Engineered Ferritin for Magnetogenetic Manipulation of ... et al. 2017... · establish subcellular control of proteins and organelles with unprecedented spatial and temporal resolution,

CommuniCation

1700189 (1 of 7) © 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.advmat.de

Engineered Ferritin for Magnetogenetic Manipulation of Proteins and Organelles Inside Living Cells

Domenik Liße, Cornelia Monzel, Chiara Vicario, John Manzi, Isabelle Maurin, Mathieu Coppey, Jacob Piehler,* and Maxime Dahan*

Dr. D. Liße, Dr. C. Monzel, Dr. C. Vicario, J. Manzi, Dr. M. Coppey, Prof. M. DahanLaboratoire Physico-ChimieInstitut Curie, CNRS UMR168, Paris-Science LettresUniversite Pierre et Marie Curie-Paris 675005 Paris, FranceE-mail: [email protected]. D. Liße, Prof. J. PiehlerDepartment of Biology/ChemistryDivision of BiophysicsUniversity of Osnabrück49076 Osnabrück, GermanyE-mail: [email protected]. I. MaurinLaboratoire de Physique de la Matière CondenséeÉcole Polytechnique91128 Palaiseau, France

DOI: 10.1002/adma.201700189

applications and for synthetic biology. In the past decade, tremendous progress has been made to exploit genetically encoded photoswitchable proteins for activation by light.[1,2] These optogenetic tools have recently been complemented by several approaches allowing magnetic actuation of protein functions.[3] Magnetic manipu-lation offers the advantage that it does not require optically transparent speci-mens, which is especially important for the application in tissues and organisms. Moreover, magnetic fields provide mul-tiple modes to control protein activity, via mechanical actuation, heating or by modulating protein concentrations within cells,[3b,4] making magnetic manipulation particularly promising for therapeutic applications.[5] Yet, a major obstacle for the breakthrough of magnetogenetics lies in equipping target proteins with magneti-cally controllable properties. This involves their conjugation with magnetic nanopar-

ticles (MNPs). The design of suitable MNPs must satisfy several physical, colloidal, and biochemical requirements. First, proper manipulation demands high magnetization of MNPs, which can be achieved by crystalline iron oxide such as magnetite or maghemite.[6] Next, proper surface biofunctionalization is needed to ensure highly efficient targeting to specific proteins while minimizing nonspecific interactions and, thus avoiding recognition by cellular degradation machineries. Finally, the total hydrodynamic diameter of fully assembled MNPs should remain below ≈50 nm to warrant efficient cellular delivery of the nanoparticles and maintain unhindered mobility in the cytoplasm.[7]

Here, we have tailored semisynthetic MNPs based on the natural protein cage ferritin as an intrinsically biocompatible nanoreactor for the synthesis of magnetite via its luminal fer-roxidase activity (Figure 1a,b). In brief, we fused monomeric EGFP (mEGFP) to the N-terminus of human heavy chain fer-ritin (HCF) to generate mEGFP::HCF protein cages densely coated with the highly stable and biologically indifferent GFP β-barrels, which served both as fluorescence marker and as specific targeting motif for proteins fused to an anti-GFP nanobody.[8] Cytosolic stealth properties of mEGFP::HCF protein cages were further optimized by chemical PEGyla-tion via surface exposed amines.[9] These MNPs combine an inherent monodispersity, biocompatibility with high saturation

Magnetogenetics is emerging as a novel approach for remote-controlled manipulation of cellular functions in tissues and organisms with high spatial and temporal resolution. A critical, still challenging issue for these techniques is to conjugate target proteins with magnetic probes that can satisfy multiple colloidal and biofunctional constraints. Here, semisynthetic magnetic nanoparticles are tailored based on human ferritin coupled to monomeric enhanced green fluorescent protein (mEGFP) for magnetic manipulation of proteins inside living cells. This study demonstrates efficient delivery, intra-cellular stealth properties, and rapid subcellular targeting of those magnetic nanoparticles via GFP–nanobody interactions. By means of magnetic field gradients, rapid spatial reorganization in the cytosol of proteins captured to the nanoparticle surface is achieved. Moreover, exploiting efficient nano-particle targeting to intracellular membranes, remote-controlled arrest of mitochondrial dynamics using magnetic fields is demonstrated. The studies establish subcellular control of proteins and organelles with unprecedented spatial and temporal resolution, thus opening new prospects for magneto-genetic applications in fundamental cell biology and nanomedicine.

