Bruton’s tyrosine kinase (BTK) is a major drug target in immune cells. The membrane-binding pleckstrin homology and tec homology (PH-TH) domains of BTK are required for signaling. Dimerization of the PH-TH module strongly stimulates the kinase activity of BTK in vitro. Here, we investigated whether BTK dimerizes in cells using the PH-TH module and whether this dimerization is necessary for signaling. To address this question, we developed high-throughput mutagenesis assays for BTK function in Ramos B cells and Jurkat T cells. We measured the fitness costs for thousands of point mutations in the PH-TH module and kinase domain to assess whether dimerization of the PH-TH module and BTK kinase activity were necessary for function. In Ramos cells, we found that neither PH-TH dimerization nor kinase activity was required for BTK signaling. Instead, in Ramos cells, BTK signaling was enhanced by PH-TH module mutations that increased membrane adsorption, even at the cost of reduced PH-TH dimerization. In contrast, in Jurkat cells, we found that BTK signaling depended on both PH-TH dimerization and kinase activity. Evolutionary analysis indicated that BTK proteins in organisms that evolved before the divergence of ray-finned fishes lacked PH-TH dimerization but had active kinase domains, similar to other Tec family kinases. Thus, PH-TH dimerization is a distinct feature of BTK that evolved to exert stricter regulatory control on kinase activity as adaptive immune systems gained increased complexity.

RNA-seq libraries from BTK-transduced and CD69 selected or input cells were prepared from TRI reagent-extracted samples as follows, beginning with the reverse-transcription. Following precipitation, RNA was resuspended in water containing 5 µM RT primer. The mixture was heated to 65ºC for 5 min, then snap cooled on ice. The RNA–oligo mixture was then mixed with 1x first-strand buffer (Thermo Fisher), 0.5 mM deoxynucleotide triphosphates, 10 mM DTT, 1 µL SuperaseIn (Thermo Fisher), and 0.5 µL Superscript III (Thermo Fisher). The final reaction mixture was incubated at 50ºC for 1 h. Followed this incubation, RNA was hydrolyzed by addition of 5 µL of 1 M NaOH and incubation at 90ºC for 10 min. The mixture was then neutralized with 25 µL of 1 M HEPES, pH 7.4, and desalted using a Micro Bio-Spin P-30 column, in tris (BioRad), eluting the cDNA in 60 µL. cDNA was amplified using PCR, which also attached the Miseq hybridizing sequences. Fastq files from MiSeq runs were aligned to the Fasta files containing the full sequences of each variant using Kallisto (Bray et al., 2016) to generate read counts for each variant.