Analytical Currents: Smaller is better for drug screening - Analytical

Analytical Currents: Smaller is better for drug screening. Anal. Chem. , 2002, 74 (5), pp 119 A–119 A. DOI: 10.1021/ac021953d. Publication Date (Web...
0 downloads 0 Views 8MB Size
news

ANALYTICAL CURRENTS PEGging the surface To form the bioaffinity sensor, some of the PEG chains are end-functionalized with biotin. The researchers then follow classic biotin– avidin or biotin–streptavidin chemistry to attach the antibody, which in this case is goat antirabbit immunoglobulin. Mixing the biotinylated PLL-g-PEG with unmodified molecules in ratios ranging from 0 to 50% varied the antibody density at the surface. Finally, the researchers see little nonspecific binding to the modified metal oxide surface, even in the presence of serum. (Langmuir 2002, 18, 220–230)

Four steps to an immunoassay. (a) The PLL-g-PEG-biotin surface resists nonspecific binding. (b) Absorbing avidin or streptavidin, (c) binding biotinylated antirabbit immunoglobulin, and (d) capturing of the target molecule, rabbit immunoglobulin. (e) OWLS measurement of surface coverage.

(a)

(b)

Biosensor surface

(c)

Biosensor surface

(d)

Biosensor surface

(e)

Adsorbed mass [ng/cm2]

Covering a metal oxide surface that acts as a waveguide with a stable and proteinresistant polymer is the formula Marcus Textor and colleagues at the Swiss Federal Institute of Technology–Zürich use to develop a novel bioaffinity sensor. In this introductory paper, the researchers describe the preparation and properties of niobium oxide surfaces coated with modified poly(L-lysine)-g-poly(ethylene glycol) (PLL-g-PEG). Coating the metal oxide surface is easy because the positively charged PLL-gPEG molecules spontaneously form a monolayer on the negatively charged surface of niobium oxide. Thanks to the optical properties of the metal oxide, surface coverage could be quantitatively determined by optical waveguide lightmode spectroscopy (OWLS).

Biosensor surface

Buffer

400 400

lgG αRlgG-biotin

400

Streptavidin

400

Mixture of PLL-g-PEG and PPB

400 0

50

100 150 Time [min]

200

250

A question of (electronic) taste Researchers have brought us even closer

polymers, one ruthenium complex, and a

water supply tests that monitor metals or

to eliminating natural taste buds for opin-

other contaminants. The natural perception

sulfonated azobenzene polymer—were

ions about taste. An efficient combination

of taste begins with the taste buds, which

chosen because they are good sensors

of sensors that mimics the four basic types

distinguish sweet, salty, sour, and bitter.

and transducers in liquid systems. Each

of taste has been presented by L. H. C. Mat-

Therefore, a good artificial sensor also

material seemed to excel at detecting a

toso and colleagues at the Brazilian Agri-

must have the ability to break down a sig-

particular type of taste. Measurements

cultural Research Corporation–São Carlos,

nal into these four basic components.

made by ac electrodes demonstrated that

Universidade de São Paulo (Brazil), and the University of Wales (United Kingdom). Artificial taste sensors could be used in the food industry for quality control, in

Unlike other strategies for electronic

this new artificial sensor is quick and can

taste sensors, Mattoso and colleagues de-

detect NaCl and sucrose solutions below

posited multiple layers of four types of

the recorded human detection thresholds

Langmuir–Blodgett and self-assembled

of 5 mM. In addition, the device detected

pharmaceuticals to test how much flavor-

ultrathin films onto gold interdigitated elec-

the suppression of a bitter taste by a sweet

ing would make a drug palatable, and in

trodes. The materials—two conducting

one. (Langmuir 2002,18, 239–245) M A R C H 1 , 2 0 0 2 / A N A LY T I C A L C H E M I S T R Y

117 A

news

ANALYTICAL CURRENTS Thinking locally in FTICR MS

Calibration error [mmu]

