To date, MRSw's have used magnetic nanoparticles (NPs) that react with molecular targets to aggregate and decrease T₂. Although MRSw's detect highly multivalent viruses or bacteria with high sensitivity,[7,8] their sensitivity for proteins is far lower.[9,10] Our goal was to improve the sensitivity of MRSw sensors for divalent antibodies, so the immune status of birds or humans might eventually be determined using a POC MR relaxometer. A monoclonal antibody recognizing the Tag peptide from a hemagglutinin of a human influenza virus was used as a target to assess strategies for improving MRSw sensitivity.

An obvious strategy for improving MRSw sensor sensitivity was to employ the equivalence principle of antibody-antigen reactions involved in precipitin formation,[11-13] reducing the concentration of Tag peptide-magnetic particles (a synthetic multivalent antigen), so that lower concentrations of anti-Tag antibodies might achieve aggregation. The Tag peptide from the influenza hemagglutinin was therefore conjugated to nanoparticles (NPs) and far larger micron-sized magnetic particles (MPs), to yield Tag-NPs and Tag-MPs. As indicated in Table 1, a Tag-MP contained 350,000 times more iron than a Tag-NP and had a correspondingly higher magnetic moment per particle. With a typical initial T₂ for MRSw assays (100 msec), the concentration of particles decreased from 2.8 × 10⁻⁹ M with Tag-NPs to 5.1 × 10⁻¹⁵ M using Tag-MPs.

Attempts to use Tag-MPs for MRSw assays indicated Tag-MPs differed from earlier NPs in two key respects. First, as shown in Figure 1, Tag-MP's underwent a reversible increase in T₂ in the 0.47 T relaxometer magnet. Micrographs indicated the T₂ increase was associated with MP aggregation resulting from magnetic attractions between Tag-MP's(see Figure 2 of reference [14]), with the randomizing effects of thermal energy causing dispersion when the field was removed. Second, Tag-MP aggregation resulted in a T₂ increase while Tag-NP aggregation resulted in a T₂ decrease,[3,15] observations consistent with the outer sphere diffusion theory used to describe the effects of magnetic particles on T₂.[16-19] This theory employs two parameters, τ_d, the diffusion time for water, and Δω, the difference in angular frequency between the magnetic field experienced by a proton at the particle (or aggregate) surface versus that in the bulk. Outer sphere diffusion theory predicts T₂ will decrease as Tag-NPs aggregate, since the motional averaging condition is fulfilled with both dispersed and aggregated materials (Δω*τ_d <1). With Tag-MP aggregation, the resulting magnetic field inhomogeneities become so few and infrequent water molecules must traverse long distances to encounter them. Here the effects on T₂ become diffusion limited (Δω*τ_d >1). The effects of NP and MP aggregation on T₂ have been discussed.[14]

Three techniques were then explored separately and in combination to increase the sensitivity of MRSw assays, see Figure 2. First, a decrease in particle concentration was achieved by replacing the magnetic nanoparticle (NP) with a micron-sized particle (MP), see Figure 2a. The second technique, magnetic field enhanced, molecular target aggregation, exploited the reversible magnetic field induced MP aggregation in the absence of a target ligand shown in Figure 1. We hypothesized that the magnetic field aggregation of Tag-MPs might enhance anti-Tag mediated aggregation (Figure 2b). Therefore solutions of Tag-MPs were exposed to anti-Tag in the relaxometer magnet, resulting in both magnetic field induced and anti-Tag antibody mediated Tag-MP aggregation. Samples were then removed from the magnetic field, to allow Tag-MPs to disaggregate which occurred only in the absence of particle crosslinking by anti-Tag. T₂ was determined in less than 30 seconds by placing samples in the relaxometer so that magnetic field induced aggregation during T₂ measurement was minimal. Finally, we employed valency enhancement, where the valency of the monoclonal anti-Tag was increased above the normal two antigen-combining sites per IgG by adding a sheep F(ab′)₂ antibody to the Fc fragment of the mouse anti-Tag monoclonal antibody (Figure 2c).

We first measured T₂ as a function of anti-Tag concentration using Tag-NPs and Tag-MPs (decreasing particle concentration but without magnetic field or valency enhancement) as shown in Figure 3a. Data were fit to a 4-parameter equation, to obtain the EC50 (midpoint), the Hill coefficient (curve slope), and computer generated maximum and minimum values of T₂. Table 2 provides EC50's, Hill coefficients, and ΔT₂'s for the curves shown in Figure 3a. ΔT₂, the T₂ for the fully dispersed state minus the T₂ for the fully aggregated state, is positive for anti-Tag induced Tag-NP aggregation but negative for anti-Tag induced Tag-MP aggregation. %ΔT₂ is the difference in T₂ in the presence and absence of anti-Tag divided by ΔT₂ and expressed in percent.

Using Tag-NPs, increasing concentrations of anti-Tag gave an EC50 of 49 nM, versus an EC50 of 0.60 nM for Tag-MPs. Further reductions in the EC50's of the MP based MRSw assays were obtained by the use of magnetic field and valency enhancement as shown in Figure 3b. Through a combination of all three techniques, the EC50's for anti-Tag were progressively decreased from a starting value of 49 nM with Tag-NPs to 0.0020 nM (2.0 pM), obtained with decreased particle concentration, magnetic field aggregation, and valency enhancement (Figure 3c).

To obtain a detection limit from these results, we determined a “projected sensitivity concentration” or PSC. Five msec was added to (or subtracted from) the value of T₂ in the absence of anti-Tag and the concentration at this T₂ determined from curve parameters (Table 2). The PSC for anti-Tag with Tag-NP sensor, the starting point of our investigation, was 26 nM. By using our three techniques for enhancing sensitivity, the PSC was reduced to 0.00014 nM or 186,000 fold. The PSC is discussed further in Supporting Information. This detection sensitivity is comparable with those of most advanced nanostructure-based antibody assays.[20]