Summary:

In this issue of Cancer Discovery, Rasool and colleagues show that TF11H/CDK7 phosphorylates the MED1 component of the Mediator complex, which enhances its interaction with androgen receptor (AR), and that this phosphorylation is increased in prostate cancer that is resistant to castration and enzalutamide. A covalent CDK7-specific inhibitor (THZ1) impairs AR-mediated MED1 recruitment to chromatin, and can suppress enzalutamide resistance in vitro and induce tumor regression in a castration-resistant prostate cancer xenograft model, suggesting a novel therapeutic approach for advanced prostate cancer.

See related article by Rasool et al., p. 1538.

Androgen receptor (AR) is critical for prostate cancer, and androgen deprivation therapy (ADT; ablation of testicular androgens by medical or surgical castration) remains the standard therapy for metastatic prostate cancer. Tumors that recur after ADT [i.e., castration-resistant prostate cancer (CRPC)] may respond to agents that further reduce androgen levels by suppressing adrenal gland and intratumoral androgen synthesis (CYP17A1 inhibitor abiraterone) or to treatment with direct AR antagonists such as enzalutamide. Moreover, the development of metastatic CRPC can be delayed by early use of these latter agents [collectively referred to as androgen signaling inhibitors (ASI)] in combination with suppression of testicular androgens, and these combined intensive ADTs are now becoming the standard of care. Unfortunately, although the majority of patients initially respond to these therapies, most relapse within a few years. Further responses may be obtained with agents including PARP inhibitors in a subset of tumors with defects in DNA-damage repair, with immunotherapy in the small subset with mismatch-repair defects, or with taxanes, but these responses are not generally durable (1).

A subset of prostate cancer that relapses after ASIs appears to be AR-independent (∼30%), as these cancers express low or undetectable levels of AR or classic AR-regulated genes, with about half of these showing evidence of neuroendocrine differentiation (2). However, AR expression and activity persist in the majority of cases, even with continued ASI treatment. Increased AR expression due to amplification of the AR gene and an upstream AR enhancer, as well as activating AR mutations, contribute to this persistent AR activity. Moreover, progression to CRPC and ASI resistance is associated with increased expression of AR splice variants that have constitutive activity due to loss of the ligand-binding domain, although the extent to which these are driving AR activity remains to be established. Although alterations in the AR itself account for some of the observed resistance in CRPC, it is clear that additional adaptations in these tumors must be contributing to persistent AR activity.

Mediator is an evolutionarily conserved multiprotein complex that plays a critical role in bridging enhancer-bound transcription factors to RNA polymerase II and the preinitiation complex (PIC) at corresponding promoters. In a simplified model, a transcription factor binds to an enhancer and recruits the Mediator complex (Fig. 1). The Mediator complex then interacts with components of the PIC including RNA polymerase II, TFIIB, TFIID, and TFIIH, and this PIC interaction is linked to loss of a CDK8 kinase module from the Mediator complex. The strongest interaction is between Mediator and the C-terminal domain (CTD) of RNA polymerase II, which contains 52 repeats of the consensus sequence Y1-S2-P3-T4-S5-P6-S7. CDK7, a component of TFIIH, then phosphorylates S5 and S7 in the CTD, which disrupts the CTD interaction with Mediator and is presumably necessary to release RNA polymerase II from promoter-proximal pausing. A second critical function of CDK7 is to phosphorylate and activate CDK9 in the pTEFb complex (CDK9/cyclin T). pTEFb is recruited to the promoter by the super elongation complex or by BRD4, but is sequestered and inhibited by binding to a 7SK snRNP complex. Upon release and activation, pTEFb phosphorylates S2 in the RNA polymerase II CTD, as well as NELF and DSIF, which together trigger the release of RNA polymerase II from the promoter and allow transcriptional elongation.

Figure 1.

Mediator complex and CDK7 mediated interactions with AR at target gene loci. AR binding to distal enhancers or superenhancers recruits the Mediator complex, which then facilitates chromatin looping to the promoter through interactions with RNA polymerase II (RNA Pol II) and other components of the PIC including TFIIH, which contains CDK7. This PIC interaction is associated with loss of the CDK8/19 kinase module from Mediator. MED1 mediates the Mediator interaction with AR, and CDK7-mediated phosphorylation of MED1 enhances MED1 interaction with AR and with the Mediator complex. In some contexts, this MED1 phosphorylation may also be mediated by additional kinases. MED1 phosphorylation may also be directly or indirectly enhanced in advanced prostate cancer by decreased expression of PP2A. CDK7 is then critical for phosphorylation of S5 and S7 in the RNA Pol II CTD, and for phosphorylation and activation of CDK9 in the pTEFb complex, which then further phosphorylates the RNA Pol II CTD at S2, resulting in promoter clearance and transcriptional elongation. Both CDK7 and CDK9 may also directly phosphorylate AR to modulate its activity, and have additional substrates (not shown) that contribute to transcription.

Figure 1.

