometry signal. This suggests that the lower than expected moment/kg observed by relaxometry can be explained by assuming that only monocrystalline particles of the correct diameter contribute to the observed relaxometry signal.

[00204] Fig. 13 shows M/M_max vs. H curves obtained by SQUID relaxometry using a 0.3 s duration pulse at room temperature. Three of the curves are nearly straight lines, indicating that much higher values of the pulsed field (not available using our current hardware) would be required to determine the saturation magnetization of the relaxometry-detected particles. Thus, in order to compare the shapes of the magnetization curves, we have normalized them using the value observed at 4 mT (M_max). Because the signal detected by relaxometry arises from a narrowly distributed subset of the particles in each sample, one might expect the M vs. H curves measured by relaxometry to exhibit nearly ideal (single μ) behavior. However, only the SHP-20 magnetization curve shows the expected shape, consistent with the Langevin function (calculated assuming 27.4 nm diameter, M = 70 J/T/kg, and T = 300 K).

[00205] Although the SHP-25 magnetization curve has slightly negative curvature, and can also be fit by the Langevin function, the resulting value of μ is 1.2x10 ¹⁸ J/T, corresponds to a particle diameter of only 17.6 nm and is not at all consistent with the observed particle size distribution. Further, this diameter implies that K must be approximately 3.5 x 10⁴ J/m³, about 3 times the bulk value, in order to obtain a 1 s relaxation time at room temperature. Because the SHP-35 magnetization curve has positive curvature, it cannot be fit by the Langevin function. Thus, the Langevin function (computed for a single value of μ) is not an adequate model for determining the magnetic moment, and therefore the particle size, that gives rise to the relaxometry signal. [002061 Moment Superposition Model Analysis. The moment superposition model (MSM ) was developed to treat the case of a distribution of non-identical, non-interacting, superparamagnetic particles. We now show that the apparently anomalous shapes of these magnetization curves can be explained by taking into account the shape of the distribution of moments from each sample. Further, the MSM analysis results in more reasonable values of K and particle diameter.

[00207] We model the magnetization observed by relaxometry in terms of a distribution f( ) of N dipole moments:

M = NJ ( )L( 5/fcr)(l - exp(-i/?M/5i' / r( , 5)) exp(-i / r( ,0)) ^ , (2)

0

where L is the Langevin function, tpulse is the duration of the magnetizing pulse, t is the time after the end of the magnetizing pulse, and the Neel relaxation time is taken to be

τ(μ, B) = r₀ εχρ((μΗ_κ 12kT)(l - B /H_K )^a ) _{i (3)}

where H_K is the anisotropy field, with H_K = 2K/M_S.

[00208] The measured nanoparticle size distributions (see Fig. 10) were converted to magnetic moment distributions by multiplying the particle volumes by the volume saturation magnetization, which was calculated using the measured mass saturation magnetization (see Table 3) for each sample multiplied by the density of magnetite (5240 kg/m³). We fit each distribution of dipole moments assumin a log-normal form:

[00209] The calculated curves in Fig. 13 (dashed lines) were then obtained from Eqs. 2 and 3 as follows: The pre-factor (τ₀) in Eq. 3 was set to 10 ¹⁰s (see A. Aharoni, Introduction to the Theory of Ferromagnetism, Oxford Science Publications, Oxford Press 2007, incorporated herein by reference, p. 94). The only parameters that were adjusted were N, the total number of nanoparticles, K (the anisotropy energy density) and σ (the exponent on the field dependence in Eq. 3). The calculated curves in Fig. 13 (dashed lines) were all obtained with K=9175 and σ=0.9, which reproduced the shape of the measured curves reasonably well. Although better fits to the data can be obtained by independently varying K and σ for each data set, the fits shown in Fig. 13 illustrate that the same parameters can be used to generate magnetization curves which differ significantly in curvature, simply due to differences in the distribution of moments. Typically, when an M vs. H curve exhibits positive curvature at small values of H (as we observe from the SHP-35 sample), it is attributed to the presence of multiple domains (i.e., ferromagnetic behavior). However, the MSM results demonstrate that in the case of M vs. H curves obtained by relaxometry, positive curvature can be obtained at low H assuming only superparamagnetic relaxation, if the average moment in the distribution is larger than the narrow range of observable moments.

