The mutational dynamics of the SARS-CoV-2 virus in serial passages in vitro

Abstract

Since its outbreak in 2019, Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) keeps surprising the medical community by evolving diverse immune escape mutations in a rapid and effective manner. To gain deeper insight into mutation frequency and dynamics, we isolated ten ancestral strains of SARS-CoV-2 and performed consecutive serial incubation in ten replications in a suitable and common cell line and subsequently analysed them using RT-qPCR and whole genome sequencing. Along those lines we hoped to gain fundamental insights into the evolutionary capacity of SARS-CoV-2 in vitro. Our results identified a series of adaptive genetic changes, ranging from unique convergent substitutional mutations and hitherto undescribed insertions. The region coding for spike proved to be a mutational hotspot, evolving a number of mutational changes including the already known substitutions at positions S:484 and S:501. We discussed the evolution of all specific adaptations as well as possible reasons for the seemingly inhomogeneous potential of SARS-CoV-2 in the adaptation to cell culture. The combination of serial passage in vitro with whole genome sequencing uncovers the immense mutational potential of some SARS-CoV-2 strains. The observed genetic changes of SARS-CoV-2 in vitro could not be explained solely by selectively neutral mutations but possibly resulted from the action of directional selection accumulating favourable genetic changes in the evolving variants, along the path of increasing potency of the strain. Competition among a high number of quasi-species in the SARS-CoV-2 in vitro population gene pool may reinforce directional selection and boost the speed of evolutionary change.

Keywords: Adaptation; Mutational dynamics; SARS-CoV-2; Serial passage in vitro; Whole genome sequencing (WGS).

Fig. 1

Maximum likelihood tree based on the WGS sequences of ten SARS-CoV-2 strains in the course of serial passages in vitro. Sequencing was carried out at passages 0, 3, 5, 7 and 10 respectively, and only high-quality sequences were used for further analysis, as shown here in 49 SARS-CoV-2 with 29,896 nucleotide sites each. Numbers at nodes indicate bootstrap support values (only values ​> ​50 are shown). Bootstrap is also given by branch colour.

Fig. 2

Replication success of ten ancestral strains of SARS-CoV-2 over ten serial passages. The replication success is defined as the development of the viral load in the course of the experiment. All strains started with viral loads of 7.6 ​× ​10⁸. Three strains did not manage to replicate and got lost within the first three passages (yellow bars), four strains maintained in culture with a decreased or nearly unchanged viral load (blue bars), three strains showed significantly higher replication success (green bars). The bars show the growth factor of the diverse strains, calculated as the percentage growth of the strain during the experiment (PFU/mL at P10 ​× ​100)/(PFU/mL at P0). The numbers above the bars represent the acquired aa substitutions during the serial passages. Numbers in parenthesis show additional temporary substitutions which did occur once in a sequence throughout the experiment but were not preserved. The first three strains, here marked with the first letter B, did not manage to replicate successfully in cell culture and were lost after at least four passages. Actual increases, thus successful adaptation to the cell cultures, were recorded in the six strains A818 to A4707, with the last three strains A7326, A6137 and A4707 achieving the highest increase in PFU/mL with growth of 4 ​× ​10³% up to 3 ​× ​10⁴%.

Fig. 3

A summary of all aa substitutions in SARS-CoV-2 in a study of the mutational dynamics of ten endemic strains of SARS-CoV-2 during serial passage. Conserved substitutions are those mutations acquired before isolation of the virus in cell culture and shown as dark-grey bars. All these conserved substitutions were sustained over the study period. Acquired substitutions evolved in the course of the serial passages. Acquired substitutions of less successful strains are shown in green, those of the highly successful strains in blue. Every temporary substitution could only be determined once in a sequence and disappeared in the course of the following passages. f, shows the frequency of each mutation (i.e., number of strains the mutation were found in). ∗∗∗, convergent mutation in all three most successful strains.

Fig. 4

Types of all acquired amino acid (aa) substitution in 11 coding sequences, acquired by SARS-CoV-2 strains in serial passages in vitro. This figure summarizes the distribution of aa substitutions on 11 coding sequences (ORF1a, ORF1b, S, ORF3a, E, M, ORF6, ORF7, ORF8, N, and ORF10). Orange bars indicate mutations acquired by the strains prior to isolation and preserved throughout the full range of serial passages, yellow bars indicate substitutions acquired by the strains throughout the in vitro serial passages, green bars show temporary mutations during one serial passage.

Fig. 5

Localization of the aa substitutions and insertions developed convergently from two or more strains during serial passage of ten SARS-CoV-2 strains in vitro. All convergently evolved mutations occurred in the regions coding for ORF1a/nsp3 (1); ORF3a (1), E (1), M (1), ORF8 (1) and spike (8). Substitutions marked with “∗∗∗” evolved convergently in all three highly successful strains.