Repository corticotrophin injection exerts direct acute effects on human B cell gene expression distinct from the actions of glucocorticoids

Summary

Repository corticotrophin injection (RCI, H.P Acthar^® gel) has been approved for use in the management of multiple autoimmune and inflammatory diseases for more than a half‐century, but its mechanism of action is not well understood. We used RNA‐Seq methods to define RCI‐regulated mRNAs in cultured human B cells under conditions of activation by interleukin (IL)‐4 and CD40 ligand. Following IL‐4/CD40L activation and RCI treatment we found up‐regulation of 115 unique mRNA transcripts and down‐regulation of 80 unique mRNAs. The effect on these RNA levels was dose‐dependent for RCI and was distinct from changes in mRNA expression induced by treatment with a potent synthetic glucocorticoid. RCI down‐regulated mRNAs were observed to include a significant over‐representation of genes critical for B cell proliferation under activating conditions. These data confirm that RCI exerts direct effects on human B cells to modulate mRNA expression in specific pathways of importance to B cell function and that, at the molecular level, the effects of RCI are distinct from those exerted by glucocorticoids.

Introduction

Corticotrophin (adrenocorticotrophin, or ACTH) is the principal regulator of production of glucocorticoid hormones from the adrenal cortex 1. The first reports of anti‐inflammatory properties of such glucocorticoids in 1950 led directly to the use of corticotrophin as a therapeutic with the intention of augmenting endogenous glucocorticoid production 2. However, subsequent clinical and experimental observations (including the recognition that cells of the immune system express receptors for corticotrophin and related peptides 3, 4, 5, 6 and that corticotrophin exerts direct effects on immune cells under conditions in which steroidogenesis cannot take place 7, 8) have suggested that augmentation of endogenously produced glucocorticoid levels might not be the sole mechanism underlying the therapeutic efficacy of repository corticotrophin injection (RCI).

Based on these observations, we recently began to re‐examine whether RCI (H.P. Acthar^® gel), a corticotrophin therapeutic approved by the Food and Drug Administration (FDA) for the treatment of multiple autoimmune and inflammatory disorders, might exert direct immunomodulatory effects independent of any action on the adrenal glands. Our previously reported experiments demonstrated that RCI exerts dose‐dependent inhibitory effects on the functions of interleukin (IL)‐4 + CD40 ligand (CD40L)‐stimulated human B lymphocytes in vitro 7. Specifically, RCI treatment results in prevention of IL‐4 + CD40L‐induced B cell proliferation and reduction of immunoglobulin production in vitro, without affecting B cell survival. At the molecular level, RCI treatment resulted in dose‐dependent reductions in the expression of mRNA encoding activation‐induced cytidine deaminase (AICDA), the key regulator of immunoglobulin heavy chain class‐switching and somatic hypermutation in B cells.

The objective of the present study was to define the effects of RCI on global patterns of gene expression in activated human B cells and to compare these effects to those exerted by glucocorticoids. We chose to examine RNA expression at one early time‐point (when immediate responses to peptide hormone action would be expected to be evident) and at one later time‐point (when immunoglobulin secretory responses to IL‐4/CD40L have maximized). We report here the patterns of mRNA expression that were altered in RCI‐treated cells and we demonstrate that the changes in specific mRNA levels in human B cells resulting from RCI treatment are largely distinct from those induced by the actions of glucocorticoids.

Methods

Human subjects, B lymphocyte preparation and culture

A total of 15 healthy volunteer subjects (five males and 10 females) donated peripheral blood samples for these studies. The ages of the subjects ranged from 24 to 65 years and averaged 41·9 ± 3·6 years. Volunteers were recruited for the study and gave informed consent to participate in the protocol approved by the Penn State/Milton S. Hershey Medical Center Institutional Review Board. Peripheral blood mononuclear cells were prepared by density gradient centrifugation on Histopaque 1077 (MP Biomedicals, Santa Ana, CA, USA) and peripheral B lymphocytes were then isolated using magnetic CD19 MicroBeads (Miltenyi Biotec, San Diego, CA, USA) and a MidiMACS Separator (Miltenyi Biotec). Recovered B cells were resuspended in complete medium (RPMI‐1640 with no phenol red, supplemented with 9% charcoal stripped fetal bovine serum, 1% GlutaMAX‐1 and 100 IU penicillin/100 µg/ml streptomycin). Stimulated cultures (with 10 ng/ml IL‐4 and 2 μg/ml recombinant human CD40L from HEK293 cells, both from R&D Systems, Minneapolis, MN, USA) and control cultures were plated at a density of 0·5–1·0 × 10⁶ cells per ml. RCI or placebo gel (provided in blinded fashion; Mallinckrodt Pharmaceuticals, St Louis, MO, USA) was added to replicate cultures at estimated ACTH analogue concentrations of 0·124, 1·24 and 2·49 µM (or equal volumes of placebo gel), to replicate the conditions under which RCI exerted biological effects in our prior studies 7. Placebo gel was identical to RCI, except that it did not contain any active pharmaceutical ingredient. In some experiments, dexamethasone (10 nM final concentration; 0·1% final volume/volume ethanol) or vehicle alone was used. Dexamethasone, a pure glucocorticoid receptor (GR) agonist, was used at concentrations well above the dissociation constant of GR (Kd ∼1 nM) to ensure GR saturation. Cells were incubated at 37°C with 5% CO₂ and cultures harvested for RNA isolation after 1 day incubation (19–24·5 h) or after 6 days of incubation. No further additions were made to cultures during either incubation period.

