Asynchronous exponentially growing cells were labeled with 20 μmol/L IdU for 20 min. They were then washed in phosphate-buffered saline (PBS), and subsequently labeled with 50 μmol/L CldU for 20 min, then cells were chased with of 200 μmol/L of thymidine for 60-90 min. Cells were trypsinized and resuspended in PBS. Cells were then embedded in pulsed-field gel electrophoresis agarose plugs to prepare high-molecular-weight genomic DNA. After proteinase K digestion, agarose plugs were washed with TE buffer. Agarose plugs were then melted in 100 mmol/L MES (pH 6.5), and digested using 2 μL of β-agarase (Biolabs). The DNA solution was poured into a Teflon reservoir and DNA was combed onto silanized coverslips (Microsurfaces, Inc.) using a combing machine. Samples have included a BAC as a molecular marker, allowing us to convert labeled signal length (in microns) to fiber length (in kb) as molecular combing was previously shown to uniform stretching 32 (link). Uniform stretching was confirmed by the uniform length of the stretched marker DNA. Coverslips were visually examined for DNA density and fiber length by a preliminary staining with YoYo1 (invitrogen). Sample preparations containing fibers that were mostly shorter than 200 kb or DNA concentration was too low or too high were discarded. The average fiber length in our studies is about 400 kb. For slide preparations that contained long fibers with appropriate density, coverslips with combed DNA were incubated at 60 °C for 2 h and denatured in 0.5 N NaOH for 20 min. IdU was detected using a mouse antibody directed against BrdU (IgG1, Becton Dickinson, cat.347580, 1:25). CldU was detected using a rat antibody directed against BrdU (Accuratechmecal, cat. OBT0030, 1:50). Single strand DNA (ssDNA) was detected using a mouse antibody directed against ssDNA (IgG 2a, Chemicon, cat. MAB3034, 1:100). Coverslips with DNA were incubated with primary antibodies for 1 h at room temperature. After washes, samples were incubated with secondary antibodies for 45 min. Secondary antibodies included Alexa Fluor 594 donkey antibody directed against rat (A21209, 1:100), Alexa Fluor 488 goat antibody directed against mouse IgG1 (γ1) (A-21121, Invitrogen, 1:100), and Alexa Fluor 647 goat antibody directed against mouse IgG2a (γ2a) (A21241, Invitrogen, 1:100). Slides were scanned with a BD pathway 855 controlled by AttoVision (Becton Dickinson). Fluorescent signals were measured using ImageJ (from the National Cancer Institute) with custom-made modifications. Only replication signals from high quality ssDNA (not from DNA bundles, not located at the end of a strand) were selected for analyses. Signals were marked for evaluation by “blind” measurers (not knowing which samples they were measuring); signal length was measured using the Image J software followed by automatic compilation of signal lengths into an Excel worksheet. Fork velocities and origin distances were calculated using Excel with a constant of 2kb/μm. Replication fork velocities were calculated using elongating fork signals only (initiating forks were eliminated; see a recent review62 (link) for a discussion of good practices in dynamic molecular combing). Experiments were performed at least in duplicate using independent biological isolations of DNA fibers for each experimental condition. Statistical analysis was performed in Prism 5 or 6 (GraphPad software) using the non-parametric Mann-Whitney rank sum test. Normality of distributions was assessed using the Kolmogorov-Smirnov test.