Co-evolving CENP-A and CAL1 Domains Mediate Centromeric CENP-A Deposition across Drosophila Species

Abstract

Centromeres mediate the conserved process of chromosome segregation, yet centromeric DNA and the centromeric histone, CENP-A, are rapidly evolving. The rapid evolution of Drosophila CENP-A loop 1 (L1) is thought to modulate the DNA-binding preferences of CENP-A to counteract centromere drive, the preferential transmission of chromosomes with expanded centromeric satellites. Consistent with this model, CENP-A from Drosophila bipectinata (bip) cannot localize to Drosophila melanogaster (mel) centromeres. We show that this result is due to the inability of the mel CENP-A chaperone, CAL1, to deposit bip CENP-A into chromatin. Co-expression of bip CENP-A and bip CAL1 in mel cells restores centromeric localization, and similar findings apply to other Drosophila species. We identify two co-evolving regions, CENP-A L1 and the CAL1 N terminus, as critical for lineage-specific CENP-A incorporation. Collectively, our data show that the rapid evolution of L1 modulates CAL1-mediated CENP-A assembly, suggesting an alternative mechanism for the suppression of centromere drive.

Figure 1. Inter-species Centromeric Localization of Drosophila CENP-A Orthologs Can Only Partially Be Explained by Phylogenetic Distance

(A) Left: phylogenetic tree of the Drosophila species analyzed in this study (mel, D. melanogaster; sim, D. simulans; ere, D. erecta; tak, D. takahashii; rho, D. rhopalia; kik, D. kikkawai; ana, D. ananassae; bip, D. bipectinata; pse, D. pseudoobscura; mir, D. miranda; wil, D. willistoni; vir, D. virilis). Divergence time and phylogenetic grouping based on Flybase. Right: schematic of CENP-A orthologs from indicated species showing relative differences in protein size. The N terminus is shown in black and the histone fold domain (HFD) is shown in gray. Loop 1 (L1) is shown shades of aqua indicative of divergence in L1. Numbers indicate amino acid positions. (B) Western blots with anti-GFP (top) and anti-lamin (loading control, bottom) antibodies of total cell extracts showing the expression of GFP-CENP-A orthologs in S2 cells used in (C) and (D). The expression of vir GFP-CENP-A was too low to visualize by western blot. (C) Immunofluorescence (IF) images of mel S2 interphase cells transiently expressing Drosophila GFP-CENP-A orthologs. Images where endogenous CENP-A is present were chosen to visualize the location of the centromere. DAPI is shown in gray, GFP in green, and mel CENP-A in red. Zoomed insets show representative centromeres with merged colors. (D) Quantification of (C). Images were manually classified as having either centromeric localization of GFP (gray bars), diffuse localization of GFP (red bars), or centromeric/diffuse (orange bars). Number of transfected cells quantified for each ortholog: 97 for mel, 114 for sim, 94 for ere, 58 for tak, 67 for rho, 36 for kik, 52 for ana, 143 for bip, 90 for pse, 212 for mir, 68 for wil, and 35 for vir. Fisher's two-tailed test p values were ***p < 0.0001 for bip CENP-A, wil CENP-A, and vir CENP-A; **p = 0.0003 for ana; **p = 0.005 for kik; and *p = 0.002 for pse CENP-A compared with mel CENP-A. These data were confirmed by one biological replicate with GFP-tagged constructs, and two additional replicates with HA-tagged constructs. See also Figures S1–S3.

Figure 2. D. melanogaster CAL1 Cannot Recruit D. bipectinata CENP-A to an Ectopic Locus

(A) Western blots of IPs with anti-CAL1 antibodies from nuclear extracts transiently expressing mel GFP-CENP-A (top) or bip GFP-CENP-A (bottom). IP was confirmed using anti-CAL1 antibody (top blot). Presence of GFP-CENP-A in CAL1 pull-downs was detected with anti-GFP antibody (bottom blot). Shown is the percentage of immunoprecipitated GFP-CENP-A relative to input. (B) Representative IF images of metaphase chromosome spreads from mel S2 lacO cells transiently co-expressing mel CAL1-GFP-LacI and HA-CENP-A orthologs: mel (top); ere (second); bip (third); pse (fourth); and wil (bottom). Chromosome spreads were quantified for the presence of HA-CENP-A at the lacO site (percentage shown in right column). Endogenous mel CENP-A is shown in red, HA in aqua, GFP in green, and DAPI in gray. Note that mel CENP-A antibodies are specific for this species and that upon expression of CENP-A orthologs that localize to mel centromeres, endogenous CENP-A levels decrease (e.g., ere CENP-A; see also Figure S2). This is not observed for mel HACENP-A, as mel CENP-A antibodies recognize this tagged protein. ***p < 0.0001 (Fisher's two-tailed test) for bip or wil CENP-A compared with mel CENP-A recruitment at the lacO. n = 13 spreads for mel CENP-A recruitment, 8 for ere, 11 for bip, 15 for pse, and 8 for wil. Yellow arrowheads indicate the lacO array. These results were confirmed by one biological replicate with the CAL1-GFP-LacI construct, and two biological replicates with GFP-CENP-A and CAL1-LacI constructs (data not shown). (C) Western blots with anti-GFP (top) and anti-fibrillarin (loading control, bottom) antibodies of whole-cell extracts showing the expression of induced mel CAL1-GFP-LacI in lacO cells shown in (B).

