Unravelling transcriptomic complexity in breast cancer through modulation of DARPP-32 expression and signalling pathways

doi:10.21203/rs.3.rs-3029600/v1

Download PDF

Article

Unravelling transcriptomic complexity in breast cancer through modulation of DARPP-32 expression and signalling pathways

https://doi.org/10.21203/rs.3.rs-3029600/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 30 Nov, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

DARPP-32 is a key regulator of protein-phosphatase-1 (PP-1) and protein kinase A (PKA), with its function dependent upon its phosphorylation state. We previously identified DKK1 and GRB7 as genes with linked expression using Artificial Neural Network (ANN) analysis; here, we determine protein expression in a large cohort of early-stage breast cancer patients. Low levels of DARPP-32 Threonine-34 phosphorylation and DKK1 expression were significantly associated with poor patient prognosis, while low levels of GRB7 expression were linked to better survival outcomes.

To gain insight into mechanisms underlying these associations, we analysed the transcriptome of T47D breast cancer cells following DARPP-32 knockdown. We identified 202 differentially expressed transcripts and observed that some overlapped with genes implicated in the ANN analysis, including PTK7, TRAF5, and KLK6, amongst others. Furthermore, we found that treatment of DARPP-32 knockdown cells with 17β-estradiol or PKA inhibitor fragment (6–22) amide led to the differential expression of 193 and 181 transcripts respectively.

These results underscore the importance of DARPP-32, a central molecular switch, and its downstream targets, DKK1 and GRB7 in breast cancer. The discovery of common genes identified by a combined patient/cell line transcriptomic approach provides insights into the molecular mechanisms underlying differential breast cancer prognosis and highlights potential targets for therapeutic intervention.

Biological sciences/Cancer/Breast cancer

Biological sciences/Cancer/Tumour biomarkers

DARPP-32, first identified four decades ago, is expressed in dopamine innervated brain regions and is highly concentrated in the basal ganglia; it is encoded by PPP1R1B^1–3. Its initial discovery was as a potent inhibitor of protein-phosphatase-1 (PP-1) ⁴, and its inhibitory action was quickly found to be dependent on its phosphorylation status, mediated through cAMP-dependent protein kinase (PKA) ⁵. Over time, DARPP-32 has been shown to play a pivotal role in regulating both biochemical and electrophysical responses to a range of physiological and pharmacological stimuli. Notably, DARPP-32 phosphorylation is modulated by dopamine, and numerous other neurotransmitters, and can also be influenced by drugs of abuse ^6,7. DARPP-32 acts as a bifunctional signalling protein that can either inhibit kinases or phosphatases dependent upon the phosphorylation of its principal phosphorylation sites, Threonine (Thr)-34 and Thr-75. Phosphorylation of DARPP-32 at Thr-34 by PKA results in potent inhibition of PP-1; while phosphorylation at Thr-75 by cyclin-dependent kinase 5 (Cdk5) inhibits PKA ⁸.

DARPP-32 signalling, a vital molecular switch in the brain, has been implicated in cancer. Notably, truncated-DARPP (t-DARPP), a cancer-specific variant of DARPP-32 lacking the Thr-34 phosphorylation site, was originally described in gastric cancer ⁹. Increased expression of both DARPP-32 and t-DARPP has been observed in breast, prostate, colon, and stomach cancer tissue when compared with normal tissue ^10,11. Furthermore, studies have indicated that DARPP-32 expression occurs after a phase of dysplasia in oesophageal squamous cell carcinomas, with tumours displaying DARPP-32 expression progressing at a slower pace than those without ¹².

The link between DARPP-32 and cancer, extends to breast tumours. High levels of a combination of both DARPP-32 and t-DARPP expression have been linked with poor survival of breast cancer patients (n = 230) ¹³ and t-DARPP expression has been shown increasing during mammary tumour development ¹⁵. Our research team conducted a comprehensive analysis of tumours from over 3000 patients showing low levels of DARPP-32 expression was associated with poor survival, especially in oestrogen receptor (ER) positive tumours ^16–19. Low PPP1R1B mRNA expression was also linked to worse prognosis of patients with ER positive tumours and CDC42 and GRB7, amongst others, were identified as related genes using Artificial Neural Network (ANN) analysis ¹⁶.

Transfection of DARPP-32 into ER positive breast cancer cell line, MCF-7, resulted in reduced cellular migration, but this effect was not observed in ER negative breast cancer cell line, MDA-MB-231, ER- that lacked expression of the adhesion receptor DDR1 ²⁰. However, co-expression of DARPP-32 and DDR1 in MDA-MD-231 cells resulted in reduced migration, which was dependent on DARPP-32 Thr-34 phosphorylation ²⁰. Further analysis revealed that cAMP elevation triggered by Wnt-5A at the plasma membrane was responsible for Thr-34 phosphorylation, leading to DARPP-32 dependent inhibition of breast cancer migration ²¹.

Emerging evidence indicates that DARPP-32 may play a role in receptor signalling in breast cancer. Specifically, t-DARPP expression in HER2 positive breast cancer cells is linked to trastuzumab resistance, a commonly used HER2 targeted therapy, through increased Bcl-2 expression ²². This resistance appears to be mediated by t-DARPP phosphorylation at Thr-75, which results in Akt phosphorylation ¹³, and subsequent cellular proliferation and growth ¹⁴. Interestingly, co-expression of DARPP-32 can reverse the effect of t-DARPP on trastuzumab resistance and Akt phosphorylation ²³, and the findings have been expanded into oesophageal adenocarcinomas ²⁴ and other HER-2 targeted agents, such as lapatinib ²⁵. In addition, t-DARPP expression can sensitise cells to epidermal growth factor receptor in the presence of trastuzumab ²⁶. RNA-sequencing (RNA-Seq) of trastuzumab-resistant breast cancer cells further identified PPP1R1B upregulation ²⁷.

Although research on the link between DARPP-32 and oestrogen is limited, DARPP-32 expression is associated with the survival of ER positive patients ¹⁶, and oestrogen stimulation increases DARPP-32 Thr-34 phosphorylation ²⁸. In contrast, the role of PKA in oestrogen signalling has been extensively studied. PKA can phosphorylate ERα at Ser-305 to redirect the receptor to new transcriptional start sites, which is associated with tamoxifen resistance in breast cancer; typically, through MYC ²⁹. Furthermore, indirect links between links between dopamine and oestrogen in the brains of women have been reported (reviewed in ³⁰), including dopaminergic activation of ER ³¹.

The precise mechanisms by which DARPP-32 exerts its influence on breast cancer remain poorly understood, Nevertheless, current evidence suggests that DARPP-32 may act as a key regulator and/or activator of several critical signalling pathways implicated in cancer development and progression. To shed light on these complex interactions, the present investigation seeks to identify the specific transcriptomic changes that arise following DARPP-32 downregulation in breast cancer.

METABRIC data set

Details of the METABRIC data set (n = 1980) have been published elsewhere ³². Cohort samples were collected at five centres in the UK and Canada between 1977–2005 with appropriate consent from the respective institutional review boards as reported in the original publication. Breast cancer specific survival was calculated as the time interval between primary surgery and death resultant from breast cancer. Almost all ER negative and lymph node positive patients received adjuvant chemotherapy, whereas ER negative and/or lymph node positive patient did not. No patients with HER2 overexpression received trastuzumab. Median follow-up was 141 months determined using the reverse Kaplan-Meier method.

Cell culture

T47D breast cancer cells were purchased from American Type Culture Collection and were maintained in Dulbecco’s Modified Eagle’s medium with high glucose (Sigma) and supplemented with 10% iron supplemented donor bovine serum (Gibco) and 1% penicillin/streptomycin (Sigma). Cell line identity was verified every 24 months using short tandem repeat (STR) verification and were routinely monitored for mycoplasma infection.

Knockdown or control cells were treated with 10nM 17β-estradiol (E2) (Sigma) dissolved in absolute ethanol for 24 hours. E2 treatment was performed in in phenol red free media containing 10% charcoal stripped FBS (Gibco, UK) to remove endogenous serum steroids and eliminate the known weak oestrogen agonistic activity of phenol red. Cells were also treated with 3µM PKA inhibitor (PKA inhibitor fragment (6–22) amide) (Tocris 1904) dissolved in water, for 24 hours. PKA inhibition was performed in Dulbecco’s modified Eagle’s medium with high glucose (Sigma, UK) and 10% iron supplemented Donor Calf Serum (Gibco, UK) without the presence of antibiotics.