[00210] The value of K obtained from the MSM analysis (9175 J/m³) implies that a particle with a relaxation time of 1 s at 300 K is 27.6 nm in diameter, in reasonable agreement with our crude estimate of 26 nm (the diameter where the 4 size distributions in Fig. 9 overlap). Further experimental studies of nanoparticles exhibiting much lower polydispersity will be required to more precisely determine the size and magnetic moment of the "ideal" particle for detection by relaxometry.

[002111 Antibody Conjugation, Colloidal Stability, and Observable Moment per Cell. Due to their superior observed moment/kg, we attempted to conjugate the SHP-25 particles to Her2/neu antibodies using the protocol recommended by the manufacturer. However, several attempts at conjugation each resulted in the precipitation of very large nanoparticle aggregates, such that the SQUID relaxometry signal from the aggregated particles was indistinguishable from the signal obtained from immobilized particles. We note that the SQUID relaxometry technique is theoretically insensitive to unbound nanoparticles in solution due to their rapid Brownian relaxation; however, if the hydrodynamic diameter of nanoparticle aggregates is sufficiently large, the Brownian relaxation time of the aggregated particles may fall within the 50 ms - 2 s time scale of the SQUID relaxometry measurement, in which case the signal is detected (even though the particles are not bound to cells). For even larger aggregates, the Brownian time constant may exceed 2 s, in which case, the observed relaxation will be dominantly due to the Neel process, and the detected signal will be difficult to distinguish from that of cell-bound nanoparticles. Thus aggregation should be avoided in order to retain an important advantage of the SQUID relaxometry technique, namely its specificity for detecting cell-bound particles. Fortunately, conjugation of the SHP-30 particles to a variety of monoclonal antibodies (against Her2/neu, CD34, and CD3) resulted in colloidally-stable nanoparticle solutions, which did not give rise to significant SQUID relaxometry signals. Therefore, the SHP-30 nanoparticles were used in the cell experiments described below.

[00212] In order to determine the cause of the stability difference between the SHP-25 and SHP-30 particles upon conjugation, further characterizations of the stock solution (as-received) SHP nanoparticle solutions were performed. Dynamic light scattering (DLS) was performed to determine the hydrodynamic size distributions of the four samples. Given that the edge-to-edge distance between cores (see Fig. 9A) is consistent with a coating thickness of only a few nanometers, the >70 nm hydrodynamic diameters suggest that there was mild aggregation of all of the SHP particles prior to conjugation, although the aggregate size was small enough not to result in a detectable SQUID relaxometry signal from the stock solutions. This mild aggregation is the greatest for the SHP-25 particles, which showed an average hydrodynamic diameter of 106 nm. Because the conjugation chemistry involves pH changes, differences in the zeta potential of the SHP-25 and SHP-30 nanoparticles can explain the different stabilities in the presence of conjugation reagents. However, only small differences in the zeta potential and electrophoretic mobility of the SHP-25 and SHP-30 nanoparticles were observed (see Table 2). For comparison, we also measured the DLS diameter, zeta potential and electrophoretic mobility of some multi-core nanoparticles (SiMAG particles, nominal 100 nm diameter, Chemicell) which showed varying degrees of stability after conjugation. Nanoparticles from SiMAG lot 0808/07 were reliably conjugated to CD34 antibodies using the same conjugation protocol described here and were then used in a number of leukemia cell experiments. Although the SiMAG 0808/07 particles (as-received) had a less negative zeta potential (-20.7 mV) compared to any of the Ocean SHP particles, they showed little aggregation by DLS prior to conjugation (129 nm average hydrodynamic diameter), and very minimal aggregation (by SQUID relaxometry) after conjugation. SiMAG Lots 1903/08 and 1803/08 showed increasing levels of aggregation by DLS (431 nm and 1760 nm average hydrodynamic diameters, respectively) that correlated with increasingly less negative zeta potentials (-18.8 mV and -8.5 mV, respectively) and an increasing tendency to aggregate after conjugation. Thus in the case of the SiMAG particles, differences in aggregation appear to be adequately explained by differences in the zeta potential. However, the zeta potential of the SHP-25 particles, which is quite negative and not very different from that of the SHP-30 particles, does not explain why the SHP-25 particles show severe aggregation upon conjugation.