Figure 3. Co-expression of CAL1 and CENP-A Ortholog Pairs Rescues Centromeric Localization

(A) Schematic of the experiments testing whether co-expression of bip CAL1 and bip CENP-A can result in the localization of bip CENP-A to mel centromeres. (B) Western blots of whole-cell extracts showing expression of HA-CAL1 constructs. Top: anti-HA. Bottom: anti-tubulin (loading control). (C) Representative IF images of interphase mel S2 cells transiently expressing HA-CAL1 orthologs. DAPI is shown in gray, HA in red, and mel CENP-A in green. The percentage of cells with centromeric HA signal is indicated in the middle column. Zoomed panels show representative centromeres with merged colors. (D) Representative IF images of metaphase chromosome spreads from S2 cells transiently expressing bip or wil GFP-CENP-A alone (first and third rows, respectively), bip GFP-CENP-A and bip HA-CAL1 (second row), or wil GFP-CENP-A and wil HA-CAL1 (fourth row). DAPI is shown in gray, GFP in green, HA in aqua, and mel CENP-A in red. White arrowheads indicate the position of the centromere. (E) Quantification of the IF shown in (D). Chromosome spreads were manually classified as having either centromeric (gray bars), diffuse (red), or centromeric/ diffuse (orange) GFP signal. n = 29 spreads for bip CENP-A alone, 73 for bip CENP-A with bip CAL1, 27 for wil CENP-A alone, and 51 for wil CENP-A with wil CAL1. ***p < 0.0001 (Fisher's two-tailed test) for the centromeric localization of bip and wil CENP-A with and without CAL1. These data were confirmed by one biological replicate using HA-tagged CAL1 constructs (data not shown), and two biological replicates using CAL1-GFP-LacI and HA-CENP-A constructs (see F in this figure for bip, and data not shown for wil). See also Figures S4 and S5. (F) Representative IF images of metaphase chromosome spreads from mel lacO S2 cells transiently co-expressing mel, bip, or ana CAL1-GFP-LacI (first, second, or third row, respectively), and bip HA-CENP-A. GFP is shown in green, HA in aqua, mel CENP-A in red, and DAPI in gray. Yellow arrowheads indicate the position of the lacO site. (G) Quantification of the images shown in (F). Cells were manually classified as having either exclusively centromeric GFP signal (gray bars), diffuse GFP signal (red bars), or centromeric and diffuse GFP signal (orange bars). n ≥ 30 cells per condition. These data were confirmed by two biological replicates (data not shown). ***p < 0.0001 (Fisher's two-tailed test). (H) Western blots with anti-GFP (top) and anti-tubulin (loading control, bottom) antibodies of whole-cell extracts showing the expression of induced bip and ana CAL1-GFP-LacI in lacO cells in F. (I) Representative IF images of metaphase chromosome spreads from mel lacO S2 cells transiently co-expressing bip CAL1-GFP-LacI and bip HA-CENP-A (aqua) showing the lacO recruitment of endogenous mel CENP-A (red) and the outer kinetochore protein Ndc80 (green). GFP fluorescence was quenched with 100% ethanol. DAPI is shown in gray. Percentage of CENP-A positive (bip, or mel and bip) lacO arrays with Ndc80 is indicated in the right column. Yellow arrowheads mark the position of lacO site. These data represent the average of three experiments (two technical replicates and one biological replicate). n = 40 HA-positive cells.

