QUANTITATIVE CONSIDERATIONS: OPSIN BIOPHYSICAL PROPERTIES
Many factors control the efficacy of microbial opsins expressed in heterologous systems, not the least of which are cell biological properties ranging from effective transcription, translation, and folding, to proper membrane trafficking and targeting. However, biophysical properties will in many cases be similarly important limiting factors. The most basic of these is the efficiency of light absorption, expressed by the extinction coefficient (εmax). For rhodopsins, εmax is typically between 50,000 and 70,000 M−1 cm−1, less than that of chlorophyll (100,000 M−1 cm−1) but more than that of most other biological chromophores such as flavins (12,000 M−1 cm−1). Another important parameter is quantum efficiency (Φ, the fraction of absorbed photons that are efficacious in driving the relevant conformational change), which varies between 0.3 and 0.7 depending on the opsin species.
Also worthy of detailed consideration is the turnover time of the photocycle, a critical figure of merit both for native function and for neuroscience applications. For the most active transporters (HR and BR), the photocycle turnover time is ∼10–20 msec. Turnover is slower for the blue PR (80–100 msec) (Wang et al. 2003), limiting the development of blue PRs for neuroscience applications. However, even these values come from measurements at zero membrane voltage; the turnover for the fast pumps at a more physiologically negative voltage slows dramatically from 10–20 msec to 100–400 msec (Geibel et al. 2001). Thus, the more negative membrane voltage becomes, the slower the photocycle becomes, and more light is needed to achieve a given increment of hyperpolarization. Interestingly, this also means that by hyperpolarizing the membrane, the pump slows down its own photocycle, an effect that will be enhanced by changes in the electrochemical gradient of the transported ion. In general, however, pump direction is not actually inverted by physiologically achievable membrane potentials or ion gradients. Although such an inversion has been reported for PR, the driving force against the normal pump direction arose from extreme negative voltage and a strong pH gradient; under these conditions, the pump can become leaky as protons are forced onto an internal acceptor (Lörinczi et al. 2009). Generally, pump inversion per se is not expected, especially in physiological settings.
For ChRs the kinetic issues are different. Ion transport (i.e., current) is coupled to occupancy of the conducting state. This, in turn, is determined by light intensity and wavelength, extinction coefficient, and quantum efficiency, as well as by a new factor, the lifetime of the resulting conducting state, dictated by the kinetics of the P520–P480 transition (Fig. 3C). For temporally precise depolarization with a single flash, wild-type ChR2 is adequate. In contrast, because of its extended conducting-state lifetime, the H134R mutation of ChR2 noted above delivers approximately twofold higher photocurrents, although temporal precision is somewhat reduced as a result of moderately slower deactivation (Nagel et al. 2005; Gradinaru et al. 2007; Lin et al. 2009). The SFO variants of ChR2, with the greatly slowed deactivation described above (e.g., C128X) (Berndt et al. 2009), achieve chronic depolarization with even brief light delivery. Moreover, in principle, these variants can achieve the same maximal current magnitudes as wild-type ChR but at much lower light levels (Schoenenberger et al. 2009). For a 10-msec lifetime open state (Ritter et al. 2008), ChR2 would have to be excited ∼100 times per second for a half-saturating stable current. Assuming an absorption cross-section of 2 × 10−20 m2, this would require 5 × 1021 photons/m2-sec (i.e., 5 × 1012 photon/mm-msec or 2 mW/mm2) chronically. However, as a result of the extended-lifetime conducting states, C128T needs 100 times less light and the C128S mutant more than 1000 times less chronic light delivery for the same activity (Berndt et al. 2009).
Accelerating steps in the photocycle can also present certain advantages. For high-frequency neuronal firing, both fast-on and fast-off kinetics are important, requiring fast formation and decay of the P520 photocycle intermediate. Only the ChETA mutations (e.g., ChR2-E123T or E123A) can provide this unique functionality at present. Total current per flash is slightly smaller than for wild-type ChR2 because of the shorter open time, but as long as good expression is achieved, this can be compensated for by slightly more intense or longer flashes (e.g., 2 msec instead of 1 msec) (Gunaydin et al. 2010). Inactivation must also be considered, which is fast but weak in ChR1 and slow but strong in ChR2. Inactivation is more pronounced at high voltage (i.e., weakly negative or positive) and approaches 70% of peak current at pH 7.5. If inactivation of the current becomes problematic, ChR-E123T (Gunaydin et al. 2010), the ChIEF variant (Lin et al. 2009), or ChR2-E90Q (Ritter et al. 2008) might be suitable.
