Nanomedicine: Nanotechnology, Biology and Medicine
Volume 6, Issue 3 , Pages 409-418, June 2010

Cellular prostheses: functional abiotic nanosystems to probe, manipulate, and endow function in live cells

  • Siyuan Lu, PhD

      Affiliations

    • Department of Physics, University of Southern California, Los Angeles, California, USA
    • ,
  • Anupam Madhukar, PhD

      Affiliations

    • Departments of Physics, Biomedical Engineering, Chemical Engineering, and Material Sciences, University of Southern California, Los Angeles, California, USA
    • Corresponding Author InformationCorresponding author: 3651 Watt Way, VHE 502, Los Angeles, California 90089-0241, USA.

Received 30 November 2009; accepted 23 January 2010. published online 29 January 2010.

Article Outline

Abstract 

A class of nanoscale (∼1–10 nm) structures designed to probe, manipulate, or endow function by direct interfacing with live cells is considered. Such a concept of cellular-level prostheses is illustrated via the example of light-activated nanoscale photodiodes capable of creating local electric fields that modulate existing voltage-gated ion channels in excitable cells. The dynamics of the membrane potential modulation by such photovoltaic functional abiotic nanosystems (PV-FANs) is modeled through an appropriate equivalent circuit. The feasibility of exceeding the typical ∼10 mV depolarization threshold for activating the action potentials is examined. In view of the continuing advances in the ability to design, synthesize, and characterize abiotic nanoscale systems that can provide desired function, several approaches to the implementation of PV-FANs are discussed. The FANs as “cellular prostheses” can provide a variety of functions in response to different stimuli and represent a paradigm-changing opportunity at the frontiers of nanomedicine.

From the Clinical Editor

A class of nanoscale (~1-10nm) structures designed to probe, manipulate, or endow live cell functions is demonstrated in this work. More specifically, light-activated nanoscale photodiodes were found capable of creating local electric fields that modulate existing voltage gated ion channels in excitable cells, thus allowing the generation of action potentials in excitable cells via external light stimulus in a controlled fashion.

Key words: Cellular prostheses, Functional abiotic nanosystem (FAN), Excitable cells, Neuronal Cells, Optical excitation, Neurobiology, Neurodegenerative Diseases

 

In recent years remarkable advances have been made in the controlled synthesis of nanoscale systems made of inorganic,1 organic,2 and biochemical3 building blocks. The incorporation of synthesized nanoscale structures into living cells and tissue for probing or modifying cell function without causing adverse side effects holds the promise of untold benefits for both basic biological science and medicine. Currently the most widely used synthesized nanostructures for biomedical purposes are the semiconductor nanocrystal quantum dots that act as fluorescent labels,4, 5 followed by magnetic nanocrystals as potential agents in magnetic resonance imaging.6 Beyond their use as imaging agents, nanocrystals are being explored as potential therapeutic agents by exploiting their simple heating functionality under light7 or alternating magnetic field8 for killing diseased cells. Indeed, the field of nanomaterials and nanostructures is poised to make innovative advances toward realizing protein size (1–10 nm) systems designed to provide a wide variety of advanced active functions that can be controlled noninvasively, say, optically or magnetically. When appropriately interfaced with live cells, such functional abiotic nanosystems (FANs)9, 10 may address different classes of diseases directly at the cellular level via probing, modifying, or even endowing new cell function. In this sense, FANs, as nanoscale prosthetic devices that replace or augment missing or impaired cellular functionality (i.e., “cellular prostheses”) represent a paradigm-changing opportunity at the frontiers of nanomedicine.

The intended application of the FAN determines its design and function. This article illustrates the potential of the FANs through the example of a well-recognized goal: optical activation of ion channels in excitable cells. To this end, in recent years light-sensitive ion channels have been introduced in neuronal cells through genetic manipulation of the ion channels of interest, and their light-activated control has been demonstrated in vitro11, 12, 13, 14 and in vivo.15 Additionally, placement of synthesized light-sensitive peptide-based ion channels in cells has been proposed16 and studied in artificial lipid membranes. In contrast to both these approaches, the FAN approach analyzed here is distinct, in that it utilizes light-activated synthesized or nature-derived photovoltaic FANs (PV-FANs)9, 10,17, 18, 19, 20 to modulate the transmembrane potential and thus impact the opening-closing of naturally occurring (i.e., existing) voltage-gated ion channels (VGICs) in excitable cells (such as sensory receptors, myocytes, and neurons). Naturally, the efficacy of such PV-FANs will depend upon their proximity and orientation with respect to the plasma membrane and ion channels. A schematic of potential arrangements is shown in Figure 1.

  • View full-size image.
  • Figure 1. 

    Schematic showing PV-FAN modified cell membrane. Two potential positions of PV-FANs are shown: (A) Transmembrane embedment (maximum excitation efficiency) and (B) attached to a membrane receptor through appropriate ligand. VECF and VCyt are the potential of the ECF (extracellular fluid) and the cytosolic region. ΔρFAN is the charge density within the excited PV-FAN. ΔρInd is the induced charge density in the surrounding medium.

In the following, we briefly capture an appropriate circuit model (Figure 2) of the combined PV-FANs/cell membrane system and estimate the capability of the PV-FANs for modulating membrane potential and activating cell function. (A detailed explanation of the model and its justification is presented in the Supplementary Material, available in the online version of this article.) Then we discuss different ways of physically implementing the PV-FANs and comment on the challenges involved in their delivery and targeted placement. Finally, we conclude by emphasizing the remarkable potential of FANs as “cellular prostheses” in clinical nanomedicine.

  • View full-size image.
  • Figure 2. 

    A zero-order equivalent circuit model of the cell membrane, PV-FAN, and their environment, together as a system. εM represents the equivalent electromotive potential directed from cytosol to ECF due to the ionic gradient across the membrane.

