High-Throughput Chip Assay for Investigating Escherichia coli

Aug 14, 2018 - Statistical analyses were performed with Graphpad Software. ..... of E. coli penetration of the blood–brain barrier, top 23 proteins ...
4 downloads 0 Views 2MB Size
Subscriber access provided by AUSTRALIAN NATIONAL UNIV

Article

A Unique High Throughput Chip Assay for Investigating Escherichia coli Interaction with the Blood-Brain Barrier Using Microbial and Human Proteome Microarrays (Dual-Microarray Technology) Yingzhu Feng, Chien-Sheng Chen, Jessica Ho, Donna Pearce, Shaohui Hu, Bochu Wang, Prashant Desai, Kwang Sik Kim, and Heng Zhu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b02513 • Publication Date (Web): 14 Aug 2018 Downloaded from http://pubs.acs.org on August 15, 2018

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 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 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.

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 41 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

Analytical Chemistry

A Unique High Throughput Chip Assay for Investigating Escherichia coli Interaction with the Blood-Brain Barrier Using Microbial and Human Proteome Microarrays (Dual-Microarray Technology) Yingzhu Feng1,3,5, Chien-Sheng Chen2,3,7, Jessica Ho3, Donna Pearce4, Shaohui Hu3, Bochu Wang1, Prashant Desai6, Kwang Sik Kim4*, and Heng Zhu3* 1

Key Laboratory of Bio-theological Science and Technology of Ministry of Education, College of

Bioengineering, Chongqing University, Chongqing 400030, PR China. 2

Department of Food Safety/Hygiene and Risk Management, Tainan City 701, Taiwan.

3

Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine,

Baltimore, Maryland 21205, U.S.A. 4

Division of Pediatric Infectious Diseases, Johns Hopkins University School of Medicine, Baltimore, Maryland

21287, U.S.A. 5

School of Life Sciences, Sun Yat-Sen University, Guangzhou 510275, China.

6

The Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine, Baltimore,

Maryland 21231, U.S.A. 7

Department of Biomedical Science and Engineering, National Central University, Taoyuan City 32001,

Taiwan.

ACS Paragon Plus Environment

Analytical 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

*Corresponding author: Heng Zhu and Kwang Sik Kim E-mail: Heng Zhu ([email protected]); Kwang Sik Kim ([email protected])

Keywords: Pathogen-host interaction / High throughput chip assay / chip-based cell probing assasy

Blood Brain Barrier (BBB) / E. coli proteome microarray / Human proteome microarray

ABSTRACT Bacterial meningitis in neonates and infants is an acute lethal disease and occurs in response to microbial

exploitation of the blood brain barrier (BBB), resulting in the intracranial inflammation. Several pathogens,

such as Escherichia coli (E. coli), can cause this devastating disease; however, the underlying molecular

mechanisms by which these pathogens exploit the BBB remain incompletely understood. To identify important

players on both the pathogen and host sides that govern the E. coli-BBB cell interactions, we took advantage of

the E. coli and human proteome microarrays (i.e., HuProt) as an unbiased, proteome-wide tool for identification

of important players on both sides. Using the E. coli proteome microarrays, we developed a unique high

throughput chip-based cell probing assay to probe with fluorescent live human brain microvascular endothelial

cells (HBMEC, which constitute the BBB). We identified several transmembrane proteins, which effectively

bound to live HBMEC. We focused on YojI protein for further study. By probing the HuProt arrays with YojI,

interferon-alpha receptor (IFNAR2) was identified as one of its binding proteins. The importance of YojI and

IFNAR2 involved in E. coli-HBMEC interactions was characterized using the YojI knockout bacteria and

2

ACS Paragon Plus Environment

Page 2 of 41

Page 3 of 41 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

Analytical Chemistry

IFNAR2-knock down HBMEC and further confirmed by E. coli binding assay in HBMEC. This study

represents a new paradigm (dual-microarray technology) that enables rapid, unbiased discovery of both

pathogen and host players that are involved in pathogen-host interactions for human infectious diseases in a

high throughput manner.

