- Email: [email protected]
Surface-enhanced Raman spectroscopic detection of Bacillus subtilis spores using gold nanoparticle based substrates
Analytica Chimica Acta 707 (2011) 155–163
Contents lists available at SciVerse ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Surface-enhanced Raman spectroscopic detection of Bacillus subtilis spores using gold nanoparticle based substrates Han-Wen Cheng, Yuan-Yuan Chen, Xin-Xin Lin, Shuang-Yan Huan ∗ , Hai-Long Wu, Guo-Li Shen, Ru-Qin Yu ∗ State Key Laboratory for Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, China
a r t i c l e
i n f o
Article history: Received 17 July 2011 Received in revised form 3 September 2011 Accepted 8 September 2011 Available online 16 September 2011 Keywords: Surface-enhanced Raman scattering Gold nanoparticles Biomarker Calcium Dipicolinate Bacillus subtilis
a b s t r a c t The detection of bacterial spores requires the capability of highly sensitive and biocompatible probes. This report describes the findings of an investigation of surface-enhanced Raman spectroscopic (SERS) detection of Bacillus subtilis spores using gold-nanoparticle (Au NP) based substrates as the spectroscopic probe. The SERS substrates are shown to be highly sensitive for the detection of B. subtilis spores, which release calcium dipicolinate (CaDPA) as a biomarker. The SERS bands of CaDPA released from the spores by extraction using nitric acid provide the diagnostic signal for the detection, exhibiting a limit of detection (LOD) of 1.5 × 109 spores L−1 (or 2.5 × 10−14 M). The LOD for the Au NP based substrates is quite comparable with that reported for Ag nanoparticle based substrates for the detection of spores, though the surface adsorption equilibrium constant is found to be smaller by a factor of 1–2 orders of magnitude than the Ag nanoparticle based substrates. The results have also revealed the viability of SERS detection of CaDPA released from the spores under ambient conditions without extraction using any reagents, showing a significant reduction of the diagnostic peak width for the detection. These findings have demonstrated the viability of Au NP based SERS substrates for direct use with high resolution and sensitivity as a biocompatible probe for the detection of bacterial spores. © 2011 Elsevier B.V. All rights reserved.
1. Introduction There are different types of bacteria, some of which constitute different pathogens to various diseases. One of the dangerous pathogens for the disease anthrax is Bacillus anthracis. This pathogen is a Gram-positive spore-forming and rod-shaped bacterium, and can be grown in an ordinary nutrient medium under aerobic or anaerobic conditions. Humans may be infected by eating or inhaling anthrax spores, breathing in material contaminated with the bacteria, or through skin exposure to infected animals. Anthrax infection is usually diagnosed by identifying the bacteria within skin blisters, blood, or other body fluids. The anthrax bacterium is also a potential biological weapon. Therefore, the rapid and accurate detection of the bacteria is very important for human health. Traditionally, PCR method has been used for bacterial identification in such pathogens. This method has the ability to discriminate between B. anthracis and other Bacillus species. However, the relative lengthy time in analysis for PCR method makes it difficult for situations requiring quick identification of the bacterial
∗ Corresponding authors. Tel.: +86 73188821916. E-mail addresses: [email protected] (S.-Y. Huan), [email protected] (R.-Q. Yu). 0003-2670/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2011.09.007
spores. There is an increasing need to develop rapid and sensitive methods for the detection of the bacterial spores. Among different methods for detecting bacterial spores, spectroscopic methods have been increasingly demonstrated as powerful tools for rapid and sensitive chemical or biological detection, especially with use of metal nanoparticles as detection probe. For example, the use of DNA-anchored gold nanoparticles has been demonstrated for colorimetric detection of the DNA from the infecting organism B. anthracis [1]. Because of nanoparticle-based enhancement effect on Raman scattering intensity, surfaceenhanced Raman scattering (SERS) technique has been widely utilized for the detection of biomolecules including proteins, DNAs, and bacteria [1–26]. Van Duyne and co-workers [3–6] recently demonstrated the use of silver film-over-nanosphere (AgFON) as SERS substrates for the detection of B. subtilis spores, harmless simulants for B. anthracis, with a portable Raman spectrometer. In this pioneering work, the AgFON substrates were shown to detect calcium dipicolinate (CaDPA), a biomarker extracted by nitric acid from B. subtilis spores. A limit of detection (LOD) of ∼103 spores, which is below the anthrax infectious dose of 104 spores, was achieved using only a 10-s data collection time, demonstrating the viability of this type of SERS substrate for rapid detection of Bacillus spores with a detection level of spore concentration of 1–2 × 10−14 M [3,4].
156
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Instead of using nitric acid extraction method as described above, simplified approaches to spore sample preparation have also attracted a great deal of interest. Chumanov and co-workers demonstrated the use of sandwich SERS substrates utilizing coupling between silver mirrors, electrochemically roughened films, and silver nanoparticles for the detection of dipicolinic acid (DPA) [11]. Interestingly, this type of SERS substrate was demonstrated for real-time monitoring the kinetics of B. subtilis endospore germination in samples that contained only several hundred endospores [12]. The germination kinetics at varying concentrations of lalanine and different temperatures were studied by monitoring the intensity and growth of the Raman peak characteristic of DPA. They further demonstrated the ability of this type of SERS detection for monitoring DPA release from a single germinating spore of B. subtilis, instead of ensemble-averaging [13]. The SERS detection limit is at the single spore level, representing a 100–1000-fold improvement. These findings further demonstrated the powerful capability of SERS in spore detection, which could parallel those with SERS detection of biomolecules or cells using labels [27–29]. In view of the limitations of silver-based substrates in many biological systems, including the tendency of surface oxidation and the propensity of surface reactivity with certain bacteria, it is desired to develop a SERS substrate that has a wider range of biocompatibility and stability against surface oxidation. We have recently demonstrated the preparation of gold nanoparticles (Au NPs) on gold substrates as a SERS substrate for the detection of dipicolinic acid (DPA), an important biomarker from the bacterial spores. In comparison with the SERS detection of DPA using silver-based substrates, our gold-nanoparticle based SERS substrates were demonstrated to exhibit an enhancement factor of ∼106 and a detection limit of 0.1 ppb [7], which also showed the ability of differentiating the biomarker signature from other possible surface reaction species [8]. The work reported herein has focused on an intriguing question whether our Au nanoparticle based SERS substrates are applicable for the SERS detection of CaDPA released from the spores (Scheme 1). If applicable, how the differences in surface adsorption for the biomarker molecules on the different SERS substrates operate for achieving the desired sensitivity. This is not a trivial extension of our prior work considering the complexity of biomarkers released from real spores, an issue that has not been addressed in previous reports. Herein we describe the findings of this investigation, along with discussing the viability of the Au NP based SERS substrates rapid and sensitive detection of biomarkers from bacterial spores. 2. Experimental 2.1. Materials 2.1.1. Chemicals Polyvinylpyrrolidone (PVP, K30), Ca(OH)2 , HAuCl4 ·3H2 O, NH2 OH·HCl, and trisodium citrate were obtained from Sinopharm Chemical Reagents (Shanghai, China). 2,6-Pyridinedicarboxylic acid (DPA, purity 99.5%) was obtained from Sigma–Aldrich. The other chemicals were all of analytical grade and used as received. 2.1.2. Bacillus subtilis Spore samples were obtained from China General Microbiological Culture Collection Center (CGMCC 1.1630). The preparation involved the use of standard protocols [30–37]. Spore concentration was determined by counting spores using a phase contrast microscopy. Based on optical microscopic images of dyeing spore suspension, the counting results showed that approximately 1 mL of sample contains 9.4 × 108 spores, which translates to spore concentration of 1.6 × 10−12 M.