In addition to conducting-state lifetime, the unitary conductance of a channelrhodopsin will also control efficacy, and estimates for this important value range from 30 fS to ∼1 pS for ChR2 (Feldbauer et al. 2009; Lin et al. 2009). In the living alga Chlamydomonas, half-maximal current is carried by 1×106 ions; if there were 10,000 ChRs per eyespot with quantum efficiency of 0.67 (as with BR), this would correspond to 100 ions per ChR (Harz et al. 1992) and a unitary conductance of 0.3 pS. More recent estimates of opsin number (∼100,000 per eyespot) (Sineshchekov et al. 2002; Berthold et al. 2008) put the ions per ChR at 10, and conductance therefore at 30 fS. Recently, this value has been supported by application of noise analysis to purified ChR2 (40 fS) (Feldbauer et al. 2009), although nonstationary fluctuation analysis had resulted in somewhat higher values (>100 fS) (Lin et al. 2009). Despite the uncertainties inherent in all of these methods, the resulting estimates are still reasonably close and together provide useful boundary conditions.
The size of the unitary conductance is also relevant to the prospect of achieving two-photon (2P)–mediated control. Conventional 2P methods pose challenges in this regard, as small excitation volumes with raster-scanning 2P laser systems will typically not recruit sufficient numbers of small-conductance, rapidly deactivating molecules to, for example, drive spiking. Two main approaches are being developed to address this issue, and 2P optogenetic control can now be achieved. First, control of light delivery can be adapted to individual imaged cells to provide spatiotemporally suitable illumination patterns that will recruit sufficient conductance within a single cell and within a temporal window set by current deactivation properties (Rickgauer and Tank 2009). Second, engineered opsins (e.g., SFOs and related mutants in which the deactivation kinetics are altered) can facilitate recruitment of threshold levels of conductance within a given temporal window. In this way, 2P-based recruitment of sparsely distributed cell populations within living tissue can become practical, pointing to the value of continuing to probe the structure–function relationships of the microbial opsins.
The action spectrum is also important, and developing absorbance at longer wavelengths than ChR2 (λmax = 470 nm) is particularly useful for depth penetration, safety, and combinatorial experiments. ChR1 is slightly red-shifted but does not express well and is pH-dependent. VChR1 (Zhang et al. 2008) absorbs at markedly red-shifted wavelengths (λmax = 540 nm), and spiking can be driven even at 589 nm in hippocampal neurons, but expression levels are significantly lower than that of ChR2 in most host cells. The slower decay of the open state (Zhang et al. 2008) partially compensates, but improvement of VChR1 expression will be very important, and a strongly expressing channelrhodopsin that absorbs beyond 550 nm will outperform and be preferable to ChR2 in the future. Indeed, as with VChR1, the channelrhodopsin variants have diverse useful properties but also corresponding limitations. For example, the SFO mutants have reduced dynamic range. Moreover, because of SFO hypersensitivity and the long lifetime of intermediates, in vitro experiments should be maintained at low light to avoid unwanted activation by ambient light. When imaging is required, illumination should be restricted to nonactivating wavelengths (>560 nm). Complete deactivation of the SFO prior to experiments with a long pulse of 590-nm light will ensure that any baseline activity is eliminated and will enable full activation of the SFO with 470-nm light at the desired time
การพิจารณาเชิงปริมาณ: OPSIN BIOPHYSICAL คุณสมบัติหลายปัจจัยควบคุมประสิทธิภาพของ opsins จุลินทรีย์ที่แสดงในระบบ heterologous ไม่น้อยซึ่งมีคุณสมบัติทางชีวภาพเซลล์ตั้งแต่ transcription มีประสิทธิภาพ การแปล และการ พับ การเมมเบรนที่เหมาะสมในการค้า และการกำหนดเป้าหมาย อย่างไรก็ตาม biophysical คุณสมบัติในหลายกรณีจะรับปัจจัยจำกัดที่สำคัญ พื้นฐานที่สุดของเหล่านี้มีประสิทธิภาพการดูดซึมแสง แสดง โดยสัมประสิทธิ์การสูญพันธุ์ (εmax) สำหรับ rhodopsins, εmax ได้ปกติระหว่าง 50000 และ 70000 M−1 cm−1 น้อยกว่าของคลอโรฟิลล์ (100000 M−1 cm−1) แต่มากกว่านั้นส่วนใหญ่อื่น ๆ chromophores ชีวภาพเช่น flavins (12000 M−1 cm−1) พารามิเตอร์ที่สำคัญอื่นประสิทธิภาพควอนตัม (Φ เศษ photons ดูดซึมที่บ็อชในการขับขี่ที่เกี่ยวข้องเปลี่ยนแปลง conformational), ซึ่งแตกต่างกันระหว่าง 0.3 และ 0.7 ตามพันธุ์ opsin ได้Also worthy of detailed consideration is the turnover time of the photocycle, a critical figure of merit both for native function and for neuroscience applications. For the most active transporters (HR and BR), the photocycle turnover time is ∼10–20 msec. Turnover is slower for the blue PR (80–100 msec) (Wang et al. 2003), limiting the development of blue PRs for neuroscience applications. However, even these values come from measurements at zero membrane voltage; the turnover for the fast pumps at a more physiologically negative voltage slows dramatically from 10–20 msec to 100–400 msec (Geibel et al. 2001). Thus, the more negative membrane voltage becomes, the slower the photocycle becomes, and more light is needed to achieve a given increment of hyperpolarization. Interestingly, this also means that by hyperpolarizing the membrane, the pump slows down its own photocycle, an effect that will be enhanced by changes in the electrochemical gradient of the transported ion. In general, however, pump direction is not actually inverted by physiologically achievable membrane potentials or ion gradients. Although such an inversion has been reported for PR, the driving force against the normal pump direction arose from extreme negative voltage and a strong pH gradient; under these conditions, the pump can become leaky as protons are forced onto an internal acceptor (Lörinczi et al. 2009). Generally, pump inversion per se is not expected, especially in physiological settings.For ChRs the kinetic issues are different. Ion transport (i.e., current) is coupled to occupancy of the conducting state. This, in turn, is determined by light intensity and wavelength, extinction coefficient, and quantum efficiency, as well as by a new factor, the lifetime of the resulting conducting state, dictated by the kinetics of the P520–P480 transition (Fig. 3C). For temporally precise depolarization with a single flash, wild-type ChR2 is adequate. In contrast, because of its extended conducting-state lifetime, the H134R mutation of ChR2 noted above delivers approximately twofold higher photocurrents, although temporal precision is somewhat reduced as a result of moderately slower deactivation (Nagel et al. 2005; Gradinaru et al. 2007; Lin et al. 2009). The SFO variants of ChR2, with the greatly slowed deactivation described above (e.g., C128X) (Berndt et al. 2009), achieve chronic depolarization with even brief light delivery. Moreover, in principle, these variants can achieve the same maximal current magnitudes as wild-type ChR but at much lower light levels (Schoenenberger et al. 2009). For a 10-msec lifetime open state (Ritter et al. 2008), ChR2 would have to be excited ∼100 times per second for a half-saturating stable current. Assuming an absorption cross-section of 2 × 10−20 m2, this would require 5 × 1021 photons/m2-sec (i.e., 5 × 1012 photon/mm-msec or 2 mW/mm2) chronically. However, as a result of the extended-lifetime conducting states, C128T needs 100 times less light and the C128S mutant more than 1000 times less chronic light delivery for the same activity (Berndt et al. 2009).Accelerating steps in the photocycle can also present certain advantages. For high-frequency neuronal firing, both fast-on and fast-off kinetics are important, requiring fast formation and decay of the P520 photocycle intermediate. Only the ChETA mutations (e.g., ChR2-E123T or E123A) can provide this unique functionality at present. Total current per flash is slightly smaller than for wild-type ChR2 because of the shorter open time, but as long as good expression is achieved, this can be compensated for by slightly more intense or longer flashes (e.g., 2 msec instead of 1 msec) (Gunaydin et al. 2010). Inactivation must also