**2. Development of enzymatic Pi biosensors for eutrophication**

In the early times of biosensor development for environmental use, there were many technological problems for practical use. To solve the problems, a wide variety of challenges have been carried out. In this section, the main breakthrough technics for the practical application in the field of biosensor development for eutrophication are described based on the previous literature [3, 10, 14, 15].

### **2.1 Standard method and consideration to practical estimation performances**

As mentioned above, the most causative nutrients of the eutrophication are phosphates, which are classified into Pi, condensed phosphates (pyro-, *meta*-, and other polyphosphates), and organic phosphates. Among the phosphates, the one with the greatest effect of eutrophication is Pi, which is an inorganic phosphate ion.

In Japan, a spectrophotometric molybdenum blue method for Pi determination is employed as the standard method [16], which is based on the method by Lowry and Lopez in 1946 [17]. By this method, Pi concentration can be determined between 0.1 and 3.0 mg/L Pi (3.2–96 μM) at a relative standard deviation (RSD) of 2–10%. The detection limit in this method is 0.03 mg/L Pi (0.32 μM Pi).

These values obtained by the standard method are barely applicable as an indication of the eutrophication (0.64 μM Pi), however, they are insufficient for the estimation of the eutrophication, which is classified into the five categories in lakes (between around 0.032–3.2 μM Pi). In addition, the maximal permissible concentration in Japanese lakes is 0.32 μM Pi for drinking and 3.2 μM Pi for environmental protection [18]. Thus, the influences of Pi on eutrophication are caused at extremely low concentrations. Therefore, it was found that the estimation of the eutrophication requires highly sensitive and wide-range Pi determination techniques.

For ideal estimation of the eutrophication, the standard method has difficulties in both accurate and high-precision measurements. In accurate measurement, dilution of sample solution needs due to that the dynamic range is narrow and causes error. In high-precision measurement, the minimum limit of determination (3.2 μM Pi) is insufficient to estimate the water quality in all categories of eutrophication.

To make ideal estimations for all categories of the eutrophication, a measurement method that was more sensitive and had a wider dynamic range (if it is possible; 0.032–3.2 μM Pi) than the standard methods was needed. However, considering the biosensor performance at that time, such high sensitivity was not realistic. Even if such highly sensitive Pi measurements cannot be made, we should be able to determine 0.32 μM Pi with a biosensor at a concentration that can confirm signs of eutrophication and monitor the quality of the lake water for drinking. In addition, we considered that the maximum limit of Pi determination obtained by the biosensor should be kept at least 3.2 μM or more as the practical dynamic range.

Like the other problems, the standard method is complicated and time-consuming to operate. Further, it requires the use of strong acid and heavy metal ions. The use of such chemicals is subject to limitations when applied to on-site monitoring and continuous monitoring because leakage of the chemicals into the environment has to be prevented. In addition, the standard method affects the influences of co-existing matters in a sample, such as Ca2+, Fe3+, NH4 + , NO2 − , NO3 − , and AsO4 3−.

As is clear here, the standard method is not sufficient to evaluate eutrophication. For this reason, the development of practical Pi biosensors has been performed.

*Developmental Studies on Practical Enzymatic Phosphate Ion Biosensors and Microbial BOD… DOI: http://dx.doi.org/10.5772/intechopen.104377*

#### **2.2 Development of enzymatic Pi biosensors**

It turns out that there were many issues that need to be resolved to actually estimate the eutrophication. Thus, we have set five requirements for the practical application of the Pi biosensors [19].

For practical use, our Pi biosensors are able to;



**Table 1.**

*Looking back on the five requirements that were set for the practical use of our automated CL-FIA systems as Pi biosensors (performed by Nakamura et al.).*


We tried to develop practical Pi biosensors that meet these requirements (**Table 1**). As a reference, the history of the studies on the Pi biosensor development has been reviewed in several articles [3, 10, 15].

### *2.2.1 Development of early time Pi biosensors*

A biosensor has characteristics that it can measure analyte using biological reactions by applying the molecular recognizing function in living organisms (**Figure 2a**). A Pi biosensor can measure directly the concentration of the bioavailable Pi existing in the environmental water. This means that the Pi biosensors have the possibility to be able to monitor the status of the water ecosystem. This section briefly explains the history of the development of Pi biosensors with such potential, which is above described [3].