2.2. Instrumentation Typically, 10-L drop of the spore or CaDPA solution was placed on the surface of AuNPs/PVP/Au substrate. The SERS substrate was then placed on the sample holder in Raman instrument. For each concentration starting from the lowest concentration, the measurement was repeated for several times. The error bars given in the results represent the standard deviations, which were based on replicates of 5–10. The measurements were performed under ambient condition. A Confocal Raman System Laboram 010 (Jobin Yvon Horiba, France) based on a 50× long working-distance objective (8 mm) was used for the Raman spectra collection. A 632.8 nm He–Ne laser excitation (0.1 mW) with slit and pinhole set at 100 and 1000 m, respectively, was used as the laser source. The acquisition of SERS spectra used a laser beam exposure time of 15 s with 3 scans, which considered a balance of signal-to-noise ratio, measurement time, and small sample volume on the SERS substrate. Usually 5–10 spots on the substrate were measured, and the measured intensities fell within the standard devotions. 2.3. Procedures 2.3.1. Extraction of CaDPA The extraction of CaDPA was achieved by adding 0.02 M HNO3 into the spore suspension followed by sonication for 12 min. The clear solution was collected for the analysis. The pH value was ∼8.5 for the spore suspension and ∼3.8 for the extracted solution. Release of CaDPA from the spore suspension was also examined as a function of storage time without adding HNO3 or sonication. As a control to the detection of biomarker released from the spore samples, CaDPA was also synthesized according to the reported method [38]. Briefly, equal moles of DPA and Ca(OH)2 were mixed in 100 mL of distilled water. After stirring for 10 min, the solution was kept under 4 ◦ C for two days to allow crystallization and precipitation of the product of CaDPA. The product was collected and dried in an oven at 70 ◦ C for 1 day. A stock solution of 1 × 10−2 M CaDPA in 0.02 M HNO3 was prepared, and the dilutions were prepared by using 0.02 M HNO3 solution. The stock solution has a pH value of 6–7, which was close to that for the 25% dilution of the spore suspension in 0.02 M HNO3 . 2.3.2. Preparation of gold nanoparticles Gold nanoparticles with an average diameter of 64.8 ± 4.4 nm were prepared according to the method of hydroxylamine reduction and seeded growth reported by Brown and Natan [39], with some modifications in precursor ratios and reaction condition, details of which were previously described [7,8]. This method involved mixing a controlled amount of Au seeds with an 100 mL solution of 0.5 mL 1% HAuCl4 and 2.1 mL 40 mM NH2 OH·HCl to produce Au nanoparticles of the desired size (∼1012 particles mL−1 ). 2.3.3. Preparation of SERS substrates The preparation of Au nanoparticles immobilized on an Au substrate involved use of PVP as an adhesive layer [40–42], details of which were recently described [7,8]. The immobilization of gold nanoparticles on an Au electrode was accomplished by immersing the electrode in an ethanol solution of 2% PVP for 5 h to allow the formation of a PVP layer first before immersing in a solution containing gold nanoparticles for 12 h, leading to the immobilization of gold nanoparticles via PVP layer on the surface of a disc gold electrode (AuNPs/PVP/Au) with a surface area of 0.0314 cm2 . The AuNPs/PVP/Au substrates were rinsed with deionized water several times before use.
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
157
Scheme 1. Illustration of a SERS substrate derived from Au NPs immobilized on a planar Au substrate surface for the detection of CaDPA released from Bacillus subtilis spores.
3. Results and discussion 3.1. Spore samples prepared by acid-extraction of CaDPA SERS spectra were first acquired for spore suspension samples in which CaDPA were extracted using HNO3 (0.02 M) as biomarkers. Fig. 1 shows a representative set of SERS spectra with an AuNPs(60 nm)/PVP/Au substrate for a series of dilutions of the extracted spore samples. Note that the stock solution of the spores used in this experiment had a spore concentration of 1.6 × 10−12 M (i.e., 9.4 × 108 mL–1 ). Major diagnostic peaks are observed at 1001 and 1030 cm−1 , regardless of the spore concentration. These bands are in agreement with those reported previously for samples in which CaDPA was extracted similarly [3–7]. In Fig. 2, the SERS intensity, ISERS (integrated area), for the diagnostic 1001-cm−1 peak is plotted as a function of the spore concentration (Cspore ) in terms of ISERS vs. Cspore and 1/ISERS vs. 1/Cspore . It appears that both ISERS vs. Cspore and 1/ISERS vs. 1/Cspore plots display good linear relationships in this spore concentration range. The observation of the two types of linear relationships in the above data is analyzed by considering the Langmuir adsorption isotherm for the adsorption of the biomarker (CaDPA) released from the spores on the SERS substrate. For simplicity, we consider the adsorption of spores in general. The coverage of the adsorbed biomarker species on the SERS substrate surface can be represented by that for the spores (), which can then be related to the concentration of spores in the solution (Cspore ), which is very small in molar concentration. Under the condition of Kspore Cspore 1, where Kspore stands for the adsorption equilibrium constant, the adsorption isotherm ( = KC/(1 + KC)) can be simplified as =
ISERS = Kspore Cspore Imax
(1)
This correlation is indeed reflected by the linear relationship of ISERS vs. Cspore in Fig. 2A, from which the slope (i.e., Kspore × Imax ) is obtained (4.6 × 1015 ). It is interesting to note that this relationship is apparently consistent with those reported previously by us for the adsorption of DPA on the SERS substrate under very low concentrations of DPA [7], suggesting the low concentration of CaDPA released from the spores by HNO3 extraction. Without any simplification, the Langmuir adsorption isotherm can be expressed in terms of a linear relationship of 1/ISERS vs. 1/Cspore , 1 1 1 1 = + × ISERS Imax × K Cspore Imax
(2)
By plotting 1/ISERS vs. 1/Cspore , the results of a linear relationship would allow one to extract the slope and intercept values, i.e., slope = 1/(Imax × Kspore ) and intercept = 1/Imax . Note that this relationship was reported by Van Duyne and co-workers [3,4] in their SERS study of spores on an AgFON substrate. From the result in Fig. 2B, the slope and intercept translate to Imax =1.4 × 104 ,
and Kspore × Imax = 9.2 × 1015 , yielding a value of 6.5 × 1011 M−1 for Kspore . The measurements were repeated several times using our AuNPs(60 nm)/PVP/Au substrates prepared in different times, which showed an overall variation of less than 15% in terms of SERS intensity. To further examine the intensity-concentration correlation in a broader and lower concentration range, Fig. 3A shows another typical set of spectra for a series of dilutions of spore samples extracted by 0.02 M HNO3 . In this case, the diagnostic peaks are observed at 999 and 1027 cm−1 , practically identical to those observed in Fig. 1. In Fig. 3B, the SERS intensity of the diagnostic 999-cm−1 peak is plotted as a function of the concentration in the spore suspension in terms of ISERS vs. Cspore and 1/ISERS vs. 1/Cspore . In this case, the 1/ISERS vs. 1/Cspore plot displays a very good linear relationship, whereas the ISERS vs.Cspore plot is linear only in a limited concentration range (>3 × 10−13 M). From the linear relationship of ISERS vs. Cspore in the higher concentration range, a value of 4.9 × 1015 is found for Kspore × Imax. From the 1/ISERS vs. 1/Cspore linear relationship, the slope and the intercept translate to Imax = 2.0 × 104 , and Kspore × Imax = 9.0 × 1015 , yielding a value of 4.5 × 1011 M−1 for Kspore . It is apparent that in the concentration range 1.6 × 10−13 –1.6 × 10−12 M (Figs. 1 and 2) and 3.1 × 10−14 –1.6 × 10−12 M (Fig. 3), similar Kspore × Imax values (4.6 × 1015 and 4.9 × 1015 ) were obtained from their ISERS vs. Cspore linear relationships, and similar Kspore values (6.5 × 1011 M−1 and 4.5 × 1011 M−1 ) were obtained from their 1/ISERS vs. 1/Cspore linear relationships. Note that a value of 1.7 × 1013 M−1 [3] or 9.0 × 1013 M−1 [4] was reported by Van Duyne and coworkers for Kspore in the 1/ISERS vs. 1/Cspore linear relationship for an AgFON substrate in the spore concentration range of 1.0 × 10−14 –6.0 × 10−13 M [3,4]. It is evident that the Kspore value for our Au NP based SERS substrate is about two orders of magnitude lower than that reported for the AgFON substrates, reflecting likely the difference in the spore biomarker’s surface adsorption between these two types of substrates.