Mediator complex and CDK7 mediated interactions with AR at target gene loci. AR binding to distal enhancers or superenhancers recruits the Mediator complex, which then facilitates chromatin looping to the promoter through interactions with RNA polymerase II (RNA Pol II) and other components of the PIC including TFIIH, which contains CDK7. This PIC interaction is associated with loss of the CDK8/19 kinase module from Mediator. MED1 mediates the Mediator interaction with AR, and CDK7-mediated phosphorylation of MED1 enhances MED1 interaction with AR and with the Mediator complex. In some contexts, this MED1 phosphorylation may also be mediated by additional kinases. MED1 phosphorylation may also be directly or indirectly enhanced in advanced prostate cancer by decreased expression of PP2A. CDK7 is then critical for phosphorylation of S5 and S7 in the RNA Pol II CTD, and for phosphorylation and activation of CDK9 in the pTEFb complex, which then further phosphorylates the RNA Pol II CTD at S2, resulting in promoter clearance and transcriptional elongation. Both CDK7 and CDK9 may also directly phosphorylate AR to modulate its activity, and have additional substrates (not shown) that contribute to transcription.

Close modal

The MED1 component of Mediator (also known as TRAP220, PBP, or DRIP205) can bind to multiple nuclear receptors, including AR. This binding was initially found to be mediated by LXXLL motifs in MED1 and the AF2 domain generated by ligand binding in nuclear receptors, but additional studies have demonstrated that the TAU1 site in the AR N-terminal domain can facilitate AR binding as well (3, 4). Previous studies have identified AKT, ERK, and DNA-PK as kinases that mediate the phosphorylation of MED1, and that this phosphorylation enhances MED1 association with the Mediator complex and its binding to AR (4–6). In this issue of Cancer Discovery, a report from Rasool and colleagues finds that MED1 phosphorylation at T1457, which enhances its interaction with AR, is mediated by CDK7 (7). Moreover, the authors show that enzalutamide resistance in CRPC models is associated with increased MED1 phosphorylation, and that a covalent CDK7-specific inhibitor (THZ1) impairs AR-mediated MED1 recruitment to chromatin. Consistent with these results, they also show that THZ1 can suppress enzalutamide resistance in vitro and induce tumor regression in a CRPC xenograft model, suggesting a novel therapeutic approach for advanced prostate cancer.

In agreement with previous data, Rasool and colleagues initially observed that AR and MED1 were corecruited to chromatin in response to androgen stimulation and were particularly enriched at superenhancers. They next found that androgen stimulation increased MED1 phosphorylation at two previously reported sites, T1032 and T1457, and that loss of the T1457 site impaired AR binding. Using a series of CDK inhibitors, the investigators established CDK7 as the relevant kinase targeting the T1457 site, and further showed that the T1457D phosphomimetic associates with AR on chromatin and was resistant to THZ1. It should be noted that T1457 phosphorylation (phospho-MED1) can also enhance MED1 association with the Mediator complex, which may contribute to the increased association with AR observed in this study. In any case, this result is consistent with AR recruitment of MED1/Mediator, with subsequent chromatin looping and interaction with CDK7/TFIIH.

Rasool and colleagues next found that AR-positive prostate cancer cells (LNCaP, VCaP, and LAPC4) had increased phosho-MED1 and were more sensitive to THZ1-induced apoptosis than AR-negative prostate cancer cells (DU145 and PC3), despite no differences in total CDK7 protein levels. THZ1 also markedly altered gene expression in LNCaP and VCaP cells, which exhibited approximately 10-fold more differentially expressed genes, including many AR-regulated genes, when compared with AR-negative DU145 cells. Significantly, a previous study had shown that CDK7/TFIIH can directly phosphorylate S515 in the AR N-terminal domain, and that this can modulate AR transcriptional activity and degradation (8). Rasool and colleagues confirmed that phospho-S515 AR levels were greatly increased by androgen, but did not see a marked decrease in response to THZ1, although their data do indicate some decrease. CDK7 may also regulate AR indirectly through activation of CDK9, which can phosphorylate a distinct site on AR (S81) and may enhance AR activity (9). Overall, the data support a significant role for CDK7-mediated phosphorylation of MED1 in regulating AR interactions with Mediator and transcriptional activity, but additional direct or indirect targets of CDK7 may also contribute to effects on AR activity.

To further interrogate MED1/CDK7 in the setting of resistance to AR antagonists, Rasool and colleagues next generated enzalutamide-resistant sublines of AR-dependent LNCaP, LAPC4, and VCaP cells. Significantly, these enzalutamide-resistant cells had restored AR activity that was associated with increases in both MED1 phosphorylation and protein stability. CDK7 protein level was not increased in the enzalutamide-resistant cells, but there was a striking decrease in the PP2A catalytic subunit, and RNAi-mediated depletion of the PP2A catalytic subunit caused an increase in phospho-MED1. Previous studies have found decreased PP2A activity in CRPC, supporting this as a physiologic mechanism contributing to the restoration of AR activity, but it is not yet clear whether CDK7 is a direct or indirect target of PP2A. Finally, the study by Rasool and colleagues shows that the CDK7 inhibitor THZ1 can suppress AR activity and proliferation of enzalutamide-resistant prostate cancer cells in vitro, and cause marked regression of castration-resistant VCaP xenografts in vivo. In future in vivo studies, it would be of interest to compare the THZ1 sensitivity of AR signaling in castration-sensitive versus castration-resistant and enzalutamide-resistant tumors.

