Screening Complex Biological Samples with ... - ACS Publications

Oct 12, 2016 - immobilization strategy in the screening of Burkholderia cepacia complex (BCC) infections in patients affected by cystic fibrosis (CF)...
0 downloads 0 Views 1016KB Size
Subscriber access provided by United Arab Emirates University | Libraries Deanship

Article

Screening Complex Biological Samples with Peptide Microarrays: the Favorable Impact of Probe Orientation via Chemoselective Immobilization Strategies on Clickable Polymeric Coatings Alessandro Gori, Laura Sola, Paola Gagni, Giulia Bruni, Marta Liprino, Claudio Peri, Giorgio Colombo, Marina Cretich, and Marcella Chiari Bioconjugate Chem., Just Accepted Manuscript • DOI: 10.1021/acs.bioconjchem.6b00426 • Publication Date (Web): 12 Oct 2016 Downloaded from http://pubs.acs.org on October 14, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Bioconjugate Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

Screening Complex Biological Samples with Peptide Microarrays: the Favorable Impact of Probe Orientation via Chemoselective Immobilization Strategies on Clickable Polymeric Coatings Alessandro Gori, Laura Sola, Paola Gagni, Giulia Bruni, Marta Liprino, Claudio Peri, Giorgio Colombo, Marina Cretich* and Marcella Chiari Consiglio Nazionale delle Ricerche, Istituto di Chimica del Riconoscimento Molecolare (ICRM) Via Mario Bianco, 9, 20131, Milano, Italy. *

To whom correspondence should be addressed:

Dr. Marina Cretich Istituto di Chimica del Riconoscimento Molecolare, ICRM, CNR Via Mario Bianco 9, I-20131 Milan, Italy. Tel: +39 02 2850 0042; Fax: +39 02 2890 1239; E-mail: [email protected]

Abstract The generation of robust analytical data using microarray platforms strictly relies on optimal ligand-target interaction at the sensor surface which, in turn, is inherently bound to the correct immobilization scheme of the interrogated bioprobes. In the present work, we performed a rigorous comparative analysis of the impact of peptide ligands immobilization strategy in the screening of Burkholderia cepacia complex (BCC) infections in patients affected by cystic fibrosis (CF). We generated arrays of previously validated Burkholderia-derived peptide probes, that were selectively oriented on polymeric coatings by means of different click-type reactions including thiol-maleimide, copper-catalyzed-azide-alkyne-cycloaddition (CuAAC) and strain-promoted-azide-alkyne-cycloaddition (SPAAC). We compared immobilization efficiency among the different chemoselective reactions and we evaluated diagnostic performances at a statistically significant level, also in contrast to random immobilization strategies. Our findings clearly support the favorable role of correct bioprobes orientation in discriminating seronegative from infected individuals and, in last analysis, in generating more reliable and reproducible data. Spacing biomolecules from the sensor surface by means of small hydrophilic linkers also positively affects the analytical performance and leads to increased statistical significance of data. Overall, all the click immobilization strategies that were considered displayed a good efficiency. Interestingly, SPAAC-mediated conjugation using DBCO cyclooctyne for some peptides resulted in sequence-dependent auto-fluorescence in the Cy5 emission range wavelength, which could be circumvented by using a different fluorescence detection channel. On the basis of our results, we critically discuss the immobilization parameters that need to be carefully considered for peptide ligands immobilization purposes. Keywords: Microarray; Polymer coating; Click chemistry; Peptide bioconjugation; Sera screening; Diagnosis; Infectious Disease; Cystic Fibrosis

1 Environment ACS Paragon Plus

Bioconjugate Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction By allowing unprecedented rapid and multi-parallel screening of biofunctional ligand libraries, the advent of microarray technology has introduced a new paradigm in the scenario of high-throughput analysis 1, 2. In the array format, where molecular probes are immobilized to discrete areas of a sensor surface, dozens to thousands of biomolecules can indeed be simultaneously tested in a single assay. Since peptides have been progressively unrevealed as key functional modulators of biological processes, peptide microarrays have witnessed an ever-increasing popularity in recent years