Back to Article Outline

Methods 

Modeling PV-FAN modulation of membrane potential 

Generically, a PV-FAN is a photoresponsive nanoscale structure in which, upon photon absorption, an electron transfers from one end of the PV-FAN to the other, generating an attendant electric field to modulate the transmembrane potential of an excitable cell. Such functionality can be realized in different materials, such as in the form of semiconductor- and metal-based PV structures or molecules of the donor-bridge-acceptor variety, as presented in the Discussion section. The effectiveness of a PV-FAN would depend upon its physical proximity and orientation with respect to the cell membrane. For maximal effectiveness in inducing membrane potential depolarization, desirable is a configuration normal to the membrane with the dipole directed from the extracellular fluid (ECF) to cytosol as shown in Figure 1 (arrangement A). However, other more easily realized configurations may be effective enough in some situations.

The basic characteristics of the PV-FANs to account for are its absorption coefficient σ and the decay rate kd of the electron-hole pair created upon photon absorption. When the PV-FAN is illuminated by photons of energy hυ at a power density P, the rate of light-induced excitation is kex = Pσ/hυ. Considering both the excitation and decay process in the PV-FANs, the areal density of the excited (with separated electron-hole pair) PV-FANs follows:

(1)
where n0 is the total areal density of PV-FANs and nex is the areal density of excited PV-FANs.

Once the charge polarization inside a PV-FAN occurs, it induces a dielectric response from the surrounding environment composed of ionic media, membrane lipids, proteins, etc. The commonly employed model of ionic medium response is the Debye model in which the induced charge redistribution and the associated electric field will decay on the scale of the Debye length away from the polarized PV-FAN. The typical Debye length is ~1 nm for physiological media of ionic strength ∼100 mM. If polarized PV-FANs are located close enough to a VGIC, a sufficiently large electric field induced by the PV-FANs in the neighborhood may exert sufficient force on the voltage-sensitive segment of the channel to cause the channel to open. We note, however, that because the cell membrane and surrounding medium are highly inhomogeneous structures, local field effects may not be adequately accounted for using bulk dielectric response description, and molecular-level simulations may be required to determine more accurately the interaction between individual PV-FAN and VGIC.

As a reasonable starting point we can account for the interaction between PV-FANs and the cell membrane by considering for the averaged membrane potential modulation caused by the total charge redistribution induced by a collection of excited PV-FANs distributed over a reasonably large area of the cell membrane. The effect of PV-FANs on the cell membrane is essentially that of an alternating current source transferring charge across the cell membrane. (For a detailed discussion of this assumption and its justification see the Supplementary Material in the online version of this article.) With the excitation (de-excitation) of an individual PV-FAN, effective charge α · e is transferred from the ECF (cytosol) to the cytosol (ECF) side of the membrane. Here e is the unit charge and α = (ΔqCyt – ΔqECF)/2e, where ΔqCyt and ΔqECF are the net charges in the cytosolic and ECF regions contributed by an individual excited PV-FAN. Thus, the total charge density transferred is α · e · nex and the equivalent areal current density iFAN can be written as:

(2)
where iFAN from the cytosolic side to the ECF side is defined as positive. The constant α depends on the geometry and location of the PV-FANs with respect to the cell membrane and affects the efficiency of PV-FANs for modulating the membrane potential. In the following we only consider PV-FANs in arrangement A (Figure 1), which is the most efficient arrangement because the PV-FAN sits across the cell membrane with the separated hole and electron residing, respectively, in the cytosolic and ECF regions, thus giving ΔqCyt ≈ e, ΔqECF ≈ –e, and α ≈ 1 (i.e., each excited PV-FAN transfers nearly one unit charge across the cell membrane).

Because our primary purpose is to assess whether the threshold for activating the cell function can be reached for reasonable characteristics and densities of the PV-FANs, we apply the well-known and commonly employed “single-compartment” and the “passive integrate-and-fire” model21 to describe the cell and consider only the subthreshold membrane potential modulation regime. It is understood that once the membrane potential crosses the threshold value corresponding to the particular function of the excitable cell, new processes come into play and require the model to be enlarged.

With such a model of PV-FANs and cell membrane as noted above, the system of PV-FAN–embedded membrane can be described using the circuit depicted in Figure 2. The cell membrane is represented by a capacitor and a resistor in parallel. The membrane capacitance cM is ∼1 μF/cm2. The typical membrane resistance rM is ∼1 to 100 kΩ∙cm2. The charging/discharging rate, kM = 1/rMcM, of the membrane is typically ∼10 to 1000 sec–1. The current density through the ion channels on the cell membrane (iM, cytosol to ECF defined as positive) can be written as:

(3)
where VM is the membrane potential and Vrest is the resting membrane potential.

Furthermore, the equivalent current density induced by the PV-FANs (iFAN) and current density through the ion channels on the cell membrane (iM) work additively to modulate membrane potential:

(4)

Finally, we apply the initial condition that at t = 0 when the light is turned on, the cell is at resting membrane potential, and none of the PV-FANs are in the excited state. Namely,

(5)

Eqs 1–5 provide a complete description of the behavior of the membrane potential as a function of time following PV-FAN excitation. The solution is straightforward, and we find that the cell membrane potential is a biexponential function of time in the subthreshold region and is given by:

(6)

We define the membrane potential modulation as ΔVM = VM – Vrest. In Figure 3 are plotted ΔVM as a function of time calculated using Eq. 6 for illustrative values of relevant parameters (kd = 100, 300, 1000 sec–1, kex = 200 sec–1, kM = 100 sec–1, n0 = 2 × 103 μm–2, cM = 1 μF/cm2).

  • View full-size image.
  • Figure 3. 

    Plots of membrane potential modulation (ΔVM) as a function of time for illustrative values of relevant parameters (kd = 100, 300, 1000 sec−1 and kex = 200 sec−1, kM = 100 sec−1, n0 = 2 × 103 μm−2, cM = 1 μF/cm2).

From Eq. 6 we find that the maximum membrane potential modulation ΔVM (max) = VM (max) – Vrest is given by

(7)
and depends on the following three parameters: (1) en0/cM, which represents the membrane potential depolarization if all PV-FANs on the membrane are excited at the same time; (2) kex/kM, the ratio of PV-FAN excitation rate to membrane charging rate, and (3) kd/kM the ratio of PV-FAN recombination (decay) rate to membrane charging rate.