The phosphorylation state of DARPP-32, and PKA activity were determined following drug treatment using ELISA and following manufacturer’s instructions (PKA activity ELISA: Thermo Fisher, EIAPKA, DARPP-32 (Thr-34): Assay Genie CBCAB00394, DARPP-32 (Thr-75): Assay Genie CBCAB0010).

DARPP-32 knockdown

DARPP-32 expression was knocked down in T47D cells using 8nM siRNA (OriGene SR313371C) or 8nM negative control siRNA (OriGene SR30004) and Lipofectamine RNAiMax (Thermo Fisher), in Opti-MEM reduced serum media (Thermo Fisher). DARPP-32 knockdown was confirmed using Western blotting and PCR. Gel electrophoresis was performed using the Invitrogen Bolt mini system, with 4–12% Bis-Tris plus gels. Lysates were prepared in Bolt LDS sample buffer and Bolt sample reducing buffer and denatured by incubation at 100°C for five minutes. Transfer to nitrocellulose (Whatman, GE Heathcare) was achieved using Bolt transfer buffer with 10% methanol. Nitrocellulose was blocked in 3% non-fat milk for one hour. Anti-beta-actin (Abcam AB8226, 1:1000) and anti-DARPP-32 (Abcam AB40801, 1:1000) antibodies were incubated on membranes overnight at 4°C. Secondary antibodies, donkey anti-rabbit immunoglobulin (Li-Cor 926-32213, 1:10000) and donkey anti-mouse immunoglobulin (Li-Cor 926-68072 1:10000) were incubated for one hour prior to visualisation on an Odyssey FC Imager (Li-Cor) using Image Studio software (V4.1). Knockdown of DARPP-32 was also confirmed using PCR, with RNA isolated using RNAprotect Cell Reagent and RNeasy Plus Mini Kit (both Qiagen). Synthesis of cDNA was achieved using RT2 First Strand kit (Qiagen) prior to real-time PCR. Real-time PCR was performed using SYBR Green ROC qPCR Mastermix (Qiagen) and specific primers (DARPP-32 primers (Qiagen 249900), and HPRT primers (forward: 5’-AAATTCTTTGCTGACCTGCTG; reverse: 5'-TCCCCTGTTGACTGGTCATT)) using Viia7TM Real-Time PCR system (Applied Biosystems) operated under Quant StudioTM Real-Time PCR software.

RNA-Seq

RNA extraction, sample quality control, mRNA library preparation and RNA-Seq was performed by Novogene. Briefly, mRNA was purified from total RNA via poly-T oligo attached magnetic beads prior to fragmentation which was followed by first strand cDNA synthesis using random hexamer primers then second strand cDNA synthesis. The library was complete following end-repair, A-tailing, adapter ligation, size selection, amplification and purification. The library was checked with Qubit and real-time PCR for quantification and bioanalyser for size detection before being sequenced on Illumina NovaSeq platforms which utilise a paired-end 150 base pair sequencing strategy. Data output was greater than 20 million read pairs per sample.

RNA-Seq assessment

Trim Galore version 0.6.7 was used to to remove the first 13 base pairs of Illumina standard adaptors (AGATCGGAAGAGC), perform quality trimming (Phred score cutoff 20), and allowed subsequent FastQC analysis. Kallisto was used to quantify the abundance of transcripts from RNA-Seq data using pseudoalignment ³³. The human reference genome was GENCODE GRCh38 version 36 for transcript identification and quantification. The differential expression of transcripts was determined using DESeq2 (1.36.0) in the statistical environment RStudio, relying on input from Kallisto using Tximport. RStudio version 2022.07.1 + 554 running R version 4.2.0, with Bioconductor version 3.15 were used for assessments ³⁴. The list of transcripts was subject to a multiple test adjustment ranked by a P < 0.05 and a greater than two-fold change. Gene enrichment analysis was performed using Qiagen Ingenuity Pathway Analysis (IPA) version 76765844.

Nottingham patient cohort and immunohistochemistry

Patients with early-stage invasive breast cancer were treated at Nottingham University Hospitals between 1998 and 2006 and underwent wither breast conserving surgery or mastectomy, which was decided by disease characteristics or patient choice, followed by radiotherapy if indicated. Nottingham Prognostic Index (NPI), ER and menopausal status determined if patients received systemic adjuvant treatment. Patients with an NPI score less than 3.4 did not receive adjuvant treatment, and patients with an NPI score of 3.4 and above were candidates for CMF combination chemotherapy (cyclophosphamide, methotrexate and 5-fluorouracil) if they were ER-negative or pre-menopausal; and hormonal therapy if they were ER-positive. No patients received trastuzumab.

Immunohistochemistry was performed on tissue microarrays that were comprised of single 0.6mm cores taken from representative tumour areas selected by a specialist breast cancer histopathologist from haematoxylin and eosin stained sections. Tissue microarray sections (4µm) were initially deparaffinised and rehydrated in sequentially in xylene, ethanol and water prior to antigen retrieval in 0.01mol L^-1 sodium citrate buffer (pH 6.0), with tissue heated in a microwave for 10 minutes at 750W, and then 10 minutes at 450W. Staining was performed using a Novolink Polymer Detection kit (Leica) using the manufacturer’s instructions. Primary antibodies wee incubated on tissue for one hour at room temperature (anti-DKK1 1:1000 (Thermo Fisher Scientific, MA5-32229); anti-GRB7 1:500 (Abcam, ab183737); anti-DARPP-32 Thr-34 phosphorylation 1:500 (Abcam ab254063).

Following staining, tissue was dehydrated in ethanol and fixed in xylene prior to mounting using DPX. For each staining run, control breast composite sections comprised of grade 1 and 2 tumours were utilised. Slides were scanned at 200x magnification using a Nanozoomer Digital Pathology Scanner (Hamamatsu Photonics). An immunohistochemical H-score technique was used to assess cytoplasmic staining, whereby the percentage area of tumour staining was classified as 0 to 3, representing, none, weak, intermediate and strong intensity staining. Nuclear staining was scored as the percentage of tumour cells demonstrating any level of staining. Greater than 30% of cores were scored by an independent assessor, with single measure intraclass correlation coefficient values above 0.7 indicating good concordance between scorers.

Statistics

Statistical analysis was performed using IBM SPSS Statistics (version 28). Cases were stratified based on breast cancer specific survival using X-Tile software and breast cancer specific survival ³⁵. All differences were deemed statistically significant at the level of P ≤ 0.05. The Pearson χ2 test of association was used to determine the relationship between categorised protein expression and clinicopathological variables. Survival curves were plotted according to the Kaplan-Meier method with significance determined using the log-rank test.

Artificial neural network

ANN analysis was used to identify a stably enriched gene set associated with PPP1R1B expression in the METABRIC patient cohort and has been published previously ¹⁶. Three PPP1R1B probes were available for assessment and have been described before; briefly, probe 1 and 3 were located in areas found in the sequence for DARPP-32 (NM_032192), and probe 2 and 3 were located in areas found in the sequence for t-DARPP (NM_181505.3) ¹⁶.

DKK1 expression in early-stage breast cancer patients

DKK1 protein expression was determined in a cohort of early-stage breast cancer patients. Tissue from 1036 patients were available for assessment, the median H-score for cytoplasmic expression of DKK1 was 110 (ranging between 10 and 260), the median H-score for nuclear DKK1 expression was 25 (ranging between 0 and 100); representative tissue staining is shown in Fig. 1.

Low nuclear and cytoplasmic DKK1 expression was significantly associated with adverse breast cancer specific survival (P = 0.002, P = 0.031 respectively) (Fig. 2). Multivariate analysis was performed using Cox’s proportional hazard method, and included tumour size, tumour grade, tumour stage, NPI, ER status, PgR status, HER2 status and vascular invasion, both nuclear DKK1 and cytoplasmic DKK1 expression was not associated with patient survival in these models (hazard ratio (HR) = 0.834, 95% confidence interval (CI) = 0.625–1.113, P = 0.217, and HR = 0.934. 95% CI = 0.667–1.304, P = 0.687).

