Tools and methods to construct the system of microarray-with-manageable-volumes (shortly, MMV) have been developed in this study, including an MMV generator, various types of MMV plates (made of gel or plastics with wells of different size and number), MMV containers for centrifugation and incubation and sample transfer/solution preparation methods.
MMV-generator and MMV
MMVs were fabricated using a home-made apparatus, MMV generator which contains DMD (digital multi-mirror device) projector LVP-XD10 (Mitsubishi, Japan) with the optical system modified (Fig. 2a). This apparatus can project any light pattern under the direction of computer and thus can polymerize gel/plastics, of which reaction can be initiated by light irradiation, in an arbitrary shape (Fig. 2b). A typical dry MMV (1024 wells/inch2) is shown with the dimension of a well (Fig. 2c) and how to fabricate an MMV is depicted (Fig. 2d). Both types of MMVs, wet and dry, were developed using acrylamide aqueous solution and acrylate (pentaerythritol tri-acrylate; Aldrich, Germany), respectively. For the generation of polyacrylamide gel (wet MMV), 18% acrylamide solution (acrylamide:bis-acrylamide = 19:1) containing 0.27 mM riboflavin and 70% sucrose, was used. For acrylate resin (dry MMV), pentaerythritol tri-acrylate, containing 0.45% bis(2,4,6-trimethylbenzoil)-phenylphosfinoxide IRGACURE®819 (Ciba, Japan) and 0.05% 2-hydroxy-2-nethyl-1-phenyl-propane-1-on DAROCUR®1173 (Ciba, Japan), was used. After a brief light irradiation using mercury-lump VLT-X10P (Osram, Germany), unpolymerized solutions were immediately removed by jet-water (in case of wet MMV) or by spinning off (in case of dry MMV). The wet MMV was further subjected to a buffer-exchanging process to remove riboflavin, acrylamide monomers and sucrose and replace with an appropriate solution. Thus, the plate of a wet MMV was usually equilibrated with a reaction solution without large molecules such as peptides and DNAs. The swelling/shrinking effect of gels during these processes was kept carefully within control. The dry MMV was finally washed in a labware detergent 7-X ES PF (Dainippon Pharm. Co., Japan) with sonication.
MMV operations
Due to the great number (~1000) and the smallness (~0.5 μL) of samples, one-by-one transfer of them is neither reasonable nor realistic. Therefore, we adopted the most direct approach for transferring samples from an MMV to an MMV by attaching two MMVs (donor MMV and recipient MMV) face-to-face and then precipitating by spinning (aperture-to-aperture transfer) and found this working. We further developed necessary techniques for handling MMVs: charging, addition, and selective addition (filtering or template method) of samples, measuring up to a volume, mixing, incubation (cultivation or enzymatic reaction), PCR, separation into solid and liquid parts.
a) Charging/Addition
Initial charging of sample solution into each well can be performed by spinning of a pool of solution placed on the top surface of MMV, resulting in filling all the wells with the solution and the remaining solution overflowed over the rim (see Fig. 3b). In case of adding another aliquot to the content of a well in an MMV, a transfer from the concerned MMV to an MMV having deeper wells (recipient MMV) was carried out in advance and then an additive transfer from an MMV (donor MMV) was made (see Fig. 3a). Further additions can be done until the recipient MMV is filled up.
(Development of a robotics is effective but not essential for most experiments.)
b) Selective addition (filtering/template method)
In those cases which require addition of a solution to particular wells, we can adopt either a filtering layer method or a template-like MMV one. The filtering layer placed between From- and To-MMVs enables the selective transfer from wells for which the filtering layer is open (penetrable). On the other hand, a template MMV can hold samples in those wells which were made open when the MMV was fabricated and thus can transfer samples selectively from those wells (see Fig. 3c).
c) Measuring up the volume
This can be done as an application of addition process (see above) if the size of the well of a donor-MMV is made unitary (of a fixed volume). N-times of addition make N units of volume input into a recipient-MMV.
