Sensors & Actuators: B. Chemical 403 (2024) 135211 Contents lists available at ScienceDirect Sensors and Actuators: B. Chemical journal homepage: www.elsevier.com/locate/snb Mycobacterium detection method combining filtration, immunomagnetic separation, and electrochemical readout in a portable biosensing device Melania Mesas Gómez a, b, Bárbara Molina-Moya f, Bárbara Camila de Araújo d, Arnau Pallarès-Rusiñol c, Jofre Ferrer-Dalmau c, Maria Valnice Boldrín Zanoni d, José Domínguez e, f, g, h, Esther Julian e, María Isabel Pividori a, b, *, 1 a Grup de Sensors i Biosensors, Departament de Química, Universitat Autònoma de Barcelona, Bellaterra, Spain b Biosensing and Bioanalysis Group, Institute of Biotechnology and Biomedicine, Universitat Autònoma de Barcelona, Bellaterra, Spain c BioEclosion SL, Avda. Can Domènech, Edifici Eureka, Universitat Autònoma de Barcelona, Bellaterra, Spain d Department of Analytical Chemistry, Institute of Chemistry, UNESP, Universidad Estadual Paulista, Araraquara, SP, Brazil e Departament de Genètica i de Microbiologia, Facultat de Biociències, Universitat Autònoma de Barcelona, Bellaterra, Spain f Institut d′Investigació Germans Trias i Pujol (IGTP), Badalona 08916, Spain g Departament de Genètica i Microbiologia, Universitat Autònoma de Barcelona, Bellaterra, Spain h CIBER Enfermedades Respiratorias, Instituto de Salud Carlos III, Universitat Autònoma de Barcelona, Bellaterra, Spain A R T I C L E I N F O A B S T R A C T Keywords: This study addresses a straightforward and highly sensitive approach for detecting Mycobacterium fortuitum. The Preconcentration strategy method involves a combination of filtration and direct immunomagnetic separation to isolate the bacteria Immunomagnetic separation retained on filters for further electrochemical magneto immunosensing in a handheld device. Unlike conven­ Mycobacterium fortuitum tional methods involving pre-enrichment by culturing, this approach employs a simplified preconcentration PEM Hemodialysis water technique, which includes filtration of large water samples, up to 100 mL, followed by a one-step process of Handheld device immunomagnetic separation and labeling. After a 60-min, during which the filter with the retained bacteria and all the reagents (including modified magnetic particles and enzymatic conjugates) are incubated, the resulting product is directly drop into a cartridge, capable of performing magnetic actuation and washing. The electro­ chemical readout is carried out on a portable battery-operated device within 30 seconds. Remarkably, the immunosensor demonstrates an outstanding limit of detection of 5 CFU mL− 1 in hemodialysis water processing 100 mL of sample. This achievement is remarkable considering the short and simplified analytical procedure, compared to the traditional isolation and culturing of mycobacteria, which typically takes 2 weeks. 1. Introduction Mycobacterium mucogenicum and Mycobacterium fortuitum can eventually progress to disseminated bacteremia, meningitis, or other syndromes. The genus Mycobacterium contains more than 170 species of non- Accordingly, [7]. Rapidly growing PEM is also responsible for health tuberculous mycobacteria (NTM) that are commonly referred to as care-related outbreaks by contamination of potable water. An example is environmental mycobacteria [1,2]. Unlike Mycobacterium tuberculosis infections in hemodialysis patients, highlighting the need to identify complex are the major responsible of infections, opportunistic NTM in­ reservoirs and minimize proliferation to prevent outbreaks [8,11]. fections are increasingly prevalent worldwide [3,4], particularly among Traditional detection of PEM in water relies on culturing [7]. Rapidly individuals with compromised immune systems or underlying lung growing PEM takes 7 days to produce observable colonies, whereas diseases. Pathogenic environmental mycobacteria (PEM) frequently slowly growing PEM need several additional days. Before culturing, cause infections, often transmitted through water [5–8], via ingestion water samples are decontaminated with alkali or acid solutions to and inhalation, particularly from aerosols [7,9,10]. Rapidly growing eliminate other bacteria and fungi that could mask the PEM colonies, PEM, including Mycobacterium abscessus, Mycobacterium chelonae, which can also have an impact on mycobacterial viability. Traditional * Corresponding author at: Grup de Sensors i Biosensors, Departament de Química, Universitat Autònoma de Barcelona, Bellaterra, Spain. E-mail address: Isabel.Pividori@uab.cat (M.I. Pividori). 1 Orcid ID: 0000-0002-5266-7873 https://doi.org/10.1016/j.snb.2023.135211 Received 14 August 2023; Received in revised form 14 November 2023; Accepted 21 December 2023 Available online 23 December 2023 0925-4005/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 culture-based PEM detection is time-consuming, labor-intensive, and handheld device operated by batteries (BioEclosion, Spain) which in­ potentially inefficient due to loss or underestimation of PEM. Nowadays, cludes the external magnet, for the quantitative amperometric readout molecular techniques play an important role for detecting mycobacterial directly on the display in 30 s, or in a bipotentiostat DRP-STAT200 species in water samples. These methods involve identifying distinct operated by DropView 2.2 for instrument control and data acquisition mycobacterial DNA sequences, often employing culture enrichment or (Dropsens, Spain). A complete filtration system of 25 mm (Product no. direct DNA isolation, followed by PCR amplification. Given the typically 073–0Q7724, Scharlab) was used to assess polycarbonate track-etched low concentration of mycobacterial DNA in environmental samples, PCR membranes (Catalogue no. 10417106, Whatman), nylon (Catalogue amplification is commonly employed [7,12,13]. Predominant reports on no. 7404–004, Whatman), cellulose