Richard Smith and colleagues at the Padesired, and that irrevocably causes large FTICR, Smith and his team found syscific Northwest National Laboratory are space-charge-induced frequency shifts, and tematic errors that aren’t accounted for correcting wrongs to make things right. poorer MMA, explain the researchers. by a “global” space-charge calibration Smith’s group proposes a new calibration As they analyzed deviations from the approach. They concluded that these erfunction that’s shown to improve mass commonly used calibration technique in rors and their dependence on charge measurement accuracy (MMA) population and post-excite 3 for all spectra by correcting radii are due to different inMM errors caused by “local” teractions among ion clouds. 2 frequency perturbations in Their corrected calibration FT ion cyclotron resonance function is based on the as1 (ICR) MS. sumption that the spaceOne of the key elements charge-induced frequency 0 in mass spectrometric measshift of each m/z ion cloud is –1 urements for biomolecules is defined by the total ion popgetting a good MMA. Curulation minus its own ion –2 rently, FTICR MS provides charge. The researchers add the best achievable mass acthat although the calibration –3 curacy over a broad m/z function is “somewhat more –4 range. But because of spacedemanding computationally,” 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 charge effects, the achievable it promises to significantly afm/z [u] MMA very much depends on fect many areas of application Calibration errors for the standard calibration function (blue line) the number of ions trapped where MMA is important, and for the corrected calibration function (red line). Summation of in the ICR cell. If one wants such as in proteomics. (J. 10 spectra using SWIFT excitation to a 0.84-cm post-excite cyclotron a large dynamic range, large Am. Soc. Mass Spectrom. radius. (Adapted with permission. Copyright 2002 Elsevier Science.) trapped ion populations are 2002, 13, 99–106)

Luminescence is a gold (cluster) rush!

lF(cps)

Lately, researchers have paid close atten- [Ru(bpy)3]2+ standard, they detected a with 1.8-nm-diam cores when they excittion to luminescence from nanoscopic luminescence of ~770 nm for a 1-µM soed the MPCs at 451 nm. The quantum metal surfaces. Now, Royce Murray and lution of tiopronin-MPC nanoparticles yield was ~0.3%. They verified that the Tao Huang at the University of emission was not the result of North Carolina–Chapel Hill some impurity from the reagents 1.2×106 a have caused some excitement. or synthesis. Instead, the emisThey report highly efficient vission definitely originated from 1.0×106 ible wavelength fluorescence the tiopronin-MPC, according b for four water-soluble, monoto the researchers. 8.0×105 c(3×) layer-protected gold clusters Murray and Huang say the (MPCs) that may offer advanquantum yield is higher than 5 6.0×10 tages in optical device and past observations of the lumibiosensor applications. nescence of gold nanocrystals. 4.0×105 d(200×) Murray and Huang describe They add that the luminescence the visible luminescence of the efficiency varies substantially with 2.0×105 MPCs at ␭max = 700–800 nm, the monolayer ligand around where the monolayer ligands are the gold core and is less for the 0.0 600 650 700 750 800 850 the thiolates of tiopronin (N-2other ligands. Their research Wavelength (nm) mercaptopropionylglycine), also explores the possible de3-mercapto-1-propanesulfonic Emission from (a) tiopronin-MPC solution, (b) mercaptosuccinic pendence of the emission on acid, mercaptosuccinic acid, acid MPC solution, (c) 3-mercapto-1-propanesulfonic acid MPC MPC core size. (J. Phys. Chem. and glutathione. Using a solution, and (d) glutathione MPC solution. B 2001, 105, 12,498–12,502) 118 A

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 2

news

Optical sensor needs its space Sometimes you just need a little space and some good light to work efficiently. Joseph Hupp and colleagues at Northwestern University may have had these ideas in mind when they used microporous, supramolecular coordination compounds to fabricate chemical sensors that operate on the principles of refraction and diffraction. All chemical sensors satisfy two requirements: They recognize the analyte in question and communicate that recognition using an externally observable signal. Hupp and his colleagues devel-

analytes, the refractive index, resulting diffraction pattern, and beam intensity all change. In principle, the photoniclattice-diffraction technique is a chemi-

tein–ligand complexes from unbound or

lecular-weight complexes, which were

300.0 nM

inactive molecules, but because Wabnitz

separated by microcentrifugation and mi-

and Loo simply separate small ligand mol-

crosize-exclusion chromatography. Un-

ecules from the protein–ligand complexes,

bound ligands were recovered from the

60.0 150.0 nM 40.0

0.0 nM

20.0

0

20.0

40.0

60.0

0 µM

Atomic force microscopy image of a micropatterned thin film on glass. (Adapted with permission. Copyright 2002 Wiley-VCH Verlag GmbH.)

