2.2.3. Three-dimensional (3D)-digital (d) PCR

The main difference of 3D-dPCR from RT-PCR is that the reaction volume is split over a high number of small partitions (from 500 up to millions) of a very small volume (currently from 6 nl down to a few picolitres). After the PCR, each partition is scored either as positive or negative (binary or digital read-out). Statistical analysis of the results is then used to determine the absolute quantity of target DNA in a sample (^{Majumdar et al., 2015; Pavšič et al., 2015}). This method shows high sensitivity with the strategy of counting a single molecule, and it also provides a high reliability and repeatability level (^{Suo et al., 2020}). Furthermore, this method has been proven to work with high sensitivity in studies related to SARS-CoV-2 (^{de Almeida et al., 2020; Suo et al., 2020}).

The probe set for N1 gene were obtained from the CDC as described above (^{CDC, 2020}). Primers used in RT-PCR analyses were also used in 3D-dPCR analyses. The probe sequence of N1 gene was FAM-5’-ACCCCGCATTACGTTTGGTGGACC-3’-BHQ1. Cycling conditions were as follows; in 10 min at 96 °C, followed by 39 cycles of 30 s at 63 °C and 2 min at 98 °C, and a final step of 63 °C per 2 min followed by maintenance at 4 °C. The chips were read in the QuantStudio 3D™ reader (Thermo, USA), and the results were interpreted in the dPCR AnalysisSuite™ app in the Thermo Fisher Connect™ Dashboard. For positive control, dilutions of (10⁻³, 10⁻⁴, 10⁻⁵ and 10⁻⁶) of a clinical isolate of SARS-CoV-2 virus (TCID50 = 10⁶/ml) RNA was used.

The positivity of the SARS-CoV-2 on PM samples was characterized by a multi-parameter decision approach. The 3D-dPCR results (copy number/ μl) were transformed in a threshold of viral particles contained in a single filter (i.e. RNA copies per filter) calculating the volume of RNA and extraction solutions. Later, viral load per filter numbers were normalized using the sampled volume of air to calculate the viral concentration in air per m⁻³.