Magnetogenetics

The astounding versatility and complexity of biological pro-cesses are fundamentally based on the cell’s ability to spatially and temporally regulate protein functions in a highly coor-dinated manner. Therefore, tools enabling acute control of protein localization and activity in living cells or organisms have huge potential, not only for unraveling the cellular determinants of biological processes, but also for medical

Adv. Mater. 2017, 29, 1700189

© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim1700189 (2 of 7)

www.advmat.dewww.advancedsciencenews.com

magnetization, and bio-orthogonal protein capturing. We dem-onstrate their efficient delivery, excellent intracellular inertness, and rapid subcellular targeting. Moreover, spatial relocaliza-tion of intracellular proteins was achieved. Finally, exploiting efficient nanoparticle targeting to organellar membranes, we demonstrated magnetically induced arrest of mitochondrial dynamics. Altogether, our magnetic nanoparticle probes enable subcellular control of protein and organelle functions with unprecedented spatial and temporal resolution, thus opening new avenues for magnetogenetic applications in fundamental cell biology and nanomedicine.

Recombinant mEGFP::HCF was produced in Escherichia coli with high yields and purified to homogeneity by thermal treat-ment followed by ammonium sulfate precipitation, size exclu-sion chromatography, and anion exchange chromatography (Figure S1, Supporting Information). After reacting surface

amines of mEGFP::HCF with poly(ethylene glycol) (PEG), MW: 2000 Da, (PEG) (mEGFP::HCFPEG2k), a magnetic core was syn-thesized at 65 °C as previously established for HCF[10] to yield the final product magnetic intracellular stealth ferritin (MagIcS HCF), further used for intracellular applications. The physico-chemical properties of HCF nanocages at different steps of the synthesis were characterized by analytical size exclusion chro-matography (aSEC), dynamic light scattering (DLS), supercon-ducting quantum interference device (SQUID) measurements, and transmission electron microscopy (TEM). The aSEC elu-tion volume observed for purified mEGFP::HCF confirmed robust cage assembly (Figure 1c), which was corroborated by DLS yielding a hydrodynamic diameter dH of 25.6 ± 0.3 nm. After PEGylation, the elution volume in aSEC was decreased as expected for dense PEG coating of the cage surface (Figure 1c). Accordingly, DLS yielded an increased dH of 27.7 ± 1.1 nm

Adv. Mater. 2017, 29, 1700189

Figure 1. Strategies and physicochemical characterization of magneto ferritin. a) Synthesis of magneto ferritin: purified mEGFP::HCF cages were chemically grafted with PEG2000 and subsequently loaded with a magnetite core exploiting the intrinsic ferroxidase activity. b) Targeting and manipula-tion inside living cells: after microinjection of magneto ferritin into the cytoplasm, soluble (I) or organelle-anchored (II) proteins fused to αGFP are rapidly targeted. A local magnetic field gradient controlled via a small magnetic tip allows for either relocalization of proteins inside the cell (I) or func-tional manipulation via magnetic forces (II). c) Analytical size exclusion chromatography of mEGFP::HCF (black), mEGFP::HCF grafted with PEG2000 (blue), and MagIcS HCF (red). d) SQUID measurement of magnetization as a function of the external magnetic field of MagIcS HCF (red) and of purified horse spleen ferritin loaded with ferrihydrate. e) TEM image of mEGFP::HCF stained with uranyl acetate. Scale bar: 100 nm. f) TEM image of MagIcS HCF (left) and size distribution of the magnetite cores within the protein cages (right). Number of analyzed cores N = 100. Scale bar: 50 nm.



