In ongoing efforts to identify mutants of C/EBPα in post-translationally modified amino acids that affect BMT (Figure 1A), we used retroviral constructs with C/EBPα fused to ER and inducible by 17ß-estradiol (E2). Rather than finding loss-of-function mutations as expected, we identified one that dramatically accelerates the pace of BMT (Figure 1B and C). This mutant (C/EBPα) contains alanine replacements of three evolutionarily conserved arginines located in the N-terminal TEI and TEII transcription activation elements of C/EBPα.

The individual contributions of the three arginines to the cells’ phenotype were studied by testing the effect of single alanine replacements for R12, R35, and R86, again using E2-inducible constructs. Results showed that C/EBPα^R35A recapitulated the phenotype induced by C/EBPα, while R12A and R86A essentially displayed no effect (Figure 1D and Figure 1—figure supplement 1A). R35A cells induced for 120 hr resembled normal macrophages, showing extensive F-actin filaments and eccentric nuclei, as well as high phagocytic capacity (Figure 1—figure supplement 1B, C). These findings indicate that replacing arginine 35 with alanine in C/EBPα dramatically increases the speed of induced BMT, generating macrophage-like cells similar to those obtained with the wild-type.

To investigate the impact of C/EBPα^R35A on gene expression, we conducted RNA-sequencing (RNA-seq) on infected B cells spanning very early to late timepoints (0, 1, 3, 6, 18, and 120 hr; Figure 1E). The analysis revealed 11,780 genes whose expression changed by more than twofold throughout BMT in either system. Principal component analysis (PCA) showed a pronounced acceleration of gene expression changes in cells infected with the mutant virus compared to the wild-type. At just 1 hpi, the gene expression patterns in C/EBPα^R35A-infected cells were similar to those of 18 hpi wild-type-infected cells, reaching similar expression levels at 120 hpi. The largest differences in gene expression between the wild-type and mutant cells were observed between 0 and 1 hpi (Figure 1F and Figure 1—figure supplement 1D).

Hierarchical clustering of the 11,780 differentially expressed genes (DEGs) resulted in eight clusters (Figure 1G). Most genes in clusters I, II, IV, and VIII showed faster activation by C/EBPα^R35A, while those in clusters IV, V, and VII showed faster silencing. Clusters I and II were enriched for macrophage-myeloid-related Gene Ontology (GO) terms and included the myeloid genes Itgam, Lyz2, Csf1r, and Cd14 (Figure 1H and Figure 1—figure supplement 1F, G). On the other hand, cluster VII was enriched for B cell-related GO terms and included the B cell genes Cd19, Pax5, Ebf1, and Rag2 (Figure 1I and Figure 1—figure supplement 1F, G). The kinetics of macrophage- and B cell-restricted genes (Figure 1 and Figure 1—figure supplement 1G, H) further highlight the acceleration of BMT induced by C/EBPα^R35A.

Through examination of thousands of differentially regulated lineage-associated genes, these results extend the distinction between C/EBPα mutant- and wild-type-induced BMT observed by FACS. Here, however, the most significant differences in gene expression were found at the earliest stages (1–3 hpi).

To examine the chromatin remodeling accompanying induced gene expression changes during BMT, we conducted an assay for transposase-accessible chromatin sequencing (ATAC-seq) (Buenrostro et al., 2015) of samples collected at multiple timepoints after activation of C/EBPα^WT and C/EBPα^R35A (Figure 2A). Our analysis identified 91,830 regions that showed significant differences in chromatin accessibility throughout wild-type- or mutant-induced BMT, representing in their vast majority TF-bound GREs (Buenrostro et al., 2015). These regions were grouped into (1) faster opening GREs induced by C/EBPα^R35A, with the highest accessibility at 120 hpi (43,429 GREs); (2) faster closing GREs, most accessible at 0 hpi (36,380 GREs); and (3) transiently opening GREs, most accessible at 18hpi (12,021 GREs) that showed little differences between the two conditions (Figure 2—figure supplement 1A,B). PCA of the differential ATAC peaks further confirmed the acceleration of chromatin accessibility changes by C/EBPα^R35A, with the coordinates of the 1–6 hpi C/EBPα^R35A samples resembling those of the 18 hpi C/EBPα^WT sample (Figure 2B).

Focusing on differentially remodeled promoter regions, we found 14,233 and these were grouped into eight clusters (Figure 2C). Promoter-associated genes in clusters I, II, and III exhibited dramatically accelerated opening dynamics for C/EBPα^R35A, while genes in clusters VII and VIII showed accelerated closing. GO analysis revealed an enrichment of macrophage terms for cluster III, which included the macrophage-restricted genes Itgam and Lyz2 (Figure 2C). C/EBPα^R35A already initiated chromatin opening at promoters and enhancers of Itgam and Lyz2 within 1 hpi, and this increased over time much more rapidly than with C/EBPα^WT cells (Figure 2—figure supplement 1C). Conversely, faster closing promoters in cluster VIII were enriched for B cell-related GO terms and included the B cell genes Cd19, Pax5, and Rag2 (Figure 2C and E). The kinetics of chromatin closing were likewise accelerated at B cell GREs. Strikingly, however, as exemplified for Cd19 and Pax5, while mutant-induced changes occurred already at 1 hpi and progressed from there onward, the same sites showed a transient chromatin opening in wild-type-induced cells (Figure 2—figure supplement 1D).

Our results show that C/EBPα^R35A is more effective than C/EBPα^WT in inducing changes of chromatin closing and opening at B- and M-GREs. However, while the mutant acts immediately to close chromatin, wild-type-induced chromatin closing is preceded by a transient chromatin opening, as is also reflected by a transient increase in gene expression (Figure 1—figure supplement 1F, G). This is in line with the recently described Gata1-induced silencing of the Kit gene enhancer during erythroid cell differentiation, which becomes transiently activated (Vermunt et al., 2023). It is also noteworthy that despite progressive alterations in chromatin accessibility at B- and M-GREs during BMT induction (Figure 2—figure supplement 1D), essentially no changes in gene expression were observed between 6–18 hpi, suggesting that chromatin remodeling is uncoupled from gene expression during this phase of the process.

To determine if the enhanced ability of the mutant to modify chromatin accessibility was due to differences in DNA binding capacity, we compared nuclear extracts of 293T transfected with C/EBPα^WT and C/EBPα^R35A. An electrophoretic mobility shift assay (EMSA) revealed no significant differences in overall DNA binding affinities between the two C/EBPα forms (Figure 3—figure supplement 1A).

We next explored whether the differences are due instead to the mutant’s preferential interaction with other proteins, performing a de novo motif discovery analyses on differentially accessible GREs. Results in Figure 3A showed that faster and transiently opening GREs were enriched for AP-1/leucine zipper family TF motifs (c-Fos, c-Jun, and JunB), known to heterodimerize with C/EBPα and activate myeloid genes (Cai et al., 2008). In contrast, faster closing GREs were mostly enriched for ETS family TF motifs (Ets1, Fli1, SpiB, and Gabpa), linked to B cell lineage differentiation and function (Eyquem et al., 2004; Hu et al., 2001; Xue et al., 2007). Of note, the Spi1 (PU.1) motif was enriched in both the accelerated chromatin opening and closing groups (Figure 3B and Figure 3—figure supplement 1B), possibly reflecting the known dual roles of PU.1 in B cells and macrophages (Scott et al., 1994; Singh et al., 1999; Heinz et al., 2010; Rothenberg, 2014). However, no PU.1 motif enrichment was seen in the transiently opening group (Figure 3—figure supplement 1C).

