The first three months of the COVID-19 pandemic was dominated by two SARS-CoV-2 lineages: A-lineages (Clade 19B) and B-lineages (Clade 19A). However, with the emergence of the Spike D614G substitution in B.1 lineages (Clade 20A), both early lineages were outcompeted and remained near-extinction from mid-2020 onwards. In early-2021, there was a re-emergence and persistence of novel A-lineage variants with substitutions in the Spike gene resembling those found in Variants of Concern (VOCs). An early A.3 variant (MD-HP00076/2020) and three A.2.5 variants (MD-HP02153/2021, MD-HP05922/2021 and CA-VRLC091/2021) were isolated and characterized for their genomic sequences, antibody neutralization, andin vitroreplication. All A.2.5 isolates had five Spike mutations relative to the A.3 variant sequence: D614G, L452R, Δ141-143, D215A, and ins215AGY. Plaque reduction neutralization assays demonstrated that A.2.5 isolates had a 2.5 to 5-fold reduction in neutralization using contemporaneous COVID-19 convalescent plasma when compared to A.3.In vitroviral characterization in VeroE6 cell lines revealed that the A.3 isolate grew faster and spread more than A.2.5. On VeroE6-TMPRSS2 cells, significant syncytia formation was also observed with the A.2.5 isolates, however Spike cleavage efficiency did not explain these differences. In human nasal epithelial cell (hNEC) cultures, the A.2.5 isolates grew significantly faster and to higher total infectious virus titers than A.3. All A.2.5 lineage isolates grew significantly faster at 37°C than at 33°C irrespective of cell type, and to higher peak titers except compared to A.3. This suggests A.2.5’s adapted to improve replication using similar mutations found in the B-lineage SARS-CoV-2 variants.ImportanceWhile both A- and B-lineage SARS-CoV-2 variants emerged and circulated together during the early months of the pandemic, the B-lineages that acquired Spike D614G eventually outcompeted all other variants. We show that the A-lineage variants eventually evolved mutations including Spike D614G and Spike L452R that improved their in vitro replication in human nasal epithelial cells in a temperature dependent manner, suggesting there are some highly selectable mutation landscapes that SARS-CoV-2 can acquire to adapt to replication and transmission in humans.