for mEGFP::HCFPEG2k. This dH is in agreement with the esti-mated particle size calculated as 12 nm HCF + 2 × 5 nm GFP (including the linker between GFP and HCF) on each site + 2 × 1–2 nm for PEG2k. Importantly, the hydrodynamic properties of the unloaded mEGFP::HCFPEG2k closely matched those of MagIcS HCF, confirming that the structural integrity of the protein cage was not affected by the harsh reaction condition during magnetic core synthesis. SQUID measurements of MagIcS HCF (Figure 1d) yielded a saturation magnetization of 87 emu g−1, comparable to the saturation magnetization of magnetite synthesized in HCF reported before.[11] In contrast, much lower saturation magnetization was observed for iron-loaded ferritin purified from horse spleen, highlighting the importance of our ex vivo synthesis strategy. TEM imaging of MagIcS HCF particles confirmed robust assembly of magnetite inside the ferritin cage (Figure 1e) with a highly monodispersed core size of 7.1 ± 0.5 nm (Figure 1f).

A key challenge for biological applications of nanoparticles is their rapid uptake by cellular degradation machineries. This effect is very pronounced in the cytosol, where NPs can be rap-idly taken up into autophagosomes.[9] This severely obstructs intracellular nanoparticle targeting and magnetic manipula-tion. We therefore investigated the intracellular stability of mEGFP::HCFPEG2k in the cytoplasm of HeLa cells. Delivery of mEGFP::HCFPEG2k into cells was conveniently accomplished by microinjection, yielding intracellular concentration in the range of 10–35 × 10−9 m, as quantified by confocal fluores-cence microscopy (Figure S2, Supporting Information). Of note, we also established simultaneous loading of multiple cells with high efficiency by means of shear forces exerted via glass beads (Figure S3, Supporting Information)[12] or using pinocytotic influx. After microinjection, highly homo-geneous distribution of mEGFP::HCFPEG2k in the cytoplasm was observed by time-lapse fluorescence imaging (Figure 2a (right); Figure S4 and Movie S1, Supporting Information). The monodispersity of MagIcS HCF in the cytoplasm was confirmed via single particle tracking (Figure S5, Supporting Information), TEM images (Figure S5, Supporting Informa-tion), and organelle targeting experiments as described below. MTT viability assays (MTT = 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) were also performed to rule out toxicity associated to MagIcS HCF within cells (Figure S6, Supporting Information).

Over a time period of 20 min, we could not observe nano-particle clustering, demonstrating that the surface coating strategy yielded intracellular stealth properties. By contrast, mEGFP::HCF without PEGylation was recognized by intra-cellular degradation machineries, as indicated by the appear-ance of bright dots on a time scale of minutes (Figure 2a (left); Figure S4 and Movie S2, Supporting Information). Further-more, those dots were colocalized with autophagosomes upon microinjection into HeLa cells overexpressing the autophagy marker mCherry::LC3B (Figure S7, Supporting Information). These results are in line with our previous studies on the autophagosomal uptake of human light chain ferritin (LCF)[9] that has been recently linked to a specific recognition mecha-nism.[13] However, we noted a significantly slower aggregation of mEGFP::HCF compared to LCF, which may be caused by the dense mEGFP coating. Overall, our results highlight the

key importance of surface PEGylation to effectively elude intra-cellular xenorecognition and degradation.

Rapid and homogeneous distribution of mEGFP::HCFPEG2k after microinjection into the cell indicated free diffusion of mEGFP::HCFPEG2k in the cytosol, a key prerequisite for effec-tive subcellular targeting of those nanoparticles. For site-specific biorthogonal recognition, we fused target proteins to a nano-body against mEGFP (αGFP).[8] Recognition of mEGFP on the surface of mEGFP::HCFPEG2k by αGFP was quantified in vitro by solid phase detection. To this end, binding of αGFP labeled with DY647 (αGFPDY647) to immobilized mEGFP::HCFPEG2k was monitored by total internal reflection fluorescence spectros-copy (TIRFS) in a flow-through system (Figure 2b).[14] From the binding curve, an association rate constant of ≈1 × 106 m−1 s−1 was obtained while no significant dissociation was observed. Consistent with the high affinity of the αGFP/GFP interaction measured in solution,[15] this result confirmed very fast, quasi-irreversible binding of αGFP to mEGFP on the surface of HCF cages.

