Direct insertion of proteins into a living cell using an atomic force microscope with a nanoneedle

Original Article

Correspondence and reprint requests to:

C. Nakamura

Research Institute for Cell Engineering (RICE),National Institute of Advanced Industrial Science and Technology (AIST),3-11-46 Nakoji,Amagasaki,Hyogo 661-0974,Japan E-mail:

chikashi-nakamura@aist.go.jp

NanoBiotechnology

Copyright ?2005 Humana Press Inc.

All rights of any nature whatsoever are reserved.ISSN 1551-1286/05/01:347–352/$30.00(Online) 1551-1294DOI:10.1385/Nano:1:4:347

Direct Insertion of Proteins into a Living Cell Using an Atomic Force Microscope with a Nanoneedle

Ikuo Obataya,1Chikashi Nakamura,1,2SungWoong Han,2Noriyuki Nakamura,1,2and Jun Miyake 1,2

1

Research Institute for Cell Engineering (RICE),National Institute of Advanced Industrial Science and T echnology (AIST),3-11-46 Nakoji,Amagasaki,Hyogo 661-0974,Japan;and 2Department of Biotechnology and Life Science,T okyo University of Agriculture and T echnology (TUAT),2-24-16 Naka-cho,Koganei,T okyo 184-8588 Japan

Abstract

We have developed a tool for directly inserting proteins into living cells by using atomic force microscopy (AFM) and an ultrathin needle,termed a nanoneedle. The surface of the nanonee-dle was modified with His-tagged proteins using nickel chelating nitrilotriaceticacid (NTA).The fluorescent proteins,DsRed2-His 6and EGFP-His 6,could be attached to and detached from the surface of the nanoneedle. These results suggest that the Ni-NTA modified nanonee-dle can successfully be used for specific delivery of proteins. The nanoneedle modified with DsRed2-His 6was able to penetrate the surface of a living HeLa cell,as confirmed by laser scanning fluorescence microscopy and monitoring an exerting force on the nanoneedle using AFM. Force curves using the nanoneedle indicated that the needle was able to penetrate at displacement speeds of 0.10–10 μm/s. These results suggest that this technique can be used to directly insert proteins into living cells and is applicable for modulation or regulation of single cell activity.

(Nanobiotechnology

DOI:10.1385/Nano:1:4:347)

Key Words:AFM; living cells; nanoneedle; LSM; protein transfer.

Introduction

Molecular genetic techniques have been used to evaluate the different functions of molecules such as DNA,RNA,and proteins of interest. Indeed,transfection of genes into cells is a general method for observing phe-notypic variations in cells and for quantifying differences in molecules of particular interest.There is,however,a growing requirement in biology to further investigate and operate sin-gle cells. In particular,it is important to understand location-specific functions inside living cells and to assess in real time the rela-tionship between external actions on cells and their corresponding cell reaction. Recent advances in the utilization of fluorescent pro-teins have promoted the analyses and opera-tion of distinctive aspects at the single-cell level (1).

Among the single-cell techniques,a microinjection technique has contributed to the introduction of solutions containing mol-ecules of interest,allowing researchers to inject a molecule or an organelle into a tar-geted living cell wherever and whenever they wish (2). For this reason the technique has often been used for the operation of oocytes or stem cells (3,4). Indeed,the technique is very powerful and versatile,but there are some problems associated with its invasive-ness and accuracy of positioning. Cell dam-age caused by inserting solid materials into living cells has not,to date,been discussed despite its effect on cell phenotype. To explore this,we have developed a cell surgery system to overcome these problems using an atomic force microscope (AFM) and an ultrathin nee-dle,referred to as a nanoneedle (5,6).

348_____________________________________________________________________________________Obataya et al.

We considered that molecules could be delivered into cells by attaching them to the surface of a needle,instead of inject-ing them through a hollow material such as a capillary. This concept,which utilizes atomic force microscopy (AFM), would allow us to optimize the shape of the inserting mate-rial for specific insertions and thereby minimize cell damage. Our previous study indicated that a flat-ended cylindrical nanoneedle with a diameter of about 200 nm is capable of penetrating living cells (6). Providing such a needle can be modified with molecules of interest,the reaction of a single cell could be analyzed just upon insertion. In order to accom-plish this operation,it is first necessary to establish a method for modifying and confirming the successful insertion of the protein-modified nanoneedle into the living cell.

In this study,we have attempted to prepare a nanoneedle modified with a fluorescent protein fused poly-histidine tag (His-tag) and to determine its insertion potential at various displacement speeds into a living cell.

Materials and Methods

HeLa cells were purchased from Dainippon Pharmaceuticals (Osaka,Japan). PSA (penicillin,streptomycin,and ampho-tericin B) solution was purchased from Kurabo (Osaka,Japan). AFM cantilevers (ATEC-CONT) were purchased from Nanosensors (Neuchatel,Switzerland). Focused-ion-beam (FIB) etching of the cantilevers was conducted by Hitachi Science Systems (Tokyo,Japan). Plasmids pDsRed2-C1, pEGFP-C1,and pEGFP-actin were purchased from Clontech (Palo Alto,CA). Glass-bottom dishes were purchased from IWAKI Glass (Tokyo,Japan). N-[5-(3′-maleimidopropy-lamido)-1-carboxypentyl]iminodiacetic acid (Maleimido-C3-NTA) was purchased from Dojindo (Kumamoto,Japan). All other materials were purchased from Wako Pure Chemicals (Osaka,Japan).

Plasmid Construction of pDsRed2-His

6

and pEGFP-His

6

Amplification of the DsRed2 and EGFP fragments (7,8). were performed using the oligonucleotide primers 5′-ATG-GTGAGCAAGGGCGAGGAG-3′(EGFP forward),5′-CTTGTACAGCTCGTCCATGCC-3′(EGFP reverse), 5′-ATGGCCTCCTCCGAGAACGTC-3′(DsRed2 forward), 5′-CAGGAACAGGTGGTGGCGGCC-3′(DsRed2 reverse) where polymerase chain reactions (PCR) were set up with 1 unit Taq EX (TAKARA Bio Inc,Shiga,Japan),dNTPs,0.5μM of each primer,about 2 ng pDsRed2-C1 or pEGFP-F DNA as template and buffer,in a total volume of 100 μL. The PCR product was cloned into an Escherichia coli expression vector containing six His residues at the C-terminus of a cloning site (pCR?T7/CT-TOPO,Invitrogen Japan KK, Tokyo,Japan) followed by transformation into E. coli DH5αstrain (TOYOBO,Osaka,Japan) as described by the manu-facturer. The ampicillin-resistant cells were grown overnight at 37°C in LB medium containing ampicillin (100 μg/mL), followed by cell harvesting by centrifugation. Plasmids were purified using the Miniprep Kit (TAKARA BIO Inc,Shiga, Japan) followed by analysis for possible inserts using agarose [1% (w/v)] gel electrophoresis. The plasmid inserts were detected by PCR and digestion.

Protein Expression of DsRed2-His

6

and EGFP-His

6

E. coli.BL21(DE3)pLysS strain was transformed with the plasmids pDsRed2-His

6

or pEGFP-His

6

and grown in LB medium supplemented with ampicillin until late log phase. Expression of the cloned gene was induced by addition of the isopropylthiogalactopyranoside (IPTG; final concentration, 1 m M). The culture was grown for another 2 h,and the cen-trifuged cell pellet was broken and suspended with a B-PER bacterial protein extraction reagent (Pierce,Rockford,IL). The His-tagged proteins were then purified using HisTrap HP (Amersham Biosciences K. K.,Tokyo,Japan) and analyzed by SDS-PAGE. The buffer was exchanged to 0.05% Tween-20/ PBS by gel filtration using PD-10 column. The protein concen-trations were determined with the BCA protein assay (Pierce). Cell Culture

HeLa cells were transfected with pEGFP-actin by using LipofectAMINE 2000 (Invitrogen). They were cultured in 10% fetal calf serum (Sigma-Aldrich Japan,Tokyo,Japan)/ Dulbecco’s minimal essential medium (Sigma-Aldrich) con-taining Glutamax (I nvitrogen),PSA,and Geneticin (Invitrogen) in a humidified incubator at 37°C with 5% CO

2

. They were plated after trypsinization onto 35-mm glass-bot-tom culture dishes coated with a collagen type-I. Preparation of Nanoneedle Modified

With Ni-NTA Moiety

The nanoneedle was prepared by FIB etching as previously described. The pyramidal silicon AFM tip was fabricated to a needle shape of about 200-nm diameter and 6–8 μm length with a flat-ended cylindrical edge. The surface of the tip was cleaned with ozone using oxygen gas and UV light followed by functionalization with 2% (3-mercaptopropyl)trimethoxysi-lane (MPTMS) in EtOH for 30 min. The obtained tip was reacted overnight with Maleimide-C3-NTA in 50% DMF/100 mM Tris-HCl (pH 7.6). The NTA groups on the surface were chelated with 10 m M NiCl

2

. Finally,the tip was incubated in bovine serum albumin (1 mg/mL in PBS) in order to cover any surfaces open to nonspecific absorption of proteins. Attachment and Detachment of His-Tagged Proteins

The surface of the AFM tip was modified with Ni-NTA as described above. The Ni-NTA modified tip was modified with DsRed2-His

6

(57 μg/mL) in PBS for 1 h at 25°C followed by washing with PBS for at least three times. The attached DsRed2-His

6

was detached by incubation with 500 m M imi-dazole solution for 1 h at 25°C followed by washing with PBS. Reattachment of EGFP-His

6

onto the surface was con-ducted using the same method as for the first attachment using the solution of 39 μg/mL EGFP-His

6

.

Direct Insertion of Proteins into a Living Cell Using an Atomic Microscope With a Nanoneedle ___________________349

Apparatus

Operation of the living cells was performed using a com-bination of a FLUOVIEW FV300/IX71 (Olympus,Tokyo,Japan) confocal laser scanning microscope (LSM) and a molecular force probe MFP-1D (Asylum Research,Santa Barbara,CA). Images of all the confocal slices were collected and reconstructed to produce cross-section images with FLU-OVIEW software. All images were obtained by double exci-tation at 488 nm and 543 nm using a 60x objective lens (oil).The emission signal of GFP and DsRed was obtained through emission filter sets of 510/530 nm and 565/610 nm. The microscope and MFP were insulated using a custom-made incubator (Tokken,Chiba,Japan) to maintain conditions of 5% CO 2and 37°C around the stage of the microscope during the operation and observation.

Results and Discussion

The apparatus used to operate the living cells relied on AFM,with microscopic observations undertaken by confocal LSM (5). In order to maintain an appropriate environment for culturing the cell,the microscope was covered with an insu-lator. The environment around the stage of the microscope was kept at 37°C with a 5% CO 2atmosphere using heaters and a CO 2regulator (Fig. 1A). Figure 1B shows a nanonee-dle fabricated into a needle shape with a diameter of about 210 nm,length of 7 μm,and a flat-ended cylindrical edge using FIB etching as previously reported.

We modified the silicon surface of the AFM tip using the nitrilotriaceticacid (NTA) group (Fig. 2). Generally,recombi-nant proteins harvested from cell cultures are subjected to affin-ity chromatography because these proteins with affinity tags can be easily purified from cell lysates. Among the current affinity tags available,NTA is one of the most useful matrixes for immobilized metal affinity chromatography (9). Proteins with a His-tag can be attached tightly onto the Ni-NTA group and can be easily detached using either high concentrations of

imidazole,low pH,or denaturants such as urea or guanidine hydrochloride (GuHCl). It is important to point out that a pro-tein immobilized on a Ni-NTA surface is likely to remain in an active state provided the His-tag position is correct and the linker length is sufficient between the His-tag and the origi-nal protein region. Furthermore,a protein on a Ni-NTA surface can be altered to produce another protein by detaching and reproducing it on the Ni-NTA surface. Therefore,we believe that a Ni-NTA surface is a useful interface for insertion of proteins into living cells.

