Effect of Alginate Lyase on Biofilm-Grown Helicobacter pylori Probed by Atomic Force Microscopy

Helicobacter pylori (H. pylori) is a microorganism with a pronounced capability of adaptation under environmental stress solicitations. Its persistence and antimicrobial resistance to the drugs commonly used in the anti-H. pylori therapy are associated with the development of a biofilm mainly composed of DNA, proteins, and polysaccharides. A fundamental step to increase the success of clinical treatments is the development of new strategies and molecules able to interfere with the biofilm architecture and thus able to enhance the effects of antibiotics. By using Atomic Force Microscopy and Scanning Electron Microscopy we analyzed the effects of the alginate lyase (AlgL), an enzyme able to degrade a wide class of polysaccharides, on the H. pylori shape, surface morphology, and biofilm adhesion properties. We demonstrated that AlgL generates a noticeable loss ofH. pylori coccoid form in favor of the bacillary form and reduces the H. pylori extracellular polymeric substances (EPS).


Introduction
Helicobacter pylori is one of the causative agents of gastritis and peptic ulcer diseases and is also considered as a risk factor for gastric carcinoma [1,2].Helicobacter pylori is a microorganism with a pronounced capability of adaptation under environmental stress solicitations [3].Its persistence and antimicrobial resistance to the drugs commonly used in the anti-H.pylori therapy are associated with genetic variability [4] and biofilm development [5][6][7][8].Helicobacter pylori is capable of forming a mature biofilm characterized by clustered bacterial cells arranged in an abundant matrix after 3 days of incubation in vitro [9][10][11].Although the biofilm matrix composition of H. pylori has not been fully elucidated, DNA, proteins, and lipids together with polysaccharides represent the most important components of the matrix extracellular polymeric substances (EPS) [12,13].
Cell aggregation and early adhesion to surfaces are both dependent on the properties of the bacterium surface that is modulated by the composition of the extracellular matrix.The chemical structure of the biofilm along with the composition of EPS produced by the bacterium not only is critical for microbial survival in diverse environments but may facilitate survival during transmission and promote persistence in vivo [14].H. pylori is able to modulate its shape from bacillary form to coccoid form to fit the environmental conditions and its virulence is strongly dependent on its cellular morphology.In general, it is regarded that the bacillary H. pylori transforms to the coccoid form under stressfully environmental conditions [15][16][17][18].However, characteristics of the coccoid H. pylori such as the ultrastructure, the viability, or the transformation process remain largely unexplored.Transformation into the coccoid form is an active, biologically led process, switched on by the bacterium as a protection mechanism.

International Journal of Polymer Science
Transformation plays a role in the transmission of the bacteria and is thought to be partly responsible for recrudescence or relapses of infection after antimicrobial treatments.
To increase the success of clinical treatments, together with the development of new antibiotics, it is crucial to exploit molecules able to interfere with the normal biofilm architecture and able to interfere with the H. pylori shape [9,19,20].To this purpose in this study we investigate the effect of alginate lyase (AlgL), an enzyme able to degrade some polysaccharides, on H. pylori biofilm formed by bacillary or coccoid shaped bacteria [21].By using Atomic Force Microscopy (AFM) we analyzed the effects of AlgL on the H. pylori biofilm adhesion properties and we measured the morphological alterations of bacteria in response to the clinical treatment.AFM, among high-resolution microscopy techniques used in microbiology and diagnostics [22][23][24][25], has the capability to determine the hydrophobic/hydrophilic character of the biofilm with great sensitivity [20,26,27].

Bacterial Strains and Culture
Conditions.The H. pylori strain used for the experiments was a clinical isolate from gastric biopsy specimens of a patient attending the Institute of Microbiology of Gemelli Hospital (Rome).Biopsy sample was cultured on Brucella agar plate containing 7% fetal calf serum (FCS), amphotericin at 5 mg/L, and vancomycin at 2 mg/L.Plate was incubated in microaerophilic atmosphere at 37 ∘ C for 3 to 5 days.Cultured bacteria were identified as H. pylori based on colony morphology, gram staining, and positive biochemical reactions such as urease, catalase, and oxidase.Clinical isolate, containing vacA and cagA genes as demonstrated by polymerase chain reaction, was stored at −80 ∘ C in Brucella broth supplemented with 20% glycerol and 7% FCS until experimental tests.