cally universal sensing technique, so the researchers are evaluating its application to a range of problems. (Angew. Chem., Int. Ed. 2002, 41, 154–157)

Smaller is better for drug screening Larger is better for some, but not for Paul

and strong inhibitor of PDF, was included

Wabnitz and Joseph Loo of Pfizer and their

to act as a reference for the other ligands.

picomole-sensitive drug screening method. Typical screening methods separate pro-

When small ligand molecules strongly inhibited Co-PDF, the result was high-mo-

their assay is faster and uses less sample.

column, further separated by capillary LC,

To demonstrate their technique, the

and analyzed by MS, thus identifying lig-

researchers chose a set of active and in-

ands with little or no inhibiting effect on

active inhibitors of cobalt(II)-peptide de-

PDF. The team also determined the rela-

formylase (Co-PDF), a substitute for the

tive order for the binding ability of the

less stable native Fe-PDF that cleaves the

ligands with Co-PDF and validated their

formyl groups from polypeptides in eubac-

results with control experiments. (Rapid

teria. Actinonin, an antibacterial agent

Commun. Mass Spectrom. 2002,16, 85–91)

Microsize-exclusion chromatography

oped a thin-film sensor that eliminates some features that other sensors need, consequently simplifying some aspects of chemosensory materials. The thin films have cavities within the nanoscale building blocks, which were prepared in one step from ReCl(CO)5 and a difunctional zinc porphyrin ligand by directed assembly. The cavity-containing materials were micropatterned as lattice structures via a “soft” lithography technique. Light diffraction by the lattices is the physical basis for the optical sensor. The sensor has a particular refractive index when it is empty, and because of its estimated 50% void volume, it creates a particular diffraction pattern and beam intensity when a helium diode laser passes through it. When the sensor contains

Incubation Microcentrifugation Protein + drug library

MicroSpin G-25 column

Elute Extract protein/ unbound drugs protein + drug complex to waste Capillary LC/MS analysis

The microsize-exclusion chromatography–capillary LC/MS technique for ligand drug screening. (Adapted with permission. Copyright 2002 John Wiley & Sons, Ltd.)

M A R C H 1 , 2 0 0 2 / A N A LY T I C A L C H E M I S T R Y

119 A

news

ANALYTICAL CURRENTS Assay resonates with intact cells

Antibodies choose the “right” catalysts If high-throughput screening is like racing

ed version of the product, and that infor-

in a Formula One car, then screening enan-

mation is used to determine enantiomeric

tioselective reactions can be like poking

excess. In both cases, the binding of the

along in an old jalopy. But Alain Wagner,

reaction product displaces a prod-

Charles Mioskowski, and colleagues at

uct–enzyme conjugate from the antibody

Université Louis Pasteur and Commissariat

binding sites. This displacement interrupts

à l’Energie Atomique (both in France) have

a colorimetric reaction, and the resulting

a way to “turbo-charge” enantioselective

decrease in the absorbance signal is re-

screening. And rather than requiring rela-

lated to the concentration of the reaction

tively expensive equipment, their method

product. Using the enantioselective reduction

relies on the cheap, simple immunoassay.

of benzoyl formic acid into mandelic acid

Actually, the researchers use two immunoassays. In one assay, an immobilized

as a model reaction, the researchers

antibody binds both enantiomers of the

screened a catalyst library of 88 candi-

reaction product, thus measuring the total

dates, running both immunoassays simul-

concentration. In the other assay, an enan-

taneously for a total of 176 reactions, and

tioselective antibody measures the con-

validated the method by testing 42 of the

centration of either the right- or left-hand-

samples using HPLC. (Angew. Chem., Int.