[00213] Instead, we suggest that significant differences in the magnetic properties of the SHP-25 and SHP-30 particles are the underlying cause of the aggregation differences. Previously Martin et al. have shown that nanoparticles with χ > 5 (volume susceptibility, in MKS units) may undergo significant clustering due to strong interparticle dipolar interactions, even in the absence of an applied magnetic field. As shown in Fig. 14, the volume susceptibility of the SHP-25 particles (not including the coating) is approximately 20, compared to χ ~ 4 for the SHP-30 particles. If one includes the approximately 3 nm thick non-magnetic coating, the overall volume susceptibility is reduced by roughly a factor of two, resulting in χ ~ 10 for the SHP-30 particles, which is still well above the χ > 5 threshold at which spontaneous clustering is predicted to occur. We therefore suggest that electrostatic repulsive forces between the SHP-25 particles are great enough to maintain colloidal stability in the as-received solution. However, when the pH is raised just prior to adding antibody, screening of the surface charge on the nanoparticles reduces the electrostatic repulsion and allows the attractive interparticle magnetic dipolar interactions to dominate, resulting in spontaneous aggregation. This suggests that increasing the thickness of the non-magnetic coating (to > 8 nm in this case) may be necessary to ensure that particles with high magnetic susceptibilities, which are desirable from the magnetic standpoint, can be conjugated to targeting ligands without inducing aggregation.

[00214] To determine the observable magnetic moment per cell, SHP-30 nanoparticles were conjugated to a monoclonal antibody that targets the CD3 cell-surface antigen, whose expression is correlated with acute transplant rejection. The CD3-conjugated nanoparticles were then incubated with 10⁷ Jurkat cells, which express high levels of the CD3 antigen on the cell surface. To assess the binding specificity, the same CD3-conjugated nanoparticles were also incubated with 10⁷ Jurkat cells pre-treated with trypsin, an enzyme that hydrolyses proteins, resulting in the digestion of cell surface antigens. Fig. 15A shows the SQUID relaxometry signal obtained during incubation of the untreated and trypsinized Jurkat cells with the CD3-nanoparticles. The incubation curves are the average of three experiments, and the error bars represent the standard error of the mean. The SQUID signal is significantly higher for cells that retain the CD3 antigen, as expected, but there is still a significant relaxometry signal (with a Neel character) observed after incubation with the trypsinized cells. Photomicrographs (Fig. 15B) suggest that some binding to the surface of the trypsinized cells did occur. However, nanoparticle aggregates, some of which are cell-sized, are visible in both the untreated and trypsinized cell experiments, and these aggregates appear to be responsible for a significant fraction of the signal observed from the trypsinized cells. We note that a cell (~10 micron diameter) is more than 100 times larger than a single antibody-conjugated nanoparticle (~70 nm diameter). Thus single nanoparticles are not visible at the magnification shown, and visible aggregates in the photomicrographs can be assumed to be of order 500 nm or larger, large enough to give rise to detectable relaxometry signals.

[00215] The number of nanoparticles that will theoretically bind to a single cell is limited by steric hindrance to roughly 150,000, assuming a single random-close-packed layer of antibody-conjugated particles (modeled as spheres of diameter 70 nm covering a 15 micron diameter cell). The observed signal due to CD3-specific binding (difference signal in Fig. 7A) is 1.9 x 10⁵ pJ/T for 10⁷ cells after 60 minutes. Given that the observable moment/kg is 0.83 J/T/kg[Fe₃0₄], this indicates specific binding of 23 pg of magnetite per cell, which is equivalent to 550,000 nanoparticles per cell. The number of specifically bound nanoparticles therefore exceeds both the steric limit for monolayer coverage and the number of CD3 receptors per cell (~100,000) determined by flow cytometry. The higher than expected number of nanoparticles per cell is certainly beneficial from a detection sensitivity standpoint, but the exact mechanism is not yet understood. The additional binding may be the result of internalization of the nanoparticles by the cells or the tendency of clusters of antigen-bound nanoparticles to attract additional nanoparticles to the cell surface. Further work may identify the cause of both the apparently enhanced antigen-specific binding and the significant non-specific signal, particularly the development of biocompatible surface coatings that minimize nanoparticle aggregation.

