A Novel Surface-Exposed Polypeptide Is Successfully Employed as a Target for Developing a Prototype One-Step Immunochromatographic Strip for Specific and Sensitive Direct Detection of Staphylococcus aureus Causing Neonatal Sepsis

Neonatal sepsis is a life-threatening condition and Staphylococcus aureus is one of its major causes. However, to date, no rapid and sensitive diagnostic tool has been developed for its direct detection. Bioinformatics analyses identified a surface-exposed 112-amino acid polypeptide of the cell wall protein NWMN_1649, a surface protein involved in cell aggregation and biofilm formation, as being a species-specific and highly conserved moiety. The polypeptide was cloned, purified, and used to immunize mice to raise specific immunoglobulins. The purified antibodies were conjugated to gold nano-particles and used to assemble an immunochromatographic strip (ICS). The developed prototype ICS detected as low as 5 µg purified polypeptide and 102 CFU/mL S. aureus within 15 min. The strip showed superior ability to directly detect S. aureus in neonatal sepsis blood specimens without prior sample processing. Moreover, it showed no cross-reaction in specimens infected with two other major causes of neonatal sepsis; coagulase-negative staphylococci and Klebsiella pneumoniae. The selected NWMN_1649-derived polypeptide demonstrates success as a promising biomolecule upon which a prototype ICS has been developed. This ICS provides a rapid, direct, sensitive, and specific option for the detection of S. aureus causing neonatal sepsis. Such a tool is urgently needed especially in resources-limited countries.


Introduction
Neonatal sepsis is a life-threatening condition that increases morbidity and mortality rates in newborns especially in developing countries [1]. The World Health Organization (WHO) reported in 2014 that almost 15% of neonatal deaths results from neonatal sepsis especially in low and middle income countries (LMICs) [2]. A traditional reliable technique for the diagnosis of this condition is

Statement of Ethical Approval
The protocol of the study was approved by the Research Ethics Committee at the Faculty of Pharmacy, Cairo University (Approval Numbers: MI 1612 and 2426).

Bioinformatics Analyses of the S. aureus Proteome to Identify Potential Targets
The PSORTdb database, v. 3.0 [19] was used to predict the localization of all proteins encoded by S. aureus Newman strain genome (Staphylococcus aureus subsp. aureus str. Newman DNA, complete genome. NC_009641.1). Proteins predicted to be localized in the cell wall were screened for S. aureusspecific proteins or peptides using the BlastP tool [20] against the non-redundant protein database excluding S. aureus proteins from query. Proteins that returned no hits or low query coverage were considered to be potentially S. aureus-specific. Amino acid (aa) sequences were retrieved and compared by multiple sequence alignment using Clustal Omega, v 1.2.4 [21] to identify highly conserved S. aureus regions. Phylogenetic analysis was performed using Unipro UGENE v. 35 platform [22]. The sequences of the potential targets were screened for conserved domains using the NCBI Conserved Domain Database [23].
Physicochemical parameters of the selected potential target were predicted using Expasy's ProtParam server [24]. Surface exposure of the target peptide was predicted by comparative modeling of the target peptide using Modeller [25] and UCSF-Chimera v. 1.14 [26]. Signal peptide position and cleavage site were predicted using SignalP 4.1 Server which employs neural networks and Markov Models (HMM) to predict signal peptide cleavage site [27]. Subcellular localization of the target peptide was predicted using CELLO online tool v. 2.5 [28]. Probable extracellular domains and transmembrane helices were identified using protein topology prediction tools TOPCONS and CCTOP server version s. 1.00 (Consensus Constrained TOPology prediction web server) [29]; to determine protein regions expected to be exposed on bacterial surface. Molecular weight prediction was performed using the Expasy bioinformatic tool [30].
2.3.2. In Vitro Assessment of the Expression of the Selected Protein Using Real-Time RT-PCR S. aureus Newman cells grown to mid-logarithmic phase were mixed with either phosphate buffered saline (PBS) or heat-inactivated human serum (obtained from healthy volunteers) to a final concentration of 10% (v/v), and were incubated for two hours at 37 • C. RNA was extracted using the RNeasy minikit (Qiagen, Hilden, Germany) and reverse-transcribed using the QuantiTect reverse transcription kit (Qiagen, Hilden, Germany). The produced cDNA was used as template for real-time PCR analysis using the GoTaq ® qPCR SYBR Green master mix (Promega, Madison, WI, USA). Primer pair 16S-Fw (5 -TGAGATGTTGGGTTAAGTCCCGCA-3 ) and 16S-Rv (5 -CGGTTTCGCTGCCCTTTGTATTGT-3 ) was used for the 16S rRNA normalizer; and primer pair SA029 (5 -CGACTGATGATGGCGTTACT-3 ) and SA030 (5 -TCCCTTTGTTTCGTTGCTTTATC-3 ) for the NWMN_1649. To ensure absence of genomic DNA contamination in the extracted RNA samples, negative RT control sample was tested.
To purify the polypeptide, overnight culture of BL21/pNWMN_1649-6xHis was diluted in fresh LB/Amp and allowed to grow to mid-logarithmic stage then induced with isopropyl β-D-1thiogalactopyranoside (IPTG) to final concentration of 1 mM for 16 h at 30 • C. Cells were collected by centrifugation at 4200× g at 4 • C, disrupted by sonication till clear solution was obtained; followed by centrifugation at 12,000× g at 4 • C. The clear supernatant was then purified using Ni-NTA resin (Qiagen, Hilden, Germany) as described earlier [32]. The polypeptide was analyzed using SDS-PAGE and its concentration was determined using QuantiPro BCA assay kit (Sigma-Aldrich, St. Louis, MI, USA).