The behavior of the ΔVM (max) response surface with varying kex/kM and kd/kM is shown in Figure 4. For the PV-FAN excitation (kex) and decay (kd) rates in the range of 0.1kM to 10kM, ΔVM (max) varies between ∼0.01en0/cM to 0.72 en0/cM. Note that ΔVM (max) scales linearly with en0/cM. For uniform PV-FAN densities in the range 1 × 102 to 1 × 104 μm–2, this gives the maximum transmembrane modulation ∼1 to 100 mV.

Back to Article Outline

Results 

Activating cell function: The case of action potential firing in nerve cells 

The results of Figure 4 provide a range of values for the maximum transmembrane potential modulation as a function of the PV-FAN density and ratios of the fundamental rate processes of the PV-FAN excitation and decay in relation to the membrane charging-discharging involved. For any given class of excitable cells, of central importance to the clinical relevance of the proposition of this work is the ability to exceed the typical threshold value relevant to activating the function of the cells for a reasonable density of PV-FANs. In the following we consider the case of neuronal cells and discuss the requirements on PV-FANs to induce action potential (AP) firing. It is well established that the depolarization of membrane potential beyond a threshold (ΔVth) of ∼10 mV causes the statistical opening of a sufficient number of voltage-gated sodium channels, which, in turn, initiates the AP to be generated and travel down the axon. In a typical neural network, AP firing at a rate of ∼10–100 Hz is considered to provide meaningful signal. This traditional macroscopic electrode-based stimulation of electrophysiology may, in appropriate situations, be replaced by spatially localized noninvasive stimulation through cell-attached PV-FANs, if these can create the needed ∼10 mV membrane depolarization beyond the threshold.

Whereas the membrane capacitance can be treated as a constant at ∼1 μF/cm2 for most biological membranes,21 the resting membrane resistance can vary quite markedly which leads to a varying kM in different types of cells. Table 1 shows a survey of the properties of different types of neuronal cells. A reasonable bracketing range for kM is ∼10–1000 sec–1. For example, for retinal ganglion cells (RGCs), kM ranges from ∼10 sec–1 to 200 sec–1. A reasonable bracketing range for the threshold membrane potential depolarization needed to fire AP (ΔVth) is ∼10–20 mV (Table 1). For RGCs, ΔVth is close to ∼15 mV. Given the range of kM and ΔVth, we choose the following three cases to estimate the requirement on PV-FANs: (a) the most relaxed situation: kM = 10 sec–1 and ΔVth = 10 mV; (b) an intermediate situation (probably the closest to the RGCs as well) kM = 100 sec–1, ΔVth = 15 mV; and (c) the most demanding situation: kM = 1000 sec–1 and ΔVth = 20mV.

Table 1. A survey of membrane charging-discharging rate (kM) and threshold of membrane potential modulation for action potential firing (ΔVth) in different types of neuronal cells
ΔVth
(mV)		∼12
Rabbit RGCs⁎		∼17
Salamander RGCs†		∼15
Turtle BON cells‡		∼11–20
Cat sensorimotor cortical neuron§
kM
(sec–1)		∼15
Salamander inner retinal neurons		∼33
Rat hippocampal neurons⁎⁎		∼81–145
Cat sensorimotor cortical neurons§		∼12–200
Cat RGCs††

BON, basic optic nucleus; RGC, retinal ganglion cells.

⁎Oesch N, et al Neuron. 2005;47:739-50.

†Mitra P, et al Vis Neurosci. 2007;24:79-90.

‡Kogo N, et al J Neurophysiol. 1997;78:614-27.

§Safstrom CE, et al J Neurophysiol. 1984;52:244-63.

Coleman PA, et al J Neurophysiol. 1989;61:218-30.

⁎⁎Johansson S, et al J Physiol. 1992;444:129-40.

††O'Brien BJ, et al J Physiol. 2002;538:787-802.

Plotted in Figure 5, A–C are the response surfaces corresponding to the minimum requirements on PV-FAN (n0, kex, kd) for AP firing (exceeding the potential modulation threshold) under the three conditions above.

  • View full-size image.
  • Figure 5. 

    Surface plots showing the minimum required photovoltaic functional abiotic nanosystem density (n0) as a function of excitation and decay rate (kex and kd) to allow action potential firing under the following three conditions bracketing typical neuronal cell characteristics. (A) Most relaxed situation: kM = 10 sec−1 and ΔVth = 10 mV; (B) intermediate situation, close to retinal ganglion cells: kM = 100 sec−1, ΔVth = 15 mV; (C) most restrained situation: kM = 1000 sec−1 and ΔVth = 20 mV.

For the intermediate case (Figure 5, B), to induce AP firing, if we keep the excitation rate (kex) ∼100 sec–1 and design the PV-FANs to have a decay rate (kd) of ∼100 sec–1, then a PV-FAN density n0 ∼3 × 103 μm–2 will be needed. If we assume that a FAN has an absorption cross-section of ∼1 nm2 for incoming photons, a photon flux of 1 × 1016 cm–2 sec–1 is necessary to achieve an excitation rate of ∼100 sec–1. For photons of energy 2 eV (620 nm, red light), this photon flux translates into ∼3 mW/cm2. Such light power density is about 3% of full-sun illumination. The requirement of the electron-hole recombination rate of ∼100 sec–1 (or excited-state lifetime ∼10 msec) will have to be met by appropriate design of the PV-FAN and the material system to be chosen. Indeed, from an energetic point of view, over their typical ∼10 msec excited-state lifetime, the photoenergy absorbed by the PV-FANs is ∼1 × 10–15 J/μm2. This is already an order of magnitude greater than the estimated total work of ∼1 × 10–16 J/μm2 needed to open all such voltage-gated sodium channels (i.e., work to move their voltage-sensitive segments; resting membrane potential is ~70 mV, the gating charge and density of voltage-gated sodium ion channels in the nodes of Ranvier are approximately 6e and ∼2000 μm–2. See Hille et al22) on the cell membrane. Similarly, in the most relaxed situation (Figure 5, A) for the same photoexcitation rate kex (∼100 sec–1) and recombination rate of kd (∼100 sec–1), a PV-FAN density of 1 × 103 μm–2 is required (about three times less than the intermediate situation) to induce AP firing. In the most demanding situation (Figure 5, C), for the same photoexcitation rate kex (∼100 sec–1) and recombination rate kd (∼100 sec–1), a FAN density of ∼2 × 104 μm–2 would be necessary (about seven times more than the intermediate situation).