To examine whether C/EBPα^WT and C/EBPα^R35A differentially interact with other proteins, we performed a proximity-dependent biotin identification assay (Figure 3B), which was previously used to identify high confidence C/EBPα interactors (Ramberger et al., 2021a). We used a mouse pre-B cell line expressing doxycycline (Dox)-inducible C/EBPα^WT or C/EBPα^R35A constructs fused with the biotin ligase variant TurboID (Branon et al., 2018). To avoid confounding interactions this cell line contained a double knockout of Cebpa and Cebpb, genes which are rapidly activated by exogenous C/EBPα (Bussmann et al., 2009). After Dox induction for 6 hr and biotin labeling, biotin-labeled proteins were identified by mass spectrometry. A comparison of the high confidence C/EBPα^WT and C/EBPα^R35A interactome revealed that C/EBPα^R35A preferentially interacted with Spi1 (PU.1), showing a 3.6-fold increase in biotin label compared to C/EBPα^WT. In addition, preferential but non-significant interactions of C/EBPα^R35A were observed with Ncor1 and Ebf1 (Figure 3B; Figure 3B, source data 1), which, however, were not further studied. The observation that most shared interactors show a skewing toward the wild-type protein suggests that C/EBPα^WT is expressed at slightly higher levels than the mutant, in line with the detection of moderately reduced levels of the mutant protein by western blot (Figure 4—figure supplement 1C). To further evaluate the interaction between C/EBPα^WT and C/EBPα^R35A with PU.1, we conducted co-immunoprecipitation (Co-IP) experiments with 293T cells co-transfected with PU.1 and either C/EBPα^WT or C/EBPα^R35A. The results showed an approximately twofold increase in the interaction between C/EBPα^R35A and PU.1 compared to C/EBPα^WT and PU.1 (Figure 3C). Finally, performing a proximity ligation assay (PLA) to study the protein’s interaction in live cells showed a significantly higher number of fluorescent nuclear dots for the combination of C/EBPα^R35A with PU.1 than with C/EBPα^WT (Figure 3D), again indicating a higher affinity between the two proteins.

In sum, motif analyses showed an enrichment of the PU.1 motif at both B- and M-GREs that were differentially accessed by the mutant. Moreover, the results obtained by Co-IP, mass spectrometry, and PLA indicate that C/EBPα^R35A shows a 1.5- to 3.5-fold higher affinity for PU.1 compared to the wild-type.

Previously, we showed that the combination between C/EBPα and PU.1 is required for converting NIH 3T3 fibroblasts into macrophage-like cells (Feng et al., 2008). To investigate if the synergy between the C/EBPα mutant and PU.1 is altered, we isolated NIH 3T3 cell lines expressing C/EBPα^WT-ER or C/EBPα^R35A-ER. These were subsequently infected or not with a PU.1 retroviral construct, and Mac-1 levels monitored at different times after E2 treatment. The results showed that in the presence of PU.1, C/EBPα^R35A activated Mac-1 more strongly than C/EBPα^WT, while cells infected with PU.1 only showed weak Mac-1 activation and C/EBPα only cells remained negative (Figure 4A, Figure 4—figure supplement 1A). Additionally, cells co-expressing C/EBPα^R35A and PU.1 displayed more striking morphological changes than those with C/EBPα^WT and PU.1 (Figure 4—figure supplement 1B). Cycloheximide treatment experiments showed that C/EBPα^R35A exhibits a similar stability than wild-type protein and under steady-state condition is expressed at 20–30% lower levels (Figure 4—figure supplement 1C). This indicates that the observed gain of function of the mutant cannot simply be explained by an increased stability or expression.

To investigate the effect of C/EBPα^WT and C/EBPα^R35A on chromatin structure and the influence of PU.1, we performed single-molecule tracking (SMT) experiments. SMT allows to observe the movements of individual TF molecules in living cells (Liu and Tjian, 2018) and has been recently used to quantify the interactions of TFs with open versus closed chromatin, by scaling to the mobility of single nucleosomes (Lerner et al., 2020; https://www.biorxiv.org/content/10.1101/2022.12.22.521655v1). For this purpose, we generated NIH3T3 cell derivatives that stably express different Halo-tagged proteins: histone H2B, Dox-inducible C/EBPα^WT, and C/EBPα^R35A. Measuring the radius of confinement and average displacements (https://star-protocols.cell.com/protocols/324) in cells expressing histone H2B-Halo served to define low versus high mobility chromatin (LoMC and HiMC), corresponding to compact versus open states, respectively (Lerner et al., 2020; Figure 4C). C/EBPα^WT-Halo and C/EBPα^R35A-Halo cultures were split and either infected with a PU.1 lentiviral construct or left uninfected (Figure 4D). Cells were then treated with Dox for 6 or 24 hr to induce C/EBPα, and SMT performed on ~50 cells per condition, randomly down-sampling 20,000 single-molecule motion tracks in triplicate.

Analyses of the C/EBPα^WT-Halo and C/EBPα^R35A-Halo motion tracks at 6 hpi revealed an interaction of both proteins with LoMC and HiMC, showing a moderate decrease of HiMC for C/EBPα^R35A compared to cells induced for C/EBPα^WT (Figure 4E and F). In contrast, C/EBPα^R35A-Halo cells co-expressing PU.1 showed a substantially decreased interaction with LoMC and increased interaction with HiMC, while PU.1 only moderately affected the distribution of C/EBPα^WT-Halo particles (Figure 4E and F). At 24 hpi, both forms of C/EBPα showed decreased interaction with LoMC and increased interaction with HiMC, although the differential effects of PU.1 co-expression were milder than at 6 hpi (Figure 4G).

Altogether, the SMT results obtained with NIH3T3 fibroblasts overexpressing C/EBPα and PU.1 demonstrate that PU.1 synergizes with C/EBPα in inducing substantial chromatin occupancy changes from closed to open chromatin states at 6 hpi. The synergism is much stronger for C/EBPα^R35A than for C/EBPα^WT and becomes less pronounced at 24 hpi, indicating that wild-type C/EBPα catches up in its ability to synergize with PU.1. These findings support the observed ability of C/EBPα and PU.1 to interact with nucleosomes in vitro and to act as a pioneer factors (Fernandez Garcia et al., 2019; Lerner et al., 2020; van Oevelen et al., 2015).

The findings described so far pointed to the possibility that PU.1 acts as a coordinator of the B cell to macrophage conversion. A plausible scenario therefore is that C/EBPα removes PU.1 from B cell GREs and induces its relocation to macrophage GREs. Moreover, through its enhanced interaction with PU.1, C/EBPα^R35A might be more efficient at relocating the factor, explaining the observed acceleration of BMT. To test this hypothesis, we conducted DNA binding experiments by chromatin immunoprecipitation followed by high-throughput sequencing (ChIP-seq) for PU.1 and C/EBPα with primary B cells expressing C/EBPα^WTER and C/EBPα^R35AER 3 hr after E2 induction. The data obtained were then visualized on the UCSC browser in parallel with our ATAC-seq data with cells from 0, 1, 3, 6, 18, and 120 hpi. The main messages of our results are encapsulated by the Cyfip2-Havcr2 locus, a 270 Kb genomic region containing both a B cell- and a macrophage-restricted gene (Figure 5A).

Of the five genes contained in this locus, two, namely Cyfip2 and Havcr2, become down- and upregulated, respectively, during BMT, with C/EBPα^R35A inducing an acceleration of both processes (Figure 5B). Mining of published data (Mullen et al., 2011) showed that PU.1 binds to four sites 50–150 kb upstream of Cyfip2 in B cells but is essentially absent at these sites in macrophages (highlighted in red). Conversely, both PU.1 and C/EBPα bind to the promoter of Havcr2 and three additional GREs in macrophages but not in B cells (highlighted in green, Figure 5C), showing that PU.1 can bind to both B- and M-GREs while C/EBPα can only bind to M-GREs.

