D. Wilson¹, E. Mortberg², S. Rubertsson², H. Zetterberg³, K. Blennow³, L. Song¹, L. Chang¹, G. Provuncher¹, P. Patel¹, E. Ferrell¹, D. Fournier¹, C. Kan¹, T. Campbell¹, A. Rivnak¹, B. Pink¹, K. Minnehan¹, T. Piech¹, D. Rissin¹, D. Duffy¹. ¹Quanterix Corporation, Cambridge, MA; ²Department of Surgical Sciences, Anaesthesia and Intensive Care, Uppsala University, Uppsala, Sweden; ³Department of Psychiatry and Neurochemistry, the Sahlgrenska Academy at the University of Gothenburg, Molndal, Sweden.

Background:
Amyloid beta (Aβ) peptides are proteolytic products from amyloid precursor protein (APP). Accumulation of Aβ in the form of extracellular plaques is a hallmark of Alzheimer’s disease (AD). Oxidative stress can increase the production of Aβ, which may potentiate AD onset and vascular dementia. It is believed hypoxic insults trigger Aβ production by activating APP proteolysis via a pathway involving β-secretase (BACE1) in which BACE1 is upregulated by the transcription factor hypoxia–inducible factor–1 (HIF–1). Animal models have helped elucidate this pathway, but a direct link between hypoxia and Aβ42 production in the human brain has not been established. We employed a new technology (Single Molecule Array, SiMoA) capable of ultrasensitive protein measurements to look for changes in serum Aβ42 in patients following cardiac arrest and resuscitation.

Methods:
26 unconscious patients with cardiac arrest were resuscitated with restoration of spontaneous circulation (ROSC). Serial blood samples were collected within 6h after cardiac arrest, and continued at intervals from 1–108h. Inclusion criteria included age, systolic BP >80mmHg after ROSC, and a Glasgow Coma Scale ≤7. Serum aliquots were frozen until assay. Samples were measured in triplicate by SiMoA Aβ42 assay, which has a limit of detection of less than 0.04 pg/mL.

Results:
Nearly all patients exhibited a significant time–dependent elevation of Aβ42. After a lag period of 10 or more hours, very clear Aβ42 elevations were observed in most cases. Elevations ranged from approximately 50% to over 30–fold, with most elevations in the range of 3–10–fold (average approximately 7–fold).

Conclusion:
These data are the first to directly link hypoxic stress to Aβ elevation in humans. The kinetic profiles may be related to time–dependence of the HIF–1/BACE1 upregulation pathway leading to a hypoxia–induced amyloid cascade. The relevance of mild chronic ischemia with upregulation of the amyloid cascade in AD pathogenesis remains to be examined.

David Wilson¹, David Duffy¹, David Rissin¹, Cheuk Kan¹, Todd Campbell¹, Stuart Howes¹, David Fournier¹, Tomasz Piech¹, Gail Provuncher¹, Purvish Patel¹, Evan Ferrell¹, Andrew Rivnak¹, Jeff Randall¹, Linan Song¹, David Walt². ¹Quanterix Corporation, Cambridge, MA, ²Tufts University, Boston, MA.

Purpose:
Develop a method for measuring prostate specific antigen (PSA) in the serum of patients who have undergone radical prostatectomy (RP). The method is based on Single Molecule Array (SiMoA) technology in which individual molecules of analyte are trapped in femtoliter microwells and counted. Measurement of PSA in post RP patients 4–6 weeks after surgery (nadir PSA levels) has potential prognostic value.