3.2. CaDPA samples from As-synthesized CaDPA In order to compare the SERS spectra of the CaDPA extracted from the spore suspension using 0.02 M HNO3 , CaDPA was also synthesized according to a reported procedure [38], details of which were described in Section 2. By comparing the Raman spectra for powder samples of DPA and CaDPA, the peak characteristics and the peak shifts are basically consistent with the literature report [43]. We note however that there were subtle differences in SERS bands between the as-synthesized CaDPA and the spore-released CaDPA, indicating the complexity of the spore-released biomarkers. While the origin of the spectral differences is not clear at this time, we believe that the presence of possible species other than CaDPA might have influenced the orientations of DPA or CaDPA
158
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Fig. 1. SERS spectra in 500–1800 cm−1 region (A) with an AuNPs(60 nm)/PVP/Au substrate for a spore suspension after extraction using 0.02 M HNO3 . (Cspore [M] (from (a) to (e)) = 1.6 × 10−13 (a), 3.9 × 10−13 (b), 7.8 × 10−13 (c), 1.2 × 10−12 (d), and 1.6 × 10−12 (e).) The spectra in the 900–1100 cm−1 region are shown in (B). (Unit for the scale bars: cps.)
adsorbed on the SERS substrate, which would be an interesting area of a further fundamental investigation. Fig. 4 shows a representative set of SERS spectra for a series of dilutions of the as-synthesized CaDPA. The main diagnostic peak is observed at 1001 cm−1 . It is interesting to note that the 1032 cm−1 peak is very small in comparison with those in the SERS spectra for CaDPA extracted from the spore suspension using 0.02 M HNO3 (Figs. 1 and 3A). The concentration dependent peak intensity is analyzed similarly to the analysis of the SERS spectra for the spore samples. Under the simplified (KCaDPA CCaDPA 1) and un-simplified conditions, the above Langmuir adsorption isotherms (i.e., Eqs. (1) and (2)) can again be used, where Cspore is substituted by the concentration of CaDPA (CCaDPA ), and Kspore by the adsorption equilibrium constant of CaDPA (KCaDPA ). Fig. 5 shows a plot of the SERS intensity of the 1001-cm−1 peak as a function of the CaDPA concentration in terms of ISERS vs. CCaDPA and 1/ISERS vs. 1/CCaDPA . In the ISERS vs. CCaDPA plot (Fig. 5A), the data do not exhibit a linear relationship. The non-linearity characteristic
is also true by dropping the two lowest and two highest concentration points (see insert). This observation is in sharp contrast to the observation of ISERS vs. Cspore linearity as shown in Figs. 1 and 3A for the spore samples. In the 1/ISERS vs. 1/CCaDPA plot (Fig. 5B), the linearity is observed only for concentrations higher than 6.0 × 10−7 M. It is evident that the two lowest concentration points deviate from the linear relationship (Fig. 5B). The slope and intercept translate to Imax = 9.2 × 105 , and KCaDPA × Imax = 1.2 × 109 , yielding a value of 1.3 × 103 M−1 for KCaDPA in the concentration range of CaDPA (6.0 × 10−7 –1.0 × 10−2 M). This KCaDPA value is consistent with that reported previously for DPA in a similar concentration range, but somewhat smaller than that reported for the AgFON based SERS substrate (KCaDPA = 9.0 × 103 M−1 ) in the concentration range of 2.0 × 10−6 –1.0 × 10−3 M [3], reflecting likely a difference between the adsorption equilibria on the two SERS substrates. In Fig. 6, the detailed SERS spectral characteristics between the spore samples extracted using nitric acid and the as-synthesized
Fig. 2. SERS intensity of the 1001-cm−1 peak as a function of Cspore : ISERS vs. Cspore (A) and 1/ISERS vs. 1/Cspore (B).
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
159
Fig. 3. (A) SERS spectra in 900–1100 cm−1 region for spore suspension extracted by 0.02 M HNO3 on an AuNPs(60 nm)/PVP/Au substrate. (Cspore [M] (from (a) to (h)) = 3.1 × 10−14 (a), 6.2 × 10−14 (b), 1.2 × 10−13 (c), 1.6 × 10−13 (d), 3.9 × 10−13 (e) 7.8 × 10−13 (f), 1.2 × 10−12 (g), and 1.6 × 10−12 (h).) (B) SERS intensity of the 999-cm−1 peak as a function of Cspore : ISERS vs. Cspore (a) and 1/ISERS vs. 1/Cspore (b).
CaDPA samples are further compared in both high and low concentrations. While the general spectral characteristics are similar, a major distinction is that the peak width for the CaDPA extracted from spore samples is narrower than that observed for the
as-synthesized CaDPA samples. This difference is believed to reflect the existence of surface competitive adsorptions of chemical species on the Au NP based substrate between the spore solution after acid extraction of CaDPA and the solution of the as-prepared
Fig. 4. SERS spectra in 500–1800 cm−1 region for CaDPA of different concentrations in the presence of 0.02 M HNO3 with an AuNPs(60 nm)/PVP/Au substrate (A). (CCaDPA [M] (from (a) to (i)) = 3.0 × 10−7 (a), 4.0 × 10−7 (b), 6.0 × 10−7 (c), 1.0 × 10−6 (d), 5.0 × 10−6 (e), 1.0 × 10−5 (f), 1.0 × 10−4 (g), 1.0 × 10−3 (h), 1.0 × 10−2 (i).) A magnified view of SERS spectra in the 900–1100 cm−1 region is shown in B. (Unit for the scale bars: cps.)
160
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Fig. 5. SERS intensity of the 1001-cm−1 peak as a function of CCaDPA : ISERS vs. CCaDPA (A) and 1/ISERS vs. 1/CCaDPA (B). (Linear regression without the two low concentration points (B), R2 = 0.99, slope = 8.3 × 10−10 , int. = 1.1 × 10−6 .)
CaDPA. We believe it is unique for our Au NP based substrates, which is possibly originated from differences in the adsorption modes for the biomarker molecules on the SERS substrates. In our previous study of DPA [7], we showed different linearity characteristics for I vs. C depending on the concentration regions. By plotting the previous data [7] in terms of 1/ISERS vs. 1/CDPA , nonlinearity was observed for CDPA < 0.01 ppm (i.e., 6.0 × 10−8 M) and CDPA < 1 ppm (i.e., 6.0 × 10−6 M) whereas a linearity was observed for CDPA > 1 ppm. It is apparent that in the high concentration range (CDPA > 1 ppm) the data display a 1/ISERS vs. 1/CDPA linear relationship. The slope and intercept (1.5 × 10−8 and 1.8 × 10−5 ) translate to Imax = 5.5 × 104 , and KDPA × Imax = 6.6 × 107 , yielding a value of 1.2 × 103 M−1 for KDPA . To ensure the reproducibility, the concentration dependence of DPA was also examined for the SERS substrate prepared in the current study. The observed main peaks at 1000 and 1030 cm−1 are in agreement with our
previous reports [7,8]. From the linear relationship of I vs. C, a value of 1.9 × 103 M−1 was obtained for KDPA . From the 1/I vs. 1/CDPA linear plot, the slope and the intercept translate to Imax = 5.5 × 104 , and KDPA = 1.2 × 103 M−1 , practically the same as 1.3 × 103 M−1 for KCaDPA obtained in the current work under the similar concentration range, but different from that reported for AgFON substrate [3,4]. Another possible factor is the difference in adsorption mode of CaDPA or DPA between Au NP and Ag NP based substrates. As we described previously [7], the 1000-cm−1 peak position for the ringbreathing mode of DPA was found to be somewhat lower than those observed for SERS substrates with Ag colloids, e.g., 1020 cm−1 [3–6] and 1010 cm−1 [9,10]. This difference is believed to reflect the difference of the surface adsorption sites and the surface orientation of the adsorbed DPA molecules between Au and Ag particles. The surface orientation of DPA adsorbed on the Au NP was believed to
Fig. 6. Comparison of SERS spectra (in 500–1800 cm−1 region) between acid-extracted spore samples (a) and as-synthesized CaDPA samples (b) under two different concentrations: (A) spore sample ((a) 1.6 × 10−13 M, extracted with 0.02 M HNO3 ) and CaDPA ((b) 6.0 × 10−7 M in the presence of 0.02 M HNO3 ); (B) spore sample ((a) 1.6 × 10−12 M extracted with 0.02 M HNO3 ) and CaDPA ((b) 1.0 × 10−5 M in the presence of 0.02 M HNO3 ).
