Dynamic Imaging of Molecular Assemblies in Live Cells Based on Nanoparticle Plasmon Resonance Coupling
基于納米粒子等離子體共振耦合的細胞中分子組裝的動態(tài)影像

Jesse Aaron, Kort Travis,‡ Nathan Harrison, and Konstantin Sokolov
Department of Biomedical Engineering, Department of Physics, UniVersity of Texas at Austin, Austin, Texas 78712, and Department of Imaging Physics, UT MD Anderson Cancer Center, Houston, Texas 77030
Received June 8, 2009; Revised Manuscript Received July 21, 2009

ABSTRACT　摘要
     We used molecular-specific gold nanoparticles to monitor epidermal growth factor receptors (EGFR) in live A431 cells over time. Dark-field hyperspectral imaging, electron microscopy, and electrodynamic modeling were used to correlate optical properties of EGFR-bound plasmonic nanoparticles with receptor regulation state. We showed that receptor trafficking resulted in a progressive red shift of greater than 100 nm in the nanoparticle plasmon resonance wavelength over a time period of 60 min. Furthermore, we demonstrated that changes in peak scattering wavelengths of gold nanoparticles from 546±15 to 574± 20, and to 597±44 nm are associated with EGFR trafficking from the cell membrane, to early endosomes, and to late endosomes/multivesicular bodies, respectively. Finally, we used the changes in scattering spectra of EGFR -bound nanoparticles and a straightforward statistical analysis of RGB-channel color images of labeled cells to create near real-time maps of EGFR regulatory states in living cells.
     在A431活細胞中我們用含有特定的金納米粒子的分子去監(jiān)控表皮生長因子受體(EGFR)。用暗場高光譜成像、電子顯微、電動模擬等手段將已綁定了等離振子納米顆粒的EGFR的光學(xué)屬性與受體的調(diào)控狀態(tài)聯(lián)系起來。在納米粒子等離振子共振超過60分鐘的條件下，我們顯示了受體的運輸導(dǎo)致了大于100納米的紅移。再者,我們證實，金納米粒子巔峰散射波長的變化(從546±15到574±20，到597±44 nm)與EGFR的運輸有關(guān)（分別從細胞膜到早期胞內(nèi)體，到晚期胞內(nèi)體/多泡體）。最后, 利用綁定納米粒子的EGFR的散射光譜的變化以及標(biāo)定細胞的三基色圖像的一個直觀數(shù)據(jù)分析，我們制作了活細胞中接近實時的EGFR調(diào)控狀態(tài)的圖像。

     Detecting and monitoring the vast number of biomolecular interactions in the cell is a central effort in biology, as these interactions largely govern the behavior of nearly all cell types. Imaging methods are an indispensable approach for measuring the spatiotemporal characteristics of protein assemblies in intact cells. In this work, we expand the application of nanoparticle plasmon resonance coupling (NPRC) and demonstrate a novel, generalized method for imaging and characterizing molecular assemblies at the nanometer length-scale in living cells.
     生物學(xué)關(guān)鍵的任務(wù)就是探測和監(jiān)測細胞中大數(shù)量的生物分子間的相互作用，因為這些相互作用在很大程度上控制著幾乎所有細胞類型的行為。成像法是必不可少的測量完整細胞蛋白組裝時空特性的方法。在本次工作中,我們將展開納米粒子等離子體共振耦合(NPRC)的應(yīng)用并展示一種新穎而又普遍的方法，利用這種方法我們可以在活細胞中以納米級別的量度來描述細胞的裝配，并為其成像。

     Over the past decades, fluorescence resonance energy transfer (FRET) has allowed many investigators to elucidate important functional associations between pairs of proteins at submicroscopic resolution, all without destroying the cell.1 More recently, techniques such as image correlation microscopy (ICM) and its variants have been widely used to characterize larger protein assemblies and clusters, including EGFR, in situ.2,3 Although all of these techniques are eminently useful, FRET is typically limited to detecting two very closely separated (5 nm) molecules of different types.1,4 While the various ICM methods can be used to evaluate associations and clustering between many molecules on the submicrometer scale, this additional information is an ensemble average and it does not directly reveal the distribution of cluster sizes or any additional information related to nanometer-scale organization of biomolecules forming the clusters.2,3 Furthermore, ICM methods are in general highly sensitive to background interference, and the resolution of the imaging system is often not fully utilized. In addition, both FRET and ICM can be critically limited by photobleaching. These limitations are avoided in electron microscopy (EM), and in particular, the immuno-gold method to image assemblies of biomolecules at the nanometer length scale.5 However, this comes at the expense of the lack of real-time, live-cell capability and of additional tedious sample preparation. For all of these reasons, new physical methods are required that extend and/or complement currently available capabilities for quantitative monitoring of molecular clusters with high spatial and temporal resolution in living cells.
     在過去的幾十年里,熒光共振能量轉(zhuǎn)移技術(shù)(FRET)的應(yīng)用使得許多研究人員能夠在亞微觀無損傷的條件下闡明細胞重要功能之間的聯(lián)系1。近來,圖像顯微技術(shù)(ICM)及其變體已經(jīng)被廣泛地用來描述大蛋白分子在原位的裝配及群集,包括EGFR。雖然所有的這些技術(shù)都是非常有用的，但是FRET技術(shù)只適用于探測不同類型的兩個非常近的(5 nm)單個分子1,4。多樣的ICM技術(shù)可以在亞顯微計量級別下用來測量眾多分子的關(guān)系和群集，而FRET附加信息只是一種統(tǒng)計平均值，不直接揭示集群的分布大小或者說任何與形成群集的毫米級別的生物分子有關(guān)的附加信息。而且ICM技術(shù)對背景干擾是高度敏感的，圖像系統(tǒng)的分辨率經(jīng)常沒有被充分利用。另外，F(xiàn)RET和ICM技術(shù)都被光漂白嚴(yán)重限制。在電子顯微鏡中這些限制因素都可以避免，特別是在納米級別免疫金法為生物分子裝配成像的條件下5。然而達到這種要求要付出一些代價，缺少了實時、活細胞性能，并需要繁冗的樣品制備過程。鑒于上述所有原因，在活細胞中用高位立體瞬時分析方法去定量監(jiān)視細胞群集時，就要求有新的物理解決方法來發(fā)展并補充現(xiàn)有可以利用的技術(shù)。

