Applicability of biosensor technologies in the detection of Coxiella burnetii infection in clinical samples

Abstract

Q fever is a zooantroponoze distributed worldwide; it is caused by obligate intracellular bacterium Coxiella burnetii. Q fever is considered challenging for diagnosis and treatment. Clinical manifestations can vary among patients, which makes differential diagnosis difficult. In addition, in animals, the infection often remains latent, thus, creating a risk of uncontrolled dissemination and transmission to humans. Hence, the need for rapid and accurate identification of the pathogen to take timely and adequate therapeutic and prophylactic measures, including on-site testing. Current strategies for the detection of C. burnetii are molecular assays, e.g. polymerase chain reaction (PCR) or next-generation sequencing (NGS), and traditional serological tests. The conventional pathogen detection methods have serious limitations that make them not suitable for on-site analysis, as it requires level three biosafety laboratory conditions for cultivation in eukaryotic cell culture and poses significant health risks. In this review, we highlight the problems related to the application of biosensors in the detection of C. burnetii infection and their place among the standard diagnostic methods, given the potential of biosensors to provide rapid and quantitative diagnosis. We consider the applicability of surface plasmon resonance (SPR)-based biosensors for C. burnetii detection. The elaboration of an SPR biochip with an immobilized structural C. burnetii protein as a recognition molecule is described. The SPR assay is based on the binding reaction ‘infectious structural protein – anti-C. burnetii antibodies’. We discuss our preliminary results addressing their application in C. burnetii detection.

Keywords:

• Biosensors
• Q-fever
• laboratory diagnostics
• surface plasmon resonance (SPR) assay

Introduction

Q fever is a zoonotic disease with global distribution and a wide host range; it is caused by intracellular gram-negative bacterium Coxiella burnetii, which spreads among people through inhalation of infected aerosols, consumption of dairy products and tick bites [1–3]. The main reservoirs for human infection are domestic ruminants [4]. Epidemic outbreaks of different sizes have been registered for years in Europe, where the infection is endemic to many countries [5], Bulgaria included. The incidence of Q fever (per 100,000 population in Bulgaria) in the periods 1961–1980, 1981–2000 and 2001–2012 was respectively from 0% to 1.5%ooo, from 0.01 to 5.64%ooo for the second period and between 0.16%ooo (2011) and 3.40%ooo (2004) and 3.74%ooo (2002) for recent years [6,7]. The last two large outbreaks in Bulgaria were registered in Etropole (2002) and in Botevgrad (2003–2004) [8,9]. The analysis of the data on the incidence of Q-fever for the period 2011–2017 shows mostly sporadic cases and two small outbreaks [10]. Serological studies among risk groups in Bulgaria (with atypical pneumonia and cardiovascular diseases) show a prevalence of 15–18% for C. burnetii [9,11].

C. burnetii has been characterized in category B of the potential bioterrorism agents by the Centers for Disease Control and Prevention (CDC) as an infectious agent that is particularly easy to spread, causing a disease and death rate and requiring specific diagnostic capacity and surveillance [1,12]. Q fever in humans has a wide variety of clinical manifestations: from asymptomatic infection to a chronic disease that can be fatal. Diagnosis of Q fever is challenging, since the clinical symptoms are very similar to those of other febrile diseases. Particularly, there have been observed self-limiting febrile illness [13], primary atypical pneumonia [14,15] or development of granulomatous hepatitis [12] in 40% of clinically manifested cases of acute Q fever infection. Furthermore, chronic Q fever is a serious medical challenge for physicians, and in some cases, the mortality rate is of up to 20–60% [16,17]. The most common clinical manifestations of chronic Q fever are infective endocarditis [16,18]. Risk factors for severe clinical form and fatal outcome are comorbidities such as hypertension, diabetes mellitus, cardiovascular diseases, neoplasm, chronic liver diseases, immunodeficiency disorders, etc. [19,20]. Chronic fatigue syndrome (CFS) can occur several years after the primary infection and can be treated ineffectively if the correct diagnosis has not been made [21].

All of this explains the increased interest in Q fever worldwide, as well as the need for an early opportunity to apply different diagnostic methods. In our previous review, we outlined the problems associated with the use of classical methods for the detection of the pathogen and antibodies against it, as well as the interpretation of their results [22].