Analyzing the PU.1 binding sites in primary B cells induced by C/EBPα for 3 hr, the factor showed higher occupancy at the B-GREs of C/EBPα^WT cells and a modestly increased binding at the M-GREs of C/EBPα^R35A cells (Figure 5D). Similarly, C/EBPα^WT strongly bound to the B-GREs and weakly to the M-GREs, while C/EBPα^R35A was selectively bound at the M-GREs (Figure 5E, Figure 5—figure supplement 1A). Since the DNA binding data only represent a snapshot in time, we also analyzed our dynamic ATAC-seq data for the Cyfip2-Havcr2 locus and found that the mutant accelerates both the closure of B-GREs and opening of M-GREs (Figure 5—figure supplement 1B). Notably, while C/EBPα^WT triggered a transient and partial opening of B-GREs at 1 hpi and closure at 18 hpi, the mutant induced an immediate and progressive closure (Figure 5F). Similar effects were also seen at GREs of B cell signature genes (Figure 2—figure supplement 1D).

Our findings at the Cyfip2-Havcr2 locus revealed that at 3 hpi both PU.1 and C/EBPα show distinct DNA occupancies at B- and M-GREs in wild-type and mutant expressing cells. These differences can best be explained by the hypothesis that C/EBPα triggers the redistribution of PU.1 bound to B-GREs to M-GREs.

The finding that the PU.1 redistribution appears to be accelerated in the mutant cells predicts that B cell enhancers show lower levels of PU.1 binding in C/EBPα^R35A-induced cells than in C/EBPα^WT-induced cells. Conversely, myeloid enhancers should show higher PU.1 occupancy. To test this, we conducted a genome-wide analysis of the ChIP-seq data, revealing 4310 regions differentially bound by PU.1 at 3 hpi. These regions were divided into sites bound either Less or More in C/EBPα^R35A cells relative to C/EBPα^WT cells. As seen from the ATAC-seq data, the majority of the observed 1647 sites that were bound Less by PU.1 became inaccessible at 18 hpi in C/EBPα^WT cells, showing a transient opening between 1 and 6 hpi. In contrast, in C/EBPα^R35A cells these sites became already partially closed at 1 hpi and fully inaccessible after 18 hpi (Figure 6A). The opposite was observed for the 2650 sites bound More by PU.1, where chromatin accessibility increased already at 1 hpi in the mutant cells, while in wild-type cells they only opened at 120 hpi (Figure 6D and E). As predicted, genes associated to sites bound Less by PU.1 showed a strong enrichment of B cell-related GO terms while More-bound sites were enriched for myeloid-related GO terms (Figure 6C).

The redistribution model further predicts that C/EBPα^R35A is released from B cell enhancers already at 3 hpi and is instead enriched at myeloid GREs, while C/EBPα^WT shows a relocation delay. We therefore compared the genome-wide DNA binding profiles of wild-type and mutant C/EBPα proteins. Of 7144 differentially occupied sites, 4352 sites were bound Less by C/EBPα^R35A. Approximately 70% of these started closing at 1 hpi in C/EBPα^R35A cells while in C/EBPα^WT cells they only closed at 120 hpi, after transiently opening between 1 and 6 hpi (Figure 6—figure supplement 1A). The remaining 30% of the sites showed a gradual but heterogeneous increase in accessibility. Finally, about half of the sites bound More by C/EBPα^R35A became more rapidly opened, as expected, while the other half showed a transient opening before closing in both wild-type and mutant cells (Figure 6—figure supplement 1B). GO analysis for the Less C/EBPα^R35A-bound sites showed an association with both B cell and myeloid terms while the rapidly opening More bound sites were associated with myeloid terms (Figure 6—figure supplement 1C).

Finally, we tested sites bound Less by both PU.1 and C/EBPα^R35A and found 783 shared sites, whereas 385 sites were bound More by both PU.1 and mutant C/EBPα. In contrast, no overlap was found between PU.1 bound Less and C/EBPα bound More or vice versa (Figure 6—figure supplement 1D) showing a strong positive correlation in the binding of the two factors. The chromatin accessibility dynamics at these commonly bound regions reflected by and large the observations described for the singly bound factors and showed the expected association with B cell and myeloid-associated GO terms (data not shown).

In conclusion, our genome-wide analyses showed that C/EBPα induces the relocation of PU.1 from B- to M-GREs, in a process accelerated by the mutant. We also observed that C/EBPα^WT transiently opens B-GREs during B cell silencing, while C/EBPα^R35A induces their immediate closing. The finding that the chromatin of B-GREs is more accessible in uninduced C/EBPα^R35A cells suggests a leakiness of the construct and could at least in part explain this finding. However, this poised state does not result in an altered expression of associated B cell genes compared to uninduced C/EBPα^WT cells. Finally, our data showed the progressive closing/opening of B/M-GREs between 1 and 6 hpi, while gene expression during this phase remains essentially unchanged (Figure 1—figure supplement 1F, G; Figure 2—figure supplement 1C, D; Figure 5—figure supplement 1B). These observations suggest that chromatin restructuring can be uncoupled from changes in gene regulation.

The observed accelerated BMT induced by C/EBPα with an alanine replacement of arginine 35 could either be due to the replacement of a charged amino acid with a hydrophobic residue or to the loss of arginine methylation, or both. To test this, we replaced the arginine with the charged residue lysine in our inducible retroviral vector and expressed it in primary B cells. These cells were then induced with E2 for 1–120 hr and analyzed by RNA-seq and ATAC-seq in comparison to C/EBPα^WT-induced cells. PCA of the data showed an accelerated BMT time course of gene expression and chromatin accessibility (ATAC peaks) for the C/EBPα^R35K-induced cells (Figure 7—figure supplement 1A,B), similar but not identical to those seen with C/EBPα^R35A cells (not shown). This suggests that the C/EBPα^R35A-induced BMT acceleration is not simply due to a lack of charge at the R35 residue.

To explore whether R35 is asymmetrically di-methylated, we used a previously described human B cell to macrophage BMT system (Rapino et al., 2013). For this we generated 4-hydroxytamoxifen (4-OHT)-inducible lines expressing C/EBPα^WT and C/EBPα^R35A fused to ERT-2, called BLaER2 and BLaER2-A, respectively. After treating the cells with 4-OHT for 24 hr and immunoprecipitating C/EBPα followed by western blotting and staining of the blot with an antibody specific for asymmetrically di-methylated arginine-containing proteins, C/EBPα could be detected in BLaER2 but not in BlaER2-A cells, as expected (Figure 7A). Similar results were obtained with transfected 293T cells, showing methylated C/EBPα in the wild-type but not mutant cells (Figure 7B).

To identify which of the Prmts known to asymmetrically methylate arginine is responsible, 293T cells were co-transfected with C/EBPα^WT or C/EBPα^R35A and with either Prmt1, Prmt3, Carm1 (Prmt4), or Prmt6. Only Carm1 induced asymmetric di-methylation of C/EBPα (as detected with an antibody specific for the modification) and that this was restricted to the wild-type protein (Figure 7—figure supplement 2A). These data indicate that Carm1 specifically and asymmetrically di-methylates R35.

We next performed Co-IP experiments to examine whether C/EBPα and Carm1 can interact, using 293T cells co-transfected with the two proteins, and found that both C/EBPα^WT and C/EBPα^R35A are capable of interacting with Carm1 (Figure 7C). Next, we performed a PLA with NIH 3T3 cell derivatives expressing either C/EBPα^WT-ER or C/EBPα^R35A-ER. After 24 hr of induction with E2, cells were stained with antibodies to C/EBPα and Carm1 and a secondary antibody to detect their interaction. This again demonstrated that both forms of C/EBPα can interact with Carm1, with C/EBPα^R35A cells showing a slightly higher signal (Figure 7—figure supplement 2B).