Figure 4. CAL1 Recognizes CENP-A via L1

(A) Schematic of mel CENP-A construct with bip L1 substituted into the mel HFD (mel GFP-CENP-A^bipL1). Blue represents the mel CENP-A protein sequence; bip CENP-A amino acids are indicated in purple. (B) Western blots of whole-cell extracts showing the expression levels and size of GFP-mel CENP-A^bipL1 chimera compared with GFP-mel CENP-A and GFP-bip CENP-A. Top: anti-GFP; bottom: anti-lamin (loading control). (C) IF images of metaphase spreads from S2 cells transiently expressing GFP-mel CENP-A^bipL1 chimera alone, or co-expressed with bip CAL1. DAPI is shown in gray, GFP in green, and mel CENP-A in red. White arrowheads indicate position of the centromere. n = 9 for mel CENP-A^bipL1 chimera alone and n = 10 for mel CENP-A^bipL1 chimera with bip CAL1. The percentage of cells with centromeric GFP signal is as indicated in the middle column. ***p < 0.0001; Fisher's two-tailed test of cells with compared with cells without bip CAL1. These data were confirmed by two biological replicates (data not shown). (D) IF images of interphase S2 cells transiently expressing GFP-mel CENP-A^bipL1 chimera alone or co-expressed with bip CAL1. DAPI is shown in gray, GFP in green, and mel CENP-A in red. Zoomed panels show representative centromeres with merged colors. (E) Quantification of the IF shown in D. GFP-CENP-A^bipL1 chimera localization was classified as centromeric (gray bars), diffuse (red), or centromeric and diffuse (orange). n = 70 cells quantified for mel CENP-A^bipL1 chimera and 83 for mel CENP-A^bipL1 chimera with bip CAL1. Error bars denote the SD of three biological replicates. ***p < 0.0001; Fisher's two-tailed test comparing cells with and without bip CAL1. (F) Schematic of bip CENP-A construct with mel L1 substituted into the HFD (bip CENP-A^melL1; HA-tagged). mel CENP-A residues are represented in blue and bip CENP-A amino acids in purple. HFDs here and in (A) are shown in darker shades of the respective colors. (G) Western blots of whole-cell extracts showing the expression levels and of the bip HA-CENP-A^melL1 chimera compared with mel HA-CENP-A and bip HA-CENP-A. Top: anti-HA; bottom: anti-lamin (loading control). (H) IF images of metaphase chromosome spreads from mel lacO S2 cells transiently expressing mel CAL1-GFP-LacI (green) and HA-tagged bip CENP-A or bip CENP-A^melL1 chimera (aqua). Endogenous mel CENP-A is shown in red, DAPI in gray. Yellow arrowheads indicate the position lacO array. n = 20 for HA-tagged bip CENP-A and 16 for HA bip CENP-A^melL1 chimera. The recruitment efficiency to the lacO site for each HA-tagged construct is indicated at the bottom of the HA panel. ***p < 0.0001; Fisher's two-tailed test. These results were confirmed by two biological replicates (data not shown).

Figure 5. Identification of CAL1 Residues Co-evolving with CENP-A L1

(A) Schematic of N-terminal CAL1-GFP-LacI constructs. For chimeras, gray indicates mel CAL1 and blue indicates bip CAL1 proteins. (B) Western blots of whole-cell extracts showing the expression and sizes of the N-CAL1-GFP-LacI constructs used in these experiments (top: anti-GFP; bottom: anti-lamin loading control). (C) IF images of metaphase chromosome spreads from mel lacO S2 cells transiently expressing the indicated N-CAL1-GFP-LacI constructs from (A), along with bip HA-CENP-A. GFP is shown in green, HA in aqua, mel CENP-A in red, and DAPI in gray. Yellow arrowheads indicate the position of the lacO array. Note that since the N terminus of CAL1 alone cannot localize to centromeres (Chen et al., 2014), these constructs are not expected to deposit bip CENP-A at the endogenous centromere. (D) Quantification of the IF shown in (C). Error bars represent the SD of three biological replicates. n = 160 spreads for mel CAL1 1–407, 82 for bip CAL1 1–420, 127 for mel CAL1^bip1–160, 127 for mel CAL1^bip1–40, and 91 for mel CAL1^bip41–160. ***p < 0.0001 when comparing mel 1–407 CAL1 recruitment of bip CENP-A with that of bip CAL1 1–420, mel CAL1^bip1–160, or mel CAL1^bip1–40 (Fisher's two-tailed test). (E) BLOSUM80 alignment of residues 1–40 of CAL1 from selected species. Shading indicates percent similarity based on the BLOSUM80 score matrix (Henikoff and Henikoff, 1992). Black, 100% similar; dark gray, 80%–100% similar; light gray, 80%–60% similar; white, less than 60% similar. Stars indicate residues that have diverged in the ananassae subgroup (red box) compared with the rest of the melanogaster group and thus are candidates for residues co-evolving with CENP-A L1. Consensus sequence is shown above the alignment. Percent identity is shown as a bar graph below the consensus sequence: green indicates highly conserved, gold indicates somewhat conserved, and red indicates unconserved.

Figure 6. Model for the Co-evolution of CENP-A L1 and CAL1 in the Context of Centromere Drive

CENP-A centromeric deposition by CAL1 requires compatibility between the CAL1 N terminus and L1 (loop 1) of CENP-A. When centromere expansion occurs as a result of unequal crossover during meiosis, the larger and thus stronger centromere will be preferentially transmitted to the next generation (centromere drive) (Henikoff and Malik, 2002; Malik and Henikoff, 2002). To restore meiotic parity, positive selection of L1 mutations (Malik and Henikoff, 2001) weakens the ability of CAL1 to assemble CENP-A into chromatin, resulting in lower levels of CENP-A being deposited. The N terminus of CAL1 co-evolves, albeit at a slower rate, and re-establishes efficient CENP-A deposition, thereby maintaining centromere identity.