PIXUL-ChIP: integrated high-throughput sample preparation and analytical platform for epigenetic studies

Abstract Chromatin immunoprecipitation (ChIP) is the most widely used approach for identification of genome-associated proteins and their modifications. We have previously introduced a microplate-based ChIP platform, Matrix ChIP, where the entire ChIP procedure is done on the same plate without sample transfers. Compared to conventional ChIP protocols, the Matrix ChIP assay is faster and has increased throughput. However, even with microplate ChIP assays, sample preparation and chromatin fragmentation (which is required to map genomic locations) remains a major bottleneck. We have developed a novel technology (termed ‘PIXUL’) utilizing an array of ultrasound transducers for simultaneous shearing of samples in standard 96-well microplates. We integrated PIXUL with Matrix ChIP (‘PIXUL-ChIP’), that allows for fast, reproducible, low-cost and high-throughput sample preparation and ChIP analysis of 96 samples (cell culture or tissues) in one day. Further, we demonstrated that chromatin prepared using PIXUL can be used in an existing ChIP-seq workflow. Thus, the high-throughput capacity of PIXUL-ChIP provides the means to carry out ChIP-qPCR or ChIP-seq experiments involving dozens of samples. Given the complexity of epigenetic processes, the use of PIXUL-ChIP will advance our understanding of these processes in health and disease, as well as facilitate screening of epigenetic drugs.

. Schematic diagram of PIXUL. PIXUL is in house-built system comprised of the following main units: (1) a transducer-lens assembly capable of focusing ultrasound in each well of a 96 well microplate, (2) a high power amplifier to drive the transducer array, (3) a chiller-reservoir-pump that circulates ultrasound coupling fluid to reduce heating of the samples, and (4) a computer to control (MATLAB) the ultrasound pulse parameters (number of cycles, treatment configurations, and treatment time).
3 Fig. S2. Analysis of PIXUL sheared genomic DNA bands using in-house agarose gel electrophoresis system, Agilent 2100 High Sensitivity DNA Bioanalyzer DNA and Fragment Analyzer. gDNA from A row of 12 wells of 96-well plate was analyzed using either in-house agarose gel electrophoresis/ MATLAB image analysis tool, Agilent High Sensitivity Agilent 2100 Bioanalyzer (Agilent) or Fragment Analyzer (Agilent). A, Ethidium bromide stained agarose gel after electrophoresis of sheared gDNA. Numbers to the left of the gels show sizes of selected ladder bands in base pair (bp). Sheared fragments were analyzed by agarose gel electrophoresis image software (Methods) and shown here as waterfall plots (right panel) that contain best-fit curves of samples 1 to 12 in sequential order. X-axis; band size in base pair (bp). Y-axis; sample from a given well. Z-axis; relative signal intensity of bands for given plate column. B, Plots of DNA bands from 12 wells analyzed with the in-house gel electrophoresis system (Methods). C, Plots of DNA bands from 12 wells (diluted to ~5ng/l and measured following manufacturer's protocol) were analyzed with Agilent 2100 Bioanalyzer. D, Plots of gDNA bands from 12 wells was analyzed with Fragment Analyzer (2l of undiluted gDNA measured following manufacturer's protocol). E, Distribution of gDNA band means (bp) assessed with the three different analysis methods. F, Table listing mean size of gDNA fragment size from each well analyzed with the three different systems. The average of the DNA fragment mean and SDEV are shown in the last two columns of the table. The results show that analysis using the traditional Agilent 2100 Bionalyzer yield higher fragment sizes than using either the in-house system or Fragment Analyzer. The reasons for the differences, which are small, are not clear. These results show that the in-house very low cost agarose gel electrophoresis system is well suited for analysis of sheared gDNA fragments given that the bands are visualized with ethidium bromide (or Sybr green) staining.