Development of an Immunochromatographic Strip(ICS)
2.5.1. Animal Immunization and Production of Immunoglobulins Ten female BALB/c mice (16-18 g) (Theodor Bilhariz Research Institute, Giza, Egypt) were immunized with 25 µg polypeptide emulsified with Complete Freund's adjuvant (Sigma-Aldrich, St. Louis, MI, USA) subcutaneously. The mice were boosted after 21 days with 12.5 µg polypeptide emulsified with Incomplete Freund's adjuvant (Sigma-Aldrich, St. Louis, MI, USA) via the same route then boosted one more time with 12.5 µg polypeptide in PBS intraperitoneally after 2 additional weeks. On day 45, mice were sacrificed, blood was collected by exsanguination of the posterior vena cava, the serum was separated and stored at −20 • C. Mice injected with pyrogen-free saline instead of the polypeptide were used as a negative control.

Specificity Evaluation of the Produced Antibodies using Enzyme Linked Immunosorbent Assay (ELISA)
Immunoglobulins in the immunized mice sera were screened with ELISA using the purified polypeptide. Then the antibodies were purified using the protein A agarose purification kit (KPL, Gaithersburg, MD, USA) according to the manufacturer's instructions. The antibody titer and specificity against the specific antigen (either in sera or purified immunoglobulins) were assessed with ELISA using 1 µg/well purified NWMN_1649 polypeptide suspended in 50 µL 0.1M carbonate buffer (0.0125M NaHCO 3 and 0.0875M Na 2 CO 3 , pH 9.5) as the coating antigen; as previously described [33].

Synthesis and Characterization of Gold Nanoparticles (GNPs)
GNPs were synthesized using the citrate reduction method [34]. The mean particle size of nanoparticles was determined by Photon Correlation Spectroscopy using Malvern Zetasizer Nano-ZS (Malvern Instruments, Malvern, UK).

Preparation of GNP-Antibody Conjugate and Assembly of ICS
The GNP-antibodies conjugate was prepared using the method described by Bailes and co-workers [35] with slight modifications. The stabilization step was done by the addition of 1% BSA (in 20 mM borax buffer, pH 9.2) at ratio of (1:10 v/v) and incubation for 1 h at room temperature, the mixture was then centrifuged at 12,000× g at 4 • C and re-suspended in 2 mM borax buffer (pH 9.2) containing 1% BSA.
ICS preparation and assembly was also performed according to Bailes and co-workers with slight modifications as the glass fiber strip (EMD Millipore, Burlington, MA, USA) (5 mm × 10 mm) was first soaked in the release agent (0.5% Tween-20, 0.5% human serum albumin in water) for 5 min and dried at 37 • C for 2 h. The Ig-gold conjugate was mixed with dispensing agent (20 mM PBS, 5% methanol and 0.1% lactose) at ratio of (4:1 v/v), loaded onto the dried, treated glass fiber strip and dried at 37 • C for 2 h. The capture antibody was the purified antibodies at concentration of 0.5 mg/mL and the control antibody was 0.5 mg/mL anti-mouse IgG antibody in 0.135 M sodium chloride, diluted (1:5) in dispensing solution to a final concentration of 100 µg/mL, then the procedures were continued as described [35]. The cellulose fiber sample pad (EMD Millipore, Burlington, MA, USA) (15 mm × 5 mm), the treated glass fiber strip, the treated nitrocellulose membrane strip (Carl Roth, Karlsruhe, Germany) (5 × 35 mm) and the absorbent pad were then all assembled.

Testing the Sensitivity and Specificity of the Constructed ICS
The ICS was tested with different amounts of the purified NWMN_1649 polypeptide in 150 µLaliquots of (10, 5, 2.5, 1.25 and 0.625 µg) and applied onto the sample pad, allowed to flow, and left for 10-15 min before reading the test result. Appearance of two bands at capture and control lines indicated a positive result, while appearance of one band at control line indicated a negative result. Sterile PBS was used as a negative control. To assess the ability of the developed ICS to detect whole S. aureus cells, it was challenged with cell suspensions containing different cell concentrations of S. aureus Newman (10 2 -10 8 CFU/mL PBS).

Assessment of the Developed ICS Using Clinical Specimens Obtained from Neonatal Sepsis Patients
A total of ten blood specimens were collected from neonates diagnosed with neonatal sepsis and the causative agents were identified by blood culture followed by VITEK 2. The samples were diluted 10 folds using sterile PBS and 150 µL of each was then loaded onto the sample pad of the assembled ICS and tested as mentioned above.

In Silico Identification of Potential Targets for ICS
Analyses of the S. aureus strain Newman proteome revealed a total of 2919 proteins distributed as follows: 38 in cell wall, 1359 in cytoplasm, 772 in cytoplasmic membrane, 113 extracellular, and 632 with unknown localization. Cell wall proteins, which are most probably surface-exposed, were chosen for further analyses as potentially good diagnostic targets.
Comparative sequence analyses using BlastP tool did not reveal S. aureus-specific entire cell wall protein. Accordingly, the search was focused on regions within those proteins that can fulfill the set criteria. Checking the sequence specificity of the first 300 aa of NWMN_1649 (BAF67921.1) of S. aureus indicated that the region between aa 46 to 157 looked distinct from the closest match (Figure 1a). Performing a BlastP search using the selected polypeptide against genus Staphylococcus while excluding S. aureus returned few hits from staphylococcal species not known to cause major human infections, if any, such as S. schweitzeri, S. argenteus, and S. sciuri ( Figure 1b) [36,37]. Multiple sequence alignment using Clustal Omega of the selected peptide and its closest matches from the BlastP analysis (Figure 1c), and phylogenetic tree analysis ( Figure S1) confirmed that the S. aureus NWMN_1649 (BAF67921.1) 112 aa (46-157) peptide is distinct from its closest matches from other species. Upon, repeating the blast search exclusively against S. aureus, the region was 100% conserved in many strains ( Figure 1d).

