Mechanism of Bidirectional Leading-Strand Synthesis Establishment at Eukaryotic DNA Replication Origins

Abstract

DNA replication commences at eukaryotic replication origins following assembly and activation of bidirectional CMG helicases. Once activated, CMG unwinds the parental DNA duplex and DNA polymerase α-primase initiates synthesis on both template strands. By utilizing an origin-dependent replication system using purified yeast proteins, we have mapped start sites for leading-strand replication. Synthesis is mostly initiated outside the origin sequence. Strikingly, rightward leading strands are primed left of the origin and vice versa. We show that each leading strand is established from a lagging-strand primer synthesized by the replisome on the opposite side of the origin. Preventing elongation of primers synthesized left of the origin blocked rightward leading strands, demonstrating that replisomes are interdependent for leading-strand synthesis establishment. The mechanism we reveal negates the need for dedicated leading-strand priming and necessitates a crucial role for the lagging-strand polymerase Pol δ in connecting the nascent leading strand with the advancing replisome.

Keywords: CMG helicase; DNA polymerase; DNA replication; leading-strand synthesis; primase; priming; replication fork; replication origin; replisome.

Graphical abstract

Figure 1

Pol ε^PIP Replisomes Are Dependent on Pol δ (A) Diagram illustrating two non-mutually exclusive pathways for connecting the 3ʹ end of the nascent leading strand to CMGE after primer synthesis by Pol α. (B) Primer extension reaction on singularly primed M13 ssDNA. (C) Schematic of the 10.1 kbp ARS306 template used for all replication reactions on naked templates. The putative products of bidirectional origin-dependent replication are illustrated. (D) Standard replication reactions performed on the template illustrated in (C). (E) Quantitation of pulse-chase experiments performed as in Figure S1 in the presence of Pol δ. Error bars represent the standard error of the mean (SEM) from 3 experiments. (F) Standard replication reaction with the indicated Pol ε mutants. Replication products were separated through 0.7% (B) and 1% alkaline agarose gels (D and F).

Figure 2

Pol δ Can Access the 3ʹ End of All Nascent Leading Strands (A) Standard replication reaction with the indicated Pol δ proteins. (B) Pulse-chase experiment performed as illustrated. When Pol δ^Cat was added during the chase, it was added immediately after the 2.5 min time point. Products were analyzed through 1% (A) and 0.8% (B) alkaline agarose gels.

Figure 3

Mapping Leading-Strand Initiation Sites at ARS306 (A) Schematic of the 5.92 kbp ARS306 template used for chromatin replication reactions. Putative replication products are illustrated, and the locations of the restriction enzymes (distances from the 5′ end of the ACS to the cleavage sites) used to truncate the Left and Right leading-strand products are shown. 5′ cleavage products of variable length will be generated if priming occurs at multiple sites (Right lead illustrated). (B) Replication reaction (60 min) on the chromatinized template shown in (A). Products were post-replicatively digested with the indicated enzymes and analyzed through a 1% alkaline agarose gel. The undigested sample was loaded twice (lanes 1 and 4) to serve as a marker. (C) Products from the reaction in (B) were analyzed through a 4% denaturing polyacrylamide gel. (D) Normalized lane profiles for the data in (C) lanes 3 and 6. (E) Normalized lane profiles for the data in (C) lanes 2 and 5. (F) Diagrammatic representation of the data in (D) and (E). Right leading strands (red) are predominantly initiated to the left of the origin and Left leading strands (blue) are mostly initiated to the right. Dotted lines indicate variable start sites for leading-strand replication. The locations of the ACS and a putative B2 element are highlighted with gray bars. The asterisk indicates that the putative B2 element for ARS306 has been assigned based on a match to a B2 consensus sequence (Chang et al., 2011).

Figure 4

Mapping Leading-Strand Initiation Sites at ARS1 (A) Schematic of the 5.76 kbp ARS1 template used for mapping experiments. Expected replication products are illustrated and the locations of the restriction enzymes used to truncate the Left and Right leading-strand products are shown. (B) Normalized lane profiles for the ARS1 mapping data in Figure S5C in the absence of RNase HII. The locations of the ACS and B2 element are highlighted with gray bars.

Figure 5

Continuous Leading Strands Are Initiated from Lagging-Strand Primers (A) Experimental design to determine directly whether leading-strand replication is initiated from leading-strand or lagging-strand primers. Delayed addition of Pol α will enable CMG to move away from the origin before synthesis is initiated. If replication is initiated from a leading-strand primer, addition of Pol α after a delay will result in a shortening of leading-strand products. Conversely, if extension of lagging-strand primers from the adjacent replisome is the mechanism used to start leading-strand replication, products will get longer when Pol α addition is delayed. (B) Reaction scheme for delayed addition of Pol α. (C) Replication reaction performed as illustrated in (B) on the chromatinized ARS306 template (Figure 3A) for 60 min. (D) Lane profiles showing the leading-strand replication products in (C). (E) Initiation-site mapping for the experiment in (C).

Figure 6

Replisomes Are Interdependent for Leading-Strand Synthesis Establishment (A) Illustration of the ARS306 CPD-containing template and the putative replication products if extension of lagging-strand primers is the sole mechanism to establish leading-strand replication. (B) 60 min replication reaction on undamaged and CPD-containing chromatinized templates. Products were separated through 1% native or denaturing agarose gels as indicated. (C) Lane profiles of the data in (B), denaturing. (D) Products from the reaction in (B) were digested with PsiI and separated through a 4% denaturing polyacrylamide gel. (E) Two-dimensional gel analysis of the products from the CPD-template reaction in (B).

Figure 7

Model Describing the Mechanism of Leading-Strand Initiation at Eukaryotic DNA Replication Origins (i) Following CMG activation, the template is unwound and Pol α is recruited to each replisome to prime on the lagging-strand template. (ii) Pol δ elongates the primers back across the origin until it catches up with the advancing CMG complexes. (iii) A polymerase switch occurs, transferring the 3ʹ end of the nascent strand to CMGE. Dashed horizontal arrows illustrate the direction of CMG movement. The colors of the nascent strands correspond to the colors of polymerases that synthesized them.