RNA isolation, library preparation and sequencing

RNA was isolated from harvested cells using the RNeasy Mini Kit (Qiagen, Germantown, MD, USA), according to the manufacturer's instructions. RNA concentration was quantitated using a Nanodrop 2000c spectrophotometer. RNA quality was determined using the Agilent 2100 BioAnalyzer provided by the Penn State College of Medicine Genome Science Core Facility. The majority of the samples had RNA Integrity Number (RIN) values above 9·0, and all RNAs had RIN values ≥ 7·1.

The cDNA library from each sample was prepared using the TruSeq Stranded Total RNA with Ribo‐Zero Gold Library Prep Kit (Illumina, San Diego, USA), as per the manufacturer's instructions. Total RNA was depleted of ribosomal RNA (rRNA) using proprietary rRNA‐depletion oligos. rRNA‐depleted RNA was subjected to fragmentation, reverse transcription, end repair, 3′‐end adenylation, adaptor ligation and subsequent polymerase chain reaction (PCR) amplification. The unique barcode sequences were incorporated in the adaptors for multiplexed high‐throughput sequencing. The final product was assessed for its size distribution and concentration using BioAnalyzer High Sensitivity DNA Kit (Agilent, Santa Clara, CA, USA) and Kapa Library Quantification Kit (Kapa Biosystems, Wilmington, MA, USA). The libraries were pooled and diluted to 2 nM in EB buffer (Qiagen, Valencia, CA, USA) and then denatured using the Illumina protocol. The denatured libraries were diluted to 10 pM by prechilled hybridization buffer and loaded onto TruSeq SR version 3 flow cells on an Illumina HiSeq 2500 (Illumina) and run for 50 cycles using a single‐read recipe (TruSeq SBS Kit version 3; Illumina), according to the manufacturer's instructions.

Quality control, mapping and quantification of RNA‐seq reads

Illumina casava pipeline version 1.8 was used to extract de‐multiplexed sequencing reads. FastQC (version 0.11.2) (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) was used to validate the quality of the raw sequence data. Additional quality filtering used FASTX‐Toolkit (http://hannonlab.cshl.edu/fastx_toolkit) using a quality score cut‐off of 20. Next, alignment of the filtered reads to the human reference genome (hg38) was performed using Tophat (version 2.0.9) 9, allowing two mismatches. Picard (version 1.102) (https://github.com/broadinstitute/picard) was used to assess proportion of mapped bases to coding, untranslated region (UTR), intronic and intergenic regions, respectively, and correctness of strand specificity. Picard was also used to find coverage across gene body to determine 5′‐ or 3′‐bias. Read counts were calculated using HTSeq 10 as provided with the Ensembl gene annotation (release 78).

Differential gene expression analysis

RUVSeq r package version 3.1 11, along with edgeR 12, was used to identify genes expressed differentially between control and RCI‐treated samples. First we normalized the raw read counts by selecting a set of ‘in‐silico empirical’ negative controls, i.e. least significantly differentially expressed genes based on a first‐pass differentially expressed analysis performed prior to normalization. Normalized read counts were applied to differentially expressed genes, using the negative binomial generalized linear model (GLM) approach implemented in edgeR. Significantly differentially expressed genes were defined to be those with corrected P‐value < 0·05 in each of the three independent experiments based on unique cell donors. Degree of up‐regulation for individual mRNAs is reported as fold increase above control (mean for three independent experiments); down‐regulation is reported as fraction of control (mean for three independent experiments).

Real‐time PCR assays for confirmation of differentially expressed mRNAs

Selected RNAs of interest that were found to be up‐ or down‐regulated in RNA Seq experiments were quantitated using an independent method (real‐time PCR). Total RNA was isolated from frozen CD19⁺ cell pellets using the RNeasy Mini Kit (Qiagen) and following a modified version of the manufacturer's protocol. The concentration of nucleic acid obtained from each preparation was measured by Nanodrop 2000c. Each RNA preparation was subjected to DNAse treatment using the DNA‐free Kit (Ambion, Foster City, CA, USA). The concentration of the RNA in the resulting supernatant was determined using the Qubit RNA HS Assay Kit (Thermo Scientific, Waltham, MA, USA) in conjunction with a Qubit Fluorometer. RNA integrity was confirmed by analysis on an Agilent 2100 BioAnalyzer.

Each DNase‐treated RNA sample was subjected to reverse transcription using a high‐capacity RNA to cDNA Kit (Applied Biosystems, Foster City, CA, USA). cDNA was amplified using TaqMan Gene Expression Master Mix and TaqMan Gene Expression Assays in the QuantStudio 12K Flex Real‐Time PCR System (Applied Biosystems) with the following parameters: 50°C for 2 min, 95°C for 10 min and 40 cycles of 95°C for 15 s and 60°C for 1 min.

TaqMan gene expression assays (Life Technologies, Carlsbad, CA, USA) are shown in Table 1.

Table 1.

TaqMan gene expression assays

IL‐4Rα	Hs00166237_m1
SLAMF1	Hs00234149_m1
SLAMF6	Hs01559920_m1
TNF‐SF11	Hs00243522_m1
LY9/SLAMF3	Hs03004330_m1
CD244/SLAMF4	Hs00175569_m1
SLAMF7	Hs00900280_m1
NR4A2	Hs00428691_m1
HDAC5	Hs00608366_m1
RORA	Hs00536545_m1

IL = interleukin; SLAMF = signalling lymphocytic activation molecule family; TNF‐SF11 = tumour necrposis factor superfamily member 11; LY9 = lymphocyte antigen 9; NR4A2 = nuclear receptor subfamily 4 group A member 2; HDAC5 = histone deacetylase 5; RORA = RAR‐related orphan receptor A.