Figure 2: Kaplan-Meier analysis of breast cancer specific survival showing the impact of low (grey line) and high (black line) protein expression: (A) DKK1 cytoplasmic expression; (B) DKK1 nuclear expression; (C) GRB7 cytoplasmic expression; (D) GRB7 nuclear expression; (E) cytoplasmic DARPP-32 Thr-34 phosphorylation; (F) nuclear DARPP-32 Thr-34 phosphorylation.

Low levels of DKK1 cytoplasmic expression was significantly associated with larger tumour size (χ² = 5.177, d.f.=1, P = 0.023), higher tumour grade (χ² = 49.148, d.f.=1, P < 0.001), marked nuclear pleomorphism (χ² = 17.603, d.f.=2, P < 0.001), more mitosis (χ² = 49610, d.f.=2, P < 0.001), intermediate NPI values (χ² = 6.592, d.f.=2, P = 0.037), ER negative tumours (χ² = 66.761, d.f.=1, P = 0.001), negative PgR status (χ² = 45.770, d.f.=1, P < 0.001), HER2 positive tumours (χ² = 6.632, d.f.=1, P = 0.010), triple receptor positive tumours (χ² = 56.570, d.f.=1, P < 0.001), and high Ki67 index (χ² = 21.096, d.f.=1, P < 0.001) at the time of presentation (Table 1). Low nuclear expression of DKK1 was significantly associated with younger patient age (χ² = 7.924, d.f.=1, P = 0.005), larger tumour size (χ² = 6.621, d.f.=1, P = 0.010), high tumour grade (χ² = 14.153, d.f.=2, P < 0.001), marked nuclear pleomorphism (χ² = 27.793, d.f.=2, P < 0.001), mitosis (χ² = 48.622, d.f.=2, P < 0.001), poor NPI values (X² = 41.645, d.f.=2, P < 0.001), ER negative tumours (χ² = 82.423, d.f.=1, P = 0.001), PgR negative tumours (χ² = 33.601, d.f.=1, P < 0.001), triple receptor positive tumours (χ² = 68.76, d.f.=1, P < 0.001), and high Ki67 index (χ² = 16.339, d.f.=1, P < 0.001) at the time of presentation (Table 1).

Table 1

Associations between the cytoplasmic and nuclear expression of DKK1 and GRB7 determined using immunohistochemistry with clinicopathological variables. The P values are resultant from Pearson χ² test of association and significant values (P < 0.05) are highlighted in bold. ER is oestrogen receptor and PgR is progesterone receptor.
	Cytoplasmic DKK1 expression			Nuclear DKK1 expression			Cytoplasmic GRB7 expression			Nuclear GRB7 expression
	Low	High	P value	Low	High	P value	Low	High	P value	Low	High	P value
Age
<50 years	247 (73.1%)	91 (26.9%)	0.055	137 (40.5%)	201 (59.5%)	0.005	371 (84.3%)	69 (15.7%)	< 0.001	334 (76.8%)	101 (23.3%)	0.008
≥50 years	469 (67.2%)	229 (32.8%)	0.055	221 (31.7%)	477 (68.3%)	0.005	878 (90.7%)	90 (9.3%)	< 0.001	794 (82.8%)	165 (17.2%)	0.008
Tumour Size
<2.0cm	414 (66.5%)	209 (33.5%)	0.023	196 (31.5%)	427 (68.5%)	0.01	773 (89.4%)	92 (10.6%)	0.326	698 (81.4%)	160 (18.6%)	0.602
≥2.0cm	302 (73.1%)	111 (26.9%)	0.023	162 (39.2%)	251 (60.8%)	0.01	476 (87.7%)	67 (12.3%)	0.326	430 (80.2%)	106 (19.8%)	0.602
Tumour Grade
1	87 (54.4%)	73 (45.6%)	< 0.001	34 (21.3%)	126 (78.8%)	< 0.001	213 (98.2%)	4 (1.8%)	< 0.001	198 (91.7%)	18(8.3%)	< 0.001
2	245 (62.3%)	148 (37.7%)		105 (26.7%)	288(73.3%)		537 (95.7%)	24 (4.3%)		495 (89.5%)	58 (10.5%)
3	384 (79.5%)	99 (25.5%)		219 (45.3%)	264 (54.7%)		499(79.2%)	131 (20.8%)		435 (69.6%)	190 (30.4%)
Pleomorphism
1	6 (40.0%)	9 (60.0%)	< 0.001	2 (13.3%)	13 (86.7%)	< 0.001	22 (100.0%)	0 (0.0%)	< 0.001	21 (95.5%)	1 (4.5%)	< 0.001
2	175 (61.6%)	109 (38.4%)		68 (23.9%)	216 (76.1%)		395 (97.8%)	9 (2.2%)		364 (91.0%)	36 (9.0%)
3	535 (72.6%)	202 (27.4%)		288(39.1%)	449 (60.9%)		832 (84.7%)	150 (15.3%)		743 (76.4%)	229 (23.6%)
Mitosis
1	280 (58.8%)	196 (41.2%)	< 0.001	120 (25.2%)	356 (74.8%)	< 0.001	656 (96.9%)	21 (3.1%)	< 0.001	605 (90.6%)	63 (9.4%)	< 0.001
2	146 (71.6%)	58 (28.4%)		66 (32.4%)	138 (67.6%)		239 (85.7%)	40 (14.3%)		217 (78.3%)	60 (21.7%)
3	290 (81.5%)	66 (18.5%)		172(48.3%)	184(51.7%)		354 (78.3%)	98 (21.7%)		306 (68.2%)	143(31.8%)
Vascular Invasion
Definite	500 (67.6%)	240 (32.4%)	0.089	250 (33.8%)	490 (66.2%)	0.409	915 (90.0%)	102 (10.0%)	0.016	832(82.6%)	175 (17.4%)	0.009
No/probable	216 (73.0%)	80 (27.0%)	0.089	108(36.5%)	188 (63.5%)	0.409	334 (85.4%)	57 (14.6%)	0.016	296 (76.5%)	91 (23.5%)	0.009
Tumour Stage
1	425 (67.5%)	205 (32.5%)	0.170	207 (32.9%)	423 (67.1%)	0.168	792 (90.4%)	84 (9.6%)	< 0.001	724 (83.3%)	145 (16.7%)	< 0.001
2	215 (73.4%)	78 (26.6%)		103 (35.2%)	190 (64.8%)		352 (90.0%)	39 (10.0%)		311 (81.0%)	73(19.0%)
3	75 (67.0%)	37 (33.0%)		47 (42.0%)	65(58.0%)		104 (74.3%)	36 (25.7%)		93 (66.4%)	47 (33.6%)
NPI
Good (≤ 3.4)	187 (55.5%)	150 (44.5%)	< 0.001	70 (20.8%)	267 (79.2%)	< 0.001	471 (97.1%)	14 (2.9%)	< 0.001	433 (90.2%)	47 (9.8%)	< 0.001
Intermediate (3.41–5.4)	399 (76.1%)	125 (23.9%)		215 (41.0%)	309 (59.0%)		601 (86.2%)	96 (13.8%)		532 (78.2%)	150 (21.8%)
Poor (> 5.4)	129 (74.1%)	45 (25.9%)		72 (41.4%)	102 (58.6%)		176 (78.2%)	49 (21.8%)		156 (69.6%)	68 (30.4%)
ER status
Negative	204 (91.5%)	19 (8.5%)	0.001	134 (60.1%)	89 (39.9%)	0.001	213 (75.8%)	68 (24.2%)	< 0.001	180 (64.7%)	98 (35.3%)	< 0.001
Positive	511 (62.9%)	301 (37.1%)	0.001	223(27.5%)	589 (72.5%)	0.001	1035 (91.9%)	91 (8.1%)	< 0.001	947 (84.9%)	168 (30.2%)	< 0.001
PgR status
Negative	346 (80.7%)	83(19.3%)	< 0.001	191 (44.5%)	238 (55.5%)	< 0.001	462 (80.5%)	112 (19.5%)	< 0.001	404 (70.9%)	166 (29.1%)	< 0.001
Positive	366 (60.9%)	235 (39.1%)	< 0.001	163 (27.1%)	438 (72.9%)	< 0.001	780 (94.4%)	46 (5.6%)	< 0.001	719 (88.1%)	97 (11.9%)	< 0.001
HER2 status
Negative	603 (67.6%)	289 (32.4%)	0.100	299 (32.5%)	593 (66.5%)	0.100	1186(97.1%)	35 (2.9%)	< 0.001	1091 (90.3%)	117 (9.7%)	< 0.001
Positive	112 (78.3%)	31 (21.7%)	0.100	58 (40.6%)	85 (59.4%)	0.100	62 (33.3%)	124 (66.7%)	< 0.001	36 (19.5%)	149 (80.5%)	< 0.001
Triple negative status
Negative	556 (64.4%)	307 (35.6%)	< 0.001	250 (29.0%)	613 (71.0%)	< 0.001	1048 (88.1%)	142 (11.9%)	0.032	953(80.9%)	225 (19.1%)	0.523
Positive	155 (93.9%)	10 (6.1%)	< 0.001	103 (62.4%)	62 (37.6%)	< 0.001	191 (93.2%)	14 (16.8%)	0.032	168 (82.8%)	35 (17.2%)	0.523
Lymph node status
Negative	425(67.5%)	205 (32.5%)	0.159	207 (32.9%)	423 (67.1%)	0.167	792(90.4%)	84 (9.6%)	0.009	724(83.3%)	145 (16.7%)	0.004
Positive	290 (71.6%)	115 (28.4%)	0.159	150 (37.0%)	225 (63.0%)	0.167	456 (85.9%)	75 (14.1%)	0.009	404 (77.1%)	120 (22.9%)	0.004
Ki67 index Groups
Negative	236 (60.7%)	153 (39.3%)	< 0.001	100(25.7%)	289 (74.3%)	< 0.001	543 (97.8%)	12 (2.2%)	< 0.001	498 (91.0%)	49 (9.0%)	< 0.001
Positive	292 (76.0%)	92 (24%)	< 0.001	151 (39.9%)	233 (60.7%)	< 0.001	442 (88.9%)	55 (11.1%)	< 0.001	404 (81.9%)	89 (18.1%)	< 0.001

