Magnetic Separation Techniques in Diagnostic Microbiology

Vol. 7, No. 1

CLINICAL MICROBIOLOGY REVIEWS, Jan. 1994, p. 43-54

0893-8512/94/$04.00+0 Copyright X 1994, American Society for Microbiology

Magnetic Separation Techniques in Diagnostic Microbiology 0RJAN OLSVIK,l.2* TANJA POPOVIC,1 EYSTEIN SKJERVE,2 KOFITSYO S. CUDJOE,2

ERIK HORNES,2 JOHN UGELSTAD,3 AND MATHIAS UHLEN4 Foodborne and Diarrheal Diseases Branch, Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia 303331; Norwegian College of Veterinary Medicine, Oslo, 2 and University of Trondheim, Trondheim, 3 Norway; and Royal Institute of Technology, Stockholm, Sweden4 INTRODUCTION-

43 44 44 44 45 IDENTIFICATION AIFTER IMS .................................................. 45 Direct Microscopy ................................................... 46 Cultivation and Enrichment ................................................... 46 Immunoassays .................................................. 46 Hybridization with Nucleic Acid Probes .................................................. 46 IMS as a Template Preparation Method for PCR .................................................. 46 MS OF NUCLEIC ACID .................................................. 46 Magnetic Bead-Based PCR .................................................. 47 Magnetic Beads as Solid Phase for Sequencing Nucleic Acid .................................................. 47 DIAGNOSTIC APPLICATIONS OF MS .................................................. 47 Bacteria .................................................. 47 E. coli ................................................... SalmoneUla spp ....................................................48 49 ShigeUla spp.................................................. 49 L. monocytogenes .................................................. 49 Staphylococcus aureus .................................................. 49 V. parahaemolyticus .................................................. 49 V. cholerae .................................................. 50 Y. enterocolitica ................................................... 50 Clostridium perfringens enterotoxin A .................................................. 50 Chlamydia trachomatis ................................................... 50 R. conorii .................................................. Environmental microorganisms .................................................. 50 50 Viruses .................................................. Enteroviruses .................................................. 50 50 Cytomegalovirus .................................................. 51 HIV-1 .................................................. Infectious pancreatic necrosis virus ................................................... 51 Parasites .................................................. 51 P. falciparum .................................................. 51 51 Schistosoma mansoni .................................................. RELEVANCE TO CLINICAL DIAGNOSTIC MICROBIOLOGY .................................................. 51 REFERENCES .................................................. 52

Immunomagnetic Purification and Concentration .................................................. Magnetic Solid Surfaces ................................................... Attachment of Ligand Molecules to Particle Surface .................................................. Enrichment Beyond Species Level ..................................................

INTRODUCTION

medical applications (69-71). Recently, this technique has also been shown to be suitable for detection of prokaryotic organisms such as bacteria and viruses. Isolation of specific bacteria bound to beads by the antigen-antibody reaction has generally been accomplished by inoculating the bead samples to cultivation broths or onto solid media selective for the target bacteria. Identification can then be accomplished by routine or conventional methods (11, 15, 18, 31, 37, 38, 44, 48, 56, 75). IMS, or immunomagnetic enrichment, is assisted by the fact that bacteria immunologically bound to magnetic beads usually remain viable and can continue to multiply if nutritional requirements are provided (68). The immunomagnetically isolated fraction can then be washed to

Immunomagnetic Purification and Concentration Immunomagnetic separation (IMS), i.e., using small su-

per-paramagnetic particles or beads coated with antibodies against surface antigens of cells, has been shown to be efficient for the isolation of certain eukaryotic cells from fluids such as blood, and this principle has found several * Corresponding author. Present address: Nosocomial Pathogens and Antimicrobial Resistance Branch, Hospital Infections Program MSGO8, Centers for Disease Control and Prevention, Atlanta, GA 30333. Phone: (404) 639-3155. Fax: (404) 639-1381.

43

44

OLSVIK ET AL.

CLIN. MICROBIOL. REV.

Primary antibody Secondary antbody

FIG. 1. IMS of microorganisms. The specific antibody can be attached directly to the magnetic bead (A) or through a secondary antibody, a species-specific precoat of sheep anti-rabbit IgG (B). The specific antibody (e.g., rabbit IgG) can also first react with the target and then be bound to the bead by precoated sheep anti-rabbit IgG (C). remove unspecifically attached organisms before it is placed on suitable growth media. Both polyclonal and monoclonal

antibodies have been employed in IMS (38, 65). These antibodies can be linked to the beads either directly or indirectly, using beads precoated with anti-mouse or antirabbit antibodies (Fig. 1) (38, 46). The IMS technique has several advantages. The target bacteria are separated from the environment and are concentrated from a large volume to a volume suitable for cultivation on plates or in broth (Fig. 2). Furthermore, removal of growth-inhibitory reagents present in the sample enhances the cultivation of beadbound bacteria. The technique is, however, limited by the requirement for antibodies directed against the surface of the target organism and the obligatory high concentration of free antigen in the test sample (51).