Experimental:
The assay starts as a sandwich ELISA except that 2.7 µm paramagnetic beads serve as the solid phase rather than the ELISA plate. The assay is initiated by mixing sample with beads coated with anti–PSA antibody, and the mixture is incubated for two hours. During the incubation, PSA molecules are captured with an excess of beads, such that the ratio of bound PSA per bead is much less than one. Following a wash in which beads are separated using a magnet, biotinylated detector antibody is incubated with the beads for 45 minutes. After a second wash, a conjugate of streptavidin–ß–galactosidase is incubated with the beads for 30 minutes to form the final enzyme immunocomplex. Upon completion, the beads are loaded onto the ends of bundles of optical fibers etched with an array of 50,000 microwells, each with a volume of 50 femtoliters (4.5 µm wide x 3.25 µm deep). Single beads are introduced into each well via a 10–minute centrifugation at 1,300 g. Excess beads are removed, and the bundles are sealed against a solution of resorufin ß–D–galactopyranoside (RDG). Wells containing an enzyme immunocomplex convert RDG to a fluorescent product, which becomes concentrated in the small volume of the wells. Concentration of the fluorescent product permits ready imaging by a CCD camera and discrimination from wells that contain no enzyme immunocomplex. Poisson statistics predict that each well will contain either one PSA molecule or no PSA molecules when the ratio of bound PSA per bead is much less than one. This results in digitized quantification. Raw signal is recorded as “% active wells”, which can be converted to “average enzymes/bead” to correct for non–Poisson behavior at higher PSA concentrations. The output is related to a standard curve and converted to a PSA concentration of the sample.

Results:
In preliminary validation studies across multiple runs and reagent batches, the AccuPSA assay exhibited a limit of detection (LOD) of 0.01 pg/mL, and a limit of quantification of less than 0.05 pg/mL. Linearity was demonstrated across a calibration range from 0 to 100 pg/mL, giving a five–log assay range relative to the LOD. The assay exhibited good agreement with a commercially available PSA test. PSA levels were measured in 60 post–RP specimens selected as being below the detection limit of a commercially available chemiluminescent immunoassay. All specimens gave measurable PSA results in the SiMoA PSA assay, with results ranging down to 0.014 pg/mL (0.4 fM).

Conclusion:
The results suggest that PSA can be reliably measured in post RP patients with SiMoA technology. Recent results by others suggest measurement of nadir PSA following surgery may have prognostic value for recurrence–free survival. Measurement of PSA in post RP patients with AccuPSA could affirm a good prognosis, reduce unnecessary adjuvant radiation, and enable earlier detection of recurrence for earlier, more effective treatment.

D.C. Duffy¹, D. M. Rissin¹, C. W. Kan¹, T. G. Campbell¹, S. C. Howes¹, D. R. Fournier¹, L. Song¹, T. Piech¹, P. P. Patel¹, L. Chang¹, A. J. Rivnak¹, E. P. Ferrell¹, J. D. Randall¹, G. K. Provuncher¹, D. R. Walt². ¹Quanterix Corporation, Cambridge, MA, ²Tufts University, Medford, MA.

Objective:
The aim of this study was to detect prostate specific antigen (PSA) in the serum of patients who had undergone radical prostatectomy (RP). To achieve this objective, an ultra–sensitive ELISA for PSA based on single molecule detection was developed and evaluated.

Methodology:
We have developed a method for detecting single immunocomplexes formed in the enzyme–linked immunosorbent assay (ELISA) using single molecule arrays (SiMoA). This digital ELISA method is based on isolating single immunocomplexes labeled with an enzyme in arrays of femtoliter wells, sealing the arrays in the presence of the enzyme substrate, and fluorescently imaging the array. Fluorescent product molecules of the enzyme–substrate reaction are confined in the femtoliter volume, giving rise to a local high concentration that can be easily detected using a standard fluorescent microscope. By using high density arrays of femtoliter wells, hundreds to thousands of single immunocomplexes can be detected simultaneously. Isolation of single immunocomplexes using SiMoA gives rise to a dramatic increase in sensitivity over bulk, ensemble detection methods. An ultra–sensitive digital ELISA for detecting PSA was developed that has a limit of detection (LOD) of 6 fg/mL (200 aM) in serum. This assay was used to measure PSA in the sera of thirty RP patients.