Cytoplasmic DKK1 expression was significantly correlated with DKK1 nuclear expression (R² = 0.483, P < 0.001), and DARPP-32 Thr-34 cytoplasmic (R² = 0.079, P = 0.015) and nuclear expression (R² = 0.144, P < 0.001) phosphorylation. Nuclear DKK1 expression was significantly correlated with GRB7 cytoplasmic (R²=-0.106, P < 0.001) and nuclear expression (R²=-0.102, P = 0.002), and nuclear DARPP-32 Thr-34 phosphorylation (R² = 0.092, P = 0.005). A correlation matrix demonstrating the relationship between DKK1 expression and other variables is shown in Fig. 3A.

GRB7 expression in early-stage breast cancer patients

GRB7 expression was determined in a cohort of early-stage breast cancer patients. Tissue from 1408 patients were assessed where the median H-score for cytoplasmic expression of GRB7 was 0 (ranging from 0 to 290), the median H-score for nuclear GRB7 expression was 0 (ranging between 0 and 90); representative tissue staining is shown in Fig. 1. Low nuclear and cytoplasmic expression of GRB7 was significantly associated with good prognosis of breast cancer patients (P = 0.012, P = 0.003 respectively) (Fig. 2). Multivariate analysis was performed using Cox’s proportional hazard method, and included tumour size, tumour grade, tumour stage, NPI, ER status, PgR status, HER2 status and vascular invasion, both nuclear GRB7 and cytoplasmic GRB7 expression was not associated with patient survival in these models (HR = 0.953, 95% CI = 0.617–1.472, P = 0.828, and HR = 1.090. 95% CI = 0.656–1.809, P = 0.740).

Low levels of GRB7 cytoplasmic expression was significantly associated with older patient age (χ² = 12.308, d.f.=1, P < 0.001), low tumour grade (χ² = 103.669, d.f.=2, P < 0.001), less nuclear pleomorphism (χ² = 51.487, d.f.=2, P < 0.001), less mitosis (χ² = 96.619, d.f.=2, P < 0.001), absence of vascular invasion (χ² = 5.833, d.f.=1, P = 0.016), low tumour stage (χ² = 32263, d.f.=2, P = < 0.001), good NPI value (χ² = 63.142, d.f.=2, P < 0.001), ER positive tumours (χ² = 58.281, d.f.=1, P = 0.001), PgR positive tumours (χ² = 65.761, d.f.=1, P < 0.001), HER2 negative tumours (χ² = 655.473, d.f.=1, P < 0.001), triple receptor positive tumours (χ² = 4.586, d.f.=1, P = 0.032), negative lymph node status (χ² = 6.784, d.f.=1, P = 0.009), and low ki67 index (χ² = 34.862, d.f.=1, P < 0.001) at the time of presentation (Table 1). Low nuclear expression of GRB7 was significantly associated with older patient age (χ² = 7.007, d.f.=1, P = 0.008), low tumour grade (χ² = 94.413, d.f.=2, P < 0.001), less nuclear pleomorphism (χ² = 41964, d.f.=2, P < 0.001), less mitosis (χ² = 88.882, d.f.=2, P < 0.001), absence of vascular invasion (χ² = 6.817, d.f.=1, P = 0.009), low tumour stage (χ² = 22.317, d.f.=2, P < 0.001), good NPI value (χ² = 48.610, d.f.=1, P < 0.001), ER positive tumours (χ² = 58.682, d.f.=1, P < 0.001), PgR positive tumours (χ² = 64.840, d.f.=1, P < 0.001), HER2 negative tumours (χ² = 521347, d.f.=1, P < 0.001), negative lymph node status (χ² = 6.81, d.f.=1, P = 0.009), and low Ki67 index (χ² = 18.637, d.f.=1, P < 0.001) at the time of presentation (Table 1).

Cytoplasmic GRB7 expression was correlated with nuclear GRB7 expression (R² = 0.768, P < 0.001), DKK1 nuclear expression (R²=-0.106, P < 0.001) and DARPP-32 Thr-34 cytoplasmic (R² = 0.141, P < 0.001) and nuclear (R² = 0.112, P < 0.001) phosphorylation. Nuclear GRB7 expression was significantly associated with DKK1 nuclear expression (R²=-0.102, P = 0.002) and DARPP-32 Thr-34 cytoplasmic (R² = 0.083, P = 0.004) and nuclear (R² = 0.090, P = 0.002) phosphorylation. A correlation matrix demonstrating the relationship between GRB7 expression and other variables is shown in Fig. 3A.

Levels of DARPP-32 Thr-34 phosphorylation in early-stage breast cancer patients

The level of DARPP-32 Thr-34 phosphorylation was determined in a cohort of early-stage breast cancer patients. Tissue from 1274 patients were available for assessment, the median H-score for cytoplasmic expression of DARPP-32 Threonine-34 phosphorylation was 100 (ranging between 0 and 290), the median H-score for nuclear DARPP-32 Threonine-34 phosphorylation expression was 25 (ranging between 0 and 110); representative tissue staining is shown in Fig. 1. Low nuclear and cytoplasmic DARPP-32 Thr-34 phosphorylation was significantly associated with adverse breast cancer specific survival (P = 0.002, P < 0.001 respectively) (Fig. 2). Multivariate analysis was performed using Cox’s proportional hazard method, and included tumour size, tumour grade, tumour stage, NPI, ER status, PgR status, HER2 status and vascular invasion, both nuclear DARPP-32 Thr-34 phosphorylation remained associated with patient survival in these models (HR = 0.658, 95% CI = 0.443–0.978, P = 0.038), but not for cytoplasmic DARPP-32 Thr-34 phosphorylation (HR = 0.763. 95% CI = 0.579–1.008, P = 0.057).

Low levels of cytoplasmic DARPP-32 Thr-34 phosphorylation was significantly associated with larger tumour size (χ² = 7.057, d.f.=1, P = 0.008), tumour cell nuclear pleomorphism (χ² = 8.778, d.f.=2, P = 0.012), the presence of vascular invasion (χ² = 5.843, d.f.=1, P = 0.016), higher tumour stage (χ² = 17.966, d.f.=2, P < 0.001), poor NPI values (χ² = 12.216, d.f.=2, P = 0.002), positive lymph node status (χ² = 11.980, d.f.=1, P < 0.001), and low Ki67 index (χ² = 6.671, d.f.=1, P = 0.010) at the time of presentation (Table 2). Low nuclear DARPP-32 Thr-34 phosphorylation was significantly associated with a moderate tumour stage (χ² = 8.644, d.f.=2, P = 0.013), and positive lymph node status (χ² = 8.541, d.f.=1, P = 0.003) at the time of presentation (Table 2).

