Patent Application Serial No. 60/116,633, filed January 20,1999, and entitled "Magnetically Responsive Particle".

Background of the Invention The invention relates to gel particles.

In humans the metabolic requirements for iron can be determined by the rate of absorption which is matched by a normally fixed rate of loss. A state of iron deficiency can occur if dietary intake of iron is insufficient, if iron absorption by mucosal membranes is impaired, or if excessive bleeding occurs. Severe cases may lead to iron deficiency anemia. Conversely, genetic defects in the absorption of iron can result in an overload of iron which can lead to chronic liver disease and premature death.

Iron is stored as a component of the protein ferritin and as part of insoluble hemosiderin. Iron is transported from the absorption and storage sites to the sites of use via specific iron-binding proteins-most notably transferrin, which is usually one-third saturated with iron.

While serum ferritin concentrations are a useful index of storage iron, they are not as informative in clinical practice when iron deficiency is often complicated by concurrent inflammatory or chronic disorders. In these situations, as well as for the assessment of patients with iron overload, the best discrimination is obtained

from combined measurements of serum iron concentration and the total iron-binding capacity (TIBC), which is a measure of both unsaturated and saturated transferrin- expressed in terms of iron concentration. From these two measurements one is able to calculate the percent transferrin saturation, a sensitive index for iron overload.

The measurement of TIBC involves first adding excess iron, usually as ferric ion, to the serum specimen and allowing the iron to bind to (e. g., saturate) all the available sites on the iron-binding proteins. Then the free (unbound) iron that remains is removed, by absorption or adsorption, and the total iron which remains bound is measured. Each step of this process causes some dilution of the serum, but it has been shown that a threefold dilution in the determination of TIBC is advantageous, since this brings the iron-saturated transferrin into the same concentration range as the original serum iron concentration. This allows better control of the instrument and chemistry parameters.

The typical absorants/adsorbants that have been used to remove excess ferric ion, without stripping transferrin-bound iron, include magnesium carbonate, amberlite resin, and activated alumina. These all usually require centrifugation, which yields a supernate that contains the protein-bound iron for analysis. A popular modification uses dry alumina columns for adsorption. In this format, the diluted serum, containing excess iron, is poured onto the column and the eluate is collected for TIBC analysis.

Summary of the Invention Accordingly, in one aspect, the invention features a gel particle that includes a scavenger, a magnetically responsive particle (MRP) and a gel. The scavenger can be, for example, a cell, a receptor, charcoal, silica or alumina. The scavenger and the MRP are disposed within the gel.

A scavenger is a particle that can have an analyte (e. g., DNA, a polynucleotide, a protein, a metalloprotein, a metal ion, or a nucleoside acid, such as cyclic AMP, cyclic GMP or their intracellular messengers) attach thereto. Examples of scavengers include a cell, a receptor, charcoal, silica and alumina.

A gel particle is a particle that includes a gel having a scavenger and an MRP disposed within the gel (e. g., the scavenger and the MRP are encapsulated in the gel, which is a shape-retaining gel).

Within the gel particles, the scavenger may or may not be attached (e. g., chemically bound) to the MRP.

Examples of gels include polyacrylamide, agarose and alginate. Gels can be gelled or cross-linked.

Gel particles can be formed by adding a scavenger and an MRP to a solution of liquefied gel (e. g., sodium alginate) to form a suspension, and then expanding the suspension through an air jet. Gel particles can also be formed using emulsification,

dripping and Rayleigh jet. Methods of preparing a gel particle containing a cell are disclosed in, for example, U. S. Patents No. 4,701,326; 5,567,451; and 5,846,530.

In another aspect, the invention features a gel particle that includes a scavenger, an MRP and a gel. The scavenger and the MRP are disposed within the gel. The gel should have a porosity that allows one or more desired analytes through while not allowing undesired analytes through. The particular undesired analytes will vary depending upon the desired analyte (s). Examples of desired analytes include DNA, a protein, a metalloprotein or a metal ion. Examples of undesired analytes include cellular debris, cell walls, proteins or antibodies.

In some embodiments, a desired analyte can be a metal ion (e. g., iron ion), and an undesired analyte can be the desired analyte bound to a protein (e. g., a metal ion, such as an iron ion, bound to a protein, such as transferrin). In certain embodiments, a desired analyte is iron ion, and an undesired analyte is iron-bound transferrin.

The porosity of gel can be, for example, about 1-1000 nanometers (nm) (e. g., about 20-200 nm, or about 200-500 nm).

In other aspects, the invention features a sample (e. g., from a patient) interacting with one of these gel particles.

