Synthetic steroids, also denoted as glucocorticoids, remain cornerstone chemotherapeutic drugs in the treatment of acute lymphoblastic leukemia (ALL). Insufficient response to glucocorticoids during induction therapy remains an important predictor for inferior outcome in pediatric ALL, particularly for T-cell ALL (T-ALL).1,2 Activation of the glucocorticoid receptor (NR3C1) leads to its cytoplasmic-to-nuclear translocation, facilitating its function as a transcription factor resulting in cell death.3 Steroid treatment may therefore provide selective pressure of leukemic cells to acquire genetic events that reduce a functional steroid response resulting in therapy failure and relapse, as was recently demonstrated in an elegant T-ALL mouse model.4 An alternative underlying mechanism may include inhibitory phosphorylation of NR3C1 that impairs its nuclear localization and transactivation potential to activate important downstream target genes including BIM and NR3C1 itself.5,6 Low NR3C1 levels, the failure to upregulate NR3C1 as a positive feedback loop following steroid exposure or epigenetic silencing of BIM have all been linked to steroid resistance.7,8 Inactivation of NR3C1 by mutations or entire gene deletions have also been identified in leukemic cell lines9 and are reported in about 6% of ALL patients at relapse.10-12 In particular, over 16% of ETV6-RUNX1-positive pre-B ALL patients have acquired NR3C1 deletions at relapse.13 Point mutations or small insertion/deletion mutations in Nr3c1 were also identified in 18% of dexamethasone-resistant relapsed samples in the previously mentioned T-ALL mouse model.4
In our previous integrated study on pediatric T-ALL patients, where we combined genetic data with clinical data including outcome and therapy response of diagnostic biopsies, activating interleukin-7 receptor (IL7R) signaling mutations were identified as an important contributor to steroid resistance.14 In addition, physiological IL7R signaling can also raise cellular steroid resistance.15 Here, we observe that NR3C1 aberrations are already recurrently present in pediatric T-ALL as early as at first diagnosis. Moreover, we confirm that NR3C1 abnormalities relate to steroid resistance and may facilitate disease selection under first-line induction therapy resulting in relapse.
Materials and methods
Primary diagnostic bone marrow or peripheral blood samples were used from a total of 146 primary pediatric T-ALL patients in this study: 72 enrolled on the Dutch Childhood Oncology Group (DCOG) protocols ALL-7/8 (n = 30)16,17 or ALL-9 (n = 42)18 and 74 patients enrolled on the German Co-Operative Study Group for Childhood Acute Lymphoblastic Leukemia study (COALL-97, n = 74)19 with a median follow up of 67 and 52 months, respectively. The patients’ parents or legal guardians provided informed consent to use leftover diagnostic material for research in accordance with the Institutional Review Board of the Erasmus MC Rotterdam and the Declaration of Helsinki. Leukemia cells were harvested from blood or bone marrow samples and enriched to a purity of at least 90%.16-19 Screening for NR3C1 mutations and loss of heterozygosity (LOH) was performed by Ion Torrent sequencing, array-based comparative genomic hybridization, and/or multiplex ligation-dependent probe amplification analyses as previously described.14 Transcriptional NR3C1 expression of primary patient material was determined using U133plus 2.0 arrays. Prednisolone LC50 was determined by performing a 4-day MTT-assay, with the prednisolone concentration ranging from 0.008 to 250 μg/mL.
Constructs and transduction
For REH cells, gateway multisite recombination (Invitrogen) was used for gateway cloning of lentiviral expression vectors by using a gateway-adapted lentiviral pLEGO-iC2 destination vector (Addgene) as previously described.14 The entry vectors existed of (1) attL1/attR5-flanked doxycycline-inducible promoter (third generation; Clontech); (2) attL5/attL4-NR3C1 (wild-type or mutant) complementary DNA sequence; (3) attR4/attR3-flanked DDK-tag followed by a stop codon, Woodchuck hepatitis virus Posttranscriptional Regulatory Element sequence, and a constitutive spleen focus forming virus promotor promoter; and (4) Tetracycline (doxycycline)-induced transcriptional activator protein-Thosea asigna virus 2A-truncated Nerve Growth Factor Receptor reporter. For the short hairpin RNA (shRNA) experiments in REH and SUP-T1 cells, lentiviral transduction was performed with PLKO.1-puro lentiviral shRNA constructs directed against the human NR3C1 gene that were selected from the MISSION T shRNA Library (Sigma-Aldrich).
Protein extraction and subsequent blotting procedure on REH or SUP-T1 cells were performed as previously described.14 Primary antibodies used for western blotting: DKDDDDDK Tag (Cell Signaling Technologies, number 2368) and β-actin (number ab6276).
