In the mid-1980s, John Kemshead and John Ugelstad came to visit me (Paul Liberti) at Jefferson Medical College in Philadelphia where I had just founded Immunicon.  Kemshead, a close friend and colleague from our days together at Mill Hill, was working at The Imperial Cancer Fund.  Ugelstad, a Norwegian chemist had recently invented a process to produce uniform polymer micro-beads and had developed an ingenious method to load them up with magnetite crystals rendering them highly magnetic. John and John had been collaborating for some time with Kemshead ‘teaching’ Ugelstad (his words) what surfaces were required to be compatible with biological systems and what kind of receptors (primarily MAbs) needed to be attached. They had finally reached a point in their collaboration where they had usable beads and with them John Kemshead had just performed the first tumor (neuroblastomas) cell depletion of bone marrow. Kemshead went on to treat numerous pediatric neuroblastoma patients by marrow depletion in combination with timed chemotherapy. The mutual interest in magnetic separation technology was based on a project we would be doing at Immunicon involving the removal of immune complexes from patients with Rheumatoid Arthritis with the goal being a new treatment approach. I was in need of one or more separation systems. Also I had been working on a colloidal magnetic nanoparticle that Immunicon had acquired from a Jefferson colleague, Charlie Owen. Owen had developed a method of taking protein along with Fe+2 and Fe+3 salts solution, co-precipitating them with base and resuspending the co-precipitate with mild sonication rendering them colloidal (100-115 nm nanoparticles). [Those weakly magnetic materials were the forerunners of the highly magnetic ferrofluids I later on developed at Immunicon.] 

The four of us spent an entire day in lively discussion (the latter part of which took place at Doc Watson’s Pub) debating the merits of small vs. large magnetic particles.  Ugelstad kept showing us his electron micrographs of these beautiful uniform beads which were indeed impressive.  Parenthetically, every time thereafter when I would hear Ugelstad speak, he almost always showed that micrograph and would smile broadly, proud of his accomplishment in creating these uniform little creatures.  Owen and I kept coming back with the fact that large beads compared to cells diameters would physically block cell surfaces.  We could isolate cells and not see any magnetic materials on their surfaces.  And, was there any advantage to them being uniform?

Both Johns got us on how difficult our colloids would be to separate as their magnetic moments were so small by comparison.  The Ugelstad beads could be separated with fairly crude magnets. We had to use internal magnetic gradient columns to separate our magnetic colloids (something my good friend Stefan Miltenyi would be perfecting and commercializing some years later at Miltenyi Biotec.)  Both Johns claimed that would bring us to a dead end.  We argued that because our materials were colloidal and had Brownian Energy – thus finding their targets quickly rather than having to shake your cells with large beads ‘like you were making Martinis” that nano-sized ferrofluids would win the day. We also added that with their enormous surface area per mass, we would be using substantially less material all of which would lead to better separations. That day and its discussion left me quite unsettled. Should Immunicon acquire rights to Ugelstad’s beautifully homogeneous magnetic microspheres which would push us more to being a magnetics company (and that’s what John Ugelstad eventually did by establishing Dynal) or should we stay the course of using magnetic and other technologies to achieve our goals in therapy and diagnostics?

While I was toying with the notion of ‘should we or shouldn’t we’, I made a startling discovery.  Even though I had made the Maxwell equation calculations many times on the magnitude of magnetic gradient required to capture our colloidal nanoparticle (>100kGauss/cm), I decided to do a ‘silly’ experiment. I placed a 10 penny nail in a magnetic field where I would induce a magnetic gradient of about 15 kG/cm at its surface and brought a tube of our magnetic colloid up onto the nail. In short order, the magnetic material had collected on the front and back surfaces of the nail that faced the magnet poles (as it should).  Realizing from Maxwell that magnetic gradient required is proportional to magnetic mass, I had an ‘aha’ moment. The only thing that could be changing was magnetic mass – so the little critters must be lining up in N-S pole chains when they in a magnetic field. Hence, mass goes way up and they are much easier to collect than expected. That was good enough for me and so I passed on John Ugelstad’s particles. I am guessing those involved over the years at Dynal and now Invitrogen and Life Technologies are happy to hear that.  [After I did my nail experiment, I shared with Ugelstad my chaining notion and several years later he verified the notion with Dynal Dynabeads by viewing them under a microscope in a magnetic field.]

Was that a wise decision on my part? That is hard for me to say. But my simple 10 penny nail experiment led me to introducing quadrupoles and other multipole open field separators to magnetic separations which is how most of the field does magnetic separations now. The end of the discussion may not yet be over as both small [particularly the high magnetic versions we developed at Immunicon] and large magnetic particle technology have proven their prowess.  With BioMagnetic Solutions’ ongoing efforts developing next generation ferrofluids and a new concept magnetic separator, we look to add some new chapters.