Introduction

Gold nanoparticles (AuNPs) have attracted great attention and show many distinct physical and chemical properties, such as localized surface plasmon resonance (LSPR) effect [1], photothermal effect [2], good electrical conductivity [3], huge surface-to-volume ratio [4], ease of biofunctionalization [5, 6] and so on. Taking advantage of LSPR effect of AuNPs, a number of label-free colorimetric biosensors have been developed [7,8,9,10]. AuNPs based colorimetric assays are becoming conventional systems for bioanalysis [11,12,13,14,15,16]. Nevertheless, there is still much room for improvement and innovation by further combining other nanomaterials or developing novel biomolecule binding events on the nanostructured metallic surface.

Theophylline belongs to methylxanthine family, which is considered as a crucial molecule for physiological processes in human metabolism [17]. It has been commonly used in the clinic for the treatment of asthma, chronic obstructive pulmonary disease (COPD) and neonatal apnea [18]. However, the serum therapeutic range is quite narrow. High concentrations of theophylline may cause lethal or neurological damages [19]. Therefore, there is an urgent need to develop convenient and fast detection methods to monitor theophylline level. So far, great efforts have been made [20,21,22,23]. For example, Stone’s group immobilized two DNA probes on the surface of AuNPs, which was linked by the aptamer against theophylline. The AuNPs aggregates were disassociated when exposed to theophylline. By monitoring the peak and intensity of UV-vis absorption spectrum, the concentration of theophylline was determined [24]. Zhang and coworkers combined the fluorescence quenching ability of AuNPs and RNA aptamer based theophylline recognition. The labelled fluorophore on the RNA was localized on the surface of AuNPs after the formation of RNA aptamer-theophylline complex. The decreased fluorescence was used to represent initial theophylline concentration [25]. Although current methods may meet the requirement of high sensitivity, the detection procedures of most methods are always complicated. In addition, expensive materials (e.g., modified nucleic acids) and instruments (e.g., fluorescence spectrophotometer) may be required.

In this contribution, we have succeeded the development of a novel method for rapid detection of theophylline. The employed recognition element of split RNA aptamers are unmodified, which makes the detection method cost-effective and convenient. On the other hand, non-crosslinking AuNPs aggregation is involved. Compared with commonly used crosslinking AuNPs aggregation, the reaction is much succinct and faster. In addition, the method is sensitive and selective. The presence of theophylline also can be determined with bare eyes. Therefore, it meets the requirements of point-of-care testing (POCT). Furthermore, the colorimetric strategy is believed to be applicable to other aptamer-based detection methods by altering the nucleic acid sequences on the AuNPs.

Experimental

Materials and instruments

Gold (III) chloride trihydrate (HAuCl4•3H2O), trisodium citrate, diethypyrocarbonate (DEPC), ethylenediaminetetraacetic acid (EDTA) were purchased from Sigma (http://www.sigmaaldrich.com/). Theophylline, caffeine, theobromine, guanine and adenine were obtained from Solarbio Science &Technology Co., Ltd. (http://solarbio.bioon.com.cn/). Human serum samples were collected from local hospitals with the approval by the Medical Ethics Committee (Suzhou, China). All other chemicals were of analytical grade and used as received. Double-distilled water was used to prepare all solutions. The resistance was 18 MΩ cm and the water was also treated with DEPC. RNA probes were synthesized and purified by Takara Biotechnology Co., Ltd. (http://www.takarabiomed.com.cn/). The sequences were listed as follows:

  • RNA probe 1: 5’-AAAGGCGAUACCAGCCGAAA-3’

  • RNA probe 2: 5’-GGCCCUUGGCAGCGUCAAAA-3’

UV-vis absorption spectra were measured by a Synergy HT multifunction microplate reader (BioTek Instruments, Inc., USA). Transmission electron micrographs were recorded by a FEI Tecnai G20 transmission electron microscope (TEM) (FEI Company, USA). Dynamic light scattering (DLS) characterization of AuNPs were performed using a Zetasizer Nano ZS90 (Malvern Instruments, UK). Photographs were taken with a mobile phone (ZTE Blade A2 Plus).