acetate (Catalogue no. 10404006, Mycobacterium detection are mostly focused on M. tuberculosis complex. Whatman), cellulose nitrate (Catalogue no. 10401106, Whatman) and While reports have increased due to the emergence of NTM infections, mixed cellulose ester (Ref. HAWP02500, Merck Millipore Ltd), in all research on PEM detection remains limited. Given PEM’s significant role instances 0.45 µm pore size and 25 mm diameter. The cyclopore poly­ in environmental systems and potential health risks, there is a need for a carbonate track-etched membranes of 0.2 µm pore size were used as a straightforward, rapid, and user-friendly PEM monitoring approach, SEM support (Catalogue no. 10417606, Whatman). The SEM images especially in low-resource settings. In this work, an electrochemical were taken with the scanning electron microscope EVO MA-10 (with magneto-immunosensor is introduced for quantifying M. fortuitum, EDS Detector, Oxford LINCA). The confocal images were collected on serving as a PEM model in this study. According to the low concentration the microscope Leica, TCS SP5 (Leica Microsystems, Germany). of this bacteria in water, the collection of 100 up to 1000 mL of sample is typical for further filtration or centrifugation, as a preconcentration step 2.2. Chemicals and biochemicals [7]. A preconcentration strategy relying on the combination of filtration and immunomagnetic separation (IMS) is used in this procedure to pull Different set of buffers were used for specific procedures in the ex­ out the bacteria directly from the filter. Moreover, an in order to simplify periments, and their composition is described in S1 (Supp. Data). All the analytical procedure, the filter is incubated at the same time with the buffers were prepared from analytical grade purchased from Merck and magnetic particles modified with anti-lipoarabinomannan antibodies Sigma and using milliQ water. The monoclonal anti-LAM antibody from and the enzymatic conjugate, performing a one-step incubation in only rabbit (Catalogue no. MA533311, ThermoFisher Scientific) was immo­ 60 min. Furthermore, the electrochemical detection takes place in an bilized on tosyl activated magnetic particles (Dynabeads™ M-450 easy-to-use and handheld device with a novel cartridge, in which the Tosylactivated, Catalogue no. 14013, Invitrogen). The anti- magnetic actuation is performed, while the excess of sample and re­ Mycobacterium polyclonal antibody (anti-TBS) from rabbit (Catalogue agents are removed. no. PAB29798, Abnova) was biotinylated with EZ-Link NHS-Biotin Kit (Catalogue no. 20217, ThermoFisher Scientific) following the manu­ 2. Experimental section facturer’s protocol, as detailed in S2 (Supp. Data), providing a bio­ tinylation degree of 6 biotin per IgG molecules (anti-TBS-BIO). The 2.1. Instrumentation excess of biotin tag was removed by desalting using the Zeba™ spin and desalting columns of 7 K cut-off (Catalogue no. 89882, Pierce Rockford, The electrochemical measurements were performed on a disposable IL). An enzymatic conjugate of anti-TBS-BIO and streptavidin-HRP cartridge (Fig. 1, panel C) for magneto-actuation, washing and reading (Catalogue no. 11089153001, Roche) was used for labelling, and pre­ (BioEclosion SL, Spain). The cartridge can be either inserted to a viously prepared by incubation for 45 min at 37ºC. Fig. 1. Schematic representation of A) filtration, B) immunomagnetic separation of the bacteria retained in the filters and labelling in one-step, followed by C) electrochemical readout for the detection of M. fortuitum in hemodialysis water performed in a cartridge and handheld device. In Panel B, experimental details are presented, showing how the filter is positioned on the Eppendorf tube for immunomagnetic separation. The photos in Panel B were captured in 1-second frame sequences. More details are presented in S8 (Sup. Materials). 2 M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 2.3. Culturing Mycobacterium fortuitum and other strains Briefly, 100 µL of sample containing M. fortuitum was incubated with anti-LAM-MPs at 107 MP mL− 1 and enzymatic conjugate (40 µg mL− 1 The strain of Mycobacterium fortuitum ATCC 6841 was growth in anti-TBS-BIO and 5 U mL− 1 strep-HRP), in a volume of 1 mL in with Tryptic Soy Broth (TSB) medium supplemented with 0.015 % (v/v) of PBST for 60 min under gentle rotation, as optimized in S4, S5 and S6 tyloxapol at 37 ºC for 72 h in continuous agitation, until reaching an OD (Supp. Data). The product is introduced in the cartridge in the “OPEN of 1.000 AU at 550 nm. The liquid culture was aliquoted and preserved position”, in which the magnetic actuation is performed, while the supplemented with 20 % (v/v) of glycerol at − 40 ºC. The concentration excess of sample and reagents are removed. 3 drops (150 µL) of the of the stock was calculated by CFU counting in Tryptic Soy Agar (TSA) cartridge solution are added. The cartridge is then put in “CLOSE posi­ plates, after the incubation for 5–7 days at 37 ºC. The concentration of tion”, and 3 drops (150 µL) added again for the further amperometric viable cells was quantified for each experiment by CFU counting, as measurement under enzyme saturation conditions at − 0.15 V and 30 s. shown in Fig. S1 (Supp data). The strains for the specificity study were: All the cartridge operation as well as the electrochemical readout were E. coli (CECT 405, ATCC 10536), L. Pneumophila (serogroup 1, Philadel­ performed following the product specifications provided by the manu­ phia 1, ATCC 33152), P. Aeruginosa (CECT116, ATCC 15542), and facturer. More details about the amperometric measurements are pro­ M. fortuitum. The concentrations of the viable bacteria were determined vided in S7 (Supp. Data). The LOD of the electrochemical magneto in all the experiments by the gold standard culturing method in solid immunosensor was calculated by processing a calibration plot ranging media and CFU counting, as detailed in Supp data. from 0 to 2.2 × 106 CFU mL− 1. 