3, 4, 5, 6

. Likewise, their synthetic accessibility and

ease of manipulation have equally contributed to make peptides such appealing tools for the development of new biosensors3,7,8. Bioprobe immobilization onto the sensor surface is probably the most crucial step in microarrays manufacturing. Indeed, an effective and reproducible immobilization strictly accounts for the reliability and the consistency of the generated analytical data. In particular, it is of utmost importance to preserve the functional properties of the immobilized biomolecule while ensuring its stable binding on the sensor surface throughout the experimental procedures. Traditional schemes for ligands immobilization relies on noncovalent random absorption of the probe on the sensor surface or, alternatively, on aspecific covalent binding, mainly involving amine and sulfhydryl groups from lysine and cysteine residues9–12. However, non specific immobilization of (peptide) ligands can result in heterogeneous presentation of the bioprobe; as a result, the analytical performance can be plagued either by low reproducibility, poor signal-to-noise ratio and high non-specific binding. Oppositely, chemoselective approaches to immobilize peptide ligands in a spatially-oriented and uniform fashion can, intuitively, prove highly beneficial to guarantee reproducible and optimal ligand exposure, thus improving affinity for the capturing targets and, ultimately, the analytical accuracy. In this scenario, the so-called “click” reactions represent valuable tools for the selective and directional conjugation even of highly functionalized molecules, such as proteins and peptides. Click reactions are indeed, by definition, characterized by high efficiency and chemoselectivity, together with fast reaction kinetics13. The click chemistry toolbox encompasses a wide range of chemoselective and bio-orthogonal reactions, whose distinctive features have been extensively reviewed14. For instance, prototypical click sulfhydryl-malemide conjugation exploits the nucleophilic character of cysteine thiols for the addition to a maleimide group that acts as a Micheal acceptor, generating a thioether bond. The selectivity of the reaction is pronounced, thiol groups are indeed much better Micheal donors than potentially competing free amine groups. Moreover, the reaction usually proceeds smoothly and fast kinetics under benign reaction conditions. Likewise, copper-catalyzed cycloaddition of azides to alkynes (CuAAC) generally proceeds with almost quantitative yields and excellent reaction kinetics. As the azido and alkyne groups involved in the cycloaddition reaction are non-native chemical handles, the reaction displays nearly full orthogonality with respect to proteins functional groups. Additional favorable properties of this reaction are the wide tolerance to different reaction conditions and the inertness of the generated triazole ring in the biological media. To 2 Environment ACS Paragon Plus

Page 2 of 18

Page 3 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

overcome limitations due to residual copper catalyst, particularly for in vivo application, strain-promoted azide-alkyne cycloaddition (SPAAC) has been developed as a catalyst-free process, where the driving force of the reaction is given by the intramolecular strain of the cyclooctyne used as the azido counterpart. Click reactions suitability to immobilize bio-ligands on microarray surfaces is now well established15, yet, to the best of our knowledge, literature is still lacking of a rigorous and critic comparative evaluation of click reaction performances in terms of binding efficiency and influence of the immobilization chemistry on the diagnostic accuracy in a serological assay. Similarly, the comparison between random vs oriented immobilization under a quantitative perspective is still to be reported, optimization of chemoselective immobilization is often restricted to one type of bioconjugation strategy 16, 17, 18, 19 and other parameters such as the effect of spacing peptide ligands from microarray surface by means of suitable spacers 5 has been so far only intuitively discussed and mostly for proteases assays 20. To provide information rapidly applicable into diagnostics, we set this investigation in the context of a real clinical assay, using a collection of previously validated peptides for the diagnosis of Burkholderia infections21,22,23,24 for the screening of Burkholderia cepacia complex (BCC) in patients affected by cystic fibrosis (CF). To this end, among a set of possible click reactions, we identified some of the most widely used, i.e. sulfhydryl-malemide conjugation and azide-alkyne-cycloaddition in both the copper-catalyzed variant (CuAAC) and the strain-promoted version (SPAAC), and applied them to the immobilization of a peptide collection on microarray surfaces conveniently modified with suitable polymeric coatings originated from the common precursor Copoly(DMA-NAS-MAPS)25. This polymer, made of N,N-dimethylacrylamide (DMA), N-acryloyloxysuccinimide (NAS), and 3(trimethoxysilyl)-propylmethacrylate (MAPS) has been extensively used as a functional coating in microarrays 26,27,28 and various biosensing applications 29, 30, for its favorable characteristics in terms of simplicity of use, high binding capacity and excellent anti-fouling properties. To introduce new functionalities for orientation control, the active esters of Copoly(DMA-NASMAPS) have been reacted with amines bearing functionalities suitable for click reactions in order to generate a family of polymeric coatings 31 for orthogonal bioprobes immobilization. Using this clickable polymeric platform in the context of fluorescence peptide microarrays, we set to evaluate: 1) the influence of random vs oriented immobilization of the peptide probes; 2) the result of spacing ligands from the polymer surface by means of short hydrophilic linkers; 3) the effect of different immobilization chemistries on binding efficiency and signal detection.