Biofilm Formation and Its
Quantification.For the biofilm formation, bacteria were harvested in Brucella broth supplemented with 7% FCS and broth cultures were incubated overnight at 37 ∘ C in a microaerophilic atmosphere (85% N 2 , 5% O 2 , and 10% CO 2 ) under shaking at 100 rev/min.
After incubation, broth culture was adjusted to an optical density at 600 nm (A600) of 0.2 which corresponded to the beginning of exponential phase.Sterilized 13-mm-diameter glass coverslips (Bioscience Tools, San Diego, CA) were placed into a standard 24-well cell culture plate (Thermo Scientific) and each well was filled with 1 mL of the precultured cell to allowadherence of H. pylori.The cultures were incubated under microaerophilic conditions at 37 ∘ C for 3 days without shaking.After incubation, the coverslips were removed, washed with phosphate-buffered saline (PBS), and placed again into a clean and sterile 24-well cell culture plate.The samples were then air-dried and stained with crystal violet (CV) for 30 s.After staining, the coverslips were rinsed with distilled water to remove excess dye and then air-dried at room temperature for 30 min.For biofilm quantification, all dye associated with the biofilms was dissolved with 1 mL of 95% ethanol and 200 L of the solutions was used to measure the absorbance at 560 nm with a microplate reader (Bio-Rad Laboratories, Hercules, CA) to determine the quantity of biofilm formation.

Alginate Lyase Treatment.
The AlgL from Flavobacterium species (28,000 U/g) was purchased as a pure substance from Sigma-Aldrich, and a stock solution was freshly prepared in sterile PBS.We used a concentration range from 0.015 U/mL to 30 U/mL of ALgL that had previously been shown to be sufficient to reduce biofilm formation in A. fumigatus [20,27].The enzyme powder was dissolved in Brucella broth and added to the wells with mature formed biofilm.Plate was incubated under microaerophilic conditions at 37 ∘ C for 16 h.A growth control well was incubated with only Brucella broth.

Atomic Force Microscopy.
A Nanowizard II atomic force microscope (JPK Instruments AG, Berlin, Germany) mounted on an optical microscope (Axio Observer Carl Zeiss, Oberkochen, Germany) was used for imaging each sample.The images were acquired using silicon cantilevers with high aspect-ratio conical silicon tips (CSC16 Mikro-Masch, Tallinn, Estonia) characterized by an end radius of about 10 nm and a half conical angle of 20 ∘ .All these cantilevers, with a nominal spring constant of about  = 0.01 N/m, were accurately calibrated as previously reported [28,29].All images were acquired at 512 × 512 pixels and processed by means of Gwyddion software.A large scan area (50 × 50 m) was imaged in order to determine the shape of the bacterium (coccoid or bacillary).In particular, the method used for cellular morphology quantification, bacillary or coccoid, involved a measurement of circularity directly from 10 different AFM images for each sample by using ImageJ software.The value of circularity is calculated as 4 * area/perimeter 2 .A value of 1.0 indicates a perfect circle.As the value approaches 0.0, it indicates an increasingly elongated shape.
To characterize the surface properties of H. pylori biofilm we calculated the Root Mean Square Roughness (RMS) directly from 10 different AFM images for each sample by using Gwyddion software.RMS is the root mean square of the vertical distance   from the mean line to the th data point: For each sample a histogram of RMS values was calculated by using 20 different images and calculating the RMS values in different areas.Each histogram was fitted by Gaussian functions by using OriginPro software.All experiments were performed in air at room temperature (∼25 ∘ C) with a measured average relative humidity of 75%.

Adhesion Force Spectroscopy.
Adhesion force spectroscopy maps were done by recording matrix of 25 × 25 force-distance curves on areas by using different magnifications.The force applied on the samples was set to 1.5 nN in all force spectroscopy points.All the force versus distance curves were obtained using the same force set point and keeping a constant speed of 1.0 m/s; the total vertical International Journal of Polymer Science displacement was set to 1.0 m.Force-distance curves were analysed using the data processing supplied with the JPK Nanowizard II AFM system.In brief, deflection data were recorded during the retraction of the tip from the biofilm surface and converted to force after multiplication with the spring constant of cantilever, whereas the real tip-substrate distance was obtained by subtracting the deflection value from the piezo movement.The point of zero distance was assigned at the intersection of the experimental curve with the straight line fitting the tail of the curve.Adhesion values were deduced from the force versus distance curves as the maximum value in the retraction curve [30].For each sample a histogram of adhesion values was calculated by using 5 different maps and was fitted by Gaussian functions.All experiments were performed in air at room temperature (∼25 ∘ C) and at average relative humidity of 75%.

