Novel surface plasmon resonance biosensor that uses full-length Det7 phage tail protein for rapid and selective detection of Salmonella enterica serovar Typhimurium

We report a novel surface plasmon resonance (SPR) biosensor that uses the full-length Det7 phage tail protein (Det7T) to rapidly and selectively detect Salmonella enterica serovar Typhimurium. Det7T, which was obtained using recombinant protein expression and purification in Escherichia coli, demonstrated a size of ∼75 kDa upon SDS-PAGE and was homotrimeric in its native structure. Micro-agglutination and TEM data revealed that the protein specifically bound to the host, S. Typhimurium, but not to non-host E. coli K-12 cells. The observed protein agglutination occurred over a concentration range of 1.5∼25 μg.ml−1. The Det7T proteins were immobilized on gold-coated surfaces using amine-coupling to generate a novel Det7T-functionalized SPR biosensor, wherein the specific binding of these proteins with bacteria was detected by SPR. We observed rapid detection of (∼ 20 min) and typical binding kinetics with S. Typhimurium in the range of 5 × 104-5 × 107 CFU.ml−1, but not with E. coli at any tested concentration, indicating that the sensor exhibited recognition specificity. Similar binding was observed with 10% apple juice spiked with S. Typhimurium, suggesting that this strategy could be expanded for the rapid and selective monitoring of target microorganisms in the environment.


Introduction
Infections involving pathogenic bacteria are a major cause of morbidity and mortality in 41 human beings and animals around world, resulting in a huge economic burden related to 42 healthcare costs and manpower/wealth losses. Rapid monitoring of pathogenic 43 microorganisms is critical for our ability to halt the fast spread of bacteria. The reliable 44 conventional methods for detecting microorganisms (e.g., culture-based biochemical and 45 serological assays) require the growth and technical manipulation of large amounts of cells, 46 and thus tend to be time-consuming (24-52 h), labor-intensive, and cost-ineffective [1][2][3]. 47 Biosensors are regarded as an attractive alternative, and have been used to detect various 48 environmental pollutants [4-6] and microorganisms [7,8]. A biosensor has two key 49 components: a biological sensing element that enables specific recognition of a pathogen; and 50 a transducer that converts this recognition to a measurable signal. Different sensing elements 51 have been exploited in the development of biosensor platforms, including antibodies [9], 52 DNA [10], RNA [11], aptamers [12], peptides [13], and carbohydrates [14]. Similarly Bacteriophages have recently gained interest as sensing elements for pathogen-detecting 57 platforms, because they are abundant in nature, stable under harsh conditions (e.g., extremes 58 of temperature, pH, ionic strength, etc.), and specific to their target host bacteria [8]. Phages 59 recognize their specific hosts through the ability of their tail proteins to bind receptors on the 60 bacterial surface. The recognition and binding of a receptor by the tail protein is highly 61 specific, which makes phages useful for bacterial typing and excellent candidates as sensing 62 elements in biosensors [21]. To date, E. coli [22][23][24][25] [27,28]. We recently showed that a fragment of tail protein from 75 phage lambda (6HN-J) bound specifically to the host, E. coli K-12, but not to other bacteria 76 [29]. We observed nonspecific transient attachment to non-host bacteria, and proposed that it 77 might be part of the mechanism through which viral tail proteins recognize their host 78 receptors. This was previously suggested by Silva et al. [30], who described that the first step 79 of bacteriophage adsorption involves random collisions between the phage and various 80 bacteria (e.g., by Brownian motion, dispersion, diffusion, and/or flow). During these 81 interactions, the phage searches for its specific receptor via reversible (transient) binding 82 [30]. 83 In order to address the issues raised with intact phages or truncated tail proteins as a 84 sensing element in biosensor, we examined whether the full-length tail protein of phage Det7 85 could be coupled with a biosensing platform to enable the specific recognition and capture of 86 S. Typhimurium. Det7 is a Salmonella phage (Myovirus) whose 75-kDa tail protein exhibits 5 87 50% overall sequence identity to the tail endorhamnosidase of Podovirus P22 [31]. Both tail 88 proteins bind octasaccharide fragments from Salmonella lipopolysaccharide, and that of Det7     for 5 min, and exposed to an X-ray film [29].