The comparative Ct method was utilized by the QuantStudio software to calculate the fold difference in expression of a target gene relative to that of the endogenous control genes of RAR‐related orphan receptor A (RORA) and histone deacetylase 5 (HDAC5) for each sample (RORA and HDAC5 mRNA levels were previously determined experimentally to be invariant between resting and activated B cells; data not shown). Statistical analysis of results from six replicate experiments using cells from unique donors was by analysis of variance (anova) with Tukey's post‐hoc test using Prism version 7.0b software for Mac (GraphPad Software, La Jolla, CA, USA). Results are reported as mean ± standard error of the mean (s.e.m.).

Immunoblots for confirmation of protein expression

Following various in‐vitro treatments, peripheral blood CD19⁺ cells were harvested and whole cell lysates were prepared in radioimmunoprecipitation assay (RIPA) buffer. Protein concentrations of the lysates were determined using the Bio‐Rad Protein Assay and a SpectraMax Plus 384 microplate reader with SoftMax Pro Data Acquisition and Analysis Software (Molecular Devices, Sunnyvale, CA, USA) to measure absorbance of dye–protein complexes at 595 nm wavelength.

Four micrograms of protein from each cell lysate was subjected to electrophoresis under reducing conditions on a Novex 10% Bis‐Tris Gel in morpholino‐propane sulphonic acid (MOPS) sodium dodecyl sulphate (SDS) running buffer containing anti‐oxidant (NuPAGE reagents from Thermo Scientific/Invitrogen). Separated proteins were transferred from the gel onto polyvinylidene fluoride (PVDF) membrane (Millipore) using NuPAGE 1X Transfer Buffer, 10% methanol and 0·1% anti‐oxidant. To assess protein transfer, the membrane was stained with SWIFT membrane stain (G‐Biosciences, St Louis, MO, USA). Afterwards, stain was removed by washing with SWIFT destain (G‐Biosciences) and water. The membrane was blocked for 1 h using 5% non‐fat dry milk in Tris‐buffered saline with 0·1% Tween 20 (TBS‐T; Cell Signaling Technology, Danvers, MA, USA) and then probed overnight using primary antibody diluted in TBS‐T containing 5% bovine serum albumin (BSA) or milk. Horseradish peroxidase (HRP)‐conjugated secondary antibody diluted in blocking buffer was used for primary antibody detection. Super Signal West Pico Chemiluminescent Substrate (Thermo Scientific) was used for signal development. Images of the blot were captured using a FluorChem SP Imaging System with AlphaEase Software (Alpha Innotech, San Leandro, CA, USA).

To prepare blot for reprobing, the membrane was stripped of antibodies by incubation in Restore Western blot stripping buffer (Thermo Scientific) and then washed in buffered saline.

The following antibodies were used for immunoblotting: HO‐1 antibody (Cell Signaling Technology 70081, 1 : 1000); goat anti‐rabbit IgG, HRP‐linked antibody (Cell Signaling Technology 7074, 1 : 2000); SLAM (E‐11) (Santa Cruz sc‐166939, 1 : 200; Santa Cruz Technology, Santa Cruz, CA, USA); and m‐IgGκ BP‐HRP (Santa Cruz, 1 : 500)

Analysis of interaction pathways among RCI‐regulated mRNAs

The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database 13, 14 was used to analyse potential relationships among the gene products modulated by RCI action. Interactions are quantitated by STRING as a combined score computed by combining the probabilities from the different evidence channels and corrected for the probability of randomly observing an interaction 15. Evidence channels include experimental data from the Biomolecular Interaction Network Database (BIND) 16, Database of Interacting Proteins (DIP) 17, Human Protein Reference Database (HPRD) 18, 19, IntAct 20, Molecular INTeraction database (MINT) database 21 and Protein Interaction Database (PID) 22. Curated data in STRING are extracted from BioCyc, gene ontology (GO) 23, Kyoto Encyclopedia of Genes and Genomes (KEGG) 24 and Reactome 25. Settings of moderate confidence (probability > 0·40) high confidence (probability > 0·7) and highest confidence (probability > 0·9) were used to characterize potential interactions among mRNAs. Over‐representation of gene products in specific GO and KEGG pathways was assessed using STRING.

Results

At the highest concentration used, RCI treatment during IL‐4/CD40L‐mediated activation of human B cells for 1 day resulted in significant and reproducible increases in the expression of 115 distinct mRNA transcripts compared to control (placebo gel‐treated) cells (Supporting information, Table S1A). Seventeen of these gene transcripts were undetectable in placebo‐treated cells but rose to levels that were detectable in RNA samples from RCI‐treated cells. Ninety‐eight genes were expressed at detectable levels in the placebo‐treated cells and their expression was increased between 3·19‐ and 320‐fold (average induction: 18·38 ± 3·53‐fold) in RCI‐treated cells (Fig. 1a). In cells treated with the intermediate concentration of RCI, we found 42 gene products to be up‐regulated significantly and reproducibly compared to control (placebo gel‐treated) cells (Supporting information, Table S1B). Average induction was 3·95 ± 0·39‐fold and the values ranged from 1·91‐ to 14·76‐fold (Fig. 1a). Twenty of the mRNAs up‐regulated by the intermediate dose were also observed to be up‐regulated by the higher RCI dose. Treatment of B cells with the lowest concentration of RCI resulted in up‐regulation of only three transcripts compared to control (placebo gel‐treated) cells (Supporting information, Table S1C). Induction ranged between 1·56‐fold and 2·69‐fold (Fig. 1a). Two mRNAs [haem oxygenase (HMOX1) and myelin protein zero (MPZ)] were observed to be up‐regulated by all three doses of RCI compared to control (placebo gel‐treated) cells. The magnitude of induction of both of these mRNA species exhibited dose dependence (Fig. 2a).