d) Mixing
Although such a tiny volume as sub-μL is rather favorable for diffusion and thermal conductance, it is very unfavorable for mixing solutions with a vortex. For this purpose, we made a target molecule (peptide) bound on a magnetic bead to lift up and drop down by magnetic force. In other cases, stirring the solution with vigorous moving of small steel balls (~0.4 mm in diameter; Super micro-ball generously provided by Toyo Seiko Co., Japan) by magnetic force was also applied, of which effect was confirmed by diffusion of dye.
e) Incubation (cultivation/enzymatic reaction)/PCR
In case of culturing of bacteria, an MMV plate was equilibrated by soaking with LB broth or Davis medium prior to the culture and a container with a large space (more than 18 cm3) equipped with a water reservoir was selected. For a PCR reaction, the MMV plate was placed in a small container (5.4 cm3) firmly sealed to avoid an evaporation loss during incubation and held in a chamber thermally controlled (Mastercycler gradient, Eppendorf Co., Germany). PCR was carried out in a special device for the sake of effective heat conductance and minimum evaporation of water (see Additional file 1). The PCR program adopted was as follows: pre-heating, 94°C, 2 min; denaturing, 94°C, 0.5 min; annealing, 55°C, 1 min; extension, 72°C, 1 min (35 cycles); post extension, 72°C, 5 min.
f) Separation into a solid and a liquid parts
By attracting magnetic beads upward out of the liquid with a magnet, the solid (beads) and the liquid were easily separated (see Fig. 3d).
g) Image processing (fluorescence/optical light)
A newly made MMV was treated to equilibrate with a permeation solution (50 mM sodium acetate, 0.1 M NaCl, pH 4.5) containing 5 μM fluorogenic substrate of cathepsin E (CE) MOCAc-Gly-Lys-Pro-Ile-Ile-Phe-Phe-Arg-Leu-Lys(DnP)-D-Arg-NH2 (Code 3200-V, Peptide Institute, Inc., Osaka), and was half-filled with a CE reaction solution (permeation solution containing 5 μM substrate and 5 pmol CE additively). This reaction mixture was transferred to a translation MMV by centrifugation (1500 rpm, 20 s). The mixture was incubated at 37°C for 10 min with a shaking incubator Bioshaker V.BR-36 (TAITEC, Saitama). Then, the fluorescence of the product in the MMV was measured with a fluoroimager GelDoc XR (BioRad, USA) using 0.008% K2CrO4 aqueous solution as an excitation filter (320-340 nm) and a glass plate as an emission filter (<360 nm cut-off).
h) In vitrotranscription and translation using an MMV
A wet MMV was treated to equilibrate with a permeation buffer for reverse transcription (13 mM magnesium acetate, 50 mM Hepes-KOH, 100 mM potassium glutamate, 20 mM creatine phosphate, 2 mM spermidine, 1 mM dithiothreitol, 2 mM ATP, 2 mM GTP, 1 mM CTP, 1 mM UTP, and 0.3 mM amino acid mix, pH 7.6). For in vitro transcription and translation, Wako pure system (Wako, Tokyo) was used and the reaction mixture was prepared following the manufacturer's instructions. Streptavidin-coated magnetic beads (TaKaRa, Kyoto) suspended in 80 μL (0.4 mg) and pre-washed repeatedly with 100 μL of water and the binding solution (TaKaRa, Kyoto) were combined and then added to every well of the MMV to half a well volume by manually pouring or controlled centrifugation. The MMV subjected to PCR (PCR MMV) was spun to precipitate solutions to the bottom before the following operation. In rare cases, some wells were readjusted to half a volume of the well with the PCR solution. Then, the PCR MMV was combined face-to-face to a translation MMV fitting the apertures in a correct phase with the 4 corners pinned and fixed. PCR solutions were transferred into each well of the translation MMV by centrifugation (1500 rpm, 20 s). The reaction solutions were stirred by moving magnetic beads up and down by attracting with a neodymium magnet (20 times). The mixtures were then incubated at 37°C for 2 h with further stirring at 20 min intervals as described above.
Polymerase Chain Reaction (PCR) in MMV
A PCR mixture was prepared as recommended by manufacture's instruction (SpeedSTAR HS DNA polymerase TaKaRa, Kyoto), added optionally with an enzyme stabilizer (Lipidure BL-802 (NOF Corp. Tokyo)), and input into wells of a wet-type MMV which had been equilibrated in advance with PCR buffer containing PCR components except template DNA, Taq polymerase, and primers. The MMV was placed within a PCR container which was hand-made from silicon rubber and stainless-steel sheets (see Additional file 1). The container was placed in a thermal cycler and PCR amplification was performed with a program (pre-denaturation, 94°C, 2 min; denaturation, 94°C, 1 min; annealing, 55°C, 1 min; extension, 72°C, 1 min (35 cycles); and post-extension, 72°C, 5 min). The MMV plate was removed from the container and the content was recovered after centrifugation. The DNA amplification with this system was confirmed by gel-electrophoresis and silver staining (see Additional file 1).
Verification of the MMV transfer operation
To demonstrate the effectiveness of the well-to-well transfer and successive reactions, we constructed and tried a model experiment. A checker-patterned MMV was filled with a PCR solution containing 2 fmole/μL of DNA coding GFP as templates, respectively. SpeedSTAR polymerase (TaKaRa, Kyoto) was used for a rapid amplification and the PCR program was utilized as follows: pre-denature, 94°C, 2 min; denature, 94°C, 20 s; annealing and extension, 1 min (25 cycles); and post-extension, 72°C, 2 min. After PCR procedures (see 'MMV operations' in Methods), the contents of the MMV were transferred to another MMV which was partly filled with a cell-free transcription/translation solution (see 'MMV operations' in Methods). Then, this MMV was incubated at 37°C for 1 h. The fluorescent image was monitored with a fluoroimager and the checkered fluorescent image was obtained. The consistence between the PCR well-pattern and the fluorescent image of GFP verified the fidelity of MMV transfer operations (see Fig. 4).
2NMethod
In the case of micro-arrays which have a 2N × 2N square well-pattern ("N" designates a natural number), its whole diversity (22N) can be generated by employing "2N" kinds of quite different (orthogonal) elements. "2N" kinds of elements can be easily prepared by using "2N" sheets of filters (by considering the symmetrical nature of filters, "N" is sufficient) which have rather simple well patterns. In the case of "N" = 2 (see Fig. 5) for example, 2 sheets of filter are sufficient to be prepared and each filter can be used in two ways, working virtually as four template plates. Four kinds of elements are transferred into a recipient microarray by centrifuge. Finally, all of the 16 wells of the microarray would have different constituents from each other. Each well can be uniquely assigned by the binary number. If we regard the digit '1' as 'exist' and '0' as 'non-exist' and if we consider the difference of the order in the number corresponds to the difference of elements in the actual experiment, then the well assigned 1001 (9 in the decimal number) should contain the elements corresponding to the elements #1 and #4 (#1-#4 corresponds to the MSB, 2nd MSB, 3rd MSB and LSB). Here, we utilized a 256-well micro-array ("N" = 4) for examining condition on re-crystallization of lysozyme (see Fig. 2f where we can confirm that the pairs of T0/T4, T1/T5, T2/T6, and T3/T7 are symmetrical).