It is important to note that the above estimation assumes the PV-FANs and ion channels are both independently and uniformly distributed on the cell membrane and the induced cell membrane modulation is estimated for PV-FAN areal density averaged over the whole cell membrane area. In reality, for an AP to be generated, only local membrane depolarization ∼15 mV is necessary to favor the opening of voltage-gated sodium channels. Thus, to minimize the required light intensity and the number of PV-FANs needed, it is desirable that the FANs be localized in the immediate proximity of the targeted VGICs.

Back to Article Outline

Discussion 

Potential implementations of PV-FANs 

For a proper perspective of the significance of the proposed notion of PV-FANs as light-based actuators of membrane potential modulation in excitable cells, it is worth recapturing here the current situation with respect to notions of light activation of cells. Among excitable cells, the ability to render normally nonphotoresponsive neuronal cells responsive to light has been recognized for some time as an important basic step to permit nonelectrical stimulation for studies of fundamental cellular processes.23, 24, 25, 26 These include single23 or two-photon24 direct stimulation of cortical neuronal cells that led to cell depolarizations up to AP thresholds; photostimulated release of caged suitably modified glutamate to chemically activate APs.25, 26 These studies are in contrast to introducing light-sensitive ion channels in the cell. Photoresponsive peptide ion channels based on isomerizable organic chromophores (e.g., azobenzene) have been synthesized and proposed to be strategically placed in neuronal cell membranes to cause photostimulation.16 Also noted is the alternative approach of genetic modification11, 12, 13, 14 of nonphotoresponsive cells to express photosensitive proteins (e.g., rhodopsin,11 melanopsin12) to cause phototriggered opening and closing of VGICs. Channelrhodopsin-2, found in green algae, when expressed in mammalian cells, forms a small-conductance nonselective cation channel that can be activated by 500-nm light (all-trans retinal changing to cis; returns to trans in the dark).13 Unlike rhodopsin and melanopsin, channelrhodopsin-2 does not require a signaling cascade to depolarize membranes. Hence it can make neurons respond to flashing light. Another approach exploits both, genetic modification of an expressed protein and biconjugated azobenzene as the ligand switch binding to the modified ion channel protein.14 However, this system requires two different wavelengths (~ 380 nm in ultraviolet and ~ 500 nm in visible) to turn on and off the molecular switch.

The FANs in general, and PV-FANs in particular, are a distinct approach that seeks to manipulate existing constituents of the cells to induce a desired response stimulated by the actuation of the FAN by external means. With the advances in the ability to synthesize and characterize nanomaterials, the PV-FANs can potentially be realized using a variety of material building blocks. Two well-suited generic categories are schematically depicted in Figures 6, A and 7, A. The first is a heterojunction between two materials that represent either semiconductor-metal (S-M) or semiconductor-semiconductor (S-S) combination (Figure 6, A), implemented in all-inorganic, all-organic, or hybrid inorganic-organic material combinations. In this category the most versatile in terms of tailoring their photoresponse over the visible to infrared wavelengths are the inorganic semiconductor- and metal-based nanocrystal heterojunctions. Although a single inorganic semiconductor quantum structure can create a photoexcited dipole, the requisite characteristics, as discussed above, needed for manipulating ion channels are more likely realized in S-M and S-S heterojunctions wherein the relevant energies of the electron states (conduction and valence bands of the semiconductor and the Fermi energy of the metal) in the two components are aligned as shown in Figure 6, B and C, respectively. This ensures that the electron-hole pair generated in the photon-absorbing component easily separates, thereby creating a dipole of strength proportional to the separation.9, 10 The absorption wavelength is tailored through the size and shape of the absorbing inorganic semiconductor—i.e., the quantum confinement effect. The subject of semiconductor heterojunctions is highly developed as it underlies much of the sophisticated semiconductor technologies for computing and communications, and thus provides a powerful platform to guide synthesis of inorganic semiconductor-based PV-FANs. Indeed, historically the simple S-M heterojunction provided the first semiconductor device—the current rectifier—that revolutionized communications over a hundred years ago in the form of telegraphy and radio.27 It is thus particularly striking to contemplate that the same basic principle—but with the advantage of over a century of technological progress that has led to the emergence of materials and tools of nanotechnology, implemented on the scale of 1 to 10 nm and rendered compatible with living cells utilizing the advances in molecular biology and biochemistry for its surface functionalization—offers today a revolutionary approach of nanomedicine as therapy for a variety of diseases involving excitable cells.

  • View full-size image.
  • Figure 6. 

    (A) Schematic showing the implementation of photovoltaic functional abiotic nanosystem using a heterojunction. Material A or B represent either semiconductor-metal (S-M) or semiconductor-semiconductor (S-S) combination. (B) Schematic band diagram of a S-M heterojunction. (C) Schematic band diagram of a type II S-S heterojunction. EC and EV represent the energy of the conduction and valence band edges of the semiconductors. Ef is the Fermi energy of the semiconductor and the metal. Φn is the height of the Schottky barrier.

  • View full-size image.
  • Figure 7. 

    (A) Schematic showing the implementation of photovoltaic functional abiotic nanosystem using a molecular donor-acceptor pair connected through a bridge molecule. (B) Schematic energy levels of donor-acceptor pair.