Table 2

Associations between the cytoplasmic and nuclear expression of DARPP-32 Thr-34 phosphorylation determined using immunohistochemistry with clinicopathological variables. The P values are resultant from Pearson χ² test of association and significant values (P < 0.05) are highlighted in bold. ER is oestrogen receptor and PgR is progesterone receptor.
	Cytoplasmic Thr-34 DARPP-32 phosphorylation			Nuclear Thr-34 DARPP-32 phosphorylation
	Low	High	P value	Low	High	P value
Age
<50 years	87 (20.7%)	334 (79.3%)	0.350	340 (80.8%)	81 (19.2%)	0.477
≥50 years	196 (23.0%)	657 (77.0%)	0.350	702 (82.4%)	150 (17.6%)	0.477
Tumour Size
<2.0cm	152 (19.7%)	619 (80.3%)	0.008	625 (81.2%)	145 (18.8%)	0.433
≥2.0cm	131 (26.0%)	372 (74.0%)	0.008	417 (82.9%)	86 (17.1%)	0.433
Tumour Grade
1	30 (15.7%)	161(84.3%)	0.064	150 (78.5%)	41 (21.5%)	0.399
2	115 (23.4%)	377 (76.6%)		403 (81.9%)	89 (18.1%)
3	138 (23.4%)	453 (76.6%)		484 (82.9%)	101 (17.1%)
Pleomorphism
1	2 (11.1%)	16 (88.9%)	0.012	15 (83.3%)	3 (16.7%)	0.836
2	59 (17.2%)	285 (82.8%)		278 (80.8%)	66(19.2%)
3	222 (24.3%)	690 (75.5%)		799 (82.2%)	162 (17.8%)
Mitosis
1	125(21.3%)	462(78.8%)	0.360	476(81.1%)	111 (18.9%)	0.48
2	67 (25.5%)	196 (74.5%)		222 (84.4%)	41 (15.6%)
3	91 (21.5%)	333 (78.5%)		344 (81.3%)	79(18.7%)
Vascular Invasion
Definite	185 (20.4%)	721(79.6%)	0.016	731 (80.7%)	175 (19.3%)	0.089
No/probable	98 (26.6%)	270 (73.4%)	0.016	311 (84.7%)	56 (15.3%)	0.089
Tumour Stage
1	147 (19.0%)	627 (81.0%)	< 0.001	613 (79.3%)	160 (20.7%)	0.013
2	90 (24.5%)	277 (75.5%)		316 (86.1%)	51 (13.9%)
3	46 (34.8%)	86 (65.2%)		112 (84.8%)	20 (15.2%)
NPI
Good (≤ 3.4)	77 (18.6%)	337 (81.4%)	0.002	325 (78.5%)	89 (21.5%)	0.072
Intermediate (3.41–5.4)	141 (21.8%)	507 (78.2%)		536 (82.8%)	111 (17.2%)
Poor (> 5.4)	65 (30.8%)	146 (69.2%)		180 (85.3%)	31 (14.7%)
ER status
Negative	67 (25.0%)	201 (75.0%)	0.220	221 (82.5%)	47 (17.5%)	0.766
Positive	216 (21.5%)	789 (78.5%)	0.220	820 (81.7%)	184 (18.3%)	0.766
PgR status
Negative	125 (23.9%)	399 (76.1%)	0.276	436 (83.2%)	88 (16.8%)	0.287
Positive	158 (21.3%)	585 (78.7%)	0.276	600 (80.9%)	142 (19.1%)	0.287
HER2 status
Negative	248 (22.5%)	852 (77.5%)	0.496	904 (82.3%)	195 (17.7%)	0.331
Positive	35 (20.2%)	138 (79.8%)	0.496	137 (79.2%)	36 (20.8%)	0.331
Triple negative status
Negative	233 (21.9%)	833 (78.1%)	0.416	868 (81.5%)	197 (18.5%)	0.469
Positive	48 (24.5%)	148 (75.5%)	0.416	164 (83.7%)	32 (16.3%)	0.469
Lymph node status
Negative	147 (19.0%)	627 (81.0%)	< 0.001	613 (79.3%)	160 (20.7%)	0.003
Positive	136 (27.3%)	363(72.7%)	< 0.001	428 (85.5%)	71 (14.2%)	0.003
Ki67 index Groups
Negative	116 (23.3%)	382 (76.7%)	0.010	412(82.7%)	86 (17.3%)	0.102
Positive	76 (16.6%)	382 (83.4%)	0.010	359 (78.6%)	98 (21.4%)	0.102

Correlations between DARPP-32 Thr-34 phosphorylation and DKK1 and GRB7 expression have already been described and shown in Fig. 3A. Cytoplasmic DARPP-32 Thr-34 phosphorylation was significantly correlated with DARPP-32 protein cytoplasmic (R² = 0.098, P < 0.001) and nuclear (R² = 0.108, P < 0.001) expression. Nuclear DARPP-32 Thr-34 phosphorylation was significantly correlated with DARPP-32 protein cytoplasmic (R² = 0.119, P < 0.001) and nuclear (R² = 0.111, P < 0.001) expression.

DARPP-32 knockdown in T47D breast cancer cell line

T47D breast cancer cells were treated with either human DARPP-32 siRNA oligo duplex or negative control siRNA to knockdown DARPP-32 expression. DARPP-32 expression was effectively reduced using siRNA knockdown; protein expression of DARPP-32 was reduced by over 95% determined using Western blotting, and over 60% at the mRNA level, using real-time PCR (Fig. 3B); negative control siRNA did not cause a reduction in DARPP-32 expression.

DARPP-32 knockdown cells were also subject to stimulation with E2 or a PKA inhibitor. The PKA inhibitor caused a dose dependent decrease in PKA activity determined by ELISA (Fig. 3C). For RNA-Seq, cells were treated with 3µM PKA inhibitor for 24 hours, which resulted in a 60% reduction in PKA activity, and a 60% reduction in DARPP-32 Thr-34 phosphorylation, which was also determined by ELISA (Fig. 3D).

RNA-Seq of DARPP-32 knockdown T47D breast cancer cells

When DARPP-32 expression was knocked down in T47D breast cancer cells, 202 differentially expressed transcripts were identified (listed in Supplementary File 1). Following stimulation with E2, 193 differentially expressed transcripts were identified between DARPP-32 knockdown cell treated with E2 versus control cell treated with E2 (listed in Supplementary File 1). When DARPP-32 knockdown T47D cells were treated with a PKA inhibitor and compared with control cells treated with a PKA inhibitor, 181 differentially expressed transcripts were identified (listed in Supplementary File 1). The lists of differentially expressed transcripts were explored for common transcripts between those identified following DARPP-32 knockdown, and then in the presence of E2 or a PKA inhibitor. PUF60, and SART3, were identified in DARPP-32 knockdown cells and when DARPP-32 knockdown cells were treated with E2. FIGNL1, TBK1, TSEN34, UBE3A, and ZCCHC7 were identified in DARPP-32 knockdown cells, and when DARPP-32 knockdown cells were treated with a PKA inhibitor. BCLAF1, CAST, CELF1, CTNBB1, KIAA1217 and SMARCE1 were common to DARPP-32 knockdown cells treated with E2 and a PKA inhibitor. RBM39 and SLC10A3 were common to all three datasets.

Qiagen IPA was used to find enriched canonical pathways, 14 were identified when DARPP-32 expression was knocked down. When DARPP-32 knockdown cells were treated with E2 105 pathways were significantly altered and 241 when DARPP-32 knockdown cells were treated with a PKA inhibitor (listed in Supplementary File 2). Expectedly dopamine-DARPP-32 feedback in cAMP signalling was one of the 14 pathways identified when DARPP-32 expression was knocked down; the tight junction pathway was common to all three assessments. Qiagen IPA assessment of upstream regulators identified on transcriptional regulator with predicted activation due to changes to target genes in the DARPP-32 knockdown dataset. Changes to CBFA2T3, CEMIP2, KDM3A, MTF2 and NOTCH4 expression indicated activation of the upstream regulator SOX2.