2.4. Covalent immobilization of anti-LAM antibodies 2.7. Novel preconcentration method. Filtration, immunomagnetic separation and electrochemical immunosensing in handheld device. Study The tailored modification of the MPs with anti-LAM antibodies was of the filtering material performed on tosyl activated magnetic particles as reported previously [11], following by blocking the unreacted tosyl groups with glycine A preconcentration strategy is integrated to the electrochemical buffer. The outline of the procedure is schematically represented in immunosensing to improve de LOD, by filtering high volume of samples Fig. S3, panel A (Supp. data). The coupling efficiency was determined as and further immunomagnetic separation to pull-out the bacteria from 98.3 %, as detailed in Fig. S3, panel B (Supp. Data). The final concen­ the filter for downstream analysis. Briefly, the procedure combines three tration of the anti-LAM-MPs was 1 x 108 MP mL− 1 and were previously steps as described in Fig. 1 and Fig. S8 (Supp. data): i) Filtration of large washed three times with washing buffer before used in each experiment. volumes (typically 100 mL) of sample followed by ii) the immuno­ magnetic separation of M. fortuitum retained in the filters with the anti- 2.5. Evaluation of the immunomagnetic separation of M. fortuitum by LAM-MPs and incubation with the enzymatic conjugate anti-TBS-BIO/ scanning electron and confocal microscopy. Specificity study strep-HRP in a one-step incubation and ii) the electrochemical readout in 30 s. The experimental details are provided in S8 (Supp. data) and the To assess the performance anti-LAM-MPs for binding M. fortuitum, figure therein. After filtration from 100 mL sample under vacuum, the scanning electron microscopy was performed. A volume of 100 µL of 25 mm diameter filter was placed on a 2.0 mL tube. 100 µL of anti-LAM- anti-LAM-MPs at 107 MP mL− 1 was incubated with 500 µL of MPs at 107 MP mL− 1, 200 µL of the enzymatic conjugate anti-TBS-BIO/ M. fortuitum at 2.2 × 108 CFU mL− 1 for 1 h at 750 rpm. Afterwards, the streptavidin-HRP were added and PBST buffer to complete a final vol­ sample was filtered in a polycarbonate membrane. Then, the filter was ume of 1 mL, to fully cover the filter and to ensure complete interaction treated as previously described [14]. In order to visualize the interaction of the MP with the bacteria retained in the filters. All reagents were used in a sandwich format of M. fortuitum with both anti-LAM-MPs and in the optimized concentrations (as described in S6, Supp. data), and anti-TBS-BIO antibody, confocal fluorescence microscopy was per­ incubated under gentle rotation at RT for 60 min. Then, the MPs labelled formed. To provide blue fluorescence to the bacteria, M. fortuitum complexes with the captured bacteria were recovered under magnetic 2.2 × 108 CFU mL− 1 was incubated with 5 µL of Hoechst for 30 min at actuation and resuspended in 200 µL of washing buffer for further 750 rpm. To provide green fluorescence to the MPs, 100 µL of transfer to the cartridge. Then, the electrochemical readout was per­ anti-LAM-MPs at 107 MP mL− 1 was preincubated with an formed as detailed described in S7 (Supp. data). anti-rabbit-ALEXA488 for 30 min. The anti-LAM-MPs were then incu­ In order to validate the approach, different materials were assessed. bated with 100 µL of M. fortuitum for 1 h at 750 rpm, and then, with Polycarbonate (PC), nylon (NY), cellulose acetate (CA), cellulose nitrate 200 µL conjugate streptavidin-Cy5/anti-TBS-BIO (80 µg mL− 1) for 1 h. (CN), and mixed cellulose ester membranes (MCE) were selected Confocal images were thereafter collected with the laser line of Hoechst because of their widely use in microbiological filtration techniques with (350–461 nm excitation, 461 nm emission), anti-rabbit Alexa Fluor® a 0.45 µm porous size and 25 mm diameter. For this study, 488 (anti-rabbit-ALEXA488) (495 nm excitation, 518 nm emission) and 100 mL M. fortuitum samples at the same concentration (2.4 ×103 CFU streptavidin-Cy5 (600–650 nm excitation, 670 nm emission). mL− 1) as well as the negative control were filtered using each membrane The specificity study was performed by magneto-actuated immuno­ and submitted to downstream analysis, in this instance magneto- assay as detailed described in S4 (Supp Data) at the optimized concen­ actuated immunoassays described in S4 and S5 (Supp. Data). tration of reagents, as shown in Fig. S4 (Supp Data). In this study, bacteria commonly found in contaminated water were included at a 2.8. Electrochemical magneto immunosensing for the quantification of high concentration range: E. coli 3.6 × 108 CFU mL− 1, L. Pneumophila M. fortuitum in hemodialisys water samples 7.2 × 107 CFU mL− 1, P. Aeruginosa 4.1 × 108 CFU mL− 1, and M. fortuitum 2.1 × 105 CFU mL− 1. The bacteria grew in solid plates were M. fortuitum is one of the most common mycobacteria in contami­ inoculated in water, adjusting the absorbance to 0.200 AU at 550 nm. nated hemodialysis waters [15] which can put immunocompromised patients at risk. Hence, it is important to continuously monitor water 2.6. Electrochemical magneto immunosensing for the quantification of supplies in hemodialysis centers, to minimize the exposure of renal M. fortuitum transplantation and immunocompromised individuals to contaminated sources. In fact, a magneto-actuated immunoassay for the detection of The electrochemical magneto immunosensing procedure involves i) M. fortuitum in hemodialysis water samples was reported previously by the immunomagnetic separation of M. fortuitum with the anti-LAM-MPs our group [11]. Therefore, a calibration curve with different concen­ and incubation with the enzymatic conjugate anti-TBS-BIO/strep-HRP trations of the bacteria in hemodialysis water samples was performed in a one-step incubation and ii) the electrochemical readout in 30 s. and the LOD was calculated. 