The second category (Figure 7, A) is molecular donor-acceptor constructs involving a bridge of variable length to create a dipole through excitation of electron in the donor and its transfer across the bridge to the acceptor. The alignment of the HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) of the donor and acceptor to achieve the desired charge transfer is as shown in Figure 7, B. The donor-acceptor complex can be constructed from inorganic, organic and biochemical molecules. Indeed, more than 2 decades ago such structures involving dimethoxynaphthalene donor and dicyanethylene acceptor separated by a rigid hydrocarbon framework of variable lengths were synthesized and their excited-state dipoles and lifetimes demonstrated to reach, respectively, 70 debye and several hundred nanoseconds.17 A variety of inorganic and organic molecule-based approaches lend themselves to the realization of charge transfer–based PV-FANs, including molecules that form liquid crystal phases but could be used in isolated distributions as PV-FANs. PV-FANs may also be implemented utilizing amino acids and peptides.18 Here guidance can be derived from biological systems in which electron transfer typically involves a combination of covalent and noncovalent interactions between donor-acceptor pairs.28 Pathways involving peptide covalent bonds typically enhance the donor-acceptor separation and correspondingly decrease their interaction. Electron transfer can be mediated via hydrogen bonds that act primarily as linkers and allow shorter donor-acceptor separation (and hence more efficient transfer). Such hydrogen-mediated transfer can also involve proton-coupled electron transfer wherein proton motion is intimately coupled to the stabilization energy in the configurational pathway of transfer. Indeed, the role of nuclear motion in impacting the rates of energy and charge transfer has been examined for some time.29, 30, 31 The photophysics of light-driven electron transfer in hydrogen-bonded donor-acceptor systems such as from porphyrin donors to acceptors was also examined early.32, 33 Thus, considerable potential resides in this class of donor-acceptor structures to be tailored for creating PV-FANs suitable for attachment to cells.

Regardless of the particular material categories employed to synthesize the PV-FANs, the ability to utilize them to modulate the cell membrane potential will depend upon their specific characteristics as well as density and distribution in the cell. The magnitude and duration of the PV effect in a heterojunction PV-FAN is decided by (a) its capability to capture photons, (b) the efficiency of photon-created electron-hole separation, and (c) the time duration before their recombination (i.e., the dipole lifetime). The efficiency of light absorption and charge separation is typically very high for inorganic nanocrystals, and thus an implementation utilizing, say, a S-M or S-S nanophotodiode is a good candidate from this perspective. The typical lifetime of the photoexcited electron-hole pair in a semiconductor nanocrystal is, however, nanoseconds. In nanocrystal S-S heterojunctions it can be several hundred nanoseconds.34 The expected duration of the electric field generated by any particular inorganic PV-FAN is short as compared with the time scale of some microseconds to milliseconds on which the VGIC protein subcomponents change their configuration to allow exchange of ions across the membrane.21 In implementations involving molecules, the added element of the configurational free-energy component may allow the added flexibility needed to attain longer dipole lifetimes of microseconds. We note, however, that it is not known for what minimum fraction of the ion channel opening time does the exciting electric field pulse have to be continuously on to set the ion channel opening in motion. Presumably a short pulse of appropriate strength is sufficient to set the ion channel's voltage-sensitive components (such as the S4 region of sodium or potassium channels) in motion, and that, in turn, is sufficient to complete the opening of the channel and thus the process of ion flow across the channel. Indeed, examining this fundamental aspect of ion channel function itself is enabled by our proposed PV-FANs. It is, however, important to recognize that their usage for transmembrane potential modulation does not require all the PV-FANs to simultaneously generate electric field (i.e., to be in the “on” state) over the time duration of the ion channel opening, but only some statistically “on” PV-FANs at the required density and distribution are sufficient to create the requisite ∼15 mV transmembrane potential change. The disparity in the PV-FAN excited-state lifetime and the ion channel opening pulse duration time scales thus has to be bridged by a combination of (1) appropriate design of the rapid separation and trapping for long times (at least microseconds) of the photoexcited electron by quantum mechanical state(s) introduced, in the case of the inorganic semiconductor-based PV-FAN, at the ligand-covered semiconductor surface of the nanoscale heterojunctions; and (2) optimization of the local spatial density and aligned orientation of the PV-FANs embedded close to the VIGC regions so that statistically at any given time a large number of these are photoexcited to collectively provide the needed magnitude and duration of the electric field generated by the trapped electron separated from the hole.

PV-FANs would be useful for a variety of biomedical applications wherever the manipulation of membrane potential and the states of VIGCs control an important aspect of the physiology of a living system. As for any externally introduced agent, the issue of delivery and biocompatibility is an important consideration for the physical implementation of PV-FANs, although the specific requirements may vary significantly according to the intended application. Simplest is in vitro applications in which PV-FANs may be used to induce localized optical excitation of excitable cells to facilitate a system-level understanding of the relevant physiology, such as the collective response of a neural network. These applications do not demand long-term biocompatibility, and the “sample” (cells) is typically physically accessible for the delivery of both the PV-FANs and the excitation light. For stabilizing the PV-FANs across the cell membrane (Figure 1, arrangement A) so that the efficiency of membrane potential modulation can be maximized, the surface functionalization of the PV-FANs has to mimic to a certain extent the naturally occurring amphiphilic transmembrane proteins, which have hydrophobic membrane-spanning domain(s) that interact with fatty acyl groups of the membrane phospholipids and hydrophilic domains extending into the aqueous medium on each side of the membrane. For this perspective the inorganic materials–based PV-FANs, although at the present time provide response to the widest tunable range of light wavelengths, including the near- and mid-infrared regime (1–20 μm), pose the greatest challenge of long-term stability and compatibility in their interfacing with live cells. Small molecule, inorganic, organic, and bio-organic building block–based PV-FANs (or FANs in general) are likely to find easier solutions to appropriate attachments and stability in live cells because of the flexibility they can provide in the configurational free energy of the combined FAN-cell-environment system.