RNA-Seq assessment and commonalities with Artificial Neural Network analysis of METABRIC cohort

The significant differentially expressed transcripts identified through RNA-Seq of DARPP-32 knockdown T47D breast cancer cells were compared with those identified as associated with PPP1R1B expression in the METABRIC patient cohort using ANN analysis to find common genes. Of the 202 transcripts identified using RNA-Seq, seven of those were common to the top 300 genes identified using ANN analyses. All seven of the common genes were identified in gene lists from the ANN of PPP1R1B probes 2 and 3. PTK7, PPFIBP2 and PACSIN2 were identified in the ANN of PPP1R1B probes 2 and 3, whilst TRAF5, KLK6, GAL3ST4 were identified within the PPP1R1B probe 2 analysis, and LIMCH1 was identified within the PPP1R1B probe 3 analysis.

Common genes were also identified between the expression of PPP1R1B probes and DARPP-32 knockdown cells with treated with E2 and when treated with a PKA inhibitor. EHMT2, KCNIP2, GOLGA2, and ADCY1, were identified in the ANN of PPP1R1B probe 1 and when DARPP-32 knockdown cells were treated with E2. GSTCD was identified in the ANN of PPP1R1B probe 2 and when DARPP-32 knockdown cells were treated with E2. EPHB6, and VASN were identified in the ANN of PPP1R1B probe 3 and when DARPP-32 knockdown cells were treated with E2.

No common genes were identified in the ANN of PPP1R1B probe 1 when DARPP-32 knockdown cells were treated with a PKA inhibitor. BMPR1B, and MTA1 were identified in the ANN of PPP1R1B probe 2 when DARPP-32 knockdown cells were treated with a PKA inhibitor and ARL6IP4 was identified in the ANN of probe 3.

PAQR6 was identified in the ANN of PPP1R1B probe 1 and probe 2 and when DARPP-32 knockdown cells were treated with E2. No genes were identified that were common to all three PPP1R1B probes between the three datasets.

In a previous study, we established an association between low DARPP-32 expression and poor prognosis of breast cancer patients, particularly those with ER positive tumours ¹⁶. However, the underlying mechanism by which DARPP-32 impacts breast cancer cell behaviour remains unknown. Using an ANN analysis we identified 18 transcripts common to expression of all three available PPP1R1B probes within the top 200 transcripts for each probe, including DKK1 and GRB7. To further investigate this finding, we conducted a large-scale cohort study to determine GRB7 and DKK1 protein expression levels in combination with determining levels of DARPP-32 Thr-34 phosphorylation in the same patient specimens. In this study we utilised the level of DARPP-32 Thr-34 phosphorylation as a proxy for full-length DARPP-32 expression. DARPP-32 Thr-34 can act as a surrogate of full length DARPP-32 phosphorylation as t-DARPP is lacking this residue; unfortunately, DARPP-32 Thr-75 phosphorylation could not be determined with sufficient accuracy to warrant further study. Our results show that low levels of DARPP-32 Thr-34 phosphorylation was significantly associated with poor patient prognosis, which was in direct alignment with our previous findings ¹⁶. Notably, this association remains significant for nuclear DARPP-32 Thr-34 phosphorylation in multivariate survival analysis. Additionally, DARPP-32 Thr-34 phosphorylation was significantly associated with advanced tumour stage and lymph node involvement. This study provides further validation that DARPP-32 may be a clinically relevant biomarker in breast cancer.

DKK1 is a secretory antagonist of the classical Wnt signalling pathway; studies have indicated that dysregulation of the Wnt signalling pathway induced via the activity of DKK1 is crucial to cancer cell migration and bone metastasis in breast cancer and prostate cancer ^36,37. In contrast, DKK1 expression reduces cellular proliferation and migration ^38,39. In this study, low expression of DKK1 was significantly associated with adverse breast cancer specific survival. In addition, low levels of DKK1 expression was associated with larger tumours, higher tumour grade, and ER and PgR negative tumours. As expected, we observed a weak but statistically significant correlation between DKK1 expression and DARPP-32 Thr-34 phosphorylation.

GRB7 is a 532 amino acid adaptor molecule with a crucial role in the activation of multiple intracellular pathways through transmission of signals from cell membrane receptors. In this study low nuclear and cytoplasmic expression of GRB7 was significantly associated with good prognosis of breast cancer patients. In addition, low levels of GRB7 expression were associated with lower tumour grade, tumour stage, ER and PgR positive tumours and HER2 negative tumours. This is aligned with previous studies that have demonstrated that low GRB7 expression is associated with improved survival of breast cancer patients (n = 638) ⁴⁰ and GRB7 is included in the 21 gene set of Oncotype DX. As expected, we observed a weak but statistically significant correlation between GRB7 expression and DARPP-32 Thr-34 phosphorylation. An association between PPP1R1B and GRB7 has been previously demonstrated in upper gastrointestinal adenocarcinomas ⁴¹.

Using the ER positive breast cancer cell line, T47D, we sought to identify transcriptomic changes that follow changes in DARPP-32 expression, in the presence and absence of oestrogen and a PKA inhibitor. The actions of E2, or 17β-estradiol, are mediated by ER, and 10nM E2 treatment in T47D cells has been shown to increase cellular proliferation with significant changes in the transcriptome, including the expression of important cancer associated genes, such as MYC^42–45. PKA inhibition with PKA inhibitor fragment (6–22) amide was used determine transcriptomic changes that occur following reduction of PKA mediated DARPP-32 Thr-34 phosphorylation; however, PKA inhibition would also be expected to alter a wide range of cell signalling effects directly through PKA. PKA inhibitor fragment (6–22) amide is a potent inhibitor of PKA that is derived from the active portion of the heat stable PKA inhibitor protein PKI. In T47D cells, 24-hour 3µM PKA inhibitor treatment resulted in a 60% inhibition of PKA, and a 60% decrease in DARPP-32 Thr-34 phosphorylation. 202 differentially expressed transcripts were identified following knockdown of DARPP-32. Following stimulation with E2, 193 differentially expressed transcripts were identified and 181 following treatment with a PKA inhibitor. Transcripts common to multiple assessments were identified, with PUF60, and SART3 common to DARPP-32 knockdown cells and DARPP-32 knockdown cells treated with E2. FIGNL1, TBK1, TSEN34, UBE3A, and ZCCHC7 were common to DARPP-32 knockdown cells and DARPP-32 knockdown cells treated with a PKA inhibitor. RBM39 and SLC10A3 were common to all three DARPP-32 knockdown RNA-Seq data sets; RBM39 encodes RNA binding motif protein 39 (RBM39), and SLC10A3 encodes solute carrier family 10 member 3.

Transcripts common to multiple assessments included a number of genes linked to breast cancer. Expression of PUF60, a spliceosome component, is associated with overall survival in breast cancer ⁴⁶, and SART3 has previously been identified as expressed in a large proportion of breast cancer cell tissue ⁴⁷. Serum levels of TBK1, are associated with the clinical outcome of breast cancer patients ⁴⁸ and ubiquitin protein ligase E3A (UBE3A) has been shown to be overexpressed in breast cancer ⁴⁹. RBM39 and SLC10A3 were common to all datasets, are linked with tumour cell behaviour ⁵⁰ and altered expression cancer ⁵¹.

The transcripts identified through RNA-Seq of DARPP-32 knockdown T47D breast cancer cells were compared with those identified as associated with PPP1R1B expression in the METABRIC patient cohort. Seven genes were identified that were common between lists, with all identified in ANN gene lists from PPP1R1B probe 2 and 3. PTK7, PPFIBP2 and PACSIN2 were identified in the ANN of PPP1R1B probes 2 and 3, TRAF5, KLK6, GAL3ST4 were identified within the PPP1R1B probe 2 analysis, and LIMCH1 was identified within the PPP1R1B probe 3 analysis. Common genes were also identified between the METABRIC ANN and in DARPP-32 knockdown cells treated with E2 and the PKA inhibitor. Of these genes, PTK7, KLK6 and LIMCH1 have interesting links with breast cancer; PTK7 is a catalytically inactive receptor tyrosine kinase in the Wnt signalling pathway, and a PTK7 targeted antibody-drug conjugate has been shown to reduce tumour-initiating cells ⁵². Both KLK6 and LIMCH1 expression has been linked to clinical outcome of breast cancer patients ^53,54

The significant associations observed between DARPP-32 and its downstream targets DKK1 and GRB7 and patient survival, underscore their importance in breast cancer. By using an approach that combines patient and cell line transcriptomics we have identified a number of common genes that provides an insight into the molecular mechanisms underlying differential breast cancer prognosis and highlights potential targets for therapeutic intervention.