3 M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 2.9. Biosafety considerations they are known to form clumps associated with biofilm formation in the natural environment [16]. In order to visualize not only the ability of the All experiments were performed in a Biosafety Class 2 environment anti-LAM-MPs in capturing mycobacteria, but also anti-TBS-BIO to required for the handling of M. fortuitum, P. aeruginosa and detect them in a sandwich-like format, confocal fluorescence micro­ L. pneumophila. All biological waste generated from the experiments scopy was performed using streptavidin-Cy5 for labelling (emitting were disposed in accordance with the local regulations for handling far-red-fluorescence at 670 nm). According to Fig. 2, panel C, biohazards. anti-LAM-MPs (incubated with ant-rabbit-ALEXA488 antibodies emit­ ting green at 518 nm) are capable of capturing aggregates of mycobac­ terium which appears in blue due to the Hoechst staining of the DNA at 2.10. Statistical analysis 461 nm recovered by red due to anti-TBS-BIO/streptavidin-Cy5 conju­ gate. The monoclonal anti-LAM antibody was selected for capturing in The statistical analyses were performed using Prism v 10.0.1 the MPs considering that the lipoarabinomannan is a lipoglycan and (GraphPad, San Diego, USA). The calibration curves were adjusted to a major virulence factor in the bacteria genus Mycobacterium. It is thus non-linear regression fitting, using the four-parameter logistic curve, expected that other bacteria will not be detected, as demonstrated in abbreviated 4PL. The limit of detection was calculated considering the Fig. 2, panel D, when challenged with a high concentration of number of negative controls for each curve, using the one-tailed test gram-negative bacteria. The signal of the negative control containing with a significance level of p > 0.05 (t Student, p > 0.05). 0 CFU mL-1 of M. fortuitum was found to be comparable with E. coli, L. Pneumophila, P. Aeruginosa (p < 0.05), suggesting a good specificity for 3. Results and discussion the assay. 3.1. Evaluation of the immunomagnetic separation of M. fortuitum by scanning electron and confocal microscopy. Specificity study 3.2. Electrochemical magneto immunosensing for the quantification of Mycobacterium fortuitum Following the incubation of the Mycobacterium -containing sample with anti-LAM-MPs, the sample was observed by scanning electron mi­ The primary objective of this study is to achieve analytical simpli­ croscopy (SEM) to confirm the effectiveness of the immunomagnetic fication, aimed at minimizing user intervention while maintaining separation. In Fig. 2, it is confirmed that tosyl-activated magnetic par­ analytical performance. The optimized assay conditions were thor­ ticles were coated with anti-LAM monoclonal antibodies to enhance the oughly tested in Sections S5 and S6 (Supp. Data). This work is also interaction between the particles and the bacteria. Fig. 2, panel A clearly focused on detecting M. fortuitum through an amperometric readout demonstrate the successful binding of the modified magnetic particles to using a user-friendly handheld electrochemical device with an inte­ the M. fortuitum bacteria, resulting in the formation of clusters due to the grated internal potentiostat. The disposable cartridge encompasses both polyvalency of both the bacteria and the magnetic particles. Large the microfluidic system and the electrode, enabling magnetic actuation corded bacterial aggregates are also observed in Fig. 2, panel B, since and efficient removal of excess samples and reagents. Notably, the same Fig. 2. Panel A and B. Evaluation of the IMS by SEM at a M. fortuitum concentration of 2.2 × 108 CFU mL-1. In all cases, identical acceleration voltage (15 KV) was used. Panel C. Confocal microscopy study of the sandwich-like format. Anti-LAM-MPs are shown in green, capturing a cluster of M. fortuitum that was stained in blue by Hoechst, and superimposed with the specific red fluorescence from anti-TBS-BIO/ streptavidin-Cy5 conjugates. Panel D. Specificity study for M. fortuitum 2.1 × 105 CFU mL− 1, E. coli at 3.6 × 108 CFU mL− 1, P. Aeruginosa at 4.1 × 108 CFU mL− 1 and L. pneumophila 7.2 × 107 CFU mL− 1. The error bars show the standard deviation for n = 3. 4 M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 solution serves a dual role in both washing and catalyzing the reaction optimal recovery rates for the bacteria. through the HRP enzyme (referred to as the cartridge solution). As a The calibration plot integrating the novel preconcentration method result, the necessity for employing diverse buffers is eliminate, leading in a one-step incubation (by filtering 100 mL of hemodialysis water), to reduced time in a simpler approach for the end-user. Additionally, the and further electrochemical immunosensor is shown in Fig. 4, panel B as cartridge design allows for the use of a small solution volume, up to green line, for the detection of the M. fortuitum from 0 to 8.7 x 104 CFU 300 μL. Fig. 3, Panel A shows the calibration plot from 0 to 2.2×106 CFU mL− 1. The data was fitted with a non-linear regression (Sigmoidal 4PL, mL− 1 for the detection of the M. fortuitum with the electrochemical GraphPad Prism Software v 10.0.1, R2 = 0.9906) and the LOD was immunosensor without the integration of the novel preconcentration calculated (n = 9), resulting in a value of as low as 5 CFU mL− 1. method. The data was fitted with a non-linear regression (Sigmoidal Compared with the electrochemical biosensor without the integration of 4PL, GraphPad Prism Software v 10.0.1, R2 = 0.9987) and the limit of the preconcentration method (Fig. 4, blue solid line), an impressive 103 detection were calculated, resulting in a value of 4,3×103 CFU mL− 1, by fold