Biosensor-based serological assay for diagnosing Helicobacter pylori infection

Helicobacter pylori causes the most common bacterial infection involving 50% of the global population. According to the World Health Organization H. pylori infection contributes to approximately 75% of the stomach cancer cases and 5.5% of all types of cancer. Therefore, timely diagnosis of the infection is highly desirable. Serological assays are widely performed for diagnosing H. pylori infection, the most frequently used one being ELISA. In the present study we showed that a serological assay can also be carried out using a biosensor based on Surface Plasmon Resonance (SPR). Unlike our previous studies where we used amplitude detection of the binding reactions, here we applied phase-sensitive detection. It was performed with a the channeled spectroscopic ellipsometer, which allowed fast measurement with high sensitivity. Thus, the detection limit achieved was more than two times lower than that of the amplitude detection. In terms of CFU, phase detection was sensitive even at 200 CFU, while amplitude detection was applicable at 3000 CFU.


Introduction
Helicobacter pylori causes the most common bacterial infection involving about 50% of the global population [1].The spread of infection depends on the socioeconomic conditions, level of urbanization, and sanitary conditions.The helical shape of the bacterium makes it easier for it to penetrate and cause an inflammatory reaction in the gastric mucosa with humoral and tissue immune reactions.Gastritis may remain stable or may progress to ulceration.
More than 90% of the people infected with H. pylori are diagnosed with gastroduodenal ulcer.According to the World Health Organization (WHO), H. pylori infection contributes to approximately 75% of the stomach cancer cases and 5.5% of all types of cancer [1].Тherefore, timely diagnosis of the infection is imperative.
Since the discovery of H. pylori numerous methods have been developed for its detection.The currently accepted 'gold standard' for diagnosing gastric H. pylori involves performing gastric biopsy culture and polymerase chain-reaction (PCR) [2,3].Thus, the advances in molecular technology avoid the potential difficulties encountered when other culture-based diagnostic methods are used.
Generally, the diagnostic methods are classified as invasive or non-invasive.The main limitation of the invasive test is that it allows only a small part of the gastric mucosa to be analyzed.Non-invasive tests fall into two categories: direct and indirect.Similar to the invasive tests, the direct ones detect direct evidence of H. pylori presence.The indirect tests detect the infection by evaluating indirect evidence, such as presence of H. pylori in the saliva, antibodies against H. pylori, or amount of labeled CO2 in the IOP Publishing doi:10.1088/1755-1315/1305/1/012019 2 breath, an expression of the urease activity of the bacterium [4].The antibody-based tests have been developed for the detection of specific anti-H.pylori immunoglobulin (IgG) antibodies in the serum of patients.Enzyme-linked immunosorbent assay (ELISA) has been used for the detection of antibodies against H. pylori antigens in serum, saliva and urine samples.
None of the above-listed tests is direct, i.e. able to detect the intact H. pylori bacteria.We have reported a direct method based on the binding reaction between BabA, the protein that builds the bacterial cell membrane and blood antigens Levis b (Le b ) [5].These molecular interactions were detected by a biosensor designed to combine a biorecognition component (bioreceptor) with a transducer, converting the biological activity into a measurable signal using surface plasmon resonance (SPR).
In the present work we explore the same binding reaction between BabA protein and blood antigens Le b , observed by phase-sensitive SPR detection.This type of detection allows the observation of a well pronounced signal even at lowest H. pylori concentrations.

Bacterial strain
Helicobacter pylori (DSM 21031) was obtained from Leibniz Institute, DSMZ-German Collection of Microorganisms and Cell Cultures, GmbH.

Cultivation procedure:
The procedure involved the following steps: i.
cultivation of the bacteria in the course of 1 -2 days at 37° C in an anaerobic container equipped with a Campylobacter gas pack to create a suitable gas culture medium.A thin layer of Brain Heart Infusion (Difco) liquid medium presents on the surface of the agar medium -Columbia Blood Agar.
ii. preparation of bacterial suspensions for treating the SPR biochip: a. preparation of a stock suspension with deionized water (density of 10 9 CFU / ml); b. preparation of ten-fold dilutions of this suspension at concentrations ranging from 10 8 to 10 2 CFU/ml.

Blood antigen:
The blood antigen used was Levis b Tetrasaccharide, L7659 purchased from Sigma Aldrich.

Evaluation of the H. pylori-Le b system:
In our experiment blood antigen Le b was the bioreceptor and a golden diffraction grating was the transducer.The biochip was prepared after the immobilization of the blood antigen on the transducer.Immobilizing the bioreceptor is a crucial step in the biosensor design.Matrix-assisted laser evaporation (MAPLE) was used for the bioreceptor deposition to avoid built-in matrix and related mediating molecules -the main causes for biosensor non-specific response.Details about the MAPLE-mediated immobilization can be found in our previously published results [6,7 ].

