Towards nano-diagnostics for bacterial infections

General

Streptococcus pneumoniae (pneumococcus) is a spherical, Gram-positive bacterium colonizing the nasal area and capable of causing life-threatening disease such as pneumonia, meningitis, bronchitis, sinusitis and otitis media in both developing countries and industrialized countries.

To date, 93 different serotypes are identified that all differ in chemical structure and immunogenicity, posing a challenge for diagnostics (5). Pneumococcal disease causes over 1.6 millions of deaths annually, mostly among children, older people and patients with immunodeficiencies (6–8). Pneumococcal infection is frequently an indication of a defect in the defense of the host, an overwhelming inoculum, an exposure to a particularly virulent micro-organism, a hematogenous spread from a distant infected site or pulmonary aspiration. Virulent factors of S. pneumoniae include the capsular polysaccharides, the pneumococcal surface protein A (PspA) and protein C (PspC), the pneumococcal surface adhesin A (PsaA) or Hic (see Figure 1). PspA is expressed in all strains of S. pneumoniae (9). The N-terminus is exposed on the bacterial surface and is classified into three families based on sequence homologies in the DNA. Ninety-five percent of all S. pneumoniae carry PspA of family 1 and 2 (PsapA1 and PsapA2 respectively). PspC has a structure related to PspA, however the N-terminus is different. This protein is present in about 75% of all pneumococci types and interferes with the complement system by binding of factor H in human plasma, resulting in protection against both complement attach and phagocytosis. PspC exists also in another variant, the factor H-binding inhibitor of complement (Hic), which is mainly found in serotype 3 (10–12).

Figure 1:

The virulent factors of E. pneumoniae [“Streptococcus pneumoniae. Textbook in Diagnosis, Serotyping, Virulence Factors and Enzyme-linked Immunosorbent Assay (ELISA) for Measuring Pneumococcal Antibodies.” (2nd version) Statens Serum Institut, Denmark].

Linder and coworkers (8) investigated the human antibody response towards PspA, PspC and Hic of 41 patients during an invasive infection and found that a strong immune response against PspA usually develops during recovery. The high degree of cross-reactivity between the PspA and PspC antibodies renders them less suitable for diagnostics (8). Wright and coworkers (13) showed that persistent antigen exposure from S. pneumonia colonization can induce protective defenses of the immune system against carriage and disease. The immunoglobin responses were directed towards various protein targets except the capsular polysaccharides (13).

Diagnostics

The diagnosis of pneumococcal disease requires the isolation of the organism from a sterile site of the human body such as blood, ascites, pleural fluid or cerebrospinal fluid (14). The current reference methods for diagnosis are culture, ELISA and polymerase chain reaction (PCR). When using rapid tests of respiratory secretions for diagnosis, an additional confirmation method is usually required to exclude the possibility of carriage in the nasopharynx area. In young children (<10 years) such carrier state is found in 30%–60%, whereas in adults the prevalence is 1%–10% (14). The conventional Neufeld “Quellung” reaction may also be used for diagnosis but is expensive, labor-intensive and prone to errors. Its developer, the German bacterist Friedrich Neufeld discovered that antibodies against specific capsular antigens of S. pneumonaie can be produced and used to distinguish between pneumococcal serotypes. After antibody binding and methylene Blue staining, the bacteria look swollen (german “Quellung”) when visualized under the microscope.