INTRODUCTION Bacterial meningitis is an important cause of mortality and morbidity in neonates and children[1,2]. The incidence of neonatal E. coli meningitis is estimated at 0.01 to 0.05 per 1,000 live births[3]. The mortality of

this serious disease could be as high as 10% ~ 30% in infants, while infection by multi-drug resistant bacteria could cause the mortality rate even higher[1-3]. Almost all microbes that are pathogenic to humans have the potential to cause meningitis, but a relatively small number of bacteria, such as E. coli, account for most cases

of acute bacterial meningitis. The basis for such association in the context of E. coli meningitis remains incompletely understood[2,3].

Several lines of evidence of human cases and experimental animal models of E. coli meningitis indicate that cerebral capillaries are the portal of circulating E. coli entry into the brain[2,4,5]. Recent studies demonstrated that successful traversal of the BBB would require actin cytoskeleton rearrangements, and

related tyrosine phosphorylation of signaling pathways in the host cells. Several microbial determinants, such

3

ACS Paragon Plus Environment

Analytical 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

as outer membrane protein A (OmpA), Invasion of brain endothelial cell protein (Ibe), Arylsulfatase (AslA),

TraJ, and Cytotoxic necrotizing factor-1 (CNF1), have been found to contribute to E. coli penetration of the

BBB; It has been shown that surface-exposed loops of OmpA contribute to binding to HBMEC and OmpA

interacts with HBMEC through N-acetylglucosamine (GlcNAc) residues in HBMEC glycoproteins, including

gp96, however, many microbial factors have not been identified to reveal the complete mechanism of interaction between E. coli and BBB[2,5-7] . Since E. coli penetration into the brain occurred in the cerebral microvasculature, we used a BBB model with HBMEC to investigate the molecular mechanism that governs E. coli penetration of the BBB[2,6,7]. We showed that E. coli exploited specific microbial and host factors to facilitate penetration of the BBB. This concept has been shown by (a) identification of microbial factors,

namely Invasion of brain endothelial cell protein A (IbeA) and Cytotoxic necrotizing factor-1 (CNF1), and

IbeA-Caspr1, CNF1-37LRP, OmpA-gp96 and FimH-CD48 are currently known interacting pairs contributing to E. coli meningitis[8,9], demonstration that such microbial factors exploit specific host cell receptors and host cell signaling molecules for E. coli penetration of the BBB[1-3]. However, these interacting pairs were identified separately using microbial genomic approaches for microbial factors and conventional proteomic approaches

such as yeast two hybridization and affinity chromatography for host factors.

4

ACS Paragon Plus Environment

Page 4 of 41

Page 5 of 41 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

Analytical Chemistry

To develop an unbiased, proteome-wide methodology for the identification of microbial and host factors

contributing to E. coli association with the human BBB, the essential step in the development of meningitis,

we developed an unique high throughput chip-based cell probing assay to probe fluorescent live HBMEC with the E. coli proteome arrays[10], which contains more than 4200 individually purified E. coli proteins, to identify bacteria proteins that could capture live HBMEC and the identified bacterial proteins were used to discover interacting host proteins by performing binding assays on the HuProt arrays[11], containing >17,000 individually purified proteins. Cell-based validation and E. coli binding assay allowed us to confirm that the

interactions between the identified E. coli and host factors contributed to the pathogenesis of E. coli meningitis,

demonstrating that our approach is feasible for investigating pathogenesis of human infectious diseases.

EXPERIMENTAL SECTION Fabrication of E. coli proteome chips A library of 4,267 strains, each carrying a unique ORF in E. coli, was kindly provided by Dr. Mori’s research

group. The ORFs were cloned into pCA24N expression vector, from which the expression of each E. coli

protein fused with N-terminal Hisx6 tag is controlled by an inducible PT5-lac promoter. All purified proteins were printed in duplicate onto Hydrogel Shot slides which is immobilized with His tag antibody[12-14,43,45]. The detailed protocols are provided in the supplementary information.

