Investigating proteins 'at work' in a living environment at atomic resolution is a major goal of molecular biology, which has not been achieved even though methods for the three-dimensional (3D) structure determination of purified proteins in single crystals or in solution are widely used. Recent developments in NMR hardware and methodology have enabled the measurement of high-resolution heteronuclear multi-dimensional NMR spectra of macromolecules in living cells (in-cell NMR). Various intracellular events such as conformational changes, dynamics and binding events have been investigated by this method. However, the low sensitivity and the short lifetime of the samples have so far prevented the acquisition of sufficient structural information to determine protein structures by in-cell NMR. Here we show the first, to our knowledge, 3D protein structure calculated exclusively on the basis of information obtained in living cells. The structure of the putative heavy-metal binding protein TTHA1718 from Thermus thermophilus HB8 overexpressed in Escherichia coli cells was solved by in-cell NMR. Rapid measurement of the 3D NMR spectra by nonlinear sampling of the indirectly acquired dimensions was used to overcome problems caused by the instability and low sensitivity of living E. coli samples. Almost all of the expected backbone NMR resonances and most of the side-chain NMR resonances were observed and assigned, enabling high quality (0.96 ångström backbone root mean squared deviation) structures to be calculated that are very similar to the in vitro structure of TTHA1718 determined independently. The in-cell NMR approach can thus provide accurate high-resolution structures of proteins in living environments.

The cell is a crowded environment in which proteins interact specifically with other proteins, nucleic acids, cofactors and ligands. Atomic resolution structural explanation of proteins functioning in this environment is a main goal of biochemical research. Recent improvements to nuclear magnetic resonance (NMR) hardware and methodology allow the measurement of high-resolution heteronuclear multidimensional NMR spectra of macromolecules in living cells (in-cell NMR). In this study, we describe a protocol for the stable isotope ((13)C, (15)N and (2)H) labeling and structure determination of proteins overexpressed in Escherichia coli cells exclusively on the basis of information obtained in living cells. The protocol combines the preparation of the protein in E. coli cells, the rapid measurement of the three-dimensional (3D) NMR spectra by nonlinear sampling of the indirectly acquired dimensions, structure calculation and structure refinement. Under favorable circumstances, this in-cell NMR approach can provide high-resolution 3D structures of proteins in living environments. The protocol has been used to solve the first 3D structure of a protein in living cells for the putative heavy metal-binding protein TTHA1718 from Thermus thermophilus HB8 overexpressed in E. coli cells. As no protein purification is necessary, a sample for in-cell NMR measurements can be obtained within 2-3 d. With the nonlinear sampling scheme, the duration of each 3D experiment can be reduced to 2-3 h. Once chemical shift assignments and NOESY peak lists have been prepared, structure calculation with the program CYANA and energy refinement can be completed in less than 1 h on a powerful computer system.

Our group has recently reported that in the immortal human HepG2 liver cell line, sphingosine 1‑phosphate (S1P) increases transcription of plasminogen activator inhibitor type‑1 (PAI‑1), the major physiological inhibitor of fibrinolysis, within 4 h. The present study aimed to elucidate the molecular mechanisms underlying this effect. PAI‑1 expression was measured by reverse transcription‑quantitative polymerase chain reaction and immunoblotting. It was demonstrated that S1P increased PAI‑1 promoter activity but did not increase the activity of promoters lacking the hypoxia responsive element (HRE) 2. In addition, S1P transiently increased the concentration of hypoxia inducible factor (HIF)‑1α, a transcription factor capable of binding to HRE. When HIF‑1α was knocked down, the induction of transcription of PAI‑1 by S1P was no longer observed. Sphingosine kinase (SPHK) activity is increased by hypoxia. It was demonstrated that increases in the concentration of the HIF‑1α protein induced by hypoxia were prevented by treatment with SPHK inhibitor or S1P receptor antagonists. Thus, modification of the induction of HIF‑1α by S1P, leading to increased transcription of PAI‑1, may be an attractive therapeutic target for thrombosis and consequent inhibition of fibrinolysis associated with hypoxia.