The genotypic and serotypic variability of pneumococci is a challenge (15, 16) that requires particular attention when developing new diagnostics. Leung and coworkers (17) presented the “sequetyping” method, a single PCR sequencing strategy for the serotyping of pneumococci by using a single primer pair. The capsulation locus (cps) of 23 different vaccine serotypes was investigated by using the alignment software ClustalX. Two well conserved primer binding sites were selected by the Primer-Finder algorithm covering the regulatory gene cpsB. This selected region at the end of the cps region is specific for S. pneumoniae. In silico predictions revealed that from the current know 92 serotypes, 84 serotypes will deliver a PCR product. The nucleotide order was determined by cycle sequencing and using the same primers. The sequences were compared with the GenBank database and a BLAST score of >98% was considered as the correct serotype. According to the in silico predictions, 46 serotypes could be identified by the selected region. The primers were used to analyze 138 pneumococcal strains that included 48 serotypes. Reproducibility was proven by the correct identification of different strains of a serotype. The 23 vaccine serotypes were identified correctly for 86%. This method is promising for diagnosis and the identification of new serotypes but not suitable for use in resource poor environments. Tuerlinckx and coworkers (18) evaluated a serotype-specific ELISA for the etiologic diagnosis in children with pneumonia acquired in community (CAP). All children (<15 years) had a CAP confirmed by a radiogram (n=163). Pneumococcal infection was confirmed by positive blood and/or pleural fluid culture (n=35) and the non-proven pneumococcal patients (n=128) were also evaluated. The Quelllung reaction was used to define the serotype and ELISA was used to diagnose the IgG and IgA antibodies specific for nine serotypes. The serological response rate with the ELISA was 82.8% and the serotyping results agreed well with the results of the Quellung reaction. With this ELISA it was possible to demonstrate the presence of pneumococcal CAP in 55% of the children with a negative culture result.

Pereira et al. (19) measured the levels of C-reactive protein (CRP), lactate, leucocytes, procalcitonin (PCT), D-dimer, cortisol and brain natriuretic peptide (BNP), in patients with severe CAP within 12 h after receiving the first antibiotic dose. The patients (n=64) were admitted to the intensive care unit and 33 had a confirmed pneumococcal infection, 13 had other bacterial causes (including E. coli, S. aurus, L. pneumophila) and 18 patients had viral infection only in particular H1N1 and H3N2. In patients with pneumococcal disease, significantly higher levels of lactate, BNP, CRP and PCT were found. PCT at a level of >17 ng/mL could identify the patients with severe pneumococcal CAP that may profit from antibiotics, while at lower PCT levels pneumococcal CAP was unlikely. Thus, multiparameter testing for pneumococcal disease is often desirable [see Salieb-Beugelaar and Hunziker (1)].

Microfluidics for S. pneumoniae

Van Heirstraeten (20) and coworkers presented a microfluidic device for analysis of community acquired lower respiratory tract infections that is capable of performing automated sample preparation, including lysis, nucleic acid purification and concentration. After isolation of the nucleic acids, reverse transcription PCR or a fluoremetric assay was performed to measure nucleic acid concentration. Sample preparation in this microdevice resulted in higher or similar concentration of bacterial DNA or viral RNA compared to the conventional benchtop experiments. The device was tested with swabs as well as cultured microorganisms and represents a step forward to the application in a point-of-care test for rapid diagnosis of community acquired lower respiratory tract infections. Dhoubhadel et al. (21, 22) developed a nanofluidic real time PCR system capable to identify 50 serotypes. 194 patients (age <5 years, culture positive, lytA PCR positive) and 140 healthy children (age <5 years, culture and lytA PCR postitive) were investigated with this method. The lytA PCR of nasopharyngeal samples was included to confirm pneumococcus. They showed that a higher bacterial load of a serotype in the nasopharynx suggests higher transmission of this serotype. In addition, the co-colonization of multiple serotypes was associated with acute respiratory infections. Figure 2 presents the bacterial load of specific serotypes of S. pneumoniae in patients and healthy children.

Figure 2:

The bacterial loads of specific serotypes in patients (red) and healthy children (blue) [With permission reused from PLoS ONE from: Dhoubhadel BG, Yasunami M, Nguyen HAT, Suzuki M, Vu TH, Thi Thuy Nguyen A, et al. PLoS One 2014;9:e110777. doi:10.1371/journal. pone.0110777 (22)].

An alternative to PCR, which requires a protocol cycling through of different temperatures, loop mediated isothermal amplification of nucleic acids (LAMP) can be used. Luo et al. (23) developed an operationally simple and cost-time effective microfluidic device for LAMP. Its capabilities were documented with the detection of K. pneumonia, M. tuberculosis and H. influenza; while S. pneumoniae lacking from the panel, its inclusion is not expected to pose major challenges). The amplification of the target was measured by the electrochemical signal of methylene blue at eight etched indium tin oxide electrochemical reactors. The potential of this approach was highlighted with the analysis of multiple genes both qualitatively as quantitatively with a limit of detection of 16, 17 and 28 copies/μL for H. influenza, K. pneumoniae and M. tuberculosis respectively.