To determine which of the 20 arginines in C/EBPα can be bound by Carm1, we conducted an in vitro methylation assay using 14 synthetic peptides spanning the entire molecule. Only the peptide containing arginine 35 generated a signal (Figure 7—figure supplement 2C). Next, we studied the enzyme’s specificity for selected targets. For this we chose as substrates a 40 amino acid long C/EBPα peptide with either an unmethylated R35, an asymmetrically di-methylated R35, or an alanine replacement. The three peptides were then treated with purified Carm1, Prmt1, Prmt3, Prmt6, or Prmt8 proteins. Only unmethylated R35 peptide could be methylated Carm1, with all other conditions being negative (Figure 7D). To examine if methylation affects the affinity of an R35 containing C/EBPα peptide for PU.1, we analyzed the data from an interactome screen (PRISMA) (Ramberger et al., 2021a). The comparison of a 15 mer with either an unmethylated or an asymmetrically di-methylated R35 showed that PU.1 interacts more strongly with the unmethylated peptide (Figure 7—figure supplement 2F).

In summary, our results indicate that Carm1, but none of several other Prmts tested, induces the asymmetric di-methylation of specifically arginine 35 of C/EBPα and that an unmethylated peptide at R35 has a higher affinity for PU.1 than its methylated form.

To evaluate the effect of Carm1 activity on BMT, we conducted gain- and loss-of-function experiments. For this, we generated two BLaER2 cell derivatives (RRC3 and RAC-1) containing 4-OHT-inducible C/EBPα^WT and C/EBPα^R35A, respectively, and a Dox-inducible Carm1 construct. For the gain-of-function experiments we treated 4-OHT-induced cells with Dox and observed a strong Carm1 band by western blot (Figure 7—figure supplement 2D) while treatment with TP064 (Nakayama et al., 2018) reduced the level of asymmetrically di-methylated BAF155 (Figure 7—figure supplement 2E), a well-described target of the enzyme (Wang et al., 2014). Carm1 overexpression resulted in a delay of both Mac-1 activation and Cd19 silencing in 4-OHT-induced wild-type but not mutant C/EBPα-containing cells (Figure 7F and Figure 7—figure supplement 2G). Conversely, treatment of 4-OHT-induced C/EBPα^WT-expressing cells with TP064 resulted in a robust acceleration of both Mac-1 upregulation and Cd19 silencing while the kinetics of C/EBPα^R35A-expressing RAC1 cells were not altered (Figure 7G and Figure 7—figure supplement 2H).

To determine if endogenous Carm1 levels correlate with BMT velocity, we analyzed a single-cell gene expression dataset of primary B cells undergoing BMT, generated in a previous study (Francesconi et al., 2019). This showed that cells with the lowest Carm1 levels transitioned to a macrophage-like identity faster than those with higher levels, with the most substantial differences occurring during the early timepoints (Figure 7H), reflecting the more pronounced acceleration of gene expression within the first hours after induction of C/EBPα^R35A (Figure 1F).

Our data showed that Carm1 overexpression slows down BMT, whereas Carm1 inhibition accelerates it, and that such perturbations did not affect C/EBPα^R35A- induced BM. This is consistent with the notion that the proportion of C/EBPα with methylated versus unmethylated arginine 35 determines the velocity of BMT.

To examine if Carm1 plays a role in myeloid cell fate specification, we first analyzed its role during myeloid differentiation. During this process, common myeloid progenitors (CMPs) differentiate into granulocyte-macrophage progenitor (GMPs), which further differentiate into granulocytes and monocyte/macrophages (Figure 8A), requiring C/EBPα (Zhang et al., 1997; Zhang et al., 2004; Ohlsson et al., 2016).

Examining the expression of Carm1 in CMPs, GMPs, granulocytes, and macrophages (Choi et al., 2019) revealed a steady decrease during differentiation (Figure 8—figure supplement 1A). We next sorted GMPs, granulocytes, and monocytes from the bone marrow and measured AsDM-BAF155 as a proxy for Carm1 activity (Wang et al., 2014) relative to total BAF155. The highest Carm1 activity was detected in GMPs, with a 3.5-fold decrease in granulocytes and a further 4.5-fold decrease in monocytes (Figure 8B and Figure 8—figure supplement 1B). This showed that not only the expression of Carm1 but also its activity decreases during myeloid differentiation.

To investigate whether Carm1 activity influences the direction of differentiation of a bipotent precursor, we conducted a colony assay with sorted GMPs in the absence or presence of the Carm1 inhibitor TP064. For this, GMPs were seeded in semisolid medium containing 0, 1, 2.5, or 10 µM TP064 and the number of granulocytic (CFU-G), monocytic (CFU-M), and mixed (CFU-GM) colonies scored after 12 days (Figure 8—figure supplement 1). The results showed a significant dose-dependent reduction of granulocyte colonies and a corresponding increase in monocyte colonies (Figure 8C). The observed bias was not due to a granulocyte-specific toxicity of the inhibitor, as the total number of colonies remained unchanged (Figure 8—figure supplement 1C).

Our findings suggest that Carm1 modulates the directionality of differentiation of GMPs, with low activities increasing the proportion of monocytic colonies while reducing the proportion of granulocytic colonies.

The finding that C/EBPα^R35A interacts more strongly with PU.1 than C/EBPα^WT could represent an artifact of overexpression. Therefore, to investigate whether the ability of endogenous C/EBPα to interact with PU.1 can be modulated by the levels or activity of Carm1, we treated the macrophage cell line RAW 264.7 for 24 hr with 0, 1, 5, or 10 µM TP064 and subjected the cells to a PLA, using antibodies against C/EBPα and PU.1. Our results showed a four- to fivefold increase in the number of nuclear dots in treated compared to untreated cells (Figure 8D), indicating that the inhibition of methylation of endogenous C/EBPα increased the factor’s affinity for interaction with PU.1. The finding that C/EBPα and PU.1 are expressed at similar levels in treated and untreated cells indicates that the inhibitor does not alter the proteins’ expression and thus could not explain the observed PLA differences (Figure 8—figure supplement 1D).

Our observations indicate that endogenous C/EBPα can interact with PU.1 and that this interaction is regulated by Carm1. Thus, inhibition of the enzyme results in a macrophage cell line induces an enhanced interaction between C/EBPα and PU.1, similar to the interaction of exogenous C/EBPα^R35A with endogenous PU.1.

As a source for the primary B cells used in our experiments, we used C57BL/6J mice. The number of female and male mice used was balanced. Mice were housed in standard cages under 12 hr light-dark cycles and fed ad libitum with a standard chow diet. All experiments were approved by the Ethics Committee of the Barcelona Biomedical Research Park (PRBB) and performed according to Spanish and European legislation.

Primary B cells were isolated from the bone marrow with a monoclonal antibody to Cd19 (BD Biosciences, Cat#553784) using MACS sorting technology (Miltenyi Biotech) as previously described (Di Stefano et al., 2016). They were cultured on gelatinized plates containing S17 feeder cells in RPMI culture medium (Gibco, Cat#12633012) containing 20%-FBS (Gibco, Cat#10270-106), 100 U/mL penicillin- 100 ng/mL streptomycin (Gibco, Cat#15140122), 2 mM L-glutamine (Gibco, Cat#25030081) and 0.1 mM 2-mercaptoethanol (Invitrogen, Cat#31350010) (further addressed as mouse B cell medium), which was further supplemented with 10 ng/mL of IL-7 (Peprotech, Cat#217-17). B cells expanded for 3 days correspond to pre-B cells. 293T human embryonic kidney cells, NIH3T3 fibroblasts (ATCC #173), RAW 2647 macrophages (ATCC #TIB-71), and mouse embryo fibroblasts were cultured in 10% FBS (Gibco, Cat#10270-106) DMEM (Gibco, Cat#12491015) medium. The final culture medium also contained 100 U/mL penicillin and 100 ng/mL streptomycin (Gibco, Cat#15140122), 2 mM L-glutamine (Gibco, Cat#25030081), and 0.1 mM 2-mercaptoethanol (Invitrogen, Cat#31350010) (further addressed as DMEM complete medium). The human BLaER1 B cell line was validated to have originated from the RCH-ACV cell line by STR (DSMZ) and contains the t(1,19) translocation based on HiC, as expected. The line and its derivatives were grown in RPMI culture medium (Gibco, Cat#22400089) containing 20%-FBS (10270-106, Gibco) (further addressed as human B cell medium). All cell lines are mycoplasma free based on periodic broad-spectrum mycoplasma testing.

Induction of transdifferentiation of primary B cells (also referred to as B cells) into macrophages was performed as previously described (Xie et al., 2004). Briefly, isolated B cells were infected with C/EBPα-ER-hCD4 retrovirus, plated at 500 cells/cm² in gelatinized plates (12 wells) onto mitomycin-C (Sigma, Cat#M0503)-treated MEFs (10 μg/mL mitomycin-C for 3 hr to inactivate MEFs). Cells were transdifferentiated in mouse B cell medium, supplemented with 10 ng/mL each of IL-7 (Peprotech, Cat#217-17), IL-3 (Peprotech, Cat#213-13), FLT-3 (Peprotech, Cat#250-31), mCSF-1 (Peprotech, Cat#315-03B), mSCF (Peprotech, Cat#250-03). One hundred nM 17b-estradiol (E2) (Merck Millipore, Cat#3301) was used to shuttle C/EBPαER into the nucleus. B cell culture medium without IL-7 was renewed every 2 days.

Fibroblast transdifferentiation into macrophage experiments were performed as previously described (Feng et al., 2008). Briefly, NIH 3T3 fibroblasts were infected with C/EBPα-ER-IRES-hCD4 wild-type or mutant retrovirus, hCD4 positive cells sorted and a cell line for each construct. Cells were plated at 200,000 cells/mL in gelatinized six-well plates and infected with PU.1∆PEST-IRES-GFP retrovirus. After 24 hr cells were re-plated at 30,000 cells/mL in gelatinized 24-well plates in DMEM complete medium supplemented with IL-3 (Peprotech, Cat#213-13) mCSF-1 (Peprotech, Cat#315-03B) and 100 nM E2 to induce C/EBPα, followed by FACS analysis for Mac-1 expression.

C/EBPα-induced transdifferentiation of BLaER and RCH-ACV cell lines was performed as previously described (Rapino et al., 2013). Briefly, RCH-ACV cells grown in human B cell medium were infected with C/EBPα-ER-IRES-GFP retroviruses (wild-type or mutant) and single GFP-positive cells sorted to generate BLaER2 and BLaER2-A lines. These lines were then infected with rtTA-puromycin retroviruses and selected with 1 µg/mL of puromycin for 1 week. Resistant cells were further infected with pHAGE-TetO-Carm1-IRES-dTomato lentiviruses. Cells were induced with 2 μg/mL of Dox (Sigma, Cat#D9891), tomato-positive cells sorted, and single-cell-derived lines RRC3 and RAC1 established. For the induction of transdifferentiation cells grown in human B cell medium supplemented with 10 ng/mL each of IL-3 (Peprotech, Cat#200-03), CSF-1 (Peprotech, Cat#315-03B) were treated with 2.5 μM 4-OHT (Sigma, Cat#H7904).

Bone marrow-derived GMPs from C57BL/6J mice were isolated by FACS, cultured in Methocult GF M3434 (03434, Stem Cell Technologies) for 14 days and scored microscopically.

293T cells were transfected by using either calcium phosphate (see below) or polyethylenimine according to the manufacturer’s protocol (Polysciences, Cat#24765-2).

Lentiviruses were produced after transfecting 293T cells with 20 μg of lentiviral vector plus 6 μg of pCMV-VSV-G, 15 μg of pCMVDR-8.91, using calcium phosphate. Briefly, calcium phosphate-DNA precipitates were prepared by pooling the three plasmids in a 2.5 M CaCl₂ aqueous solution. While vortexing, one volume of HBS 2× (HEPES-buffered saline solution pH = 7.05, 280 mM NaCl, 0.05 M HEPES, and 1.5 mM Na₂HPO₄) was added dropwise to an equal volume of the calcium phosphate-DNA solution. The mixture was incubated for 15 min at room temperature and added dropwise to 293T cells grown in DMEM complete medium onto gelatin-coated 100 mm dishes. After 8 hr of incubation at 37°C, the transfection medium was replaced with fresh medium and the supernatant collected after 24 hr. The medium was replaced again, and a second round of supernatant was collected after another 24 hr and mixed with the previous batch. The combined supernatants were centrifuged for 5 min at 300 relative centrifugal force (rcf) and filtered through 0.45 μm strainers to remove cell debris. Lentiviral particles were then concentrated by centrifugation for 2 hr at 20,000 relative centrifugal force (rcf) (Beckman Coulter, Optima L-100K) in round bottom polypropylene tubes (Beckman Coulter, Cat#326823). After discarding the supernatants, the lentiviral pellets obtained from one 150 mm dish were thoroughly re-suspended in 80 μL of PBS. 10⁶ fresh cells were then collected in 900 μL of the respective culture medium and 10 μL of lentiviral suspension were added. Subsequently, the virus-cell mixture was centrifuged at 1000 rcf for 2 hr at 32°C (Beckman Coulter, Allegra X-30R). Infected cells were then cultured as described above and subsequently FACS-sorted for the establishment of clonal cell lines.

For the production of retroviruses used to infect mouse cells (Bussmann et al., 2009) we transfected platinum E cells (Cell Biolabs, Cat#RV-101) with the respective constructs. Platinum A cells (Cell Biolabs, Cat#RV-102) were transfected for use in the infection of human cells as previously described (Di Stefano et al., 2014).

To inhibit Carm1 enzymatic activity we used TP064 (Bio-Techne RD Systems, Bristol, UK) as previously described (Nakayama et al., 2018). For experiments with B cells, these were pre-incubated with 5 µM of TP046 24 hr prior to induction with E2, and treatment with the inhibitor continued during the time of induction. For the colony forming assay with GMPs, 1–10 µM of TP064 was added to the medium at the time of plating.

Mouse bone marrow cell extraction was performed as previously described (Di Stefano et al., 2014). Briefly, femurs and tibias of C57BL/6J mice were extracted and crushed on a mortar in PBS supplemented with 4%FBS and 2 mM EDTA and filtered through 0.45 μm strainers (Merck Millipore, Cat#SLHV033RB). For B cells, bone marrow-derived cells were incubated with sequentially 0.1 µg per 1 million cells of both Fc block and CD19-biotin antibody for 10 and 20 min, respectively, followed by 10 µL of magnetic streptavidin microbeads (Miltenyi, Cat#130-048-101) for an additional 20 min. CD19+ cells were sorted using LS columns (Miltenyi, Cat#130-042-401). For BMT CD19+ B cells were infected with C/EBPα-ER-IRES-hCD4 (wild-type and mutants) and cultured over MEF feeder cells for 4 days. Cultured B cells were incubated sequentially with 0.1 µg per 1 million cells of both Fc block and hCD4-biotin antibody for 10 and 20 min, respectively, followed by 10 µL of magnetic streptavidin microbeads (Miltenyi, Cat#130-048-101) for an additional 20 min. hCD4+ cells were enriched with LS columns (Miltenyi, Cat#130-042-401).

For GMPs, bone marrow cells were lineage-depleted using a Lineage Cell Depletion Kit (Miltenyi, Cat#130-090-858). Lineage negative cells were then incubated with CD34-APC, cKit-APC-Cy7, Sca1-PE-Cy7, and CD16/32-FITC for 1.5 hr. Sca1- cKit+ CD34+ CD16/32+ cells (GMPs) were sorted using either FACS Aria or Influx cell sorters. For granulocytes and monocytes, bone marrow-derived cells were incubated sequentially with 0.1 µg per 1 million cells of both Fc block and Mac1-biotin antibody for 10 and 20 min, respectively, followed by 10 µL of magnetic streptavidin beads (Miltenyi, Cat#130-048-101) for an additional 20 min. Mac1+ cells were sorted using LS columns (Miltenyi, Cat#130-042-401) and incubated with Mac1-PE and Ly6g-APC for 20 min. Mac1+ Ly6g- (monocytes) and Mac1+ Ly6g+ cells (granulocytes) were sorted using either FACS Aria or Influx cell sorters.

For 3T3 NIH fibroblasts cells infected with C/EBPα-ER-IRES-hCD4 (WT and R35A) were incubated with 0.1 µg per 1 million cells of both Fc block and hCD4-biotin antibody (BD Pharmingen, Cat#555347) for 10 and 20 min, respectively, followed by 10 µL of magnetic streptavidin beads (Miltenyi, Cat#130-048-101) for an additional 20 min. hCD4+ cells were purified using LS columns (Miltenyi, Cat#130-042-401).

For B lymphoblastic leukemia cells (RCH-ACV) cells stably infected with C/EBPα-ERT2-IRES-GFP, rtTA-Puro and TetO-Carm1-IRES-TdTomato were induced with 1 µg/mL of Dox (Sigma-Aldrich, Cat#D9891). GFP+ and TdTomato+ cells were single cell-sorted using either FACS Aria or Influx cell sorters.

In co-cultures between B cells and feeder cells, non-adherent cells were collected, joined with trypsinized adherent cells, and centrifuged at 300 rcf for 5 min. Cells were re-suspended in 100 μL PBS containing 1 μg/mL of mouse Fc block for 10 min. Conjugated primary antibodies were added to the blocking solution and cells were further incubated at 4°C in the dark for 20 min. Cells were washed with additional 1 mL of PBS and centrifuged at 300 rcf for 5 min. The supernatant was discarded and cells were re-suspended in 500 μL of PBS containing 5 μg/mL of DAPI. Samples were processed in a FACS analyzer (LSR II, BD; Fortessa, BD) with DiVa software and data analyzed using FlowJo software. Antibodies used for cell sorting and flow cytometry are listed in Supplementary file 3.

After BMT, cells were removed from feeder cells through differential adherence to tissue culture dishes for 40 min. Around 200,000 of the resulting B cells (or induced macrophages) were plated in each well of a 24-well plate containing 0.01% poly-L-lysine-treated coverslips (Corning, Cat#354085) in 10% FBS-DMEM supplemented with IL-3 (Peprotech, Cat#213-13), mCSF-1 (Peprotech, Cat#315-03B) and cultured at 37°C overnight in the presence of 1:1000 diluted blue fluorescent carboxylated microspheres (Fluoresbrite, Cat#17458-10). Cells were centrifuged at 1000 rcf for 5 min to improve attachment to the coverslips. The supernatant was removed and the cells were washed once with PBS.

For fixation, 4% PFA was added to the wells for 20 min, cells were washed twice with PBS and cell membranes permeabilized with 0.1% Triton X-100 PBS (0.1% PBST) for 15 min at room temperature. Cells were blocked using 0.1% PBST with 3% bovine serum albumin (BSA) for 30–45 min. Cells were washed twice in PBS. Actin filaments were subsequently stained with 1:100 diluted red phalloidin (Alexa Fluor 568, Thermo Fisher Scientific, Cat#A12380) while DNA was stained with a 1:500 diluted yellow probe (Quant-iT PicoGreen dsDNA Assay Kit, Thermo Fisher Scientific, Cat#P7589). Cells were incubated with the two dyes in 0.1% PBST containing 1% BSA at room temperature for 1 hr in the dark and washed twice with PBS afterward. Coverslips carrying the attached cells in the well were then recovered with tweezers and mounted upside-down onto a charged glass slide containing a 14 μL drop of mounting medium (7 μL Dako+7 μL 0.1% PBST). Coverslips were sealed with nail polish and imaged in a Leica TCS SPE inverted confocal microscope.

C/EBPα- and C/EBPβ-deficient mouse v-abl-transformed B cells were transduced with pRetroX vectors containing Dox-inducible wild-type C/EBPα or R35A C/EBPα constructs C-terminally fused to TurboID or only TurboID ligase as a control (Branon et al., 2018). Cells were selected with 1 µg/mL puromycin for 2 weeks in DMEM supplemented with 10% charcoal-treated FBS, 1% penicillin/streptomycin, 1% HEPES, and 50 mM β-mercaptoethanol. One µg/mL Dox was added for 6 hr and proximity labeling performed by addition of 500 µM biotin for 30 min. Cells were then harvested by centrifugation, washed three times in ice-cold PBS, lysed, biotinylated proteins pulled down and subjected to LC-MS/MS as previously described (Ramberger et al., 2021b).

Raw data was analyzed using MaxQuant (version 1.6.3.3) against mouse proteome (Uniprot, 2019 version) with isoforms using standard settings. Additional options included fast LFQ, ‘match between run’, and lysine biotinylation (Cox et al., 2014). The data were analyzed using R-statistical software package (version 3.6.1). The protein files from MaxQuant were filtered for contaminants, reverse hits and proteins identified only by site. Intensity values were log2 transformed and filtered to allow only proteins with valid values in all four replicates in at least one of the experimental groups. Remaining missing values were imputed from normal distribution with width of 0.3× standard deviation and downshift of 1.8 from an observed mean. Significant interactors were defined as >2-fold enriched against TurboID and no-Dox controls and to pass the significance cutoff (adj. p-value <0.05) in both t-tests against the two controls. Multiple testing correction was applied using the Benjamini-Hochberg procedure. Differential interactors were scored based on the mean log2FC (WT/R35A)±3 s.d. cutoff. Data from each experimental group (C/EBPα^WT and C/EBPα^R35A at 6 hpi) were obtained from four replicates.

PLA was performed using Duolink Orange Kit (Sigma-Aldrich, Cat#DUO92007). Briefly, after sorting or culturing desired cell populations, 8000–100,000 cells per well were seeded into 24-well plates containing 0.01% poly-L-lysine (Sigma) treated coverslips in appropriate medium, centrifuged at 1000 × g for 5 min and fixed with 4% PFA for 15 min. Subsequent steps were performed according to the kit’s protocol with antibody concentrations identical to those used for immunofluorescence. Coverslips were mounted using Fluoroshield mounting medium with DAPI (Abcam, Cat#ab104139) and imaged in a Leica TCS SPE confocal microscope.

After antibody staining of cell surface markers, cells were fixed in 4% BSA for 10 min at room temperature in a rotating wheel. Fixation was stopped with two washes in PBS. Cells were permeabilized in 0.1% PBST at room temperature in a rotating wheel for 10 min. Cells were blocked using 0.1% PBST with 3% BSA for 30–45 min. Cells were washed twice in PBS. Cells were incubated with primary antibodies and secondary antibodies diluted at the stated concentrations in 0.1% PBST with 1% BSA for 2 and 1 hr, respectively, with two washes in PBS in between and after. Cells were resuspended in PBS and processed in a FACS analyzer (LSR II, BD; Fortessa, BD) with DiVa software and data analyzed using FlowJo software.

Antibodies used for cell surface immunofluorescence and intracellular staining for flow cytometry are listed in Supplementary file 1.

Preparation of whole cell lysates and immunoprecipitation of wild-type or mutant C/EBPα proteins were performed as previously described (Kowenz-Leutz et al., 2010). Briefly, cells were lysed (20 mM HEPES pH 7.8, 150 mM NaCl, 1 mM EDTA pH 8, 10 mM MgCl₂, 0,1% Triton X-100, 10% glycerol, protease inhibitor cocktail [Merck], 1 mM DTT, 1 mM PEFA bloc [Böhringer]). Immunoprecipitation was performed with appropriate antibodies as indicated for 2 hr at 4°C. Immunoprecipitated proteins were collected on Protein A Dynabeads (Invitrogen, Cat#100001D) or Protein-G Dynabeads (Invitrogen, Cat#10004D), separated by SDS-PAGE (Mini PROTEAN TGX, 4–15%, Bio-Rad #5671084). For western blotting, samples were loaded in 10% Mini-PROTEAN TGX gels (Bio-Rad) and resolved by electrophoresis in running buffer (Supplementary file 2). Protein samples were transferred to a methanol pre-activated PVDF membrane (Bio-Rad, Cat#1620177) by running them in transfer buffer (TBS) (Supplementary file 2) for 1 hr at 300 mA and 4°C. Membranes were rinsed in milliQ water and protein transfer was checked by Ponceau staining (Sigma). Transferred membranes were washed once with TBS and three times with TBS-Tween (TBST) followed by 5% milk in TBST for 45 min. Membranes were then incubated with primary antibodies (Supplementary file 3) in 5% milk TBST, rotating overnight at 4°C, then washed three times with TBST followed by incubation with the secondary antibodies conjugated to horseradish peroxidase in 5% milk TBST for 1 hr. After three TBST washes, proteins were detected using enhanced chemiluminescence reagents (Amersham ECL Prime Western Blotting detection) in an Amersham Imager 600 analyzer or visualized by ECL (GE Healthcare, UK). Quantification of band intensity from scanned blots was performed with Fiji software.