In a further aspect, the invention features a method of removing an analyte from a sample. The analyte can be, for example, a protein, DNA, a metal ion, a

metalloprotein, a steroid, or a metal-containing analyte. The method includes contacting the sample with a gel particle that a magnetically responsive particle. The method also includes removing the analyte from the sample.

Methods, reagents, and reaction mixtures of the invention provide advantageous methods of determining TIBC. TIBC methods of the invention can (a) eliminate centrifugation, (b) be rapid, (c) use small amounts of serum (e. g,. less than about 500 uL), and/or (d) can be adapted to on-line automation or semi-automation.

All publications, patent applications, and patents cited herein are incorporated by reference. Other features and advantages of the invention will be apparent from the following description of the figures, detailed description, and from the claims.

Brief Description of the Figures Fig. 1 is a graph of a regression analysis of a correlation study; and Fig. 2 is a graph of a regression analysis of a linearity study.

Detailed Description Sample The preferred sample is serum, but plasma can also be used. Suitable sample sizes can be less than 1,000 microliters (e. g., less than about 500 microliters, such as about 200 microliters or 100 microliters). They can be 10-500 microliters (e. g., 5-10

microliters, or 50-75 microliters). The sample can be from a patient (e. g., a human patient) or from a culture (e. g., tissue culture or fermentation process.

Sequestration Embodiments of the invention can allow, for example, the determination of TIBC in a sample, without a centrifugation step.

The free iron in a sample can be sequestered by contacting the sample with a gel particle containing a scavenger (e. g., alumina) and an MRP. The gel particle can serve to separate the sample into a TIBC phase and a phase which includes sequestered Fe. The partition into phases can be effected without centrifugation and/or without removing either phase from the container which holds the sample.

Scavengers Scavengers should be capable of binding an analyte of interest. Scavengers include cell receptors, ion absorbers (e. g., alumina, silica or charcoal), metal complexing agents (e. g., crown ether, and dendrimers) and ion exchange resins.

Other scavengers include polyacrylic acid, deferoxamine and other iron-binding iminocarboxylic acids (e. g., EDTA, EGTA, NTA and/or HEDTA), and other iron- binding carboxylic acids (e. g., tartaric acid, malonic acid, succinic acid, citric acid and/or tricarballic acid). For embodiments in which the analyte is iron, the scavenger does not need to be specific for iron as long as it absorbs Fe++ and/or Fe.

In certain embodiments, the scavenger can be neutral grade alumina, such as commercially available from Scientific Adsorbents, Inc. (Atlanta, GA), J. T. Baker (Phillipsburg, NJ), or Sigma Chemical Co. (St. Louis, MO).

Magnetically Responsive Particles (MRP's) MRP's are particles which respond to a magnetic field. In embodiments in which iron is the analyte, MRP's which include an iron binding ligand can be included in the gel particle. The free iron passes through the gel, and is bound to the MRP's. A magnetic field is applied to the sample to partition the gel particles (e. g., by bringing them to the bottom of the sample container) and to form a phase which is free of MRP's, the TIBC phase. The iron in the TIBC phase (e. g., supernatant phase) is then measured.

An MRP can be made by the entrapment or precipitation of iron, nickel or cobalt salts in the particle rendering the particle paramagnetic. The particles should not release or leach free iron. A suitable example is a particle made from styrene co- polymerized with acrylic acid in the presence of an iron, nickel or cobalt salt which are encapsulated with an additional co-polymerized surface layer. Similar particles are available commercially as ESTAPOR superparamagnetic particles. Another suitable example is a particle is polyacrolein polymerized with precipitated metal salts (e. g., iron salts, such as magnetite), and then ground to size. Such particles are commercially available from, for example, Cortex Biochem, Inc., under the

trademark MagAcroleini. Other suitable particles are commercially available from Polysciences, Inc. (Warrington, PA, trademark PolyMag and Bangs Laboratories (Fishers, IN).

Other suitable particles are those which are the product of the polymerization of monomers such as methacrylate, divinylbenzene, acrylamide, acrylic acid, or polyethylene in various combinations and proportions. Polymerization should be performed in the presence of iron, nickel or cobalt salts and an additional polymerization step should be used to encapsulate the particles.

Other MRP's include particles of cellulose, agarose (e. g., type 1A agarose, commercially available from Sigma Chemical Co., located in St. Louis, MO, or commercially available from FMC Bioproducts, located in Rockland, ME), or other polysaccharides with an iron, nickel or cobalt salt precipitated within the particle.