160
Micro-agglutination assay 161 The micro-agglutination assay [27] was performed with Det7T and bacteria in microtiter  The Det7 phage appears to offer advantages over other viral sensing elements for Salmonella 201 detection, as it has a wide-ranging host specificity against various Salmonella sp.
[32], and its 202 tail protein exhibits good heat stability [31]. Here, we set out to use the full-length wild-type 203 Det7 tail protein (Det7T) as a sensing element for SPR biosensor construction. A construct 204 encoding full-length Det7T was PCR amplified to yield a product of ~2130 base pairs. We 205 fused the purified PCR product with a (His) 6 -tag by inserting it into a linearized vector in 206 which the coding sequence was under the control of the IPTG-inducible T7lac promoter. In 207 our system, Det7T was found to be overproduced in the soluble fraction, as assessed by SDS-208 PAGE (Fig 1A).  Although we initially applied this soluble fraction to a HisTALLON affinity column, we 220 found that Det7T did not bind to this Ni column. This may reflect that the His-tagged N-or 221 C-termini of Det7T are buried in its 3D structure, even under the slightly denaturing 12 222 conditions (6 M urea) used in our work. It has been reported that the surface area of the 223 monomer was buried in the trimer of the amino-terminally shortened Det7 tail protein,

224
Det7tspΔ 1-151 [31]. Therefore, we heat treated the soluble fraction at 80℃ for 6 min to 225 denature other proteins, and then purified our target protein using DEAE and a Sephadex 226 column. The purified Det7T had an apparent size of ~75 kDa on SDS-PAGE (Fig 1A), which 227 was consistent with the predicted size. Micro-agglutination assay 237 The ability of the purified Det7T to bind S. Typhimurium or E. coli was tested by a micro-238 agglutination assay (Fig 2). 1.5 μg.ml -1 of Det7T was incubated with S. Typhimurium cells (Fig. 2, 7 th well on left of the 255 upper panel); the degree of agglutination increased up to ~25 μg.ml -1 (Fig 2) and plateaued 256 thereafter. These results indicate that Det7T formed a tight bond with S. Typhimurium cells.

257
The minimum concentration of Det7T that yielded detectable cell agglutination (1.5 μg.ml -1 ) 258 was 2-fold lower than the minimal agglutinating concentration of P22 tail protein (3 μg.ml -1 ) 259 [27]. There was no detectable specific binding of Det7T with E. coli (Fig 2), indicating that 260 Det7T specifically agglutinated with S. Typhimurium. deactivation of excess active groups respectively. The grid was then washed several times 267 with PBS, exposed to the microorganism, and observed by TEM. Consistent with the results 268 of our micro-agglutination assays, we observed the binding of Det7T to S. Typhimurium cells 269 treated with Det7T (Fig 3A), but not to Det7T-treated E. coli (Fig 3B) or to non-treated S. 270 Typhimurium cells (Fig 3C).  CFU.ml -1 (Fig 4A).  We obtained a significant signal of bacterial capture with as little as ~5 x 10 4 CFU.ml -1 S. 305 Typhimurium (Fig 4B). We previously obtained a comparable detection limit of 2 x 10 4 306 CFU.ml -1 for E. coli K-12 with immobilization of N-terminally (His) 6 -tagged 6HN-J on a Ni-307 coupled chip [29]. Our results resemble those of a report in which the detection limit of S. 308 aureus was found to be 10 4 CFU.ml -1 with the lytic phage SPR-based SPREETA TM sensor 16 309 [19]. The SPR response representing S. Typhimurium binding was found to be dose-310 dependent ( Fig 4C). In contrast, we did not observe any binding response to non-host E. coli 311 (Fig 4D) [30]. Here, we found that full-length Det7T showed rapid recognition and binding to the cell These data suggest that the SPR biosensor system described herein could be exploited for 340 rapidly and selectively monitoring pathogenic microbes in food and the environment. Det7T to a CM5 chip through amine coupling to generate a novel Det7T-functionalized SPR 347 biosensor. We observed rapid detection of (~ 20 min) and typical binding kinetics with host 348 S. Typhimurium in the range of 5 x 10 4 -5 x 10 7 CFU.ml -1 , but not with non-host E. coli, 349 indicating that our biosensor exhibited selective recognition. The binding of Det7T was also 350 similarly observed in 10% apple juice spiked with S. Typhimurium. The results suggest that