Selection of Aβ aptamers
The DNA construct for in vitro transcription/translation (see Fig. 9b) was amplified using the following primers in the PCR reaction: P1, 5'GATCCCGCGAAATTAATACGACTC ACTATA3'; and P2, 5'GGCTCGCGAATACTGCGAAGGAGTGAGATC3'. A part of the translation mixture and the magnetic beads were removed from the translation MMV using a magnet. On the other hand, Aβ peptides were trapped on magnetic beads as follows: an aliquot containing 50 μL Magnotex SA particle (TaKaRa, Kyoto) was mixed with biotin-labeled Aβ peptide solution (5 μg/50-500 μL) in a binding buffer. After incubation for 20 min at room temperature, the beads were washed and re-suspended with 500 μL PBS buffer (pH 7.0). For each selection round, 2 mg of beads, which have approximately 500 pmoles of bound Aβ (1-42), was used. Magnetic beads-Aβ conjugates were inputted into a fresh MMV to make each well containing 0.050 pmoles (0.5 μL) of the conjugate. Then, the conjugate suspension was spun down into a translation MMV and the solution in the wells was mixed by moving the magnetic beads up and down with a magnet. The suspension was incubated at 37°C for 1 h with several 20 min-interval mixings. With fixing the magnetic beads with a magnet on the bottom of well, the solution was transferred to a fresh MMV by centrifugation. The fluorescence of GFP bound on beads was monitored with a fluoroimager. After selection of Aβ-binding peptides by fluorescence intensity, the separated supernatants were used as a template solution for PCR for the next round of selection and DNA sequencing to confirm the Aβ-binding peptides.
Library construction and selections
a) Library construction
Combinatorial DNA and IVV (in vitro virus)-peptide libraries were constructed according to the previously reported method of YLBS [35] and cDNA display [36].
b) Preparation of CE-immobilized beads used for selection
Purified cathepsin-E (CE) was immobilized on NHS-activated sepharose beads (GE Healthcare, USA) by using the amine coupling chemicals to form a chemically stable amide bond in accordance with the manufacturer's instructions. The enzyme-coupled beads were stored in 100 mM phosphate buffer (pH 7.4) at 4°C until further use. The coupling efficiency was calculated by comparing the absorbance of uncoupled enzyme with that of free enzyme.
c) Affinity-based selection
The IVV-peptide library in 100 μL of Selection buffer (50 mM Tris-HCl, 100 mM NaCl, 5 mM MgCl2, pH 7.2) was mixed with 5 μL of CE-immobilized beads and incubated at 25°C for 30 min. The beads were washed with 200 μL of Selection buffer, Washing buffer-1 (50 mM Tris-HCl, 0.5 M NaCl, pH 7.2) and Washing buffer-2 (50 mM Tris-HCl, 1 M NaCl, pH 7.2) in accordance with the protocol of washing repeat count, which varied along the selection round, i.e., 2:2:1 in Round 1; 2:2:2 in Round 2; 3:3:3 in Round 3. Finally, the beads were suspended in 200 μL of Elution buffer-1 (50 mM Tris-HCl, 1 M NaCl, 10 mM MgCl2, pH 7.2) and incubated at 37°C for 5 min. After centrifugation (1500 rpm, 1 min), the beads were washed again with Elution buffer-1. The combined supernatant was stored (Sup1). The beads were suspended with Elution buffer-2 (50 mM Tris-HCl, 2 M NaCl, 10 mM MgCl2, pH 7.2) and incubated at 95°C for 5 min. After incubation, the beads were removed and the supernatant was stored (Sup 2). The IVVs in Sup1 and Sup 2 were purified with a Bio-Spin column (Bio-Rad Laboratories, USA).
d) Function-based selection
The DNA library resulting from Round 3 of the affinity-based selection was inserted in the IVV-SF-link DNA construct as a variable region [36]. The IVVs were prepared and incubated as described in affinity-based selection. The protocol of washing repeat count with Selection buffer, Washing buffer-1 and Washing buffer-2 was carried out according to the selection round program (F1 to F9) as follows: 3:3:3 in F1, 4:5:3 in F2, 5:5:5 in F3, 5:7:5 in F4, 5:7:7 in F5, 5:10:7 in F6, 5:10:10 in F7, 5:15:10 in F8, and 5:15:15 in F9.