Nanodiagnostics for S. pneumoniae

Shi et al. (24) presented a rolling cycle amplification of DNA using gold nanoparticles as surface plasmon resonance sensors. The probes included long oligonucleotides, whereby boundary sequences were complementary to the adjacent sequences (also called padlock probes). These probes were coupled to the gold nanoparticles and were specific for bacterial pathogen sequences in 16S rDNA. Hybridization of the target with the probes brings the two ends in contact, resulting in circularization. The advantage of these probes is the flexibility for test development and the potential for targeting a variety of organisms (for further reading, see Szemes et al. (25). Six different pathogens were investigated: E. coli, S. dysenteriae, S. epidermidis, A. aureus, E. faecalis and S. pneumonia. The device identified the six pathogens with 0.5 pM probe quantities and a limit of detection of 0.5 pg/μL genomic DNA in clinical samples.

An electronic nose to detect the complex metabolites that are produced by microorganisms during an infection was presented by Tang et al. (26) following prior work by Schmid (27). The miniaturized gas sensing and battery powered device has 8 sensors incorporated (manufactured of polymer-carbon nanocomposites) and includes a learning kernel. The capability to identify ventilator-associated pneumonia (VAP) was verified in clinical trial where 74 infected samples with confirmed Klebsiella (n=35) and Pseudomas aeruginosa (n=39) including 43 controls were analyzed. Upon usage of the learning kernel, the accuracy was for the infected patients improved from 91.89 to 100%. Even though not all presented examples include the detection of S. pneumoniae, it is clear that the detection of microbiological pathogens is moving into the field of nanodiagnostics.

General

Waterborne diseases are affecting thousands of lives. Poor sanitation and limited access to clean water are the main causes. Salmonella enterica serotype Typhi is a gram-negative rod shaped and flagellated bacterium responsible for typhoid fever, an infection restricted to humans. Globally, 22 million individuals are infected annually of which ∼200.000 die (28). S. typhi and the related bacterium S. paratyphi A, which leads to a similar clinical syndrome, belong to the large family of Salmonella bacteria. Over 2000 serotypes are known which are classified by their differential surface “H” (flagella) and “O” (lipopolysaccharides, LPS) antigens. Types with a common “O” antigen are grouped together. S. typhi belongs to serogroup D that has O9 and O12 in common (29). Infection with S. typhi begins with the ingestion of contaminated food, followed by invasion of the gastrointestinal mucosa by the bacteria and translocation to the lymphoid follicles (Figure 3) where the bacteria can survive and even replicate inside macrophages. Subsequently, they are spread via the bloodstream to the spleen, liver and intestinal lymph nodes (30). The typical incubation period is between 8 and 14 days (31). Symptoms include fever, weakness, anorexia, abdominal pain and potential complications are gastrointestinal bleeding, intestinal perforation, and encephalopathy (32).

Figure 3:

Dissemination of S.typhi in the host during a systemic infection [With permission reused from PLoS Pathogen from: de Jong HK, Parry CM, van der Poll T, Wiersinga WJ, PLoS Pathog 2012;8:e1002933. doi:10.1371/journal.ppat.1002933].

Immune cells and antibody-mediated immune responses play a role in controlling and clearing S. typhi infection (33). Vaccines are available, but not very effective, although with the emergence of multidrug resistant bacteria, the development of new vaccines would be very urgent (34).