Acknowledgements

We thank the Nottingham Health Science Biobank and Breast Cancer Now for the provision of tissue samples. S.J.S. was funded through the University of Nottingham’s Research Vision as a Nottingham Research Fellow, and the University of Nottingham Life Cycle 6 Campaign. Megan-Rose Greener is funded through the Wellcome Trust, grant number RS3684.

Author contributions

B.S., performed immunohistochemistry, knockdown experiments and statistical analysis; R.V, and R.D performed second scoring; M-R.G, and A.I. assisted with gene enrichment analysis; A..G, E.R., I.E. provided tissue and clinical datasets, G.B performed ANN, S.G.M and S.J.S were responsible for study design; S.J.S conceived the study, secured research funding, performed statistics and bioinformatics and wrote the draft manuscript. All authors were provided the manuscript for comment.

Competing interests

The authors declare no competing interests.

Data sharing

The data have been deposited in NCBI’s Gene Expression Omnibus (GEO) and are accessible through GEO Series accession number GSE204836. The METABRIC data is publicly available https://www.ebi.ac.uk/ega/studies/EGAS00000000098. Code related to the analysis can be found on github at https://github.com/sarahstorr.

Research involving human participants

Ethical approval for the patient cohort was granted by Nottingham Research Ethics Committee 2, under the title ‘Development of a molecular genetic classification of breast cancer’ (C202313). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All samples collected from Nottingham used in this study were pseudo-anonymised; those collected prior to 2006 did not require informed patient consent under the Human Tissue Act, after 2006 informed consent was obtained from all individual participants included in the study.

