The optics module houses two solid state lasers (488 and 639 nm excitation) and a spectrophotometer with a charge-coupled device (CCD) camera for simultaneous laser-induced fluorescence detection of two dye colors in each capillary. The emission photons are detected by an ∼3 × ∼25 mm CCD camera (Hamamatsu, Bridgewater, NJ, USA) that is thermoelectrically cooled.
Depicted in Supplementary Figure 2 is the CE cartridge, which is replaceable and houses the capillary array. An integral chip included in the cartridge body collects data on capillary configuration, serial number, expiration date, and usage. After each CE cycle, the capillary array is automatically purged of used gel and refilled with fresh gel. The cartridge contains 48 (and up to 96) electrodes that complete the electrical circuit. The capillaries periodically contact the PCR enabling the electrokinetic injection from the sample, followed by CE separation of the amplicons. Within the cartridge, the capillaries are arranged into an optical window where the laser impinges for excitation of the fluorescently labeled amplicons. The window is designed to eliminate capillary-to-capillary crosstalk by incorporating opaque barriers between the capillaries.
PCR assays and real-time CE
To demonstrate the capability of ICEPlex, an in-house prototype multiplex assay, the ViraQuant assay, which simultaneously detects and quantifies viral loads of cytomegalovirus (CMV), Epstein-Barr virus (EBV), BK virus (BKV), and human herpes viruses-6B (HHV6) and -7 (HHV7) along with an extraction control, was developed. For EBV detection, a dual approach was followed to allow accurate virus particle quantification while maximizing detection sensitivity. For this purpose, two different genomic regions were targeted: one primer pair, designed for virus particle quantification and designated EBVq, amplifies an unique region of the EBV genome and a second primer pair, designed for sensitivity and designated EBVs, targets a genomic repetitive DNA element with a strain-specific copy number variation of 5–11 copies/genome (10). Amplicon sizes and primer sequences for the various targets are shown in Table 1. Primers for the ViraQuant assay were designed with melting temperatures (Tm) of approximately 60°C for all targets and controls in the assay. In order to minimize the potential for primer-dimer formation in a multiplex reaction, primer candidate sequences were screened with in-house developed software. This software analyzed all primer candidates for sequence interactions and output problematic individual primer pairs. Potential sequence interactions were ranked according to strength, expressed in ΔG values, and location (i.e., either internally or at the 3′-end of one of the primers, the latter having a higher probability for sequence extension during PCR). Primer candidates with values of ΔG < -6 and extendable 3′- ends were eliminated from the primer candidate pool. This approach generated in silico matched primers for use in multiplex reactions.
PCR reactions contained 2× Qiagen Multiplex PCR Master mix with HotStarTaq DNA polymerase (Qiagen, Valencia, CA, USA), gene-specific primer pairs (250–600 nM), where one primer of each pair was labeled with 6-carboxyfluorescein (FAM), and the plasmid clones of quantification calibrators and detection sensitivity controls (SCs). The SCs varied in size from their cognate wild-type targets (Table 1), and were amplified using the identical set of wild-type target primers, included at approximately 100 copies per reaction. In the absence of a viral target, the amplification of the SC ensures proper functioning of its cognate primer pair. Thermal cycling parameters included an initial 15-min hot-start activation of the HotStarTaq DNA polymerase, followed by 95°C for 30 s, 62°C for 90 s, and 72°C for 60 s for each of 41 cycles. Sampling for CE was performed 12 times during alternate cycles, beginning with cycle 19 and concluding with cycle 41.
Quantification calibrators consisted of a set of three engineered plasmids, designated DNA A, DNA B, and DNA C, each producingamplicons of unique sizes (113 bp for DNA A, 304 bp for DNA B, and 350 bp for DNA C) from an identical primer pair and each added at a predetermined concentration of 50,000 copies for DNA A, 5000 copies for DNA B, and 1000 copies for DNA C. These were used to generate internal calibration threshold cycle (Ct) values for every sample by comparing the amounts of each calibrator input with the resulting Ct. The amplification curves for each target analyte amplicon were reconstructed using ICEPlex analysis software by quantifying the electropherogram peak areas over successive PCR cycles (8). Ct values for each target in a sample were then determined and compared with the calibrator Ct values, to arrive at the number of target copies initially present in each reaction: Nt = Nc/Ec[ΔCt] [Where Nt = number of test targets; Nc = number of calibrators; Ec = efficiency of doubling, which here = 2.0; and ΔCt = Ct(test) – Ct(ref)].