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
161
be possibly an edge-on mode via its aromatic ring interacting with the surface, which may account for the significant enhancement for the ring-breathing mode. Since DPA is adsorbed on the surface via carboxylate groups, the interaction between carboxylate groups and alumina or silver surfaces of the AgFON substrate [3–5] could be stronger than the interaction between DPA and the Au NP substrate because of the ionic character of the bonding between the negatively charged carboxylates and the positively charged Al(III) or Ag(I) surface species. In the case of DPA on Au surface, the adsorption occurs likely through a coordination interaction via its pyridine ring with Au surface [8], which should be weaker than the electrostatic interaction. Such a difference in the adsorption mode is likely responsible for the difference between KCaDPA (1.3 × 103 M−1 ) determined in this work and KCaDPA (9.0 × 103 M−1 ) reported for AgFON in the previous work [3]. 3.3. Assessment of extraction efficiency and detection limit Table 1 summarizes the data determined from the above linear relationships, including Kspore and KCaDPA . In these experiments spore samples were diluted from a stock solution with a concentration of 1.6 × 10−12 M and spore density of 7.9 × 1010 spores g−1 . Using Kspore = 4.5 × 1011 M−1 and KCaDPA = 1.3 × 103 M−1 , moles of CaDPA extracted from the spores can be obtained from the ratio of Kspore /KCaDPA , which yields 3.5 × 108 moles of CaDPA extracted from the spores. Since the mass of CaDPA (WCaDPA = 3.5 × 108 × (167 + 40)) is 7.2 × 1010 g and the mass of spores (Wspore = 6.02 × 1023 /(7.92 × 1010 spores g−1 )) is 7.6 × 1012 g, an extracted percentage (7.2 × 1010 g/7.6 × 1012 g) of 1.0% is obtained, which gives an extraction efficiency (1.0%/8.9%) of 11%. A slightly different value of 15% was also obtained for the extraction efficiency from another set of data (Table 1), indicating some degree of variability of the determined extraction efficiency value. In the early report [3] for spore samples with a spore concentration of 3.1 × 10−13 M and spore density of 5.6 × 1010 spores g−1 , the calculation using Kspore (1.7 × 1013 M−1 ) and KCaDPA (9.0 × 103 M−1 ) gave a value of 1.9 × 109 M CaDPA. Using WCaDPA (1.9 × 109 × (167 + 40)) = 3.9 × 1011 g and Wspore 23 (6.02 × 10 /(5.6 × 1010 spores g−1 )) = 1.1 × 1013 g, the calculated extracted percentage (3.9 × 1011 g/1.08 × 1013 g) and the extraction efficiency (3.0%/8.9%) give 3.0% and 34%, respectively. Note that in a later report by the same authors [4], Kspore = 9.0 × 1013 M−1 and KCaDPA = 4.9 × 104 M−1 were obtained. In this case, Kspore /KCaDPA ratio is 1.8 × 109 , which is similar to that reported in their early work [3]. Apparently, the extraction efficiency (34%) is somewhat larger than what was observed in our experiment (11%), probably due to difference in experimental condition for the preparation of spore samples. The limit of detection (LOD) was also estimated from the SERS data for the lowest concentration by considering a minimum signal being above three times of the noise level. Based on the data corresponding to the lowest spore concentration (3.1 × 10−14 M) (not shown), the estimate yielded a value of 2.5 × 10−14 M for LOD. By linear extrapolation in this region with the consideration of 3 times of noise level, this value corresponds to LOD for the detection of the spores, which translates to 3.0 × 103 spores per 0.2 L sample solution in comparison with the LOD value reported for the spore detection using AgFON substrates (1.1–2.1 × 10−14 M or 1.4–2.6 × 103 spores per 0.2 L) [3,4]. For the as-synthesized CaDPA samples corresponding to the lowest CaDPA concentration (3.0 × 10−7 M) (not shown), the linear extrapolation in this region with the consideration of 3 times of noise level yielded a value of 1.3 × 10−7 M for the LOD. This value is one order of magnitude lower than the LOD reported for the CaDPA detection using AgFON substrates (3.1 × 10−6 M in 0.2 L [3] or 1.9 × 10−6 M in 0.2 L [4]). In Table 2, the LOD data from
Fig. 7. Comparison of SERS spectra for spore samples aged for different lengths of time without using reagent for extraction: (a) blank (0.02 M HNO3 ), (b) 2 months, (c) 5 months (supernatant after centrifugation), (d) 5 months (after stirring), (e) 1 year (50% dilution) and (f) fresh spore suspension extracted by 0.02 M HNO3 . The concentration of stock solution of spores: 1.6 × 10−12 M for (b), (c), and (d), 2.6 × 10−12 M for (e), and 1.6 × 10−12 M for (f).
our SERS detection using Au NP based substrates are summarized, in comparison with the LOD values reported for SERS detection of spores and CaDPA using AgFON substrates [3,4]. In view of the relatively large difference of adsorption equilibrium constants between our Au NP substrates and the previous AgFON based substrates, it is remarkable that the difference in LOD in terms of spore concentration is relatively small, with a value of 3.0 × 103 spores per 0.2 L for Au NP substrates, and 1.4–2.6 × 103 spores per 0.2 L for AgFON substrates [3,4]. In comparison, the difference in LOD in terms of as-prepared CaDPA concentration is much greater, with a value of 1.3 × 10−7 M for Au NP substrates, and 1.9–3.1 × 10−6 M for AgFON substrates [3,4]. 3.4. Detection of spores samples without extraction acid In addition to extraction of CaDPA from spore samples using 0.02 M HNO3 , the possible self-release of CaDPA from the spore suspension was also examined as a function of storage time without adding HNO3 or sonication. CaDPA was found to be released slowly from spore suspension. In Fig. 7, a representative set of SERS spectra for several samples of spores are compared, including spore samples after storage under 4 ◦ C for different time lengths. While there are small differences of the detected bands among the samples aged for different time lengths, these samples showed the diagnostic band at ∼1000 cm−1 characteristic of CaDPA. The peak width is similar to that observed for the spore samples after acid extraction. Since there is no intentional means being applied for the extraction, the detected CaDPA was termed as “self-released” CaDPA from the spores. The estimated concentrations of the equivalent spores corresponding to the naturally released CaDPA after ageing for different time lengths are shown in Table 3.
162
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Table 1 A summary of Kspore and KCaDPA , and extraction efficiency for CaDPA from spore samples. K (M−1 )
Expt.