Magnetic Solid Surfaces Several magnetic solid phases in particle format are commercially available for magnetic separation (MS) of biological organisms, organelles, or molecules. Common to all of these particles is that specific binding molecules can be attached to them. The most frequently used particles are naturally different forms of immunoglobulins. Most particles are super-paramagnetic; i.e., they are magnetic in a magnetic

Target organism

Attachment of Ligand Molecules to Particle Surface Proteins may bind to hydrophobic surfaces, such as different forms of plastic polymers, and make a monolayer that is resistant to washing, even with mild detergents. This observation led to the development of enzyme-linked immunosorbent assays (ELISAs) and other solid-surface immunoassays. In most cases, it is desirable to have covalent binding between the particle surface and the protein. This is achieved through specific groups at the particle surface which bind to -NH2 or -SH groups on the proteins. Table 1 lists the most common surface groups involved in activation of magnetic particles, the activating agent, and the active derivative that results. Note that particles with epoxy groups on the surface (-CH2-CH-CH2) do not need to be activated for binding protein. Particles activated with toluene sulfonyl are stable for several months in water, but the binding rate of protein is very slow at room temperature (70). Activation with tresyl chloride, on the other hand, gives high protein reactivity but lower stability. A favorable balance between stability and reactivity seems to be obtained by the use of pentafluorbenzene sulfonyl chlorides (69, 70). Synthetic peptides can be coated just as well as whole natural

proteins (9).

*Magnetic parficle

10 ml crude sample

field but are nonmagnetic as soon as the magnetic field is removed. This is important because, once separated by a magnet, particles should not attach to each other through intermagnetic force but should go directly back into suspension. Physical parameters, i.e., the shape and size of the particles, are also important. In order to perform identically in a suspension, with respect to sedimentation and kinetics of binding to other molecules, identical size and form of the particle are preferable. Some commercially available particles consist of treated flakes of magnetic oxides of various sizes and shapes that have a layer of different chemical groups on the surface. In other cases, the particles are prepared by mixing small grains of magnetic oxides with natural or synthetic polymer, followed by procedures to achieve appropriate particle size. Magnetic oxides might also be produced in situ in the polymer phase of the process. Magnetic particles have also been prepared by dispersing magnetic oxides in a mixture of highly water-insoluble compounds and vinyl monomers. From this dispersion is prepared an aqueous dispersion of droplets, containing magnetic oxides, which can be made into magnetic particles by polymerization of the monomer. All of these polydisperse particles are magnetic in a magnetic field but do not maintain the same magnetic force because their sizes are different (35, 61). The monomeric, uniform magnetic spheres most commonly used are made by mixing Fe2" with monosized porous polymer beads. A new polymer surface layer will then close the pores and keep the iron inside the particles (69). Finally, chemical composition of the particle surface is critically important for their successful use in bioseparation (70). An inert surface that does not bind to biological elements other than the specific binding molecule and the target for IMS is naturally desirable. Considerable attention has been given to this problem by the manufacturers of magnetic particles for biological affinity separation.

0.5 ml pure sample

FIG. 2. Outline of the purification and concentration effect and MS (enrichment) of microorganisms from a large crude volume to a purified smaller volume.

Enrichment Beyond Species Level In clinical diagnostic microbiology, it could be vital to differentiate pathogenic or virulent strains from nonpathogens of the same species (33, 48, 55, 79, 85). Selective

VOL. 7, 1994

MAGNETIC SEPARATION TECHNIQUES

TABLE 1. Protein coupling to magnetic particles Functional group(s) at particle surface

Activating

-COOH

R-N=C=R'

-NH2 -CONH2

OHC(CH2)3CHO OHC(CH2)3CHO

-OH -OH

CNBr

-OH

Cl-S-R

o

I

0

Active derivative

reagent

/-COO-C/NHR N+HR' -N=CH(CH2)3CHO -CON=CH(CH2)3CHO

-o /C-NH -O/ 0

1

0SR

o~o

--CH2--CH-CH2 None growth media are usually less efficient for isolation of pathogenic variants when the sample also contains large numbers of nonpathogenic variants of the same species (22). In the case of Eschenchia coli, traditional cultivation of food samples in selective enrichment broths favors E. coli strains of environmental origin over those of human origin (22). Thus, a disease-inducing strain of human origin might not be

45

detected in a food sample. However, IMS provides the potential for selective isolation of strains possessing specific surface epitopes, such as fimbriae, associated with the ability to induce disease (25, 33, 38, 39). Specific 0 antigens found on strains of members of the family Enterobactenaceae have been used as targets for IMS (18, 29, 37, 46, 51). This technology can facilitate epidemiologic investigations by examining samples for specific serotypes of the species associated with the outbreaks of a disease (67).