As proof-of-concept experiment, we explored targeting of MagIcS HCF to Tom20 fused to αGFP and mCherry (Tom20::mCherry::αGFP). Tom20 is a subunit of the protein translocase in the outer membrane (Tom) complex of mito-chondria, which are organelles with key functions in cellular energy metabolism. The C-terminus of the protein (where the tags are located) is positioned on the cytosolic side of the mito-chondrial membrane. After microinjection of MagIcS HCF into HeLa cells transiently overexpressing Tom20::mCherry::αGFP, very rapid targeting to mitochondria was observed (Figure 2c). Highly homogeneous co-localization of MagIcS HCF and Tom20::mCherry::αGFP confirmed unhindered mobility of MagIcS HCF inside the cytosol. Indeed, binding of MagIcS HCF to Tom20::mCherry::αGFP on mitochondria was extremely fast and already completed during microinjection. We there-fore tested microinjection of substoichiometric quantities of MagIcS HCF, which limited binding to Tom20::mCherry::αGFP of mitochondria surrounding the injection point (Figure S8, Supporting Information), corroborating very fast nanoparticle targeting properties. TEM imaging confirmed the localization at the mitochondrial outer membrane of targeted MagIcS HCF, visible as characteristic dark circular spots (Figure 2d). These spots, which were not observed in control cells (Figure S9, Sup-porting Information), correspond to the magnetite cores of MagIcS HCF and had an average density along the membrane of 8 ± 3 µm−1 (N = 5). Considering the 50 nm thickness of TEM slices, the labeling density is comparable to the surface den-sity (≈100 µm−2) previously reported for Tom20 complexes.[16] To further substantiate specificity and versatility of subcellular targeting, we used a model protein that could be shuttled from the endoplasmic reticulum (ER) to the Golgi apparatus (GA) uti-lizing the retention using selective hooks system (RUSH).[17] We employed tumor necrosis factor alpha (TNFα) fused to αGFP and mCherry as well as a streptavidin binding peptide (SBP) (SBP::αGFP::mCherry::TNFα) as a target protein. This protein is retained within the ER due to its interaction with streptavidin (SR) fused to Ii (SR::li), an isoform of the human invariant chain of the major histocompatibility complex, which comprises an ER retention motif. Release from the ER is triggered by addition of 40 × 10−6 m biotin, which out-competes SBP and dissociates

Adv. Mater. 2017, 29, 1700189



SBP::αGFP::mCherry::TNFα from SR::li. Thus, trafficking of SBP::αGFP::mCherry::TNFα along the natural secretory pathway is initiated to reach the GA after about 40 min (Figure S10, Sup-porting Information). Using this system, we explored live cell targeting of MagIcS HCF to SBP::αGFP::mCherry::TNFα before and after its release from the ER. For both ER and GA locali-zation of SBP::αGFP::mCherry::TNFα, characteristic distribu-tion and very high co-localization of MagIcS HCF was observed (Figure 2e,f). As for mitochondria, targeting was extremely fast and highly specific, highlighting the possibility to target MagIcS HCF to different subcellular structures with high efficiency.

Next, we explored the potential of MagIcS HCF to control localization of proteins in the cytoplasm. For magnetic manipu-lation, we employed localized magnetic field gradients gener-ated by a fine magnetic tip.[3b] The magnetic force exerted for an MNP at ≈10 µm distance of the tip extremity is in the femto-Newton range, as determined using the Boltzmann law from the profile of MagIcS HCF distribution close to the tip (Figure S11, Supporting Information). The force can also be calculated using an estimate ≈104 T m−1 of the magnetic gra-dient (as previously described[4a]) together with the MagIcS HCF magnetic moment inferred from SQUID measurements.