     Recently, plasmonic-resonant nanoparticles have been explored as molecular-specific probes for highly sensitive detection6-16 and for photothermal damage of cells.17-20 When two or more plasmonic nanoparticles are in close proximity they exhibit the effect of nanoparticle plasmon resonance coupling (NPRC) that manifests as a spectral shift in the optical cross sections of the metal nanoparticle assemblies when compared to those observed from the isolated nanoparticles.21,22 This spectral shift is strongly dependent on the interparticle distance,23 and the sensitivity can be conveniently adjusted by modifying the particle size. NPRC is significant for particle center-to-center distances of less than about three times the particle radius, thus providing a useful range of detectable interaction distances out to tens of nanometers. Because of this behavior, NPRC has recently been demonstrated as an attractive analog to FRET, as evidenced by its use in the detection of DNA-DNA,24-26 DNA-protein,27 and protein-protein binary interactions.28 These investigations showed that NPRC is not limited by photobleaching and that it can extend the range of detectable distances by more than an order of magnitude in comparison to FRET.
     最近,等離子共振的納米粒子已經(jīng)被開發(fā)成特定的分子探針，用在高靈敏度探測6-16和細胞的光熱破壞上17-20。相比觀察單個納米粒子，當(dāng)兩個等離振子納米粒子間距非常近時，它們就會顯示納米粒子等離振子共振耦合效應(yīng)(NPRC)，表現(xiàn)為在金屬納米粒子裝配光學(xué)橫截面上的一個光譜的漂移21,22。光譜漂移很大程度上取決于粒子間的距離，靈敏度可以通過改變粒子的大小來方便地調(diào)整。對中心間距小于3倍粒子半徑的粒子群，NPRC是非常重要的，為上萬計的納米提供一個有效的可探測交互距離的范圍。正是因為這個效應(yīng)，用NPRC來演示一個有吸引力的FRET模擬，作為探測DNA-DNA,24-26 DNA-蛋白質(zhì),27 和 蛋白質(zhì)-蛋白質(zhì)的二態(tài)交互作用28的有效證明。這些研究顯示NPRC不受光漂白的影響，而且相比FRET其擴展探測范圍不在同一數(shù)量級上。

     We have shown previously that NPRC provides a powerful cancer diagnostic method by facilitating the detection of growth factor receptor clustering in cancerous tissue.6,29 In the present work, we expand the applicability of NPRC to the study of molecular clustering and regulation mechanisms in living cells. We have developed a novel computational
framework to gain nanometer-scale information from NPRC spectra and, in addition, a practical image-analysis method to infer growth factor receptor regulation state.
     我們之前已經(jīng)證明NPRC提供了一種強有力的癌癥診斷方法,通過更便利的檢測聚集在癌變組織中的生長因子6,29�，F(xiàn)在,我們擴大了NPRC的適用范圍，研究活細胞內(nèi)分子的聚集和調(diào)節(jié)機制。我們已經(jīng)開發(fā)了一種新型的計算框架來從NPRC光譜中獲得信息，除此之外，用一個實用的圖像分析方法來推斷生長因子受體調(diào)節(jié)狀態(tài)。

     We monitor the trafficking mechanisms of the epidermal growth factor receptor (EGFR), a receptor tyrosine kinase (RTK) that controls some of the most fundamental cellular processes including DNA replication and cell division.30 EGFR is also an important cancer biomarker with regulatory pathways that have profound implications for the development of new cancer therapeutics.31,32 The EGFR internalization process is a key regulatory pathway that determines cell behavior. It is well known that membrane-bound EGFR elicits growth-promoting signals upon ligand binding. However, it has recently been shown that signaling can continue even after internalization of EGFR in early endosomes and that the signaling activity finally ceases after entry into lysosomes.31 Further, processes involving EGFR recycling and nuclear translocation add to the complexity of EGFR’s role.30,33 Thus, it is vital that an effective EGFR imaging strategy not only involves detection of the protein, but also includes the nanometer-scale details of its organization, aggregation, and sequestration within cellular compartments. In this study, we demonstrate that NPRC is a convenient method that facilitates the retrieval of this information using dark-field color imaging of live cells labeled with EGFR specific gold nanoparticles.
     我們觀察了表皮生長因子受體（EGFR）的交易機制，一個酪氨酸激酶（RTK）受體控制了一些最基本的細胞過程,包括DNA復(fù)制和細胞分裂30。在治療手段中EGFR也是重要的腫瘤標(biāo)志物,對開發(fā)新型癌癥治療法有深遠的影響31,32。EGFR內(nèi)在過程是一個關(guān)鍵的調(diào)節(jié)方式,決定了細胞的行為。眾所周知,綁定EGFR的膜能引起配位基綁定促進信號的增長。然而,最近發(fā)現(xiàn)，即使在早期的胞內(nèi)體EGFR內(nèi)化之后信號可以繼續(xù)發(fā)放，最終在溶酶體進入后才停止31。 涉及EGFR的再循環(huán)和細胞核的轉(zhuǎn)移過程都增加了EGFR的復(fù)雜性30,33。因此問題的關(guān)鍵就明顯了，有效的EGFR不僅涉及蛋白質(zhì)的檢測,也包括細胞局部區(qū)域內(nèi)的組織、聚合與分離的納米級別細節(jié)。在本研究中,我們證實NPRC是一個方便的方法,有利于檢索暗場條件下貼有EGFR特定金納米�；罴毎�的顏色圖像信息。

Synthesis of Molecular Specific Gold Nanoparticles 特定金納米粒子分子的合成
     Near-spherical gold nanoparticles (with average diameter of 25 nm) were synthesized using the method described by Frens, whereby chloroauric acid (HAuCl4) is chemically reduced using sodium citrate.34 By altering the Au3+ to citrate ratio, the size of the resulting nanoparticles can be controlled. For the size of particles used in the present work, this protocol produces gold colloid with a particle concentration of approximately 6 × 1011 particles per milliliter.
     近球形的金納米粒子(平均直徑25nm)使用了Frens描述的方法來合成，用檸檬酸鈉的化學(xué)方法減少氯金酸34。通過改變Au3 +與檸檬酸鹽的比率、可以控制產(chǎn)生納米粒子的大小。此次工作中的粒子的大小，擬定金膠質(zhì)的粒子聚集濃度大約為每毫升6×1011顆粒。

