The human immunodeficiency virus (HIV) attacks the immune system to destroy T-helper cells , a type of white blood cell, as well as rapidly duplicating itself. Thus, someone with HIV virus may likely be infected other critical viruses such as hepatitis B virus (HBV) and hepatitis C virus (HCV) [1, 2]. For example, about 25 % of HIV-infected people are coinfected with HCV. In general, people with both HIV and HCV need to be treated for both diseases with appropriate medicine. Thus, medical doctors should be very cautious when prescribing HIV and HCV drugs carefully to avoid any critical side effect(s) due to drug-to-drug interaction.

Recently, many research groups are interested in developing analytical methods for the detection of HIV as well as HCV [3, 4, 5, 6, 7]. Currently, it is possible to quantify trace levels of HIV and/or HCV using biosensors with various detection methods such as chemiluminescence [6], colorimetric [4], electrochemical [3], fluorescence [7], and mass spectroscopy [5]. Unfortunately, it is difficult to consecutively or simultaneously quantify HIV and HCV in a sample using the current analytical tools because the analytical time required for the quantification of HIV is different from that of HCV.

It is expected that 1,1’-oxalyldiimidazole chemiluminescence (ODI-CL) detection may solve this problems. It has been reported that 1,1’-oxalyldiimidazole chemiluminescence (ODI-CL) detection is more sensitive than other optical detections [8, 9, 10, 11]. Thus, it was possible to develop a rapid biosensors using the highly sensitive ODI-CL detection because the time necessary to quantify a biomarker in a sample is determined by the sensitivity of the detection system [12, 13]. For example, the biosensor with ODI-CL detection was able to consecutively and rapidly quantify coagulation factors IIa (thrombin) and Xa in plasma and whole blood [14]. This is because the highly sensitive ODI-CL detection can analyze trace levels of luminophores formed from the short incubation (e.g., 2 min in plasma, 4 min in whole blood) for the hydrolysis reaction between IIa (or Xa) and a specific fluorogenic substrate.

HIV-1 protease (PR)[15, 16] and HCV PR [17, 18] are widely applied as biomarkers for diagnosing HIV and HCV infections. In order to analyze HIV-1 PR and HCV PR, various fluorogenic substrates have been developed [15, 16, 17, 18]. Thus, based on the previous research results for the analyses of coagulation factors IIa and Xa, [14] it is possible to cost-effectively diagnose HIV and HCV infections if a biosensor with ODI-CL detection, capable of consecutively quantifying HIV and HCV in a sample, is developed.

The functions and resolution of the camera in a smartphone are as good as those of expensive digital single-lens reflex (DSLR) camera. For example, it is possible to capture clear photo images in low light using the extension of exposure time of the smartphone camera lense. In other words, it is possible to sense dim light emitted from chemiluminescence reaction using the functions such as high resolution and control of exposure time of a smartphone camera.

Recently, cost-effective and rapid 3D printers have become commercially avaiable. Using the 3D printers, it is possible to fabricate prototypes at a low cost before manufacturing the final products.

Using the advantages of ODI-CL detection, smartphone camera, and 3D printer, we were able to propose two hypotheses for the consecutive or simultaneous detection of HIV and HCV in a sample.

Hypothesis 1: “If ODI-CL detection can quantify trace levels of luminophore from the hydrolysis reaction of HIV-1 PR (or HCV PR) and a fluorogenic substrate during a certain incubation time, a highly sensitive biosensor can be developed for the consecutive quantification of HIV and HCV in a sample.”

Hypothesis 2: “If a portable device, fabricated with a 3D printer and operated with a smartphone camera and ODI-CL detection can capture dim light, it is possible to qualitatively and simultaneously diagnose HIV and HCV infections using light emitted from the two reaction wells of the device with each well containing HIV-1 PR or HCV PR.”

With a research plan based on the two hypotheses, we devised two types of biosensors with ODI-CL detection. First, we developed a quantitative biosensor for the determination of HIV-1 PR and HCV concentrations. Then, we fabricated a qualitative biosensor for positive and negative tests of HIV and HCV infections. The detailed research results and their application were reported in this manuscript.