IDENTIFICATION AFTER IMS Direct Microscopy Antibody-coated beads can agglutinate bacteria in pure culture, forming microscopically visible rosettes or agglutinates (30). Under optimal conditions, one to six bacteria attach to each bead, depending on the size of the bacterium and the bead (54). To demonstrate viability, bacteria attached to the beads can be visualized by acridine orange staining and examined under fluorescence microscopy (38). A second antibody with a fluorescent label directed against epitopes on the target bacteria can also be used to ensure the correct identity of the bacteria bound to the beads (15). Figure 3 shows scanning electron microscopy of Yersinia enterocolitica attached to magnetic beads with rabbit-anti-Y enterocolitica surface antigens as the second antibody. However, the time and workload involved in preparing the

FIG. 3. Scanning electron microscopy of Y enterocolitica serogroup g bound to Dynabeads M450 coated with sheep anti-rabbit IgG, activated with rabbit anti-Y enterocolitica serogroup 9-specific rabbit IgG. The photograph is presented through the courtesy of Emanuel M. Garcia, Animal Disease Research Institute, Nepean, Ontario, Canada.

46

OLSVIK ET AL.

sample make electron microscopy unsuited for diagnostic purposes. Cultivation and Enrichment Bacteria bound by antibodies to the surface of beads are generally viable and will grow under appropriate conditions. Bacteria do not need to be detached from the beads, as attachment apparently has no effect on their growth. Both solid and liquid media have been used for cultivation of several bacterial species immunologically bound to magnetic beads (10, 18, 31, 32, 43, 66, 77). However, enumeration of CFU must take into consideration that each colony is not always the product of a single cell. Several cells might be attached to a cluster of beads to initiate a single colony (51, 66).

Immunoassays Both intact bacteria and their soluble antigenic determinants can be detected after magnetic extraction from the test sample, using a second antibody in a sandwich format. The large surface of the beads enables substantially more solidphase antibodies to be involved in the reaction with antigen than in the ELISAs performed in microtiter plates (12, 19, 45, 56). Magnetic particles prepared from macroreticular polymer particles have a large surface area of 0.5 to 2 cm2I,ug and can be coated with more than 0.3 ,ug of immunoglobulin G (IgG) per cm2 (70). The kinetics of an antigen-antibody reaction is enhanced when the solid-phase-bound element, such as antibody-coated beads, is not fixed. As a result, both the incubation time for reactions with antigens in solutions and consequent, nonspecific binding are reduced. Multiple intensive washing steps, as in the traditional ELISA microdilution plate format, are generally unnecessary (12, 45). Hybridization with Nucleic Acid Probes DNA-based methods provide new tools to the clinical laboratory and improve both the sensitivity and the efficiency of several tests (41, 47, 48, 80). Hybridization techniques that detect either species-specific genes or genes encoding certain virulence factors and their locations on plasmids or on the genome are valuable diagnostic tools. These tests often require cultivation and isolation of the target organism before hybridization, but IMS can be used directly to isolate sufficient numbers of cells. Bacteria on the magnetic beads can be placed directly onto membranes, where the organisms are lysed, and their DNA can be made single stranded, fixed to membrane surfaces, and hybridized with the appropriate probes. The total test time can be reduced from 2 days to 3 to 5 h. The combination of IMS with DNA and RNA hybridization has been proven to be useful and rapid (39, 51). IMS as a Template Preparation Method for PCR The ability of PCR to amplify specific DNA elements drastically reduces the need for the large quantities of test DNA required for hybridization assays. In theory, one copy of the target gene is sufficient for successful amplification. In many ways, the extreme sensitivity of PCR can be compared with cultivation of bacteria on nonselective media, when a single live bacterium can be detected upon initiation of a colony (3, 57). However, certain disadvantages limit the technique for diagnostic use (36). The sample volume tradi-

CLIN. MICROBIOL. REV.

tionally used in PCR ranges from