Figure 1.

RNA‐Seq reveals modulation of gene expression by repository corticotrophin injection (RCI) in interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B lymphocytes in vitro. (a) mRNAs up‐regulated by RCI (solid symbols) or dexamethasone (open symbols): each RNA that was up‐regulated significantly in each of three independent experiments analysed by RNA‐Seq is represented by one symbol. The level of expression is shown as the average fold‐induction above the level observed in control cells [RCI placebo gel control for RCI and ethanol vehicle control for dexamethasone (Dex)] for three mRNA transcripts that were increased reproducibly by treatment with low‐dose RCI [estimated adrenocorticotrophin (ACTH) analogue concentration, 0·124 μM] for 42 mRNAs whose levels were increased by medium dose RCI (estimated ACTH analogue concentration 1·24 μM) and for 98 mRNAs whose levels were increased by high‐dose RCI (estimated ACTH analogue concentration 2·49 μM). The relative increases in 55 distinct mRNAs whose levels were increased by treatment of IL‐4/CD40L‐activated human B cells with 10 nM Dex, a synthetic glucocorticoid, are also shown. Each data point represents the mean induction level of an individual mRNA observed in three independent experiments. Horizontal bars indicate mean of all induction values for each treatment. (b) mRNAs down‐regulated by RCI (solid symbols) or dexamethasone (open symbols): the level of expression is shown as the fraction of the level observed in control cells (RCI placebo gel control for RCI and ethanol vehicle control for dexamethasone) for two mRNA transcripts that were decreased reproducibly by treatment with low‐dose RCI (estimated ACTH analogue concentration, 0·124 μM) for 69 mRNAs whose levels were decreased by medium‐dose RCI (estimated ACTH analogue concentration 1·24 μM) and for 80 mRNAs whose levels were decreased by high‐dose RCI (estimated ACTH analogue concentration 2·49 μM). The fractional expression of 18 distinct mRNAs whose levels were decreased by treatment of IL‐4/CD40 ligand‐activated human B cells with 10 nM Dex, a synthetic glucocorticoid, are also shown. Each data point represents the mean value relative to baseline from three independent experiments. Horizontal bars indicate mean of all suppression values for each treatment.

Figure 2.

RNA‐Seq reveals dose–response relationships for selected genes whose expression was modulated by repository corticotrophin injection (RCI) in interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B lymphocytes. (a) The level of expression [mean ± standard error of the mean (s.e.m.) of three independent experiments] is shown as fold‐induction above the level observed in placebo gel‐treated control cells for 20 mRNA transcripts that were increased reproducibly by treatment with medium‐dose RCI [estimated adrenocorticotrophin (ACTH) analogue concentration 1·24 μM; shaded grey bars] as well as by high‐dose RCI (estimated ACTH analogue concentration 2·49 μM); black bars). Two mRNAs [haem oxygenase 1 (HMOX1) and myelin protein zero (MPZ)] additionally exhibited significant up‐regulation in cells treated with low‐dose RCI (estimated ACTH concentration 0·124 μM; clear bars outlined in black). (b) The level of expression [mean ± standard error of the mean (s.e.m.)] of three independent experiments) is shown as the fraction of the level observed in placebo gel‐treated control cells for 22 mRNA transcripts that were decreased reproducibly by treatment with medium‐dose RCI (estimated ACTH analogue concentration 1·24 μM; shaded grey bars) as well as by high‐dose RCI (estimated ACTH analogue concentration 2·49 μM; black bars).

A distinct subset of mRNA transcripts was down‐regulated reproducibly in RCI‐treated cells by 1 day. The highest dose of RCI resulted in reproducible down‐regulation of 80 unique gene transcripts to as low as 2·8% of control values (Supporting information, Table S1D). On average, these transcripts were suppressed to 18.03 ± 0.01% of control values (Fig. 1b). One transcript was suppressed to undetectable levels. Treatment with the intermediate RCI dose resulted in down‐regulation of 69 mRNAs (Supporting information, Table S1E) to an average of 36·23 ± 0·12% of control values (Fig. 1b). Two transcripts were suppressed below detectable levels by the intermediate RCI dose. Levels of 22 mRNA transcripts were suppressed significantly by both the intermediate and high doses of RCI; all but one of these mRNAs demonstrated greater suppression with the higher dose of RCI (Fig. 2b). The lowest dose of RCI resulted in reduced levels of only two mRNAs (Supporting information, Table S1F). Levels of these two mRNAs were reduced to 58·5% of baseline (Fig. 1b).

Differences in mRNA expression levels between RCI‐ and placebo gel‐treated cells had largely disappeared by the sixth day of cell culture (Supporting information, Table S2). At that time‐point, in cells treated with high‐dose RCI, no mRNAs were increased and only 16 mRNAs were decreased compared to placebo‐treated cells. Treatment with intermediate dose RCI resulted in one mRNA at levels higher than placebo control cells and four mRNAs at levels lower than placebo control. Cells treated with lowest dose of RCI revealed two down‐regulated mRNAs at the 6‐day time‐point.