improvement was achieved, from 4.8×103 CFU mL− 1 to 5 CFU processing 0.1 mL of sample. This LOD is comparable with a magneto- mL− 1. Moreover, the LODs for the magneto-actuated electrochemical actuated immunosensor previously reported by our research group but immunosensor in water (Fig. 3) and hemodialysis water was found to be for Salmonella, in which a LOD of 7.5 × 103 CFU mL− 1 in milk was the same, 4.8×103 (n = 8). achieved in 50 min without any pretreatment. [17]. In order to further This approach allows the detection of low concentrations in high- improve the LODs, a new approach to handle higher volume of sample, volume samples that is one of the bottlenecks in the detection of envi­ combining filtration and IMS, was presented in the next section. ronmental bacteria, in around 60 min (including filtration, which takes only 2 min). A similar LOD (13 CFU mL-1) was previously reported by our research group in a magneto-ELISA, but in 3 h [11]. However, the 3.3. Filtration, magnetic actuation and electrochemical immunosensing. new method offers several key advantages over to time-consuming and Study of the filtering material labor-intensive ELISAs. It streamlines the process by integrating immunomagnetic separation and labeling into a single step, significantly The LOD achieved by the electrochemical magneto immunosensor reducing the analysis time and minimizing washing steps. User inter­ (4,3×103 CFU mL− 1) by processing 0.1 mL of sample are above the vention as pipetting is minimized through a user-friendly cartridge limits required by most legislations. In order to improve the LODs, a new design, enhancing accessibility and reproducibility. The incorporation approach to handle higher volume of sample (up to 100 mL), combining of handheld electrochemical readout makes on-site testing possible, filtration and direct immunomagnetic separation (IMS) of the bacteria offering portability and immediacy, crucial for point-of-care and retained in the filters, is proposed in this work, as depicted in Fig. 1. field-based applications. The readout time is further reduced through Selecting the appropriate filter material is crucial to ensure a swift several steps: the direct introduction of the modified-MPs particle con­ filtration workflow, a good removing of the bacteria retained in the jugate into the cartridge, the subsequent washing, the addition of filters by the MPS, while minimizing nonspecific adsorption of all the enzymatic substrates (using the same solution) within the cartridge, and reagents (including the enzymatic conjugate) during the one-step in­ the subsequent electrochemical measurement, accomplished in less than cubation on the filters. 30 seconds. This represents a 10-fold reduction in time in comparison A collection of filter materials frequently employed in microbiology, with an immunoassay with optical readout, where the substrates require such as polycarbonate, nylon, cellulose acetate, cellulose nitrate, and a minimum incubation time between 15 and 30 min [11]. The table S9 mixed cellulose esters, was chosen. These materials share a hydrophilic (Supp data) provides a more comprehensive overview of the proposed surface to prevent excessive retention and thereby facilitate immuno­ method in relation to other methods including standard culturing and magnetic separation and further labelling. The results obtained from PCR, enhancing the main features. The preconcentration strategy not filtering 100 mL water containing M. fortuitum and negative controls only eliminates preenrichment and DNA extraction or amplification were compared for each material, as illustrated in Fig. 4, panel A. While techniques but also enables detection of intact mycobacteria in a the negative controls yielded similar results across all materials, nylon streamline process. In earlier studies, we explored several strategies to emerged as the choice due to its significant difference in signal improve LOD, such as incorporating a short preerichment step [17] or compared to other materials. In this study, nylon filters exhibited the amplifying the genetic material after the IMS [14]. In the current study, highest signal intensity, thereby promoting interaction among different we have achieved a remarkable 103-fold improvement in the LOD, biorecognition elements during the one-step incubation and yielding Fig. 3. Panel A. Calibration plot for the for the detection of M. fortuitum 0 to 2.2×106 CFU mL− 1 with the magneto-actuated electrochemical immunosensor in water samples, without the integration of the novel preconcentration method (R2 = 0.9987). (n = 3). Panel B. Raw data obtained from the amperometric measurements, performed at applied potential = − 0.150 V (vs. Ag/AgCl). The current value at the steady current (30 s) was used for the calibration plot. 5 M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 Fig. 4. Panel A. Study of the filtering material by the filtration of 100 mL samples containing M. fortuitum (2.4 x 102 CFU mL− 1) as well as the negative controls (0 CFU mL− 1). All the filters were 0.45 µm pore size and 25 mm of diameter. The filtering materials were polycarbonate (PC), mixed cellulose ester (MCE), cellulose nitrate (CN), nylon (NY) and cellulose acetate (CA). Panel B) Calibration plot for the magneto-actuated electrochemical immunosensor in hemodialysis samples. The blue solid line shows the calibration plot ranging from 0 to 8.7×104 CFU mL− 1 without the integration of the novel preconcentration method (R2 = 0.9967). The green solid line shows the calibration plot ranging from 0 to 8.7×104 CFU mL− 1 (R2 = 0.9906) with preconcentration by filtering 100 mL hemodialysis samples with nylon membranes. In all cases, a one one-step incubation of the reagents and the electrochemical readout in the cartridge with the handheld device was performed. The error bars show the standard deviation for n = 3. reducing