5

ACS Paragon Plus Environment

Analytical 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

E. coli proteome chip assay

To investigate HBMEC proteins binding to E. coli, HBMECs were labeled with Vybrant CFDA SE Cell Tracer

(Life Technology, USA) and then applied to the E. coli proteome chip to proceed with chip assay. Cultivated

HBMEC were washed with PBS twice, and then lifted with 3 mL of 0.02% EDTA. After incubating at 37 °C under 5% CO2 atmosphere for 15 mins, PBS was added up to 15 mL. HBMEC were aliquoted into 1x106 cells to be stained by 1µl of Vybrant CFDA SE Cell Tracer in a total 1 mL solution for 2.5 mins at 37 °C, PBS was added

to a final 10 mL. HBMEC were then washed and collected through spin down and resuspended in 500 µl of Hanks’ Balanced Salt Solution (HBSS) with Ca2+ and Mg2+. The HBMEC were ready for probing E. coli

proteome chips at this point. After the proteome chip was probed with the CFDA stained HBMEC for 40 mins,

the chip was washed with HBSS, and stood vertically at room temperature (RT) for drying. Finally, the binding

signals of the protein arrays were acquired with microarray scanner (GenePix 4000B). Each experiment was

performed in triplicates.

Flow cytometry assays To validate the interaction of HBMEC with candidate proteins, proteins were labeled with fluorescein

isothiocyanate (FITC) and extra FITC was removed by spin columns. 120 µl of labeled proteins were mixed with 1 mL of 107 HBMEC in PBS containing 2% FBS for 1 hr at room temperature at dark. After mixing, the HBMEC-protein complexes were washed with PBS containing 2% FBS twice[15,16]. Finally, the washed complex solution was filtered for FACS. Each experiment was performed in triplicates. 6

ACS Paragon Plus Environment

Page 6 of 41

Page 7 of 41 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

Analytical Chemistry

Assay for HBMEC binding YojI coupled fluorescent microspheres

Bead binding assay was conducted to observe the binding affinity between candidate protein YojI and HBMEC.

YojI was extracted and purified as N-terminal Hisx6 tag fusion protein in E. coli using pre-loaded Ni-NTA

resins. In brief, the Carboxylated fluorescent microspheres (0.5 µm in diameter) (Life Technology, USA) was

covalently coupled with protein candidate according to the manufacturer’s protocol. Because the YojI protein

possesses amino groups that can react with carboxyl groups of the Carboxylated fluorescent microspheres

which made the YojI protein covalently coupled with microspheres through dehydration synthesis. Confluent

cultures of HBMEC were grown in collagen-coated 24-well plates and were incubated with either protein candidates-coupled beads (107/mL)(The concentration of this bead-conjugated YojI can be measured the

fluorescence according to the calibration curve of fluorescence with bead amount) or no protein-coupled beads as negative control for 1hr at 37 °C in dark followed by adding 10% glycine to quench the reaction[17]. After washing three times with ice-cold PBS to remove unbound beads, the fluorescence intensity of bound beads in

each well was measured with a fluorescence spectrometer equipped with plate reader (Molecular Devices). Each

experiment was performed in triplicates.

Assay for HBMEC interactions with yojI∆ strains

YojI knockout (i.e., yojI∆) bacteria was generated by one-step PCR approach[31]. After electroporation into E. coli, the yojI gene was replaced with the chloramphenicol cassette via homologous recombination, which was

7

ACS Paragon Plus Environment

Analytical 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

facilitated by use of theλred system. Subsequently, chloramphenicol antibiotic was used to identify yojI∆ mutants, and the correct yojI replacement was confirmed by PCR with screening primers and by sequencing of

PCR products.

Confluent cultures of HBMEC grown in 24-well plates were incubated with 107 of E. coli wild type, YojI∆ for 1hr at 37 °C. HBMEC were washed with RPMI 1640 to remove unbound E. coli. The bound E. coli were

enumerated by plating on sheep blood agar plates (BAP), and counted on the next day. The association

frequencies were calculated by dividing the number of bound bacteria by the number of the original inoculums.

The results are presented as a relative association. Each experiment was performed in triplicates.