Adv. Mater. 2017, 29, 1700189

Figure 2. Stabilization and site-specific targeting of mEGFP::HCF. a) Aggregation of mEGFP::HCF after 20 min of microinjection (left, N = 7) and after 20 min of microinjection of PEG2000 grafted mEGFP::HCF (right, N = 6) in HeLa cells. Scale bar of 10 µm applies to both images. b) Binding kinetics of αGFPDY647 to immobilized mEGFP::HCFPEG2k (black) and to a surface solely functionalized with PEG2k (blue) as monitored by TIRFS. The association rate constant was derived from a monoexponential fit to the mEGFP::HCFPEG2k binding kinetics (red). c) Site-specific targeting of MagIcS HCF to mito-chondria (red: Tom20::mCherry::αGFP, green: MagIcS HCF, N = 24). The white line depicts the cellular boundaries as obtained from bright-field images (not shown). Scale bar: 15 µm. d) Transmission electron microscopy image of MagIcS HCF bound to Tom20::mCherry::αGFP. Single particles are indi-cated by white arrows. Scale bar: 100 nm (N = 5). e) Site-specific targeting of MagIcS HCF to the Golgi apparatus (red: SBP::αGFP::mCherry::TNFα, green: MagIcS HCF, N = 5). Scale bar: 15 µm. f) Site-specific targeting of MagIcS HCF to the endoplasmic reticulum (red: SBP::αGFP::mCherry::TNFα, green: MagIcS HCF, N = 10). Scale bar: 15 µm.



At first, we tested the response of freely diffusing MagIcS HCF after microinjection into HeLa cells. Strikingly, when approaching the magnetic tip within a few micrometers from the cell membrane (Figure 3a and Movie S4, Supporting Infor-mation), MagIcS HCF accumulated at the cytoplasmic side close to the tip within several seconds. Upon removing the magnetic tip, nanoparticles diffused back toward the cell inte-rior and could be repeatedly attracted and released. Evaluation of attraction and relaxation time constants yielded typical values of τA = 8.9 s and τR = 2.5 s, respectively (Figure 3a), demon-strating that fast spatial control of MagIcS HCF was possible.

We therefore explored the capability to control the subcellular localization of a cytosolic protein captured to the nanoparticle surface. As a model protein for probing spatial manipulation

of large size fusion proteins, we employed the catalytically active domain of T-cell lymphoma invasion and metastasis-inducing protein (TIAM), fused to mCherry and αGFP. MagIcS HCF were microinjected into HeLa cells expressing the ≈91 kDa TIAM::mCherry::αGFP construct, which was rapidly captured to the nanoparticle surface via the GFP–αGFP inter-action. Application of a magnetic tip resulted in the attraction of the magnetic particles, yielding effective relocalization of TIAM::mCherry::αGFP within several seconds (Figure 3b and Movie S5, Supporting Information). Protein–particle complexes were efficiently moved to different positions in the cell (Movie S5, Supporting Information), demonstrating that fast and revers-ible protein manipulation was possible. Typical MagIcS TIAMHCF attraction time constants were τA = 17 s and relaxation

Adv. Mater. 2017, 29, 1700189

Figure 3. Magnetic manipulation of MagIcS HCF inside living cells. a) Left: magnetic manipulation of MagIcS HCF after microinjection into HeLa cells (left). False-color intensity scale of mEGFP fluorescence. Scale bar: 5 µm. Right: particle attraction and relaxation kinetics (black) and fit (red) with monoexponential function (right). b) Left: magnetic manipulation of MagIcS TIAMHCF after microinjection into HeLa cells transiently overexpressing TIAM::mCherry::αGFP. False-color intensity scale of mCherry fluorescence. Scale bar: 5 µm. (N = 16) Right: comparison of attraction kinetics observed for free MagIcS HCF (black) and MagIcS TIAMHCF captured to MagIcS HCF (gray) and respective by a monoexponential function (red) (N = 6). c) Top: magnetic manipulation of MagIcS HCF targeted to mitochondria (green: MagIcS HCF, red: Tom20::mCherry::αGFP). Scale bar: 15 µm (N = 5). Bottom: detailed images illustrate changes in mitochondrial dynamics before and after magnetic field application. d) Autocorrelation function of mitochondria dynamics displayed for 100 randomly selected pixels before and after magnetic perturbation. e) Monitoring of changes in mitochondria membrane potential via TMRM intensity changes. Average TMRM intensity of mitochondria in MagIcS HCF targeted cell (gray) and adjacent nontargeted cell (black) and their ratio (red) before, during, and after magnetic field application.