C
Spore – (a) Spore – (b) CaDPA In Ref. [3]
7.9 × 10 spore g 7.9 × 1010 spore g−1 3.0 × 10−7 –1.0 × 10−2 M CD 5.6 × 1010 spore g−1 3.1 × 10−6 –1.3 × 10−3 M CDc
−1
10
6.5 × 10 4.5 × 1011 M−1 1.3 × 103 M−1 1.7 × 1013 M−1 9.0 × 103 M−1 11
E.% a
E.E.% b
1.4% 1.0%
15% (a) 11% (b)
3.0%
34%
Note: Spore conc. = 1.6 × 10−12 M (9.4 × 108 spores mL−1 ); spore density = 7.9 × 1010 spores/g; laser power = 0.1 mW. K refers to Kspore or KCaDPA . a E% represents % of extraction. b E.E.% refers to extraction efficiency. c CD refers to CaDPA. Table 2 A summary of LODs from our SERS detection using Au NP based substrates in comparison with those reported using AgFON substrates. LODCaDPA a
SERS substrate
LODspore a
AgFON [3] AgFON [4] AuNPs/Au
2.6 × 10 spore (2.1 × 10 M) 1.4 × 103 spore (1.1 × 10−14 M) 3 −14 3.0 × 10 spore (2.5 × 10 M) 3
−14
−6
3.1 × 10 M 1.9 × 10−6 M 1.3 × 10−7 M
Plaser (mW) 50 50 0.1
Note: a 0.2 L 0.02 M HNO3 .
For a fresh solution of spore samples, the self-released CaDPA (e.g., after 2–5 months) corresponded to one third or fourth of the equivalent spore concentration achieved by the acid extraction. For an aged spore solution, the self-released CaDPA (e.g., after 1 year) corresponded to a spore concentration which is almost the same as the equivalent spore concentration achieved by the acid extraction. To understand the mechanism responsible for the self-release of CaDPA from spores and its implication for the SERS detection, we considered two possible pathways. The first involved considering the possibility of laser-induced release. However, several control experiments with different time lengths of exposures by the laser power used for the SERS measurement did not reveal significant release of measurable signals. This ruled out the possibility of laserinduced release. The second pathway involved considering some possible conditions for germination, denaturation and inactivation. It is known that the spore inner membrane could become permeable during germination in which ion fluxes and CaDPAs are released from inside the spore as a result of spore re-hydration. In a recent report [44], some non-nutrient agents and nutrient germinants bound to specific receptors located in the spore’s inner membrane were shown to trigger the release of ∼10% spore (dry weight) of DPA which is located exclusively in the spore’s central region or core as a 1:1 chelate with divalent cations (CaDPA). CaDPA release triggers hydrolysis of the spore’s peptidoglycan (PG) cortex, which then allows swelling of the spore core and further water uptake, resulting in a core water content similar to that in growing cells. In another recent report [45], it is pointed out that spores can rapidly return to active growth through germination followed by outgrowth although spores of bacteria of the Bacillus species
Table 3 Estimated concentrations of the equivalent spores corresponding to the naturally released CaDPA after ageing for different time lengths (based on linear relationship in Fig. 3B(b)). Spore samples fresh stock suspension (b) 2 months (c) 5 month (cent’d) (d) 5 month (stirred) (e) 1 year (50%)b
SERS peak & Ia −1
999 cm , 7962 999 cm−1 , 3306 1000 cm−1 , 4397 1000 cm−1 , 4856 999 cm−1 , 7461
Conc. of spore (M) 1.6 × 10−12 4.4 × 10−13 6.2 × 10−13 7.1 × 10−13 1.3 × 10−12
Note: The stock solution was not subjected to acid extraction, and was not homogenized in the bottle. a I refers to integrated area. b Curve (e) in Fig. 7 is 50% of the stock solution (2.6 × 10−12 M) after ∼1 year.
are very resistant to a variety of harsh conditions and can survive for many years. Wet-heat treatment in spore killing is associated with protein denaturation and enzyme inactivation, which results in the release of the spore core’s CaDPA through rupture of the spore’s inner membrane. It is therefore believed that a combination of non-nutrient/nutrient induced germination and wet-heat induced denaturation could have played an important role in the release of CaDPA. 4. Conclusions In conclusion, the Au NP based SERS substrates have been shown to function as an effective probe for the highly sensitive detection of biomarker CaDPA released from B. subtilis spores. The LOD (1.5 × 109 spores L−1 ) for our Au NP based SERS substrates has been found to be quite comparable with that reported for Ag NP based SERS substrates. This finding is remarkable in view of the fact that the surface adsorption equilibrium constant for the Au NP based SERS substrates is smaller by a factor of 1–2 order magnitude than that for Ag NP based SERS substrates. The results have also revealed the viability of SERS detection of CaDPA released from the spores under ambient conditions without extraction using any reagents. This is an intriguing finding in view of the detection of CaDPA released from the spores under the ambient condition without any extraction procedures. One important implication of this finding is the viability of the Au NP based SERS substrates as a biocompatible probe with high resolution and sensitivity for rapid and sensitive detection of bacteria spores with minimized interferences from possible co-existent species in biological samples. Acknowledgements This work is financially supported by the National Natural Science Foundation of China (Grant Numbers: 21035001, 20775023, 20605007, and 20675005) and “973” National Key Basic Research Program of China (2007CB310500, and 2011CB911000). References [1] S.J. Park, T.A. Taton, C.A. Mirkin, Science 295 (2002) 1503–1506. [2] C.L. Haynes, A.D. McFarland, R.P. Van Duyne, Anal. Chem. 77 (2005) 338A–346A. [3] X.Y. Zhang, M.A. Young, O. Lyandres, R.P. Van Duyne, J. Am. Chem. Soc. 127 (2005) 4484–4489. [4] X.Y. Zhang, J. Zhao, A.V. Whitney, J.W. Elam, R.P. Van Duyne, J. Am. Chem. Soc. 128 (2006) 10304–10309.
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163 [5] X.Y. Zhang, N.C. Shah, R.P. Van Duyne, Vib. Spectrosc. 42 (2006) 2–8. [6] C.R. Yonzon, X.Y. Zhang, J. Zhao, R.P. Van Duyne, Spectroscopy 22 (2007) 42–56. [7] H.W. Cheng, S.Y. Huan, H.L. Wu, G.L. Shen, R.Q. Yu, Anal. Chem. 81 (2009) 9902–9912. [8] H.W. Cheng, W.Q. Luo, G.L. Wen, S.Y. Huan, G.L. Shen, R.Q. Yu, Analyst 135 (2010) 2993–3001. [9] S.E.J. Bell, J.N. Mackle, N.M.S. Sirimuthu, Analyst 130 (2005) 545–549. [10] J.W. Lee, H.B. Lee, K. Kim, Anal. Bioanal. Chem. 397 (2010) 557–562. [11] J.K. Daniels, G. Chumanov, J. Phys. Chem. B 109 (2005) 17936–17942. [12] J.K. Daniels, T.P. Caldwell, K.A. Christensen, G. Chumanov, Anal. Chem. 78 (2006) 1724–1729. [13] D.D. Evanoff Jr.J., T.P. Heckel, K.A. Caldwell, G. Christensen, Chumanov, J. Am. Chem. Soc. 128 (2006) 12618–12619. [14] J. He, X.F. Luo, S.W. Chen, L.L. Cao, M. Sun, Z.N. Yu, J. Chromatogr. A 994 (2003) 207–212. [15] L.H. Harrison, J.W. Ezzell, T.G. Abshire, S. Kidd, A.F. Kaufmann, J. Infect. Dis. 160 (1989) 706–710. [16] L.J. Kricka, P. Fortina, Clin. Chem. 47 (2001) 1479–1482. [17] S.E.J. Bell, N.M.S. Sirimuthu, Chem. Soc. Rev. 37 (2008) 1012–1024. [18] K. Kneipp, H. Kneipp, J. Kneipp, Acc. Chem. Res. 39 (2006) 443. [19] J. Kneipp, H. Kneipp, K. Kneipp, Chem. Soc. Rev. 37 (2008) 1052–1060. [20] M.D. Porter, R.J. Lipert, L.M. Siperko, G. Wang, R. Narayanana, Chem. Soc. Rev. 37 (2008) 1001–1011. [21] S. Nie, S. Emory, Science 275 (1997) 1102. [22] J.D. Driskell, R.J. Lipert, M.D. Porter, J. Phys. Chem. B 110 (2006) 17444–17451. [23] P.N. Njoki, S. Lim, D. Mott, H.Y. Park, B. Khan, S. Mishra, R. Sujakumar, J. Luo, C.J. Zhong, J. Phys. Chem. C 111 (2007) 14664–14669. [24] S. Lim, P.N. Njok, H.Y. Park, X. Wang, L. Wang, D. Mott, C.J. Zhong, Nanotechnology 19 (2008) 305102. [25] A. Gultekin, A. Ersoz, D. Hur, N.Y. Sariozlu, A. Denizli, R. Say, Appl. Surf. Sci. 256 (2009) 142–148.