A complication of some cell-based protein assays is that the cells aren’t always intact, which means that the proteins may not be in their native environment any longer. So Karolina Lundin, Christer Lindqvist, and colleagues at Åbo Akademi University and Wallac Oy (both in Finland) developed an assay that uses time-resolved fluorescence resonance energy transfer (FRET) to detect proteins on intact cells. 340 nm Eu3+

FRET 665 nm Cy5

Yield determination E

E

Wash

Transfer

Step 2: binding competition

Yellow color

E

Step 3: staining

= BF = MA

Cat E Step 1: catalysis

= Anti-racemate antibody

= Enzyme–product conjugate

= Enantiospecific antibody ee determination

E

E

Wash

Yellow color

E

Transfer Step 2: binding competition

Step 3: staining

Schematic depicting the use of an enzyme immunoassay for the high-throughput screening of enantioselective catalysts. (Adapted with permission. Copyright 2002 Wiley-VCH Verlag GmbH.) 120 A

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 2

The fluorescence resonance energy transfer system. When the Cy5-labeled antibody binds to the receptor, energy is transferred from the Eu3+ donor, resulting in a signal. (Adapted with permission. Copyright 2001 Academic Press.)

FRET assays are often used to identify the interaction of two molecules. One molecule is labeled with a fluorescence acceptor, which is excited only when a molecule—usually a binding partner— bearing a fluorescence donor is nearby. In general, the energy transfer from the donor to the acceptor works well only over very small distances, although some donor–acceptor pairs, such as Eu3+–Cy 5, are good over relatively long distances. In the new assay, the cells are infected with a virus that bears a receptor of interest—the interleukin-2 receptor, in the test case. The receptor is then expressed on the cells’ surfaces. In the meantime, the surfaces of the cells are

news

biotinylated, and Eu3+ donor molecules linked to streptavidin are added; the resulting biotin–streptavidin binding traps the fluorescent donor molecules. When an antibody labeled with a Cy5 acceptor molecule binds to the receptor, the energy transfer to the acceptor yields a signal, which can be measured in a timeresolved manner.

The maximum S/N ratio obtained was 4.1. The researchers determined that too much biotin labeling can interfere with the binding of the antibody to the receptor and that the presence of free biotin can inhibit the energy transfer. The assay’s specificity was verified in a competitive assay with unlabeled antibody. (Anal. Biochem. 2001, 299, 92–97)

Tracing the pattern of aminochromes Catecholamines may be the culprits, but aminochromes are accessories after the fact, and now a research group has developed an analytical tool to find the evidence. Francisco Lemos-Amado and colleagues at the University of Aveiro, the Universidade do Porto, and the Instituto Superior de Ciências da Saúde (all in Portugal) establish a general pattern for the collision-induced-decay (CID) fragmentation mechanism of protonated aminochromes. Both types of compounds are implicated in pathological oxidative toxic effects. High levels of catecholamines

cause cardiotoxicity, and oxidation of the catecholamine dopamine seems to contribute to Parkinson’s disease. Oxidative stress results from the first of two oxidation reactions on a catecholamine to produce aminochrome end products, 2,3-dihydroindole-5,6diones. Aminochromes, on the other hand, oxidize other compounds to cause toxic repercussions both in vivo and in vitro. Adrenochrome can induce cell damage, disrupt metabolic pathways, and cause heart arrhythmia and even sudden death. Because little is known about the oxidation mechanism of catecholaR R mines, the researchers focused on O O aminochromes that they synthesized C+ from five catecholamines. They found N+ N+ HO HO that different substituents resulted in ′′R ′′R significantly different electrospray ionization tandem MS (ESI-MS/MS) –CO spectra. Isoprenochrome and dopaR chrome lose propene and formic acid, R H H respectively. These aminochromes folC+ O H O C+ R′ low the same general fragmentation N N pattern as the other three in the study—the consistent, consecutive ′′R ′′R loss of two carbonyl groups. –CO The team also used a commercial adrenochrome, separately from the exR perimentally generated adrenochrome, H3C for additional CID experiments to R′ clarify the fragmentation mechanism. N +HC They conclude that their HPLC/ESIR′′ MS/MS procedure could be useful in metabolite screening and for oxidation General mechanism for the fragmentation of studies of catecholamines in vivo or in protonated aminochromes from ESI-MS/MS. vitro. (Rapid Commun. Mass Spec(Adapted with permission. Copyright 2002 John trom. 2002, 15, 2466–2471) Wiley & Sons, Ltd.) M A R C H 1 , 2 0 0 2 / A N A LY T I C A L C H E M I S T R Y