HIV-1 PR recombinant was purchased from Prospec Protein Specialists. Fluorogenic substrate (DABCYL - GABA - Ser - Gln - Asn - Tyr - Pro - Ile - Val - Gln – EDANS) of HIV- 1 PR was purchased from CRC Scientific. EDANS (5 - [(2 - aminoethyl)amino]naphthalene - 1 - sulfonic acid) is a luminophore. HCV NS3/4A PR 1b, recombinant, and fluorogenic substrate (DABCYL-gamma-Abu-Ser-Gln-Asn-Tyr-Pro-Ile-Val-Gln-EDANS) of HCV NS3/4A PR were purchased from Anaspec. Bis(2,4,6-trichloro)phenyl oxalate (TCPO) and 4-Methylimidazole were purchased from TCI America. Glucose, glucose oxidase, horseradish peroxidase (HRP), fluorogenic substrate (Benzoyl-Phe-Val-Arg-AMC, HCl) of thrombin (IIa), and IIa were purchased from Sigma Aldrich. Coagulation factor Xa was purchased from Thermo Scientific. Fluorogenic substrate of factor Xa (CH₃SO₂-D-Cha-Gly-Arg-AMC, AcOH) was purchased from Cryopep. 3% and 30 % H₂O₂ were purchased from VWR. Deionized H₂O (HPLC grade), Ethyl acetate, Isopropyl alcohol, and high concentration of PBS (pH 7.4, 20×), TBS (pH 7.4 × 10), phosphate buffered saline with Tween-20 (PBST) and Tris buffered saline with Tween-20 (TBST) were purchased from EMD. 8-well EIA/RIA strip-well plate was purchased from Costar. Prostate specific antigen (PSA) ELISA kit was purchased from Monobind.

As shown in Scheme 1, ODI was formed from the reaction of bis(2,4,6-trichloro)phenyl oxalate (TPCO) and 4-methylimidazole in ethyl acetate at room temperature. ODI in ethyl acetate was stable for at least 8 hours at room temperature. A certain concentration of luminophore (10 μl) was injected in a borosilicate test tube (12 mm × 75 mm). The tube was inserted into the sample holder of a luminometer with two syringe pumps (Lumat 9507, Berthold) to measure CL emission. 100 mM H₂O₂ (25 μl) dissolved in isopropyl alcohol was dispensed through the first syringe pump of the luminometer. With the addition of ODI (25 μl) using the second syringe pump, the luminometer measured immediately the strength of the light emitted in the test tube.

As shown in Scheme 1, the color of light emitted in the test tube was dependent on the chemical and physical properties of luminophore [19, 20]. This is because luminophore (L*), formed after receiving energy from the high-energy intermediate (X), formed from the reaction of ODI and H₂O₂, emits light.

In order to determine the rapid reaction condition between HIV-1 PR and fluorogenic substrate of HIV-1 PR, we used three different pH buffers such as sodium phosphate buffer (pH 5.1), sodium phosphate buffer (pH 5.7) and sodium acetate with EDTA (pH 4.7). This is because other research groups reported that the range of pH optimized for the reaction of HIV-1 PR and fluorogenic substrate is 4.7 ~ 6.0.

In order to determine an appropriate buffer solution for the analysis of HCV PR with fluorogenic substrate, Tris-HCl (pH 8), Tris-HCl (pH 8.5) and Tris-HCl (pH 7.5) with 50 % Glycerol were used.

A certain concentration of protease (HIV-1 PR or HCV PR, 50 μl) and 50 μM fluorogenic substrate (50 μl) mixed in a 1.5-ml centrifuge tube were incubated for 10 min at 37 °C. After the incubation, each sample (10 μl) injected into a borosilicate tube was inserted into the luminometer to measure the ODI-CL.

In order to determine whether HIV-1 PR and HCV PR are in a sample, we designed a portable device operated with a smartphone (e.g. LG V10, Samsung Note 4) using a online software (https://www.onshape.com) for 3D printing (see Fig. 1). Finally, the device was printed with a 3D printer (Monoprice Select) and black polylactic acid (PLA). Two strip-wells were inserted into the two sample holders of the device. HIV-1 PR and HCV PR samples (25 μl each) injected in one of the two strip-wells reacted with fluorogenic substrates (25 μl each) in two strip-wells for 3 min. After the reaction, H₂O₂ (50 μl) in isopropyl alcohol was added in two strip-wells. Finally, using LG V10 smartphone, we took images of blue light emitted immediately in the two strip-wells with the addition of ODI through the two holes of the device. The picture of the light emitted in the device was obtained using the exposure function of the smartphone camera. The exposure time was 10 sec. Fig. 2 shows the device operated with a smartphone.