As for their potential applications in vivo, PV-FANs may be directly interfaced with the retinal neurons (such as RGCs) and serve as therapeutic agents to restore vision for the treatment of retinal degeneration in which photoreceptor cells are lost.9, 10 For such an application, retinal neurons can be accessed by intraocular injection for the purpose of PV-FAN delivery. The optics of the eyes naturally allows the excitation light to be projected onto the PV-FAN–embedded cells. Furthermore, cortical application of PV-FANs similar to those envisioned for genetically introduced light-sensitive channels15 may also be considered, although inherent to such applications are the challenging issues of achieving noninvasive delivery of the 1- to 10-nm PV-FANs (overcoming the blood-brain barrier) and the optical excitation (through the skull) to the targeted cortical region. Indeed, for any specific potential in vivo application of PV-FANs, experiments would have to be undertaken to optimize their targeted delivery and retention time while minimizing their toxicity. The results of such experimental work would ultimately determine whether a certain design of PV-FANs is suitable or not for a specific application. The tasks involved, however, are leveraged enormously by the recent advances in the investigation of nanoscience-based drug delivery systems35 and the toxicity of nanomaterials.36 For instance, targeted delivery of PV-FANs to a specific set of cells could be facilitated by using carrier nanoparticles functionalized with cell-specific ligands35; the toxicity analysis of the PV-FANs could benefit from the molecular-level understanding of the nanomaterials-induced toxicity, which is being aggressively examined by the community.36

To conclude, advances in the ability to create protein-size nanoscale functional devices made out of inorganic, organic, and bio-organic building blocks, and the parallel developments in the ability to surface-functionalize these FANs to allow their attachment and coexistence within the living-cell environment, open the possibility of their usage as cellular-level probes, manipulators, and therapeutic agents. In the first two roles, FANs provide a critically important step forward for advancing understanding of biology at the molecular level. Appropriately designed and interfaced FANs open the door for providing means of quantitative data acquisition while monitoring and/or manipulating simultaneously multiple intracellular and cellular processes, and thus pursuing a systems-level approach and perspective. Applied to biological processes speculated or known to underlie onset of disease, FANs open a new paradigm and bridge toward molecular-level studies and understanding of physiology and pathology. In their role as providing a specific function, either lost to the cells of interest or endowed anew to offset loss of function elsewhere, FANs are a powerful therapeutic agent and open a new frontier of nanomedicine. In this article we have illustrated the potential of this approach of “cellular prostheses” by considering the classes of cells possessing VGICs and the modulation of the transmembrane potential in such cells utilizing synthetic PV-FANs that provide local electric fields in response to excitation via light. Such PV-FANs can be realized in all major classes of materials—inorganic, organic, and bio-organic—utilizing appropriate building blocks and advances in the techniques and methodologies of nanoscience and nanotechnology. Most importantly, for the particular example of generating local (i.e., intracellular level) electric fields to modulate the transmembrane potential enough to activate action potentials in neuronal cells, the most challenging element in the design and implementation of the PV-FANs is controlling their excited dipole lifetime (through proper internal dynamics in the case of organic and bio-organic PV-FANs). With targeted delivery of appropriate PV-FANs, a task that is enormously important but fortunately helped by the advances in targeted delivery of drugs and other therapeutic agents, our modeling and simulation results presented here indicate that it should be possible to achieve sufficient local density of such PV-FANs to photoactivate AP firing in neuronal cells. Among a variety of applications, this would be of value for probing the physiology of the neural circuit and for the treatment of retinal degeneration caused by loss of photoreceptors. We hope the views and analysis presented here will induce interest in the study of FANs.

Back to Article Outline

Appendix A. Supplementary Material 

Download file

download text

• The Physical Processes Involved in the PV-FAN/Cell Membrane System and the Circuit Model

The basic function of the PV-FAN is to generate a dipole upon absorption of light in the appropriate wavelength regime and for the attendant electric field to contribute to the depolarization of the transmembrane potential of the excitable cell. As such, the effectiveness of a PV-FAN would depend upon the physical proximity and orientation of the FANs with respect to the cell membrane. For maximal effectiveness in inducing charge polarization across the membrane, desirable is a configuration normal to the membrane with the dipole directed from the extracellular matrix (ECM) to cytosol as shown in the schematic of Figure 1 arrangement (a) (see main text). However, other more easily realized configurations may be effective enough in some situations and the modeling presented here is not constrained to a specific arrangement. To analyze the efficacy of PV-FANs, we consider an equivalent circuit model of the FANs, the cell, and their environment, together as a biological system, as depicted in Figure 2 (See main text).

The key elements of each of the physical-biological components are represented, as appropriate, via equivalent capacitance, resistance, etc. as discussed below.

• Excitable Cell Membrane

The standard equivalent circuit of the cell plasma membrane used in the literature is depicted in purple in Figure 2 (see main text).S1 The cell membrane is 3-5nm in average thickness and has a capacitance cM ∼1μF/cm2, almost independent of cell type. For excitable cells in the resting condition (no massive opening of voltage or ligand gated ion channels), the typical resistance rM of the membrane is on the order of 1-100kΩΩcm2, depending on cell type. The charging/discharging rate, kM = 1/rMcM, of the plasma membrane is typically on the order of ∼10-1000sec-1. The ionic gradient across the plasma membrane, created by the ATP driven iontransporters, leads to an electromotive potential directed from inside (cytosol) to outside (extracellular matrix) of the cell.20 Conventionally, the potential inside the cell membrane with respect to outside the membrane is defined as the membrane potential (VM). The membrane potential for cell in the resting condition, denoted as the resting potential (Vrest), is typically ∼ - 70mV. For a typical cell of membrane area of ∼105 μm2 this corresponds to ∼ 5 × 108 unit charges across the membrane distributed on the whole cell (a uniform charge areal density of 5 × 103 e/μm2). Upon electric disturbance of the excitable cell, if the membrane potential change (ΔVM) exceeds a certain threshold (ΔVth) value typical of the particular class of excitable cells, then the corresponding function is activated. Most popularly studied are, of course, the neuronal cells for which the modulation threshold to activate a transient spike of the membrane potential, the action potential, is typically ∼10-15 mV above the resting potential.