     To conjugate monoclonal antibodies to the surfaces of the gold nanoparticles, a heterobifunctional linker was utilized, as described in ref 35. Briefly, anti-EGFR monoclonal antibodies (clone 29.1.1 Sigma) were purified from ascites fluid using a 100 kDa MWCO centrifugal filter and resuspended in 40 mM HEPES solution (pH 7.5) at a concentration of 1 mg/mL. Sodium periodate (NaIO4) was added to the antibody solution at a final concentration of 10 mM for 30 min under mild agitation and protected from the light. Sodium periodate oxidizes the hydroxyl moieties located on
the glycosylated Fc region of antibodies to aldehyde groups. Then, the solution was diluted 50-fold using 1× PBS, and 200 μM of a heterobifunctional linker was added (Sensopath Technologies, Inc.).
     利用異型雙功能連接器把單克隆抗體結(jié)合到黃金納米粒子表面,參考35。簡單地說, 抗EGFR的單克隆抗體(克隆29.1.1σ)要用一個100 kDa的MWCO離心式過濾器來從腹水中剔除，并在40 mM濃度為1 mg/mL的HEPES溶液(pH 7.5)中懸浮處理。添加高碘酸鈉(NaIO4)到抗體溶液中，并最終在弱風(fēng)潮,防止光線條件下保存30分鐘。濃度在最后10毫米為30分鐘,。高碘酸鈉將位于抗體糖化基區(qū)域的羥基氧化成醛基。然后,用1× PBS將溶液稀釋50倍，然后添加200μM的異型雙功能連接器(Sensopath科技有限公司)。

     The linker consists of a PEG chain containing two sulfhydril groups on one end, and a hydrazide moiety on the other. The newly formed aldehydes on theantibodies condense with the hydrazide group on the linker, leaving the thiols available for binding to the gold surface. Excess unreacted linker was removed via centrifugal filtration as described above, and the antibody-linker solution was stored at 1 mg/mL in 1× PBS for up to 4 weeks. Prior to cell labeling, the antibody-linker solution was first diluted 100-fold in 40 mM HEPES solution to which an equal volume of 25 nm gold nanoparticles was added. The two components were allowed to mix under mild agitation for 20 min, followed by addition of 5 kDa monofunctional thiolated polyethylene glycol (mPEG-SH, Nektar) to 10-6
M final concentration. Five minutes after addition of mPEG thiol, antibody-nanoparticle conjugates were spun down at 3400 × g for 30 min, the supernatant removed, and the conjugates resuspended in phenolphthalein-free DMEM cell culture media (Gibco) containing 10% FBS.
     這個連接器有一個PEG鏈和一個酰肼組成，PEG鏈包含兩個在一端的sulfhydril群組。這個新合成的醛在連接器上與酰肼群組被濃縮,殘余的硫醇可用在綁定黃金表面上。多余的未反應(yīng)的連接器如上所述通過離心過濾被移除, antibody-linker（抗體連接器）溶液儲存在1mg/ mL, 1×PBS中4周時間。在這之前，要標(biāo)記細胞, 抗體連接器溶液首先在40毫米的HEPES中稀釋100倍, 相當(dāng)于添加等體積25 nm金納米粒子。這兩個部件混在一起柔和攪拌20分鐘,緊隨其后的是增加5 kDa單功能硫醇鹽的聚乙二醇(mPEG-SH, Nektar)，達到10-6 M的最終濃度。在添加mPEG硫醇五分鐘后,抗體納米粒子連接器在3400 × g的條件下減速旋轉(zhuǎn)30分鐘，然后除去上層清夜。將連接器在含有10% FBS的無酚酞細胞培養(yǎng)基中再次懸浮處理。

     Gold particles of 25 nm diameter were used because they provide high contrast in detection of EGFR overexpressing A431 cells.29 The particles were conjugated with antibodies that do not block receptor activation by EGF.36 We have demonstrated in previous studies that nanoparticles conjugated with either monofunctional thiolated polyethylene glycol (mPEG-SH) or a nonspecific antibody do not interact with EGFR expressing cells. Further, specificity of EGFR binding by the gold conjugates was shown by displacement and competition assays in the presence of free anti-EGFR antibodies.29
     使用直徑25 nm金粒子是因為在探測EGFR表譯A431 細胞29時可以提供很高的對比度。結(jié)合了抗體的粒子，不會通過EGF36來阻止受體反應(yīng)。我們在先前的研究已經(jīng)證明, 不管是單功能硫醇鹽的聚乙二醇(mPEG-SH)還是非特異性抗體與納米離子結(jié)合都不會影響EGFR表譯細胞。與單功能硫醇鹽的共軛粒子聚乙二醇(mPEG-SH)或非特異性抗體不與表皮生長因子受體表達細胞。而且在無抵抗EGFR抗體條件下檢測出的取代和競爭現(xiàn)象顯示了綁定金連接物的EGFR的特性29。

     Dynamic Imaging of Live Cells. A431 keratinocytes were cultured on 22 mm square glass coverslips using DMEM/ F-12 (50/50) growth medium, supplemented with 5% FBS (Gibco) in a 37 C and 5% CO2 environment. For live imaging experiments, A431 cells were seeded onto optical imaging flow chambers (ibidi μslide chambers, Integrated BioDiagnostics) and allowed to adhere overnight. Then, the chambers were separated into two samples: chambers where cells were exposed to 10 μM AG1478 (Calbiochem), a EGFR phosphorylation inhibitor,37 and control chambers where cells were not exposed to the inhibitor. The samples were then kept for another 24 h under serum-starvation conditions. Chambers were placed under the microscope and connected to a peristaltic pump supplying either DMEM (containing 10% FBS) media or gold nanoparticle conjugates in complete DMEM at low flow rates (typically 1 mL/min). Temperature control of solutions was accomplished through the use of an in-line resistive heater (Harvard Apparatus) to 37 C before being introduced into the chamber.
     活細胞的動態(tài)影像。A431角質(zhì)化細胞培養(yǎng)在22毫米平方的蓋盒內(nèi)，F(xiàn)-12(50 / 50)細胞培養(yǎng)基增長介質(zhì)在37 C 、 5% CO2條件下輔以5%的FBS來培養(yǎng)。對于活影像實驗,將A431細胞植入到光學(xué)成像流動腔中(ibidi μslide chambers, Integrated BioDiagnostics)，讓其附著過夜。然后將腔分入兩個樣品中-腔和控制腔。腔中細胞暴露在10 μM AG1478 (Calbiochem 一種新的磷酸化抑制劑)下，控制腔內(nèi)的細胞不暴露在抑制劑下。然后再將樣品在血清饑餓條件下保存24小時。把腔放在顯微鏡下，連接到蠕動泵上以低流速(通常 1 mL/min)供給DMEM (含10% FBS)介質(zhì)或者供給純DMEM金納米連接物。用內(nèi)置電阻加熱器來控制溫度，在引進腔之前，先將溫度提升到37度。