121 A

news

RESEARCH PROFILES Dry proteins may quench thirst for microarrays To compare the results with traditional the substrate’s surface. Compared with the torrent of DNA miThe only other parallel microarray fab- enzyme-linked immunoassays (ELISAs), croarrays, protein microarrays have been the researchers conducted “threshold” imrication technique is microcontact printlimited to a trickle. But a method develmunoassays, in which the plasma sample ing. However, the ESD mask method oped by Victor Morozov and colleagues uses proteins more economically—requir- was diluted serially (1:300 to 1:300,000, at the Russian Academy of Sciences and in this case), and the threshold at which ing only 1 µL of sample for ~100,000 the National Hematology Research Centhe signal could be differentiated from spots—and allows easier control of the ter (Russia) might open the floodgates. the background was determined using a spot size, Morozov says. In this issue of Analytical Chemistry (pp fluorescent microscope. In 88% of 927–933), the researchers describe (a) (b) the cases, the microarray and ELISA the parallel fabrication of 1200 proAP HRP LaP Coc AP AP results agreed qualitatively. In 80% tein microarrays, which are then BSA GSA HB1 HB2 MDp MFa of the cases, the sensitivities of both used for immunoassays. methods were comparable; in the Whether “spotting” with pins, βGa mhCon Pap Ova IgG IgG A remaining examples, the microarray stamping with microcontact printTrI a-m a-h Avi Trs Mbw results differed by 10- to 1000-fold. ers, or spraying from piezoelectric IgG IgE Comparisons with the ELISA (inkjet) dispensers, keeping the HRP Mbh HE HK Ure AP WL data also provided preliminary eviproteins in a microarray active can AP HRP AP AcE ADH HRP dence that the microarray results be a challenge. Even proteins that (c) (d) could be quantified. If this is upheld are active after deposition may not by further testing, it could eliminate survive storage unless they’re kept the need for serial dilutions. cold or dry. In a few instances, the reMorozov and colleagues took searchers noted spots for proteins the dry route, using electrospray other than those with which the deposition (ESD) in a low-humidimice were immunized. Sometimes, ty environment to dehydrate prothese hits—which also appeared teins as they were deposited. “Durin the ELISAs—were attributed to ing electrospraying, drying happens cross-reactivity, but the the patterns in a few milliseconds, before the were so reproducible that they (a) Locations of the antigens on a protein microarray. (b) microdroplets reach the surface,” could be used to identify which says Morozov. In many cases, “The Results from a mouse immunized against concavalin A, mouse provided the sample. trypsin inhibitor, and sperm whale myoglobin. (c–d) Diluprotein literally doesn’t have time Morozov suggests that these imtions of plasma from a mouse immunized against a mixto unfold.” munological “fingerprints” might be ture of 10 proteins. ESD also has been used for more than a curiosity. For example, protein microarrays by Dutch rethey might indicate that the mouse searchers, who sprayed proteins from had become immunized against this proThe researchers conducted a series of capillaries positioned close to the substrate immunoassays using 6 ⫻ 6 arrays (spot di- tein (or one like it) through a previous ill(Anal. Chem. 2001, 73, 307 A). Moroness. In the future, such fingerprinting ameter = 30–40 µm) of 28 proteins that zov and colleagues used a mask—a mica might be applied to people, he adds, to served as antigens. The substrate bearing plate with 1200 conical holes drilled in it—instead of capillaries. Then they placed all 1200 arrays was cut into 3-mm squares reveal “a history of illnesses or the genetic construction of [a person’s] antibody of 4 arrays each. The squares were glued a prepared aluminum substrate on a movmachinery—even without knowledge of to special holders, and one square was put able stage beneath the mask, loaded the a particular sensitivity.” into a microplate well. Thus, a full assay ES capillary with a protein sample mixed These findings add up to greater conwas performed in each well. with sucrose, and sprayed, laying down venience, says Morozov. The ability to Although some experiments were one spot in each of 1200 arrays simultanemake thousands of microarrays in parously. After repeating the process for all of conducted with human plasma, most used mouse plasma, which is safer to allel, to multiplex assays in a microwell the proteins, the researchers immobilized handle. To approximate the variation plate, and to quantify the results would the dry proteins by treating the substrate found in human plasma, the mice had give protein microarrays a big boost. Get to form covalent links between the amino been immunized with up to 10 proready; there just might be another flood. a groups of the proteins and the carbonyl —Elizabeth Zubritsky teins, Morozov says. groups of oxidized dextran molecules on

122 A

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 2

news

Protein plays matchmaker for QD–antibody conjugation For a team of researchers at the Naval Research Laboratory (NRL) in Washington, D.C., conjugating luminescent quantum dots (QDs) with antibodies became an adventure into perseverance and exciting possibilities. In the summer of 1999, the group experimented with common covalent chemistry to form QD–antibody bioconjugates but ended up with useless aggregated dots. It was like “running into the brick wall,” recalls George Anderson, one of several researchers on the project. “We tried all the conventional covalent techniques over and over and were just kind of frustrated day after day with that.” But the NRL researchers began collaborating with other scientists and came up with an engineered adaptor protein to conjugate the QDs with antibodies. It was a perfect fit. The result is a new kind of fluoroimmunoassay that has the potential to be a powerful tool in medical diagnostics and biowarfare detection, say Ellen Goldman and Hedi Mattoussi. When they realized the fruits of their labor, the research “went from frustrating to fun very quickly,” says Anderson. In the Feb. 15 issue of Analytical Chemistry (pp 841–847), Anderson, Goldman, Mattoussi, and colleagues describe how they attached the immunoglobulin G (IgG)-binding ß2 domain of a modified streptococcal protein G (PG) to the highly luminescent semiconductor CdSe–ZnS core-shell QDs via electrostatic self-assembly. The product was a QD–antibody conjugate (QD–IgG) with an emission maximum of 570 nm. The electrostatic interaction arises from negatively charged dihydrolipoic acid (DHLA)-capped QDs capturing positively charged basic leucine zipper domains appended to the C-termini of PGs. The researchers also developed a purification tool using the maltose binding protein’s basic leucine zipper (MBPzb) on the mixed-surface QDs to remove unbound antibody from the QD–IgG product via affinity chromatography. The scientists can vary the number of anti-

bodies conjugated to a single QD by simply changing the ratio of the molecular adaptor protein, the PG basic leucine zipper (PG-zb), and the MBP-zb that are incubated with the QDs. They used the QD–IgG conjugates in direct and sandwich fluoroimmunoassays to successfully detect the protein

QD

PG-zb molecular adaptor protein

MBP-zb purification tool

Immunoglobulin G (IgG)

Cartoon of a mixed-surface composition QD–antibody conjugate.

toxin staphylococcal enterotoxin B and the small-molecule explosive 2,4,6-trinitrotoluene. The researchers say their method is different from previously reported conjugation techniques, such as those using avidin–biotin technology or covalent cross-linking. People still use the covalent bond approach, but aggregation remains an issue with that technique, says Mattoussi. In the NRL strategy, however, which involved help from Moungi Bawendi’s group at the Massachusetts Institute of Technology, aggregate-free conjugates worked well in fluoroimmunoassays. “We were really surprised,” says Anderson. “One of the first observations was when we mixed the proteins with the dots, there was a great jump in luminescence output.” The most trying moments of the project involved purifying the molecular adaptor protein and making the mixed-