The first Pi biosensor was studied by Guilbalt and Nanjo in 1975. Using the Clark-type DO electrode, they studied an enzymatic Pi biosensor that was based on the inhibition of alkaline phosphatase activity by Pi [22]. By employing the inhibitory reaction, this biosensor lacked sensitivity with the detection limit of 0.1 mM Pi. In addition, the inhibitory reaction is low selectivity in general. In 1990, d'Urso and Coulet studied a two-enzyme Pi biosensor using nucleoside phosphatase and xanthine oxidase [23]. Enzymatically generated hydrogen peroxide (H2O2) by the existence of Pi was electrochemically measured. In this study, the Pi biosensor could have a dynamic range between 0.1 and 10 μM Pi. In 1992, Wollenberger et al. improved the two-enzyme system to a multiple-enzyme system for Pi recycling [24]. The Pi biosensor was fabricated by incorporating a Clarktype DO electrode into a flow injection analysis (FIA) system. As the result, this Pi biosensor realized an excellent detection limit at 25 nM Pi. However, this biosensor was not suitable for practical use due to its short lifetimes by using an unstable enzyme. As another reason, it was reported that inosine used as another substrate was also unstable [25].

#### *2.2.2 Development of our practical Pi biosensors*

Karube et al. have studied several kinds of enzymatic Pi biosensors using pyruvate oxidase (PO) from *Pediococcus* sp. (PO*P*sp) [9, 26–28] and genetically engineered PO from *Lactobacillus plantarum* (PO*Lp*) [29], *Aerococcus viridans* (PO*Av*) [8], and maltose phosphorylase (MP) [11, 20]. These studies were performed for the practical use of an automatic FIA system (desktop-type; **Figure 2b**) [8] and a remote-controlled automatic continuous Pi monitoring system (submersible buoy-type; **Figure 2c**) [9].

The Pi biosensor using PO is superior to other Pi biosensors because it requires only one step of a catalytic reaction for selective Pi detection. Using PO*P*sp, the first Pi biosensor was studied by Kubo et al. in 1991 [26]. The Pi biosensor was batch type and used a DO electrode to measure the consumption of DO by the PO*P*sp catalytic reaction under the existence of Pi, two cofactors (thiamin pyrophosphate; TPP and flavin adenine dinucleotide; FAD) and an enzyme activator (Mg2+). However, the detection limit of 12 µM Pi was insufficient for environmental water control.

#### *Developmental Studies on Practical Enzymatic Phosphate Ion Biosensors and Microbial BOD… DOI: http://dx.doi.org/10.5772/intechopen.104377*

In 1996, Ikebukuro et al. examined the combination of a luminol chemiluminescence (CL) reaction and a FIA system for enzymatic Pi biosensor (CL-FIA system) [27]. At that time, the CL reaction was known as a highly sensitive reaction, and the FIA system was also known as a highly repeatable measurement system [30, 31]. The reason why the latter repeatability contributes to high-precision analysis was that the closer the multiple measurement results obtained from the same standard solution were to their average values, the smaller the standard deviation value. Along with this, by increasing the significant difference between the standard measurement value and the blank value, the standard measurement value closer to the blank value could be set as the detection limit. As a result, by combining CL and FIA technics, it was possible to realize a high-precision analytical method due to the synergistic effect of both. This meant that the CL-FIA system can turn the enzymatic Pi biosensors into highly sensitive analytical instruments. Thus, the CL-FIA biosensor systems have been widely studied [32].

In the PO*P*sp catalytic reaction, hydrogen peroxide was produced and subsequently consumed by a luminol-peroxidase reaction (peroxidase from horseradish, HRP). The CL light resulting from the presence of Pi was detected at a photomultiplier tube (PMT). By the purification of TPP to remove residual phosphates, a detection limit of 74 nM Pi was obtained, although the TPP purification was unsuitable for practical use [28].

In 1997, Nakamura et al. used highly sensitive luminol catalyzing peroxidase from *Arthromyces ramosus* (ARP) [9]. This CL-FIA system was improved to the remotecontrolled automatic continuous Pi monitoring system and compactly integrated into a submersible buoy to monitor dam water for drinking (**Figure 2c**) [9, 10]. Then, PO*P*sp was immobilized onto chitin-chitosan beads filled in a stainless-steel column. In this study, 160 nM Pi was detected without purification of TPP, and by keeping PO*P*sp activity in the stainless-steel column, 0.32 μM Pi was detected for 48 days. This result was sufficient for practical use of the CL-FIA biosensor system for Pi monitoring. Unfortunately, the manufacturing of PO*P*sp was stopped.