These particles can be washed in strong acid to remove surface iron, nickel or cobalt which could leach from the particle. An example of such a particle is sold by Cortex under the name MAGCELL.

Iron, nickel or cobalt can also be covalently attached to a particle.

Iron chelating polymers can be used as iron binding ligands in methods of the invention. U. S. Patents No. 5,487,888 and 3,899,472 describe a variety of chelating polymers. These polymers can be used to encapsulate magnetic particles for use in methods described herein.

In certain embodiments, the particle is preferably less than about 1,000 microns in diameter, and more preferably less than about 10 microns, or less than about 1 micron in diameter. In some embodiments, the particle is about 1 micron in diameter.

Measurement of Iron Iron in a sample can be measured (e. g., with a colorometric method). For example, the sample can be reacted with a dye (e. g., ferene or ferrozine, or other pyridyl or phenanthroline dyes which form colored complexes with iron).

Determination of TIBC The following method can be used to measure TIBC.

(1) Mix a solution containing Fe at a concentration sufficient to saturate the iron binding proteins in the solution with serum. Under these conditions the added iron binds to the serum iron binding proteins to"saturate"any available binding sites; (2) incubate for five minutes; (3) add a slurry of gel particles; (4) incubate for five minutes;

(5) put a magnet into contact with the sample tube, and the iron which did not attach to the serum proteins (which is sequestered by the gel particles) is pulled from suspension/solution; and (6) measure the supernatant Fe to determine the TIBC of the sample.

The method need not be performed in any particular sequence of steps.

For example, the Fe solution can be added to the sample (e. g., serum) first, and allowed time to saturate all the available sites. After this step, the gel particles can be added to remove excess iron, the sample can be incubated, a magnetic field can be applied, and the measurement of supernatant iron can be made.

In a preferred embodiment, the gel particles and high Fe concentration are mixed, and used as a single reagent. The sample (e. g., serum) is added to this reagent and allowed time for the iron to transfer to the available sites on the serum proteins, a magnetic field is applied, and a measurement of supernatant iron is made.

In this embodiment, the affinity of iron for the sites in the sample (e. g., serum protein sites) is much greater than the affinity of iron for the solid-phase ligand.

In another embodiment, the gel particles and the sample (e. g., serum) are first combined, then the high Fe concentration solution is added (wherein the serum proteins and the gel particles compete for the excess iron sample), allowed to incubate, a magnetic field applied, and a measurement of supernatant iron made.

The following examples are illustrative and not be construed as limiting.

Examples In the following examples, the alumina was neutral grade alumina purchased from Scientific Adsorbents, Inc. (Atlanta, GA), the agarose was type 1A purchased from Sigma Chemical Co. (St. Louis, MO), and the paramagnetic particles were MagAcrolein polyacrolein paramagnetic particles purchased from Cortex Biochem, Inc. (San Diego, CA).

Correlation Study Experiments were conducted to obtain a comparative evaluation between the magnetic separation TIBC methodology and the alumina column separation TIBC methodology. 103 patient samples were taken, spanning from 107 to 570 ug/dL TIBC per patient.

The magnetic separation TIBC methodology used an iron saturating reagent that contained 24 mg/L ferric chloride (i. e., 800ug/dL Fe) in 1 milliMolar (mM) citric acid solution (Reference Diagnostics product number 4021) and a magnetic separation reagent that contained 300 mg/mL magnetically responsive alumina particles in 20mm MOPS buffer, pH 7.5 (Reference Diagnostics product number 4022).

The alumina column separation TIBC methodology used an iron saturating solution that contained not less than 500 ug/dL in Imm citric acid solution and

disposable columns that were prefilled with 300mg of alumina (J&S Medical Associates, Inc., product numbers ICM 117 and ICM 127).

Serum iron reagents and a calibrator were used with ferene chromogen at 40 mM (DMA, Inc. product number 1580-300).

All iron measurements we performed on a Roche COBAS Mira Plus chemistry analyzer using the DMA serum iron reagent.

A regression analysis of the results was performed for an objective correlation. The results of the regression analysis are shown in Fig. 1 (y = 1.02x- 5.7; with a correlation coefficient of 0.99).