Hemmings, H. C., Jr., Nairn, A. C., Aswad, D. W. & Greengard, P. DARPP-32, a dopamine- and adenosine 3':5'-monophosphate-regulated phosphoprotein enriched in dopamine-innervated brain regions. II. Purification and characterization of the phosphoprotein from bovine caudate nucleus. J Neurosci 4, 99–110 (1984).
Walaas, S. I. & Greengard, P. DARPP-32, a dopamine- and adenosine 3':5'-monophosphate-regulated phosphoprotein enriched in dopamine-innervated brain regions. I. Regional and cellular distribution in the rat brain. J Neurosci 4, 84–98 (1984).
Ouimet, C. C., Miller, P. E., Hemmings, H. C., Jr., Walaas, S. I. & Greengard, P. DARPP-32, a dopamine- and adenosine 3':5'-monophosphate-regulated phosphoprotein enriched in dopamine-innervated brain regions. III. Immunocytochemical localization. J Neurosci 4, 111–124 (1984).
Hemmings, H. C., Jr., Greengard, P., Tung, H. Y. & Cohen, P. DARPP-32, a dopamine-regulated neuronal phosphoprotein, is a potent inhibitor of protein phosphatase-1. Nature 310, 503–505, doi:10.1038/310503a0 (1984).
Hemmings, H. C., Jr., Nairn, A. C. & Greengard, P. DARPP-32, a dopamine- and adenosine 3':5'-monophosphate-regulated neuronal phosphoprotein. II. Comparison of the kinetics of phosphorylation of DARPP-32 and phosphatase inhibitor 1. J Biol Chem 259, 14491–14497 (1984).
Fienberg, A. A. & Greengard, P. The DARPP-32 knockout mouse. Brain Res Brain Res Rev 31, 313–319, doi:10.1016/s0165-0173(99)00047-8 (2000).
Greener, M. R. & Storr, S. J. Exploring the Role of DARPP-32 in Addiction: A Review of the Current Limitations of Addiction Treatment Pathways and the Role of DARPP-32 to Improve Them. NeuroSci 3, 494–509 (2022).
Bibb, J. A. et al. Phosphorylation of DARPP-32 by Cdk5 modulates dopamine signalling in neurons. Nature 402, 669–671, doi:10.1038/45251 (1999).
El-Rifai, W. et al. Gastric cancers overexpress DARPP-32 and a novel isoform, t-DARPP. Cancer Res 62, 4061–4064 (2002).
Beckler, A. et al. Overexpression of the 32-kilodalton dopamine and cyclic adenosine 3',5'-monophosphate-regulated phosphoprotein in common adenocarcinomas. Cancer 98, 1547–1551, doi:10.1002/cncr.11654 (2003).
Wang, M. S. et al. Overexpression of DARPP-32 in colorectal adenocarcinoma. Int J Clin Pract 59, 58–61, doi:10.1111/j.1742-1241.2004.00305.x (2005).
Ebihara, Y. et al. DARPP-32 expression arises after a phase of dysplasia in oesophageal squamous cell carcinoma. Br J Cancer 91, 119–123, doi:10.1038/sj.bjc.6601899 (2004).
Hamel, S. et al. Both t-Darpp and DARPP-32 can cause resistance to trastuzumab in breast cancer cells and are frequently expressed in primary breast cancers. Breast Cancer Res Treat 120, 47–57, doi:10.1007/s10549-009-0364-7 (2010).
Vangamudi, B. et al. t-DARPP regulates phosphatidylinositol-3-kinase-dependent cell growth in breast cancer. Mol Cancer 9, 240, doi:10.1186/1476-4598-9-240 (2010).
Christenson, J. L. & Kane, S. E. Darpp-32 and t-Darpp are differentially expressed in normal and malignant mouse mammary tissue. Mol Cancer 13, 192, doi:10.1186/1476-4598-13-192 (2014).
Kotecha, S. et al. Dopamine and cAMP-regulated phosphoprotein 32 kDa (DARPP-32) and survival in breast cancer: a retrospective analysis of protein and mRNA expression. Sci Rep 9, 16987, doi:10.1038/s41598-019-53529-z (2019).
Martin, S. G. et al. Dopamine and cAMP-regulated phosphoprotein 32kDa (DARPP-32), protein phosphatase-1 and cyclin-dependent kinase 5 expression in ovarian cancer. J Cell Mol Med 24, 9165–9175, doi:10.1111/jcmm.15553 (2020).
Saidy, B. et al. PP1, PKA and DARPP-32 in breast cancer: A retrospective assessment of protein and mRNA expression. J Cell Mol Med 25, 5015–5024, doi:10.1111/jcmm.16447 (2021).
Saidy, B. et al. Retrospective assessment of cyclin-dependent kinase 5 mRNA and protein expression and its association with patient survival in breast cancer. J Cell Mol Med 24, 6263–6271, doi:10.1111/jcmm.15268 (2020).
Hansen, C., Greengard, P., Nairn, A. C., Andersson, T. & Vogel, W. F. Phosphorylation of DARPP-32 regulates breast cancer cell migration downstream of the receptor tyrosine kinase DDR1. Exp Cell Res 312, 4011–4018, doi:10.1016/j.yexcr.2006.09.003 (2006).
Hansen, C. et al. Wnt-5a-induced phosphorylation of DARPP-32 inhibits breast cancer cell migration in a CREB-dependent manner. J Biol Chem 284, 27533–27543, doi:10.1074/jbc.M109.048884 (2009).
Belkhiri, A. et al. Expression of t-DARPP mediates trastuzumab resistance in breast cancer cells. Clin Cancer Res 14, 4564–4571, doi:10.1158/1078-0432.CCR-08-0121 (2008).
Gu, L., Waliany, S. & Kane, S. E. Darpp-32 and its truncated variant t-Darpp have antagonistic effects on breast cancer cell growth and herceptin resistance. PLoS One 4, e6220, doi:10.1371/journal.pone.0006220 (2009).
Hong, J. et al. Regulation of ERBB2 receptor by t-DARPP mediates trastuzumab resistance in human esophageal adenocarcinoma. Cancer Res 72, 4504–4514, doi:10.1158/0008-5472.CAN-12-1119 (2012).
Christenson, J. L., Denny, E. C. & Kane, S. E. t-Darpp overexpression in HER2-positive breast cancer confers a survival advantage in lapatinib. Oncotarget 6, 33134–33145, doi:10.18632/oncotarget.5311 (2015).
Denny, E. C. & Kane, S. E. t-Darpp Promotes Enhanced EGFR Activation and New Drug Synergies in Her2-Positive Breast Cancer Cells. PLoS One 10, e0132267, doi:10.1371/journal.pone.0132267 (2015).
Murad, R. et al. Transcriptome and chromatin landscape changes associated with trastuzumab resistance in HER2 + breast cancer cells. Gene 799, 145808, doi:10.1016/j.gene.2021.145808 (2021).
Auger, A. P., Meredith, J. M., Snyder, G. L. & Blaustein, J. D. Oestradiol increases phosphorylation of a dopamine- and cyclic AMP-regulated phosphoprotein (DARPP-32) in female rat brain. J Neuroendocrinol 13, 761–768, doi:10.1046/j.1365-2826.2001.00700.x (2001).
de Leeuw, R. et al. PKA phosphorylation redirects ERalpha to promoters of a unique gene set to induce tamoxifen resistance. Oncogene 32, 3543–3551, doi:10.1038/onc.2012.361 (2013).
Del Rio, J. P. et al. Steroid Hormones and Their Action in Women's Brains: The Importance of Hormonal Balance. Front Public Health 6, 141, doi:10.3389/fpubh.2018.00141 (2018).
Olesen, K. M. & Auger, A. P. Dopaminergic activation of estrogen receptors induces fos expression within restricted regions of the neonatal female rat brain. PLoS One 3, e2177, doi:10.1371/journal.pone.0002177 (2008).
Curtis, C. et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486, 346–352, doi:10.1038/nature10983 (2012).
Bray, N. L., Pimentel, H., Melsted, P. & Pachter, L. Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol 34, 525–527, doi:10.1038/nbt.3519 (2016).
Anders, S. & Huber, W. Differential expression analysis for sequence count data. Genome Biol 11, R106, doi:10.1186/gb-2010-11-10-r106 (2010).
Camp, R. L., Dolled-Filhart, M. & Rimm, D. L. X-tile: a new bio-informatics tool for biomarker assessment and outcome-based cut-point optimization. Clin Cancer Res 10, 7252–7259, doi:10.1158/1078-0432.CCR-04-0713 (2004).
Pang, H. et al. Effects of DKK1 overexpression on bone metastasis of SBC-3 cells. Oncol Lett 15, 6739–6744, doi:10.3892/ol.2018.8160 (2018).
Thudi, N. K. et al. Dickkopf-1 (DKK-1) stimulated prostate cancer growth and metastasis and inhibited bone formation in osteoblastic bone metastases. Prostate 71, 615–625, doi:10.1002/pros.21277 (2011).
Chen, L., Li, M., Li, Q., Wang, C. J. & Xie, S. Q. DKK1 promotes hepatocellular carcinoma cell migration and invasion through beta-catenin/MMP7 signaling pathway. Mol Cancer 12, 157, doi:10.1186/1476-4598-12-157 (2013).
Gonzalez-Sancho, J. M. et al. The Wnt antagonist DICKKOPF-1 gene is a downstream target of beta-catenin/TCF and is downregulated in human colon cancer. Oncogene 24, 1098–1103, doi:10.1038/sj.onc.1208303 (2005).
Nadler, Y. et al. Growth factor receptor-bound protein-7 (Grb7) as a prognostic marker and therapeutic target in breast cancer. Ann Oncol 21, 466–473, doi:10.1093/annonc/mdp346 (2010).
Maqani, N. et al. Molecular dissection of 17q12 amplicon in upper gastrointestinal adenocarcinomas. Mol Cancer Res 4, 449–455, doi:10.1158/1541-7786.MCR-06-0058 (2006).
Strom, A. et al. Estrogen receptor beta inhibits 17beta-estradiol-stimulated proliferation of the breast cancer cell line T47D. Proc Natl Acad Sci U S A 101, 1566–1571, doi:10.1073/pnas.0308319100 (2004).
Dubik, D., Dembinski, T. C. & Shiu, R. P. Stimulation of c-myc oncogene expression associated with estrogen-induced proliferation of human breast cancer cells. Cancer Res 47, 6517–6521 (1987).
Katchy, A., Edvardsson, K., Aydogdu, E. & Williams, C. Estradiol-activated estrogen receptor alpha does not regulate mature microRNAs in T47D breast cancer cells. J Steroid Biochem Mol Biol 128, 145–153, doi:10.1016/j.jsbmb.2011.10.008 (2012).
Yu, S., Kim, T., Yoo, K. H. & Kang, K. The T47D cell line is an ideal experimental model to elucidate the progesterone-specific effects of a luminal A subtype of breast cancer. Biochem Biophys Res Commun 486, 752–758, doi:10.1016/j.bbrc.2017.03.114 (2017).
An, J. et al. Identification of spliceosome components pivotal to breast cancer survival. RNA Biol 18, 833–842, doi:10.1080/15476286.2020.1822636 (2021).
Suefuji, Y. et al. Expression of SART3 antigen and induction of CTLs by SART3-derived peptides in breast cancer patients. Br J Cancer 84, 915–919, doi:10.1054/bjoc.2000.1690 (2001).
Liu, H., Li, H., Zhang, J., Meng, Q. & Ma, L. Correlation of TBK1, AR, and other serum cancer-related biomarkers in breast cancer patients: An observational study. Medicine (Baltimore) 101, e29996, doi:10.1097/MD.0000000000029996 (2022).
Deng, S. et al. Over-expression of genes and proteins of ubiquitin specific peptidases (USPs) and proteasome subunits (PSs) in breast cancer tissue observed by the methods of RFDD-PCR and proteomics. Breast Cancer Res Treat 104, 21–30, doi:10.1007/s10549-006-9393-7 (2007).
Puvvula, P. K. et al. Inhibiting an RBM39/MLL1 epigenomic regulatory complex with dominant-negative peptides disrupts cancer cell transcription and proliferation. Cell Rep 35, 109156, doi:10.1016/j.celrep.2021.109156 (2021).
Tian, S., Li, J., Xiang, J. & Peng, P. The Clinical Relevance and Immune Correlation of SLC10 Family Genes in Liver Cancer. J Hepatocell Carcinoma 9, 1415–1431, doi:10.2147/JHC.S392586 (2022).
Damelin, M. et al. A PTK7-targeted antibody-drug conjugate reduces tumor-initiating cells and induces sustained tumor regressions. Sci Transl Med 9, doi:10.1126/scitranslmed.aag2611 (2017).
Haritos, C. et al. Kallikrein-related peptidase 6 (KLK6) expression differentiates tumor subtypes and predicts clinical outcome in breast cancer patients. Clin Exp Med 18, 203–213, doi:10.1007/s10238-018-0487-4 (2018).
Alifanov, V. V., Tashireva, L. A., Zavyalova, M. V. & Perelmuter, V. M. LIMCH1 as a New Potential Metastasis Predictor in Breast Cancer. Asian Pac J Cancer Prev 23, 3947–3952, doi:10.31557/APJCP.2022.23.11.3947 (2022).

No competing interests reported.

Download PDF

Journal Publication

published 30 Nov, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
20 Oct, 2023
Reviews received at journal
13 Oct, 2023
Reviewers agreed at journal
02 Oct, 2023
Reviewers invited by journal
23 Jun, 2023
Editor assigned by journal
23 Jun, 2023
Editor invited by journal
09 Jun, 2023
Submission checks completed at journal
09 Jun, 2023
First submitted to journal
06 Jun, 2023

You are reading this latest preprint version

Unravelling transcriptomic complexity in breast cancer through modulation of DARPP-32 expression and signalling pathways

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Methods

METABRIC data set

Cell culture

DARPP-32 knockdown

RNA-Seq

RNA-Seq assessment

Nottingham patient cohort and immunohistochemistry

Statistics

Results

Artificial neural network

DKK1 expression in early-stage breast cancer patients

GRB7 expression in early-stage breast cancer patients

Levels of DARPP-32 Thr-34 phosphorylation in early-stage breast cancer patients

DARPP-32 knockdown in T47D breast cancer cell line

RNA-Seq of DARPP-32 knockdown T47D breast cancer cells

RNA-Seq assessment and commonalities with Artificial Neural Network analysis of METABRIC cohort

Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1