163
[26] A. Dhawan, Y. Du, F. Yan, M.D. Gerhold, V. Misra, T. Vo-Dinh, IEEE Sens. J. 10 (2010) 608–616. [27] G.C. Zhu, Y.J. Hu, J. Gao, L. Zhong, Anal. Chim. Acta 697 (2011) 61–66. [28] M.A. Woo, S.M. Lee, G.S. Kim, J.H. Baek, M.S. Noh, J.E. Kim, S.J. Park, A.M. Tehrani, S.C. Park, Y.T. Seo, Y.K. Kim, Y.S. Lee, D.H. Jeong, M.H. Cho, Anal. Chem. 81 (2009) 1008–1015. [29] J. Kneipp, H. Kneipp, A. Rajadurai, R.W. Redmond, K. Kneipp, J. Raman Spectrosc. 40 (2009) 1–5. [30] J.A. Lindsay, T.C. Beaman, P. Gerhardt, J. Bacteriol. 163 (1985) 735–737. [31] P.S. Margolis, A. Driks, R. Losick, J. Bacteriol. 174 (1993) 528–540. [32] R.E. Marquis, G.R. Bender, J. Bacteriol. 161 (1985) 789–791. [33] W.L. Nicholson, P. Setlow, Sporulation, germination and outgrowth, in: C.R. Harwood, S.M. Cutting (Eds.), Molecular biological methods for Bacillus., John Wiley and Sons, Chichester, England, 1990, pp. 391–450. [34] M. Paidhungat, P. Setlow, J. Bacteriol. 182 (2000) 2513–2519. [35] D.L. Popham, S. Sengupta, P. Setlow, Appl. Environ. Microbiol. 61 (1995) 3633–3638. [36] Y. Rotman, M.L. Fields, Anal. Biochem. 22 (1967) 168. [37] R. Tennen, B. Setlow, K.L. Davis, C.A. Loshon, P. Setlow, J. Appl. Microbiol. 89 (2000) 1–10. [38] G.F. Bailey, S. Karp, L.E. Sacks, J. Bacteriol. 89 (1965) 984–987. [39] K.R. Brown, M.J. Natan, Langmuir 14 (1998) 726–728. [40] Q. Zhou, Q. Fan, Y. Zhuang, Y. Li, G. Zhao, J.W. Zheng, J. Phys. Chem. B 110 (2006) 12029–12033. [41] X.J. Liu, S.G. Huan, Y. Bu, G.L. Shen, R.Q. Yu, Talanta 75 (2008) 797–803. [42] G.K. Liu, J. Hu, P.C. Zheng, G.L. Shen, J.H. Jiang, R.Q. Yu, Y. Cui, B. Ren, J. Phys. Chem. C 112 (2008) 6499–6508. [43] A.A. Kolomenskii, H.A. Schuessler, Spectrochim. Acta A 61 (2005) 647–651. [44] L.X. Peng, D. Chen, Y.Q. Li, P. Setlow, Anal. Chem. 81 (2009) 4035–4042. [45] P.F. Zhang, L.B. Kong, P. Setlow, Y.Q. Li, Appl. Environ. Microbiol. 76 (2010) 1796–1805.
Contents lists available at SciVerse ScienceDirect
Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca
Surface-enhanced Raman spectroscopic detection of Bacillus subtilis spores using gold nanoparticle based substrates Han-Wen Cheng, Yuan-Yuan Chen, Xin-Xin Lin, Shuang-Yan Huan ∗ , Hai-Long Wu, Guo-Li Shen, Ru-Qin Yu ∗ State Key Laboratory for Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha 410082, China
a r t i c l e
i n f o
Article history: Received 17 July 2011 Received in revised form 3 September 2011 Accepted 8 September 2011 Available online 16 September 2011 Keywords: Surface-enhanced Raman scattering Gold nanoparticles Biomarker Calcium Dipicolinate Bacillus subtilis
a b s t r a c t The detection of bacterial spores requires the capability of highly sensitive and biocompatible probes. This report describes the findings of an investigation of surface-enhanced Raman spectroscopic (SERS) detection of Bacillus subtilis spores using gold-nanoparticle (Au NP) based substrates as the spectroscopic probe. The SERS substrates are shown to be highly sensitive for the detection of B. subtilis spores, which release calcium dipicolinate (CaDPA) as a biomarker. The SERS bands of CaDPA released from the spores by extraction using nitric acid provide the diagnostic signal for the detection, exhibiting a limit of detection (LOD) of 1.5 × 109 spores L−1 (or 2.5 × 10−14 M). The LOD for the Au NP based substrates is quite comparable with that reported for Ag nanoparticle based substrates for the detection of spores, though the surface adsorption equilibrium constant is found to be smaller by a factor of 1–2 orders of magnitude than the Ag nanoparticle based substrates. The results have also revealed the viability of SERS detection of CaDPA released from the spores under ambient conditions without extraction using any reagents, showing a significant reduction of the diagnostic peak width for the detection. These findings have demonstrated the viability of Au NP based SERS substrates for direct use with high resolution and sensitivity as a biocompatible probe for the detection of bacterial spores. © 2011 Elsevier B.V. All rights reserved.
1. Introduction There are different types of bacteria, some of which constitute different pathogens to various diseases. One of the dangerous pathogens for the disease anthrax is Bacillus anthracis. This pathogen is a Gram-positive spore-forming and rod-shaped bacterium, and can be grown in an ordinary nutrient medium under aerobic or anaerobic conditions. Humans may be infected by eating or inhaling anthrax spores, breathing in material contaminated with the bacteria, or through skin exposure to infected animals. Anthrax infection is usually diagnosed by identifying the bacteria within skin blisters, blood, or other body fluids. The anthrax bacterium is also a potential biological weapon. Therefore, the rapid and accurate detection of the bacteria is very important for human health. Traditionally, PCR method has been used for bacterial identification in such pathogens. This method has the ability to discriminate between B. anthracis and other Bacillus species. However, the relative lengthy time in analysis for PCR method makes it difficult for situations requiring quick identification of the bacterial
∗ Corresponding authors. Tel.: +86 73188821916. E-mail addresses: [email protected] (S.-Y. Huan), [email protected] (R.-Q. Yu). 0003-2670/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.aca.2011.09.007
spores. There is an increasing need to develop rapid and sensitive methods for the detection of the bacterial spores. Among different methods for detecting bacterial spores, spectroscopic methods have been increasingly demonstrated as powerful tools for rapid and sensitive chemical or biological detection, especially with use of metal nanoparticles as detection probe. For example, the use of DNA-anchored gold nanoparticles has been demonstrated for colorimetric detection of the DNA from the infecting organism B. anthracis [1]. Because of nanoparticle-based enhancement effect on Raman scattering intensity, surfaceenhanced Raman scattering (SERS) technique has been widely utilized for the detection of biomolecules including proteins, DNAs, and bacteria [1–26]. Van Duyne and co-workers [3–6] recently demonstrated the use of silver film-over-nanosphere (AgFON) as SERS substrates for the detection of B. subtilis spores, harmless simulants for B. anthracis, with a portable Raman spectrometer. In this pioneering work, the AgFON substrates were shown to detect calcium dipicolinate (CaDPA), a biomarker extracted by nitric acid from B. subtilis spores. A limit of detection (LOD) of ∼103 spores, which is below the anthrax infectious dose of 104 spores, was achieved using only a 10-s data collection time, demonstrating the viability of this type of SERS substrate for rapid detection of Bacillus spores with a detection level of spore concentration of 1–2 × 10−14 M [3,4].