Real-time quantitative polymerase chain reaction
RNA was isolated with TRIzol reagent (Thermo Fisher Scientific). Real-time quantitative polymerase chain reaction were performed by using the DyNAmo HS SYBR Green qPCR Kit (Thermo Fisher Scientific) under 1X conditions, supplemented with 4 mM magnesium chloride using the CFX384 Touch-Time polymerase chain reaction Detection System (Bio-Rad). Expression levels were calculated relative to the expression of the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) household gene. Primers used are GAPDH Fw primer 5′-GTCGGAGTCAACGGATT-3′, GAPDH Rev primer 5′-AAGCTTCCCGTTCTCAG-3′, GILZ Fw primer 5′-TGGCCATAGACAACAAGAT-3′, GILZ Rev primer 5′-TTGCCAGGGTCTTCAA-3′, FKBP5 Fw primer 5′-GAATGGTGAGGAAACGC-3′, FKPB5 Rev primer 5′-ATGCCTCCATCTTCAAATAA-3′, SGK1 Fw primer 5′-GGAGCCTGAGCTTATGAAT-3′, and SGK1 Rev primer 5′-TTCCGCTCCGACATAATA-3′.
Associations between NR3C1 mutational/deletion status and transcriptional NR3C1 expression or in vitro prednisolone LC50 were calculated by using the Mann-Whitney U test. Survival analysis for relapse-free survival in our patient cohort was calculated using the log-rank (Mantel-Cox) test. For all tests, P < 0.05 was used as significance level.
Recurrent NR3C1 aberrations in T-ALL patients at diagnosis
In our initial patient cohort for which we combined results from sequencing and LOH analyses,14 we identified 6 out of 69 (9%) diagnostic patient biopsies that harbor NR3C1 deletions (n = 4) or carried heterozygous missense mutations (n = 2; G323V and K777N). To substantiate these findings, we screened an additional cohort of 77 diagnostic pediatric T-ALL patient samples (totaling 146 patients) for NR3C1 mutations and/or LOH using Ion Torrent sequencing, array-based comparative genomic hybridization, and/or multiplex ligation-dependent probe amplification analyses. We identified 2 additional patients with NR3C1 deletions and 3 patients who harbored NR3C1 mutations. A heterozygous single nucleotide insertion resulting in a premature frameshift in the protein coding domain (E116fs) was identified in 1 patient, whereas a second patient had a heterozygous G568W missense mutation. The third patient had acquired 3 mutations including 2 missense mutations (N130D and R386L) and a nonsense mutation (G371X) that truncates the NR3C1 protein before the DNA binding and ligand-binding domains. Thus, we identified 6 patients having NR3C1 deletions (4%) and 5 other patients with newly identified missense or nonsense NR3C1 mutations (3%) in a total of 146 diagnostic T-ALL patient samples (Figure 1 and Supplemental Table 1, http://links.lww.com/HS/A123).
Four out of 6 NR3C1-deleted patients were classified as early thymic progenitor acute lymphoblastic leukemia (ETP-ALL) based on unsupervised cluster analysis of gene expression data.20 These NR3C1 deletions occurred in the context of larger 5q deletion (Figure 2A), which is consistent with the observation that ETP-ALL patients frequently harbor interstitial or terminal 5q deletions.21 One patient (number 1944) had a TLX1 translocation and co-clustered with the TLX cluster, and in contrast to ETP-ALL patients had a relative small deletion on 5q that includes the NR3C1 locus. For 1 patient, its cluster had not been determined.
In relation to outcome, none of the NR3C1 deleted patients relapsed, while 3 out of 5 patients with a NR3C1 mutation relapsed. In addition to small patient numbers, no significant difference in survival was observed compared to patients that lack NR3C1 aberrations (Figure 2B). As a consequence of NR3C1 deletions, NR3C1 expression levels were significantly lower for NR3C1 deleted patients compared to wild-type patients (Figure 2C, P = 0.0017). Interestingly, no correlation was found between the relative NR3C1 expression and in vitro steroid response as measured for 83 treatment-naïve patient samples (Figure 2D). This indicates that relative basal NR3C1 messenger RNA (mRNA) expression levels in patient biopsies are not predictive for steroid responsiveness of primary T-ALL patients at diagnosis. This is in line with previous observations that steroid responsiveness is largely determined by the ability to upregulate steroid response genes including NR3C1 itself and pro-apoptotic BIM following steroid exposure.22,23 Therefore, steroid responsiveness seems independent of a certain NR3C1 expression threshold. Upon measuring actual steroid-induced cytotoxicity, we found that patients with NR3C1 aberrations overall had a significantly inferior in vitro steroid response compared to NR3C1 wild-type patients (Figure 2E, P = 0.0078). Four NR3C1-aberrant patients were completely resistant in vitro, and 4 had an intermediate in vitro response to prednisolone.