it from 4.8×103 CFU mL− 1 to 5 CFU mL− 1. The Advancement electrochemical biosensing applications. Medical Instrumentation (AAMI) [18], the European Pharmacopeia (EP) (ISO13959, 2014) or the Japanese Society for Dialysis Therapy (JSDT) CRediT authorship contribution statement [19] establish as maximum level of heterotrophic bacteria 100 CFU mL-1 of total viable counts (TVC) in conventional hemodialysis water. Melania Mesas Gómez: Data curation, Investigation, Validation, On the other hand, the Centers for Medicare and Medicaid Services Formal analysis, Writing – original draft. Bárbara Molina-Moya: Inves­ (CMS), Department of Health and Human Services set the regulations in tigation, Validation. Bárbara. Camila de Araújo: Investigation, Valida­ 200 CFU mL-1, while urgent action should be taken at 50 CFU/mL [20]. tion. Arnau Pallarès-Rusiñol: validation. Jofre Ferrer-Dalmau: These values underscore the effectiveness of the method as a monitoring validation. Maria Valnice Boldrin Zanoni: Supervision. José. Domí­ tool for conventional dialysis water and dialysis fluids. Notably, the nguez: Supervision. Esther Julian: Conceptualization, Methodology, calculated limit of detection is as low as 5 CFU mL− 1, and a remarkable Supervision. María Isabel Pividori: Conceptualization, Methodology, signal-to-background ratio of 6.8 is achieved for 86 CFU mL− 1. This is Supervision, Funding acquisition. achieved using a portable device powered by batteries in only 1 h and requiring minimal intervention from the end user. Declaration of competing interest 4. Conclusions The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence The presence of non-tuberculous mycobacteria in water distribution the work reported in this paper. systems constitutes a critical global health concern, owing to its link with an increased incidence of NTM-associated pulmonary diseases and Data availability related complications. Instances of these bacteria being detected in drinking water and hemodialysis systems were reported several times The datasets generated during this study are available in the CORA [5,21–24]. The innovative aspects of the strategy presented here for RDR repository https://dataverse.csuc.cat. M. fortuitum integrates key components aiming to achieve simplification and portability while maintaining outstanding analytical performance. Acknowledgments In conclusion, the method offers a preconcentration strategy that com­ bines magnetic particles and filtration, facilitating efficient capture and This research was funded by the Ministry of Science, Innovation and concentration of bacteria via immunomagnetic separation. In the stan­ Universities, Spain (Projects PID2019-106625RB-I00/AEI/10.13039/ dard microbiological method, after sample filtration, the retained bac­ 501100011033 and PID2022-136453OB-I00) and from Generalitat de teria are resuspended or extracted from the filter using special buffers, Catalunya (2017-SGR-220, 2021SGR-00124, 2021SGR-00092, vortexing, centrifugation or ultrasonication. In this study, the pre­ SGR2021-1267). Also, Ministry of Universities (Grant FPI/BES-2017- concentration strategy is integrated to the electrochemical biosensing. 080748) are gratefully acknowledged. The SEM and confocal images By filtering high volumes of samples and subsequently employing were taken in the Servei de Microscòpia of Universitat Autònoma de magnetic actuation to pull out the bacteria directly from the filter, the Barcelona, Spain. The hemodialysis water was kindly provided by need for resuspension or additional steps for downstream analysis was Hospital Germans Trias i Pujol (Can Ruti, Badalona, Spain). eliminated. Other further source of analytical simplification relies on the integration of immunomagnetic separation and labeling into a single Appendix A. Supporting information step in only 60 min, significantly reducing analysis time in contrast to the conventional two-week timeframe needed for traditional isolation Supplementary data associated with this article can be found in the and culturing methods. The user-friendly cartridge design minimizes online version at doi:10.1016/j.snb.2023.135211. user intervention, enhancing ease of use and reproducibility. Moreover, the handheld electrochemical readout powered by batteries enables on-site and point-of-care testing, marking a significant advance in 6 M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 References water distribution system, Rev. Inst. Med. Trop. Sao Paulo. 51 (2009) 37–43, https://doi.org/10.1590/S0036-46652009000100007. [23] R. Briancesco, M. Semproni, S. Della Libera, M. Sdanganelli, L. Bonadonna, Non- [1] T. Fedrizzi, C.J. Meehan, A. Grottola, E. Giacobazzi, G. Fregni Serpini, tuberculous mycobacteria and microbial populations in drinking water distribution S. Tagliazucchi, A. Fabio, C. Bettua, R. Bertorelli, V. De Sanctis, F. Rumpianesi, systems, Ann. Ist. Super. Sanita 46 (2010) 254–258, https://doi.org/10.4415/ M. Pecorari, O. Jousson, E. Tortoli, N. Segata, Genomic characterization of ANN_10_03_05. Nontuberculous Mycobacteria, Sci. Rep. 7 (2017), https://doi.org/10.1038/ [24] M.R. Maleki, H.S. Kafil, N. Harzandi, S.R. Moaddab, Identification of srep45258. nontuberculous mycobacteria isolated from hospital water by sequence analysis of [2] Y. Matsumoto, T. Kinjo, D. Motooka, D. Nabeya, N. Jung, K. Uechi, T. Horii, T. Iida, the hsp65 and 16S rRNA genes, J. Water Health 15 (2017) 766–774, https://doi. J. Fujita, S. Nakamura, Comprehensive subspecies identification of 175 org/10.2166/wh.2017.046. nontuberculous mycobacteria species based on 7547 genomic profiles, Emerg. Microbes Infect. 