3 Environment ACS Paragon Plus

Bioconjugate Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Results and Discussion “Clickable” polymeric platform In order to enable peptides conjugation onto microarray surface, we exploited a versatile polymeric platform obtained by modifying the same polymer precursor, Copoly(DMA-NAS-MAPS) 31. The polymer Copoly(DMA-NAS-MAPS) bears active esters that easily react with nucleophiles such as amines, thus offering the possibility of introducing new functionalities suitable for click reactions. Specifically, Copoly Maleimide was obtained by N-(2-aminoethyl)maleimide trifluoroacetate salt, Copoly Alkyne by propargylamine, and Copoly Azide by azido-1-propanamine (Scheme 1). The four polymers are used to coat microarray silicon chips by an easy dip-and-rinse protocol. Copoly Maleimide, Copoly Alkyne and Copoly Azide were used for selective conjugation of the peptide pool by click reactions whereas Copoly(DMA-NASMAPS) was used for non-directional peptide attachment (Scheme 1).

Scheme 1: Modifications of the polymer precursor Copoly(DMA-NAS-MAPS) to generate a "clickable" polymeric platform for peptide binding on microarrays. a) N-(2-aminoethyl)maleimide trifluoroacetate salt / DIPEA; b) propargylamine; c) 3-azido-1-propanamine. Resulting coated slides were then reacted with respective click counterparts to yield peptide-functionalized microarray surface.

4 Environment ACS Paragon Plus

Page 4 of 18

Page 5 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

Peptide synthesis All peptides were assembled on a 2-CTC resin by a fully automated microwave-assisted Fmoc-SPPS protocol32 (See Experimental Section). For each peptide, upon complete assembly of the amino acidic sequence, resin was split in order to allow divergent modification with diverse functional groups at the Nterminal region (Figure 1). Specifically, peptides were produced as: 1) unmodified (P1a-P5a) for random immobilization on Copoly(DMA-NAS-MAPS) coated chips via aspecific covalent binding, mainly involving amine and sulfhydryl groups from lysine and cysteine residues; 2) cysteine functionalized (P1b-P5b) and 3) extended with short-chain PEG spacers (O2Oc) bearing a terminal cysteine (P1c-P5c) for oriented immobilization on Copoly Maleimide via thiol-maleimide reaction; 4) functionalized with cyclooctyne (DBCO) (P1d-P5d) for subsequent immobilization on Copoly Azide exploiting SPAAC; 5) modified with short-chain PEG spacers (O2Oc) bearing a terminal propargylglicine (P1e-P5e) or azidoalanine (P1f-P5f) for oriented immobilization via CuAAC on chip surfaces coated by Copoly Azide and Copoly Alkyne respectively.

Figure 1. Peptidic probe sequences (P1-P5) and N-terminus modifications introduced for different immobilization strategies.

Peptide immobilization We investigated the immobilization yields of each peptide on the corresponding clickable polymeric surface using the IRIS (Interferometric Reflectance Imaging Sensor) platform33. In this technique, the microarray can be imaged before the fluorescence assay to provide, by the use of previously determined calibration factors, a quantification of the amount of probe mass that is bound at each spot location. Binding yields for each peptide-surface conjugation are reported in nanograms/square millimeter in Figure 2.