Results Patient Alumina Column Magnetic Particle Number TIBC TIBC 1 343 361 2 402 411 3 367 391 4 438 450 5 371 392 6 186 170 7 230 229 8 248 262 9 236 230 10 168 153 11 177 177 12 168 163 13 219 211 14 187 184 15 159 154 16 195 195 17 186 174 18 417 424 19 399 408 20 399 403 21 381 382 22 351 340 23 365 359 24 371 368 25 363 364 26 381 365 27 365 374 28 279 266 29 264 267 30 91 107 46131470 32 502 512 47733503 34 331 330 35 235 243 36 277 280 37 252 230 30938308 39 213 210 40 217 219 41 344 345 42 206 197 31243303 44 327 322 45 392 404 46 340 334 47 341 332 35848361 49 365 346 50 289 286 32951329 37452390 53 272 270 54 296 292 17755185 56 362 370 57 342 58 325 338 59 302 309 296 316 44561421 62 273 283 63 326 325 64 373 368 65 234 244 66 392 380 34667339 29968305 31369313 70 238 213 71 277 276 72 176 166 27973274 74 204 207 57075554 76 192 190 26977276 78 183 183 79 339 359 80 363 366 81 328 330 43882423 83 219 213 25584259 85 423 430 86 261 259 87 176 173 22088211 89 339 352 90 377 378 28691274 41492403 93 411 417 17794173 24395234 96 255 256 97 145 146 98 374 381 99 357 379 100 399 404 101 392 408 102 137 141 103 101 108

Precision Study A study was conducted to test the precision (both within run and between runs) of the magnetic TIBC method. Precision was evaluated with quality control materials at two TIBC concentrations. All iron measurements were performed on the Roche COBAS Mira Plus chemistry analyzer using DMA iron reagents.

With run precision was determined by performing 25 magnetic separations for each control and measuring the iron content of each supernate (all in the same run). For an average level of 314 ug/dL, the standard deviation was 6.1 ug/dL and the coefficient of variation (c. v.) was 1.9%. For an average level of 158 ug/dL, the standard deviation was 4.5 ug/dL and the c. v. was 2.8%.

Between run precision was determined by performing a magnetic separation for each control 25 different times over about two weeks, and each time the iron content of the two supernatant solutions were measured in a different run. For an average level of 302 ug/dL, the standard deviation was 13.2 ug/dL and the c. v. was 4.4%. For an average level of 155 ug/dL, the standard deviation was 10.3 ug/dL and the c. v. was 6.6%.

Linearity Study A study was conducted to determine whether a linear response exists across the expected range of TIBC measurements.

Two serum specimens, at the high and low ends of the typical patient distribution range, were each measured using the magnetic TIBC method. They were then combined in different proportions to produce four additional samples. The proportions (high/low) were: 80/20; 60/40; 40/60; and 20/80. The samples were also measured using the magnetic alumina TIBC method and compared to the expected TIBC concentrations.

The results are shown in Fig. 2 (y = 1. Olx + 7.4; with a correlation coefficient of 0.999). Expected Observed 812 812 658 693 503 526 349 358 202 216 40 40

Interference Study An interference study was conducted to determine whether certain substances interfered with the magnetic TIBC measurements.

Bilirubin, hemoglobin, and ascorbic acid were added to samples in various amounts to determine whether these substances caused significant interference.

Triglyceride interference was determined by diluting a high triglyceride specimen (Tg=1466 mg/dL, TIBC=340 ug/dL) with a specimen that had much lower triglyceride and TIBC concentrations (Tg=72 mg/dL, TIBC=94 ug/dL). The two specimens were mixed in varying proportions (80/20; 40/60; 60/40; and 20/80) to yield four additional samples with intermediate triglyceride and TIBC concentrations. Each sample was tested with the magnetic TIBC method and the results were compared to calculated (expected) TIBC values.

These tests showed that the magnetic TIBC method is not interfered by: bilirubin at a level of up to at least 35 mg/dL; ascorbic acid at a level of up to at least 50 mg/dL; hemoglobin at a level of up to at least 200 mg/dL; and triglyceride at a level of up to at least 1450 mg/dL. Bilirubin TIBC (ug/dL) _ 42.0 236.5 34.4 247.5 24.6 247.0 11.9 248.0 6.1255.0 0.3 247. 5 HemoglobinTIBC (mg/dL) (ug/dL) 500 312.0 300 287.5 200 274.0 150 279.0 50 268.0 25 267.5 273.5 Ascorbic TIBC Acid (ug/dL) (mg/dL) 50 318.0 40 322.7 30 316.3 20 311.7 10 314.7 0 319. 7 Total Expected Observed Percent Triglycerides TIBC TIBC Recovery (mg/dL) (ug/dL) (ug/dL) 1466 340 340 100.0 1206 291 282 96.9 920 242 235 97.1 642 192 184 95.8 364 143 141 98.6 72 94 94 100. 0

Other embodiments are in the claims.

What is claimed is: