Registered report: the androgen receptor induces a distinct transcriptional program in castration-resistant prostate cancer in man

Sampling

• Each experiment consists of two cohorts:

  • Cohort 1: untreated LNCaP cells

  • Cohort 2: LNCaP cells treated with R1881 (synthetic androgen)

• The experiment will be performed 3 times

  • This experiment is exploratory in nature, and thus no power calculations are necessary.

Procedure

Notes:

1. All cells will be sent for mycoplasma testing and STR profiling

2. Plate 1 × 10⁵ LNCaP cells in 6 cm tissue culture dishes and let grow overnight to adhere.

   1. LNCaP cells are expanded in RPMI-1640 supplemented with 10% FBS at 37 °C/5% CO₂. When cells have reached approximately 70% confluence, media will be removed, cells washed twice with 1x PBS and replaced with RPMI media supplemented with 10% charcoal dextran treated FBS.

   2. Prepare 2 separate plates; one for treatment, one for control.

3. Cells are kept in steroid depleted media for 72 h prior to androgen stimulation or vehicle control treatment. Treat cells with R1881 or DMSO for 0 and 24 h.

   1. For treatment, replace media with fresh media containing 1 nM R1881.

      1. R1881 is dissolved in DMSO at >10 mg/ml.

   2. For control, add vehicle (DMSO) only, at the same total volume as for treated cells.

4. ^# Lyse cells at 0 and 24 h time points with TRIzol reagent, according to manufacturer’s instructions.

   1. Use mechanical disruption (i.e., cell scraping) to dislodge cells, if necessary.

5. #Immediately flash-freeze and store cell/TRIzol mixture at −80 °C until ready to perform RNA extraction in Protocol 3.

6. Repeat experiment an additional two independent times.

Deliverables

• Data to be collected:

  • None applicable

• Samples delivered for further analysis:

  • Homogenized cells for use in Protocol 3, stored at −80 °C

Confirmatory analysis plan

• See Protocol 3 for analysis of gene expression

Known differences from the original study

• This experiment was not performed in the original study; it is an additional experiment added by the PCFMFRI core team in order to explore a the key finding that cultured cells possess a different AR-mediated gene signature for the core 16-gene set than in situ tumor tissue (Sharma et al., 2013). The FGC clone of the LNCaP cell line was not used in the study by Tao and colleagues, but rather low passage LNCaP cells.

Provisions for quality control

All data obtained from the experiment—raw data, data analysis, control data and quality control data—will be made publicly available, either in the published manuscript or as an open access dataset available on the Open Science Framework (https://osf.io/84tu2/).

Sampling

• The experiment will use at least 5 mice per group, for a power of at least 85.92%.

  • See Power Calculations for details.

  • To buffer against unexpected mouse deaths, 6 mice per cohort will be used.

• The experiment consists of two cohorts:

  • Cohort 1: non-castrated tumor-bearing mice

  • Cohort 2: castrated tumor-bearing mice

    • 70% of castrated mice develop castration-resistant tumors (C Massie, pers. comm., 2015). Thus, to generate at least 6 castration-sensitive tumors (or 30% of total castrated mice) and at least 6 castration-resistant tumors, 22 mice will be castrated.

  • Total mice:

    • 28 mice injected with LNCaP cells

    • 22 injected mice will be surgically castrated

Procedure

Notes:

1. All cells will be sent for mycoplasma testing and STR profiling

2. Culture LNCaP cells in RPMI-40 supplemented with 10% FBS at 37 °C/5% CO₂.

   1. Trypsinize to dissociate cells

   2. Centrifuge at ≤1,000 rpm and remove the supernatant

   3. Use a P200 pipette to gently dissociate the cells. Pipette up and down several times.

   4. Count cells

   5. Resuspend 2 × 10⁶ dissociated cells in 0.1 ml cold PBS, then mix with an equal volume of cold Matrigel. Total volume should be 0.2 ml per injection.

      1. Approximately 6 × 10⁷ cells will be needed for all 28 injections

3. Subcutaneously inject mice in the rear flank. Each mouse should receive a single injection.

   1. Mice should be microchipped prior to injection, so that they can be easily monitored throughout the duration of the study.

   2. Inject 0.2 ml cell/matrigel mixture per mouse subcutaneously into the flank of the mice using a 29G insulin syringe.

4. Measure tumor volume with manual calipers ^#twice weekly.

   1. Note time for tumors to form as well as tumor diameter and volume.

      1. #Measurement 1 will be width, measurement 2 will be the length, and measurement 3 will be the height of the tumor.

      2. #Since the growth of the tumor is not uniform, we will use the formula V = length × width × height × 0.5326 to obtain tumor volume.

5. Let tumors grow to approximately 100 mm³. Tumor growth takes approximately 4 weeks.

6. Randomly assign mice to either the intact or castrate cohort (ratio ∼1:5).

   1. ^#At the time of randomization (when the tumor growth in majority of mice is 100 mm³), mice without any measurable tumors and those with tumor volume between approximately 400–500 mm³ will be eliminated.

   2. ^#The remaining mice will be arranged in ascending order based on the tumor volume and group numbers will be assigned in a serpentine order in such way that the average tumor volume in each group will have equal or similar tumor volumes.