Here, we focus on the applicability and feasibility of biosensor technology for fast and specific Q fever diagnosis.

Main challenges in Q fever diagnosis

The diagnosis of Q fever is difficult owing to the lack of а distinct clinical manifestation. This, together with the social significance of the disease, determines the need for sensitive, selective, rapid, cost-effective, real-time laboratory and on-site (portable) diagnostic methods. The greatest challenge to clinicians is diagnosing the disease during the acute and/or chronic phase of infection [1,12]. Misdiagnosis can occur because of the non-specific clinical symptoms, the low levels of the infectious agents in the clinical samples and the absence of specific laboratory confirmation due to the long period of seroconversion (up to 15–26 d). The difficulty in establishing a diagnosis is also determined by the stage of the disease (acute or chronic) and the choice of appropriate clinical samples [22,23].

At present, the most commonly used methods are the serological methods, the immunofluorescence assay (IFA) (gold standard for diagnosing Q fever), the indirect enzyme-linked immunosorbent assay (ELISA) and the immunohistochemistry (IHC) analysis of paraffin-embedded tissues [24,25]. It is well known [20] that the sensitivity of IFA is 95–100%, while the sensitivity of ELISA is a little bit lower. Regarding the specificity, IFA demonstrates 75–100% specificity, whereas ELISA about 10% lower. For both methods it depends on the used antibodies (IgG or IgM) [26].

Acute Q fever could be confirmed by the interpretation of the serological results based on the differential reactivity to C. burnetii antigens. The number of phase II antibodies rises in the blood, and they are first to be detected. In chronic Q fever, endocarditis included, phase I antibody is constantly expressed, while phase II antibody can be either expressed or not expressed. An acute Q fever is diagnosed in the early phase of infection by polymerase chain reaction (PCR) (≤7 d) and has been considered to be a confirmatory test [23]. Phase I and II antibodies are detectable 7 d following onset, and a fourfold rise in the antibody titre against phase II antigen must be observed [27]. Seroconversion occurs in about 90% of the patients by the third week of illness, whereas the development of antibodies typically takes 7–15 d after onset of symptoms. In the second week of acute illness, phase II IgG antibodies increase almost simultaneously with phase II IgM antibodies. Phase II IgG antibodies might remain detectable for a long time or even have lifetime occurrence in both symptomatic and asymptomatic infection [28]. Therefore, serum samples from the acute phase of Q fever do not aid in making immediate treatment decisions. In addition, high cross-reactivity with Legionella, Bartonella and other Coxiella species may occur [29,30]. An advantage of serology tests is that they allow the differentiation of acute from chronic Q fever by interpretation of results based on differential reactivity to antigens (phase 1 and phase 2) of C. burnetii. This is important for the development of a correct therapeutic approach, since the treatment of the acute form of the disease and chronic Q fever, differs.

Another approach for isolation of intracellular bacteria (like C. burnetii) from clinical samples is cultivation performed in embryonated eggs or in cell cultures [2]. Culture-based assays for this pathogen require biosafety level 3 (BSL-3) laboratories and are relatively time-consuming [31]. The sensitivity of cell cultures is low [32].

In the case of chronic Q fever, the efficient diagnostic tools are the combination of clinical symptoms, risk factors, serological tests and PCR detection of C. burnetii DNA [3,33]. The persistence of high levels of phase I IgG and IgA antibodies or the reoccurrence of these antibodies after treatment is likely to result in development of chronic Q fever [34].

More recently, PCR-based diagnostic assays have been successfully developed for the direct detection of C. burnetii DNA in cell cultures and clinical samples and are considered a confirmatory test for acute Q fever [23,28,35]. These assays are based on different modifications of the PCR technique assisted by LightCycler systems, SYBR green, TaqMan chemistry and loop-mediated isothermal amplification (LAMP) [36–39].

PCR-based identification of C. burnetii has been performed by targeting various genes, including 16S rRNA; sodB, a superoxide dismutase; singular chromosomal genes like com1, isocitrate dehydrogenase (icd). Moreover, a transposon-like repetitive region (Trans), the transposase gene of the IS1111 insertion element of the C. burnetii genome, and the heat shock operon encoding two heat shock proteins (htpA and htpB) also have been used for Q fever diagnosis [40,41]. The selection of an appropriate clinical material is essential for the detection of C. burnetii DNA, since the specificity of PCR assays depends on the target gene under amplification. Sensitivity in the range 33–67% has been reported for PCR testing involving serum from patients with acute Q fever [42,43]. However, lack of sensitivity has been reported for detection of C. burnetii in serum obtained from patients and conserved frozen for long periods [37]. A disadvantage of PCR protocols is that the identification of bacterial DNA does not indicate the viability of the bacteria [44,45].

In recent years, next-generation sequencing (NGS) has found an increased application in medical practice, especially in the diagnosis of infectious diseases requiring complex microbiological diagnostics. NGS has demonstrated its clinical utility in identifying a broad range of pathogens, C. burnetii included, with a high sensitivity and a short turn-around time [46,47]. Kondo et al. [46] described the application of NGS of microbial cell-free DNA (mcfDNA) in a patient’s plasma in diagnosing endocarditis caused by C. burnetii, as well as the use of strain typing by BLAST46. NGS is used to decipher Q fever outbreaks in the initial stages of infection, when the specific antibodies needed for serological analyses have not been formed yet, and it is utilized for early diagnosis of chronic Q fever as well [48]. The availability of whole C. burnetii genome-sequences was a major step to evaluate the genomic features, genetic diversity, evolution, as well as genetic determinants of antibiotic resistance, pathogenicity and ability to cause outbreaks of Q fever [49].

Both diagnostic platforms, the detection of C. burnetii antibodies by serological tests or its etiological diagnosis, respectively, require the availability of expensive laboratory equipment or reagents, qualified personnel, as well as a long technological time for analysis. Therefore, the search for new diagnostic platforms for the detection of C. burnetii is topical and urgent.

Biosensors for detection of C. burnetii in clinical specimens

The first biosensor for detection of the agent of Q fever was developed by Koo et al. [50]. It is based on a silicon micro ring resonator (SMR) interrogated optically by a feed wave-guide [51]. SMR transduces the binding reaction of biomolecules immobilized on its surface whenever the refractive index near the waveguide changes, which in turn changes the resonant wavelength. The assay involves isothermal DNA amplification (IDA) where recombinase polymerase amplification (RPA) is applied. A highly sensitive DNA amplification combined with smart SMR detection overcomes problems like occurrence of a false-negative reaction and low sensitivity due to inhibitors present in the blood plasma, such as hemoglobin, IgG fraction, heparin, etc.

The sensing chip used by Koo et al. [50] was functionalized by an amine group immobilizing C. burnetii DNA primers. These primers and the reverse primers simultaneously co-assembled with the recombinase enzymes. Thus, elongation of DNA and its detection occurred on the SMR structure. This is a winning strategy of sensing, since two problems are overcome: the low concentration of the pathogen and the effect of inhibitors in clinical samples.

The DNA samples of C. burnetii used by Koo et al. [50] were obtained from frozen formaldehyde-fixed paraffin-embedded tissue and frozen blood plasma specimens from patients with acute Q fever. The infectious agent C. burnetii was detected within 20 min, with a sensitivity of 81–93% and with a detection limit 10 times higher than that of the PCR method.

The same sensing strategy was applied in the improved version of the sensor, where a specific method for resonant wavelengths monitoring was used [52]. The optimization aimed to solve one of the main problems in diagnosing acute Q fever – to distinguish it from other disorders of fever of unknown origin (FUO). The optimization focused also on diagnosing Q fever in blood plasma specimens without the need of purification from various inhibitors, which was previously required for reliable C. burnetii detection. The specificity (89.5%) was demonstrated by the reliably proven differential diagnosis of Q fever and other febrile illnesses. The detection time for diagnosing Q fever (10 min) was faster than that of the previous sensor version.

The need for a rapid diagnostic test was the main motivation in developing a new modification of the SMR sensor [53]. The new approach consisted in the excitation of the sensory structures placed in the focus of a ball lens optical fibre. This eliminated the need for placing an optical fibre near the resonators, light tunnelling was not used, which increased measurement range and sensitivity. The light emitted from the sensing structure was gathered into the output ball-lens optical fibre connected to a photodiode. A signal acquisition process allowed for the automatic analysis of the detection of the resonant wavelength shift. The main advantage is a stable, rapid, automatic and user-friendly measurement of DNA. There are no differences in the methods of functionalization of the sensor surface and the sensor performance, as well as in their accuracy and selectivity. This modification of the SMR sensor was later integrated in an automated sample-to-answer diagnostic system that provides rapid and sensitive pathogen identification [54].

An electrochemical biosensor for Q fever was recently described [55]. This biosensor was based on the immobilization of CBU_1718 (GroEL) (UniProtKB – P19421 (CH60_COXBU) protein onto a gold electrode modified by self-assembled monolayer. Immobilization of the antigen on the gold electrode was optimized by varying several parameters such as protein concentration and incubation time to achieve best performance. GroEL was chosen for immobilization since it showed a higher yield in expression (0.65 mg/mL) compared to genetic region Com1. The sensor was treated with PBS, sera from blood donors and from patients with acute and chronic Q fever infections. The sensor showed the required discrimination ability allowing for accurate serological detection of chronic Q fever, which is not surprising, given the period of seroconversion.

SPR biosensors as a prospective serological test

Our research was conducted at National Centre of Infectious and Parasitic Diseases (NCIPD) and the Bulgarian Academy of Sciences (BAS), Sofia, Bulgaria. The aim of the study was to demonstrate development of the rapid, sensitive, quantification and real-time diagnostic test for detection of C. burnetii infection based on molecular interaction and formation of multimolecular complex ‘structural C. burnetii protein – specific anti - C. burnetii antibody’.

According to To et al. [56] and Varghees et al. [57], C. burnetii immunogenic antigens, such as Hsp60, Com1, Cbmip, P1 and AdaA provoke the anti-C. burnetii seroresponse. Chen et al. [58] reported that the immunodominant antigen Com1 may have the potential to improve the detection of Q fever specific antibodies.

For SPR assay elaboration recombinant C. burnetii 60 kDa chaperonin protein (Biorbyt, UK, Cat. No: orb1477544) from Hsp60 family which plays an essential role in assisting protein folding was immobilized on biochips. We provided immobilization via matrix-assisted pulsed laser evaporation (MAPLE) technique. The key point of this technique is the absence of molecules mediating the fixation of the recognition molecules to the transducer surface, which guarantees the highest possible detection specificity. A study by Dyankov et al. [59] gives details regarding the MAPLE technique and the parameters of the immobilization procedure. For antigen deposition a frozen solution target involving 5–7% protein in deionized water was used. The functionalization of the sensor biochip surface is critical for sensitivity enhancement and for reliability of antibody binding.

Antigen-biofunctionalized SPR biochips were incubated with specific anti-human polyclonal C. burnetii CBU_1314 antibody (cat. # DZ41196, Boster Biological Technology, Pleasanton, CA). As a stock solution, we used the one supplied by the manufacturer. Working concentrations of the stock solution were dissolved in fresh deionized water in the concentration range 1–12 µL/mL.

The first group of ten biochips was incubated for 20 min with polyclonal antibody at different concentrations. All SPR chips of the first group possessed well-expressed displacement of the plasmon resonance (Figure 1), as a result of antigen-antibody interaction. After the SPR chips were functionalized, the plasmon resonances were measured (dark curve in Figure 1). The spectral positions of these resonances were taken as a reference against which the spectral displacement due to the antigen–antibody interaction was registered (red curve in Figure 1).

Figure 1. The spectral displacement due to the bimolecular interaction ‘structural C. burnetii protein – specific anti-C. burnetii antibody’ registered against the functionalized biochip; black curve: the SPR signal of functionalized biochip; red curve: the SPR signal after antigen binding.

The binding property of C. burnetii antigenic protein to the specific anti-C. burnetii antibody was assessed. The spectral shift increased with increasing antibody concentration. Although this concentration dependence was well expressed, it was not convincing enough to draw quantitative conclusions about antibody concentration. The reason was that the detection was at the limit of the measurement error.

The second group of biochips were incubated with bovine serum albumin (BSA, cat. No 9048-46-8, Merck, Kenilworth, NJ) at different concentrations in the range of 1–20 µg/mL for the purpose of checking the specificity of the immobilized antigen. Spectral displacements were observed below the limit of detection in the chips from the second group. Thus, no shift due to biomolecular interaction was detected, indicating the high specificity of the immobilized antigen.

The SPR serology assay reaches detection limits in the low µg/mL range employing specific polyclonal anti-C. burnetii antibody. During acute Q fever, antibody levels start rising in the blood 5 d after infection and become detectable. Hence, SPR assay application follows the serological response of acute Q fever [60] (Figure 2).

Figure 2. Application of serological and molecular-biology methods for detection of acute Q fever.

Based on the preliminary results, achieved with the SPR-assay performance we have demonstrated a model for a new rapid test for identification and quantification of anti-C. burnetii antibodies, based on biosensor technology. We have implemented the biosensor assay with a standardized anti-C. burnetii polyclonal antibody and infectious structural protein, which allow further comparison with other serological assays such as ELISA and/or IFA in the following aspects:

• Study of the cross-reactivity with Legionella, Bartonella and other infectious agents that may occur2).

• Study of the possibility of diagnosing the acute from chronic phase of the Q fever disease by quantitative evaluation of the specific antibodies to antigens for phase 1 and phase 2.

SPR provides quantitative detection, and the sensitivity can reach a hundred femtomoles, as we have shown in our studies [61]. Given this, the application of SPR detection could be competitive with ELISA and IFA and its application for medical diagnosis is promising.

From a technological point of view, SPR detection is a rapid, more convenient, on-site applicable technique that does not require skilled personnel and is easy-to-use and cost-effective.

Conclusions

Our aim here was to review the recent advances in novel biosensor technologies used for diagnosis of C. burnetii in clinical samples. Early diagnosis not only increases the chance of rapid recovery, but also helps prevent the spread of infection. The biosensors based on SMR are the only ones available so far for detection of C. burnetii in blood samples. The lack of other types of sensors results from the aforementioned difficulties in Q fever detection: low levels of the infectious agents in clinical samples, long seroconversion period and cross-reactivity. The detection principle of the SMR biosensor is very effective, since it detects part of bacterial DNA (no need for high biosafety level laboratory), the detection is based on the combination of IDA and a sensitive DNA amplification that overcomes the problem of the low levels of the infectious agents. RPA used as an isothermal assay decreases the detection time up to 10 min, which allows for real-time and quantification detection. Compared to the PCR assay, the reported sensitivity of the SMR sensor is higher, while the specificity is comparable. We developed an SPR biosensor based on structural C. burnetii antigen directly immobilized on the SPR transducer. Our experiments confirmed that the immobilized infectious antigen bind to used specific anti-C. burnetii antibodies. While immobilization methods involving an SMR sensor surface and the combination of IDA have proven to be a winning strategy, the performance of SPR as compared to ELISA in terms of selectivity and sensitivity remains to be proven.

Author contributions

PGK and GD, conceptualized and designed the study, wrote the original draft, reviewed and supervised the manuscript; contributed to funding; KS, TV and SK contributed to the study design, experimental studies, interpretation of data and contributed to funding acquisition; PF contributed to manuscript review and editing the data analysis. MB and IT contributed to the clinical studies, manuscript editing, literature research and contributed to funding acquisition. All authors have read and approved the final version of the manuscript and agree to be equally accountable for the integrity of the entire study.

Disclosure statement

The authors declare no conflict of interest.

Data availability statement

The data that support the findings from this study are available from the corresponding author [PGK] upon reasonable request.

Additional information

Funding

This work was supported by the Bulgarian National Science Fund under Grant number KP-06-N33/3, from 13 December 2019, entitled: ‘Molecular-genetic identification and creation of an archive genomic bank of the circulating human and animal C. burnetii genotypes and determination of their role as particularly dangerous infectious agents causing epidemiological outbreaks on the territory of Bulgaria’ and by the Bulgarian Ministry of Education and Science (MES) in the frames of Bulgarian National Recovery and Resilience Plan, Component ‘Innovative Bulgaria’, the Project № BG-RRP-2.004-0006-C02 ‘Development of research and innovation at Trakia University in service of health and sustainable well-being’.