In Silico Identification of Potential Targets for ICS
Analyses of the S. aureus strain Newman proteome revealed a total of 2919 proteins distributed as follows: 38 in cell wall, 1359 in cytoplasm, 772 in cytoplasmic membrane, 113 extracellular, and 632 with unknown localization. Cell wall proteins, which are most probably surface-exposed, were chosen for further analyses as potentially good diagnostic targets.
Comparative sequence analyses using BlastP tool did not reveal S. aureus-specific entire cell wall protein. Accordingly, the search was focused on regions within those proteins that can fulfill the set criteria. Checking the sequence specificity of the first 300 aa of NWMN_1649 (BAF67921.1) of S. aureus indicated that the region between aa 46 to 157 looked distinct from the closest match (Figure 1a). Performing a BlastP search using the selected polypeptide against genus Staphylococcus while excluding S. aureus returned few hits from staphylococcal species not known to cause major human infections, if any, such as S. schweitzeri, S. argenteus, and S. sciuri (Figure 1b) [36,37]. Multiple sequence alignment using Clustal Omega of the selected peptide and its closest matches from the BlastP analysis (Figure 1c), and phylogenetic tree analysis ( Figure S1) confirmed that the S. aureus NWMN_ 1649 (BAF67921.1) 112 aa (46-157) peptide is distinct from its closest matches from other species. Upon, repeating the blast search exclusively against S. aureus, the region was 100% conserved in many strains (Figure 1d).
Searching the NCBI Conserved Domain Database revealed that the parent protein NWMN_1649 (BAF67921.1) (WP_001050548.1) is a bacterial surface protein involved in cell aggregation and biofilm formation [43]. The protein consists of multiple domains, N-terminal side includes the YSIRK type signal peptide and microbial surface components recognizing adhesive matrix molecules (MSCRAMM) family adhesin SdrC domain. The peptide of choice is part of MSCRAMM domain (Figure 2a). The signal peptide cleavage site was predicted by SignalP 4.1 server to be between positions 37 and 38. The target peptide subcellular localization was predicted to be extracellular by CELLO (v2.5).
Protein topology prediction using TOPOCONS confirmed the signal peptide location, while the targeted region is predicated to be extracellular (Figure 2b). Homology model showed that the selected peptide is surface-exposed and accessible (Figure 2c). These analyses indicated that this region could be a potential good target for the proposed diagnostic kit.
Searching the NCBI Conserved Domain Database revealed that the parent protein NWMN_ 1649 (BAF67921.1)(WP_001050548.1) is a bacterial surface protein involved in cell aggregation and biofilm formation [43]. The protein consists of multiple domains, N-terminal side includes the YSIRK type signal peptide and microbial surface components recognizing adhesive matrix molecules (MSCRAMM) family adhesin SdrC domain. The peptide of choice is part of MSCRAMM domain (Figure 2a). The signal peptide cleavage site was predicted by SignalP 4.1 server to be between positions 37 and 38. The target peptide subcellular localization was predicted to be extracellular by CELLO (v2.5).
Protein topology prediction using TOPOCONS confirmed the signal peptide location, while the targeted region is predicated to be extracellular (Figure 2b). Homology model showed that the selected peptide is surface-exposed and accessible (Figure 2c). These analyses indicated that this region could be a potential good target for the proposed diagnostic kit. Topology prediction of the NWMN_1649 whole protein as determined using the protein topology prediction tools TOPCONS and CCTOP (Consensus Constrained TOPology prediction web server); black boxes indicate the signal peptide, the gray boxes indicate transmembrane helix (in to out), and the blue lines indicate that this sequence is predicted to be extracellular. (c) Homology modelling of the NWMN_1649 whole protein with the target peptide (46-157) highlighted in blue being predicted to be surface-exposed and accessible. The modeling was performed using Modeller and UCSF-Chimera.

Successful Cloning, Expression, and Purification of NWMN_1649 (46-157) Polypeptide
It was crucial to ensure that NWMN_1649 protein is expressed by S. aureus under relevant conditions. RT-RT-PCR analyses indicated that NWMN_1649 is expressed either when incubated in PBS or heat-inactivated serum. However, there was a significant expression increase with more than 18-fold (18.51 ± 4.137) when S. aureus was incubated in the presence of serum (Figure 3a). Given the of the NWMN_1649 whole protein as determined using the protein topology prediction tools TOPCONS and CCTOP (Consensus Constrained TOPology prediction web server); black boxes indicate the signal peptide, the gray boxes indicate transmembrane helix (in to out), and the blue lines indicate that this sequence is predicted to be extracellular. (c) Homology modelling of the NWMN_1649 whole protein with the target peptide (46-157) highlighted in blue being predicted to be surface-exposed and accessible. The modeling was performed using Modeller and UCSF-Chimera.

Successful Cloning, Expression, and Purification of NWMN_1649 (46-157) Polypeptide
It was crucial to ensure that NWMN_1649 protein is expressed by S. aureus under relevant conditions. RT-RT-PCR analyses indicated that NWMN_1649 is expressed either when incubated in PBS or heat-inactivated serum. However, there was a significant expression increase with more than 18-fold (18.51 ± 4.137) when S. aureus was incubated in the presence of serum (Figure 3a). Given the results of expression and topology prediction, the NWMN_1649 (46-157) polypeptide is surface-exposed and expressed under normal growth conditions, thus it can be considered a promising diagnostic target.
The DNA fragment encoding the selected peptide was amplified using primer pair AA726/AA727 resulting in the predicted DNA fragment size of 342 bp (Figure 3b and Figure S2A). The DNA fragment of pET-15b vector containing the T7 promotor, the His-Tag, and the multicloning site was amplified with primer pair AA667/AA666 resulting in a 490 bp fragment, upon NdeI digestion of this fragment, it resulted into a 440 bp fragment (Figure 3b). Upon joining the two fragments via DNA ligation followed by PCR amplification with primer pair AA667/AA727, the resultant DNA fragment was obtained with a predicted size of 782 bp (Figure 3b and Figure S2A). The AA667/AA727 fragment was cloned into the rapid TA cloning vector pLUG-Prime ® -TA-Cloning Vector II and transformed into E. coli. Resultant colonies were screened using AA437/AA438 primers which bind on the plasmid backbone outside the cloning site. The clones which gave PCR bands with the predicted fragment size of 965 bp (Figure 3c and Figure S2B) were considered potential right clones. DNA sequencing confirmed the right sequence of the insert and its orientation.
The cloned DNA fragment encoded for a 133 aa polypeptide; 112 aa derived from the NWMN_1649 protein and 21 aa from the fragment cloned from pET-15b including an N-terminal 6xHis-tag fragment. The whole fragment was predicted to be 13.76 kDa with the 112 aa of NWMN_1649 corresponding to 11.45 kDa and the 21 aa from pET-15b correspond to 2.31 kDa ( Figure S2C). Upon eluting the His-tagged polypeptide from nickel column and running the fraction in SDS-PAGE gel, prominent band showed up around~28 kDa (Figure 3d). This band most probably corresponds to the dimeric form of the purified polypeptide.
Biomolecules 2020, 10, x 7 of 14 results of expression and topology prediction, the NWMN_1649 (46-157) polypeptide is surfaceexposed and expressed under normal growth conditions, thus it can be considered a promising diagnostic target. The DNA fragment encoding the selected peptide was amplified using primer pair AA726/AA727 resulting in the predicted DNA fragment size of 342 bp (Figure 3b and Figure S2A). The DNA fragment of pET-15b vector containing the T7 promotor, the His-Tag, and the multicloning site was amplified with primer pair AA667/AA666 resulting in a 490 bp fragment, upon NdeI digestion of this fragment, it resulted into a 440 bp fragment (Figure 3b). Upon joining the two fragments via DNA ligation followed by PCR amplification with primer pair AA667/AA727, the resultant DNA fragment was obtained with a predicted size of 782 bp (Figure 3b and Figure S2A). The AA667/AA727 fragment was cloned into the rapid TA cloning vector pLUG-Prime ® -TA-Cloning Vector II and transformed into E. coli. Resultant colonies were screened using AA437/AA438 primers which bind on the plasmid backbone outside the cloning site. The clones which gave PCR bands with the predicted fragment size of 965 bp (Figure 3c and Figure S2B) were considered potential right clones. DNA sequencing confirmed the right sequence of the insert and its orientation.
The cloned DNA fragment encoded for a 133 aa polypeptide; 112 aa derived from the NWMN_1649 protein and 21 aa from the fragment cloned from pET-15b including an N-terminal 6xHis-tag fragment. The whole fragment was predicted to be 13.76 kDa with the 112 aa of NWMN_1649 corresponding to 11.45 kDa and the 21 aa from pET-15b correspond to 2.31 kDa ( Figure  S2C). Upon eluting the His-tagged polypeptide from nickel column and running the fraction in SDS-PAGE gel, prominent band showed up around ~28 kDa (Figure 3d). This band most probably corresponds to the dimeric form of the purified polypeptide.

