Pseudohyphal differentiation in Komagataella phaffii: investigating the FLO gene family (FAIRE-seq)

The protein production host Komagataella phaffii has possess the ability to differentiate into pseudohyphal form when cultivated at slow growth rates (µ=0.05 h-1) in glucose-limited chemostats. In this study, we investigated the K. phaffii FLO gene family in the context of pseudohyphae formation. Transcriptional analysis helped us identify 3 possible responsible genes, FLO11, FLO400 and FLO5-1, all of which are under control of Flo8, a transcription factor whose disruption prevents pseudohyphae formation. Knocking out FLO11 revealed that this is not the sole protein responsible for this phenotype. Strikingly, the expression of FLO400 and FLO5-1 was negatively correlated with pseudohyphae formation, and shown to be under epigenetic control by FAIRE-Seq analysis. Knock outs of these two genes completely inhibited the appearance of pseudohyphal cells and prevented the expression of FLO11. Even though the mechanism is unclear at present, we propose that in K. phaffii Flo400 and/or Flo5-1 act as upstream signals that lead to the induction of FLO11 expression upon severe glucose limitation in chemostats at slow growth rate, and that the expression of FLO400 and FLO5-1 is controlled by epigenetic silencing, which acts independently from the general activation of FLO gene expression by the transcriptional regulator Flo8. Overall design: To investigate possible changes in chromatin accessibility that occur at a genome-wide level of fast and slow growing K. phaffii, we carried out Formaldehyde Assisted Isolation of Regulatory Elements (FAIRE-Seq) analysis of samples collected during glucose-limited chemostat cultivation after 5 residence times at different growth rates; fast (µ=0.1 h-1), slow (µ=0.05 h-1) and finally again switching back to the fast (µ=0.1 h-1) growth rate. The cultivation was carried out for both wild-type CBS7435 as well as the CBS7435 strain with a disrupted FLO8 gene. In both cases samples were taken in biological replicates, except for the ?flo8 sample at the slow growth rate and after switching back to the fast growth rate, where due to technical problems the chemostat had to be terminated and only one sample was available for these time points. An input control sample was used which was a formaldehyde fixed sample which was decrosslinked overnight before carrying out the phenol/chloroform extraction