The observed alterations in B cell gene expression in RCI‐treated cells were not attributable to any effect exerted by the placebo gel. The experiments utilized pairwise comparisons between RCI‐treated cells and cells treated with identical volumes of the placebo gel. Furthermore, additional pairwise comparisons of mRNA expression between IL‐4/CD40L‐activated B cells with no other additions to culture medium and IL‐4/CD40L‐activated cells treated with the highest volume of placebo gel preparation revealed only minor differences (data not shown).

Alterations in mRNA expression induced by RCI were distinct from those observed when human B cells under the same activating conditions were treated with glucocorticoids. Dexamethasone treatment of IL‐4/CD40L‐activated human B cells for 1 day resulted in increased levels of 56 unique mRNA transcripts (Fig. 1a and Supporting information, Table S3A) and reduction in levels of 18 mRNAs (Fig. 1b and Supporting information, Table S3B). Of the dexamethasone up‐regulated gene products, 55 were detectable in control IL‐4/CD40L‐activated human B cells and rose, on average, by 5·07 ± 0·64‐fold. TSC22D3, a transcript known to be regulated positively by glucocorticoids (glucocorticoid‐inducible leucine zipper protein; GILZ) rose by an average of almost sevenfold in the three independent experiments conducted with dexamethasone treatment. Only one gene product, the mRNA encoding glutathione S‐transferase µ3 (GSTM3) was up‐regulated by both RCI (11·2‐fold) and dexamethasone (4·4‐fold). Transforming growth factor (TGF)‐β1 mRNA was up‐regulated (3·2‐fold) by glucocorticoid therapy but suppressed (to 4% of levels in control activated cells) by RCI. Eighteen mRNAs were found to be down‐regulated to an average of 38 ± 0·02% of control in dexamethasone‐treated cells (Fig. 1b). Three gene transcripts were found to be down‐regulated in both RCI and dexamethasone‐treated cells: macro domain containing protein 2 (MACROD2), prostate androgen‐regulated mucin‐like protein 1 (PARM1) and tumour necrposis factor superfamily member 11 (TNFSF11) (RANK ligand).

Independent real‐time PCR Taqman gene expression assays were used to provide confirmation of the observed modulation of mRNA expression in RCI‐treated and dexamethasone‐treated cells after 1 day. We selected candidate genes of interest from among signalling lymphocyte activation molecule (SLAM) family members and TNFSF members as well as the IL‐4 receptor alpha subunit and the orphan nuclear receptor nuclear receptor subfamily 4 group A member 2 (NR4A2) for these analyses. Four of the mRNAs had exhibited increased expression and four mRNAs had exhibited decreased expression in the RNA‐Seq data. Three of the four mRNAs observed to be up‐regulated by RCI were confirmed in the reverse transcription (RT)–PCR assays (Fig. 3). SLAMF3, SLAMF7 and nuclear receptor subfamily 4 group A member 2 (NR4A2) mRNAs were all increased significantly in RCI‐treated cells, while dexamethasone did not alter levels of these mRNAs. The SLAMF4 assay used did not reveal significant changes mRNA with either RCI or dexamethasone treatment. The RCI down‐regulated mRNAs selected for confirmation by RT–PCR showed concordant results between the two data sets for all four mRNAs examined, including IL‐4 receptor alpha subunit (IL‐4RA), SLAMF1, SLAMF6 and TNF‐SF11 (RANKL) mRNA. Levels of each of these transcripts were found by both RNA‐Seq and Taqman gene expression assay to be reduced significantly in RCI‐treated cells compared to IL‐4/CD40L‐stimulated cells (Fig. 4). Dexamethasone treatment resulted in suppression of mRNA in only one of these Taqman assays – SLAMF6 – but the suppression was significantly less than that observed with RCI (P < 0·0001 by anova). Dexamethasone treatment showed a slight stimulatory effect on IL‐4RA expression, in distinction to the suppression observed with RCI (Fig. 4). We confirmed parallel changes in gene expression at the protein level in RCI‐treated cells for two mRNAs of interest – HMOX1, which was the most robustly up‐regulated transcript (greater than 80‐fold induction), and SLAMF1 (CD150), whose mRNA was among those reduced most profoundly (by 80%) in RCI‐treated cells. Immunoblots (shown in Supporting information, Fig. S1) revealed findings concordant with the changes in mRNA levels encoding each protein.

Figure 3.

Real‐time polymerase chain reaction Taqman assays confirm patterns of up‐regulated expression of selected mRNAs in interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B cells at 1 day (19–24 h). Effects of placebo gel (PBO), repository corticotrophin injection (RCI) gel and dexamethasone (Dex) are shown on expression of four mRNAs identified in RNA‐Seq data as being up‐regulated in RCI‐treated cells. Data represent mean ± standard error of the mean (s.e.m.) of six independent experiments using cells from unique individual donors. Statistical analysis was performed using analysis of variance (anova) with Tukey's multiple comparison test.

Figure 4.

Real‐time polymerase chain reaction Taqman assays confirm patterns of down‐regulated expression of selected mRNAs in IL‐4‐CD40 ligand (CD40L)‐activated human B cells at 1 day (19–24 h). Effects of placebo gel (PBO), repository corticotrophin injection (RCI) gel and dexamethasone (Dex) are shown on expression of four mRNAs identified in RNA‐Seq data as being down‐regulated in RCI‐treated cells. Data represent mean ± standard error of the mean (s.e.m.) of six independent experiments using cells from unique individual donors. Statistical analysis was performed using analysis of variance (anova) with Tukey's multiple comparison test.

