Introduction

In addition to the changing environmental factors, more and more harmful effects on the human body caused by bacteria and viruses appear. Fortunately, pharmaceutical research and modern medicine had undergone significant changes by the end of the twentieth century. In addition to advances in medicine, after introduction of antibiotic therapies, child mortality decreased and contributed to an increase in life expectancy [1,2,3,4,5,6,7,8,9,10,11,12,13][14]. After the discovery of penicillin, despite the initial successes, it had become known that due to the genetic mutation ability of bacteria, a chance of survival against the available antibiotics is offered to some mutant strains due to moderate antibiotic sensitivity and resistance [7]. Long before the discovery and use of antibiotics, gene mutations responsible for natural resistance to certain classes of antibiotics already existed in nature [1, 6,7,8, 15,16,17].

Due to the immunization of the body to microorganisms and the fight against pathogenic bacteria, development of new antibiotics is essential due to moderate sensitivity and the emergence of new multidrug resistant strains. While the twentieth century was called the era of the discovery of antibiotics, in the following decades, only existing antibiotics were modified [2, 18,19,20].

There are efforts to set stricter application conditions for the further usability of existing antibiotics. If antibiotics are applied properly, we can contribute to reducing the development of multidrug resistant strains. Unfortunately, in the case of an unknown pathogen, current clinical protocols use combined broad-spectrum antibiotic treatments “blindly,” often unnecessarily, at the start of treatment in order to control sepsis caused by bacteria.

In addition to the stricter application of human medicine, many forums have tried to regulate the uncontrolled use of antibiotics in the crop and livestock sectors [8], which indirectly contributes to the mutation of bacteria.

The treatment of septic symptoms is mainly determined by the clinical appearance and the microbiological culture of the pathogens based on the laboratory results. Prompt, aggressive antibiotic therapy and appropriate surgical treatment are essential to ensure a good prognosis. Despite rapid diagnosis and treatment, there is still a high mortality rate [21,22,23], which in many cases can be decisively influenced by appropriately chosen antibiotic therapy.

Based on some studies, it would be an important criteria that, if possible, the patient should not receive systemic antibiotics before sampling [24,25,26,27] and that the patient only receives the therapeutic or prophylactic dose of antibiotics afterward [28].

One of the main goals could be a fast, reliable method to identify bacterial strains and drug resistance during surgery. In the early management of periprosthetic joint infections, under certain criteria, there is a chance for preserving the implant by applying DAIR protocol (debridement, antibiotics and implant retention), followed by prolonged antibiotic treatment. However, this therapeutic approach has a relatively high failure rate, reported between 10% and 40%. It is known that Staphylococcus aureus or coagulase-negative staphylococci (CNS) are accounted for two-third of all PJI cases and susceptibility to methicillin (methicillin-resistant Staphylococcus aureus [MRSA] and methicillin-sensitive Staphylococcus aureus [MSSA]) could definitely improve the efficacy of antibiotic management [29].

In order to increase effectiveness, Trampuz et al. recommend the sonication procedure during which, after the removal of the implant, the bacteria are extracted from the biofilm in an ultrasonic sonicator, which is later examined by microbiological methods [30, 31].

In order to reduce periprosthetic joint softening, under the leadership of Deirmengian and Parvizi, the alpha defensin protein was recommended to be subjected to further tests in the diagnosis of periprosthetic joint infection [32]. As a development of this work, a rapid test is available that can also be used in clinical practice (Synovasure® Alpha Defensin Test), which clearly shows αD positivity after 10 min [26, 27].

Despite efforts to reduce drug resistance, the identification and isolation of the causative agent and the early use of targeted antibiotic therapy are still crucial for the success of treating bacterial infections.

Objective: We aimed to demonstrate the practical utility of differential scanning calorimetry (DSC) used in our research project [9,10,11] to isolate the pathogen bacteria in joint and periprosthetic infections.

Materials and methods

In accordance with the norms of the responsible Regional and Institutional Research Ethics Committee (number of permission: 6737) and the Helsinki Declaration of 1975, revised in 2008, using our previous research results [9,10,11], in order to verify and isolate the identity of the most common pathogens in locomotor septic pathologies. The research was performed in the DSC laboratory of the Institute of Biophysics, University of Pécs, in the frame of our interdisciplinary working group.