Surface plasmon resonance method
The method was successfully applied for SARS CoV-2 structural proteins detection [8,9].The optical set-up used for amplitude SPR detection is shown in figure 1. Spectral readout for SPR observation was used.The white light was collimated and passed through a polarizer controlling the polarization of the incident light.After its reflection frоm the transducer, the light was focused on a fiber In the present study we modified the SPR method used until now in terms of mode of interrogation: instead of amplitude detection of the resonance we applied phase-sensitive detection [10].
Our motivation was to improve detection sensitivity through decreasing the limit of detection (LOD).The main problem of the widely-used amplitude-sensitive SPR technology consists in the existence of a physical limit of LOD.As it has been demonstrated, the detection limit problem can be solved by the involvement of phase properties of light reflected under SPR.This problem is solved thanks to the sharp phase jump of reflected light in the very dip of the SPR amplitude curve.Figure 2 shows SPR measured on the golden diffraction grating (black curve) using the set-up shown in figure 1, and phase change provided by ellipsometric measurement on the same grating (red curve).The sharp phase jump ensures one order of magnitude higher sensitivity to the refractive index variation provoked by the biomolecular interactions.Phase control implies inherent relative measurement with respect to a reference beam which is the s -polarized component of light.Using a reference beam makes it possible to lower the LOD measured.Despite the above-stated distinct advantages of phase-sensitive SPR sensors, the majority of SPR sensors still rely on amplitude interrogation performed by angular or spectral readout.One reason is that phase retrieval requires more complicated optical instrumentation.In the present study we demonstrated that phase readout can be effectively accomplished by the channeled spectroscopic ellipsometer (CSE) proposed in [11].The CSE advantages include no requirement for mechanical or active components for phase retrieval, possessing a simple optical configuration that can be miniaturized, and ability to immediately determine the phase only from a single spectrum.
The principle of the spectroscopic ellipsometer is schematically illustrated in figure 3. A polychromatic light from a tungsten lamp transmits a polarizer P and is obliquely incident on the grating with immobilized Le b .The transmission axis of P is oriented at 45 • to the plane of incidence.The beam of light reflected from the grating passes through the CSE that consists of multiple order birefringent retarders R1 and R2 and an analyzer A. The fast axes of R1 and R2 are oriented at 0 • and 45 • to the fast axis, respectively, and the transmission axis of A is aligned with the fast axis.The light transmitted by A is analyzed by a spectrometer.
The CSE components are united in one holder incorporated in the set-up shown in Fig. 1 and placed in front of the lens 6.By means of Fourier transform methods Stokes parameters are found from the recorded spectrum and respectively the phase is retrieved.This study was performed in accordance with the Declaration of Helsinki.

Results and comments
A flow cell was made of polycarbonate capable of containing a biochip.Firstly, it was washed with deionized water to provide baseline measurement: the resonance wavelength for both amplitude and phase-sensitive detection was established.Then, bacterial suspensions of specified concentrations were applied to the biochip.Subsequently, the biochip was washed with running deionized water to remove any non-specific binding.After that, the resonance wavelength corresponding to the bacterial concentration was measured by both amplitude and phase-sensitive detection.The procedure was repeated for all concentrations of the bacterial suspension, using a fresh biochip for each concentration.
Figure 4 shows the results of the SPR measurement of the H. pylori-Leb binding reaction for the maximum and minimum bacterial concentrations.The measurements were provided by amplitude (solid curves) and phase-sensitive detection (dashed curves).The resonance appears as a smooth dip in the reflectivity spectrum at amplitude detection.As a result, accurate determination of the resonance wavelength was difficult.The dashed curves show the spectral distribution of the phase extracted from 5 the Stokes components.The phase jumps are well expressed and their spectral positions can be evaluated with higher accuracy.The data analysis aimed at calculating the spectral shift and was performed by subtracting the baseline resonance wavelengths from the values of the resonance wavelengths collected from both the amplitude and phase-sensitive measurements.Each point measurement was carried out in 20 repetitions, and the mean result was calculated to exclude influence of any noise caused by the environment or the detector itself -this was the basis for establishing measurement accuracy.Le b -H.pylori binding was well pronounced for concentrations ranging between 10 3 -10 8 CFU/ml: the spectral displacement was in the range 1 -7 nm.The interaction with 10 3 and 10 2 CFU/ml of H. pylori generated spectral displacement slightly above the measurement accuracy of the spectrometer which is better 1 nm.The overall measurement accuracy was evaluated approximately at ± 1 nm for amplitude detection, while LOD was close to 1.8 nm -figure 5 (a).However, the concentration 10 2 CFU/ml could not be distinguished from 10 3 .The minimum H. pylori concentration with a reliably observable binding to Le b was approximately 3 000 CFU.
For phase-sensitive detection the overall measurement accuracy was evaluated approximately at ± 0.3 nm while LOD was lower than 1 nm figure 5 (b).Now, a reliably observable binding occurred at approximately 200 CFU, i.e. more than one order of magnitude improvement in sensitivity was achieved.

Conclusions
The results obtained allow us to draw the following conclusions: • SPR biosensors that exploit Le b -H.pylori binding reaction are a reliable diagnostic tool because they detect intact H. pylori bacteria; • Le b -H.pylori binding is well expressed with high affinity; • SPR phase -sensitive detection is very effective, since it improves sensitivity by more than one order of magnitude.Considering the high sensitivity reached, we hope that one of the main problems in the detection of H. pylori in saliva -the low concentration -has been solved.The infection diagnosis in saliva is important to answer the question of whether the oral cavity is a potential reservoir of H. pylori infection.In this case, it is important to know whether the saliva does not inhibit the Le b -H.pylori binding, to which the following experiments will be devoted.
As far as the Le b -H.pylori binding reaction is concerned, it is not clear whether its affinity depends on the bacterial strain.This will be a subject of a future investigation.

Figure 4 .
Figure 4. SPR amplitude (black curves) and phasesensitive detection (red curves) of H. pylori-Le b binding reaction for two different bacterial concentrations (continuous and dashed curves).

Figure 5 Figure 5 .
Figure 5. А spectral shift as a result of H. pylori binding to Le b evaluated by: (a) amplitude detection; (b) phase-sensitive detection.