In 1999, Nakamura et al. reexamined the development of the CL-FIA system for an enzymatic Pi biosensor using a new enzyme, PO*Av*, which was purified from recombinant *Escherichia coli* [8]. Then, PO*Av* was densely immobilized onto *N*-hydroxysuccinicacidimido (NHS)-gel beads. Here, the velocity of the enzymatic reaction strongly depends on the temperature. This means that the temperature must be kept constant by precise control to obtain repeatable signals based on the enzymatic reaction. On the other hand, accurate (i.e., absolute) temperature control dose is not essentially needed due to that the enzymatic activity immobilized in the biosensor changes. Thus, in this study, a desktop-type automatic CL-FIA system involving a precise temperature control system was fabricated as a trial enzymatic Pi biosensor instrument (**Figure 2b**). Then, the detection limit was 96 nM Pi. In addition, with sufficient performance at a short measurement time of 2 minutes, a wide linear range of calibration (96 nM and 32 μM Pi) was obtained with an average relative standard deviation (RSDav) of 2.3% (eight points, *n* = 5) at 25.0 C. In addition, this system kept making calibration curves from 0.16 to 32 μM Pi (five points, *n* = 3; averaged correlation, *r* = 1.00) for at least 2 weeks. Thus, the practical utility of the Pi biosensor system was demonstrated.

On the other hand, in 1995, Conrath et al. studied a new electrochemical Pi biosensor using an analyte recycling system consisting of four enzymes, maltose phosphorylase (MP), acid phosphatase (AcP), mutarotase (MUT), and GOD [33]. The four enzymatic system enabled the successful detection of 10 nM Pi. However, the system was too complicated, and we thought it would be difficult to reproduce in manufacturing as well as other multiple-enzyme systems [24]. In 1999, Nakamura

et al. modified the MP system and applied it to our CL-FIA biosensor system using a tri-enzymatic reaction of MP-MUT-GOD without the analyte recycling by AcP [20]. Then, we could obtain the same results with Conrath et al. at a detection limit of 10 nM Pi. In addition, an excellent calibration between 10 nM and 30 μM Pi was obtained and stability to detect 1.0 μM Pi was observed for at least 2 weeks.

In 2003, Nakamura et al. improved the MP-MUT-GOD system for freshwater measurements [11]. Our previous studies revealed that the Pi biosensors employing the CL-FIA system were affected by the cations contained in the real sample solutions. Then, we have tried to examine several pretreatment methods to remove the cations from the sample [8, 11, 20]. As the result, we could find and establish a pretreatment method using a cation-exchanging resin. A total of 31 samples of freshwaters were taken from the river and pond. These real samples were pretreated by our method and measured for Pi determination by both this Pi biosensor and the conventional molybdenum-blue method. The results showed that the value from the conventional method was 2.78 times higher than that from the Pi biosensor. One reason for this outcome was considered that there were large differences in the reacting conditions between the Pi biosensor (natural pH) and the conventional method (under strong acid conditions). Therefore, an enzymatic Pi biosensor may determine the more accurate and realistic Pi concentration as free and bioavailable Pi, which is needed to understand the water ecosystem.

Furthermore, Nakamura et al. studied a pyrophosphate ion (PPi) biosensor in 2004 [21]. As well as the Pi concentration, the PPi concentration is also an indicator of the eutrophication and the organic pollution. For the enzymatic PPi biosensor, inorganic pyrophosphatase (IP) was added to the PO*Av* reaction, and the possibility of this PPi biosensor to environmental water was also shown. In another study, Nakamura et al. also examined the development of a disposable electrochemical Pi biosensor chip for on-site monitoring [34]. Then, a commercially available self-monitoring blood glucose (SMBG) chip was applied to the Pi measurements. Pi was measured with a PO*Av* or several MP systems coupled with ferricyanide ion (Fe3+) as an electrochemical mediator.

Here, an author looks back on the five requirements that were set for the practical use of our Pi biosensors [8, 9, 11, 20, 21]. In the first requirement, the CL-FIA systems were successfully applied to the Pi biosensors and made it possible to highly sensitive and practical Pi measurements [8, 9, 11, 20, 21]. In the second requirement, our Pi biosensor was able to continue to make practical calibration curves as the stability tests and demonstrate the practicability [8, 9]. In the third requirement, the sample pretreatment method was finally established by the countermeasures against interfering matters [11]. In the fourth requirement, the establishment of the sample pretreatment method was enabled to perform real sample applications [11]. In the fifth requirement, the automation of the Pi and PPi biosensor systems was realized by employing the CL-FIA systems [8, 9, 21].