NR3C1 levels determine steroid response levels
To further investigate the relationship between NR3C1 expression and steroid responsiveness, we used the REH cell line that lacks expression of a functional glucocorticoid receptor.24 For this, we stably transduced REH cells with a doxycycline-inducible NR3C1 expression construct (denoted as REHNR3C1). Restoration of wild-type NR3C1 expression following doxycycline induction effectuated a highly sensitive prednisolone response in REHNR3C1 cells (Figure 3A). To explore whether reductions in NR3C1 levels would proportionally diminish the steroid response, we knocked-down NR3C1 expression in REHNR3C1 cells by introducing 4 isopropyl β- d-1-thiogalactopyranoside (IPTG)-inducible NR3C1-directed shRNA lentiviral constructs. Partial knockdown of NR3C1 upon IPTG treatment in doxycycline-induced REHNR3C1 cells reduced the sensitivity to prednisolone treatment. For these derivative REHNR3C1 lines, knockdown by short hairpin constructs led to pronounced NR3C1 knockdown and steroid insensitivity (Figure 3, B and C). As a control, no reduction in steroid cytotoxicity was observed for REHNR3C1 cells that were transduced with shRNA construct (sh5) directed against the 3′UTR of the normal NR3C1 gene, which was absent in the lentiviral NR3C1 expression construct.
To further visualize the relationship between NR3C1 protein level and steroid sensitivity in a T-ALL context, we transduced steroid-sensitive SUP-T1 T-ALL cells with NR3C1-directed shRNA lentiviral constructs sh2 or sh3. Timed activation of these shRNA constructs by addition of IPTG (eg, 4 or 1 d prior or at the start of prednisolone treatment) demonstrated that a gradual increase in NR3C1 knockdown augmented steroid resistance (Figure 3, D and E). These results highlight an interdependency between NR3C1 expression and steroid response within a specific cellular context.
NR3C1 mutations and steroid response
We then used the same REH model to study the functional consequences of the identified NR3C1 missense (K777N, G323V, or N130D) or G371X nonsense mutation on steroid responsiveness. REH cells were stably transduced with either of these mutant NR3C1 expression constructs, resulting in expression of mutant NR3C1 receptors following doxycycline induction (Figure 4A). Using these bulk transduced lines, expression of K777N, G323V, or N130D mutant molecules elicited an efficient transcriptional steroid response, as demonstrated by the upregulation of specific NR3C1 target genes such as GILZ, FKBP5, and SGK1 after steroid treatment (Figure 4B). Moreover, the functionally of these mutations was reflected by their steroid responsiveness since these mutations were equally sensitive steroid response compared to REHNR3C1 cells (Figure 4C). As expected, expression of the truncating G371X mutant was totally ineffective to restore a steroid response (Figure 4, B and C).
In this study, we found that relative NR3C1 mRNA expression levels in primary diagnostic patient samples do not correlate with steroid response. Interindividual differences in other factors that contribute to steroid resistance may influence the steroid response at diagnosis. For example, IL7R signaling mutations, epigenetic silencing of the BIM locus, or overexpression of BCL2 have all been associated with steroid resistance.8,14,25-27 Thus, the relative NR3C1 mRNA level is not useful as a predictive biomarker for clinical steroid responsiveness for patients at diagnosis per se. However, for experiments performed within a specific and controlled cellular context such as REH or SUP-T1 cells, we observed a strong interdependency between NR3C1 levels and steroid response levels, in line with previous findings by others.4,28
Loss of NR3C1 expression due to mutations, deletions, or other mechanisms that occur under the pressure of steroid treatment4 explain the relative increased incidence of NR3C1 deletions and mutations at disease relapse.10,11,13 However, our study demonstrates that heterozygous NR3C1 deletions or truncating mutations are already present in diagnostic biopsies of approximately 5% of pediatric T-ALL patients and associate with steroid resistance. This may suggest that natural steroid hormones that normally shape the immune system29,30 already elicit a strong selection pressure on (pre)leukemic cells before actual diagnosis of disease.