     Samples were continuously exposed to conjugates at approximately 1012 particles/mL under low flow conditions for 45-60 min during time-lapse imaging. Cell samples were imaged using a Leica DM6000 upright microscope, a 100 W halogen light source, and an oil-immersion darkfield condenser (Leica) with an N.A. of 1.2. A long-working distance (1.2 mm), 63× objective (Leica) with 0.75 NA was used for collection of scattering signal and the images were recorded by a high-sensitivity 12-bit color-mosaic CCD camera (SPOT Pursuit XS, Diagnostic Instruments). To ensure cell viability over the duration of live imaging experiment, we carried out Calcein AM assay (Molecular Probes) after 2 h of cell exposure to anti-EGFR nanoparticle conjugates. Results (figure available in Supporting Information) showed that there are no statistically significant changes in cell viability between cells treated with gold nanoparticles and unlabeled controls.
     在延時影像過程中，樣品以大約1012粒子/ mL的低流量在連接器中持續(xù)暴露45-60分鐘。用萊卡DM6000直立顯微鏡像，輔以100瓦鹵素光源和數(shù)值孔徑為1.2的油浸暗場鏡頭(Leica) 采集圖。用一個長工作距離(1.2 mm)、 0.75 NA、63倍的物鏡(Leica)來采集散射的信號，用高敏感度的12字節(jié)的色拼CCD照相機(SPOT Pursuit XS, Diagnostic Instruments)來記錄圖像。為確保整個活動影像實驗的細胞活力，在細胞暴露在抗EGFR納米顆粒連接物中2小時后，我們執(zhí)行鈣黃綠素AM實驗(分子探針)。結(jié)果顯示金納米粒子標(biāo)定的細胞與未標(biāo)定的細胞的穩(wěn)健性在統(tǒng)計上沒有顯著的改變。

Figure 1. Dynamic imaging of live cells. (A) A schematic of EGFR trafficking upon ligand binding. This process was monitored in real time in live cells labeled with anti-EGFR gold nanoparticles under the microscope. (B) Live A431 cells and (C) live cells pretreated with 10 μM AG1478, which reduces endocytosis. From left to right, images show color and intensity changes in the gold nanoparticle optical signal at 0, 15, 30, and 50 min after initial exposure to nanoparticles. Dark-field images of untreated cells (B) display changes in color from blue to yellow-orange that are not present in treated cells (C). The arrow in panel B, far right, indicates the presence of a filopodium emanating from an adjacent cell. Images were acquired using transmitted dark-field illumination and a longworking distance 63×, 0.75 NA objective.
圖一活細胞動態(tài)影像。(A)依據(jù)配合基綁定的EGFR交易圖。在標(biāo)定抗EGFR金納米粒子活細胞中這個進程是在顯微鏡下實時監(jiān)控的。(B)活A(yù)431細胞和(C)用 10 μM AG1478（降低內(nèi)吞作用）預(yù)處理的活細胞。從左到右分別為從初始暴露到納米粒子到0, 15, 30和50 分鐘后圖像顯示金納米粒子光學(xué)信號的顏色和強度的變化。未經(jīng)處理的暗場圖像的色彩從藍到黃橘色的變化，這是經(jīng)過處理的細胞顯示不出來的。圖B中最右邊的箭頭顯示鄰近細胞放射出來的絲線。圖像的獲取要求使用透射暗場照明、長工作距離、63倍、0.75 NA的物鏡。

     Dynamic imaging of live cells labeled with anti-EGFR nanoparticles exhibited a progressive color change from green to yellow, and finally, to orange-red (Figure 1B). The color changes are well-correlated in time with known trafficking dynamics of EGFR (Figure 1A).33,38 Initially, EGFR molecules dimerize and aggregate in the plasma membrane, followed by endocytosis into early endosomes. These early endosomes can then either recycle to the cell surface or proceed through formation of late endosomes and MVBs within 20-60 min.33,38 The effect of continuous color changes was disrupted by a potent EGFR inhibitor, AG1478, which interferes with EGFR trans-phosphorylation and internalization (Figure 1C).37 These data indicate that the observed color changes are associated with EGFR reorganization within the cell both as a result of ligand binding, and also due to intracellular trafficking in vesicles. In images taken from untreated samples, the relative intensity of the red channel increases from 27 to nearly 34% (Figure 1B) whereas for samples treated with AG1478 inhibitor, thisvalue increases only from 28 to 30% (Figure 1C).
     標(biāo)定抗EGFR納米粒子的動態(tài)活細胞影像顯示了連續(xù)的顏色變化，從綠到黃，最終到橘紅(圖 1B)。這些顏色的改變同步關(guān)聯(lián)著已知的EGFR交易動力。起初EGFR的二聚和聚合是在質(zhì)膜中，隨后內(nèi)吞成為早期的胞內(nèi)體中。這些早期的胞內(nèi)體可以再循環(huán)到細胞表面或者在20-60分鐘內(nèi)繼續(xù)行進形成后期胞內(nèi)體或MVB。AG1478（一種強有力的EGFR抑制劑）將打斷連續(xù)顏色變化這種效果，因為它干擾EGFR的橫穿磷化作用和內(nèi)化作用(圖 1C)37。這些數(shù)據(jù)表明已測的顏色變化與EFGR在細胞內(nèi)的重組有關(guān)系，是配合基綁定造成的，也是由于細胞內(nèi)的囊泡交易。在未經(jīng)處理樣品的圖像里， 紅色的相對強度從27%提到了34%(圖 1B)，相反，對用AG1478抑制劑處理的樣品這個數(shù)值的增長就只是28 % 到30%了(圖 1C)。