As described above, we studied several types of Pi biosensors for the estimation of eutrophication, and finally, two trial Pi biosensor instruments of the desktop and the submersible buoy were developed. Although thorough examinations using the real samples were necessarily employing the automatic Pi biosensor instrument for practical use and commercialization, various factors made it impossible to conduct the study (laboratory relocation, running costs, stop production of enzymes, etc.). In conclusion of our study for practical use of the enzymatic Pi biosensors, the highest risk was the stop production of the enzymes. As the general concerns, the development and subsequent practical use of the enzymatic biosensor tend to be expensive running costs for the enzyme. In particular, the FIA system consumes a lot of reagents and our CL-FIA system needs to contain the enzyme (HRP or ARP) in the CL reagent.

*Developmental Studies on Practical Enzymatic Phosphate Ion Biosensors and Microbial BOD… DOI: http://dx.doi.org/10.5772/intechopen.104377*

### **2.3 Subsequent development of Pi biosensors and a new insight into the future perspectives of the Pi biosensor fields**

Subsequent development of the Pi biosensors by other groups was introduced here. In 2005, Kwan et al. studied a screen-printed electrochemical Pi biosensor using PO. The enzymatic Pi biosensor had a linear calibration range from 75 to 625 μM Pi [35]. The PO reaction needs TPP, which releases Pi. For practical use, an investigation of the storage condition of TPP would be needed. In 2001, Mousty et al. studied an electrochemical tri-enzymatic Pi biosensor employing an MP-MUT-GOD reaction system [36]. This system had a linear calibration range from 1 to 50 μM and was stable for at least 2 weeks. In 1998, Fernandez et al. studied another type of electrochemical tri-enzymatic Pi biosensor [37]. The three enzymes, a substrate, a cofactor, and a mediator, were incorporated into hydrogels. The enzymatic Pi biosensor had a detection limit of 2 mM Pi. Quite complicated principles and unstable enzymes would be a problem for practical use. These enzymatic Pi biosensors had insufficient sensitivity for actual eutrophication. For practical use, the influences of reducing matters in a sample solution on the biosensor response cannot be ignored. In 2013, Lawal and Adeloju studied measures against reducing matters in electrochemical Pi biosensors [38]. For this aim, they used conductive polymer of polypyrrole to the electrochemical bi-enzymatic Pi biosensors and could not observe the obvious influences of uric and ascorbic acids on both amperometry and potentiometric methods. However, the sensitivity of the Pi biosensor was insufficient for natural waters, therefore, the effects of the polypyrrole on the Pi measurement might be unclear. In 2020, Korkut et al. also used the polypyrrole to a PO-Pi biosensor and successfully performed accurate Pi determination of eutrophied water at 91% [39].

The enzymatic Pi biosensors have been developed not only for environmental use but also for medical use (urine Pi) [40, 41] and food use [42]. In the case of food use, in 2020, He and Liu successfully developed a highly sensitive PO-Pi biosensor employing "Nano-Enabled Biosensing" techniques that combined with gold nanorods (AuNRs) as nanomaterials and conductive materials. The PO-Pi biosensor was able to detect 0.4 nM Pi, which is sufficient to estimate eutrophication. On the other hand, these conductive materials have the potential to perform well in both food and medical applications, where samples are high in reducing matters [38, 39, 42].

On the other hand, studies using molecular recognition elements other than enzymes have also been carried out. The molecular recognition elements can be categorized into two types. One is bacterial phosphate-binding proteins (PBPs) and another is an ionophore that is made of lipid or polymer. The PBP is a component of a phosphate transport system and a highly selective recognition element of Pi (*K*d = 1 μM) [43]. In bacteria, PBP has the function of transporting Pi into cells and organelles and does not bind other inorganic or organic ions except Pi and arsenate.

In 2002, Kubo et al. studied an electrochemical PBP Pi biosensor. They extracted the PBP from *E. coli* of wild type and detected Pi potentiometrically at a range from 0.1 to 1.5 mM [44]. In 2004, Lyndon et al. applied a mutant PBP for fluorometric Pi biosensor. Then, fluorophore was site-specifically labeled to the mutant PBP. This Pi biosensor showed excellent selectivity and was detected around sub-μM Pi [45], although the stability of PBP should be addressed. In 2019, Sarwar et al. successfully applied a fluorometric PBP-Pi biosensor (mutant PBP-labeled fluorophore), which was purchased from Thermo Fisher Scientific Co., to the estimation for actual eutrophication employing a smartphone camera detection system [46]. They prepared a cassette system for the smartphone using a 3D printer and the paper-based disposal