We identified NR3C1 missense mutations in 4 T-ALL patients (including 1 patient that also harbored a truncating mutation). The missense mutations tested in REH cells demonstrated an efficient transcriptional steroid response and seemed equally effective as the wild-type NR3C1 molecule in effectuating steroid-induced apoptosis. This is in contrast to a recent functional screening using a similar experimental approach in REH cells, demonstrating that (distinct) relapse-enriched NR3C1 mutations in ALL did confer steroid resistance.31 This indicates that NR3C1 missense mutations found at relapse are selected during therapy and cause resistance towards synthetic steroids, while mutations at diagnosis do not necessarily drive steroid resistance per se. The N130D mutation co-occurred with the truncating G371X mutation in one of our patients and may therefore represent a passenger mutation rather than reflecting a steroid resistance driving mutation.
The steroid-sensitive phenotype of our mutant NR3C1 overexpressing cells somewhat contradict the intermediate or steroid-resistant phenotype of corresponding primary patient blasts. Since induced expression of mutant NR3C1 molecules exceeds the physiological level as normally expressed in ALL blasts, we cannot exclude the possibility that subtle effects of the studied NR3C1 mutations on steroid responsiveness are masked. This may be exemplified by the T-ALL patient harboring the K777N mutation since this mutation was also preserved at relapse. In contrast to the controlled context of REH cells, many other factors are influential in the steroid sensitivity of leukemic blasts. The presence of mutant NR3C1 may therefore synergize with other (epi)genetic aberrations at disease diagnosis. Regardless of a supporting or causal role, the presence of NR3C1 mutations related to steroid resistance in our cohort, and 3 out of 4 patients with missense NR3C1 mutations relapsed during therapy.
In conclusion, approximately 7% of patients that are diagnosed with T-ALL already harbor NR3C1 inactivating events that influence their leukemic response to steroid treatment.
Sources of funding
This study was supported by grants of the Foundation “Kinderen Kankervrij” (KiKa): KiKa-82 (WS) and KiKa-219 (JCGvdZ).
1. Riehm H, Reiter A, Schrappe M, et al. [Corticosteroid-dependent reduction of leukocyte count in blood as a prognostic factor in acute lymphoblastic leukemia in childhood (therapy study ALL-BFM 83)]. Klin Padiatr. 1987; 199:151–160.
2. Griffin TC, Shuster JJ, Buchanan GR, et al. Slow disappearance of peripheral blood blasts is an adverse prognostic factor in childhood T cell acute lymphoblastic leukemia: a Pediatric Oncology Group study. Leukemia. 2000; 14:792–795.
3. Weikum ER, Knuesel MT, Ortlund EA, et al. Glucocorticoid receptor control of transcription: precision and plasticity via allostery. Nat Rev Mol Cell Biol. 2017; 18:159–174.
4. Wandler AM, Huang BJ, Craig JW, et al. Loss of glucocorticoid receptor expression mediates in vivo dexamethasone resistance in T-cell acute lymphoblastic leukemia. Leukemia. 2020; 34:2025–2037.
5. Miller AL, Garza AS, Johnson BH, et al. Pathway interactions between MAPKs, mTOR, PKA, and the glucocorticoid receptor in lymphoid cells. Cancer Cell Int. 2007; 7:3.
6. Piovan E, Yu J, Tosello V, et al. Direct reversal of glucocorticoid resistance by AKT inhibition in acute lymphoblastic leukemia. Cancer Cell. 2013; 24:766–776.
7. Gruber G, Carlet M, Türtscher E, et al. Levels of glucocorticoid receptor and its ligand determine sensitivity and kinetics of glucocorticoid-induced leukemia apoptosis. Leukemia. 2009; 23:820–823.
8. Jing D, Huang Y, Liu X, et al. Lymphocyte-specific chromatin accessibility pre-determines glucocorticoid resistance in acute lymphoblastic leukemia. Cancer Cell. 2018; 34:906–921.e8.
9. Schmidt S, Irving JA, Minto L, et al. Glucocorticoid resistance in two key models of acute lymphoblastic leukemia occurs at the level of the glucocorticoid receptor. FASEB J. 2006; 20:2600–2602.
10. Mullighan CG, Phillips LA, Su X, et al. Genomic analysis of the clonal origins of relapsed acute lymphoblastic leukemia. Science. 2008; 322:1377–1380.
11. Irving JA, Bloodworth L, Bown NP, et al. Loss of heterozygosity in childhood acute lymphoblastic leukemia detected by genome-wide microarray single nucleotide polymorphism analysis. Cancer Res. 2005; 65:3053–3058.
12. Irving JA, Enshaei A, Parker CA, et al. Integration of genetic and clinical risk factors improves prognostication in relapsed childhood B-cell precursor acute lymphoblastic leukemia. Blood. 2016; 128:911–922.