Fig.S3. Comparison of shearing efficiencies of Bioruptor and PIXUL. Serumdeprived HCT116 96-well cultures.
A, Cells were crosslinked directly in the 96-well plate, quenched with glycine, and washed with PBS. PBS was then replaced with lysis buffer. Buffer is each well of row A was pipetted up and down several times, was then transferred to 0.5ml Eppendorf tubes and sheared using the Bioruptor. The rest of the plate was treated with PIXUL. B, After proteinase K digestion and reversal of cross-linking, DNA was analyzed by agarose gel electrophoresis. Numbers to the left of the gels show sizes of selected ladder bands in base pairs (bp). Sheared fragments were analyzed by agarose gel electrophoresis image software (Methods) and shown here as waterfall plots (MATLAB) that contains best-fit curves in sequential order of culture plate column of samples 1 to 12. X-axis; band size in base pairs (bp). Y-axis; sample from a well of a given column. Z-axis; relative signal intensity of bands for given plate column. C, The average size of Bioruptor sheared fragments was 394±64bp (69.9±14.1% in 200-600bp range) compared to 280±13bp ( 83.6±3.5% in 200-600bp range) with PIXUL (mean+SDEV, n=12).

Fig. S4. Matrix ChIP analysis of mouse liver chromatin prepared using PIXUL and
Bioruptor. Fragments of flash frozen mouse livers were cross-linked and then sheared either in a 96-well plate using PIXUL (12min) or Bioruptor (45min). A, Agarose gel electrophoresis of sheared chromatin after proteinase K digestion and reversal of crosslinking. Numbers to the left of the gels show sizes of selected ladder bands in base pairs (bp). Sheared fragments were analyzed by agarose gel electrophoresis image software (Methods) and shown here as waterfall plots (MATLAB) that contains best-fit curves of samples 1 to 12 in sequential order, X-axis; band size in base pair (bp). Y-axis; sample from a well of a given column. Z-axis; relative signal intensity of bands for given lane on the gel. B, Average size of sheared fragments. C, cartoon showing location of PCR primers along the genes. D, Sheared chromatin samples from control and septic mice livers (ref 40) were analyzed by Matrix ChIP with Pol II and H3 antibodies. Data represent mean+SEM (n=6 mice) expressed as a fraction of input. The results show consistent shearing of liver chromatin with either Bioruptor or PIXUL. Bioruptor yielded smaller fragments (by ~100bp). Matrix ChIP signal was higher with PIXUL (blue) compared to Bioruptor (green).