Diagnostics

The diagnosis of a S. typhi infection can be done by the cultivation of blood, stool, urine or duodenal contents. The specificity of a culture is considered as 100%. The reported sensitivities in literature vary. Gotuzzo and coworkers (35) showed that the organism could be isolated from 98% of bone marrow cultures of patients, while only 70% were blood culture positive. Even after five days of antibiotic therapy, such bone marrow cultures can remain positive (36). Within the first week of onset, cultures from blood, stool and intestinal secretion are positive in ∼85%–90% of the cases. In a later stage of the infection, this is 20%–30% lower. The use of stool cultures alone yielded a sensitivity of <50% and for urine this is even less (37). Another method to diagnose is detection of target bacterial DNA in samples by PCR. However, the number of bacteria circulating in blood of patients with bacteremia is <1 CFU/mL blood (2). Thus, the human DNA is dominating, which may result in false negative results due to reduced sensitivity or false positive results due to non-specific binding of the primers. Zhou et al. (38) developed a method to enrich the target bacterial fliC-d gene. Human DNA is removed from the blood samples. Ox bile is used to lyse the human blood cells, while S. typhi is resistant. The released Human DNA is subsequently degraded by Micrococcal nuclease. Experiments with spiked blood samples showed that this method enhanced the PCR sensitivity by 1000 fold. Urine may also be used to diagnose typhoid fever by PCR as developed by Kumar et al. (39). This two-step PCR yielded a sensitivity and specificity of 90.9% and 100% respectively for blood samples, 95.5% and 100% respectively for urine samples and 68.1% and 85% for stool samples. Blood cultures had a sensitivity of 31.8% and a specificity of 100%. An intriguing observation was reported by Ahirwar et al. (40). They exposed urine, blood and stool samples of acute typhoid fever patients (n=90) and chronic typhoid carriers (n=36) to acidic pH. Isolation after the acidic exposure yielded 77.7% of the acute patients to be positive for S. typhi, compared to 8.8% provided with the conventional method. In chronic carriers provided 42% positive samples for S. typhi that was 5.5% when the conventional method was applied. The acid shock enhances the multiplication of the bacteria and as a consequence increases the isolation rates from samples.

The application of serological tests for S. typhi is limited as a result of false positives due to prior infections or cross-reaction with non-Salmonella bacteria. Therefore, a positive result always requires an additional confirmation method as a positive culture or PCR. A common method is the agglutination test introduced in the late nineteenth century by Widal and Secard, where specific antibodies in patient serum result in the aggregation of a cell suspension. In case of a S. typhi infection, these antibodies (the agglutinins) are directed against the H (flagellar) and the O (somatic) bacterial antigens. Bakr and coworkers (41) compared the sensitivity, specificity and accuracy of 4 different brands of Widal tests for the anti-H and O antibodies. 150 clinically suspected patients and a negative control group (n=25) were investigated. As reference test an IgM anti-LPS ELISA was used (91/150). Serum samples were diluted (1/80, 1/160 and 1/320). They concluded that a four-fold rise of the antibody titer was not demonstrable (clinical diagnosed patients) and H agglutinins were less specific and sensitive when comparing to O agglutinins. Even though the Widal tests cannot be used to confirm an infection due to the low sensitivity and specificity, in the absence of other methods in resource poor areas, this test is a possible alternative. In addition, a negative test is a good indication for the absence of the disease as concluded by Andualem et al. (42). They investigated a Widal test and found a sensitivity of 71.4%, a specificity of 68.4%, a positive predicitive value of 5.7% and a negative predictive value of 98.9%. For further reading, see Olopoenia et al. (43), who reviewed this test since its introduction in 1896.

Another rapid test is the Tubex method, to detect IgM specific for the O9 LPS in patient sera. Ley and coworkers (44) investigated the use of Tubex among hospitalized children (n=139) in Tanzania with blood culture as reference method (n=33), culture confirmed non-typhi infections (n=49) and other non salmonella infections (n=57). The non-typhi samples were used as control, resulting in a sensitivity of 79% and a specificity of 89%. They concluded that apart from a shorter time to obtain results, results of the Tubex test are comparable to the Widal test.

Keddy et al. (45) investigated the performance of various rapid antibody diagnostic tests with blood culture as the reference method in sub-Saharan African sites. They concluded that the single-tube Widal test and semi-quantitative slide agglutination test performed poorly. The included Tubex and Typhidot tests performed better but still not comparable tot the blood culture. The Tubex had a sensitivity and specificity of 73% and 69% respectively, Typhidot IgM 75% and 60.7% respectively and Typhidot IgG 69.2% and 70.4% respectively. It is thus evident that the available tests for diagnosis of salmonella infection leave a large unmet need for simple, inexpensive, reliable and fast point-of-care use. The search for new salmonella biomarkers is ongoing; Liang and coworkers (46) performed an protein microarray with 2724 S. typhi antigens (∼63% of all predicted proteins) and identified IgM antibodies against 77 and IgG antibodies against 16 novel antigen targets in the serum of acute typhoid patients and healthy controls. The antigens specific for IgMs produced a 97% & 91% sensitivity and specificity respectively, whereas the antigens specific for IgGs produced a 97% & 80% sensitivity and specificity respectively. This type of investigation contributes to both the understanding of the immune responses of the human body during infection and may lead to the development for novel rapid tests for diagnosis of typhoid fever.