When creating the septic joint model, the sterilely collected synovial fluid samples were permanently stored at − 70 °C under experimental conditions, taking into account the physiological environment and basic reproduction conditions of the bacteria. On the day the experimental model was assembled, taking into account the rules of sterility and the risk of cross infection, during septic musculoskeletal infections, we monitored the proliferation of bacteria in isothermal mode at 37 °C. We used the most common bacteria (Escherichia coli, MRSA (methicillin-resistant Staphylococcus aureus), Pseudomonas aeruginosa, Staphylococcus aureus, Staphylococcus epidermidis and Streptococcus pyogenes [10, 12]) under in vitro conditions, by superinfecting the synovial fluid with a concentration of 103 or 105 CFU mL−1. Depending on the tested concentration (103/105 CFU mL−1), in the case of the used bacterial strains, their heat production was recorded during their reproductive cycle, and the maximum of the heat rates was monitored [10, 11].

Similarly to our previous research project, synovial samples were collected in cases without arthrosis and, in cases of different degrees of arthrosis, based on the modified Kellgren–Lawrence classification [13] (Table 1). Then, we examined and analyzed the factors influencing the reproduction of bacterial strains with isoperibol calorimetry, all for unequivocal test results. The reproductive cycle of bacterial strains that occur more often in septic locomotor pathologies was investigated with the SETARAM Micro DSC-II calorimeter, based on the chosen experimental model, using SETARAM produced conventional and mixing batch vessels (see Fig. 1).

Fig. 1
figure 1

For the denaturation (left) as well as for the proliferation test used, specially designed mixing cells, in closed and open (after mixing) positions (SF in the lower, bacteria in the upper compartment)

Full size image

Results

Denaturation results of the examined synovial samples

Similar to our previous results [9,10,11], the denaturation curves of the synovial samples are influenced by the progression of the degree of arthritis, their chemical composition and primarily their protein content (see Fig. 2). As arthrosis progresses, an important parameter change can be detected, which can be seen in the change in the half-value width (T1/2) of the denaturation curve (see Table 2), which is the temperature range of the structural transformation at half of the maximum heat flow. (It gives information about the cooperativity among the different thermal domains being the constituent elements of the tested sample.)

Fig. 2
figure 2

The denaturation scans of different grades of synovitis (the curves are average of the independent experiments)

Full size image

We performed a deconvolution of the scans too, because on the basis of the shape and “shoulders” of the curves it is obvious the presence of a multidomain structure. (We have taken into consideration the components with highest contribution to the total enthalpy, see Fig. 3 and Table 3.)

Fig. 3
figure 3

Deconvoluted DSC scans of SF with different severity (Gr.0-Gr.IV.). Colors in figures are; black: experimental average curve, red: sum of the individual transitions, blue stands for the main contributors of denaturation (1st-albumin, 2nd and 3rd to α1-α2, β and γ globulins [33,34,35]). Endotherm deflection upwards

Full size image

We have checked the effect of more aggressive bacteria on the denaturation behavior of the infected SF too. After the mixing, we have waited the end of the proliferation, and in the same arrangement, we denatured the infected SF. The results are shown in Fig. 4 and Table 4, after deconvolution in Fig. 5 and Table 5.

Fig. 4
figure 4

Denaturation scans of infected SF samples. Endotherm deflection downward

Full size image
Fig. 5
figure 5

Deconvoluted DSC scans of SF infected with different bacteria. Symbols in figures are: A experimental average curve, B sum of the individual transitions, the C-D-E curves stand for the main contributors of denaturation (C-albumin, D and E to α1-α2, β and γ globulins [33,34,35]). Endotherm deflection downward

Full size image

Results of the proliferation cycle of the tested bacteria

After reaching the thermal equilibrium between the sample mixing vessel (filled with the tested bacterial strain and the separated medium synovial fluid) and the reference vessel (in both compartment the same amount of SF), after the mixing, the heat released during the proliferation process of the bacterial strains was recorded as a function of the time elapsed, which at the tested concentrations (103–105 CFU mL-1) was characteristic of the given bacterial strain, had different time courses as well as different total heat production (see Table 6) [9,10,11,12].

The maximum heat flow time and the calorimetric enthalpy are bacteria-dependent parameters. The run of the scans prohibited a multistep proliferation, so we performed a deconvolution (see Figs. 6 and 7, and Table 7 and 8). In the tables we took into consideration only three main steps in the heat production. (The smaller contributions which were in the vicinity of the detectable limit of the instrument were neglected in the repeated deconvolution.)

Fig. 6
figure 6

Isoperibolic heat flow scans of less effectively proliferating bacteria

Full size image
Fig. 7
figure 7

Isoperibolic heat flow scans of more effectively proliferating bacteria

Full size image

During the calorimetric identification of most frequently infecting bacteria, the time elapsed to reach the maximum of the proliferation heat flow could be a monitor, as shown in Table 9. After mixing the proliferation starts maximum in 18–20 min. The quickest proliferation’s peak is around 3 h 30 min, the smallest at 12 h 40 min.

Discussion

The aim of our current study was to demonstrate the applicability of thermal analysis in evaluation of the grade of osteoarthritis or the infection of knee joint using synovial fluid samples. Our primary goal was to gain reliable information about the bacterial growth in a relatively short time.

Based on the literature, two positive cultures or the presence of a sinus tract are considered as major criteria for diagnosis of periprosthetic joint infection [36]. While tasking samples with swabs are a widely used practice, they are less reliable and nowadays more sophisticated methods are recommended for detections of pathogens [37]. Generally, intraoperatively collected tissue samples, sonication and synovial fluid should be utilized instead of swabs.

Besides standard culturing techniques, matrix-assisted LASER desorption/ionization time-of-flight (MALDI-TOF) spectroscopy and an automated antimicrobial susceptibility testing systems can be used to identify pathogens and test antibiotic sensitivity [38].

While the established microbiological tests are still the “gold-standards” in the detections of pathogens, these methods are also time-consuming. However, in case of treatment of periprosthetic infections, rapid diagnosis and availability of antibiotic sensitivity are essential. Therefore, one of the greatest advantages of thermal analysis that it can be performed in a few hours, instead of days.

The early diagnosis of osteoarthritis would be a key factor to prevent further progression of these degenerative diseases. Beside the conventional imaging methods, such as X-rays, CT-scan and MRI, only a few alternative tests are found in the literature for the detection of a developing osteoarthritis. To find a reliable biomarker to diagnose the osteoarthritis in the early phase, Ab Aziz et al. have recently utilized attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy to test serum samples of patients with symptomatic primary knee osteoarthritis [39]. Another possible target could be the biochemical analysis of synovial fluid, due to the inflammatory processes and oxidative stress occurring in arthritis [40] (Table 1).

Table 1 Description of Kellgren–Lawrence and the modified Kellgren–Lawrence classification systems [13]
Full size table

In regard of osteoarthritis samples, Fig. 2 and Table 2 clearly demonstrate that the alteration of DSC scans is a good fingerprint of the severity of disease. The T1/2 increases in all stages (the biggest is in case of Grade II, similar the change in calorimetric enthalpy) compared to the healthy control, demonstrating, that the cooperativity among the fluid components decreases in all cases [9,10,11].

Table 2 Denaturation parameters of synovial fluids in different severity of arthritis
Full size table

During the denaturation, the actual stage of SF proteins is the key factor. In a healthy SF, it is in the range of 9–22 g L-1, but in inflammatory stage, it can reach 44 g L-1. Roughly 60% of the protein content is albumin, the others are globulins, proteoglycans and hyaluronic acid [41,42,43]. The denaturation scans of SF exhibit a good similarity to the blood plasma curves [33, 44,45,46,47,48,49], where 4–5 thermal domains can be separated by deconvolution of scans. The albumin gives the biggest contribution. Generally, the enthalpy resolution of a commercial DSC equipment is in the range of 5%; therefore, we followed the Briere method [34], where in case of SF only three main compound was used (see Fig. 3, where the fitting was better than 95%). Table 3 shows that the albumin contribution (Tm) was the biggest one, and exhibited a severity dependence, compared to the control.

Table 3 Maximum denaturation temperature (Tm) values of the three main overlapping transitions (made by deconvolution) and relative contribution of each one to the total enthalpy (using average DSC scans of different synovial fluids with altering degree of severity)
Full size table

Those SF samples which undergone to bacterial infection produced strong bacterial dependence (see Fig. 4 and Table 4). The melting temperature and calorimetric enthalpy exhibit a fingerprint character. Having a big DSC scan pool, with the aid of a good informatic database, it could be used as a quick test of the type of infecting bacterium (within a few hours scanning the Tm can be reached for the identification). The deconvolution of the DSC curves (see Fig. 5 and Table 5) also demonstrate the shifting of denaturation temperature of the biggest compound albumin to the higher range, representing a strong structural rearrangement.

Table 4 Denaturation parameters of SF after inoculation with different bacteria
Full size table
Table 5 Maximum temperature (Tm) values of the three main overlapping transitions (made by deconvolution), and their relative contribution to the total enthalpy
Full size table

The proliferation experiments performed with the six more frequently infecting bacteria exhibited also the ability to use the isoperibol technique to identify the infecting agent. On the basis of Table 6, we can see that the smallest calorimetric enthalpy is produced by Streptococcus pyogenes, but it quickest to reach the maximum heat flow, while the biggest ΔH by Pseudomonas aeruginosa. Two groups can be separated in the bacterium line: moderately proliferating (see Fig. 6 and Table 7) as well as more effectively proliferating (on the basis of total calorimetric enthalpy) bacteria (see Fig. 7 and Table 8). This can be also a relatively quick method to identify the bacterium and start the antibiotic treatment.

Table 6 Thermal proliferation characteristics of bacteria causing frequent joint infections
Full size table
Table 7 The proliferation time maxima (tms) and relative abundance (in %, higher than 5%, rounded to one decimal) to the total enthalpy of the deconvoluted curves in case of the bacteria that produced the smallest calorimetric enthalpy during the proliferation
Full size table
Table 8 The proliferation time maxima (tms) and relative abundance (in %, rounded to one decimal) to the total enthalpy (higher than 5%) of the deconvoluted curves in case of the bacteria that produced the highest calorimetric enthalpy during the proliferation
Full size table

Table 9 presents data for a further identification of bacteria. To reach the maximum heat flow during the proliferation Streptococcus pyogenes leads, and the slowest is Escherichia coli. It is interesting that in case of Streptococcus pyogenes, Staphylococcus epidermidis and Pseudomonas aeruginosa at 105 CFU mL-1 infecting concentration the proliferation heat flow reaches the maximum value later than in case of 103 CFU mL-1. Comparing the synovial media with similar but still varying protein content, we did not notice any significant differences in the bacterial reproduction cycle, the chemical composition of the synovial samples mainly did not affect the initial reproduction kinetics of the bacteria. The sharp decrease in the growth rate of the tested bacteria can be influenced by the depleting culture medium level.

Table 9 Ranking of the bacteria according to the time of reaching the maximum heat rate
Full size table

During the analysis of the reproduction cycle of bacterial strains, when differentiating the different proliferation kinetics, in some cases, sample superinfection also arises, which results in the heat flow an irregular run, the sum of the heat flow can be produced by two or more bacteria during a unit of time. With the present investigation, despite the sterility and the strict treatment of the bacterial strains, we could not confirm or exclude possible superinfection of the samples.

Conclusions

Given that the number of antibiotics discovered and used in clinical practice in recent decades is significantly exceeded by the number of resistant strains that have developed [28], we need a significant strategic change in approach in order to be effective in treatment with the existing set of antibiotics. Despite the high sensitivity and specificity of the biomarkers, antibiotic sensitivity and resistance remain an open question, for the detection of which the isoperibolic calorimetric test is a reliable tool, based on the work of Trampuz and Braissant [50, 51] and our research.

Until now, in clinical practice, we have been able to achieve effective treatment based on the combined use of serological markers, synovial tests and microbiological vaccinations [28]. However, it could be revolutionized by utilizing antibiotic sensitivity specified with microcalorimetrical test.

Our experimental series using DSC and isoperibolic calorimetry proves that it is an excellent method for verifying the identity and reproductive capacity of the pathogen. In future studies, the effectiveness of broad-spectrum antibiotics could be tested in our experimental settings in a relatively short time duration.