Adhesion assay using complemented YojI∆ strain (YojI/C)

The yojI overexpress plasmid (pCAN24N) from the ASKA library was purified and transformed (via electroporation) into the YojI∆ strain (Keio YojI∆) from the Keio collection to complement YojI∆ strain (YojI/C strain)[18,21]. HBMEC were seeded into 24-well plates and grown until confluence. Bacteria cultures were grown in appropriate selective media. The yojI/C strain was pretreated for 2 hours with 1 mM IPTG. All

bacteria strains were then incubated with HBMEC (MOI of 100) for 1½ hours (IPTG was also added to the

wells during incubation of YojI/C with HBMEC). Following incubation, HBMEC were washed extensively to

remove unbound bacteria, cells were lysed, and bacteria were plated for enumeration. The experiment was

performed in triplicates.

8

ACS Paragon Plus Environment

Page 8 of 41

Page 9 of 41 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

Analytical Chemistry

Assay for WT E. coli binding to HBMEC in the presence of YojI proteins

This assay was conducted to observe the association between MG1655 and HBMEC after HBMEC was first

blocked with YojI proteins. YojI proteins were purified and pre-incubated with HBMEC for 1 hr before adding

MG1655 at MOI 100 to HBMEC. Association of the bacteria to HBMEC was measured as described above.

We performed the experiments in triplicates.

RNA profiling in HBMEC

Total RNA was extracted from HBMEC using TRIZOL (Invitrogen) before and after the cells were treated for

90 mins with purified YojI (100 µg/mL), IbeC (100 µg/mL), and buffer. Hybridization of the purified RNA samples was performed using a commercial DNA microarrays following manufacturers’ instruction

(Eukaryote Total RNA Pico Series II; Agilent Technologies, Inc.).

Protein-protein interaction assays performed on the HuProt arrays

The HuProt arrays, each comprised of 17,000 human proteins individually purified as N-terminal GSTX6 tag

fusions, and printed in duplicate onto Hydrogel Shot slides which is immobilized with GST tag antibody, were

fabricated and provided by CDI Laboratories Inc. (CDI, USA). To identify potential host receptors that might

facilitate adhesion of E. coli, purified YojI proteins were labeled and probed to the HuProt arrays using a

previously established protocol. The HuProt arrays were first blocked with 1% BSA in TBST for 1 hr at RT.

After a brief wash, Cy5 (Abbkine)-labeled YojI proteins were incubated on the HuProt in TBS-T (0.05% Tween 9

ACS Paragon Plus Environment

Analytical 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

20) and 1% BSA in the hybridization chamber with shaking for 1 hr at RT. Finally, the HuProt arrays were washed with TBS-T and distilled water for several times[10,11,19,20]. After the final wash, the arrays were dried by centrifugation at 201 g and then scanned with a microarray scanner (Axon GenePix® 4000B). Each experiment

was performed in triplicates.

Generation of stable IFNAR2-knockdown cell line in HBMEC

IFNAR2 shRNA and GFP control packaging lentivirus (TRCN000058783; Sigma) were constructed following

manufacturers’ instruction. After HBMEC was cultured in 6-well plates to reach 70% confluence, medium

containing lentiviruses and polybrene (10 µg/mL; Life Technology) was added at a multiplicity of infection

(MOI) of 10 and mixed with the cells. Polybrene was used to improve infection efficiency. After incubation for

24 hrs, supernatants in the wells were replaced by DMEM containing FBS and puromycin (10 µg/mL) for

selection for 2 days twice. The generated HBMEC stable cell line was used for subsequent RT-PCR analyses[22].

Total RNA was isolated from the cells in each group with an RNA simple Total RNA kit (Tiangen Biotech Co., Ltd.) and reverse transcription was carried out using SuperScriptTM III First-Strand Synthesis System. RT-PCR

was performed using SYBR-Green Master Mix (QPK-201, YOBO Co.Ltd) MyiQTM single color real-time

PCR Detection system (Bio-Rad). The reaction conditions were set as follows: 95 °C for 10 min; 40 cycles of

10

ACS Paragon Plus Environment

Page 10 of 41

Page 11 of 41 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

Analytical Chemistry

95 °C for 10 sec, 58 °C for 20 sec and 72 °C for 30 sec. Gene expression levels were calculated using the 2-∆∆ Ct

method as previously described. GAPDH was used as the normalization internal control.

Designed primer pairs for RT-PCR for detecting IFNAR2 shRNA1 knocking down efficiency:

Primer pairs for detecting IFNAR2 gene:

F: 5’-CAGATCACAGCTTCTTCCCA-3’; R: 5’-CCACCCATTCTCAGGGTAGT-3’

Primer pairs for detecting control GADPH gene:

F: 5’-ACATCAATGAGTGGCTCCAA-3’; R: 5’-TTGCTGTACCCGATCTTGAA-3’

The association between IFNAR2-knockdown HBMEC cells and WT E. coli or YojI∆ cells was quantified using the same method as described above. Each experiment was performed in triplicates.

HBMEC and WT E. coli association assays in the presence of IFNAR2 antibodies

HBMEC was blocked with anti-IFNAR2 antibodies at several concentration (Santa Cruz, USA). Anti-IFNAR2

antibodies were added to 6-well plate at final concentrations of 0.02, 0.2 and 2 µg/mL for each well, respectively, and incubated for 12 hrs at 37 °C. The bacterium association assay was quantified using the same

method as described above. Each experiment was performed in triplicates.

Statistical analysis

11

ACS Paragon Plus Environment

Analytical 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

Data were analyzed with the Student’s t test and *P ˂ 0.05 and **P ˂ 0.01 were considered to be statistically

significant. Statistical analyses were performed with Graphpad Software.

RESULTS AND DISCUSSION Chip-based cell probing assay for identifying E. coli proteins that interact with HBMEC To identify bacterial proteins as potential ligands for interacting with the HBMEC using an unbiased,

proteome-wide approach, we probed fluorescently labeled, live HBMEC to an E. coli proteome microarray, comprised of ~4,200 E. coli K-12 proteins individually purified as Hisx6 fusions[10]. After incubating for 40 mins at 37 °C, the E. coli proteome arrays were carefully washed to remove those unattached HBMEC from

the E. coli proteome arrays. After drying, the captured cells were visualized by a microarray scanner. Although

HBMEC cells are of large size (e.g., 50 µm in diameter), which made this assay not easy to handle, because the

HBMEC will get off the chip easily due to the big size of HBMEC cells while doing the wash step. So, the key

to have this assay successful is to wash gently without shaking. The chip has to get in and out the wash buffer

surface vertically to avoid shearing force. However, we were able to detect some E. coli proteins that could

repeatedly bind to the HBMEC (Figure 1). Of the three replicated binding assays, 23 E. coli proteins were found to capture HBMEC in at least two of the three assays. Eight of the 23 proteins are annotated as

membrane proteins, including Diacylglycerol kinase alpha (DgkA), Glycerol uptake facilitator protein (GlpF),

12

ACS Paragon Plus Environment

Page 12 of 41

Page 13 of 41 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

Analytical Chemistry

Serine/threonine-protein phosphatase 4 catalytic subunit (Ppx), Phosphate transport system permease protein

(PstC), A putative transporter protein (YbaE), Peptide transport system permease protein (Sap C), ABC

transporter ATP-binding/permease protein (YojI), and UPF0382 family inner membrane protein (YgdD).

Interestingly, GlpF, PstC, YbaE, SapC and YojI are all annotated to encode transporter activity (Table S-1). Because the function of the membrane proteins, YojI and YbaE, were not characterized in previous researches,

they were chosen for further validation and characterization. It is interesting to note that bovine histone H4

proteins, printed as landmarks at the same corner of each of the 48 printed protein blocks, also strongly

interacted with the cells. One plausible explanation is that the observed interactions were also likely due to

electric static force as histone proteins are highly positively charged and the cell surfaces are generally

negatively charged.

13

ACS Paragon Plus Environment

Analytical 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 1. Utilization of E. coli and human proteome arrays to identify YojI and interferon-alpha

receptor (IFNAR2) mediated bacterium-host cell interactions. Live HBMEC was labeled with Vybrant CFDA SE cell tracer and probed to the E. coli proteome microarrays. As one of the E. coli membrane proteins

that could capture the labeled HBMEC, YojI’s interaction with HBMEC was further validated using additional

methods, such as FACS. To identify the host receptors of YojI, purified YojI proteins were fluorescently

labeled and probed to the HuProt arrays, comprised of ~17,000 human individually purified proteins. One of

the identified human proteins, IFNAR2, was validated with cell adhesion assays. Each assay was performed in

triplicates.