     Cell Labeling at Different Temperatures. In order to establish a relationship between the scattering behavior of EGFR-bound plasmonic nanoparticles and the dynamics of EGFR regulation in live cells, we carried out labeling at 4, 25, and 37 C. Temperature control allows the activationand trafficking mechanisms of EGFR to be arrested at critical points. EGFR internalization is inhibited in cells at 4 C; internalization proceeds to formation of early endosomes at 25 C; and finally, complete EGFR regulation proceeds through the formation of multivesicular bodies (MVB) and late endosomes within an hour at 37 C.33,38,39 For the temperature-based experiments, cells were seeded onto coverslips at between 50 000-100 000 cells/mL and left to adhere overnight. Then, cells were washed in 1× PBS, and 1 mL of gold nanoparticle bioconjugates in DMEM supplemented with 5% FBS was added. Coverslip-adherent cell samples containing nanoparticles were placed in 4, 25, and 37 C environments and allowed to interact for 60 min. Nanoparticles were then aspirated from the sample, followed by washing in 1× PBS, and fixation with freshly prepared 4% formaldehyde at 4 C for 15 min. Color images of cells labeled at the three temperature points showed a readily observable changes from green (4 C, Figure 2a) to yellow (25 C, Figure 2b) and finally to orange-red (37 C, Figure 2c) color in the scattering signal.
     在不同的溫度下標(biāo)定細胞。為了在等振納米粒子的散射反應(yīng)和活細胞中EGFR機制的動力之間建立關(guān)系，我分別在4、25和37 C下進行標(biāo)定。溫度控制可以讓EGFR的激活和交易機制在關(guān)鍵的點停止。在C時細胞的EGFR內(nèi)化被抑制。在25 C細胞的內(nèi)化進行形成早期的胞內(nèi)體，最終在37 C條件下1小時內(nèi)通過形成多泡體和后期胞內(nèi)體獲得完整的EGFR調(diào)控機制33,38,39。對基于溫度的實驗，細胞要以50 000-100 000個/mL植入蓋玻片，并讓其附著過夜。然后用1× PBS和添加了5% FBS金納米粒子生物連接物的補充培養(yǎng)基沖洗細胞，然后將包含納米粒子細胞樣品的蓋玻片被放入4, 25, 和 37 C條件下，讓其作用60分鐘。隨后納米粒子從樣品中被吸走，接著在1× PBS中沖洗細胞，并用已備的新鮮的4%甲醛在4 C條件下固定15分鐘。在三種不同溫度下標(biāo)定的細胞的色彩圖像顯示了散射信號的明顯變化，從綠(4 C, 圖 2a)到黃(25 C, 圖 2b)最終到橘紅(37 C, 圖 2c)。

Figure 2. Cells labeled at different temperatures. Dark-field images of A431 cells labeled with 25 nm anti-EGFR gold nanoparticle conjugates at 4 (a), 25 (b), and 37 C (c). Dark-field microscopy reveals light scattering from the samples. Controlling temperature arrests the normal EGFR regulatory processes at critical points with receptors remaining on the cell membrane at 4 C and endosomal internalization and multivesicular body sorting at 25 and 37 C, respectively. TEM of identical samples (d, 4 C; e, 25 C; and f,
37 C) shows the nanometer-scale EGFR rearrangements that correspond to the optical images. In (g), a schematic illustrates the qualitative relationship between EGFR regulation state and optical signature of the associated gold nanoparticles.
圖2 在不同溫度標(biāo)定細胞。圖中顯示標(biāo)定了25 nm抗EGFR的金納米粒子連接物的A431細胞在4 (a)、 25 (b)和37 C (c)條件下的暗場圖像。暗場顯微鏡揭示樣品的光散射。通過控制溫度來阻止正常EGFR調(diào)控程序。在4 C時受體停留在細胞膜表面，在25和37 C時分別進行胞內(nèi)體內(nèi)化、多泡體分類揀選。(d, 4 C; e, 25 C; 和 f, 37 C) 完全相同樣品的TEM顯示了納米級別EGFR的新排列，這種TEM圖像與光學(xué)圖像是對應(yīng)的。圖(g)，顯示EGFR的調(diào)控階段與有關(guān)金納米粒子光學(xué)信號之間的定性關(guān)系。

     The corresponding transmission electron microscopy (TEM) images of labeled cells (Figure 2d-f) showed the temperature-dependent changes in the nanometer-scale organization of EGFR that had been reported by others.39 At 4 C, nanoparticles are located on the cell membrane in small clusters. Increasing the temperature to 25 C results in a more three-dimensional, volumefilling aggregate morphology that is consistent with endosomal uptake. At 37 C, progression of EGFR trafficking leads to appearance of an even more complicated aggregate structure corresponding to MVBs, which are destined for EGFR degradation.
     標(biāo)定細胞的透射電子顯微圖像(圖 2d-f)顯示了納米級別EGFR的調(diào)控與溫度有關(guān)的變化，這曾經(jīng)被報道過39。在4 C時，納米粒子被安放在小群集的細胞膜上。溫度提高到25 C時會導(dǎo)致形成更三維、更飽滿的結(jié)合形態(tài)，這與胞內(nèi)體的攝取有關(guān) 。在37 C時，EGFR的交易進程導(dǎo)致出現(xiàn)了一個非常復(fù)雜的與MVB相一致的聚合結(jié)構(gòu)。

     We used hyperspectral imaging to quantify the nanoparticle scattering changes in relation to EGFR behavior (Figure 3, a-i). Hyperspectral images were acquired using the PARISS system (Lightform, Inc.). Scattering peaks in the range of 530-550 nm are present in images obtained at 4 C (Figure 3d); at this temperature EGFR is predominantly located on the cytoplasmic membrane. Increased red-shifting of the peaks becomes apparent at 25 C with EGFR located predominantly in early endosomes (Figure 3e), while a wide range of scattering peak positions is evident at 37 C (Figure 3f). At 4 and 25 C (Figure 3g and h, respectively), labeled cells display peak wavelength distributions with mean and standard deviation values of 546 ±15 and 574 ± 20 nm, respectively. Interestingly, at 37 C (Figure 3i), there is a large degree of heterogeneity in peak scattering wavelengths (mean and standard deviation, 597 ± 44 nm); this is consistent with the simultaneous presence of multiple regulation/ trafficking stages of EGFR within the cell.
     我們用高光譜成像來量化與EGFR行為有關(guān)的納米粒子散射變化(圖 3, a-i)。高光譜成像要求使用PARISS系統(tǒng) (Lightform公司)。散射的巔峰在530-550 nm之內(nèi)的圖像采集于4 C 條件下(圖 3d)：在這種條件下，EGFR明顯位于細胞質(zhì)的膜上。在25 C時隨著EGFR主要集中在早期胞內(nèi)體時紅色漂移開始增多(圖 3e)。然而在37 C時寬譜波段的巔峰位置很明顯(圖 3f)。在4 和 25 C(圖 3g 和 h)，波峰分布的平均值和標(biāo)準(zhǔn)差分別為546 ±15 和 574 ± 20 nm。有趣的是，在37 C (圖 3i)散射光譜的波峰有很大程度的偏移（平均值和標(biāo)準(zhǔn)差為597 ±44 nm）；這與細胞內(nèi)存在EGFR多樣機制和交易階段是一致的。