Optimization of NaCl concentration

Freshly prepared AuNPs were kept stable for at least 3 months. However, after the introduction of certain amount of salt like NaCl, non-crosslinking aggregation of AuNPs occurred immediately. The color and UV-vis absorption spectrum were changed accordingly. The concentration of NaCl was a critical parameter, thus it was optimized. First, standard solutions of NaCl with a series of concentrations were prepared. Next, 0.5 mL of concentrated AuNPs were mixed with 0.5 mL of NaCl with various concentrations. UV-vis absorption spectra of the mixtures were then measured from 350 to 700 nm. The absorbance values at 520 nm and 650 nm were recorded (A 520 and A 650). The ratios of R 650/520 (A 650/A 520) were calculated, which were related to the NaCl concentration.

Colorimetric analysis of theophylline

AuNPs were mixed with RNA probe 1 (5 μM) and RNA probe 2 (5 μM), respectively. The two solutions (AuNPs-R1 and AuNPs-R2) were then purified by centrifuging at 10, 000 rpm for 20 min to remove excess RNA, respectively. Next, 245 μL of AuNPs-R1 and 245 μL of AuNPs-R2 were blended. The mixture was then treated with 10 μL of theophylline with various concentrations. Subsequently, UV-vis absorption spectra were measured to check the salt-tolerances of the solutions in the presence of NaCl (70 mM). Specifically, the absorbance values at 520 nm and 650 nm were obtained for the calculation of R 650/520. The relationship between this parameter and the concentration of initial theophylline was then studied.

Interference investigation and serum sample analysis

To verify the selectivity of this colorimetric method, interfering molecules such as caffeine, theobromine, guanine and adenine with similar chemical structures were employed. Some common biomolecules like bovine serum albumin (BSA) and duplex DNA were also checked. These molecules interacted with AuNPs-R1 and AuNPs-R2 instead of target theophylline. Briefly, 490 μL of the mixture of AuNPs-R1 and AuNPs-R2 was blended with 10 μL of the interferents with the final concentration of 100 μM. After treated with 70 mL NaCl, corresponding UV-vis absorption spectra were recorded. R 650/520 values were calculated, which were then compared with that of theophylline (10 μM).

To examine the utility of this colorimetric method in complex biological fluids, human serum samples were challenged for theophylline analysis. Two independent serum samples were collected and diluted for 10 times. After that, 5, 10, 15 μM theophylline were spiked, which were then detected in the colloid system in the presence of 70 mM NaCl. The concentrations of theophylline were obtained and recoveries and relative standard deviations (RSDs) were calculated.

Results and discussion

Working principle

Schematic illustration of the working mechanism for the detection of theophylline is shown in Scheme 1. A 33-mer sequence of the theophylline aptamer is splitted into two shorter fragments, which are made up to 20-mer (RNA probe 1 and RNA probe 2) [26]. The binding property is not affected. The two split RNA aptamers are able to remain stable during the experimental process with shorter RNA fragments. Single-stranded RNA probes interact with AuNPs via their exposed nitrogen-containing bases, which can protect AuNPs from salt-induced non-crosslinking aggregation [27]. Therefore, the mixture of AuNPs-R1 and AuNPs-R2 shows high salt-tolerance and the UV-vis absorption spectra barely change after treated with NaCl of relatively high concentration. However, in the presence of target theophylline, RNA probe 1, RNA probe 2 and theophylline self-assemble into an intact RNA aptamer-theophylline complex through a substantial change in the secondary structure. As a result, without the protection of RNA probes, AuNPs cannot resist the salt-induced aggregation any more, which is reflected by the color changes and UV-vis absorption spectra. A colorimetric strategy for the detection of theophylline is thus established.

Scheme 1
scheme 1

Illustration of colorimetric detection of theophylline based on RNA aptamers recognition and non-crosslinking AuNPs aggregation

Full size image

Salt-induced non-crosslinking AuNPs aggregation

TEM images are taken to characterize the size and distribution state of AuNPs. Bare AuNPs disperse well in water with the diameter around 13 nm (Fig. S1a). The negatively surface charge of bare AuNPs is due to the capped trisodium citrate (Fig. S2). However, if salt is added to the colloid system, the outer shell of the electrical double layer of AuNPs is compressed. The electrostatic repulsions among AuNPs are thus decreased, which leads to the non-crosslinking aggregation (Fig. S1b) [28]. The sizes of dispersed and aggregated AuNPs are well distinguished by DLS characterizations (Fig. S3). Moreover, since LSPR of AuNPs is strongly correlated with their distribution state, UV-vis absorption spectrum may also be used to indicate the changes. As shown in Fig. 1, bare AuNPs have an obvious absorbance peak at 520 nm (curve a). If the nanoparticles aggregate after adding salts, the peak at 520 nm decreases sharply and the absorbance at 650 nm increases significantly (curve b). The phenomenon is in good accordance with previous reports [29]. Nevertheless, with the existence of RNA molecules like RNA probe 1 on the surface of AuNPs, salt-induced aggregation can be effectively inhibited even with high levels of salt. The UV-vis absorption spectrum is restored to that of AuNPs without salt (curve c). The results indicate that by monitoring the changes of UV-vis absorption spectra, the salt-tolerance of AuNPs adjusted by RNA on their surface can be well determined.

Fig. 1
figure 1

UV-vis absorption spectra of a AuNPs, b AuNPs with NaCl, c AuNPs-R1 with NaCl. The concentration of AuNPs is 5 nM and the concentration of NaCl is 70 mM

Full size image

Preliminary experiments

Since both of the absorbance values at 520 nm and 650 nm change, we have used the value at 650 nm to that at 520 nm (R 650/520) to reveal the distribution state of AuNPs [30]. We have obtained the R 650/520 values of bare AuNPs in the presence of NaCl with a series of concentrations. With the increase of NaCl, the AuNPs aggregate more drastically. As a result, the larger R 650/520 value will be (Fig. 2). The relationship between NaCl concentration and R 650/520 is fitted with a Boltzmann sigmoid equation as follows:

$$ y=0.1156+0.8983/\left(1+\exp \left(x-{x}_0\right)/ dx\right)\Big)\ \left({x}_0=51.7018, dx=4.1820,{R}^2=0.9961\right) $$

in which y is R 650/520, x is the NaCl concentration (mM). When the AuNPs were treated with NaCl of the concentration of x 0 (51.7018 mM), the slope of the calibration curve is the largest. In another word, R 650/520 is most sensitive at this point to the changes of NaCl concentration. In addition, when NaCl concentration is increased to 70 mM, the R 650/520 value reaches a plateau, which is used as the optimized NaCl concentration in the following experiments. The concentration of RNA probes used to protect AuNPs from salt-induced aggregation is also investigated. The optimized concentration is 5 μM after comparing the values of R 650/520 (Fig. S4).

Fig. 2
figure 2

The relationship between R 650/520 of bare AuNPs and the concentration of added NaCl. Error bars represent standard deviations of three measurements

Full size image

Color changes

We have observed the color changes of AuNPs after different treatments. AuNPs before and after mixing with RNA probes or further theophylline exhibit wine-red color, indicating that AuNPs disperse well in water (Fig. 3a–e). After further incubating with NaCl, bare AuNPs aggregate and the color turns to blue (Fig. 3f). In the presence of RNA probe 1, the AuNPs color does not change, demonstrating the excellent salt-tolerance of RNA protected AuNPs (Fig. 3g). However, in the colloid system of AuNPs-R1 and AuNPs-R2, theophylline is able to remove the RNA probes on the surface of AuNPs, forming RNA aptamer-theophylline complex. The remained AuNPs cannot resist salt-induced aggregation any more, thus the color turns to blue (Fig. 3h). Since the aptamer is splitted into two fragments, none of the RNA probe can interact with theophylline alone. AuNPs-R1 or AuNPs-R2 colloid system does not respond to theophylline, and the solution colors remain wine-red (Fig. 3i–j). These color changes demonstrate that the strategy works well.

Fig. 3
figure 3

Solution colors of a AuNPs, b AuNPs-R1, c mixture of AuNPs-R1, AuNPs-R2 and theophylline, d AuNPs-R1 and theophylline, e AuNPs-R2 and theophylline. f–j are the corresponding top cases after adding NaCl. The concentration of AuNPs is 5 nM; the concentration of NaCl is 70 mM; the concentration of theophylline is 50 μM

Full size image

Sensitive detection of theophylline

Standard theophylline solutions with a series of concentrations are prepared which are added to the mixture of AuNPs-R1 and AuNPs-R2. Then, the solutions are treated with NaCl with the concentration of 70 mM. UV-vis absorption spectra are recorded which are shown in Fig. 4a. With the increase of theophylline level, more RNA probes are stripped, and the salt-tolerance of AuNPs decreases, which can be reflected by the decrease of A 520 and the increase of A 650. The calibration curve representing the relationship between R 650/520 and theophylline is depicted in Fig. 4b and the linear relationship is established between 0.1 and 20 μM. The linear fitting equate is as follows:

$$ y=0.1093+0.0439\ x\ \left(n=3,{R}^2=0.9952\right) $$

in which, y stands for R 650/520, x is the concentration of theophylline (μM). Previous report has indicated that the safe theophylline concentration in the plasma of medicinal theophylline users is no larger than 111 μM [31]. The range of this method is quite acceptable for theophylline assay in real samples. Moreover, a quite low limit of detection (LOD) is calculated to be 0.067 μM. The analytical performances are excellent compared with most previously reported methods (Table S1).

Fig. 4
figure 4

a UV-vis absorption spectra of the mixture of AuNPs-R1 and AuNPs-R2 after treated with different amount of theophylline in the presence of NaCl. b Calibration curve of theophylline detection with a series of concentrations. Inset shows the linear relationship between R 650/520 and the concentration of theophylline. Error bars represent standard deviations of three measurements

Full size image

Selectivity investigation

To investigate the selectivity of the theophylline detection method, we have employed several molecules with similar chemical structures as interferents including caffeine, theobromine, guanine, and adenine. We also compared the response of this method to common biomolecules like BSA and duplex DNA. The R 650/520 values in the cases of 10 μM theophylline and 100 μM interferents are compared in Fig. 5. Only in the presence of target theophylline, significant increase R 650/520 value is observed. Caffeine, theobromine, guanine, adenine, BSA and DNA with 10 times higher concentrations only contribute limited R 650/520 values. The results demonstrate the high selectivity of the method.

Fig. 5
figure 5

Selectivity of theophylline measurement (10 μM) over other interfering molecules (100 μM)

Full size image

Determination of theophylline in human serum samples

To demonstrate the practical application of the present colorimetric method, the concentrations of theophylline are determined in human serum samples. The samples are firstly diluted for 10 times and different amount of theophylline are added to the solutions. Then, the samples are measured by the method. By comparing with the standard curve achieved in Fig. 4, the concentrations of added theophylline are calculated. As listed in Table 1, the values are quite consistent. In addition, the recoveries are among 99 and 107%, and all RSDs are less than 5%, verifying the method may have good potential utility in clinical screening applications. One can also distinguish the existence of theophylline in serum samples with bare eyes (Fig. S5).

Table 1 Detection of theophylline in serum samples
Full size table

Conclusions

In summary, we have successfully fabricated a novel detection method for theophylline analysis with high sensitivity. Non-crosslinking AuNPs aggregation phenomenon is involved in the strategy and the presence of target theophylline can be distinguished even visually. The satisfactory linear concentration range makes the method suitable for therapeutically relevant theophylline levels application. Although the introduction of RNA molecules may demand the use of RNase-free water and pipette tips, the split RNA aptamers promise the high selectivity. The method does not respond to interfering molecules like caffeine even with quite high concentrations. Moreover, the operation is quite convenient. The detection procedure only involves mixing the split aptamers, AuNPs, samples and NaCl solution. Thus, a convenient protocol is presented for fast and cost-effective detection of theophylline in serum samples. In addition, the strategy can also be expanded to other analytical applications of RNA aptamers and AuNPs.