First,we tested the attachment and detachment efficiency of the His-tagged protein using silicon AFM that was not fab-ricated. The silicon surface was oxidized and cleaned with ozone,then a thiol group was derivatized on the surface using 2% MPTMS. The thiol group reacted with the maleimide group of Maleimido-C3-NTA to produce a NTA group on the surface. Thereafter,a nickel (II) ion was chelated onto the NTA group in a specific buffer. To avoid any nonspecific interactions between the surface and the proteins,the surface was blocked with 1 mg/mL bovine serum albumin. Finally,the AFM tip was reacted with DsRed2-His 6for 1 h following washing with PBS. The tip with DsRed2-His 6was observed by LSM in PBS using excitations at 488 nm and 543 nm. As shown in Fig. 3A,DsRed2-His 6successfully bound to the Ni-NTA group when the fluorescence was of sufficient intensity.The intensity was comparable to that of fluorescent dyes such as AlexaFluor488 or AlexaFluor546 under the same condi-tions. However,the fluorescence profile on the surface was heterogeneous,suggesting that further chemical optimization was necessary during the chemical derivatization process.Figure 3B depicts AFM tip after exposure to 500 m M imida-zole for 1 h. The fluorescent intensity of the surface substan-tially decreased compared to the surface before treatment. In general,denaturants such as urea or GuHCl are used for com-plete detachment; however,such denaturants were not used to evaluate the remaining protein after treatment in this study.

Fig.1.(A ) Schematic representation of the apparatus for the cell operation.The microscope and AFM unit are in the insulator to maintain a suitable environment for the cell.(B ) The nanoneedle fabricated from silicon AFM tip using FIB.Scale bar indicates 2 μm.

350_____________________________________________________________________________________Obataya et al.

Independent experiments revealed that the remaining fluo-rescence totally diminished after treatment with 8 M GuHCl,which was due to both detachment and denaturation of the protein. Figure 3C displays the LSM image after modifica-tion with EGFP-His 6. The results show that the green fluo-rescence of EGFP was dominant on the surface,suggesting that the proteins on the surface of the silicon tip can be changed in a few steps. Thus,it is practical to use a Ni-NTA surface on a silicon AFM tip for protein modification.

The nanoneedle fabricated from the silicon AFM tip was modified with DsRed2-His 6via the Ni-NTA group in the same

manner as used for the pyramidal tip. Figure 4A shows a stack

Fig.2.

Procedure for modification of the silicon surface with His-tagged protein.

Fig.3.(A) LSM image of modified AFM tip after modification with DsRed2-His 6protein.(B)Image after treatment with 500 m M imida-zole.(C)Image after attachment of EGFP-His 6protein.λex =488 nm and 546 nm,λem =510–530 nm (green signal) and 565–610 nm (red signal).

Direct Insertion of Proteins into a Living Cell Using an Atomic Microscope With a Nanoneedle ___________________351

of fluorescence image slices of a living HeLa cell visualized by transient expression of EGFP-actin,and with the AFM can-tilever in the culture medium,where the tip of the cantilever was aimed at the nucleus of the HeLa cell. Figure 4B displays a cross-sectional image reconstructed from the stack shown in Fig. 4A. The position of the nanoneedle was lowered using a piezo device onto the AFM while the exerting force on the nanoneedle was monitored. After the insertion signal (a sud-den decrease in force or relaxation),the images were collected by LSM to construct a cross-sectional image (Fig. 4C). The image demonstrated that the nanoneedle was lowered hori-zontally to the targeted position and penetrated through the cell surface even when the height of the HeLa cell was only 5 μm. The fluorescence intensity on the needle surface remained constant for a 30-min period while the needle was held in the cell. These results imply that the Ni-NTA linker was stable and the protein was intact during the operation.We also investigated the influence of displacement speed on the force curves during the insertion process. Whether the needle can penetrate or not can be judged by observing a decrease or relaxation of an existing force. It must be noted,however,that the chemical structure of the needle surface greatly affects the force-curves. Therefore,the needle modi-fied with the protein must travel at an appropriate displace-ment speed,otherwise the penetration event cannot be detected by force monitoring using AFM. Figure 5 shows the force curves during insertion of the nanoneedle modified with

DsRed2 via the Ni-NTA group at various insertion speeds.