The experimental data obtained with the luminometer was analyzed and evaluated with the statistical tools of 2016 Microsoft Excel.

High energy intermediate (X) formed from the reaction of ODI and H₂O₂ transfer energy to luminophore based on the principle of internal chemiluminescence resonance energy transfer (inter-CRET) [19]. In general, luminophore after the inter-CRET emit bright light. However, luminophore conjugated with peptides and a quencher, fluorogenic substrate, cannot emit light in the absence of protease after the inter-CRET between luminophore and X.

This is because the energy of luminophore in the excited state transfers immediately to the quencher as shown in Fig. 3(A) due to the intra-CRET. Thus, the relative CL intensity under the condition of Fig.3 (A) was as same as the background in the absence of protease and fluorogenic substrate.

As shown in Fig. 3(B), fluorogenic substrate is separated in the presence of protease due to the hydrolysis reaction between fluorogenic substrate and protease. Thus, luminophore can emit bright light when ODI-CL reagents are added in the solution containing free luminophore after the hydrolysis reaction. For example, Fig. 3(C) shows that the relative CL intensity after the hydrolysis reaction between fluorogenic substrate and HCV PR was much stronger than that in the absence of HCV PR.

As shown in Fig. 4(A), the best buffer solution for the quantification of HIV -1 PR in a sample sodium phosphate (pH 5.1), prepared with the addition of HCl in sodium phosphate buffer (pH 7). Relative CL intensity after the hydrolysis reaction in sodium phosphate 5.1 was about 2.5-fold higher than that in sodium acetate with 1 mM EDTA (pH 4.7) applied to measure fluorescence [21]. The relative CL intensity in sodium phosphate buffer (pH 5.1) was higher than that in sodium phosphate buffer (pH 5.7). In addition, we confirmed that relative CL intensity in sodium phosphate (pH 5.1) with 1 mM EDTA was much lower than that in sodium phosphate buffer only. EDTA may be a quencher in ODI-CL reaction. Thus, we selected sodium phosphate (pH 5.1, 10 mM) for the quantification and monitoring HIV-1 PR using ODI-CL detection.

Fig 4(B) indicates that the best buffer is Tris-HCl (pH 8.5). This is because relative CL intensity after the hydrolysis reaction in Tris-HCl (pH 8.5) was much higher than those in other buffers. Also, we didn’t add glycerol in Tris-HCl (pH 8.5) for the analysis of HCV PR using ODI-CL detection even though it was used for measuring fluorescence in the previous report. This is because we confirmed that glycerol is a strong quencher in ODI-CL reaction. Thus, we have used Tris-HCl buffer (pH 8.5) in this research.

As shown in Fig. 5, relative CL intensity was enhanced with the increase of fluorogenic substrate concentration. The concentration of HIV-1 fluorogenic substrate used in the hydrolysis reaction for the analysis of HIV-1 PR was much higher than that for the detection of HCV PR. The results indicate that the hydrolysis reaction between HIV-1 fluorogenic substrate and HIV-1 PR in sodium phosphate buffer (pH 5.1) is slower than that between HCV fluorogenic substrate and HCV PR in Tris-HCl buffer (pH 8.5).

As shown in Fig. 6, relative CL intensity measured after the hydrolysis reaction between HCV fluorogenic substrate and HCV PR in Tris-HCl buffer (pH 8.5) was enhanced with the increase of the incubation time. The highest relative CL intensity was measured after 3 min of incubation. Then, relative CL intensity after an incubation time longer than 3 min was similar to that of the 3-min incubation with acceptable error range (± 5 %). The incubation time for the hydrolysis reaction of HIV-1 fluorogenic substrate (108 μM) and HIV-1 PR (100 nM) was also studied. We confirmed that the incubation time for the measurement of the highest CL intensity was as short as 3 min. Thus, we selected the 3-min incubation for the hydrolysis reaction to develop a biosensor capable of rapidly and simultaneously sensing HIV-1 PR and HCV PR.

Fig. 7 shows that the biosensor with ODI-CL detection can quantify trace levels of HIV-1 PR and HCV PR with the calibration curves. The dynamic range of calibration curve for the analysis of HIV-1 PR is from 12 to 92 nM. The limit of detection (LOD = Background + 3σ) of the biosensor capable of sensing HIV-1 PR was as low as 4.8 nM. σ is the standard deviation of the background determined after the 20-time measurement (N = 20). The dynamic range of calibration curve for the quantification of HCV PR was from 2.8 to 44 nM. LDO of the biosensor for the analysis of HCV PR was as low as 0.45 nM.

Fig. 9 shows that the portable device operated with the camera of LG V10 smartphone to catch images of light emitted from ODI-CL reaction. The exposure time was as long as 10 sec to get the images of CL emission in the presence of HCV PR (40 nM, left) and HIV-1 PR (90 nM, right). Fig 9(A) indicates that the first sample contains just HCV PR, whereas the results of Fig. 9(B) shows that HIV-1 PR without HCV PR is in the second sample. Fig. 9 (C) shows that the third sample have HCV PR as well as HIV-1 PR. Thus, bright blue light was detected at the two sample holders of the device. However, the light blue emitted in the presence of HCV PR in Tris-HCl buffer (pH 8.5) was different from the blue mixed with light green in the presence of HIV-1 PR in sodium phosphate buffer (pH 5.1). These results indicate that the color of EDANS emitted from ODI-CL reaction is dependent on the pH of buffer solution.

Like HIV-1 PR and HCV PR, coagulation factors IIa and Xa react with a specific fluorogenic substrate conjugated with 7-Amino-4-methylcoumarin (AMC). Free AMC formed after the 5-min hydrolysis reaction of a fluorogenic substrate and IIa (or Xa) emit blue light. The color of AMC emitted after the hydrolysis reaction in TBST (pH 7.5) for detecting coagulation factor IIa in TBST (pH 7.5) is the same as that in PBS (pH 7.4) for sensing coagulation factor Xa. The results indicate that the color of luminophore emitted in ODI-CL reaction is dependent on the pH of buffer solution rather than the components of the buffer solution.

Fig. 11 shows that the portable device operated with the LG V10 Smartphone and ODI-CL reaction can detect glucose in a sample. As shown in Fig. 11, the brightness of CL is dependent on the concentration of glucose in a sample. With the increase of glucose concentration, the brightness of CL emission was enhanced. After obtaining the images, we were able to read relative CL intensity using the software (ImageJ) provided by the National Institute Health (NIH). The relative CL intensity (223 ± 21) of the right spot was about 1.5-fold brighter than that (145 ± 17) of the left spot. The results indicate that the device can be used to diagnose and manage diabetes.

Fig. 12 shows that the portable device can to be applied as the tool for ODI-CL immunoassay capable of diagnosing prostate cancer.

PSA (25 μl), biomarker of prostate cancer, and detection antibody conjugated with HRP (100 μl) were added in the strip-well coated with capture antibody. The mixture in the strip-well was incubated for 30 min at room temperature. After the incubation, the well was washed 3 times with PBST. The strip-well was inserted into the sample holder of the portable device. Then, the mixture (50 μl) of Amplex Red (non-luminescent dye, substrate), H₂O₂, and 4-iodophenol in PBS (pH 7.4) was added and incubated for 10 min in the strip-well. After the incubation, H2O2 (50 μl) in isopropyl alcohol was added in the strip-well. Finally, ODI was injected into the strip-well through the hole using the multi-pipette to capture the image shown in Fig. 12.

As shown in Fig. 12, the brightness of CL is dependent on the concentration of PSA in a sample. In other words, red light emitted in the presence of 50 ng/ml PSA was brighter than that that in the presence of 5 ng/ml.

We developed for the first time a biosensor with ODI-CL detection capable of consecutively and rapidly quantifying HIV-1 PR and HCV PR in a sample. In other words, we expect that the biosensor can be applied to early diagnose and rapidly monitor HIV as well as hepatitis C infections.

Based on the development of the quantitative analytical method, we also were able to design and fabricate a portable device operated with a smartphone and ODI-CL reaction for the qualitative diagnosis (positive and negative) of HIV and hepatitis C infections. We also confirmed that the device can be applied to early diagnosis and monitor various human diseases such as diabetes, blood clot disorder, cancer, cardiac ailments, and infectious diseases.

As future work, it is necessary for the development of a fully automated portable device based on the prototype, fabricated with 3D printer, operated with a smartphone and ODI-CL reaction. We expect that the automated device operated with a professional App installed to smartphone will be applied to various fields such as biochemistry, environmental science and engineering, homeland security, medical science, and toxicology.