Adv. Mater. 2017, 29, 1700189

time constants were τR = 13 s (Figure 3b and Figure S12, Sup-porting Information). The slow down by a factor of ≈2 com-pared to MagIcS HCF alone can be ascribed to the substantially increased hydrodynamic diameter caused by binding of up to 24 TIAM::mCherry::αGFP molecules/particle. Overall, these results demonstrate that a biocompatible magnetic nanocarrier based on ferritin can be exploited for fast, reversible, and flex-ible relocalization of proteins inside cells.

Finally, we explored the capabilities of MagIcS HCF for magnetic manipulation of biological functions with subcel-lular resolution. We chose the fusion and fission dynamics of mitochondria as an important organellar function implicated in quality control and energy metabolism[18] with emerging relevance in neurodegenerative and metabolic disorders as well as aging.[19] Following the injection of MagIcS HCF and their targeting to Tom20::mCherry::αGFP (as described above), time-lapse imaging revealed that the typical fusion and fission dynamics of mitochondria were maintained (Figure 3c and Movie S6 and S7, Supporting Information). However, upon approaching the magnetic tip close to the cell, fusion and fission processes were progressively arrested throughout the cell (Figure 3c and Movie S6, Supporting Information). The arrest was maintained when removing the magnetic field, suggesting that the magnetically induced perturbation had been converted into a biochemically sus-tained response. In contrast, mitochondrial dynamics was not altered in control cells targeted by mEGFP::HCFPEG2k without a magnetic core (Movie S7, Supporting Informa-tion), by MagIcS HCF without application of a magnetic field (Movie S8, Supporting Information), or in untargeted cells (Movie S9, Supporting Information). We quantified the change in mitochondrial dynamics of MagIcS HCF targeted cells before and after magnetic field application via temporal autocorrelation analysis and found a reduction in autocorre-lation amplitudes by more than 85% and an increase in the characteristic relaxation times from 48 to 76 s after magnetic perturbation (Figure 3d and Figure S13, Supporting Infor-mation). Loss of mitochondrial dynamics upon application of a magnetic field was accompanied by slight structural agglomeration as well as a decrease in the potential of the inner mitochondrial membrane, monitored by following the intensity of tetramethylrhodamine methyl ester (TMRM) probes (Figure 3e).[20] While this indicates a decline in mito-chondrial respiration, a complete loss of TMRM intensity and inhibition of respiration was not observed. By contrast, the membrane potential of mitochondria decorated with non-magnetic mEGFP::HCFPEG2k remained constant upon mag-netic field application (Figure S14, Supporting Information). In all cases, cellular morphology remained nearly unchanged during the experiment, corroborating the localized and selec-tive perturbation of mitochondrial performance. We fur-ther examined possible causes for the magnetically induced mitochondrial arrest. Importantly, suppression of mitochon-drial dynamics was observed either when applying a mag-netic field gradient via a magnetic tip (Movie S6, Supporting Information) or when using the uniform field (≈400 mT) of a magnet placed above the cell chamber (Movie S10, Sup-porting Information). Thus, we rule out that the arrest results from mechanical pulling on the magnetically labeled proteins

in the mitochondrial membrane. We then considered the potential role of magnetically induced interactions between nanoparticles, similar to what had been reported in other bio-logical contexts.[21a,b] The free energy required to separate two ferritins bound by dipole–dipole interactions was calculated as previously described,[21] yielding a value of ≈0.01 kBT. Thus, dipole–dipole interactions alone seem too low to explain our observations, and more complex, multi-nanoparticle effects may be involved. While the mechanism of changes in mito-chondria dynamics remains to be deciphered, a potential enhancer for mitochondrial interlacing and motion reduction is the multivalency of MagIcS HCF.