Figure 3. Quantitative relationship between scattering from EGFR-bound gold nanoparticles and EGFR regulatory stages. Hyperspectral darkfield microscopy of cells labeled at 4 C (left column, a,d,g), 25 C (middle column, b,e,h), and 37 C (right column, c,f,i). The data was acquired using the microscope settings described in Figure 1. Representative single-pixel spectra from a cell image display scattering peaks at ca. 546 (a), 576 (b), and 601 nm (c). Cell images in (d-f) are color-coded according to the scattering peak position at each pixel in the field of view. Pixels that did not have an identifiable peak in a corresponding spectrum were not assigned a color. Distributions of the peak scattering wavelengths indicated in (d-f) are shown in (g-i). Electrodynamic simulations of scattering from nanoparticle aggregates are shown in (j-l) for the following three typical structures: a 4-particle chainlike structure (j,m), a 19-particle disklike structure (k,n), and a 130-particle volume-like structure (l,o). The total scattering cross section is plotted for each structure, alongside a rendering of the corresponding detailed particle arrangement (m-o). Distinct red shifting and broadening of the scattering spectra is due to both the affect of the increased number of contributing particles and the effect of the transition from a 2D to a more 3 D volume-filling morphology.
圖3.  綁定金納米粒子EGFR的散射與EGFR調(diào)控階段之間的定量關(guān)系。在4 C (左邊, a,d,g), 25 C (中間, b,e,h), 和 37 C (右邊, c,f,i)標(biāo)定細胞的暗場高光譜顯微。用圖1中描述的顯微裝置來獲取數(shù)據(jù)。在細胞圖像中代表性的單像素光譜顯示了在大約 546 (a), 576 (b)和 601 nm (c)散射的波峰。(d-f)圖像是根據(jù)視場范圍內(nèi)每一個像素的散射波峰位置標(biāo)定的顏色。如果一個像素沒有可以辨認(rèn)的波峰與光譜相對應(yīng)的像素，就不給分配顏色。(g-i) 顯示了(d-f)中頂峰散射波長的分布。(j-l)為下面三個典型結(jié)構(gòu)是納米粒子聚合物光譜的電動模擬：一個4-粒子的鏈狀結(jié)構(gòu)(j,m)、一個19-粒子的盤狀結(jié)構(gòu)(k,n)，一個130-粒子的卷裝結(jié)構(gòu)(l,o)。全部的散射截面都標(biāo)給每一種結(jié)構(gòu)標(biāo)繪的，旁邊附帶相應(yīng)的粒子分布細節(jié)(m-o)。受有效粒子數(shù)目增長的影響和從二維到三維的形態(tài)變化的作用產(chǎn)生了明顯的紅色漂移和散射光譜的增寬。
Electrodynamic Modeling of Light Scattering from Metal Nanoparticle Aggregates. 金屬納米粒子聚合物光散射電動模型。

     In order to further explore the relationship between nanoparticle aggregate morphology and the associated optical cross sections, we implemented detailed electrodynamic simulations. The computational electrodynamics codes used for the nanoparticle aggregate simulations are part of a new T-matrix code-base implemented entirely in C++, which has been developed over the past several years.40 A hybrid multithreaded messagepassing- interface (MPI) task model is used throughout, allowing optimal utilization of modern cluster-computing resources. The specific T-matrix formulation used, as extended to aggregates, is based primarily on the methods discussed in Mackowski, Mishchenko,41-43 and Stout.44 The specifics of the permittivity functions used are described in Aaron et al.45 In all cases, incident illumination is unpolarized with wave-vector directed into the page as the rendered structures are shown (Figure 3m-o). Gold spheres are 25 nm in diameter, surrounded by water, with mean interparticle center-to-center spacing of 2.4 times the particle radius. The particles are normally distributed in 3D about ideal lattice-points with a fractional standard deviation of 0.07.
For each of the aggregate morphologies, simulation structures were constructed using typical values for aggregate particle number, interparticle spacing, and overall morphology as indicated from the TEM images.
     為了進一步發(fā)掘納米粒子聚合物形態(tài)與聯(lián)合光學(xué)截面之間的關(guān)系，我實施詳細的電動模擬。用于計算納米聚合模擬的電動代碼是完全應(yīng)用基于C++的部分新型T矩陣代碼，它是在過去的幾年中被開發(fā)出來的。自始至終都使用混合的多線程的信息通道截面(MPI)工作模型，采用最佳的現(xiàn)代群組計算資源。具體的擴展至聚合物的T-矩陣構(gòu)架主要基于Mackowski, Mishchenko41-43 和 Stout44探討的方法。介電常數(shù)函數(shù)的細節(jié)在Aaron et al上有陳述。在所有的實驗中入射光都是非偏振的，波動矢量像圖 3m-o一樣指向頁面。金球體直徑為25nm，由水環(huán)繞，粒子中心間隔為2.4倍粒子半徑。粒子的3D理想晶格點分布的小數(shù)標(biāo)準(zhǔn)偏差為0.07。每一個聚合形態(tài)都用TEM圖像中描述的聚合顆粒的數(shù)目、顆�？臻g、以及整個形態(tài)的典型值來構(gòu)造。