13. Kuster L, Grausenburger R, Fuka G, et al. ETV6/RUNX1-positive relapses evolve from an ancestral clone and frequently acquire deletions of genes implicated in glucocorticoid signaling. Blood. 2011; 117:2658–2667.
14. Li Y, Buijs-Gladdines JG, Canté-Barrett K, et al. IL-7 receptor mutations and steroid resistance in pediatric T cell acute lymphoblastic leukemia: a genome sequencing study. PLoS Med. 2016; 13:e1002200.
15. Delgado-Martin C, Meyer LK, Huang BJ, et al. JAK/STAT pathway inhibition overcomes IL7-induced glucocorticoid resistance in a subset of human T-cell acute lymphoblastic leukemias. Leukemia. 2017; 31:2568–2576.
16. Kamps WA, Bökkerink JP, Hählen K, et al. Intensive treatment of children with acute lymphoblastic leukemia according to ALL-BFM-86 without cranial radiotherapy: results of Dutch Childhood Leukemia Study Group Protocol ALL-7 (1988-1991). Blood. 1999; 94:1226–1236.
17. Kamps WA, Bökkerink JP, Hakvoort-Cammel FG, et al. BFM-oriented treatment for children with acute lymphoblastic leukemia without cranial irradiation and treatment reduction for standard risk patients: results of DCLSG protocol ALL-8 (1991-1996). Leukemia. 2002; 16:1099–1111.
18. Veerman AJ, Kamps WA, van den Berg H, et al. Dexamethasone-based therapy for childhood acute lymphoblastic leukaemia: results of the prospective Dutch Childhood Oncology Group (DCOG) protocol ALL-9 (1997-2004). Lancet Oncol. 2009; 10:957–966.
19. Escherich G, Tröger A, Göbel U, et al. The long-term impact of in vitro drug sensitivity on risk stratification and treatment outcome in acute lymphoblastic leukemia of childhood (CoALL 06-97). Haematologica. 2011; 96:854–862.
20. Homminga I, Pieters R, Langerak AW, et al. Integrated transcript and genome analyses reveal NKX2-1 and MEF2C as potential oncogenes in T cell acute lymphoblastic leukemia. Cancer Cell. 2011; 19:484–497.
21. La Starza R, Barba G, Demeyer S, et al. Deletions of the long arm of chromosome 5 define subgroups of T-cell acute lymphoblastic leukemia. Haematologica. 2016; 101:951–958.
22. Bouillet P, Metcalf D, Huang DC, et al. Proapoptotic Bcl-2 relative Bim required for certain apoptotic responses, leukocyte homeostasis, and to preclude autoimmunity. Science. 1999; 286:1735–1738.
23. Ramdas J, Liu W, Harmon JM. Glucocorticoid-induced cell death requires autoinduction of glucocorticoid receptor expression in human leukemic T cells. Cancer Res. 1999; 59:1378–1385.
24. Bachmann PS, Gorman R, Papa RA, et al. Divergent mechanisms of glucocorticoid resistance in experimental models of pediatric acute lymphoblastic leukemia. Cancer Res. 2007; 67:4482–4490.
25. Bachmann PS, Piazza RG, Janes ME, et al. Epigenetic silencing of BIM in glucocorticoid poor-responsive pediatric acute lymphoblastic leukemia, and its reversal by histone deacetylase inhibition. Blood. 2010; 116:3013–3022.
26. Jing D, Bhadri VA, Beck D, et al. Opposing regulation of BIM and BCL2 controls glucocorticoid-induced apoptosis of pediatric acute lymphoblastic leukemia cells. Blood. 2015; 125:273–283.
27. Autry RJ, Paugh SW, Carter R, et al. Integrative genomic analyses reveal mechanisms of glucocorticoid resistance in acute lymphoblastic leukemia. Nat Cancer. 2020; 1:329–344.
28. Grausenburger R, Bastelberger S, Eckert C, et al. Genetic alterations in glucocorticoid signaling pathway components are associated with adverse prognosis in children with relapsed ETV6/RUNX1-positive acute lymphoblastic leukemia. Leuk Lymphoma. 2016; 57:1163–1173.
29. Sacedón R, Vicente A, Varas A, et al. Early maturation of T-cell progenitors in the absence of glucocorticoids. Blood. 1999; 94:2819–2826.
30. Pazirandeh A, Xue Y, Prestegaard T, et al. Effects of altered glucocorticoid sensitivity in the T cell lineage on thymocyte and T cell homeostasis. FASEB J. 2002; 16:727–729.
31. Li B, Brady SW, Ma X, et al. Therapy-induced mutations drive the genomic landscape of relapsed acute lymphoblastic leukemia. Blood. 2020; 135:41–55.