In conclusion, we have successfully engineered semisyn-thetic magnetic nanoparticles with tailored properties for remote-controlled intracellular manipulation of molecules and organelles. The versatile applicability of MagIcS HCF was achieved by carefully combining a small hydrodynamic size with high magnetic response, biocompatible stealth coating, and highly specific and ultrafast targeting moieties. Altogether, these properties yielded the capability of mag-netically controlling the subcellular localization of target proteins captured to the nanoparticle surface in live cells, or the spatiotemporal dynamics of intracellular organelles. We thus envision broad applicability of our new semisynthetic magnetic biomaterials, for instance as magnetic switch of intracellular signaling networks or of structural remodeling processes, with exciting perspectives both in fundamental research and in nanomedicine.

Supporting InformationSupporting Information is available from the Wiley Online Library or from the author.

AcknowledgementsD.L. and C.M. contributed equally to this work. D.L. acknowledges Oliver Beutel for fruitful discussions in the early phase of the project. C.M. acknowledges financial support from German Academic Exchange Service (DAAD) and Fonds der Chemischen Industrie. M.D. acknowledges funding from French National Research Agency (ANR) Paris-Science-Lettres Program (No. ANR-10-IDEX-0001-02 PSL). D.L. and M.D. acknowledge funding from Labex CelTisPhyBio (No. ANR-10-LBX-0038). This project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under Grant Agreement No. 686841 (MAGNEURON). The authors thank Karin Busch for fruitful discussions, and Gaëtan Cornilleau, Fanny Cayrac, Su Jin Paik, Thibaut Lagny, and Aurélie di Cicco for technical assistance and are grateful to Franck Perez and Gaëlle Boncompain for sharing constructs of the RUSH system, and to Maria Carla Parrini and Jacques Camonis for the mCherry::LC3B plasmid. D.L., C.M., C.V., J.M., and I.M. performed the experiments. D.L. and C.M. analyzed the data. D.L., J.P., and M.D. conceived the project. The manuscript was written by D.L., C.M., J.P., and M.D. The authors have no competing financial interests.

Conflict of InterestThe authors declare no conflict of interest.



Adv. Mater. 2017, 29, 1700189

Keywordsferritin, intracellular protein manipulation, magnetic nanoparticles, magnetogenetics, manipulation of mitochondrial dynamics

Received: January 10, 2017Revised: May 11, 2017

Published online: September 28, 2017

[1] T. N. Lerner, L. Ye, K. Deisseroth, Cell 2016, 164, 1136.[2] a) K. Muller, W. Weber, Mol. BioSyst. 2013, 9, 596; b) K. Zhang,

B. Cui, Trends Biotechnol. 2015, 33, 92.[3] a) S. A. Stanley, J. E. Gagner, S. Damanpour, M. Yoshida,

J. S. Dordick, J. M. Friedman, Science 2012, 336, 604; b) F. Etoc, D. Lisse, Y. Bellaiche, J. Piehler, M. Coppey, M. Dahan, Nat. Nanotechnol. 2013, 8, 193; c) S. A. Stanley, J. Sauer, R. S. Kane, J. S. Dordick, J. M. Friedman, Nat. Med. 2015, 21, 92; d) M. Rotherham, A. J. El Haj, PLoS One 2015, 10, e0121761; e) M. A. Wheeler, C. J. Smith, M. Ottolini, B. S. Barker, A. M. Purohit, R. M. Grippo, R. P. Gaykema, A. J. Spano, M. P. Beenhakker, S. Kucenas, M. K. Patel, C. D. Deppmann, A. D. Guler, Nat. Neu-rosci. 2016, 19, 756.