[00216] Example embodiments and applications - Magnetic Nanoparticles for In-Vivo Detection and Localization of Disease. Embodiments of the present invention can be used in ultra-sensitive biomagnetic methods for: (1) early detection and localization of disease, (2) image-guided therapy to treat disease, (3) monitoring and controlling treatment. Fig. 16 is an illustration of magnetic relaxometry for in-vivo detection of disease. Note, in relation to Fig. 16, (1) Superparamagnetic nanoparticles (~24 nm) coated with antibodies specific to disease cells; (2) Particles attach to target cells in-vitro or in-vivo; (3) Magnetizing field (~40 Gauss); applied with short pulse (~0.75 sec); (4) Delay for induced fields settling; (5) SQUID sensors measure; decaying field from nanoparticles; (6) Particles bound to cells decay by Neel mechanism (~1 sec), unbound particles decay by Brownian (~^sec).

[00217] Fig. 17 is an illustration of an example Superconducting Quantum Interference Device (SQUID) sensor system for relaxometry. Note, in relation to Fig. 17, a liquid helium dewar for cooling low temperature SQUIDs, seven 2^nd order gradiometers for sensors, coils for magnetizing fields, 3-D nonmetallic stage for holding cells and animals, larger coils for human subjects.

[00218] Fig. 18 is an illustration of results with particular cells. Note, in relation to Fig. 18, that results for breast, ovarian, T-cells, and leukemia have been obtained; there is a large number of nanoparticles per cell, with the number dependent on the number of ligands per cell; unbound particles give no moment; some phagocytosis occurs; magnetic relaxometry gives moment/cell, sites/cell, number of cells in the sample; antibody sites per line: (a) MCF7 breast 11 x 10⁶, (b) SK- OV-3 ovarian, 6.39 x 10⁶ , (c) BT-474 breast, 2.75 x 10⁶ , (d) MCF7 breast 0.18 x 10⁶ , (e) MDA-MB-231 0.11 x 10⁶, (f) non-breast/ovarian < 4000. Fig. 19 is an illustration of biomagnetic sensitivity compared to other methods.

[00219] Fig. 20 is an illustration of results with animal models. Note, in relation to Fig. 20, (1) Human cancer cells injected into SCID mice and tumors allowed to grow; (2) Nanoparticles injected into Mice; (3) Mice imaged by SQUID sensor system as a function of time; (4) Tumor locations obtained and verified; 1 - 2 mm accuracy; (5) Cell numbers obtained and verified by histology.

[00220] Fig. 21 is an illustration of the application to a magnetic biopsy needle. Note, in relation to Fig. 21, (1) Magnetic Nanoparticles - CD34 labeled to target Leukemia Cells; (2) Added to bone marrow; (3) Needle with small magnets inserted; (4) After 2 min needle extracted; (5) Cells collected placed under SQUID; (6) M D determined. [00221] Fig. 22 is an illustration of issues associated with production and characterization of nanoparticles. Note, in relation to Fig. 22, (1) Size must be ~24 nm; (2) Ideal particles are monodispersed; (3) Maximum moment/mg (Fe); (4) Commercial products unreliable; (5) Particles must have biocompatible coatings.

[00222] Fig. 23 is an illustration of susceptometry measurements of nanoparticle properties. Fig. 24 is an illustration of properties before and after addition of PEG.

[00223] Other information and references. Understanding of the present invention can be facilitated by the following references, each of which is incorporated herein by reference:

A. Aharoni, Phys. Rev. B 7, 1103-1107 (1973);

Adolphi, N.L., Huber, D.L., JaetaoJ.E., Bryant, H.C., Lovato, D.M.,Fegan, D.L., Venturini, E.L., Monson, T.C., Tessier, T.E., Hathaway, H.J., Bergemann, C, Larson, R. S., Flynn ER, Characterization of magnetitie Nanoparticles for SQUID-relaxometry and magnetic needle biopsy, JMMM 321 (2009) 1459-1464, PMID 20161153;

Amikam Aharoni, Introduction to the Theory of Ferromagnetism, Oxford, 1996;

ASTM E394-00, Standard Test Method for Iron in Trace Quantities Using the 1,10-Phenanthroline