Known interactions among the proteins encoded by the transcripts up‐ or down‐regulated by RCI at 1 day were explored using the STRING database 14. Analysis of the 115 up‐regulated transcripts revealed statistically significant over‐representation of gene products in several GO and KEGG protein pathways (Supporting information, Table S4A,B). A graphical representation of identified relationships among the RCI‐up‐regulated gene products is shown in Fig. 5. The most prominent functionally related cluster includes 12 different gene products involved in the temperature and unfolded protein responses. Also over‐represented were genes in the mitogen‐activated protein (MAP) kinase signalling pathway [KEGG 04010; false discovery rate (FDR) = 0·0162), including growth arrest and DNA damage inducible proteins (GADD45A and GADD45B), MAP kinase 10, heat shock protein family (HSP)A2 and HSPB1, proto‐oncogene C‐Fos (FOS) and dual specificity phosphatase 10 (DUSP10)]. Examination of the messenger RNAs that were down‐regulated in the RCI‐treated cells revealed significant over‐representation of mRNAs that encode proteins involved in a variety of immune response pathways in both the GO Biological Response database (Supporting information, Table S4C) and in the KEGG protein pathway database (Supporting informationm Table S4D). A graphical representation of identified relationships among the RCI down‐regulated gene products is shown in Fig. 6.

Figure 5.

Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) network analysis of mRNAs reproducibly up‐regulated by repository corticotrophin injection (RCI) treatment of interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B cells in culture at 19–24 h. Individual nodes represent proteins (including all isoforms produced by a single gene locus). Large nodes indicate proteins with known or predicted three‐dimensional structure. Lines represent known interactions between proteins. The thickness of the line represents the confidence of interactions calculated by STRING. Thick lines indicate confidence > 0·90; intermediate thickness lines indicate confidence > 0·70; thin lines indicate confidence > 0·4. Highlighted by darker shading are 12 gene products included in either or both of the significantly over‐represented gene ontology (GO) pathways: ‘response to temperature stimulus’ [GO 0009266; false discovery rate (FDR) = 0·02] and ‘response to unfolded protein’ (GO 0006986; FDR = 0·0488).

Figure 6.

Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) network analysis of mRNAs reproducibly down‐regulated by repository corticotrophin injection (RCI) in interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B cells in culture at 19–24 h. Individual nodes represent proteins (including all isoforms produced by a single gene locus. Large nodes indicate proteins with known or predicted three‐dimensional structure. Lines represent known interactions between proteins. The thickness of the line represents the confidence of interactions calculated by STRING. Thick lines indicate confidence > 0·90; intermediate thickness lines indicate confidence > 0·70; thin lines indicate confidence > 0·4. Highlighted by the darker shading are 17 gene products included in either or both of the significantly over‐represented gene ontology (GO) pathways: ‘positive regulation of immune system process’ [GO 0002684; false discovery rate (FDR)] = 0·000294) and ‘positive regulation of immune response’ (GO 0050778; FDR = 0·000294).

STRING database analysis of the gene products that appeared to be most sensitive to RCI regulation (i.e. those modulated by both lower and higher doses of RCI) confirmed the findings of up‐regulation of temperature/unfolded protein response pathway genes and down‐regulation of immune response pathway genes. Among the 20 mRNAs up‐regulated by lower as well as higher doses of RCI, over‐representation (five mRNAs) of the GO pathway 0009266, ‘response to temperature stimulus’; FDR = 0·00974) was observed. Among the 22 mRNAs down‐regulated by RCI treatment, over‐representation (seven mRNAs) of the pathway ‘positive regulation of immune system process’ (GO 0002684; FDR = 0·0479) was noted.

As expected from the observed minimal overlap between RCI and dexamethasone‐regulated transcripts, pathways analysis of the gene transcripts up‐ and down‐regulated by dexamethasone in human B cells activated in vitro did not reveal any similarity to the results observed in RCI‐treated cells (Fig. 7). The set of dex up‐regulated genes did not reveal statistically significant enrichment of any GO biological process or KEGG protein pathways. The dexamethasone down‐regulated gene set did not reveal significantly more interactions than expected (P = 0·187) and no significant enrichment of any GO or KEGG pathways was observed.

Figure 7.

Comparison of mRNAs modulated by repository corticotrophin injection (RCI) and dexamethasone (Dex) in interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B cells in culture at 19–24 h. (a) Gene products upregulated by RCI or Dex. Circles denote the universe of transcripts up‐regulated by each hormone, with overlap denoting mRNAs up‐regulated by both. Within each circle are tabulated the number of interactions (structural, functional, genetic or experimentally observed) among these gene products, as well as the significantly over‐represented gene ontology (GO) biological process pathways and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. (b) Gene products down‐regulated by RCI or Dex. Circles denote the universe of transcripts up regulated by each hormone, with overlap denoting mRNAs downregulated by both. Within each circle are tabulated the number of interactions (structural, functional, genetic or experimentally observed) among these gene products, as well as the significantly over‐represented gene ontology (GO) biological process pathways and KEGG pathways.

Discussion

In these experiments, we sought evidence for direct effects of RCI on isolated human B lymphocytes by characterization of a molecular signature of an early cellular response to the drug at the mRNA level. Our experiments were carried out under conditions in vitro that activate B cells to undergo late steps of differentiation leading to active IgG secretion. We found evidence for dose‐dependent alterations in specific B cell mRNAs that are largely consistent with our previous observations of suppressive effects of the RCI preparation on B cell function in vitro 7. The effects of RCI at the mRNA level are clearly distinct from those exerted by a potent glucocorticoid (dexamethasone) under identical conditions of activation of purified isolated B cells.