Validation using flow cytometry assay 14

ACS Paragon Plus Environment

Page 14 of 41

Page 15 of 41 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

Analytical Chemistry

To validate the interactions between HBMEC cells and the above-mentioned E. coli membrane

proteins, we chose the flow cytometry (i.e., FACS) analysis as an orthogonal method. The seven membrane

proteins were individually purified from E. coli in large quantity using Ni-NTA resins and labeled with FITC,

followed by neutralization and removal of the extra dye molecules. The FITC-labeled proteins were then

eluted and their quantity and quality were examined using Coommassie stain (Figure S-1). FITC-labeled YojI and YbaE protein samples were added to 107 HBMEC cells at three final concentrations (i.e., YojI: 10, 50, and

250 µg/mL; YbaE: 10, 50, and 250 µg/mL) in a volume of 1 mL and incubated for 1 hr at room temperature. After washing, HBMEC-protein complexes were analyzed using a FACS machine. Among the seven tested

membrane proteins, both YojI and YbaE showed specific association with HBMEC in a dose-dependent fashion (Figure 2A-C), indicating that YojI and YbaE are likely to interact with the HBMEC (Table S1)[23-25, 27]

. These findings suggest that YojI might be involved in mediating the binding events between E. coli and

HBMEC[26,27], which cause our attention.

Validation of the association of YojI with HBMEC using cell-based association assays To further characterize the interactions between the bacterial protein YojI and HBMEC, we performed a

series of cell-based association assays. First, we employed a previously published protocol to couple purified YojI protein to fluorescent microspheres[28,29]. To better mimic the in vivo conditions, we incubated these

YojI-coupled fluorescent microspheres to plate-attached HBMEC at a final concentration of 10 µg/mL. As a

15

ACS Paragon Plus Environment

Analytical 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

negative control, unconjugated fluorescent microspheres were processed in the same fashion in parallel. After

washing under stringent conditions, fluorescent signals obtained from YojI-coupled fluorescent microsphere

group exhibited significantly (P < 0.05) higher fluorescent signals compared with those from the negative

control group (Figure 2D). Secondly, we examined whether YojI played an important role in pathogen-host interactions at the level of cell-cell association using a previously reported method[27, 30]. Briefly, plate-attached HBMEC were

separately incubated with the same number of the wild-type (WT) E. coli and yojI knockout (i.e., yojI∆) bacteria generated on the same genetic background by one- step PCR approach[31,32]. After incubation for 1 hr at 37 °C, unbound E. coli cells were removed by stringent washes, and those E. coli cells that tightly bound to

HBMEC were recovered (EXPERIMENTAL SECTION). To quantify the physical association between HBMEC and E. coli cells, we plated the bacterial cells recovered from each assay on sheep blood agar plates

and counted the number of E. coli colonies formed on the following day. The association of yojI∆ cells with HBMEC was dramatically decreased by as much as ~80% as compared with WT cells (Figure 2E), suggesting that YojI protein plays an important role in the E. coli-HBMEC interactions at the cellular level.

To further confirm the specificity of YojI in host cell recognition, we employed a competition assay in a

similar setup. Plate-attached HBMEC were incubated with WT E. coli cells in the presence or absence of purified YojI proteins at a final concentration of 10 µg/mL[33]. Using the same colony counting approach, we observed that the presence of YojI proteins in the cell adhesion assay significantly (P< 0.01) reduced the

16

ACS Paragon Plus Environment

Page 16 of 41

Page 17 of 41 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

Analytical Chemistry

number of attached E. coli WT cells to plate-attached HBMEC (Figure 2F). Taken together, the above bacterial-host cell interaction assays support that YojI plays a crucial role in E. coli interaction with HBMEC.

Figure 2. Validation of YojI and HBMEC interactions. (A) The threshold gating of FACS analysis to detect YojI, YbaE and HBMEC interactions (FSC: forward scatter; SSC: side scatter). (B) HBMEC was separately

incubated with purified and labeled YojI proteins at concentrations of 10, 50 and 250 µg/mL, respectively. Shadowed area denotes the negative control fraction without YojI (FITC: fluorescein isothiocyanate). (C)

HBMEC was separately incubated with purified and labeled YbaE proteins at concentrations of 10, 50 and 250

µg/mL, respectively. Shadowed area denotes the negative control fraction without YbaE.