8 (2019) 1043–1053, https://doi.org/10.1080/ 22221751.2019.1637702. Melania Mesas Gómez: She is a phD student in the Autonomous University of Barcelona [3] D.R. Prevots, T.K. Marras, Epidemiology of human pulmonary infection with (UAB, Spain). She finished the university degree in Chemistry in 2016, and the MSc nontuberculous mycobacteria a review, Clin. Chest Med. 36 (2015) 13–34, https:// in“Insdustrial Chemistry and Introduction to Chemical Research” in 2017 at UAB.During doi.org/10.1016/j.ccm.2014.10.002. their degree and master’s thesis she started to work in the development of immunoassays [4] J.R. Honda, R. Virdi, E.D. Chan, Global environmental nontuberculous and electrochemical biosensors for the detection of breast cancer cells using magnetic mycobacteria and their contemporaneous man-made and natural niches, Front. particles in the group of Sensors and Biosensors (UAB, Spain). She has had the opportunity Microbiol. 9 (2018), https://doi.org/10.3389/fmicb.2018.02029. to collaborate in two European projects, concerning the development of a biosensor for [5] M.J.M. Vaerewijck, G. Huys, J.C. Palomino, J. Swings, F. Portaels, Mycobacteria in tuberculosis (INNOVA4TB) in Sweeden and novel biomarkers for the stratification of fever drinking water distribution systems: ecology and significance for human health, in children in low-resource settings (EChiLiBRiST). During her doctoral dissertation, she FEMS Microbiol. Rev. 29 (2005) 911–934, https://doi.org/10.1016/J. explores new methods to enhance rapid diagnostic tests for the detection and quantifi­ FEMSRE.2005.02.001. cation of pathogenic bacteria in environmental and clinical samples and the design of [6] B. Adrados, E. Julián, F. Codony, E. Torrents, M. Luquin, J. Morató, Prevalence and biosensors for these targets. concentration of non-tuberculous mycobacteria in cooling towers by means of quantitative PCR: a prospective study, Curr. Microbiol. 62 (2011) 313–319, Dr. Bárbara Molina Moya: She graduated in Biotechnology at the Universitat Autònoma https://doi.org/10.1007/S00284-010-9706-2. de Barcelona (UAB) in 2011 and did the MSc in Developmental Biology and Genetics at the [7] S. Pedley, Pathogenic mycobacteria in water: a guide to public health Universitat de Barcelona (UB) in 2011–2012. In 2012, she started to be part of Dr. Dom­ consequences, monitoring and management, World Health Organ. (2013), https:// inguez’s research group at the Institut Germans Trias y Pujol, where she did the PhD in doi.org/10.2166/9781780405834. Microbiology at the UAB (2016). The phD consisted in the development and optimization [8] J.O. Falkinham, Nontuberculous mycobacteria: community and nosocomial of molecular methods for drug resistance detection and genotyping in M. tuberculosis, waterborne opportunistic pathogens, Clin. Microbiol. Newsl. 38 (2016) 1–7, while she carried out secondments in the Institut de Génétique et Microbiologie (Institut de https://doi.org/10.1016/J.CLINMICNEWS.2015.12.003. Biologie Cellulaire et Intégrative, France) and at the Odessa National University/Odessa [9] R. Briancesco, M. Semproni, R. Paradiso, L. Bonadonna, Nontuberculous Regional Tuberculosis Hospital (Ukraine). During the post-doctoral researcher stage, she mycobacteria: an emerging risk in engineered environmental habitats, Ann. continued in the line of development of methods for the diagnosis of tuberculosis and other Microbiol. 64 (2014) 735–740, https://doi.org/10.1007/S13213-013-0708-8/ respiratory diseases and participated in the development of European H2020 projects. TABLES/3. Until 2023, she was project manager of the H2020-MSCA-RISE-2018 INNOVA4TB project. [10] C. Ruis, J.M. Bryant, S.C. Bell, R. Thomson, R.M. Davidson, N.A. Hasan, J. van Ingen, M. Strong, R.A. Floto, J. Parkhill, Dissemination of Mycobacterium abscessus via global transmission networks, Nat. Microbiol. 6 (2021) 1279–1288, Dr. Bárbara Camila de Araújo Souza: PhD in Chemistry with emphasis on Analytical https://doi.org/10.1038/s41564-021-00963-3. Chemistry from São Paulo State University. She completed a sandwich internship at the [11] M.F. Brugnera, R. Bundalian, T. Laube, E. Julián, M. Luquin, M.V.B. Zanoni, M. Autonomous University of Barcelona, where she specialized in molecular biology tech­ I. Pividori, Magneto-actuated immunoassay for the detection of Mycobacterium niques for the construction of electrochemical biosensors for pathogenic bacteria. During fortuitum in hemodialysis water, Talanta 153 (2016) 38–44, https://doi.org/ her master’s degree, carried out at UNESP, she developed external academic research to 10.1016/j.talanta.2016.02.041. improve the quality of life of immunocompromised patients. She studied a bachelor’s [12] E. Torvinen, P. Torkko, A. Nevalainen, H. Rintala, Real-time PCR detection of degree in chemistry at the Federal University of Bahia. She has experience in topics environmental mycobacteria in house dust, J. Microbiol. Methods 82 (2010) involving photoelectrocatalysis, water treatment, molecular biology and microbiology. 78–84, https://doi.org/10.1016/J.MIMET.2010.04.007. She is currently a researcher associated with the Bioenergy Research Institute developing [13] R. Dibaj, H. Shojaei, T. Narimani, Identification and molecular characterization of bioelectrochemical systems with the aim of producing high value-added products from mycobacteria isolated from animal sources in a developing country, Acta Trop. 204 agro industrial waste using Pseudomona aeruginosa. (2020), 105297, https://doi.org/10.1016/J.ACTATROPICA.2019.105297. [14] S. Liébana, A. Lermo, S. Campoy, J. Barbé, S. Alegret, M.I. Pividori, Magneto Dr. Arnau Pallarès Rusiñol: He is currently working at Bioeclosion SL as researcher () immunoseparation of pathogenic bacteria and electrochemical magneto and project manager. He holds a B.A. degree in Chemistry (2016) and MSc in "Industrial genosensing of the double-tagged amplicon, Anal. Chem. 81 (2019) 5812–5820, Chemistry and Introduction to Chemical Research" (2017) at the "Universitat Autònoma de https://doi.org/10.1021/ac9007539. Barcelona" (UAB), Catalonia, Spain. He received his PhD degree in Chemistry (2022) [15] F.G. Sartori, L.F. Leandro, L.B. Montanari, M.G.M. De Souza, R.H. Pires, D.N. Sato, working in Sensors and Biosensors Group, Department of Chemistry at UAB. His thesis C.Q.F. Leite, K. De Andrade Prince, C.H.G. Martins, Isolation and identification of work focused on the development of in vitro diagnostic tests (IVDs) for the detection and environmental mycobacteria in the waters of a hemodialysis center, Curr. monitoring of breast cancer disease by using extracellular vesicles, or exosomes, as targets. Microbiol. 