Method, 2000. ;

Bryant, HC, Sergatskov DA, Lovato D, Adolphi N, Larson RS, Flynn ER, Magnetic needles and superparamagnetic cells, Phys. Med. and Biol. 52 (2007) 4009-4025, PMID 17664592

C. Mikkelsen, M. Hansen, H. Bruus, J. Magn. Magn. Mater. 293 (2005) 578;

D. Eberbeck, F. Wiekhorst, U. Steinhoff, et al., J. Magn. Magn. Mater. 321 (2009) 1617;

D. Eberbeck, F. Wiekhorst, U. Steinhoff, et al., J. Phys.: Condens. Matter 18 (2006) S2829;

D. Eberbeck, S. Hartwig, U. Steinhoff, L. Trahns, Magnetohydrodynamics 39 , 77 (2003), "Description of the magnetisation decay in ferrofluids with a narrow particle size distribution";

E. Heim, F. Ludwig, and M. Schilling, J. Magn. Magn. Mater. 321 (2009) 1628;

E. Romanus, D.V. Berkov, S. Prass, et al., Nanotechnology 14 (2004) 1251;

F. Gazeau, M Levy, and C. Wilhelm, Nanomed. 3 (2008) 831;

F. Ludwig, E. Heim, M. Schilling, et al., J. Appl. Phys. 103 (2008) 07A314;

F. Ludwig, E. Heim, M. Schilling, J. Appl. Phys. 101,113909 (2007) "Characterization of

superparamagnetic nanoparticles by analyzing the magnetization and relaxation dynamics using fluxgate magnetometers";

Flynn ER, Bryant, HC, Bergemann C, Larson RS, Lovato D, Sergatskov DA, Use of a SQUID array to detect T-cells with magnetic nanoparticles in determining transplant rejection, JMMM, 311 (2007) 429-435, PMID 18084633; Flynn ER, Detection and Treatment Possibilities of Disease with Magnetic Nanoparticles, 6th International Conference on the Scientific and Clinical Applications of Magnetic Carriers, Mayl7-20, 2006, Krems, Austria, Invited Talk;

Flynn ER., Bryant, HC, A SQUID based system for in-vivo cancer imaging, Phys. Med. and Biol. 50 (2005) 1273-1293, PMID 15798322;

G. Basso, F. Lanza, A. Orfao, et al., J. Biol. Regul. Homeos. Agents 15 (2001) 68;

G. R. Gamble, Journal of Cotton Science 8 (2004) 198.;

H. C. Bryant, N.L. Adolphi, Dale Huber, Danielle Fegan, Todd C. Monson, Trace E. Tessier, Flynn ER, Magnetic Properties of Nanoparticles Useful for SQUID Relaxometry in Biomedical Applications, submitted to JMMM, April, 2010;

H.C. Bryant, D.A. Sergatskov, D.A Lovato, et al., Phys. Med. Biol. 52 (2007) 4009;

H.-U. Worm, Geophys. J. Int. 133, 201-206 (1998) , "On the superparamagnetic-stable single domain transition for magnetite, and frequency dependence of susceptibility". See eq 5.;

J. E. Jaetao, K. S. Butler, N. L. Adolphi, D. M. Lovato, H. C. Bryant, I. Rabinowitz, S. S. Winter, T. E.

Tessier, H. J. Hathaway, C. Bergemann, Flynn ER, and R. S. Larson, Enhanced Leukemia Cell Detection

Using a Novel Magnetic Needle and Nanoparticles, Cancer Res 69 (21)2009, 8310-8316, PMID

19808954;

J. E. Mayer and M. G. Mayer, Statistical Mechanics, p. 347, John Wiley, London, 1940.;

J.E. Jaetao, D.M. Lovato, N.L. Adolphi, et al., Cancer Res. 2009 Nov l;69(21):8310-6;

J.E. Martin, E.L. Venturini, and D.L. Huber, J. Magn. Magn. Mater. 320 (2008) 2221;

K. Enpuku, T. Tanaka, T. Matsuda, et al., J. Appl. Phys. 102 (2007) 054901;

L. D. Landau and E. M. Lifshitz, Statistical Physics, p. 77, Second Revised and Enlarged Edition,

Addison-Wesley (1969);