Fig.4.(A)

Fluorescent image where the AFM tip was aimed at the nucleus of a HeLa cell.(B)Cross -section view before insertion of the nanoneedle modified with DsRed2-His 6.(C)Cross -section view after insertion of the needle.Scale bar indicates 20 μm for A and 10 μm for B and C .

Fig.5.Force curves when the protein-modified nanoneedle was inserted and pulled out.The nanoneedle operated with a speed of 10 μm/s (A),5.0 μm/s (B),1.0 μm/s (C),and 0.10 μm/s (D).

352_____________________________________________________________________________________Obataya et al.

Here,the force was zero before contacting on the cell surface. After contact,the force increased during the indentation of the cell surface,followed by a sudden drop. This drop in force indicated the point of penetration through the cell surface. The needle was found to successfully penetrate the cell sur-face at all the displacement rates examined in this study. At a displacement rate of 10 μm/s,the force before penetration ranged from 1 to 2 nN using the protein-modified nanoneedle (Fig. 5A). The force at penetration decreased as the speed of the needle decreased to slower speeds. The force during inser-tion also decreased as the needle speed decreased. As a result, the relaxation forces were detectable for all of the displace-ment rates examined in this study.

When the nanoneedle was moved to a higher rate,an over-lapping oscillation became noticeable. This oscillation was pre-sumed to originate from thermal noise or ambient vibrations. If the frequency of the oscillation noise was too low,it could mask or affect measurements at the point of penetration. When the nanoneedle was reduced to less than 5 μm/s,the dropping force was distinguishable from the oscillation noise because the apparent oscillation frequency was higher compared to the traveling rate. The period from the point of contact to the relax-ation point indicates when the needle indents the cell surface, i.e.,cell deformation. Obviously,a small deformation is favor-able for living cells. Although this deformation period was longer at higher rates of displacement than at lower rates,the indentation depth was,at most,only 1 μm,suggesting that cell deformation was small for all of the displacement rates. Taken together,the needle could therefore successfully penetrate the cell surface and the relaxation force at the penetration site was distinguishable from 0.10 to 10 μm. We consider that the prac-tical displacement speed was from 1 to 5 μm/s with respect to the force sensitivity of the current AFM apparatus.

In summary,we have demonstrated the direct insertion of a protein bound to the surface of a nanoneedle into a living cell. By applying a Ni-NTA group as a linker,the protein bound successfully onto the needle surface. The protein mod-ified nanoneedle could penetrate the cell surface to access the inside of a living cell. Because His-tagged proteins can be bound and unbound using simple procedures,this technique should offer much scope and potential for other His-tagged proteins.

Acknowledgments

We would like to thank the Industrial Technology Research Grant Program provided by the New Energy and Industrial Technology Development Organization (NEDO) of Japan.

I. O. acknowledges Industrial Technology Fellowship Program from NEDO.

References

1.Sawano,A.,Takayama,S.,Matsuda,M.,and Miyawaki,A.

(2002),Dev. Cell3,245–257.

2.King,R. (2004),Methods Mol. Biol.245,167–174.

3.Wakayama,T.,Perry,A. C.,Zuccotti,M.,Johnson,K. R.,and

Yanagimachi,R. (1998),Nature394,369–374.

4.Tsulaia,T. V.,Prokopishyn,N. L.,Yao,A.,et al. (2003),J. Biomed.

Sci.10,328–336.

5.Obataya,I.,Nakamura,C.,Han,S. W.,Nakamura,N.,and

Miyake,J. (2005),Nanoletters5,27–30.

6.Obataya,I.,Nakamura,C.,Han,S.,Nakamura,N.,and Miyake,J.

(2005),Biosens. Bioelectron.20,1652–1655.

7.Matz,M. V.,Fradkoc,A. F.,Labas,Y. A.,et al. (1999),Nat.

Biotech.17,969–973.

8.Chalfie,M.,Tu,Y.,Euskirchen,G.,Ward,W. W.,and Prasher,

D. C. (1994),Science263,802–805.

9.Hochuli,E.,Dobeli,H.,and Schacher,A. (1987),J. Chromatogr.

411,177–184.