67 (2013) 107–111, https://doi.org/10.1007/S00284-013-0341-6. After his PhD, he has continued with research on IVDs at Bioeclosion SL, a UAB spinoff. [16] L. Hall-Stoodley, H. Lappin-Scott, Biofilm formation by the rapidly growing mycobacterial species Mycobacterium fortuitum, FEMS Microbiol. Lett. 168 (1998) 77–84, https://doi.org/10.1111/J.1574-6968.1998.TB13258.X. Dr. Jofre Ferrer-Dalmau: He holds a Ph.D. in Biomedicine, as well as a degree in [17] S. Liébana, A. Lermo, S. Campoy, M.P. Cortés, S. Alegret, M.I. Pividori, Rapid Biotechnology and Biochemistry from the Autonomous University of Barcelona. He has detection of Salmonella in milk by electrochemical magneto immunosensing, successfully completed an MBA at the Escuela de Administración de Empresas (EAE, Biosens. Bioelectron. 25 (2009) 510–513, https://doi.org/10.1016/j. Spain). He serves as the co-founder and Chief Executive Officer of Bioeclosion, Spain. Since bios.2009.07.022. 2016, he has been actively engaged with BioEclosion while concurrently managing the [18] Association for the Advancement of Medical Instrumentation, 2006. Water Analytical & Bioanalytical business unit at Anapath Research SAU. treatment equipment for hemodialysis applications (ANSI/AAMI RD62:2006). Arlington, 2007. Dr. Maria Valnice Boldrin Zanoni: She is Professor of Analytical Chemistry at the [19] H. Kawanishi, T. Akiba, I. Masakane, T. Tomo, M. Mineshima, T. Kawasaki, Institute of Chemistry, UNESP, Araraquara. Doctor of Chemical Sciences from the Institute H. Hirakata, T. Akizawa, Standard on microbiological management of fluids for of Chemistry of the University of São Paulo, São Carlos (1989), Post-doctoral at the Uni­ hemodialysis and related therapies by the Japanese Society for Dialysis Therapy versity of Loughborough, England; University of Wisconsin, Madison, USA and Oxford 2008, Ther. Apher. Dial. 13 (2009) 161–166, https://doi.org/10.1111/J.1744- University. She was advisor of the Research Development from UNESP from 2009–2017. 9987.2009.00674.X. She is currently coordinating of “National institute for alternative technologies of detec­ [20] A.D. Coulliette, M.J. Arduino, Hemodialysis and water quality, Semin. Dial. 26 tion, toxicological evaluation and removal of micropollutants and radioaticves”. She has (2013) 427, https://doi.org/10.1111/SDI.12113. experience in electroanalysis of micropollutants, photoelectrocatalytic processes and [21] Z. Hussein, O. Landt, B. Wirths, N. Wellinghausen, Detection of non-tuberculous environmental electrochemical. In 2014 was honored with the Elsevier/CAPES Award, mycobacteria in hospital water by culture and molecular methods, Int. J. Med. which highlighted 10 Brazilian researchers, which works had higher impact in the sci­ Microbiol. 299 (2009) 281–290, https://doi.org/10.1016/J.IJMM.2008.07.004. entific community. [22] L.B. Montanari, F.G. Sartori, M.J.O. De Cardoso, S.D. Varo, R.H. Pires, C.Q.F. Leite, K. Prince, C.H.G. Martins, Microbiological contamination of a hemodialysis center Dr. José Domínguez: Dr José Domínguez is a researcher at Institut d′Investigació Ger­ mans Trias I Pujol (IGTP) in Badalona, Spain, and group leader in Innovation in Respi­ ratory Infections and Tuberculosis Diagnosis. He is also assistant professor of the Departament de Genètica i Microbiologia at Universitat Autònoma de Barcelona.He also contributes to CIBER Enfermedades Respiratorias, Instituto de Salud Carlos III. His work 7 M. Mesas Gómez et al. Sensors and Actuators: B. Chemical 403 (2024) 135211 primarily focuses on diagnosing respiratory diseases and tuberculosis, integrating clinical Dr. María Isabel Pividori: She is Professor in the Chemistry Department, and group and translational research in the activities. leader in the Institut de Biotecnologia i de Biomedicina (IBB), Autonomous University of Barcelona (UAB), Catalonia, Spain. She holds a B.A. degree in Biochemistry with First Class honours from the Universidad Nacional del Litoral, Argentina. She received her PhD de­ Dr. Esther Julián Gómez: She obtained her BSc in Biology (1994) from the University of gree in Chemistry from the Autonomous University of Barcelona, in 2002. Her research is Barcelona, and her MSc in Microbiology (1997) and PhD in Biological Sciences (2002) particularly focused on the design of in vitro diagnostic tests, biosensors and bio­ from the Autonomous University of Barcelona (UAB). She then joined the Immunology instrumentation. Her research expertise is related with the improvements in terms of the Department at the University of Stockholm (2004) where she did her postdoctoral training. biorecognition event as well as in the simplification of the biosensing procedures, the In December 2009, Esther Julián started as an Associate Professor and Principal Investi­ integration of nanomaterials and the enhancement of the analytical signal. Regarding the gator of the Mycobacteriology Research Laboratory in the Genetics and Microbiology application fields, she is mainly focused on affordable emerging technologies appropriate Department at the UAB. From June 2023 she is Full Professor at the same Department. Her at community and primary-care level in healthcare and food safety in low resource main scientific interests are in the field of mycobacteriology, in particular, the use of settings. mycobacteria as immunomodulatory and immunotherapeutic agents. 8