Together these studies show that CDK7 enhances AR activity at least in part through phosphorylation of MED1, that CDK7-mediated increases in MED1 phosphorylation contribute to the restoration of AR activity in advanced prostate cancer, and that CDK7 is a potential therapeutic target in CRPC. However, a major concern with targeting any of the CDKs is toxicity. In developing a CDK7/MED1-targeted therapy, it will be important to determine whether all or a subset of ASI-resistant prostate cancers are hypersensitive to CDK7 inhibition, and whether a hypersensitive subset can be identified through biomarkers (e.g., phospho-MED1 levels). Additional studies characterizing the interaction between AR and MED1, and in particular assessing MED1 interaction with AR splice variants, would also be of benefit. Finally, although these studies have focused on MED1 and CDK7, the broader implication is that enhanced AR interaction with the Mediator complex may contribute to advanced prostate cancer. Interestingly, CDK8 and its paralog CDK19 are components of the Mediator complex that are associated with AR expression and with more aggressive prostate cancer, and may thereby be worthy of further exploration as therapeutic targets. Moreover, recent studies are providing a more dynamic model of Mediator function, and may reveal additional therapeutic approaches (10).

No potential conflicts of interest were disclosed.

This work was supported by NIH grants P01 CA163227 and P50 CA090381, Department of Defense Prostate Cancer Research Program W81XWH-16-1-0431 and Early Investigator Research Award PC170570, the A. David Mazzone Research Awards Program, and Prostate Cancer Foundation Young Investigator and Challenge Awards.

1.
Einstein
DJ
,
Arai
S
,
Balk
SP
. 
Targeting the androgen receptor and overcoming resistance in prostate cancer
.
Curr Opin Oncol
2019
;
31
:
175
82
.
2.
Bluemn
EG
,
Coleman
IM
,
Lucas
JM
,
Coleman
RT
,
Hernandez-Lopez
S
,
Tharakan
R
, et al
Androgen receptor pathway-independent prostate cancer is sustained through FGF signaling
.
Cancer Cell
2017
;
32
:
474
89
.
3.
Ren
Y
,
Behre
E
,
Ren
Z
,
Zhang
J
,
Wang
Q
,
Fondell
JD
. 
Specific structural motifs determine TRAP220 interactions with nuclear hormone receptors
.
Mol Cell Biol
2000
;
20
:
5433
46
.
4.
Jin
F
,
Claessens
F
,
Fondell
JD
. 
Regulation of androgen receptor-dependent transcription by coactivator MED1 is mediated through a newly discovered noncanonical binding motif
.
J Biol Chem
2012
;
287
:
858
70
.
5.
Belakavadi
M
,
Pandey
PK
,
Vijayvargia
R
,
Fondell
JD
. 
MED1 phosphorylation promotes its association with mediator: implications for nuclear receptor signaling
.
Mol Cell Biol
2008
;
28
:
3932
42
.
6.
Foulds
CE
,
Feng
Q
,
Ding
C
,
Bailey
S
,
Hunsaker
TL
,
Malovannaya
A
, et al
Proteomic analysis of coregulators bound to ERalpha on DNA and nucleosomes reveals coregulator dynamics
.
Mol Cell
2013
;
51
:
185
99
.
7.
Rasool
RU
,
Natesan
R
,
Deng
Q
,
Aras
S
,
Lal
P
,
Sander Effron
S
, et al
CDK7 inhibition suppresses castration-resistant prostate cancer through MED1 inactivation
.
Cancer Discov
2019
;
9
:
1538
55
.
8.
Chymkowitch
P
,
Le May
N
,
Charneau
P
,
Compe
E
,
Egly
JM
. 
The phosphorylation of the androgen receptor by TFIIH directs the ubiquitin/proteasome process
.
EMBO J
2011
;
30
:
468
79
.
9.
Gordon
V
,
Bhadel
S
,
Wunderlich
W
,
Zhang
J
,
Ficarro
SB
,
Mollah
SA
, et al
CDK9 regulates AR promoter selectivity and cell growth through serine 81 phosphorylation
.
Mol Endocrinol
2010
;
24
:
2267
80
.
10.
El Khattabi
L
,
Zhao
H
,
Kalchschmidt
J
,
Young
N
,
Jung
S
,
Van Blerkom
P
, et al
A pliable mediator acts as a functional rather than an architectural bridge between promoters and enhancers
.
Cell
2019
;
178
:
1145
58
.