5 Environment ACS Paragon Plus

Bioconjugate Chemistry

3,5 3 Unmodified

2,5

SH ng/mm2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 18

2 PEG-SH 1,5

PEG-N3 PEG-yne

1

PEG-DBCO 0,5 0 P1

P2

P3

P4

P5

Figure 2: binding efficiency (ng/mm2) for each peptide and immobilization strategy, assessed by IRIS labelfree platform. Unmodified peptides, spotted on Copoly(DMA-NAS-MAPS) provided the lowest immobilization yield. The introduction of chemoselective groups and the use of a orthogonal click strategies enhanced binding efficiency for the whole set of probes. Standard Deviation is indicated by the error bars.

In general, given a peptide spotting concentration of 2 mg/mL and a common optimized printing buffer, the amount of bound probe ranges from 0.1 to 3.2 ng/mm2 depending on peptide sequence and conjugation strategy. Unmodified peptides, immobilized randomly on Copoly(DMA-NAS-MAPS), provided the lowest binding as compared to the chemoselective strategies. The mere addition to the peptide sequences of a terminal cysteine residue and its selective reaction with Copoly Maleimide surface remarkably improved the binding efficiency; in contrast, the use of a PEG linker did not show any effect on the efficiency of the sulfhydryl-maleimide conjugation. As for the azide-alkyne-cycloaddition, both copper catalyzed and copper free variants provided improved reactivity over sulfhydryl-maleimide conjugation except for peptide P5. Specifically, the three peptides P2e, P3e and P4e modified by -yne group resulted in the highest immobilization yield on the Copoly Azide surface. The copper free SPAAC conjugation via DBCO modified probes exhibited the most uniform binding among the different peptides (P1d-P5d). This "equalizing" effect among the bio-probes is even more evident looking at the number of peptide molecules bound per surface unit (Supporting Information Figure 1-S) and is a convenient feature of this conjugation strategy. In large arrays of bioprobes differing each other for binding vocations and physicochemical properties, surface binding uniformity is indeed a critical issue, especially in clinical applications of microarrays where multiplex quantitative data are required.

6 Environment ACS Paragon Plus

Page 7 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

We found that, for click based conjugations, a peptide spotting concentration of 2 mg/mL (in the range of 1 mM for the whole set of peptides) saturated the surface available binding sites for all polymeric surfaces tested; increasing peptide concentration in the spotting solution not only was ineffective in increasing bound mass but, for copper catalyzed reactions, negatively affected binding reproducibility and yield. The reproducibility of probe binding was evaluated for each conjugation strategy. The intra-chip and interchip CV% are reported in the Supporting Information (Figure 2-S and Figure 3-S). In this study, except for P5 a dramatic effect of the aspecific reactivity of unmodified peptides on Copoly(DMA-NAS-MAPS) was observed on binding reproducibility. In contrast, any introduced chemoselective strategy drastically cut down the variability in peptide binding, reducing the CV to 10-20% and lower, values commonly observed in microarray experiments. Notably, sulfhydryl-maleimide conjugation, using cysteine-PEG linker resulted in the most reproducible immobilization strategy among the ones here investigated. In this work, for consistency, we have used a common spotting buffer (25 mM Na/Acetate pH 4.8, 15 mM trehalose) for all peptides and binding strategies except for the presence of 100uM CuSO4, 400uM THPTA and 6.25 mM Ascorbic Acid in CuACC based conjugation. We confirmed that no significant oxidation of the peptide probes occurs during CuAAC immobilization process by running control experiments in solution (see Supporting Information). The acidic buffer was selected to facilitate the solubility of all peptides. Indeed, dissolution of unmodified peptides at a neutral or alkaline buffer provided sub-optimal solubility for some probes and, in any case, didn't improve the binding mediated by amino groups on the Copoly(DMA-NASMAPS) surface (data not shown). In this sense, the wide tolerance to any pH condition of click chemistry reactions is of great value for peptide handling in the context of microarray immobilization. Experimental Section summarizes the conjugation conditions. Beside conjugation efficiency, the label-free array imaging provided by IRIS platform (upper panel of Figure 3) allowed us to evaluate the morphology of the peptide spots after binding. In general, the spot diameters and their homogeneity vary, depending on peptide sequence and immobilization method. P1, for example, generates small spots in any condition except when modified by DBCO. It is known that printing buffers strongly influence the spot morphology: detergents and additives such as glycerol or sugars can control the spreading of the droplets and facilitate an even evaporation of the spot. The presence of copper catalyst in CuACC based conjugation slightly affected the homogeneity of some spots. On the whole, P1c-P5c on Copoly Maleimide provided the most homogeneous spots in terms of diameter and morphology. We have also observed that the presence of the PEG linker is in general beneficial for spot morphology (data not shown).

