Research in Dr. Nichols' lab involves study of steroid hormone receptors, primarily the estrogen receptors alpha and beta (ERa, ERß), and their role in normal, as well as in cancer tissue. Dr. Nichols has developed in vitro model systems that allow (a) the analysis of the effect of estrogen receptor (ER) mutations (that we found in breast cancers) on antihormone resistance and (b) screening of novel compounds for ER-subtype selective ligands. Better understanding of ligand activation of ERs may lead to improved endocrine therapies for treating and perhaps preventing breast and other estrogen responsive cancers.
Tamoxifen is one of the most effective drugs for treatment and prevention of breast cancer yet a substantial number of breast cancers (30%) fail to respond to tamoxifen or will become resistant, even if they have estrogen receptor (ER+). Dr. Nichols tests the hypothesis that the estrogen receptor, its co-regulator proteins, or their interaction is altered in breast lesions where tamoxifen is ineffective. Tamoxifen therapy (or other selective estrogen receptor modulators- SERMs) requires a functional ER, yet the clinical use of immunohistochemistry is unable to determine function. Coactivators amplify transcription via ER and are the target of antiestrogen inhibition. Dr. Nichols has found changes in ER affecting coactivator binding sites and tamoxifen-induced structure from tumor tissue. One recovered mutation makes tamoxifen a better agonist than estradiol.
Figure 1: Helix 12, key for transcriptional activity, has an active hormone-bound position, or an inactive antihormone bound position that blocks subsequent coactivator binding, depending on the nature of the bound ligand. (figure from Tanenbaum et al, PNAS USA 95, 5998-6003,1998).
We have examined human breast cancer samples to find altered ER alpha alleles that may change in vivo responses to ligands. We have isolated a number of ER alpha mutations, including several at the Helix 12 region of the hormone inducible AF-2. Mutations of ER alpha (M543V) have been found which cause the SERM, 4-hydroxytamoxifen (ZOHT) to be more active than estradiol (E2) in transcription from an estrogen-response-element ERE. This is opposite the character of the wild-type ER alpha and is consistent with the clinical observation that some tumors seem to grow in the presence of tamoxifen, and will regress if treatment is halted. Though we do not find this mutation in a high proportion of tumors examined, it is direct evidence for the possibility that tamoxifen resistance could arise by such a mechanism. Tamoxifen inhibits growth signaling through ER alpha by binding directly to its hormone-binding pocket, blocking coactivator interaction, so mutations in ER alpha to escape this block should logically be one mechanism of resistance, as demonstrated for AR in prostate cancer. Breast epithelial cells with an ER alpha variant that responds positively to tamoxifen may spawn a population of cells that escape from growth repression by tamoxifen. In turn, their production of paracrine growth factors (TGF alpha, IGF) could lead to outgrowth of cells that form an eventual tumor.
Unexpectedly, the M543V mutation showed greater transcriptional activation by ZOHT than E2, while the M543I mutation maintained the normal E2 > ZOHT response. While these two may seem to be subtle amino acid substitutions at the same position, our results demonstrate the greater degree of hydrophobicity/ length of the amino acid side chains (isoleucine > valine) on the interior of Helix 12 as a critical component for transcriptional activation by E2 and agonists, as opposed to SERM inhibition. We tested the double mutation M543A/L544A in similar transcription experiments and found that consistent with this premise, the reduction in hydrophobicity/ side chain length with alanine also correlated with SERM activation > E2. We propose that the degree of hydrophobicity of the underside of Helix 12 is critical for holding its ligand-activated position over an E2 bound pocket, much like a “Velcro” patch. With a reduction in the length of the aliphatic side chains such as occurs with M543V, the underside of Helix 12 will not bind as tightly over the ligand occupied pocket and allows the “antiestrogen” or alternative position of Helix 12, perhaps even dissociated from the surface of the hormone binding domain. However co-expressing the p160 coactivator AIB1 with the M543V ER alpha recovers relatively strong transcriptional activity with E2, consistent with increased p160 binding to again stabilize the “agonist” position of Helix 12. This model predicts a general feature of all nuclear hormone receptors, namely that mutations to reduce the hydrophobic side-chains of specific conserved amino acids on the undersides of H12 (amino acid alignment equivalents of ER alpha 539/540 and 543/544) of any of the nuclear hormone receptor family will “invert” the transcriptional activity by agonists and antagonists of that receptor. (Nichols et al., 2010).
Dr. Nichols has a collaboration with Endece Pharmaceuticals for mechanistic analysis of several anticancer drugs. He is testing a parent compound and several of its metabolites in experiments to assess their growth-inhibitory and estrogenic or anti-estrogenic character. They have high likelihood to interact with steroid receptors (most likely estrogen receptors) and with enzymes that metabolize steroids, e.g. aromatase.
Dr. Nichols also works with small molecule compounds, 1,1-dichloro-2,2,3-triarylcyclopropanes (DTACs), exhibiting antiestrogenic properties in several assays. He has demonstrated that a subset of these compounds have selectivity for ER alpha or ER beta. In collaboration with Dr. Bino John, we have begun work on an RNA fingerprint of breast cancer that involves computational and experimental work to look at microRNAs and breast cancer. He also develops methods of synthesis for small interfering RNA libraries to discover gene products related to alterable phenotypes.
Nichols M. New directions for drug-resistant breast cancer: the CDK4/6 inhibitors. Future Med Chem 7(12):1473-1481, 2015 (doi:10.4155/fmc.15.86).