[4] a) F. Etoc, C. Vicario, D. Lisse, J. M. Siaugue, J. Piehler, M. Coppey, M. Dahan, Nano Lett. 2015, 15, 3487; b) D. Seo, K. M. Southard, J. W. Kim, H. J. Lee, J. Farlow, J. U. Lee, D. B. Litt, T. Haas, A. P. Alivisatos, J. Cheon, Z. J. Gartner, Y. W. Jun, Cell 2016, 165, 1507.

[5] Y. Pan, X. Du, F. Zhao, B. Xu, Chem. Soc. Rev. 2012, 41, 2912.[6] E. Tombacz, R. Turcu, V. Socoliuc, L. Vekas, Biochem. Biophys. Res.

Commun. 2015, 468, 442.[7] A. S. Verkman, Trends Biochem. Sci. 2002, 27, 27.[8] A. Kirchhofer, J. Helma, K. Schmidthals, C. Frauer, S. Cui,

A. Karcher, M. Pellis, S. Muyldermans, C. S. Casas-Delucchi,

M. C. Cardoso, H. Leonhardt, K. P. Hopfner, U. Rothbauer, Nat. Struct. Mol. Biol. 2010, 17, 133.

[9] D. Lisse, C. P. Richter, C. Drees, O. Birkholz, C. You, E. Rampazzo, J. Piehler, Nano Lett. 2014, 14, 2189.

[10] M. Allen, D. Willits, J. Mosolf, M. Young, T. Douglas, Adv. Mater. 2002, 14, 1562.

[11] E. Fantechi, C. Innocenti, M. Zanardelli, M. Fittipaldi, E. Falvo, M. Carbo, V. Shullani, L. Di Cesare Mannelli, C. Ghelardini, A. M. Ferretti, A. Ponti, C. Sangregorio, P. Ceci, ACS Nano 2014, 8, 4705.

[12] P. L. McNeil, E. Warder, J. Cell Sci. 1987, 88, 669.[13] J. D. Mancias, X. Wang, S. P. Gygi, J. W. Harper, A. C. Kimmelman,

Nature 2014, 509, 105.[14] M. Gavutis, S. Lata, P. Lamken, P. Müller, J. Piehler, Biophys. J.

2005, 88, 4289.[15] M. H. Kubala, O. Kovtun, K. Alexandrov, B. M. Collins, Protein Sci.

2010, 19, 2389.[16] C. A. Wurm, D. Neumann, M. A. Lauterbach, B. Harke, A. Egner,

S. W. Hell, S. Jakobs, Proc. Natl. Acad. Sci. USA 2011, 108, 13546.[17] G. Boncompain, S. Divoux, N. Gareil, H. de Forges, A. Lescure,

L. Latreche, V. Mercanti, F. Jollivet, G. Raposo, F. Perez, Nat. Methods 2012, 9, 493.

[18] a) S. Haroon, M. Vermulst, Curr. Opin. Genet. Dev. 2016, 38, 68; b) M. Roy, P. H. Reddy, M. Iijima, H. Sesaki, Curr. Opin. Cell Biol. 2015, 33, 111.

[19] a) T. A. Weber, A. S. Reichert, Exp. Gerontol. 2010, 45, 503; b) D. H. Cho, T. Nakamura, S. A. Lipton, Cell Mol. Life Sci. 2010, 67, 3435.

[20] R. C. Scaduto, L. W. Grotyohann, Biophys. J. 1999, 76, 469.[21] a) S. Chen, Q. Wu, C. Mishra, J. Kang, H. Zhang, K. Cho,

W. Cai, A. A. Balandin, R. S. Ruoff, Nat. Mater. 2012, 11, 203; b) R. J. Mannix, S. Kumar, F. Cassiola, M. Montoya-Zavala, E. Feinstein, M. Prentiss, D. E. Ingber, Nat. Nanotechnol. 2008, 3, 36.

Documents

Engineered Ferritin for Magnetogenetic Manipulation of ... et al. 2017... · establish subcellular control of proteins and organelles with unprecedented spatial and temporal resolution,