Microfluidics for S. typhi

Limited work has been done on the application of microfluidic devices for the detection of antigens or antibodies specific for S. typhi. Lafleur and coworkers (47) developed a multilayered microfluidic immunoassay card for the detection of IgM antibodies against S. typhi LPS and P. falciparum HRPII antigen in blood discussed in Salieb-Beugelaar and Hunziker (1). The available samples were limited (n=9) and included positive and negative clinical samples. The detection of S. typhi IgM on the card was compared by laboratory ELISA and bench-top rapid flow-through membrane immunoassay. The volume of each sample was not enough to perform the tests in duplicate. The ELISA was done in duplicate. The intensity response of the card was comparable to both the bench-top assay and the ELISA. The clinical utility of the cards could not be evaluated as a result of the limited number of samples.

Nanodiagnostics for S. typhi

Das et al. (48) recently published an electrochemical DNA sensor for the detection of the S. typhi Vi gene. A self-assembled layer of organosilane 3-mercaptopropyltrimethoxysilane (MPTS) on top of a screen printed electrode was used to electrochemically deposit gold nanoparticles. On the surfaces of the nanoparticles, thiol-modified DNA probes were immobilized to detect target nucleic acid of S. typhi. After the characterization of the sensor by atomic force microscopy, cyclic voltammetry and electrochemical impedance spectroscopy, the performance of the sensor was measured by differential pulse voltammograms, using methylene blue as a hybridization indicator. In more than 98% of the isolates of patients, this gene was expressed. Linearity was from 1.0.10^–11 to 0.5×10^–8 M. The limit of detection was ∼50 pM. Typhi detection in real samples appears to be an ongoing project of this promising early work.

General

The gram negative, rod shaped and facultative anaerobic bacterium E. coli is responsible for outbreaks of both waterborne and foodborne diseases. Physiologically, this bacterium is present in huge numbers in the intestinal lumen of human and warm-blooded animals. While usually innocuous, a breakup of the gastrointestinal barriers even otherwise non-pathogenic E. coli strains can cause an infection. The bacterium is used as an indicator of fecal contamination in water quality assessment (49). Clinical syndromes associated with diarrheagenic E. coli include (1) watery diarrhea caused by enterotoxic E. coli (ETEC), (2) infantile diarrhea caused by enteropathogenic E. coli (EPEC), (3) hemorrhagic colitis and hemolytic uremic syndrome caused by enterohemorrhagic E. coli (EHEC), (4) dysentery caused enteroinvasive E. coli (EIEC) and (5) persistent diarrhea in children and patients infected with HIV caused by enteroaggregative E. coli (EAEC) (50, 51). EHEC is responsible for isolated outbreaks of diarrhea, bloody diarrhea and a cause of postdiarrheal hemolytic uremic syndrome (HUS) (52) where the O157:H7 serotype is the globally dominating cause (53). The diagnosis of EHEC is based on the detection of the Shiga toxin in feces and the isolation and characterization of the strain, but the causative E. coli often is no longer detectable at the time of onset of HUS, rendering diagnosis difficult. In addition, HUS may also be caused by other Shiga toxin producing serotypes including the S. dysenteriae serotype 1 (54). According to the WHO, ETEC is the leading bacterial cause of diarrhea in the developing world and is the most common cause of travelers’ diarrhea. It is estimated that approximately 210 million cases per year with around 380.000 deaths are caused by ETEC (55, 56). The strain is known to produce colonization factors that enhance adherence to the intestinal mucosa (57), in addition, the bacteria produce two types of toxins, the heat stable (ST) and/or the heat labile (LT) enterotoxin. The latter leads to an immune response in humans (58). For further reading, we recommend the review of Isidean and coworkers (59). Vaccines are available, but their the protection is limited in potency and usually abates with a few months. Steinsland et al. (60) reported that natural infections provide also protection as reflected by the decreasing rates of diarrhea with age and/or the lower ratios of symptomatic infections with increasing age. This suggests the hypothesis that immunization in early life might represent an effective preventive strategy.

Diagnostics

For the diagnosis of these infections, stool cultures or PCR tests are the gold reference methods. There is ongoing PCR tests development as shown by the recently published multiplex PCR of Andrade and coworkers (61) for the detection of both typical and atypical EAEC. Typical EAEC expressing the AggR regulon is a global cause of childhood diarrhea (62), whereas atypical forms of EAEC do not express this regulon.

Andrade et al. (61) designed 4 amplicons of 4 different genes; aggR (346 bp), aaiA (476 pb), aatA (630 bp) and aaiG (782 bp). A multiplex PCR was performed with 406 E. coli isolates that were collected in an earlier study (63). The final collection included 199 non-pathenogenic E. coli, 3 EHEC, 17 ETEC, 16 EIEC, 81 DAEC, 32 EPEC (18 atypical, 14 typical) and 58 EAEC (20 atypical, 38 typical). From all strains, 75 were detected by the PCR with at least one PCR product, of which 55 belonged to the EAEC pathotype (17 atypical, 38 typical). A sensitivity of 94.8% and a specificity 94.3% were found with a detection limit of 125 ng of purified genomic DNA.

Rapid tests are used mostly for toxin detection in food samples or bacterial cultures. The Duopath Verotoxin rapid test was developed for the detection of Shiga toxin 1 and 2 (Stx1 and Stx2 respectively) of EHEC in human stool samples and were evaluated by Grif et al. (64). Compared with a commercial ELISA and PCR, the rapid test showed a lower sensitivity and specificity. For Stx1 the sensitivity and specificity for the rapid test was 46.15% and 79.8% respectively while the PCR was reported to be 100%. For Stx2 the sensitivity and specificity for the rapid test was 79.8% and 75% respectively while also here the PCR was declared to be perfect sensitivity and specificity. The EHEC-ELISA was not able to distinguish between the two toxins and provided a sensitivity of 94.6% and specificity of 96.3% compared to the PCR.

Specific antibodies against the lipopolysaccharides (LPS) of the bacteria can also be detected, however there is a chance of false positive due to the presence of cross-reactive antibodies against epitopes of LPS of different bacteria (including other E. coli strains) (65).

Recently, bacterial glycoengineering technology was used to generate recombinant glycoproteins that were composed of a polysaccharide (O121, O145 or O157) coupled to a carrier protein (66). They demonstrated that these glycoproteins could be used as antigen in ELISA to discriminate between confirmed infected patients (O157, O145 or O121) and healthy children (total n=71). Of all positive samples, the infection was confirmed by positive isolation of the bacteria from stools, detection of fecal Stx on Vero cells by using antibodies (Stx1 & Stx2) and/or PCR specific for the stx 1/2 genes by using a sample from the MacConky Sorbitol agar plate. Three different groups were composed and analyzed with the new glyco-ELISA. The first group included patients that were diagnosed for another infectious disease (n=14). All obtained samples were serologically negative for the three glycoprotein-antigens and hereby confirming their high specificity. The second group included patients with a diagnosis of HUS or bloody diarrhea (n=13). 92.3% of the analyzed samples were serologically positive for the glycoprotein-antigens (7.7% O121, 30.8% O145 and 53.8% O157). The third group included patients with a clinical diagnosis of HUS or bloody diarrhea without a confirmation of the infection by stool culture of a specific PCR (n=44). 79.5% of these samples were serologically positive (2.3% O121, 15.9% O145 and 56.8% O157). They also identified a specific IgM response in almost all samples indicating that it might be possible to diagnose an infection in the early stages.

Microfluidics for E. coli

Agrawal and coworkers (67) developed a PDMS multiplexed microfluidic device for the quantitative or the multiplexed detection of waterborne pathogens. The capability of the device was shown with the detection of E. coli and S. typhimurium. In Figure 4, a schematic representation of the immuno assay is presented. Each microfluidic channel contains three different zones: the pre-capture zone, the capture zone and the detection zone respectively. In each zone, a magnet is embedded. In the (pre)capture zones, the immunomagnetic separation of the target bacteria from the sample was achieved through the specific binding of the antibody conjugated magnetic nanoparticles with the bacteria. Subsequently, the bacteria-magnetic beads complexes are moved towards the detection zone where specific antibody labeled Quantum Dots are used to visualize the presence of the target bacteria. The working range for both types of bacteria was 10³–10⁷ CFU/mL. The capabilities of the device might be further investigated by the detection of E. coli in diluted stool samples.

Figure 4:

A. The microfluidic channels contain three different zones. The antibody conjugated magnetic nanoparticles are entrapped with a permanent magnet. In B and C the target bacteria is captured and in D the quantum dot labeled antibody is used to visualize the presence of the target bacteria [With permission reused from Springer from Agrawal S, Morarka A, Bodas D, Paknikar KM. Appl Biochem Biotechnol. 2012;167:1668–77 (67)].

Jian et al. (68) developed a high throughput microfluidic device for the detection of airborne bacteria including E. coli. Bioaerosols (made by using an aerosol generator) with numbers of E. coli ranging from 10⁴- 10⁶ CFU/mL where used to determine the sensitivity of the microfluidic chip. E. coli, C. koseri, S. aureus, K. pneumoniae, E. faecalisand P. aeruginosa were used to validate the system. For each organism an amplicon was designed such that the difference in between was ∼40 bp in order to differentiate between the PCR product by gel electrophoresis. The bacteria were first captured and then enriched by using microfluidic chip containing a staggered herringbone mixer structure, and consecutively they were amplified by the use of a high-throughput continuous-flow PCR chip. Amplified products were analyzed by 1.5% agarose gel electrophoresis. No cross-contamination between different channels was found, and proof of concept for efficient mixing of reagents and sample and finally precise control of fluids through microvalves (loading and mixing) was achieved.

Nanodiagnostics for E. coli

Pandey and coworkers (69) presented 3D cystine flowers and palm like structures that are used to develop high performance electrochemical immunosensors. The capabilities of these sensors were investigated by the detection of E. coli bacterium (O157:H7). A linear range from 10 to 3×10⁹ CFU/mL with a LOD of 4.7 CFU/mL was determined for the 3D flower and for the palm leafs the linear range was 10³–3×10⁹ CFU/mL and the LOD 9.64×10² CFU/mL. This difference could be explained by the nanostructured surface of the 3D flowers (confirmed by the increased surface roughness), leading to an increased surface area and consequently a larger available surface for the immobilization of antibodies. The shelf live of the sensors was 35 days at 4°C in PBS.

Cheng et al. (70) presented a novel amperometric immunosensor based on functionalized four-layer magnetic nanoparticles for the detection for E. coli (O157:H7). The core is magnetic Fe₃O₄, the second layer is Prussian blue, the third layer is N-(2-aminoethyl)-3-aminopropyltrimethoxylsilane and the final layer is the gold (Au) nanoparticle shell. Antibodies specific for O157:H7 were bound on the nanoparticle surface by using Au-SH bonds. (Turning of the electricity can regenerate the sensor). Heat killed E. coli was detected with a linear range of 3.6×10³–10⁶ CFU/mL. The investigators published the development of another novel sensor for the amperometric detection of E. coli (O157:H7) (71). Here, graphene oxide was used as a nanocarrier for the electrostatic adsorption of SiO₂ nanoparticles (Au coated) and immobilization of thionine (see Figure 5). On this surface, large amounts of G-quadruplex DNA and signal DNA were immobilized. After addition of hemin, a hemin/G-quadruplex is formed and acts as a signal tag for the detection of the eaeA gene of E. coli (O157:H7). The developed sensor is capable to detect the eaeA gene with a linear range of 0.02–50 nM and with a limit of detection of 0.01 nM. The latest discussed work opens the door to the development of nanodiagnostics for the detection of any pathogen.

Figure 5:

The fabrication of the nanocomposite DNA biosensor and the detection of the sample [With permission reused from Elsevier from Li Y, Deng J, Fang L, Yu K, Huang H, Jiang L, et al. Biosens Bioelec 2015;63:1–6 (71)].