Assessment of the Obtained Anti-Sera and Construction of the ICS
The sera of the mice immunized with the NWMN_1649 polypeptide produced titers ranged between 6.4 ×10 4 -2.56 × 10 5 . A diagrammatic sketch of the strip design is presented in Figure 4a. The developed strip was challenged with aliquots of purified polypeptide andwith S. aureus cell suspensions. The ICS showed positive signal with as low as 5 µg polypeptide (Figure 4b) and as low as 10 2 CFU/mL S. aureus (Figure 4c).
Biomolecules 2020, 10, x 8 of 14 elution fractions "E" of the purification of the cell lysates of the E. coli cells carrying the pNWMN_1649-6xHis. The first lane contained a protein marker for molecular mass estimation.

Assessment of the Obtained Anti-Sera and Construction of the ICS
The sera of the mice immunized with the NWMN_1649 polypeptide produced titers ranged between 6.4 ×10 4 -2.56 × 10 5 . A diagrammatic sketch of the strip design is presented in Figure 4a. The developed strip was challenged with aliquots of purified polypeptide andwith S. aureus cell suspensions. The ICS showed positive signal with as low as 5 µg polypeptide (Figure 4b) and as low as 10 2 CFU/mL S. aureus (Figure 4c).

The Developed ICS Specifically Detects S. aureus in Neonatal Sepsis Samples
Blood was collected from neonatal sepsis cases diagnosed with different microbial infections. The developed ICS gave only positive results with samples infected with S. aureus (Figure 5a), while those infected with K. pneumoniae or coagulase-negative staphylococci (CoNS) showed negative

The Developed ICS Specifically Detects S. aureus in Neonatal Sepsis Samples
Blood was collected from neonatal sepsis cases diagnosed with different microbial infections. The developed ICS gave only positive results with samples infected with S. aureus (Figure 5a), while those infected with K. pneumoniae or coagulase-negative staphylococci (CoNS) showed negative results (Figure 5b). These results indicate that the developed ICS can specifically detect the target microorganism (S. aureus) in this limited pool of real neonatal sepsis samples.
Biomolecules 2020, 10, x 9 of 14 results (Figure 5b). These results indicate that the developed ICS can specifically detect the target microorganism (S. aureus) in this limited pool of real neonatal sepsis samples. Figure 5. The developed ICS specifically detects S. aureus in neonatal sepsis blood samples. (a) Photographs of two representative ICS upon testing using two blood samples from neonatal sepsis cases infected by S. aureus. (b) Photographs of two representative ICS upon testing using two blood samples from neonatal sepsis cases infected by K. pneumoniae or coagulase-negative staphylococci (CoNS). One drop (~15 µL) of the blood samples was diluted to 150 µL using sterile PBS prior to loading on the sample pad and allowing it to flow for 10-15 min at room temperature.