Fig. S5 Comparison of exome sequencing libraries prepared from human genomic DNA sheared by either PIXUL or Covaris LE220 Ultrafocused Sonicator.
Procedure: Four biological replicates consisting of 500ng gDNA (Promega, Madison, WI) were sheared in an off-the shelf 96-well plate using PIXUL (18min) or in a 96 well AFA microtube plate (Cat. 520168) using the Covaris LE 220 Ultrafocused Sonicator (450 watts, Duty Cycle 20%, 1000 cycles per burst for 3.5 min with dithering), (Covaris Inc, Woburn, MA). A 2X AMPure XP bead (Cat. A63882, Beckman Coulter, Indianapolis, IN) cleanup followed shearing to remove the smallest fragments, as well as to normalize sample volumes. Shotgun library was prepared using the Kapa HTP Library Prep Kit (Cat. KK8235, Roche, Indianapolis, IN) according to the manufacturer's instructions. End Repair was done by dA-taiing, adapter ligation, a dual SPRI size selection cleanup and 6 cycles of PCR using the Kapa HiFi Polymerase (Cat. KK2616). Adapters contained dual 8bp indices to facilitate multiplexing. The Quant-it dsDNA Assay Kit (Cat. Q33120I, Invitrogen, Carlsbad, CA) and the Agilent 2100 Bioanalyzer (Santa Clara, CA), were used to determine library quantity and quality, respectively. Libraries were pooled equally by mass into a single 8-plex capture pool, which hybridized for 64 hours with the Nimblegen V2 Exome probe set (Cat. 05860504001, Roche, Indianapolis, IN). Exome capture was followed by a stringency wash using the Nimblegen Hybridization and Wash kit (Cat. 05634253001) according to the manufacturer's recommendations. The final library pool was amplified a second time using the Kapa HiFi Polymerase in order to generate sufficient mass for sequencing. The final library pool was loaded on an Illumina MiSeq, using V3 chemistry and paired end 75bp reads, targeting 10 million reads for the pool (Illumina, San Diego, CA). Results: Comparison between PIXUL and Covaris instruments are shown in the following panels: A, Base distribution (A, red; T, pink; C, green; G, yellow) by sequencing cycle. The colored bars on the right expanded view to show order of the bases. B, Mean quality of the base calls across the human chromosomes. C, Distribution of library insert sizes. D, Library construction and Sequencing metrics. PreCapture Fragment Size was determined using the Agilent 2100 Bioanalyzer. Sequencing Insert Size was determined by mapping to Hg19. The percent of reads on target is defined as the number of reads overlapping the Nimblegen V2 Exome probe region. This comparison shows that the yield and insert size are comparable for exome sequencing libraries prepared from gDNA sheared by either PIXUL or Covaris instruments. Panels A and B demonstrate that shearing by PIXUL produces fragments that are not biased by base composition. Shearing by PIXUL resulted in library inserts that were approximately 20bp longer than library inserts resulting from Covaris shearing. This likely led to a decreased number of reads mapping to the V2 exome target, as the sequencing read length was not sufficient to cover the entire insert.

Fig.S6. Parallel PIXUL-RNA and PIXUL-ChIP analysis of mouse heart, kidney, liver, and lung. A,
Pieces of frozen heart, kidney, liver and lung samples well placed in wells of columns in 96-well plate containing 100l TRIZOL and were treated in PIXUL (1min). RNA was isolated, reversed transcribed (using oligo dT) and analyzed by qPCR. Data represent mean+SEM (n= 3 mice) expressed as a ratio to the transcript levels of housekeeping ribosomal protein gene, L32. B, Pieces from the same frozen organs as in A were cross-linked in wells of a 96-well plate and then sonicated in PIXUL (16min). PIXUL-sheared chromatin samples were simultaneously analyzed for Pol II levels at indicated organ-specific genes using Matrix ChIP. ChIP DNA was analyzed by qPCR. Data represent mean+SEM (n=3 mice) expressed as a fraction of input. These results demonstrate that PIXUL can be used to simultaneously processes samples from several organs for both chromatin and RNA isolation.

Fig. S7. Scatter plots of comparative analysis of PIXUL-ChIP-seq and ENCODE data sets.
HCT116 cells were grown to the density of ~200,000 cells per well, cross-linked, and sonicated using 96-well PIXUL. ChIP was performed with the indicated antibodies and libraries were generated from a single PIXUL well using Active Motif's Low Cell ChIP-Seq Kit. Libraries were sequenced on a NextSeq 500 (Methods). deepTools (ref 46) was used for each set of ChIP-seq data to count the number of reads for each 5kb-sized bin across the human genome for ENCODE data and PIXUL-ChIP-seq data. Read counts were normalized in each bin to total read counts in each file. empty bins were filtered out, and plotted the normalized read counts (log2 counts per million mapped reads) comparing ENCODE and PIXUL-ChIP-seq data. Each dot represents a genomic bin. r = Pearson's correlation coefficient. Scatter plots demonstrate good correlation between PIXUL-ChIPseq and ENCODE datasets. The differences between PIXUL-ChIP-seq and ENCODE data sets may reflect use of ChIP antibodies from different sources, growth conditions and the lower number of HCT116 cells (~200,000 for PIXUL-ChIP-seq) compared to ENCODE (>10 6 ).