L. Neel, Adv. Phys. 4 (1955)191.;

M. L. Shliomis, Sov. Pys.JJspekhi (Engl, transl.), 17, (2 ),153-169(1974), "Magnetic Fluids".;

N.L. Adolphi, D.L. Huber, H.C. Bryant, T.C. Monson, D.L. Feganc, J.K. Lim, J.E. Jaetao, T.E. Tessier, D.M. Lovato, K.S. Butler, P.C. Provencio, H.J. Hathaway, S.A. Majetich, R.S. Larson, and E.R. Flynn, Characterization of Single-core Magnetite Nanoparticles for Magnetic Imaging by SQUID- relaxometry, submitted to Phys. Med. and Biol., June, 2010.

N.L. Adolphi, D.L. Huber, J.E. Jaetao, et al., J. Magn. Magn. Mater. 321 (2009) 1459;

R. E. Rosensweig, Journal of Magnetism and Magnetic Materials 252 (2002)370-374, "Heating magnetic fluid with alternating magnetic field".;

R. Kotitz, W. Weitschies, L. Trahms, et al., J. Magn. Magn. Mater. 201 (1999) 102;

R. Tietze, R. Jurgons, S. Lyer, et al., J. Magn. Magn. Mater. 321 (2009) 1465; R. W. Chantrell, S. R. Hoon, B. K. Tanner, J. Magn. Magn. Matter 38, 133-141 (1983), "Time- dependent magnetization in fine-particle ferromagnetic systems".;

R.M. Ferguson, K.R. Minard, and K.M. Krishnan, J. Magn. Magn. Mater. 321 (2009) 1548.

R.W. Chantrell, S.R. Hoon, B.K. Tanner, J. Magn. Mater. 38 (1983) 133;

Ricardo Aragon, Phys Rev B 46, 5334-5338 (1992), "Cubic magnetic anisotropy of nonstoichiometric magnetite".;

S. A. Majetich and M Sachen, J. Phys. D: Appl. Phys. 39 (2006) R407-R422, "Magnetostatic interactions in magnetic nanoparticle assemblies: energy, time and length scales";

S. Ge, X. Shi, J.R. Baker, et al., Phys. Med. Biol. 54 (2009) N177;

S.H. Chung, A. Hoffmann, S.D. Bader, et al., Appl. Phys. Lett. 85 (2004) 2971;

W F Brown, Phys Rev 130, 1677-1686 (1963) , "Thermal Fluctuations of a Single -Domain Particle";

Wyn Williams, Adrian R. Muxworthy, Greig A. Paterson, J. Geophys. Res. , 111, B12S13 (2006),

"Configurational anisotropy in single-domain and pseudosingle domain grains of magnetite";

Y.R. Chemla, H.L. Grossman, Y. Poon, et al., Proc. Natl. Acad. Sci. 97 (2000) 14268;

[00224] The present invention has been described as set forth herein in relation to various example embodiments and design considerations. It will be understood that the above description is merely illustrative of the applications of the principles of the present invention, the scope of which is to be determined by the claims viewed in light of the specification. Other variants and modifications of the invention will be apparent to those of skill in the art.

Claims

What is claimed is:

1) A method of extracting cells of a first type from a sample, comprising:

a. providing targeted nanoparticles, each comprising a magnetic nanoparticle conjugated with a targeting agent that preferentially binds to cells of the first type;

b. introducing the targeted nanoparticles to the sample in a manner that allows bonding of the targeting agents to cells of the first type; and

c. subjecting the bound targeting nanoparticles to the operation of a magnetic device to extract from the sample cells to which are bound targeted nanoparticles.

2) A method as in claim 1, wherein the targeted nanoparticles comprise magnetic nanoparticles conjugated an antibodies specific to cells of the first type.

3) A method as in claim 1, wherein the magnetic device comprises an elongated member having disposed thereon a plurality of magnetic regions.

4) A method as in claim 3, wherein the magnetic device further comprises a nonmagnetic sheath removably mounted over the elongated member.

5) A method as in claim 3, wherein the elongated member comprises a stainless steel rod, and wherein the plurality of magnetic regions comprises a plurality of permanent magnets.

6) A method as in claim 5, wherein the permanent magnets comprise NdFeB magnets.