• PV-FANs

Generically, a PV-FAN is a photo-responsive nanoscale structure in which, upon absorption of a photon, an electron transfers from one end of the PV-FAN to the other results in a charge redistribution inside the PV-FAN: , where , are, respectively, the wavefunctions of the hole and the electron and e is the unit charge. Such functionality can be realized in different ways and materials such as in the form of more traditional semiconductor and metal based photovoltaic structures or molecules of the donorbridge-acceptor variety, as discussed in Sec.IV. It is reasonable to assume that the charge separation is instantaneous as in most physical implementations it is indeed likely rapid compared to all other relevant time scales of motions. Other than the positions of the separated electron and hole inside the PV-FAN with respect to the cell membrane, for the purposes of modeling the cell membrane potential modulation in response to perturbation by such a structure, the most basic characteristic of the PV-FANs to account for is its absorption coefficient σ and the decay rate kd of the electron-hole pair created upon absorption of the photon. When the PVFAN is illuminated by photons of energy hυ at a light power density P impinging continuously in time, the rate of light induced excitation is, kex = Pσ /hυ. Considering both the excitation and decay process in the PV-FANs, the areal density of the excited (with separated e-h pair) PVFANs follows the equation

(1)

where n0 is the total areal density of PV-FANs and nex is the areal density of PV-FANs in the excited states.

• Interaction between PV-FAN and the cell membrane.

Once the charge polarization inside the ith PV-FAN, ΔρFANi , occurs upon photon absorption, it induces a dielectric response from the environment surrounding it and composed of ionic media, membrane lipids, proteins, etc.. Such dielectric response occurs typically on the order of ps to ns, significantly faster than the typical e-h recombination time in a PV-FAN as well as the cell membrane time constant of milliseconds, It can thus be practically considered as instantaneous. Let the induced charge redistribution be ΔρFANi . In the Fourier space, the induced charge redistribution can be conveniently expressed as , where is the dielectric function of the environment. The photon absorption induced charge redistribution in the PV-FAN and the redistribution in the environment together create a change in the local electrical field . The commonly employed model of physiological ionic medium response is the Debye model in which the induced charge redistribution and thus the associated electric field will decay on the scale of the Debye length away from the polarized PV-FAN. The typical Debye length for physiological conditions of ionic strength ∼100mM is on the order of 1 nm. If polarized PV-FANs are located close enough to a voltage-gated ion channel, a sufficiently large electric field induced by the PV-FANs in the neighborhood may exert sufficient force on the voltage sensitive segment of the channel to cause the channel to open. We note however that as the cell membrane and surrounding medium are highly inhomogeneous structures, local field effects may not be adequately accounted for in such bulk dielectric response description and molecular level simulations may be required to determine more accurately the interaction between individual PV-FAN and voltage-gated ion channel.

As a reasonable starting point, we can consider the interaction between PV-FANs and the cell membrane by accounting for the averaged membrane potential modulation caused by the total charge redistribution induced by a collection of excited PV-FANs distributed over a reasonably large area of the cell membrane. Let {N} be the set of PV-FANs that are photoexcited at a given instant of time. Within the dielectric response time (ps to ns) following the excitation of the set {N}, the ion current through the ion channels on the membrane can be neglected since the ion channel opening time constant is on the order of millisecond. Thus the net charge redistribution across the cell membrane is contributed only from charge polarization inside the PV-FANs and the net induced charge in the medium surrounding the FAN is zero when integrated over either the cytosol or ECM region. The net charge contributed from jth individual PV-FAN in the cytosolic region and in the ECM region can be written as,

where the integrations, respectively, are over the volume fraction of the jth FAN within the cytosol region and ECM region. The cell membrane potential modulation reached due to the charge redistribution induced by the set {N} of PV-FANs is thus,

where S and cM is the area and the areal capacitance of the membrane. For simplicity, we assume that all PV-FANs are in the same arrangement so that ΔqjCyt and ΔqjECM are independent of j. Let α = (ΔqjCyt-ΔqjECM)/(2e), so that ΔV0M = α·e·nex/cM where nex is the total number of the excited PV-FANs in collection {N} divided by the membrane area S, i.e. the areal density of photoexcited FANs.

Thus, from the point of view of a circuit model, the effect of PV-FANs on the cell membrane is essentially an AC current source transferring charge across the cell membrane to modulate membrane potential as schematically depicted in Figure 2 (see main text). With the excitation (de-excitation) of an individual FAN, effective charge α ⋅e is transferred from the ECM (cytosol) to the cytosol (ECM) side of the cell membrane. The total charge density transferred is α·e·nex. Thus the equivalent areal current density iFAN (t) can be written as:

(2)

where the iFAN (t) from the cytosolic to the ECM side of the cell membrane is defined as positive. The constant α depends on the geometry and location of the PV-FANs with respect to the cell membrane and depicts the efficiency of PV-FANs for modulating the membrane potential. As shown in Figure 1 (see main text), arrangement (a) is the most efficient arrangement as the PVFAN sits across the cell membrane with the separated hole and electron residing, respectively, in the cytosolic and ECM regions thus giving ΔqCyt ≈ e, ΔqECM ≈ −e and α ≈ 1 i.e. each excited PV-FAN transfers nearly one unit charge across the cell membrane to modulate the membrane potential.

Reference:

S1. Dylan P, Abbott LF. Theoretical neuroscience: computational and mathematical modeling of neural systems. MIT Press, Cambridge, MA 2001.