156
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Instead of using nitric acid extraction method as described above, simplified approaches to spore sample preparation have also attracted a great deal of interest. Chumanov and co-workers demonstrated the use of sandwich SERS substrates utilizing coupling between silver mirrors, electrochemically roughened films, and silver nanoparticles for the detection of dipicolinic acid (DPA) [11]. Interestingly, this type of SERS substrate was demonstrated for real-time monitoring the kinetics of B. subtilis endospore germination in samples that contained only several hundred endospores [12]. The germination kinetics at varying concentrations of lalanine and different temperatures were studied by monitoring the intensity and growth of the Raman peak characteristic of DPA. They further demonstrated the ability of this type of SERS detection for monitoring DPA release from a single germinating spore of B. subtilis, instead of ensemble-averaging [13]. The SERS detection limit is at the single spore level, representing a 100–1000-fold improvement. These findings further demonstrated the powerful capability of SERS in spore detection, which could parallel those with SERS detection of biomolecules or cells using labels [27–29]. In view of the limitations of silver-based substrates in many biological systems, including the tendency of surface oxidation and the propensity of surface reactivity with certain bacteria, it is desired to develop a SERS substrate that has a wider range of biocompatibility and stability against surface oxidation. We have recently demonstrated the preparation of gold nanoparticles (Au NPs) on gold substrates as a SERS substrate for the detection of dipicolinic acid (DPA), an important biomarker from the bacterial spores. In comparison with the SERS detection of DPA using silver-based substrates, our gold-nanoparticle based SERS substrates were demonstrated to exhibit an enhancement factor of ∼106 and a detection limit of 0.1 ppb [7], which also showed the ability of differentiating the biomarker signature from other possible surface reaction species [8]. The work reported herein has focused on an intriguing question whether our Au nanoparticle based SERS substrates are applicable for the SERS detection of CaDPA released from the spores (Scheme 1). If applicable, how the differences in surface adsorption for the biomarker molecules on the different SERS substrates operate for achieving the desired sensitivity. This is not a trivial extension of our prior work considering the complexity of biomarkers released from real spores, an issue that has not been addressed in previous reports. Herein we describe the findings of this investigation, along with discussing the viability of the Au NP based SERS substrates rapid and sensitive detection of biomarkers from bacterial spores. 2. Experimental 2.1. Materials 2.1.1. Chemicals Polyvinylpyrrolidone (PVP, K30), Ca(OH)2 , HAuCl4 ·3H2 O, NH2 OH·HCl, and trisodium citrate were obtained from Sinopharm Chemical Reagents (Shanghai, China). 2,6-Pyridinedicarboxylic acid (DPA, purity 99.5%) was obtained from Sigma–Aldrich. The other chemicals were all of analytical grade and used as received. 2.1.2. Bacillus subtilis Spore samples were obtained from China General Microbiological Culture Collection Center (CGMCC 1.1630). The preparation involved the use of standard protocols [30–37]. Spore concentration was determined by counting spores using a phase contrast microscopy. Based on optical microscopic images of dyeing spore suspension, the counting results showed that approximately 1 mL of sample contains 9.4 × 108 spores, which translates to spore concentration of 1.6 × 10−12 M.
2.2. Instrumentation Typically, 10-L drop of the spore or CaDPA solution was placed on the surface of AuNPs/PVP/Au substrate. The SERS substrate was then placed on the sample holder in Raman instrument. For each concentration starting from the lowest concentration, the measurement was repeated for several times. The error bars given in the results represent the standard deviations, which were based on replicates of 5–10. The measurements were performed under ambient condition. A Confocal Raman System Laboram 010 (Jobin Yvon Horiba, France) based on a 50× long working-distance objective (8 mm) was used for the Raman spectra collection. A 632.8 nm He–Ne laser excitation (0.1 mW) with slit and pinhole set at 100 and 1000 m, respectively, was used as the laser source. The acquisition of SERS spectra used a laser beam exposure time of 15 s with 3 scans, which considered a balance of signal-to-noise ratio, measurement time, and small sample volume on the SERS substrate. Usually 5–10 spots on the substrate were measured, and the measured intensities fell within the standard devotions. 2.3. Procedures 2.3.1. Extraction of CaDPA The extraction of CaDPA was achieved by adding 0.02 M HNO3 into the spore suspension followed by sonication for 12 min. The clear solution was collected for the analysis. The pH value was ∼8.5 for the spore suspension and ∼3.8 for the extracted solution. Release of CaDPA from the spore suspension was also examined as a function of storage time without adding HNO3 or sonication. As a control to the detection of biomarker released from the spore samples, CaDPA was also synthesized according to the reported method [38]. Briefly, equal moles of DPA and Ca(OH)2 were mixed in 100 mL of distilled water. After stirring for 10 min, the solution was kept under 4 ◦ C for two days to allow crystallization and precipitation of the product of CaDPA. The product was collected and dried in an oven at 70 ◦ C for 1 day. A stock solution of 1 × 10−2 M CaDPA in 0.02 M HNO3 was prepared, and the dilutions were prepared by using 0.02 M HNO3 solution. The stock solution has a pH value of 6–7, which was close to that for the 25% dilution of the spore suspension in 0.02 M HNO3 . 2.3.2. Preparation of gold nanoparticles Gold nanoparticles with an average diameter of 64.8 ± 4.4 nm were prepared according to the method of hydroxylamine reduction and seeded growth reported by Brown and Natan [39], with some modifications in precursor ratios and reaction condition, details of which were previously described [7,8]. This method involved mixing a controlled amount of Au seeds with an 100 mL solution of 0.5 mL 1% HAuCl4 and 2.1 mL 40 mM NH2 OH·HCl to produce Au nanoparticles of the desired size (∼1012 particles mL−1 ). 2.3.3. Preparation of SERS substrates The preparation of Au nanoparticles immobilized on an Au substrate involved use of PVP as an adhesive layer [40–42], details of which were recently described [7,8]. The immobilization of gold nanoparticles on an Au electrode was accomplished by immersing the electrode in an ethanol solution of 2% PVP for 5 h to allow the formation of a PVP layer first before immersing in a solution containing gold nanoparticles for 12 h, leading to the immobilization of gold nanoparticles via PVP layer on the surface of a disc gold electrode (AuNPs/PVP/Au) with a surface area of 0.0314 cm2 . The AuNPs/PVP/Au substrates were rinsed with deionized water several times before use.
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
157
Scheme 1. Illustration of a SERS substrate derived from Au NPs immobilized on a planar Au substrate surface for the detection of CaDPA released from Bacillus subtilis spores.