Nuclear extracts were prepared from transfected 293T cells by a mininuclear extract protocol (Schreiber et al., 1989). EMSA were performed as previously described (Kowenz-Leutz et al., 1994) using double-stranded IRDye Oligonucleotides containing a C/EBP-binding site: IRD800-GACACTGGATTGCGCAATAGGCTC and IRD800-GAGCCTATTGCGCAATCCAGTGTC (Metabion). Briefly, binding reactions with nuclear extracts (2.5 µg) and double-stranded IRD800 oligos (20 pmol) were incubated for 15 min on ice, orange loading dye (Li-Cor, Cat#P/N 927-10100) was added and protein-DNA complexes were separated on a 5% native polyacrylamide gel in 0.5× TBE at 25 mA at room temperature. EMSA results were visualized and quantified (Odyssey scanner, Licor, channel 800 nm).

Methylation of peptides (PSL, Heidelberg, Germany, Supplementary file 4) was performed using the bioluminescence-based MTase-Glo Assay (Promega, Cat#V7601) according to the manufacturer’s protocol. Assay conditions: 200 ng of enzyme was incubated with 5 μM peptide, 10 μM S-adenosyl-L-(methyl)-methionine as methyl donor (SAM) and 6× methyltransferse-Glo reagent at 23°C for 60 min. S-adenosylhomocysteine generated during the reaction was converted to ADP as a proportional reaction product dependent of substrate methylation by the enzymes. Subsequent incubation with the Methyltransferase-Glo Detection Solution at 23°C for 30 min converts ADP to ATP that is used in a luciferase/luciferin-based reaction and determined as relative light units in a Berthold luminometer (Hsiao et al., 2016).

RNA was extracted with a miRNeasy mini kit (217004, QIAGEN), quantified with a NanoDrop spectrophotometer and its quality examined in a fragment Bioanalyzer (Aligent 2100 Bioanalyzer DNA 7500 assay). CDNA was synthesized with a High-Capacity RNA-to-CDNA kit (4387406, Applied Biosystems). For RNA-seq, libraries were prepared with a TruSeq Stranded mRNA Library Preparation Kit (Illumina) followed by single-end sequencing (50 bp) on a HiSeq2500 instrument (Illumina), obtaining at least 40 million reads per sample.

Quality control of FASTQ reads was performed using FastQC version v.0.11.3. Reads were mapped aligned to the mm10 genome using STAR version 2.5.0a (Dobin et al., 2013). Gene expression was quantified using STAR (--quantMode GeneCounts). Normalized counts and differential gene expression analysis were carried out using DESeq2 version 1.14.1 (Love et al., 2014). For each transdifferentiation experiment, timepoint 0 hr was set as a reference point and any gene that exhibited a statistically significant change in expression (log2FC ≥0.5849625 and p-value ≤0.05) at a later timepoint was isolated. For PCA, log2 DESeq2 normalized counts of DEGs averaged across replicates were used. The R prcomp() command with scale = T was used. Pheatmap version 1.0.12 was used to visualize changes in gene expression for all the isolated DEGs with the following clustering options: clustering_distance_rows=‘correlation’, clustering_method=‘ward.D2’, scale=‘row’.

DEGs were determined for each timepoint as described in the ‘Materials and methods’. The union of identified DEGs in the WT and R35A systems per timepoint were used to generate scatterplots depicting the log2FC changes of the aforementioned genes for each transdifferentiation system. A regression line, colored in red, was fit for each scatterplot using the geom_smooth(method = lm) R command. The identity line (y=x line) is depicted in green. The spearman correlation coefficient (cor(method = ‘spearman’) function in R) and the number of DEGs are also depicted per scatterplot.

Functional analyses by GO were performed with the R package ‘g: profiler2’ version 0.2.0 (Raudvere et al., 2019). Baloonplots depict all pathways associated with a specific keyword that were found enriched in at least one cluster. Metaplots for each cluster depict the average log2FC values of genes per timepoint and per cluster. Shaded background corresponds to the mean values ± 1.644854 s.d. Gene expression analysis of signature genes was performed using the individual values of the B cell genes Pax5, Ebf1, Foxo1, Ikzf1, Rag1, Rag2, Bcl11a, Spib, Ikzf3, CD2, CD19, Igll1, Vpreb1, Vpreb2, Vpreb3, Pou2a1, Blk, CD79a, CD79b, Lef1 and the macrophage genes C1qc, Fcer1g, Sell, Ccr1, Mitf, Tlr2, Csf1r, Trem2, Fam20c, Adam8, Batf2, Fes, Itgam, Ccl3, CD300lf, Tnsf9, Tyrobp, CD14, Ifitm6, Csf3r , normalized to timepoints o hr for B cell genes and 120 hr for macrophage genes.

ATAC-seq was performed as published (Buenrostro et al., 2015). Briefly, cells were harvested at the mentioned timepoints, feeder-depleted and lysed and 50,000 cells used per condition. Immediately, transposition was performed using Nextera Tn5 Transposase (15027865, Illumina) at 37°C for 30 min. Chromatin was then purified using QIAGEN MinElute PCR Purification Kit (28004, QIAGEN). DNA was then amplified using NEBNext High Fidelity PCR Master Mix (M0541S, New England Biolabs Inc) and barcoded primers. qPCR was performed to determine the optimal number of cycles for each condition to stop amplification prior to saturation. Quality was analyzed by gel electrophoresis and in a fragment Bioanalyzer (Agilent 2100 Bioanalyzer DNA 7500 assay).

Read quality was assessed with FastQC version v.0.11.3. Adaptors were removed using Cutadapt (version 0.4.2_dev) TrimGalore! In paired end mode (--paired –nextera) (Martin, 2011). Reads were aligned to the mm10 genome using bowtie2 (v 2.2.4) in paired end mode with standard parameters. Output SAM files were converted to BAM files using samtools (v 0.1.19) (Li et al., 2009). BAM files were sorted and indexed using the samtools commands sort and index, respectively. Low-quality reads and reads associated with a not primary or supplementary alignment SAM flag were filtered out using the samtools command ‘samtools view -F 2304 -b -q 10’. PCR duplicates were removed with Picard MarkDuplicates (version 2.3.0) with the following options: ‘REMOVE_DUPLICATES = true ASSUME_SORTED = true VERBOSITY = WARNING’.

Filtered BAM files were indexed with samtools index and were used as input in the bamCoverage command of deeptools (v3.0.1) (Ramírez et al., 2016) in order to generate bigwig files. bamCoverage was used with the options –binSize 1 –normalizeUsing RPGC –effectiveGenomeSize 2150570000 –extendReads –outFileFormat bigwig. In order to call peaks, bam files of each timepoint and experiment were merged using the samtools merge command. Resulting merged bam files were indexed, and peaks were called using MACS2 with the options -f BAMPE –nolambda –nomodel -g mm -q 0.05.