17

ACS Paragon Plus Environment

Analytical 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

(D) YojI-coupled fluorescent microspheres showed binding signals to plated HBMEC. Using the un-coupled

empty fluorescent microspheres as a negative control, YojI-coupled microspheres showed a small but

significant increase in fluorescent signals (P=0.05). (E) YojI-mediated bacterium-host cell interaction is

confirmed with cell adhesion assays. WT and YojI∆ bacterial cells, generated on the same genetic background of the WT E. coli, were separately incubated with plated HBMEC. After washing, the adhered bacterial cells

recovered from HBMEC were platted on agar plates and numbers of formed bacterial colonies counted. Using

the colony numbers as a quantification method to estimate bacterium-host cell interactions, knocking YojI

significantly reduced the bacterial adhesion by ~80% (P=0.0001). (F) Presence of YojI proteins reduced E.

coli-HBMEC interactions. To demonstrate specificity of YojI-mediated bacterium-host cell interactions, plated

HBMEC was first incubated with purified YojI proteins at a final concentration of 10 µg/mL, followed by adding WT E. coli cells to HBMEC. Using the same cell adhesion assays, the presence of YojI protein

significantly reduced bacterial cell adhesion to HBMEC by >30% (P=0.004). Single (*) and double (**)

asterisks indicate the P values 89%. Again, we employed the cell-based assay to evaluate the E. coli cell association

efficiency to HBMEC under different conditions. Separately plate-attached WT and shRNA1-expressing (i.e.,

IFNAR2 knockdown) HBMEC were incubated with WT or yojI∆ E. coli cells at 37 °C. After unbound bacterial cells were removed with several washes, the bound E. coli cells to HBMEC were recovered and

plated on sheep blood agar plates, followed by counting the numbers of bacterial colonies. Using the colony

numbers recovered from WT E. coli cell attachment to WT HBMEC plated as a normalization point, we

observed a small but significant reduction (i.e., 89%) when IFNAR2 was knocked down in HBMEC cells

(Figure 4A). The attachment efficiency, however, was dramatically reduced to the lowest level of 33% , when yojI was also knocked out, suggesting an additive effect between IFNAR2 and YojI. As a comparison,

knocking out yojI alone reduced the attachment efficiency to 37.4%, representing a small but significant

difference.

To demonstrate the specificity of IFNAR2-mediated cell-cell interactions between the bacteria and host,

we employed a commercially available anti-IFNAR2 antibody in the same bacterial cell association assay.

After a pre-incubation step with the antibody of various amounts (i.e., 0.02, 0.2 and 2.0 µg) to plate-attached

24

ACS Paragon Plus Environment

Page 24 of 41

Page 25 of 41 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

Analytical Chemistry

HBMEC, WT E. coli cells were added to the plates. Using the mock control (i.e., no antibody added) as the

normalization point for measuring bacterial cell attachment, we observed the attachment efficiency was

reduced to 88.5%, 83.9% and 77.5% by adding anti-IFNAR2 antibody at final concentrations of 0.02, 0.2 and

2 µg/mL, respectively (Figure 4B). The observed dose-dependent blocking effect indicated that IFNAR2 played a role in mediating bacterial-host cell interactions.

Figure 4. Validation of the YojI-IFNAR2-mediated E. coli and host cell interactions. (A) Knocking down of IFNAR2 (shRNA1) in HBMEC significantly reduced WT E. coli adhesion to HBMEC as compared with

untreated cell (i.e., HBMEC WT) (P=0.049). More importantly, knocking out YojI (yojI∆) in WT E. coli cells

25

ACS Paragon Plus Environment

Analytical 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

further reduce the bacterial cell adhesion to shRNA1-treated HMBEC to 33% as compared with the WT cell

controls ( yojI∆/WT, P=0.0003; yojI∆/shRNA1, P=0.0001 ). Single (*) and double (**) asterisks indicate the P values