Scanning Electron Microscopy.
For SEM imaging, samples on coverslips were processed as explained before [31].Briefly, samples were washed three times with 0.1 M sodium cacodylate buffer (pH 7.4) and then dehydrated serially in 30%, 50%, 70%, 90%, and 100% ethanol.Finally samples were fixed for 1 hour with 2% osmium tetra-oxide (OsO4).The SEM procedures were completed by drying the samples and sputtering 8 nm gold layer.Micrographs were collected with a Zeiss Supra 25 microscope (Germany).

Results
The clinical strain of H. pylori used in our study was demonstrated to be a high biofilm producer by crystal violet staining (data not shown).In the inset of Figure 1  However, the interaction of the 30 U/mL AlgL with the biofilm has a significant effect on the distribution shape inducing a decrease of its mean value from ∼0.78 down to a value of ∼0.63.This decrease clearly reflects a higher statistical occurrence of the more elongated conformations instead of the more compacted ones.
In Figures 2(a) and 2(b) two representative topological images showing bacteria enveloped in the EPS before and after the 30 U/mL AlgL treatment are reported.In Figure 2(e) several features resembling fractures (see white arrow) are clearly visible within the EPS surface.This finding demonstrates that, together with the direct effect of the AlgL on the H. pylori shape, the enzyme produces a significant alteration in the biofilm architecture forming fractures in the EPS matrix that envelope the bacteria.In Figures 2(b) and 2(c) the relative deflection images of the same two samples before and after the AlgL treatment highlight the structural EPS differences and the loss of homogeneity produced by the enzymatic activity.
To quantify the structural modification in the EPS structure we computed the biofilm surface roughness (RMS) by using (1) from ten different topological images obtained from 3 independent samples.The biofilm RMS distributions acquired before and after the AlgL treatment are shown in Figures 2(g (g and h) the roughness distribution and Gaussian fit (black lines) are shown for each sample.The distributions were obtained following the procedure described in Section 2. Distributions of the friction forces between AFM tip and three different surfaces: clean glass, biofilm, and biofilm treated with 30 U/mL AlgL (i).
after the AlgL treatment the distribution is much more spread encompassing two different peaks at ∼26 nm and ∼55 nm (see histogram in Figure 2(h)).The occurrence of two peaks reflects the presence of two morphologically different regions in the EPS matrix.The first peak at ∼26 nm rises from all those samples areas not affected by the enzymatic EPS digestion whereas the second peak at ∼55 nm rises from the areas where the EPS layer is dramatically reduced showing fractures of different sizes, from few nanometres up to several microns.In Figures 2(c) and 2(f) the relative torsional images, produced by the torsion of the AFM cantilever, obtained from the two samples before and after the AlgL treatment, are also reported.Torsional images are able to quantify the friction forces between the sample surface and the AFM silicon tip.In Figure 2(c) the friction force shows a homogeneous value all over the EPS surface whereas in Figure 2(f) the friction forces significantly change inside and outside the fractures.In Figure 2(i) three different friction force distributions are reported.The average friction force between the AFM tip and the biofilm decreases from a value of ∼1.2 nN to 0.5 nN after the AlgL treatment and several areas of the sample showed a value of about 0.2 nN, a value very similar to the friction force between the cantilever tip and the clean glass coverslip.These results clearly evidence a difference in the holes surface chemical properties due to the contact of the AFM tip with the glass coverslip under the EPS.
A direct evaluation of the biofilm chemical properties can be achieved by mapping the adhesion properties of the biofilm surface by atomic force spectroscopy.In Figures 3(a) and 3(d) two representative magnifications of a H. pylori bacterium surrounded by EPS before and after the AlgL treatment are shown and in Figures 3(b) and 3(f) the corresponding adhesion maps (25 × 25 pixels) with the bacteria perimeter highlighted in red are reported.The same colour scale has been used in both cases.By comparing the two maps, a huge increase of biofilm adhesion can be clearly detected after the AlgL treatment.This increase is mainly located in the region outside the bacterium, whereas the increase in the region occupied by the bacterium appears to be less pronounced.
A more quantitative comparison between the EPS adhesion properties before and after the AlgL treatment is shown in Figures 3(c) and 3(e), where the statistical distribution of the adhesion values is reported for both cases.Before the AlgL treatment, the biofilm adhesion displays a bell shaped distribution with adhesion values ranging between 0 nN and 21 nN and with a peak value of about ∼ 6nN (see histogram in Figure 3(c)).In this condition, the H. pylori biofilm has a hydrophobic behaviour along the whole sample even if it is slightly more hydrophobic outside the bacteria.Conversely, after the AlgL treatment a wider distribution can be detected with adhesion values up to 30 nN.A Gaussian analysis of the distribution highlights the presence of tree peaks approximately positioned at 8 nN, 18 nN, 27 nN.The wider adhesion values distribution and the occurrence of three distinct peaks highlight once again that the AlgL treatment induces deep structural modifications in the EPS matrix.In this regard, worthy of note is the arising of a third peak centred at 27 nN.This peak encompasses value ranging between 24 and 30 nN that are consistent with the measured adhesion value of the silica tip with the bare glass coverslip (∼29 nN) [28,29].This result clearly confirms that, aside from the modification of the bacterium shape, the AlgL treatment induces the formation of nano-and microfractures of the EPS matrix and therefore can be highly effective in the disruption of the bacterial biofilm.
In Figure 4(b) the effect of five different concentrations of AlgL on the H. pylori biofilm structure is reported.As the enzyme concentration increases, the cell circularity decreases and the surface of fractured EPS increases showing a clear dose-dependent enzymatic effect of the AlgL on the H. pylori biofilm.Only for the lower concentration, AlgL = 0.015 U/mL, we did not observe any difference between the treated and the untreated samples.When the AlgL has a concentration of 0.15 u/mL some biofilm fractures and several elongated cells are observable (see Figure 4(a)).At 30 U/mL almost half of the surface is clearly lysed and the H. pylori bacillary form becomes predominant (see Figure 4(c)).