7. For the castration cohort, surgically castrate mice.

   1. ^#Castration procedure:

      1. Anesthetize animal with 100 mg/kg ketamine/5 mg/kg Xylazine

      2. Place on dorsal side facing the tail toward the surgeon. Prep the abdominal area by shaving the hair and swabbing with the surgical swab.

      3. An abdominal incision will be made and vas deferens and spermatic blood vessel is exteriorized and cauterized and testis tissue removed.

      4. The tissue is replaced into abdominal cavity and the incision will be closed with wound clips or absorbable sutures.

      5. After surgery, the mice will be given an analgesic, Buprenorphin 0.1 mg/kg (IP or IM) and a topical antibiotic mixture containing bacitracin zinc, neomycin sulfate, and polymyxin b sulfate.

      6. Castrated animals will be monitored daily for 7 days post castration.

   2. Around 70% of castrated mice will develop castration resistant tumors, which, after an initial regression, will regrow.

8. Continue monitoring tumor volume in all cohorts ^#twice weekly until tumor reaches 10% of body weight.

9. Once tumor volume is ≥10% of body weight, sacrifice mouse and excise tumor.

   1. Euthanize mice under isoflurane anesthesia.

   2. Spray tumor-bearing area on the flank with 70% isopropanol.

   3. Make a small incision on the skin of the flank, and peel skin to expose the subcutaneous tumor.

   4. Excise tumor using surgical scissors and forceps.

   5. Excess non-tumor tissue will be cleaned from tumor and the tumor will be flash frozen in liquid nitrogen and stored in −80 °C until further processing.

10. Dissociate tissue ^#with bead homogenizer with 1.00 mm silicon-carbide beads in TRIzol and store homogenized tissue at −80 °C until RNA extraction in Protocol 3.

Deliverables

• Data to be collected:

  • All mouse health records, including date of injection, date of castration, date of sacrifice, reason for sacrifice

  • All tumor volume measurements for all mice

• Samples delivered for further analysis:

  • ^#Homogenized tissue in TRIzol for use in Protocol 3, stored at −80 °C

Confirmatory analysis plan

• See Protocol 3 for analysis of gene expression

Known differences from the original study

All known differences in reagents and supplies are listed in the materials and reagents section above, with the originally used item listed in the comments section. Tumor volume will be measured twice weekly instead of daily as suggested by the replicating lab. All differences have the same capabilities as the original and are not predicted to alter experimental outcome.

Provisions for quality control

All data obtained from the experiment—raw data, data analysis, control data and quality control data—will be made publicly available, either in the published manuscript or as an open access dataset available on the Open Science Framework (https://osf.io/84tu2/).

Procedure

1. ^#Extract total RNA from TRIzol-preserved samples using RNeasy kit according to manufacturer’s instructions.

   1. Report total concentration and purity of isolated total RNA.

2. Reverse transcribe ^#0.5 ug RNA from cell and tissue samples derived from Protocols 1 and 2 into cDNA using the ^#qScript cDNA synthesis kit according to the manufacturer’s protocol.