In order to pinpoint regions of interest, peaks of all timepoints and all experiments were merged using the bedtools suite command bedtools merge. Read counts falling within the merged peak regions were calculated using the Rsubread package and the featurecounts command with the options isPairedEnd = T, strandSpecific = 0, useMetaFeatures = F. For each transdifferentiation experiment, DESeq2 was used in order to compare all timepoints with timepoint 0 hr. Any peak showing a logFC ≥1, an adjusted p-value ≤0.05, and average counts across timepoints ≥5 was termed as a differentially accessible region and was kept for further analyses. The total number of peaks isolated was 91,830. Variance stabilized counts were calculated for the isolated regions using the DESeq2 command varianceStabilizingTransformation and the options ‘blind = T’’, fitType=‘’parametric’’. Variance stabilized counts were averaged across timepoint replicates by raising them at the power of 2, extracting their average and log2 transforming the resulting mean. PCA was applied to this dataset using the R prcomp command, with ‘scale = T’.

To group peaks, PCA was initially applied and PC1 and PC2 values for the 91,830 regions were used in order to arbitrary cluster peaks into three groups depending on the sign of their PC1 and PC2 values. Values for each of the three groups were visualized using the pheatmap package. Visual examination of the three main groups showed different trends: peaks whose accessibility higher at 120 hr (43,429 peaks), lower at 120 hr (36,380 peaks), and higher at 18 hr (12,021 peaks).

Peaks from the three different groups were centered and extended 50 bp upstream and downstream. Nucleotide sequences for each centered peak were extracted using bedtools getfasta. Sequences were submitted into MEME-ChIP with the following parameters: -dna -seed 49 -meme-nmotifs 20 -meme-minw 5 -meme-minsites 2 -meme-minw 4 -meme-maxw 12. TOMTOM was run using the output meme.txt file in order to identify matches of known TF motifs to the de novo discovered motifs. For each TOMTOM output a series of additional filtering steps were undertaken:

Genomic coordinates of mm10 genes were downloaded from the UCSC table browser (RefSeq genes). A single promoter region was assigned to each gene. The region consisted of 1 kb upstream and downstream of the transcription start site of the largest transcript of each gene. Counts for each timepoint and each transdifferentiation experiment were assigned to each promoter as described above. DESeq2 was used in order to identify differentially accessible promoters as described above with the following differences regarding the cutoffs used: -foldChange ≥ 1.5 and p-value ≤ 0.05. Variance stabilized counts were extracted for each differentially accessible promoter, a mean value per replicate was extracted, and the values were plotted using pheatmap. Promoters were then grouped into eight clusters. Baloonplots depict all pathways associated with a specific keyword that were found enriched in at least one cluster.

For each promoter cluster and each promoter, log2FC changes were extracted by comparing expression levels (DESeq2 normalized counts) of every timepoint with the corresponding timepoint 0 hr of the experiment.

Thirty-thousand NIH 3T3 cells inducible for C/EBPα^WT-Halo or C/EBPα^R35A-Halo (3T3aER-R and 3T3aER-A, respectively) were seeded in eight-well plates (Lab-Tek 155049), and induced for 6 or 24 hr with 1 µg/mL Dox, with or without prior infection for 24 hr with TetO-FUW-PU.1 lentivirus. Right before imaging, cells were treated with 5 nM of Janelia Fluor 549 (JF549) Halo Tag ligand (a kind gift from Luke Lavis, HHMI) for 15 min. Cells were subsequently washed three times in PBS at 37°C, and GlutaMAX medium (Thermo Fisher, Cat#21063029) was added to each well. All imaging was carried out with a Nanoimager S super resolution microscope (Oxford Nanoimaging Limited) under highly inclined thin illumination (HILO) conditions (Tokunaga et al., 2008). A scientific complementary metal oxide semiconductor camera was used with a 2.3 electrons rms read noise at standard scan, a 100×, 1.49 NA oil immersion objective, and a 561 nm green laser. For imaging experiments, one frame was acquired with a 100 ms exposure time at 10 Hz to measure the intensity of fluorescence of the nuclei, and for SMT experiments, 5000 frames were acquired with an exposure of 10 ms (100 Hz).

Quantification and statistical analysis of SMT was performed as previously described (Lerner et al., 2020). In brief, TIF stacks SMT movies were analyzed using MATLAB-based SLIMfast script (Teves et al., 2016), a modified version of MTT (Sergé et al., 2008), with a maximal expected diffusion coefficient (DMax) of 3 μm²/s. The SLIMfast output.txt files were reorganized by the homemade csv converter.m MATLAB script (available in Lerner et al., 2020) in.csv format for further analysis. The single molecule tracking.csv files (see previous section) were first classified by the homemade SMT_Motion_Classifier.m MATLAB script. Single molecule trajectories (or tracks) with a track duration shorter than five frames were discarded from the analysis. Motion tracks are classified by the script in different groups: tracks with α≤0.7 were considered as Confined; motion tracks with 0.7 < α < 1 as Brownian; and motion tracks with α ≥ 1 as Directed. In addition, the motion tracks showing a behavior similar to a levy-flight (presenting mixed Confined and Directed/Brownian behavior) were detected by the presence of a jump superior to the average jump among the track + a jump threshold of 1.5, and classified as ‘Butterfly’. Butterfly motion tracks were segmented into their corresponding Confined and Directed/Brownian sub-trajectories for posterior analysis. As an additional filtering step of confined motions (including confined segments of Butterfly tracks), we defined a jump threshold of 100 nm, to filter out motion tracks with an average frame-to-frame jump size larger than 100 nm. All scripts are publicly available.

To obtain samples for the ChiP-seq experiments biological duplicates of primary B cells expressing either C/EBPα^WT ER or C/EBPα^R35A ER were induced for 3 hr with E2 and 100,000 cells per sample processed by ChIPmentation, as described (Schmidl et al., 2015) (see also: https://www.nature.com/articles/nmeth.3542). Three µL of anti-C/EBPa antibody (Santa Cruz sc-61 14AA) or anti-PU1 (Santa Cruz sc-352 T21) were used per immunoprecipitation. Tagmentation of immobilized C/EBPa or PU1-enriched chromatin was performed for 5 min at 37 degrees in a 25 µL transposition reaction mix. Libraries were prepared with an Illumina TruSeq ChIP Library Preparation Kit and after quality control on Bioanalyzer sequenced (1 × 50 bp) on an Illumina HiSeq2500 instrument.

ChIP-seq data were processed similarly to ATAC-seq data (refer to ‘determination of differentially accessible ATAC peaksanalysis of ATAC-seq data’ section) with the following modifications: bamCoverage was used with: --binSize 1 --normalizeUsing RPKM --effectiveGenomeSize 215057000. Peak calling for each experiment was conducted using macs2 callpeak and the options -f BAMPE -g mm.

Duplicates from C/EBPα^WT and C/EBPα^R35A peaks were merged using bedtools merge. The number of reads per peak were quantified and differentially occupied peaks identified as described in the ‘Differentially accessible ATAC-seq peaks’ section. Peaks with an average of at least 20 reads, p-adjusted ≤ 0.05 and |logFC| ≥ 1 were classified as differentially occupied peaks. The same process was repeated for PU.1 peaks in the C/EBPα^WT and C/EBPα^R35A systems.

Statistical analyses were performed using Prism 9 software. To calculate significance, samples from at least three biologically independent experiments were analyzed. Two biological replicates were used for RNA- and ATAC- sequencing experiments and statistics applied to the expression of a collection of genes. For samples with n≥3, values shown in the figures represent mean ± s.d. Box plots represent median with quartiles and whiskers and individual values are shown. One-way, two-way ANOVA (with the corresponding multiple comparison analyses) and Student’s t-tests were applied accordingly. p-Values appear indicated in each figure only when ≤0.05. In time course experiments, p-values of differences between conditions by two-way ANOVA are shown. In box plots, p-values of each individual timepoint as well as p-values of differences between conditions by two-way ANOVA are shown.

All constructs and cell lines listed can be requested from the corresponding authors.