ARTICLE pubs.acs.org/bc
A Bioluminescent Probe for Salivary Cortisol Sung Bae Kim,*,† Yasuhiro Takenaka,‡ and Masaki Torimura† †
Research Institute for Environmental Management Technology, National Institute of Advanced Industrial Science and Technology (AIST), 16-1 Onogawa, Tsukuba 305-8569, Japan ‡ Department of Endocrinology and Diabetes, Saitama Medical University, 38 Morohongo Moroyama-machi, Iruma-gun, Saitama 350-0495, Japan
bS Supporting Information ABSTRACT: Cortisol is a classical biomarker for the stress levels of human beings. We fabricated highly sensitive bioluminescent probes for salivary cortisol. The following strategies were contrived in the molecular design. Gaussia princeps luciferase (GLuc) was dissected into two fragments, between which an N-terminal-extended ligand binding domain of glucocorticoid receptor (GR HLBD), named Simgr4, was inserted. First, this unique single-chain probe was then situated downstream of a glucocorticoid response element (GRE) promoter in a reportergene system for constructing two ONOFF switches for cortisol. Second, a circularly permutated (CP) variant of Simgr4 was formulated. The reporter-gene system exerted an improved signal-to-background (S/B) ratio of 8.5 to cortisol. Furthermore, a circularly permutated (CP) variant of Simgr4 exerted a 10 enhanced detection limit to cortisol and a long dynamic range from 109 to 106 M cortisol, covering all of the normal clinical ranges of serum, urine, and saliva. This optimized probe successfully determined daily fluctuations of salivary cortisol and the correlations with those by ELISA. This study is the first to investigate the contribution of the HLBD of a nuclear receptor and multiple ONOFF switches for molecular probes and salivary cortisols.
’ INTRODUCTION Glucocorticoids, called the “stress” hormones, ubiquitously affect almost every cellular, molecular, and physiologic network of the living subjects.1 The levels of glucocorticoids in physiological samples are elevated by various stimulators, including psychological stress and even by 2-deoxy-D-glucose (2DG),2 and direct the signs and symptoms of diseases, such as Cushing’s Syndrome and Addison’s Disease.3 An acute elevation of glucocorticoids impairs the retrieval of long-term memory.4 Among glucocorticoids, cortisol has been classically targeted as a “biomarker” of the stress levels of human beings, and is related directly or indirectly with inflammation, autoimmune disease, and cancers.5 The normal clinical levels of cortisol range from 0.8 to 6.4 107 M in serum, from 0.6 to 7.5 107 M in urine, and from 0.2 to 2.8 108 M in saliva.3 This clinical level of salivary cortisol is approximately 10 times less than those of urine and serum, and thus requires improved sensitivity for salivary cortisol. The low levels of salivary cortisol were conventionally estimated with an enzyme-linked immunosorbent assay (ELISA) and radio immunoassay (RIA).6 However, these approaches carry intrinsic demerits of time consumption, labor intensity, cost, a long protocol starting from sample dilution, and a narrow linear range. We recently introduced a smart single-chain probe illuminating intracellular molecular events.7,8 The unique molecular design for an intracellular complementation of split-luciferase was proven to be effective for determining steroids, second r 2011 American Chemical Society
messengers, and protein phosphorylation in living mammalian cells.810 We also represented a single-chain bioluminescent probe for stress hormones using split-Gaussia princeps luciferase (GLuc), where appropriate consensus motifs of coactivators and corepressors were intensively examined in a circularly permutated (CP) probe backbone.11 However, these approaches are still insufficient even for a direct determination of the low levels of salivary cortisol, and are poorly demonstrated in real samples. To address the conventional limitations, we first formulated a new single-chain probe. Upon design of the single-chain probe, the N-terminal end of the ligand binding domain of glucocorticoid receptor (GR HLBD) was extended from 527 to 777 to 486777 AA. The GR HLBD was sandwiched between the two fragments of GLuc, named “Simgr4”. Through this molecular design, we intended (i) to approximate the N- and C-terminal ends within 8 Å according to the previous consideration on protein fragment reconstitution12 and (ii) to exert a weak transcriptional effect on the host probe as observed previously in a GR truncation study.13,14 On the basis of the molecular design, we first fabricated a new bioluminescent probe system, where the system carries two ON OFF switches exerting multiple recognitions of salivary cortisol, i.e., the glucocorticoid-response element (GRE) promoter for determining Received: May 4, 2011 Revised: July 13, 2011 Published: August 13, 2011 1835
dx.doi.org/10.1021/bc200220k | Bioconjugate Chem. 2011, 22, 1835–1841
Bioconjugate Chemistry
ARTICLE
Figure 1. (A) X-ray crystallographic structure of GR LBD and GR DBD in the complex with cDNA comprising a glucocorticoid response element (GRE). The mark of scissors indicates the N-terminal end of GR HLBD (486777 AA). (B) Schematic structures of cDNA constructs encoding Simgr3 and Simgr4. Abbreviations: GLuc-N, the N-terminal fragment of GLuc; GLuc-C, the C-terminal fragment of GLuc; GR LBD, the ligand-binding domain of glucocorticoid receptor; sDBD, the shortened DNA binding domain; Hinge, the hinge region of GR; LXXLL motif, a LXXLL motif originated from GRIP1 ID3. (C) The working mechanism of the probe with multiple recognizing ligands. An exogenous ligand first activates expression of a bioluminescent probe at the downstream of the GRE promoter (first ONOFF switch). The ligand is recognized again by the expressed probe (second ONOFF switch).
transcriptional activity (first switch) and a glucocorticoid-sensitive single-chain probe downstream of the GRE promoter (second switch; Figure 1). A CP modification of Simgr4 was second made for improving the signal-to-background (S/B) ratio, named “cSimgr4”. This molecular design was surprisingly effective to determine very low levels (