7) A method as in claim 4, wherein the sheath comprises a polyimide sheath.

8) An apparatus for the extraction of cells of a first type from a sample, comprising a plurality of targeted nanoparticles, each comprising a magnetic nanoparticle conjugated with a targeting agent that preferentially binds to cells of the first type; a sample holder configured to contain a sample which can include cells of the first type; and a magnetic device configured to be disposed in relation to the sample such that the magnetic device acts on targeted nanoparticles that have been bound to cells of the first type in the sample holder and extracts said cells from the sample.

9) An apparatus as in claim 8, wherein the targeted nanoparticles comprise magnetic nanoparticles conjugated an antibodies specific to cells of the first type.

10) An apparatus as in claim 8, wherein the magnetic device comprises an elongated member having disposed thereon a plurality of magnetic regions.

11) An apparatus as in claim 10„ wherein the magnetic device further comprises a nonmagnetic sheath removably mounted over the elongated member.

12) An apparatus as in claim 10, wherein the elongated member comprises a stainless steel rod, and wherein the plurality of magnetic regions comprises a plurality of permanent magnets.

13) An apparatus as in claim 12, wherein the permanent magnets comprise NdFeB magnets. 14) An apparatus as in claim 11, wherein the sheath comprises a polyimide sheath.

15) A method of detecting, measuring, or a combination thereof, cells of a predetermined type in a first sample comprising:

a. preparing a second sample by combining the first sample with a plurality of targeted nanoparticles;

b. subjecting the second sample to an applied magnetic field; and

c. measuring the relaxation of the magnetic field induced in the bound nanoparticles, and from the measurement determining the presence, concentration, prevalence, or other characteristic of cells of the predetermined type.

16) A method as in claim 15, wherein subjecting the second sample to an applied magnetic field comprises subjecting the second sample to a uniform magnetic field of strength and duration sufficient to impose a uniform magnetization on the nanoparticles in the second sample that are bound to cells of the determined type.

17) A method as in claim 15, wherein the targeted nanoparticles comprise magnetic

nanoparticles bound to targeting agents, wherein the cells of the predetermined type have binding sites or other affinities for the targeting agents.

18) A method as in claim 17, wherein the targeting molecules comprise antibodies.

19) A method of detecting the presence of cancer in a patient, comprising:

a. obtaining a first sample of fluid, serum, tissue, or other substance collected from the patient, wherein the presence of a predetermined substance in the sample indicates the presence of cancer in the patient;

b. providing targeted nanoparticles, comprising magnetic nanoparticles conjugated with targeting agents, wherein a targeting agent comprises an agent that preferentially binds with the predetermined substance;

c. providing a second sample by subjecting the first sample to the targeted

nanoparticles under conditions that allow binding of the targeting agents to the predetermined substance;

d. imposing a known magnetization on the nanoparticles in the second sample;

e. measuring the relaxation of the magnetization of the nanoparticles after imposition of the known magnetization; and

f. determining the presence of the predetermined substance in the second sample from the relaxation of the magnetization. 20) A method as in claim 19, wherein the predetermined substance is cells of a specific type of cancer, and wherein the targeting agent comprises one or more antibodies that bind to those cells.

21) A method as in claim 20, wherein the predetermined substance is PSA, and wherein the targeting agent comprises a PSA-specific antibody.

22) A method as in claim 20, wherein the predetermined substance and targeting agent comprise one or more or the pairs set forth in the specification.

23) A method of localizing a recurrence or metastasis of cancer in a patient, comprising: a. detecting the presence of cancer according to any of the methods of claims 1-9; b. injecting the patient with targeted nanoparticles, wherein the targeted nanoparticles preferentially bind with cells indicative of the recurrence or metastasis of cancer; c. scanning the patient with a magnetic relaxation instrument, and identifying when the measured magnetic relaxation indicates bound nanoparticles; and

d. indicating the presence of recurrence or metastasis of cancer responsive to the

identification of bound nanoparticles.

24) A method as in claim 23, wherein the cells indicative of recurrence or metastasis

comprise cancer cells.

25) A method as in claim 24, wherein the targeted nanoparticles comprise magnetic

nanoparticles conjugated with antibodies that preferentially bind with cancer cells.