3. Perform qPCR analysis of target genes in 384-well plates. ^*Run each reaction in triplicate.

   1. Use SYBR green mastermix with the following primer sets according to manufacturer’s protocol:

      1. PECI:

         1. F: GCCGGTTGAACATGATCTTT

         2. R: ATGGGCTGAGGTTGTTTGTC

      2. TNFSF10

         1. F: TGTGTCAGGGCTCTACTGTGA

         2. R: ATTCCCAGGGTAGGAGGAGA

      3. ABHD12

         1. F: TAGCCCAGGCGTGTAATAGG

         2. R: CTGGCCTTGAAGCAACATCT

      4. XRCC3

         1. F: GCAATCACAGCCAGAACAGA

         2. R: CAGAAGCAGAGTGTCCCACA

      5. MAD1L1

         1. F: GCACCCCTGTTGTTTTCATT

         2. R: ATGCCTGTTCCTTTGTGACC

      6. SEC61A1

         1. Replicating lab will design and optimize primers for this gene

      7. GFM1

         1. F: AAAAGGACTCCCTTCCCTCA

         2. R: ACGCGAGGAAAAAGAGAGTG

      8. TSPAN13

         1. Replicating lab will design and optimize primers for this gene

      9. STIL

         1. F: TCGACCAATCCCAAGTCTTC

         2. R: ATAGAGCACTTCCGGCTTCA

      10. TRMT12

          1. F: CTCAAGCAAGCGCATCAATA

          2. R: GGGCTTCCCACTTTCTCTCT

      11. EIF2B5

          1. F: CAGACAGATCGGGTTCCAAT

          2. R: TTCCATTGAGCGCTGATTTT

      12. TM4SF1

          1. F: TGCATTCATTTGGATTGGAA

          2. R: GAAAATCCGACAACCTGGAA

      13. NDUFB11

          1. Replicating lab will design and optimize primers for this gene

      14. SLC26A2

          1. F: GGAAAGGGAAGGAAAGGAAG

          2. R: TAGCCACAGCCAGTCACATC

      15. AGR2

          1. F: CAGCCATTCAAATCCCTTGT

          2. R: AAGAGTTCGTGGGGAAATCA

      16. NT5DC3

          1. F: ATGCACATCTTGGGAAGGTC

          2. R: TCCCTCCCCTTTTCCTCTTA

      17. CAMKK2 (known AR target as additional control; Massie et al., 2011)

          1. F: TGAAGACCAGGCCCGTTTCTACTT

          2. R: TGGAAGGTTTGATGTCACGGTGGA

      18. GAPDH control primers (Wang et al., 2012)

          1. F: GAAGGTGAAGGTCGGAGTC

          2. R: GAAGATGGTGATGGGATTTC

   2. Reaction mixture:

      1. Set up according to the manufacturer’s protocols for SYBR Green mastermix

   3. Run qPCR reactions on an ^#ABI StepOnePlus qPCR machine with the following cycling parameters:

      1. 2 min at 50 °C

      2. 10 min at 95 °C

      3. 40 cycles of:

         1. 15 s at 95 °C

         2. 1 min at 60 °C

   4. Analyze qPCR data using the ΔCt method normalized using the Actin control.

Deliverables

• Data to be collected:

  • Quality control data for total RNA and synthesized cDNA

    • A₂₈₀/A₂₆₀ and A₂₆₀/A₂₃₀ ratios for both

  • Efficiency calculations for each primer pair

  • Melt curve analysis and optimization data for primer pairs

  • Raw and normalized (to control gene) qRT-PCR values

  • Data analyzed with the ΔΔ Ct method

Confirmatory analysis plan

This replication attempt will perform the statistical analyses listed below, compute the effects sizes, compare them against the reported effect size in the original paper and use a meta-analytic approach to combine the original and replication effects, which will be presented as a Forest plot.

• Statistical Analysis of the Replication Data:

  • Comparison of average gene expression for the core 16 genes in xenografts from full, castration-sensitive and castration-resistant tumors

    • One-way ANOVA followed by planned pairwise comparisons using the Bonferroni correction to account for multiple comparisons:

  • Full vs. castration-sensitive

  • Castration-sensitive vs. castration-resistant

• Meta-analysis of original and replication attempt effect sizes:

  • This replication attempt will perform the statistical analysis listed above, compute the effects sizes, compare them against the reported effect size in the original paper and use a meta-analytic approach to combine the original and replication effects, which will be presented as a forest plot.

• Additional Exploratory Analysis of the Replication Data:

  • MANOVA with the average gene expression for 2 groups of genes in cultured cells as dependent variables and R1881 exposure as the independent variable followed by the following Bonferroni corrected comparisons:

    • Cohort 1: genes with no in vitro regulation by androgens (EIF2B5, NDUFB11, TNFSF10, XRCC3, TRMT12, STIL, AGR2, ABHD12, TSPAN13, MAD1L1) in R1881 treated vs. controls

    • Cohort 2: genes that showed up-regulation by androgens (SLC26A2, GFM1, SEC61A1, TM4SF1, NT5DC3, PECI) in R1881 treated vs. controls

  • Comparison of relative gene expression averaged within each sample (in xenografts and in cultured cells) for the core 16 genes: Two-way ANOVA with Cell Type (Xenograft vs Culture) and Treatment (Androgen exposure/Full or Control/Castration-sensitive as main effects followed by planned pairwise comparisons using the Bonferroni correction to account for the following comparisons:

    • Control LNCaP cells vs. R1881 treated LNCaP cells

    • Full xenograft vs. R1881 treated LNCaP cells

    • Castration-sensitive xenograft vs. Control LNCaP cells

Known differences from the original study

The original study used multiple methods for gene expression analysis, including Illumina expression arrays and qRT-PCR. This replication will only use qRT-PCR to analyze and compare gene expression of the core 16-gene set. This study will include a positive control by measuring expression of a known androgen target (CAMKK2). All known differences in reagents and supplies are listed in the materials and reagents section above, with the originally used item listed in the comments section. All differences have the same capabilities as the original and are not predicted to alter experimental outcome.

Provisions for quality control

All data obtained from the experiment—raw data, data analysis, control data and quality control data—will be made publicly available, either in the published manuscript or as an open access dataset available on the Open Science Framework (https://osf.io/84tu2/).

Summary of original data

• Note: data were estimated from published heat map.

  • A greyscale intensity value was generated for each gene in each sample using Image J. That value was normalized with respect to white (set to 0) and color (positive numbers for red, negative numbers for blue).

  • These scores were averaged across all 16 genes to generate a total score for each sample.

  • The samples in each group were averaged and a mean and SD generated.

  • See the Data Estimation worksheet (https://osf.io/9xhqs/?view_only=fac583dc5ea3471c88d581f64feda258) for details.

Test family

• One-way ANOVA on the mean scores per group followed by Bonferroni corrected comparisons on the following tumor groups:

  • Full vs. castration-sensitive

  • Castration-sensitive vs. castration-resistant

Power calculations

• Power calculations were performed using GraphPad PRISM v6 for Mac and G*Power v. 3.1.7 (Faul et al., 2007).