PBP-Pi biosensor chip using a Xerox printer. This PBP-Pi biosensor had high sensitivity of 1.1 ppb (12 nM) Pi and a response time of only 4 seconds, considering the effect of fluorescence photo-bleaching. One of the notable results in this study is the high reproducibility of the results obtained from the paper-based disposal PBP-Pi biosensor chips using Xerox printing. In general, reproducibility of the results obtained by the disposable type single-use biosensor was one of the problems in practical development. The author has been also energetically studying this problem in the practical study of the SMBG biosensor chips, therefore the author understand the difficulty [47]. In 2020, Franz et al. developed a PBP-Pi biosensor using a mutant PBP-labeled fluorophore and established a new principle of thermophoresis [48]. For Pi detection, they combined the advantages of a biological Pi-receptor based on the PBP with the principle of thermophoresis. The PBP-Pi biosensor enabled the sub-nanomolar Pi determination in sample volumes ≤10 μL.

In the cases applying the ionophores to the Pi biosensors, Carey and Riggan in 1994 applied cyclic polyamine to an ion-selective electrode (ISE) for Pi [49]. This Pi sensor specifically detected dibasic phosphate (HPO4 2−) and obtained a linear response between 1.0 μM and 0.1 M Pi. For the real sample application, the influences of sample pH have to be concerned. The ionophores from microbes were used for the Pi biosensors [50, 51]. As an example, the Pi biosensor studied by Wygladacz et al. obtained a linear calibration range from 1.0 μM to 2.5 mM Pi. However, the lifetime was the order of days [51].

Other principles for the Pi biosensors were also studied. In 2001, Schreiter et al. studied a Pi bioavailability assay employing a luminescent cyanobacterial reporter strain for the replacement of the conventional AGP test. The method enabled the highly sensitive detection of Pi from 0.3 to 8 μM, although it took 8 hours of incubation [52]. In 2003, Dollard and Billard studied a *pho A*::*lux*-based bacterial biosensor using a Pi-sensing plasmid and demonstrated the possibility for the assessment of Pi bioavailability [53]. However, the use of genetically engineered Pi biosensor elements is unsuitable in the cases of these leakages into the environment.

In this section, the author explained the representative studies on the development of Pi biosensors and Pi bioavailability assays for eutrophication. The biomaterials, such as enzymes, PBPs, and ionophores, were used for the Pi biosensors. These biomaterials need purification from living organisms. Therefore, the issue of cost and risk of production outage will be a challenge to the practical application of the Pi biosensors in this field. Further, in 2021, Becker et al. calculated the complete environmental factor (E-factor) of the enzyme [54]. The complete E-factor, including required waste and water, was calculated as 37,835 g-waste/g-enzyme. Therefore, the use of such biomaterials will require environmental consideration and not be neglected for the sustainability assessment of bioprocesses in the future. Furthermore, new technologies will be introduced one after another in this field, such as smartphones, 3D printers, nanomaterials, miniaturization, and automation.

### **3. Development of microbial BOD biosensors for organic pollution**

#### **3.1 Standard method and consideration to practical estimation performances**

To estimate the degree of organic pollution, several indicators of total organic carbon (TOC), COD, BOD, and DO have been employed. In general, COD is employed for closed water bodies of both natural water and seawater, and BOD is employed for

*Developmental Studies on Practical Enzymatic Phosphate Ion Biosensors and Microbial BOD… DOI: http://dx.doi.org/10.5772/intechopen.104377*

flowing water, such as rivers. The differences in the usage of COD and BOD estimation methods are simply determined by the presence or absence of the flux, which depends on the decomposition rate of organic matters dissolved in the water body by aerobes, although they are often employed as the references in each field.

In the estimation methods for organic pollution, only BOD involves the results obtained by a biological reaction. The conventional standard method of the BOD estimation is referred to as the 5-day BOD (BOD5) method [55, 56]. It requires 5 days of incubation to obtain the results. The BOD value (mg O2/L) is calculated from the amount of DO consumed by the aerobic decomposition of organic matters during incubation (primary fermentation). From this principle, BOD is also called biochemical oxygen consumption.

At the peak of organic pollution in advanced countries in the 1960s, the standard method of BOD was limited to the BOD5 method [55, 56]. The BOD5 method has several problems that prevent it from satisfying the needs for practical use in wastewater control, that is, this method is time-consuming and requires tedious operations. Therefore, a method that can be used to monitor the BOD value in real-time or continuous was urgently needed.

#### **3.2 Development of microbial BOD biosensors**

In 1960', there were many issues that need to be resolved to actually estimate the BOD. To solve these problems, many kinds of BOD biosensors have been studied and developed for practical use [57].