Back to Article Outline

References 

  1. Cozzoli PD, Pellegrino T, Manna L. Synthesis, properties and perspectives of hybrid nanocrystal structures. Chem Soc Rev. 2006;35:1195–1208
  2. Riess G. Micellization of block copolymers. Prog Polym Sci. 2003;28:1107–1170
  3. Sarikaya M, Tamerler C, Jen AKY, Schulten K, Baneyx F. Molecular biomimetics: nanotechnology through biology. Nat Mater. 2003;2:577–585
  4. Fu AH, Gu WW, Larabell C, Alivisatos AP. Semiconductor nanocrystals for biological imaging. Curr Opin Neurobiol. 2005;15:568–575
  5. Gao XH, Chan WCW, Nie SM. Quantum-dot nanocrystals for ultrasensitive biological labeling and multicolor optical encoding. J Biomed Opt. 2002;7:532–537
  6. Lee JH, Huh YM, Jun Y, Seo J, Cho EJ, Yoon HG, et al. Artificially engineered magnetic nanoparticles for ultra-sensitive molecular imaging. Nat Med. 2007;13:95–99
  7. O'Neal DP, Hirsch LR, Halas NJ, Payne JD, West JL. Photo-thermal tumor ablation in mice using near infrared-absorbing nanoparticles. Cancer Lett. 2004;209:171–176
  8. Jordan A, Scholz R, Wust P, Fahling H, Felix R. Magnetic fluid hyperthermia (MFH): cancer treatment with AC magnetic field induced excitation of biocompatible superparamagnetic nanoparticles. J Magn Magn Mater. 1999;201:413–419
  9. Lu S. Some studies of nanocrystal quantum dots on chemically functionalized substrates (semiconductors) for novel biological sensing. PhD Dissertation, University of Southern California, 2006. Chapter 7.
  10. Lu S, Madhukar A, Humayun M, inventors; University of Southern California. Functional abiotic nanosystems. United States Utility Patent Application US 2009/0088843 A1. 2009.
  11. Zemelman BV, Lee GA, Ng M, Miesenbock G. Selective photostimulation of genetically charged neurons. Neuron. 2002;33:15–22
  12. Melyan Z, Tarttelin EE, Bellingham J, Lucas RJ, Hankins MW. Addition of human melanopsin renders mammalian cells photoresponsive. Nature. 2005;433:741–745
  13. Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K. Millisecond-timescale, genetically targeted optical control of neural activity. Nat Neurosci. 2005;8:1263–1268
  14. Kramer RH, Chambers JJ, Trauner D. Photochemical tools for remote control of ion channels in excitable cells. Nat Chem Biol. 2005;1:360–365
  15. Arenkiel BR, Peca J, Davison IG, Feliciano C, Deisseroth K, Augustine GJ, et al. In vivo light-induced activation of neural circuitry in transgenic mice expressing channel rhodopsin-2. Neuron. 2007;54:205–218
  16. Higuchi M, Minoura N, Kinoshita T. Photoinduced structural and functional changes of an azobenzene-containing amphiphilic sequential polypeptide. Macromolecules. 1995;28:4981–4985
  17. Warman JM, De Haas MP, Paddon-Row MN, Cotsaris E, Hush NS, Oevering H, et al. Light-induced giant dipoles in simple model compounds for photosynthesis. Nature. 1986;320:615–616
  18. Yasutomi S, Morita T, Imanishi Y, Kimura S. A molecular photodiode system that can switch photocurrent direction. Science. 2004;304:1944–1947
  19. Winter J, Schmidt C, Korgel B. Optimization of quantum dot–nerve cell interfaces. Mat Res Soc Symp Proc. 2004;789:N6.2.1–N6.2.4
  20. Kuritz T, Lee I, Owens ET, Humayun M, Greenbaum E. Molecular photovoltaics and the photoactivation of mammalian cells. IEEE T Nanobiosci. 2005;4:196–200
  21. Dylan P, Abbott LF. Theoretical neuroscience: computational and mathematical modeling of neural systems. Cambridge, Massachusetts: MIT Press; 2001;
  22. Hille B. Ion channels of excitable membranes. Sunderland, Massachusetts: Sinauer Associates, Inc.; 2001;
  23. Fork RL. Laser stimulation of nerve cells in Aplysia. Science. 1971;171:907–908
  24. Hirase H, Nikolenko V, Goldberg JH, Yuste R. Multiphoton stimulation of neurons. J Neurobiol. 2002;51:237–247
  25. Lester HA, Nerbonne JM. Physiological and pharmacological manipulations with light flashes. Annu Rev Biophys Bioeng. 1982;11:151–175
  26. Denk W. Two-photon scanning photochemical microscopy: mapping ligand-gated ion channel distributions. Proc Natl Acad Sci USA. 1994;91:6629–6633
  27. Pearson GL, Brattain WH. History of semiconductor research. Proc IRE. 1955;43:1794–1806Bose JC, inventor. Detector for electrical disturbances. United States patent US 755840. 1904
  28. Piotrowiak P. Photoinduced electron transfer in molecular systems: recent developments. Chem Soc Rev. 1999;28:143–150
  29. Ratner MA, Madhukar A. Role of nuclear motions in electron and excitation transfer—importance of transfer-integral dependence upon nuclear coordinate. Chem Phys. 1978;30:201–215
  30. Barbara PF, Meyer TJ, Ratner MA. Contemporary issues in electron transfer research. J Phys Chem. 1996;100:13148–13168
  31. Gray HB, Winkler JR. Electron transfer in proteins. Ann Rev Biochem. 1996;65:537–561
  32. De Rege PJF, Williams SA, Therien MJ. Direct evaluation of electronic coupling mediated by hydrogen-bonds—implications for biological electron-transfer. Science. 1995;269:1409–1413
  33. Turro C, Chang CK, Leroi GE, Cukier RI, Nocera DG. Photoinduced electron-transfer mediated by a hydrogen-bonded interface. J Am Chem Soc. 1992;114:4013–4015
  34. Kumar S, Jones M, Lo SS, Scholes DG. Nanorod heterostructures showing photoinduced charge separation. Small. 2007;3:1633–1639
  35. Allen TM, Cullis PR. Drug delivery systems: entering the mainstream. Science. 2004;303:1818–1822
  36. Nel A, Xia T, Madler L, Li N. Toxic potential of materials at the nanolevel. Science. 2006;311:622–627

 Funded by Defense Advanced Research Projects Agency/Air Force Office of Scientific Research–funded Defense University Research Initiative in Nanotechnology (DURINT) program grant no. F49620-01-1-0474.

PII: S1549-9634(10)00011-0

doi:10.1016/j.nano.2010.01.004

Nanomedicine: Nanotechnology, Biology and Medicine
Volume 6, Issue 3 , Pages 409-418, June 2010

Article Tools

* Image Usage & Resolution