Discussion
Helicobacter pylori is a bacillary shaped Gram negative bacterium, which colonizes the gastric mucosa where pH falls to ∼1.5 in the absence of food and to ∼5.0 during the digestive phase [1].H. pylori infection has an estimated prevalence of about half the world's population [32], making this pathogen one of the most successful human pathogens.Infection by this bacterium has been associated with increased risk for the development of gastritis, peptic ulcer disease, and gastric adenocarcinomas [33,34].Similar to other microorganisms, H. pylori is able to form biofilm structure which has been observed in vitro at air/liquid interfaces in media with a high carbon/nitrogen ratio [35].Even if the significance of biofilm in pathogenesis of H. pylori infections is not yet fully understood, to increase the success of clinical treatments, the development of new strategies able to interfere with the biofilm architecture and thus able to enhance the effects of antibiotics is fundamental.In this work we analyzed the effects of the alginate lyase on the H. pylori shape, surface morphology, and biofilm adhesion properties.We demonstrated that AlgL generates a noticeable loss of H. pylori coccoid form in favour of the bacillary form and reduces the H. pylori extracellular polymeric substances (EPS).H. pylori is morphologically divided into two types: a bacillary and a coccoid form [3]. The bacillary forms are viable and culturable in vitro, while the coccoid forms are viable and less culturable [36].
It is well known that the helical forms of H. pylori are transformed into coccoid forms as a result of many factors, like ageing and stress conditions such as starvation or the presence of antibiotics.Reversion from the coccoid state to the bacillary form, although less frequent, was however highlighted in some studies [37].Coccoid form of H. pylori, initially considered quiescent bacterium, retains an essential role in relapse and gastritis complications and is capable of reverting to the bacillary state [38,39].In our data, AlgL activity resulted in a response of H. pylori in terms of morphology change.A possible explanation could be that the enzyme does not lead to a stress condition for the bacteria but its effect on the EPS's polysaccharides could promote, in response to the well-known quorum sensing signalling, the bacillary form morphology able to actively produce EPS and leave the biofilm community as planktonic bacterium.
Production of biofilm by H. pylori has important effect in promoting its resistance to antimicrobials and host defense factors [40].It is generally accepted that the H. pylori dormant states of coccoid forms can survive for a long time in the gastric environment.Coccoid forms are metabolically quiescent and have been shown, in some cases, to become selectively resistant to antibiotics.These forms probably contribute to the therapeutic failure and relapses of infections.Treatment with AlgL seems to prevent the transformation to coccoid forms retaining most of the bacteria in the spiral shape, hence creating a favourable condition to future combined treatment with antibacterial drugs [19,41].
Treatments with standard antibiotics are often ineffective at eradicating biofilm-related infections.Because the role of H. pylori extracellular matrix in delaying the penetration of the drugs in the colony is still unknown; the best antibacterial drug should not be selected only on the basis of its efficiency to kill in vitro grown bacteria but also on its ability to penetrate EPS.
In this work, to study the permeability properties of an EPS reconstituted in vitro that mimics the in vivo EPS, we used AFM, which is an excellent method for the determination of the hydrophobic/hydrophilic character of the biofilm with great sensitivity [20,26,27].
We demonstrated that the presence of AlgL can greatly reduce the extent of the H. pylori biofilm forming several nano-and microfractures, which can allow the drugs transit and thus enhance their effect on the cell walls.Finally, since to devise novel and efficient pharmacological strategies based on the synergistic activity of drugs and AlgL a spatial and temporal detailed knowledge of the AlgL ability to break up and disrupt the biofilm structures is fundamental, AFM can be applied to evaluate the "in vitro" effects of a specific treatment on adhesions or other physical aspects that are crucial for the establishment and persistence of microbial biofilm.