     The optical behavior of nanoparticle aggregates has a complicated, nonlinear dependence on details of the aggregate morphology, in addition to a greatly enhanced optical cross section that varies, in general, quadratically as the number of particles in the aggregate.29,40 The details of aggregate morphology may include receptor/nanoparticle number, spacing, and overall arrangement in space (such as 2D planar or a more 3D volume-filling configuration). In general, the effect of interparticle coupling produces a red shift and broadening of the resonance peak. Figure 3(j-l) shows scattering cross sections from three typical structures (Figure 3m-o) that were designed to approximate the aggregate morphology indicated in Figure 2d-f. Note that the simulated spectra show the overall redshifting and peak broadening, consistent with spectra shown in Figure 3a-c.
     納米粒子聚合物的光學(xué)反應(yīng)復(fù)雜非線性地依賴聚合物的形態(tài)細節(jié)，而且很大提高光學(xué)截面群集粒子數(shù)量的變化29,40。聚合物的形態(tài)細節(jié)可以顯示受體/納米粒子的數(shù)量、空間和總體的空間分布（比如二維平面或者三維立體結(jié)構(gòu)）。通常情況粒子間耦合的影響會產(chǎn)生一個紅色漂移和諧振波峰的變寬。圖3(j-l)顯示三種典型結(jié)構(gòu)(圖 3m-o)的散射橫截面，這三種模型被設(shè)計成近似聚合物形態(tài)，圖2d-f。注意的是模擬光譜顯示總體紅移和波峰增寬，與圖3a-c描述的一致。

     For random aggregates, due to the high-order, nonlinear dependence of the plasmonic resonance coupling on the interparticle spacing, small statistical fluctuations in particle position about the mean have a significant effect on spectral shift.40 In addition, three-dimensional more volume-filling aggregates, such as those seen in endosomes, experience significantly more spectral shift than two-dimensional aggregates. This is noteworthy because it indicates that NPRC can detect differences in morphology associated with internalization events due to corresponding changes in the nanometer-scale organization of EGFR.
     對于隨機的聚合物，由于在粒子空間內(nèi)高位非線性地依賴等振粒子共振耦合，所以粒子位置平均值波對光譜漂移有明顯影響。而且像那些在胞內(nèi)體看到的三維volume-filling 比二維聚合物經(jīng)歷著更多的光譜漂移。這是值得關(guān)注的，因為它說明NPRC可以探測到形態(tài)上面的差別與內(nèi)化作用有關(guān)，這種內(nèi)化作用是由相應(yīng)納米級別EGFR組織的變化導(dǎo)致的。

     Monitoring of EGFR Trafficking. Having established the relationship between the scattering properties of EGFR-bound nanoparticles and EGFR regulatory stages, we examined the degree to which RGB imaging could be effectively employed to monitor EGFR trafficking. It appeared that the increase in the relative intensity of the red channel in color images of live cells (Figure 1, B) was a pronounced indicator of EGFR activation and endocytosis. Therefore, we used this parameter to correlate color changes in live cell images with EGFR regulatory stages. To accomplish this, the hyperspectral data presented in Figure 3 were grouped into three categories, corresponding to the associated EGFR regulatory states, (1) surface receptors, (2) early endosomes, and (3) late endosomes/MVBs. Pixel-data from at least five complete cells were analyzed in each group, representing at least 1000 spectra per group. The resulting spectra were used to obtainstatistically distinct distributions for intensity-normalized red channel values characteristic for each EGFR state (Figure 4). Then, for each nonbackground pixel in the time-lapse images (shown in Figure 1), we determined the probability that its normalized red channel value falls within each ofthe distributions associated with EGFR states (Figure 4). These paired comparisons resulted in three p-values that were used as weighting factors in assigning a pseudocolor to each pixel in live cell images. This analysis resulted in images where color represents a statistical probability of a regulatory stage of EGFR (Figure 5).
     監(jiān)視EGFR交易。 既然已經(jīng)建立了綁定納米粒子的EGFR散射性質(zhì)與EGFR的調(diào)控階段之間的關(guān)系，我們檢驗三基色圖像等級也可以有效地用來監(jiān)視EGFR交易�；罴毎�顏色圖像的紅帶相對強度出現(xiàn)增長(圖 1, B)是EGFR激活和內(nèi)吞作用的一個明顯標(biāo)志。因此，我用這個參數(shù)將活細胞圖像顏色變化與EGFR的控制階段聯(lián)系起來。為了達到這個目標(biāo)，圖3提供的高光譜數(shù)據(jù)被分成三個類別，分別對應(yīng)相關(guān)聯(lián)的EGFR的調(diào)控階段。（1）表皮受體，（2）早期胞內(nèi)體，（3）晚期胞內(nèi)體/MVB。每組至少要分析5個完整的細胞被，每組至少代表1000光譜。用得出的結(jié)果光譜被來為每一個EGFR區(qū)域標(biāo)準(zhǔn)化紅帶強度值獲得統(tǒng)計上的明顯的區(qū)域分布（圖4）。然后對每一個無背景像素的延時圖像，我們分析標(biāo)準(zhǔn)紅帶值屬于每一個相關(guān)EGFR階段的幾率。這些成對的比較導(dǎo)致3個假定值，在為活細胞圖像每個圖像設(shè)定偽色時這些假定值被用作權(quán)定因數(shù)。這個分析可以形成圖像，在圖像中顏色代表著EGFR控制階段的統(tǒng)計概率(圖 5)。

Figure 4. Relative red channel intensity distributions for gold nanoparticle labeled EGFR molecules on the cell surface (blue), in early endosomes (green), and in late endosomes (red). For the CCD used, the red channel collects signal in the 600-700 nm wavelength range. Increasing plasmon resonance coupling in each case results in an overall increase in red channel intensity. Live cell images from Figure 1 were analyzed on a pixel by pixel basis and compared with the three distributions shown via a statistical z-test.
圖4.標(biāo)定金納米粒子EGFR分子的相關(guān)紅帶強度分布——細胞表面（藍）、早期胞內(nèi)體（綠）、后期胞內(nèi)體（紅）

     由于CCD的使用，紅帶在600-700 nm波長范圍內(nèi)收集信號。每種情況等離子共振耦合的增加都導(dǎo)致紅帶強度整體的增加。用像素原理對圖1活細胞圖像進行一個像素的分析，并與一個數(shù)據(jù)z-測試顯示的三種分布相比較。