surface QDs. While Mattoussi and Anderson were examining the chemistry involved in attaching the antibodies to QDs, Goldman and J. Matthew Mauro had been working on chimeric recombinant proteins with combinations of functional domains, which eventually led to using an adaptor protein for the conjugation. In the beginning, they had a lot of nucleic acid contamination in their adaptor protein preparations. They had to rethink their strategy and eventually solved that issue by using a denaturing preparation, says Goldman. Another hurdle was making the QDs themselves. They had to be water-soluble and compatible for making bioconjugates, says Mattoussi. The mixed-surface QDs were prepared by incubating DHLA-capped QDs with various molar ratios of purified PG-zb dimer mixed with purified E. coli MBP containing the C-terminal peptide linker and the positive leucine zipper. The researchers point out that luminescent QDs as inorganic fluorophores have the potential to circumvent some of the functional limitations encountered by organic dyes in biotechnological applications. They are resistant to photobleaching and can be excited over a continuous range of wavelengths. The researchers are working on optimizing immunoassay sensitivity using different colored dots to simultaneously look for multiple toxins or small molecules in the same sample, say the researchers. But they would like to make the QDs more stable at lower pHs, which would make them more useful for environmental analysis. Currently, the QDs only work at pH >7. They’d also like to change the type of surface charge and improve the IR imaging capability of the QDs to make them ideal for use with tissue, says Mattoussi. For the researchers, the future is bright and hopeful. “This is science, so things change,” adds Mattoussi. “Maybe in two years, there will be some other ideas that we never thought of.” a —Cheryl M. Harris

M A R C H 1 , 2 0 0 2 / A N A LY T I C A L C H E M I S T R Y

123 A

news

RESEARCH PROFILES Faster analysis of PBDEs and PCBs Even though polybrominated diethyl ethers (PBDEs) prevent fires, they appear responsible for some serious environmental and health concerns. In the Feb. 15 issue of Analytical Chemistry (pp 790–798), Adrian Covaci and his colleagues at the University of Antwerp (Belgium), the Netherlands Institute for Fisheries Research, and Health Canada report a method to quantify PBDEs in human adipose (fat) tissue samples that cut analysis times by at least 50% while maintaining resolution. The method is also applicable to complex mixtures, including polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs). PCBs are toxic industrial pollutants that are well known to accumulate in the body and the environment. Over

Abundance

Abundance

Abundance

the past 20 years, PBDEs have been used extensively as flame retardants. So far, only nanograms of PBDEs are found per gram of lipid in the human body—a much lower concentration than PCBs— but PBDEs have begun to appear more frequently in environmental and human samples. For example, the PBDE levels in a Swedish human breast milk study conducted over the past 30 years doubled every 5 years until 1997, when they leveled off. Recent toxicology data suggest that some PBDEs appear to be endocrine-disrupting chemicals. Some countries have recently started monitoring PBDEs in fish and humans. In 2000, Covaci’s collaborators Jacob de Boer and John Jake Ryan participated in a study of PBDE levels in fish and sediments with 16 other labs in Europe and North America in the (a) first interlaboratory test pro13C-BDE 47 4000 gram. In the current work, Covaci, de Boer, Ryan, and 13C-BDE 99 13C-BDE 153 their colleagues validated 2000 their method and report data from Belgium’s first study of PBDEs in humans. 0 In this human study, ab6.00 7.00 8.00 9.00 10.00 Time (min) (b) dominal adipose cells were 28 47 removed postmortem, ex77 99 6000 75 tracted with a hot Soxhlet technique for ~2 h, and 4000 71 66 119 cleaned up with a dual sys154 100 153 2000 tem of solid-phase extraction 138 85 cartridges. For PCBs and 0 OCPs, the resulting extract 6.00 7.00 8.00 9.00 10.00 is concentrated enough to Time (min) (c) analyze 1 µL by GC. Large8000 volume injection (LVI) re47 quires that the low-level 6000 PBDE samples are first con153 4000 centrated. The researchers 99 used a type of LVI that can 100 2000 28 accept up to 100 µL of sam0 ple introduced in 5-µL incre6.00 7.00 8.00 9.00 10.00 Time (min) ments into an empty liner where solvent is eliminated EILR-MS chromatograms of brominated diethyl ethers via gas flow through a split (BDEs). (a) Standard mixture of 13C BDEs. (b) BDE stanline. This technique “is exceldards. (c) BDE congeners identified from human adipose lent for measuring both types tissue. 124 A