Okrongly, D. Quanterix Corporation, Cambridge, MA

Quanterix was founded in 2007 based on single molecule detection of enzyme labels and fits well with the theme of the 2010 ISE Meeting: New Roles for Old Molecules: Enzymes in Personalized Medicine. The ability to observe the behavior of single enzyme molecules was first described by Rotman in 1961, using β–D–galactosidase to cleave a fluorogenic substrate dispersed as microdroplets in silicone oil. This provocative finding was followed up over the years by others using various detection systems, and has led to a better understanding of the heterogeneity of enzymatic activity displayed at the single molecule level. In 2005, Rissin and Walt demonstrated the analytical potential of single enzyme molecule systems. They demonstrated that single molecules of streptavidin covalently coupled to β–D–galactosidase (SBG) could be dispersed into an optical array composed of 24,000 individual femtoliter reaction chambers coated with biotin. After sealing the chambers with a fluorogenic substrate solution trapped inside, a digital readout of the array was obtained by counting the number of fluorescent reaction vessels, indicating that the well had captured an SBG molecule. The number of fluorescent wells was demonstrated to be proportional to the concentration of SBG present in the original solution and was consistent with a Poisson distribution at low concentrations where single molecules were in each well and Gaussian distribution at higher concentrations where more than one enzyme molecule was in each vessel. The sensitivity was an astounding 2.6 aM (2.6 x 10^–18 M). This observation paved the way for more complex systems to be developed, where a capture antibody–ligand–detector antibody complex is sequentially formed in the vessel, and the detector antibody (coupled with biotin), is labeled by the SBG. Quanterix is utilizing this technology to develop high sensitivity assays for the clinical diagnostics market. Data for several representative assays with over 1000x more sensitivity than standard immunoassays will be presented, along with the implications for personalized medicine.

Rissin DM¹, Kan CW¹, Campbell TG¹, Howes SC¹, Fournier DR¹, Song L¹, Piech T¹, Patel PP¹, Chang L¹, Rivnak AJ¹, Ferrell EP¹, Randall JD¹, Provuncher GK¹, Walt DR², Duffy DC¹. ¹Quanterix Corporation, Cambridge, MA, and ²Tufts University, Medford, MA.

We describe a method for detecting single immunocomplexes formed in the enzyme–linked immunosorbent assay (ELISA); we call this method digital ELISA. This method is based on isolating single immunocomplexes labeled with an enzyme in arrays of femtoliter wells, sealing the arrays in the presence of the enzyme substrate, and fluorescently imaging the array. Fluorescent product molecules of the enzyme–substrate reaction are confined in the femtoliter volume, giving rise to a local high concentration that can be easily detected using a standard fluorescent microscope. By using high density arrays of femtoliter wells, hundreds to thousands of single immunocomplexes can be detected simultaneously. Isolation of single immunocomplexes in this way gives rise to a dramatic increase in sensitivity to enzyme labels over bulk, ensemble detection methods. This method was used to detect yoctomole levels of β–galactosidase enzyme label. The enzyme sensitivity using these single molecule arrays – which we term SiMoA – is ~10⁵ times greater than the same enzyme measured on a fluorescent plate reader and over 100 times greater than detection of alkaline phosphatase using chemiluminescence, the gold–standard of sensitive ELISA detection. Ultra–sensitive immunoassays for detecting clinically–relevant proteins have been developed using digital ELISAs that can detect sub–femtomolar concentrations of the proteins in serum. These assays have been used to detect proteins in clinical samples at concentrations that are well below the detection limits of current immunoanalyzers. With the aim of detecting very low concentrations of nucleic acids without recourse to PCR (or other target amplification), a single molecule assay for DNA was also developed. Attomolar limits of detection for direct detection of single molecules of DNA in a sandwich assay were achieved. This single molecule detection technology couples directly to the back–end of established clinical immunoassays providing a two to four log improvement over the detection limit of current clinical methods. The method overcomes the complexity associated with existing ultra–sensitive protein detection methods, bringing single molecule sensitivity to clinical immunodiagnostics that should lead to downstream benefits in patient care.