Figure 5. Pseudocolor images of live cells representing distribution of EGFR regulatory stages. Statistical analysis of color time-lapse images shown in Figure 1 results in color-coded maps of EGFR distribution in untreated cells (A), and in cells after treatment with 10 μM AG1478 (B); a blue color indicated high probability that EGFR is located on the cytoplasmic membrane, green indicates location in the early endosomes, and finally, red indicates location in the late endosome/MVBs. Note that treatment with AG1478 markedly reduces endocytosis of EGFR. The arrow in panel A, far right, indicates a filopodium emanating from an adjacent cell, which displays heavy EGFR trafficking along its axis.
圖5    代表EGFR調(diào)控階段分布的活細胞偽顏色圖像
對圖1顯示的延時圖像色彩統(tǒng)計分析決定了要為EGFR的分布標(biāo)定顏色—未處理細胞(A)、用10 μM AG1478處理的細胞(B)。藍色表明EGFR位于胞質(zhì)膜的可能性很高，綠色表示位于早期胞內(nèi)體，紅色表示位于晚期胞內(nèi)體/MVB。注意AG1478處理明顯地降低了EGFR的內(nèi)吞作用。A圖最右邊的箭頭表明絲線來自鄰近的細胞，顯示了圍繞其軸線的大量EGFR交易。

     In contrast to previous studies that employed fluorescent tags such as quantum dots46 to monitor EGFR trafficking, the use of plasmonic resonant nanoparticles provides additional information about proximity and nanometer-scale organization of biomolecules. It has been shown that the optical changes corresponding to interparticle distance between nanoparticle pairs can be precisely quantified23 and work is under way to develop a similarly quantified model for nanoparticle aggregates.40
     與以前的用熒光標(biāo)記的研究（例如量子點46）相比，等振離子納米粒子的使用提供了鄰近納米級別生物分子組織的附加信息。已經(jīng)證實，與粒子間距相對應(yīng)的光學(xué)變化可以被精確地量化，而且為納米粒子連接物開發(fā)一個類似的量化模型40工作正在進行。

     The application of NPRC described here can be extended to other RTKs and G-protein-coupled receptors that are involved in signal transduction.33 Other molecular systems for investigation may include integrin clustering and immunological synapse formation. This method can potentially be applied to heteromolecular interactions such as interactions between different RTKs if the interacting molecules are labeled with plasmonic nanoparticles that exhibit distinct optical properties. Theoretical simulations that are described here can be used to optimize the design of nanoparticles that would produce distinct optical changes in the case of homoand heterointeractions.
     這里描述NPRC的應(yīng)用可以擴展到其他的RTK和涉及信號轉(zhuǎn)換的G蛋白耦合受體33。其它調(diào)控分子的系統(tǒng)包含整合蛋白群集免疫學(xué)突觸形狀的信息。如果相互作用的分子被標(biāo)定顯示明顯光學(xué)特征的等離振子納米粒子，這種方法可以用來探測異性分子的交互作用，例如探測不同RTK間的交互作用。這里的理論模擬可以用來優(yōu)化納米粒子的設(shè)計，這種設(shè)計在單一異性交互作用的情況下將產(chǎn)生明顯的光學(xué)變化。

     A primary strength of NPRC as a biosensing tool is associated with the complex optical behavior of plasmonic nanoparticle assemblies. The dramatic changes in optical properties associated with nanometer-scale changes in organization facilitate the development of statistical associations with the biological processes under observation. Further understanding of NPRC will be vital for its development as a quantitative biosensing tool. Issues such as particle shape heterogeneity, as well as the more nuanced effects of nanoparticle aggregate morphology on spectral characteristics (such as scattering peak width) are still under study and promise to reveal a wealth of additional information.
     作為生物感應(yīng)工具NPRC主要的優(yōu)點是聯(lián)系著等離子納米粒子裝配的復(fù)雜光學(xué)反應(yīng)。觀察下，組織中與納米級別變化有關(guān)的光學(xué)特征數(shù)據(jù)的改變促進了生物進程統(tǒng)計聯(lián)系的發(fā)展。為其作為一個定量生物感應(yīng)工具的發(fā)展，進一步了解NPRC是非常重要的。粒子形狀的不均勻性以及納米粒子聚合物對光譜特征的微妙影響（例如散射波峰寬度）還在研究當(dāng)中，有望能夠揭示大量的有用的附加信息。

     NPRC can be used as an independent method or it can potentially be integrated into already in-use biophysical methods such as FRET,1,4 ICM,2,3 or EM5 to provide additional complementary information. TEM with immuno gold labels has been used to study 3D organization of proteins with high spatial resolution in fixed samples. The method presented here allows extending this TEM imaging approach to live-cell dynamic optical imaging of molecular assemblies with unambiguous information about the proximity and nanometer- scale organization of biomolecules. While not strictly a super-resolution method, NPRC can provide sub diffraction characterization capability over distances ranging from a few tens of nanometers, which significantly exceed distances available through FRET. In addition, NPRC exhibits sensitivity to geometrical characteristics of protein clusters such as a planar 2D or a volume-filling 3D aggregation in livecell imaging and it is capable of following cluster dynamics in real time.
     NPRC可以被用作獨立的方法，也可以整合到先前使用過的生物物理方法上（例如FRET,1,4 ICM,2,3 或者 EM5）來提供附加的完善信息。有免疫金標(biāo)簽的TEM已經(jīng)被用在研究固定樣品高分辨率的三維蛋白質(zhì)組織。這里使用的方法是將TEM圖像擴展到活細胞動態(tài)光學(xué)圖像，這些影像是分子組裝（關(guān)于鄰近和納米級別生物分子組織）的清晰信息。然而對那些不要求超分辨率的方法，NPRC可以提供描述輔助衍射特征的能力，排列距離超過成千上萬納米，這明顯超過了FRET的有效距離。此外，NPRC展現(xiàn)了其對蛋白質(zhì)族幾何特征的敏感性，例如活細胞圖像中平面二維和立體三維聚合物，并且能夠?qū)崟r的追蹤族的活力。

     The sensitivity of NPRC to nanometer-scale geometry of individual clusters can potentially provide invaluable complementary information to ensemble population measurements such as cluster density and the average number of receptors per cluster that are afforded by ICM methods. Further, gold nanoparticles themselves can make excellent probes for ICM, due to their brightness and photostability, which will facilitate the direct combination of these techniques. We anticipate that NPRC will continue to gain interest among biological and biomedical researchers as a method to produce unique insight into biomolecular interactions.
     NPRC對納米級別的個別群簇的敏感性可以提供潛在的非常有價值的補充總體種群測量值信息，例如ICM方法可以提供群密度、每個種群受體的平均數(shù)量。而且，納米粒子本身也可以成為ICM極好的探針，因為其亮度和耐光性都可以促進這些技術(shù)的直接聯(lián)合。我們預(yù)期NPRC將會不斷引起生物及生物醫(yī)藥研究者的興趣，這種方法對生物分子內(nèi)作用有獨特的洞察力。