Other earlier studies have reported activity of both biologically derived and synthetic ACTH preparations on peripheral B lymphocyte function in vitro. Biologically derived ACTH preparations have been reported to suppress antibody production by murine spleen cells in vitro in response to both T cell‐dependent and ‐independent antigens administered in vivo 26. Synthetic ACTH was found to suppress IgE production in vitro by cultures of human peripheral blood mononuclear cells but not by highly purified B cell populations 27. Human tonsillar B cells activated in vitro by IL‐2 or IL‐4 exhibited enhanced cellular proliferation and antibody production in response to ACTH treatment 28, 29. Differences in corticotrophin preparations, cellular culture systems and the use of different B cell‐activating conditions do not allow direct comparisons among these various reports.

A number of mRNAs whose levels were altered during treatment of B cells with RCI have been identified previously as important in either normal B cell function or in conditions of humoral autoimmunity. The IL‐4 receptor alpha subunit was down‐regulated in RCI‐treated cells to approximately one‐third of the level observed in control IL‐4/CD40L‐activated B cells. Comparable reductions in IL‐4Rα expression have been shown to be associated with diminished responsiveness to IL‐4, associated with attenuated antibody secretory responses in neonatal human B cells 30. SLAMF3 (Ly9), which was found to be up‐regulated by RCI treatment in the current experiments (by more than sixfold in RNA‐Seq experiments and by 9·02 ± 1·36‐fold in RT–PCR experiments) has been reported to function as an inhibitory cell surface receptor capable of disabling autoantibody responses 31. SLAMF7 (CS1, CD319) was noted to be up‐regulated by more than eightfold in our RNA‐Seq data and by 2·18 ± 0·30‐fold in our RT–PCR experiments. This cell surface receptor has been found previously to be up‐regulated during B cell activation 32; how further augmentation of levels by RCI could contribute to suppressive effects on B cell proliferation or antibody production remains to be investigated. Two other SLAM family members, SLAMF1 (CD150) and SLAM F6 (CD352), showed significant down‐regulation in both RNA‐Seq data and in RT–PCR confirmation experiments. SLAMF1 mRNA was reduced to 20% of control levels by RCI in RNA‐Seq data and to 6 ± 1·4% of control levels in RT–PCR experiments. SLAMF6 mRNA fell to 17% of control in RNA‐Seq experiments and 6 ± 4% of controls in RT–PCR data. Mice with targeted deletions of SLAMF1 and SLAMF6 exhibited augmented B cell antibody responses, consistent with an inhibitory action of these receptors 33. How the apparent RCI‐induced down‐regulation of these mRNAs is related to the observation of diminished cellular proliferation and antibody production by RCI‐treated human cells in vitro also remains a subject for further investigation.

One mRNA exhibited discordant results between RNA‐Seq observations and RT–PCR confirmation. SLAMF4 (CD244) protein has been reported previously not to be expressed on B cell surface at any developmental stage 34, 35. In RNA‐Seq experiments we found low‐level mRNA expression but up‐regulation by RCI (19‐fold); in RT–PCR experiments we found that RCI treated cells expressed 45 ± 13% of control levels. The reasons for these discordant findings are unclear, but could include artefacts due to isoform‐specific misamplification in the PCR experiments.

Two other candidate gene products of interest showed significant modulation with RCI treatment. NR4A2 [nuclear receptor related 1 protein (Nurr1), nerve growth factor IB (NGFI‐B)] is an orphan nuclear receptor known to function in forkhead box protein 3 (FoxP3) induction in T cells and in the determination of T cell fate 36, 37. We found vigorous up‐regulation of B cell expression of NR4A2 (20‐fold in RNA‐Seq experiments and 67 ± 17·84‐fold in RT–PCR assays. TNF‐SF11 (RANK ligand) has been identified as a B cell product that exerts effects on osteoclasts to promote bone resorption 38. RCI‐treated B cells activated by IL‐4 and CD40L showed reduction of TNF‐SF11 mRNA levels to 8% of control in the RNA‐Seq experiments and to 1 ± 0·6% in RT–PCR experiments.

In addition to effects on specific candidate gene products in B cells, pathways analyses revealed that RCI treatment of B cells under activating conditions was associated with significant changes in functionally related groups of mRNAs. First, we noted prominent up‐regulation of components of the unfolded protein response pathway with RCI treatment. During the induction of high‐level immunoglobulin secretion by B lymphocytes, as occurs under the IL‐4/CD40L‐stimulated conditions in our cultures, this cellular pathway is normally activated to accommodate the increase in nascent antibody flux through the endoplasmic reticulum 39, 40, 41. As RCI action on B cells under these activating conditions in vitro results in reduced antibody secretion 7, a further compensatory increase in the expression of unfolded protein response pathway components might result, but such an interpretation of the data remains to be explored formally. A KEGG pathway including MAP kinase 10 (JNK3), GADD45A and GADD45B was the second functional pathway with multiple components over‐represented significantly in the mRNAs up‐regulated in RCI‐treated, IL‐4/CD40L‐activated B cells. GADD45 gene transcript levels are known to be increased under conditions of environmental stress or growth arrest and GADD45A may function as suppressor of autoimmunity 42, 43, 44. Increased expression of these transcripts would be consistent with the growth arrest we have observed in association with RCI treatment 7.