Figure 1 :
Figure 1: Quantitative (a-b) analysis of H. pylori biofilm cell morphologies in (c-d) circularity distribution is shown for each sample.The distribution was calculated by means of ten different AFM images for each sample and following the measuring system described in Section 2. Inset images show representative AFM images (20 × 20 m) for each sample, which highlights the different shapes.
two representative AFM topological images of the H. pylori before (Figure1(a)) and after (Figure1(b)) the treatment with 30 U/mL of AlgL are reported.It appears that the AlgL treatment deeply modifies the bacterium shape inducing a noticeable loss of the coccoid rounded form in favour of the more elongated bacillary form.This result is further confirmed by the circularity distributions obtained before (Figure1(a)) and after (Figure1(b)) the AlgL treatment.Distributions were calculated as described in Materials and Methods by means of ten different AFM images for each sample.The two distributions appear to be very spread out since H. pylori assumes a large ensemble of coexisting conformations in both cases.

Figure 2 :
Figure 2: Representative zoom AFM images (10 × 10 m) of H. pylori biofilm are shown before treatment (a-c) and after treatment with AlgL (e-g).Height images (a and d) show different cells from each experimental condition and the presence of fractures in biofilm generated by AlgL treatment (d).Deflection and lateral images (b, e, c and f) reveal more details in biofilm texture.In Figures(g and h) the roughness distribution and Gaussian fit (black lines) are shown for each sample.The distributions were obtained following the procedure described in Section 2. Distributions of the friction forces between AFM tip and three different surfaces: clean glass, biofilm, and biofilm treated with 30 U/mL AlgL (i).

Figure 3 :
Figure 3: Adhesion force maps of H. pylori biofilm: (a) representative high-resolution image (2.3 × 2.3 m) of H. pylori before treatment (a) and after treatment with AlgL (d).Representative adhesion force maps (25 × 25 pixels) with shadows of bacteria (red lines) were shown for H. pylori before treatment (b) and after treatment with AlgL (e).Histograms and Gaussian fit (black lines) obtained from five different maps for H. pylori before treatment (c) and after treatment with AlgL (f).

Figure 4 :
Figure 4: H. pylori circularity and percentage of surface covered by the biofilm for 5 different concentrations of AlgL (0.015, 0.15, 7.5, 15, and 39 U/mL) (b).Representative SEM images of H. pylori biofilm after the treatment with 0.15 U/mL of AlgL (a) and with 30 U/mL of AlgL.(c) Red arrows show the lysed EPS produced by the H. pylori.Scale bar 400 nm.