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 2

of chemicals in the same extract” but is limited by the complex matrix presented by human samples, according to Covaci. Using a narrow-bore (0.1-mm-i.d.) GC column reduced analysis time by 50% without sacrificing resolution, says Covaci. Moreover, tri- to hexacongeners of PBDE eluted in 6.0–9.8 min. Results show levels of certain PBDE congeners in the human tissue are 2.2–11.7 ng/g of lipid weight, which compares well with other European results. When PCB congeners were analyzed by this approach, retention times dropped by 85%, from 50 to 8 min. Electron impact low-resolution MS (EI-LRMS) was used to identify and quantify common congeners. According to Covaci, a classically designed LR mass spectrometer with splitless injection and a normal column i.d. (0.25 mm) has a detection limit of 1–2 ng/g lipid for PBDE analysis in human adipose samples. Using LVI and a narrowbore capillary column may reduce the detection limit to 0.1–0.3 ng/g lipid in adipose tissue, he adds. But why use LRMS when high-resolution MS (HRMS) is available? Because, Covaci says, HRMS instruments can cost up to 5 times more and aren’t easily available. He says, “One would like to set up viable methods with the instruments existing in his lab.” LRMS can be run on a bench-top machine, which is easier to maintain and has greater versatility than a high-resolution mass spectrometer, he adds. Although EI isn’t sensitive enough to determine low concentrations of highly brominated diphenyl ethers, such as the hepta- to decacongeners, these congeners have not been reported in human samples to date. At this stage, the method isn’t sensitive enough for serum samples. “We would need some 20 mL of serum—or 10 mL of milk—per person in order to have measurable peaks,” says Covaci. However, they are currently refining the method for human milk, blood, and other matrixes. a —Rachel Petkewich

news

PEOPLE 2002 ACS Awards Several scientists in the analytical chemistry community will receive the 2002 American Chemical Society awards at the 223rd National Meeting in Orlando, Fla., in April. Alan G. Marshall, professor at Florida State University, will receive the ACS Award in Analytical Chemistry, sponsored by Fisher Scientific. The award is given in recognition of outstanding contributions to pure or applied analytical chemistry. Marshall is known for his work in the analytical applications of FT ion cyclotron resonance ultrahighresolution MS.

biological processes to aid in, for example, the elucidation of posttranslational modifications. Takeshi Oka, professor at the University of Chicago, will receive the E. Bright Wilson Award in Spectroscopy, sponsored by Rohm and Haas. The award recognizes fundamental contributions in all fields of spectroscopy. Oka is best known for his discovery of H +3, the most abundant molecular ion in the universe and the initiator of interstellar chemistry.

Edward Cussler, professor at the University of Minnesota, will receive the ACS Award in Separations Science and Technology, sponsored by IBC Advanced Technologies and Millipore. The award recognizes outstanding accomplishments in the field of fundamental or applied separation science and technology. Cussler is known for his work with membranes for facilitated transport, temperature-sensitive gels, hollow-fiber extractions, and an artificial gill for keeping a dog alive underwater.

Ed Yeung, professor at Iowa State University, will receive the ACS Award in Chromatography, sponsored by Supelco. The award recognizes specific achievements in the field of chromatography. Yeung is known for his many contributions to our fundamental understanding of chromatography and the development of novel instrumentation, including single-molecule adsorption interactions and multiple capillary systems for DNA sequencing and high-throughput screening. Brian Chait, Camille and Henry Dreyfus Professor at the Rockefeller University, will receive the Frank H. Field and Joe L. Franklin Award for Outstanding Achievement in Mass Spectrometry, sponsored by Bruker Daltonics. Chait is known for his work in the development of mass spectrometric tools for studying

M A R C H 1 , 2 0 0 2 / A N A LY T I C A L C H E M I S T R Y

125 A