Consistent with our previously published 7 observation of RCI's direct inhibitory effects on human B cell proliferation and immunoglobulin production during IL‐4/CD40L‐mediated activation in vitro, we found that the set of mRNAs down‐regulated by RCI exhibited significant over‐representation of transcripts encoding components of several immune response pathways. The central components of this group, MAP kinase 11, MAP kinase 13 and myocyte enhancer factor 2C (MEF2C) have been demonstrated to regulate the cellular proliferation required in the terminal differentiation of antibody‐secreting cells 45. The proliferative arrest induced by RCI action on IL‐4/CD40L‐activated B cells 7 was observed after several days in culture. As no proliferative response has yet occurred in control IL‐4/CD40L‐activated cells at the early time‐point (19–24 h) of the current study, it seems possible that inhibition of the p38 MAP kinases and/or MEF2C expression might be central to suppressive effect of RCI gel on the function of activated B cells. RCI action by this mechanism might also be antagonistic to another known hormonal action on lymphocytes. MAP kinase p38 has been found previously to be up‐regulated by prolactin action in Nb2 T lymphoma cells and in human leucocytes 46, and consistent with the known stimulatory effect of prolactin on lymphocyte proliferation 47, 48.

We note several limitations to our study. First, we were not able to examine a wide range of time‐points and we selected only two for our studies. The early time‐point was chosen because of previously noted robust effects of RCI on some specific gene expression, but this is before any cellular proliferative or IgG secretion in response to the IL‐4/CD40L stimulus is observable 7. A later time‐point was selected to be after maximal IgG secretion into culture medium has occurred, but near the end of the life of culture B cells under these stimulating conditions. More detailed analyses of the time–course of changes in expression of specific mRNAs identified by these studies might now be of interest. Secondly, it is not technically feasible to confirm independently the changes identified by RNA‐Seq for each individual mRNA, and we did note one discrepancy with RT–PCR (RNA‐Seq did not identify activation‐induced cytidine deaminase (AICDA) as a RCI suppressed mRNA, although our previous report did so). We may thus have missed some regulated gene products. Our approach was designed to minimize type I error by the use of stringent criteria for reproducibility throughout several experiments using unique cell donors. As a consequence, some gene products of interest may have been excluded from our analysis. For example, the mRNA encoding activation‐induced cytidine deaminase (AICDA, the key enzyme for the initiation of immunoglobulin heavy chain gene class‐switch recombination and somatic hypermutation in activated B cells) was found in our previous study 7 to be suppressed by RCI treatment. In the current RNA‐Seq experiments, however, although AICDA transcripts were reduced by almost 80%, on average, the change observed in two subjects was of borderline statistical significance, and AICDA was excluded from the analysis. Thirdly, while our studies were designed to explore the direct effects of RCI treatment on human B cells, the effects of RCI in vivo would be expected to be more complex, as indirect actions mediated through other cells of the immune system and through the glucocorticoid‐producing cells of the adrenal cortex would probably also be operative. Finally, because RCI is a naturally derived product containing a complex mixture of porcine ACTH analogues, it is not yet known whether the effects of this product on gene expression can be attributed to effects of full‐length ACTH alone. Expression of melanocortin receptors (MCRs) 49 on human B cells has been reported previously using studies at the mRNA 50 and protein levels 4, although the role of these specific MCRs in mediating functional humoral immune responses is not well understood.

These experiments have taken an unbiased approach to the identification of specific gene products and specific functional pathways that are modulated by the action of RCI on human B cells. The data support the contention that RCI exerts direct specific effects on the development of cells of the humoral immune system in humans, and that these effects are clearly distinct from any actions mediated by induction of steroidogenesis and consequent glucocorticoid action in vivo. This information will serve as a basis for advancement of our understanding of the immunomodulatory effects of RCI, and its mode of action for the treatment of autoimmune diseases.

Supporting information

Additional Supporting Information may be found in the online version of this article at the publisher's website:

Fig. S1. Immunoblots confirm modulation of protein expression by repository corticotrophin injection (RCI) in interleukin (IL)‐4/CD40 ligand (CD40L)‐activated human B lymphocytes in vitro. (a) Haem oxygenase 1 (HMOX1) expression is up‐regulated at the protein level in RCI‐treated cells. Lysates from isolated human CD19⁺ cells were activated with IL‐4 and CD40L for 20 h without additions (left lane), with placebo gel (PBO; middle lane) or with RCI (right lane) at the high dose level used in RNA‐Seq and reverse transcription–polymerase chain reaction (RT–PCR) experiments. Proteins were separated by electrophoresis on denaturing gels, transferred to polyvinylidene (PVDF) fluoride membranes, immunoblotted with a specific polyclonal rabbit antibody directed against HMOX1 and a peroxidase‐conjugated secondary goat anti‐rabbit antibody. The blot was imaged using chemiluminescent substrate. Findings were confirmed with a replicate immunoblot experiment using cells from a second unique donor. (b) Signalling lymphocytic activation molecule 1 (SLAMF 1) (CD150) expression is down‐regulated at the protein level in RCI‐treated cells. Lysates from isolated human CD19⁺ cells were activated with IL‐4 and CD40L for 20 h without additions (left lane), with placebo gel (PBO; middle lane) or with RCI (right lane) at the high dose level used in RNA‐Seq and RT–PCR experiments. Proteins were separated by electrophoresis on denaturing gels, transferred to PVDF membranes, immunoblotted with a specific mouse monoclonal antibody directed against CD150 and peroxidase‐conjugated mouse immunoglobulin (Ig)G kappa binding protein. The blot was imaged using chemiluminescent substrate. Findings were confirmed with a replicate immunoblot experiment using cells from a second unique donor.

Table S1. Repository corticotrophin injection (RCI)‐modulated mRNAs at 19–24 h

Table S2. Repository corticotrophin injection (RCI)‐modulated mRNAs at 6 days

Table S3. Dexamethasone‐modulated mRNAs at 19–24 h

Table S4. GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways