Sun_Supplemental Figure 6
PLASMIDS AND SUPPLEMENTAL METHODS
Supplementary Methods are organized by chronologic appearance in the manuscript.
Plasmids
CALML5 rescue, ZNF750 overexpression, and BioID fusion protein keratinocyte expression was performed by cloning onto the pLEX MCS vector (Thermo Scientific).
Short hairpin RNAs were cloned into the pLKO.1 vector (www.addgene.org or Sigma-Aldrich) and stable gene depletion achieved by keratinocyte infection in 1 µg/ml polybrene and 48 hour puromycin (1 µg/ml) selection.
Short hairpin targeting sequences in this study:
CALML5 shA / CTTTGCCCTGCAGTGGAATGACALML5 shB / CCCGCATTGATCTAAATAAAG
SFN shRNA / CCTGCTCTCAGTAGCCTATAA
For CRISPR/Cas9 genome editing, five candidate CALML5 targeting sites were cloned into the pLentiCRISPR vector (Feng Zhang lab, available from www.addgene.org). Of the five short guides against CALML5, the most efficient vector (#3) was used. The control scramble sequence below consists of sgRNA #3 sequence scrambled to a predicted non-targeting sequence.
CALML5 sgRNA #1 / TCTCCGCGGTTGACACGGACALML5 sgRNA #2 / GGCTTTCTCCGCGGTTGACA
CALML5 sgRNA #3 / ACGGGCAAGAACCTCTCGG
CALML5 sgRNA #4 / GTTTCCATCCGTGTCAACCG
CALML5 sgRNA #5 / GGCCTTCCTTGCCGCCGTC
CALML5 scramble / CTGACGCGACGGTAGCACA
RNA-seq analysis (Fig. 1A-B)
Samples were prepared as described in main Methods. Sample libraries were sequenced twice: The first sequencing was performed to a target of 30 million reads /sample (actual mapped reads ranged from 29-44M reads/sample). To improve potential representation of genes expressed at lower levels, repeat sequencing was performed to a target of 100 million reads/sample (actual mapped reads ranged from 71-93M reads/sample). These two sequencing runs are annotated as seq1 and seq2 in Supplemental Table 1 and results were combined between both sequencing runs. Reads were aligned to the transcriptome using the Tophat v.1.3.3, and reads per kilobase per million (RPKM) was assigned to each gene, then averaged between the corresponding progenitor or differentiated strata of the two biologic replicates.
To compare relative enrichment of gene abundance in progenitor vs. differentiated strata, an enrichment score was devised. Because some genes displayed zero RPKM counts in either progenitor or differentiated compartments, this precluded a simple fold-change calculation because of the potential for zero demoninators. Therefore, each gene count was re-scored as RPKM+1, to generate all non-zero numbers. Then, difference between log2 of gene RPKMs was designated as an enrichment score. The rank-ordered plot of enrichment scores for all genes is displayed in Fig. 1B.
Immunoblotting (Fig. 1D and others)
Protein lysates were resolved by SDS-PAGE gel electrophoresis, transferred to nitrocellulose membranes, then co-incubated with primary target and control antibodies overnight at 4°C. After washing three times with PBS-T, membranes were incubated with fluorescent secondary antibodies (IRDye 800CW and IRDye 680, LiCor) at 1:10000 dilutions for 45 minutes at room temperature. After washing three times with PBS-T and once with PBS, blots were visualized on a LiCor Odyssey CLx.
Immunofluorescence (Fig. 1E and others)
Seven micron tissue sections were mounted onto polysine slides (Thermo Scientific). Sections were fixed in 100% methanol or acetone for 10 min, followed by blocking in PBS-Tween (0.1%) with 5% horse serum for 45 min. Sections were incubated with primary antibodies overnight at 4°C, washed with PBS-Tween (0.1%) three times, then incubated with Alexa Fluor secondary antibodies at 1:500 dilution (488 and 555 Alexa Fluor dyes, Life Technologies) for 1 hour. Slides were washed with PBS-Tween (0.1%) three times, then mounted with Prolong Gold with DAPI (Life Technologies).
RNA harvesting and quantitative RT-PCR (Fig. 2A and others)
RNA was harvested using the RNeasy Plus Mini Kit (Qiagen) according to manufacturer’s instructions. Complementary DNA was generated with iScript reverse transcriptase according to manufacturer’s instructions (Bio-Rad). Quantitative PCR was performed on a Roche Lightcycler 480 with the following cycling parameters: 94°C for 10 seconds, 60°C for 10 seconds, 72°C for 10 seconds, 40 cycles. Gene expression level was quantiatated relative to the RPL32 gene as a loading control.
Xenografts (Fig. 2F)
To generate epidermal tissue xenografts, keratinocytes were seeded at a density of 500,000 cells on 1.4 cm2 section of devitalized human dermis cultured at the air-liquid interface in KGM (keratinocyte growth medium). After four days, tissue was grafted to size-matched surgical defects generated the dorsum of 8-week old female NOD/scid gamma mice (Jackson ImmunoResearch Laboratories), then wrapped with petroleum-impregnated gauze, Tegaderm (3M), and Coban (3M). Xenografts were unwrapped 18 days after initial grafting. All animal protocols were approved by Stanford Panel on Laboratory Animal Care (Protocol 9863).
Transepidermal water loss (Fig. 2G)
To measure transepidermal water loss, a Vapometer (Delfin Technologies) fitted with the small adapter was applied to xenografts on live mice 28 days post-grafting, according to the manufacturer’s instructions. All mice were housed in adjacent cages under similar conditions, and were measured consecutively in the same environmental setting. Four measurements were taken for each xenograft sample.
Chromatin immunoprecipitation (Fig. 3G)
Chromatin immunoprecipitation was performed as described previously (Boxer et al. 2014).
siRNAs and RNA interference (Fig. 3D, 3J, 4G, and others)
RNA interference was performed by introduction of 1 nmol of siRNA to 1 x 106 cells using nucleofection (Lonza) according to the manufacturer’s recommendations for keratinocytes.
Short interfering RNAs were obtained from the following providers. Specific siRNA is denoted by catalog number: CALML5 (Thermo Scientific J-013401-05), CHMP2B (Origene SR308637C), SFN (Origene SR301875C), SNAP29 (Life Technologies AM16708-137715), UBE2M (Origene SR305963B), WIBG (Origene SR313474). TINCR and ZNF750 siRNA was described previously (Kretz et al. 2013; Boxer et al. 2014).
Actinomycin D mRNA stability assay (Fig. 3L)
Control or TINCR-depleted keratinocytes were plated to confluence and differentiated in 24-well plates for 96 hours in 1.2 mM calcium. Media was then replenished with addition of 5 µg/ml of Actinomycin D to arrest transcription (timepoint zero). RNA was harvested at designated timepoints by aspiration of the media and direct application of RLT Plus Buffer and RNA isolation using the Rneasy Plus kit (Qiagen). RNA was subjected to reverse transcription using iScript RT (Bio-Rad), and relative target mRNA abundance measured by quantitative RT-PCR compared to 18SRNA control at each timepoint. Primers for each gene are listed in the Supplemental Table.
Far Western blotting (Fig. 4D and S4B)
The following recombinant proteins were used: CALML5 (Origene, TP304372), SFN (Origene, TP304045), GST (Sigma, SRP5348). CALML5 and GST were spotted on nitrocellulose membranes without denaturation. Recombinant SFN was incubated with the membrane at 1 µg/ml concentration, overnight at 4°C in TBS-T buffer with or without 1.2 mM calcium chloride and EGTA, then washed 4 times in TBS-T. Membranes were then incubated with primary antibody to SFN (1:200 dilution) for 1 hour at room temperature, washed 4 times, then incubated with secondary anti-mouse HRP (1:2500 dilution) for 45 minutes at room temperature. Blots were visualized with chemiluminescence. Ponceau S (Sigma) was used to stain a replicate blot as a protein loading control.
Proximity ligation assay (Fig. 4E, S4G)
Proximity ligation assays were performed using the DuoLink Mouse/Rabbit Orange kit (Sigma) according to the manufacturer’s instructions.
Phenocopy assay (Fig. 4G)
The 15 genes used in the CALML5 phenocopy depletion assay were AADACL2, PSORS1C2, FLG, LCE3D, LOR, ALOXE3, PDZK11P1, HOPX, IGFBP3, SPINK5, TGM1, ELOVL4, CDSN, BP1FC, ALOX12B. These genes were selected for their consistent repression at the 72hr time endpoint of CALML5 siRNA treatment in vitro, to allow parallel experimental conditions with which to compare the other candidates. Sources of siRNAs are listed above.
1 nmol siRNA was introduced to 1 x 106 keratinocytes using a nucleofector (Lonza) according to manufacturer’s recommended protocol. 12 hours after nucleofection, cells were trypsinized, counted, and plated to confluence on 24-well plates. After cells adhered, calcium chloride was added to final concentration of 1.2 mM to promote differentiation. Total RNA was harvested 72 hours after induction of differentiation and subjected to RNA isolation and qRT-PCR.
Nile Red staining (Fig. S1A)
Seven micron sections of epidermal tissue were sectioned onto polysine slides, and outlined with a hydrophobic pen. A Nile Red (Sigma Aldrich) stock solution was generated at 500 µg/ml concentration in acetone, then diluted to a working solution of 2.5 µg/ml in 75% glycerol. Slides were fixed in 4% paraformaldehyde for 5 minutes, rinsed briefly in PBS, then incubated with Nile Red working solution for 10 minutes at room temperature. Slides were rinsed twice in PBS, then dried and mounted.
TUNEL assay (Fig. S1B)
Assays were performed with the In Situ Cell Death Detection Kit-TMR Red (Roche) according to manufacturer’s instructions.
PolyA length assay (Fig. S2H)
CALML5 polyA length was measured with a polyA tail-length assay (Affymetrix) according to the manufacturer’s instructions. Primers used were:
Reverse primer:CATGGAACTCGGCAGTCCTTTA
Forward primer 1:CAGACCTTGGGCAGAAGGAG
Forward primer 2:AGACCTTGGGCAGAAGGAGG
Subcellular fractionation (Fig. S4A)
Keratinocytes were differentiated in vitro by plating to confluence and adding calcium to 1.2 mM for 96 hours. Cells were harvested and processed with the NE-PER nuclear cytoplasmic extraction kit (ThermoFisher) according to the manufacturer’s instructions.