Discussion
Neonatal sepsis is a leading cause of death in both developed and developing countries [2,44]. Immunochromatographic strip tests are well known for bacterial and viral antigens detection [8,45]. Yet so far, no ICS has been developed to directly detect S. aureus in sepsis samples.
Bioinformatics tools have been widely used for drug targets and vaccine candidates' prediction in various microorganisms [46]. In the current study, an extensive bioinformatics approach managed to identify the target polypeptide. The analyses collectively predicted that this polypeptide is a promising target being species-specific, conserved among many sequenced S. aureus strains, and predicted to be surface-exposed, and accessible.
Many S. aureus sequence types (ST) have been found in the literature to be commonly associated with morbidity and mortality of neonates in neonatal intensive care units, hospital-acquired infections, septicemia and neonatal sepsis including ST 1,ST 8,ST 59,ST 72,ST 93,ST 239,. The % identity within the first 500 hits in the BLAST analysis was very high (97.92 to 100%), indicating that this target is highly conserved in S. aureus strains in general including those belonging to sequence types known to be involved in fatal infections.
Real-time RT-PCR was used to ensure the optimal expression of the chosen protein candidate under normal growth conditions that would allow the use of this antigen as a target [53]. This tool confirmed the expression of our target protein, NWMN_1649, and its expression up-regulation upon exposure to human serum. On cloning and expression of the selected polypeptide as a 6xHis fusion protein, the produced protein showed up on SDS-PAGE around the size of a dimer rather than a monomeric size. It appeared so with or without using excessive reducing conditions and upon using washing and elution buffers with lower pH and also with high NaCl concentration in washing buffer. This phenomenon has been previously found with several proteins [54,55]. Wu and co-workers

Discussion
Neonatal sepsis is a leading cause of death in both developed and developing countries [2,44]. Immunochromatographic strip tests are well known for bacterial and viral antigens detection [8,45]. Yet so far, no ICS has been developed to directly detect S. aureus in sepsis samples.
Bioinformatics tools have been widely used for drug targets and vaccine candidates' prediction in various microorganisms [46]. In the current study, an extensive bioinformatics approach managed to identify the target polypeptide. The analyses collectively predicted that this polypeptide is a promising target being species-specific, conserved among many sequenced S. aureus strains, and predicted to be surface-exposed, and accessible.
Many S. aureus sequence types (ST) have been found in the literature to be commonly associated with morbidity and mortality of neonates in neonatal intensive care units, hospital-acquired infections, septicemia and neonatal sepsis including ST [47][48][49][50][51][52]. The % identity within the first 500 hits in the BLAST analysis was very high (97.92 to 100%), indicating that this target is highly conserved in S. aureus strains in general including those belonging to sequence types known to be involved in fatal infections.
Real-time RT-PCR was used to ensure the optimal expression of the chosen protein candidate under normal growth conditions that would allow the use of this antigen as a target [53]. This tool confirmed the expression of our target protein, NWMN_1649, and its expression up-regulation upon exposure to human serum. On cloning and expression of the selected polypeptide as a 6xHis fusion protein, the produced protein showed up on SDS-PAGE around the size of a dimer rather than a monomeric size. It appeared so with or without using excessive reducing conditions and upon using washing and elution buffers with lower pH and also with high NaCl concentration in washing buffer. This phenomenon has been previously found with several proteins [54,55]. Wu and co-workers showed that the His-tag presence in recombinant purified protein could mediate dimer formation [56]. They suggested that the His-tag could have affinity to an amino acid patch at the protein surface. They gave evidence of His-tag-dependent protein dimerization which should be taken into consideration while analyzing Hig-tagged proteins biochemical properties [56]. Another study has purified recombinant YukE protein, which is a secreted protein of the WXG100 superfamily in Bacillus subtilis, and by size exclusion chromatography they showed that it forms a dimer in vitro; the dimer band remained upon changing pH and the monomer was not observed with changing conditions [57]. ELISA has been widely used in previous studies for evaluation of binding of the produced antibodies to their specified antigens [45,58,59]. In this study, we have assessed the specificity of the produced antibodies in the sera of immunized mice with ELISA using purified specific NWMN_1649 polypeptide; which showed good titers.
Several approaches have been employed before for the detection of S. aureus. For instance, PCR was used to amplify the nuc gene that encodes a thermostable nuclease as a rapid diagnostic tool for S. aureus infections by direct testing of the clinical samples. However, performing PCR would require specific equipment and well-trained personnel [60]. Another group developed a method depending on measuring the resonance of the light-scattering signal of aptamer-conjugated-gold nanoparticles; it is low-cost and highly sensitive but it needs a relatively long time (1.5 h) and requires well-trained personnel [61]. Latex agglutination assays were widely used for the detection of S. aureus infections as rapid, simple, and sensitive methods; however, they suffered from low specificity especially in the presence of S. saprophyticus [62].
A very recent study reported the development of an ICS for detecting S. aureus in milk samples [63]. However, they used a ribosomal protein as a target. Accordingly, the detection required excessive sample processing including cell wall digestion and cell lysis prior to application on ICS which is a major drawback that has been avoided in the current study. Moreover, their limit of detection was as high as 10 4 CFU/mL. In addition, in 2014, Niu and co-workers have developed an ICS to detect S. aureus in simulated food samples and had a limit of detection as low as 10 3 CFU/mL [64]. This ICS was based on a commercially available monoclonal antibody (ab37644)that cross-reacts with CoNS S. epidermidis according to the vendor's description (Abcam, Cambridge, MA, USA)and this mAb production has been discontinued in June 2020. In addition, other studies were directed toward S. aureus detection in food specimens. However, they also required extensive sample treatment before testing [65,66]. This was avoided in the current study and we could directly test blood samples without prior treatment which would save time and effort. The designed strip had reaction time of 10-15 min which is similar to those described in previous studies [7,67]. Another study designed ICS for S. aureus detection in respiratory samples [58]; although direct sample detection was possible, the detection limit was too high 10 6 CFU/mL and the reaction toward S. epidermidis was not evaluated, which is a major blood contaminant in neonatal sepsis and can easily cross-react with antibodies to S. aureus antigens. It is worth mentioning that the ICS developed in the current work was non-reactive to a cell suspension of S. epidermidis, which is an additional advantage to the current tool. Accordingly, an additional advantage of the current ICS is its low limit of detection, as low as 10 2 CFU/mL of S. aureus, which is at least 10-fold less than previously reported systems.
Potential limitations of the current study may include the lack of sensitivity of the developed ICS to some S. aureus strains belonging to sequence types which were not represented in the 500 hits checked in the BLAST analysis. This could be possibly addressed in future studies by expanding the range of the isolates tested to ensure good coverage of all the possible sequence types.

Conclusions
In this study, we have applied an innovative bioinformatics approach to identify a specific, surfaceexposed biomolecule to be used as a successful target for the development of a sensitive, specific, simple, and rapid ICS for detecting S. aureus causing neonatal sepsis. The ICS successfully provides a promising alternative prototype for S. aureus detection especially in resources-limited countries. Future plans include large scale clinical testing of the developed prototype and usage of the selected antigen to generate specific monoclonal antibodies to maintain a sustainable source for production of this ICS.
Supplementary Materials: The following are available online at http://www.mdpi.com/2218-273X/10/11/1580/s1, Figure S1: Phylogenic tree of the blast data in tree representation showing target peptide relationship to its nearest blast hits. Figure S2: Confirmation of the NWMN_1649-6xHis fusion protein construct. Table S1: Results of BLAST analysis of the DNA sequence encoding the target peptide with results of the strains belonging to different STs known to be involved in neonatal infections highlighted.