be considered, which is fast but weak in ChR1 and slow but strong in ChR2. Inactivation is more pronounced at high voltage (i.e., weakly negative or positive) and approaches 70% of peak current at pH 7.5. If inactivation of the current becomes problematic, ChR-E123T (Gunaydin et al. 2010), the ChIEF variant (Lin et al. 2009), or ChR2-E90Q (Ritter et al. 2008) might be suitable.In addition to conducting-state lifetime, the unitary conductance of a channelrhodopsin will also control efficacy, and estimates for this important value range from 30 fS to ∼1 pS for ChR2 (Feldbauer et al. 2009; Lin et al. 2009). In the living alga Chlamydomonas, half-maximal current is carried by 1×106 ions; if there were 10,000 ChRs per eyespot with quantum efficiency of 0.67 (as with BR), this would correspond to 100 ions per ChR (Harz et al. 1992) and a unitary conductance of 0.3 pS. More recent estimates of opsin number (∼100,000 per eyespot) (Sineshchekov et al. 2002; Berthold et al. 2008) put the ions per ChR at 10, and conductance therefore at 30 fS. Recently, this value has been supported by application of noise analysis to purified ChR2 (40 fS) (Feldbauer et al. 2009), although nonstationary fluctuation analysis had resulted in somewhat higher values (>100 fS) (Lin et al. 2009). Despite the uncertainties inherent in all of these methods, the resulting estimates are still reasonably close and together provide useful boundary conditions.
The size of the unitary conductance is also relevant to the prospect of achieving two-photon (2P)–mediated control. Conventional 2P methods pose challenges in this regard, as small excitation volumes with raster-scanning 2P laser systems will typically not recruit sufficient numbers of small-conductance, rapidly deactivating molecules to, for example, drive spiking. Two main approaches are being developed to address this issue, and 2P optogenetic control can now be achieved. First, control of light delivery can be adapted to individual imaged cells to provide spatiotemporally suitable illumination patterns that will recruit sufficient conductance within a single cell and within a temporal window set by current deactivation properties (Rickgauer and Tank 2009). Second, engineered opsins (e.g., SFOs and related mutants in which the deactivation kinetics are altered) can facilitate recruitment of threshold levels of conductance within a given temporal window. In this way, 2P-based recruitment of sparsely distributed cell populations within living tissue can become practical, pointing to the value of continuing to probe the structure–function relationships of the microbial opsins.
The action spectrum is also important, and developing absorbance at longer wavelengths than ChR2 (λmax = 470 nm) is particularly useful for depth penetration, safety, and combinatorial experiments. ChR1 is slightly red-shifted but does not express well and is pH-dependent. VChR1 (Zhang et al. 2008) absorbs at markedly red-shifted wavelengths (λmax = 540 nm), and spiking can be driven even at 589 nm in hippocampal neurons, but expression levels are significantly lower than that of ChR2 in most host cells. The slower decay of the open state (Zhang et al. 2008) partially compensates, but improvement of VChR1 expression will be very important, and a strongly expressing channelrhodopsin that absorbs beyond 550 nm will outperform and be preferable to ChR2 in the future. Indeed, as with VChR1, the channelrhodopsin variants have diverse useful properties but also corresponding limitations. For example, the SFO mutants have reduced dynamic range. Moreover, because of SFO hypersensitivity and the long lifetime of intermediates, in vitro experiments should be maintained at low light to avoid unwanted activation by ambient light. When imaging is required, illumination should be restricted to nonactivating wavelengths (>560 nm). Complete deactivation of the SFO prior to experiments with a long pulse of 590-nm light will ensure that any baseline activity is eliminated and will enable full activation of the SFO with 470-nm light at the desired time
การแปล กรุณารอสักครู่..