Sample preparation
a) GFP-expressing Escherichia coli
The buffer components in a gel-MMV plate (1 × SSC) were exchanged with LB broth and the affluent solution in wells was removed by centrifugation. Overnight culture of E. coli cells harbouring green fluorescence protein (GFP) was diluted with LB broth to the concentration of one cell per well (0.5 μL) and then put into wells of an MMV by centrifugation. The MMV was layered on a wet tissue paper in a petri dish and subjected to the incubation at 37°C. The fluorescence of GFP expressed in E. coli cells was monitored with a fluoroimager, Molecular Imager FX (Bio-Rad Laboratories, USA).
b) A DNA library for cathepsin E-inhibitory peptides
A DNA library was constructed, of which DNA encodes a different species of a cathepsin E-binding peptide (8-28 amino acids in size) which is derived from the preceding in vitro selection experiment [36] (see ' Library construction and selection' and Fig. 11).
c) Lysozyme solutions
Lysozyme provided in lyophilized powder (protein content, ~95%; ~50,000 units/mg protein (Sigma-Aldrich, Germany)) was dissolved in distilled water (250 mg/mL) using a vortex mixer with a special care so as not to leave visible flocs and then spun down. The supernatant was used for the crystallization experiment.
Generation of combinatorial conditions (2Nmethod)
We applied the 2N method (see '2N Method' and Fig. 5) using a 256-well MMV to generating 256 species of conditions aimed for crystallization of lysozyme. The ionic strength and pH were modulated using the following NaCl and sodium acetate-HCl buffers, respectively. Each template MMV of T0~T7 was charged with 0.1 μL of 2 M NaCl (T0), 3 M NaCl (T1), 4 M NaCl (T2), 5 M NaCl (T3), pH 3 (T4), pH 4 (T5), pH 5 (T6), and pH 7 (T7), respectively. To make the final solution volume equal in each well, the complementary volume of water was added to those wells which were not charged with any of T0~T7 solutions, using the complementary templates T0~T7 (well/non-well relations inverted). Finally, a lysozyme solution (250 mg/mL) was added (0.1 μL). The MMV was covered with a piece of transparent adhesive tape Titer Stick (Wakenyaku, Japan) to prevent evaporation and was kept at 20°C for several days. Each well was monitored by an inverted microscope IM (Olympus, Tokyo).
Multistep operations of MMVs
a) In vitroselection of Aβ-binding peptides
The whole procedure is described in detail above, in which the construct and the selection of a peptide-fused GFP is described (see Fig. 9). Peptides were expressed by successive reactions of PCR, in vitro transcription and in vitro translation as written above. Then, Aβ protein (Aβ42) linked to a magnetic bead was added by an MMV-transfer (i.e., transfer from an MMV to another MMV by centrifugation). Those peptides bound to Aβ were separated from the remaining unbound ones by the next MMV-transfer (see 'MMV operation'), washed with a buffer (10 mM Tris-HCl, 1 mM EDTA, 1 M NaCl, 0.1% NaN2), and then subjected to the fluorescence monitoring using fluoroimager Molecular Imager FX (Bio-Rad Laboratories, USA) at the excitation wavelength of 488 nm.
b) In vitro selectionof cathepsin E-inhibitory peptides
A set of peptides which have a cathepsin E-inhibition activity selected in the preceding study [34] were further screened using MMVs (see Fig. 10). The DNA molecules encoding those peptides were diluted and input in an MMV (~50 copies per well) and then subjected to PCR, in vitro transcription/translation, and protease Xa digestion. Then, the magnetic beads were removed from the MMV. The remaining solutions were used to detect the inhibition activity of the peptide in each well. The volume of the reaction mixtures were, if too much, reduced to around a half-well volume by sucking with filter paper. The measurement of the inhibition activity in each well was carried out as described in Methods (Image processing).