7 Environment ACS Paragon Plus

Bioconjugate Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3: imaging of the peptide microarray by label-free IRIS platform (upper panel) and by fluorescence after completion of a bioassay (lower panel). All peptide microarrays were analyzed in the Cy5 channel except for the DBCO-modified peptides that were scanned for fluorescence in the Cy3 channel. Spot morphology varies depending on peptide sequence and binding chemistry. On the whole, cysteine-PEG modified peptides on Copoly Maleimide provided the most homogeneous spots.

Negative controls to exclude aspecific peptide adsorption and demonstrate the role of functional polymer groups for covalent binding were run as follows: unmodified peptides dissolved in the printing buffer for CuACC conjugation were spotted on slides coated by Copoly Azide and Copoly Alkyne and on slides coated by Copoly(DMA-NAS-MAPS) and Copoly Maleimide previously blocked by ethanolamine and beta mercaptoethanol respectively (See Experimental Section). The chips were imaged by IRIS immediately after spotting and upon overnight incubation and washing. A minimal binding was detected for p5 on Copoly Azide and Copoly Maleimide and for p4 and p5 on Copoly Alkyne; the other peptide spots were undetactable. Overall these results demonstrate a negligible role of aspecific adsorption of the peptide sequences on the polymer surfaces. Similarly, spots of the printing buffers not containing peptides were undetactable confirming no artefacts in peptide spot mass quantification (Figure 4-S in the Supporting Information).

8 Environment ACS Paragon Plus

Page 8 of 18

Page 9 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

Peptide microarray analytical performance Peptide microarrays produced using the different conjugation strategies discussed above were compared in the screening of Burkholderia cepacia complex (BCC) infections in patients affected by cystic fibrosis (CF). Peptide sequences were previously validated as diagnostic probes for Burkholderia infections; the peptidic epitopes are used to specifically capture human IgG in patients affected by BCC thus revealing the presence of the infection. Our aim was to test different click-based immobilization strategies and to verify the role of immobilization parameters, also in contrast to random immobilization, in the real diagnostic context of the discrimination of seronegative vs seropositive patients. The arrays were probed with 12 serum samples from BCC positive CF patients (diagnosed by microbiological culture and PCR) and 12 healthy donors. The peptide specific IgG content in each sample was evaluated by fluorescence detection using Alexa-fluor 647 labelled anti-human IgG except for the DBCO modified peptide arrays where Cy3 labelled anti-human IgG was used. Throughout the experiments we have indeed observed peptide-dependent auto-fluorescence in the Cy5 fluorescence channel (Excitation 635 nm; Emission 676/37 nm) for the DBCO probes (see details and Figure 5-S in Supporting Information). It is worth underlining that this effect, to the best of our knowledge, has not been previously reported and that this should draw particular attention in the use of DBCOfunctionalized probes, as it could result in misleading false positive. Given the peptide-dependent nature of the recorded auto-fluorescence, we speculate that the effect may be due to the peculiar chemical environment that surrounds the DBCO moiety in the context of different peptide sequences. However, a closer investigation on the origin of this effect was out of the scopes of the current work and it is currently on-going in our lab. A representative fluorescence result obtained with a BCC positive serum on the PEG-modified peptide arrays immobilized on clickable surfaces is reported in Figure 3 (lower panel). In general, CF patients positive for BCC showed patterns of fluorescence signals (corresponding to seroreactive peptides) that were significantly different from patterns detected in healthy individuals, mostly providing lower fluorescence signals. The ability of each peptide to capture specific antibodies, thus distinguishing between controls and patients, was evaluated performing the unpaired t Test (significative p