3. Results and discussion 3.1. Spore samples prepared by acid-extraction of CaDPA SERS spectra were first acquired for spore suspension samples in which CaDPA were extracted using HNO3 (0.02 M) as biomarkers. Fig. 1 shows a representative set of SERS spectra with an AuNPs(60 nm)/PVP/Au substrate for a series of dilutions of the extracted spore samples. Note that the stock solution of the spores used in this experiment had a spore concentration of 1.6 × 10−12 M (i.e., 9.4 × 108 mL–1 ). Major diagnostic peaks are observed at 1001 and 1030 cm−1 , regardless of the spore concentration. These bands are in agreement with those reported previously for samples in which CaDPA was extracted similarly [3–7]. In Fig. 2, the SERS intensity, ISERS (integrated area), for the diagnostic 1001-cm−1 peak is plotted as a function of the spore concentration (Cspore ) in terms of ISERS vs. Cspore and 1/ISERS vs. 1/Cspore . It appears that both ISERS vs. Cspore and 1/ISERS vs. 1/Cspore plots display good linear relationships in this spore concentration range. The observation of the two types of linear relationships in the above data is analyzed by considering the Langmuir adsorption isotherm for the adsorption of the biomarker (CaDPA) released from the spores on the SERS substrate. For simplicity, we consider the adsorption of spores in general. The coverage of the adsorbed biomarker species on the SERS substrate surface can be represented by that for the spores (), which can then be related to the concentration of spores in the solution (Cspore ), which is very small in molar concentration. Under the condition of Kspore Cspore 1, where Kspore stands for the adsorption equilibrium constant, the adsorption isotherm ( = KC/(1 + KC)) can be simplified as =
ISERS = Kspore Cspore Imax
(1)
This correlation is indeed reflected by the linear relationship of ISERS vs. Cspore in Fig. 2A, from which the slope (i.e., Kspore × Imax ) is obtained (4.6 × 1015 ). It is interesting to note that this relationship is apparently consistent with those reported previously by us for the adsorption of DPA on the SERS substrate under very low concentrations of DPA [7], suggesting the low concentration of CaDPA released from the spores by HNO3 extraction. Without any simplification, the Langmuir adsorption isotherm can be expressed in terms of a linear relationship of 1/ISERS vs. 1/Cspore , 1 1 1 1 = + × ISERS Imax × K Cspore Imax
(2)
By plotting 1/ISERS vs. 1/Cspore , the results of a linear relationship would allow one to extract the slope and intercept values, i.e., slope = 1/(Imax × Kspore ) and intercept = 1/Imax . Note that this relationship was reported by Van Duyne and co-workers [3,4] in their SERS study of spores on an AgFON substrate. From the result in Fig. 2B, the slope and intercept translate to Imax =1.4 × 104 ,
and Kspore × Imax = 9.2 × 1015 , yielding a value of 6.5 × 1011 M−1 for Kspore . The measurements were repeated several times using our AuNPs(60 nm)/PVP/Au substrates prepared in different times, which showed an overall variation of less than 15% in terms of SERS intensity. To further examine the intensity-concentration correlation in a broader and lower concentration range, Fig. 3A shows another typical set of spectra for a series of dilutions of spore samples extracted by 0.02 M HNO3 . In this case, the diagnostic peaks are observed at 999 and 1027 cm−1 , practically identical to those observed in Fig. 1. In Fig. 3B, the SERS intensity of the diagnostic 999-cm−1 peak is plotted as a function of the concentration in the spore suspension in terms of ISERS vs. Cspore and 1/ISERS vs. 1/Cspore . In this case, the 1/ISERS vs. 1/Cspore plot displays a very good linear relationship, whereas the ISERS vs.Cspore plot is linear only in a limited concentration range (>3 × 10−13 M). From the linear relationship of ISERS vs. Cspore in the higher concentration range, a value of 4.9 × 1015 is found for Kspore × Imax. From the 1/ISERS vs. 1/Cspore linear relationship, the slope and the intercept translate to Imax = 2.0 × 104 , and Kspore × Imax = 9.0 × 1015 , yielding a value of 4.5 × 1011 M−1 for Kspore . It is apparent that in the concentration range 1.6 × 10−13 –1.6 × 10−12 M (Figs. 1 and 2) and 3.1 × 10−14 –1.6 × 10−12 M (Fig. 3), similar Kspore × Imax values (4.6 × 1015 and 4.9 × 1015 ) were obtained from their ISERS vs. Cspore linear relationships, and similar Kspore values (6.5 × 1011 M−1 and 4.5 × 1011 M−1 ) were obtained from their 1/ISERS vs. 1/Cspore linear relationships. Note that a value of 1.7 × 1013 M−1 [3] or 9.0 × 1013 M−1 [4] was reported by Van Duyne and coworkers for Kspore in the 1/ISERS vs. 1/Cspore linear relationship for an AgFON substrate in the spore concentration range of 1.0 × 10−14 –6.0 × 10−13 M [3,4]. It is evident that the Kspore value for our Au NP based SERS substrate is about two orders of magnitude lower than that reported for the AgFON substrates, reflecting likely the difference in the spore biomarker’s surface adsorption between these two types of substrates.
3.2. CaDPA samples from As-synthesized CaDPA In order to compare the SERS spectra of the CaDPA extracted from the spore suspension using 0.02 M HNO3 , CaDPA was also synthesized according to a reported procedure [38], details of which were described in Section 2. By comparing the Raman spectra for powder samples of DPA and CaDPA, the peak characteristics and the peak shifts are basically consistent with the literature report [43]. We note however that there were subtle differences in SERS bands between the as-synthesized CaDPA and the spore-released CaDPA, indicating the complexity of the spore-released biomarkers. While the origin of the spectral differences is not clear at this time, we believe that the presence of possible species other than CaDPA might have influenced the orientations of DPA or CaDPA
158
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Fig. 1. SERS spectra in 500–1800 cm−1 region (A) with an AuNPs(60 nm)/PVP/Au substrate for a spore suspension after extraction using 0.02 M HNO3 . (Cspore [M] (from (a) to (e)) = 1.6 × 10−13 (a), 3.9 × 10−13 (b), 7.8 × 10−13 (c), 1.2 × 10−12 (d), and 1.6 × 10−12 (e).) The spectra in the 900–1100 cm−1 region are shown in (B). (Unit for the scale bars: cps.)
adsorbed on the SERS substrate, which would be an interesting area of a further fundamental investigation. Fig. 4 shows a representative set of SERS spectra for a series of dilutions of the as-synthesized CaDPA. The main diagnostic peak is observed at 1001 cm−1 . It is interesting to note that the 1032 cm−1 peak is very small in comparison with those in the SERS spectra for CaDPA extracted from the spore suspension using 0.02 M HNO3 (Figs. 1 and 3A). The concentration dependent peak intensity is analyzed similarly to the analysis of the SERS spectra for the spore samples. Under the simplified (KCaDPA CCaDPA 1) and un-simplified conditions, the above Langmuir adsorption isotherms (i.e., Eqs. (1) and (2)) can again be used, where Cspore is substituted by the concentration of CaDPA (CCaDPA ), and Kspore by the adsorption equilibrium constant of CaDPA (KCaDPA ). Fig. 5 shows a plot of the SERS intensity of the 1001-cm−1 peak as a function of the CaDPA concentration in terms of ISERS vs. CCaDPA and 1/ISERS vs. 1/CCaDPA . In the ISERS vs. CCaDPA plot (Fig. 5A), the data do not exhibit a linear relationship. The non-linearity characteristic
is also true by dropping the two lowest and two highest concentration points (see insert). This observation is in sharp contrast to the observation of ISERS vs. Cspore linearity as shown in Figs. 1 and 3A for the spore samples. In the 1/ISERS vs. 1/CCaDPA plot (Fig. 5B), the linearity is observed only for concentrations higher than 6.0 × 10−7 M. It is evident that the two lowest concentration points deviate from the linear relationship (Fig. 5B). The slope and intercept translate to Imax = 9.2 × 105 , and KCaDPA × Imax = 1.2 × 109 , yielding a value of 1.3 × 103 M−1 for KCaDPA in the concentration range of CaDPA (6.0 × 10−7 –1.0 × 10−2 M). This KCaDPA value is consistent with that reported previously for DPA in a similar concentration range, but somewhat smaller than that reported for the AgFON based SERS substrate (KCaDPA = 9.0 × 103 M−1 ) in the concentration range of 2.0 × 10−6 –1.0 × 10−3 M [3], reflecting likely a difference between the adsorption equilibria on the two SERS substrates. In Fig. 6, the detailed SERS spectral characteristics between the spore samples extracted using nitric acid and the as-synthesized
Fig. 2. SERS intensity of the 1001-cm−1 peak as a function of Cspore : ISERS vs. Cspore (A) and 1/ISERS vs. 1/Cspore (B).
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
159
Fig. 3. (A) SERS spectra in 900–1100 cm−1 region for spore suspension extracted by 0.02 M HNO3 on an AuNPs(60 nm)/PVP/Au substrate. (Cspore [M] (from (a) to (h)) = 3.1 × 10−14 (a), 6.2 × 10−14 (b), 1.2 × 10−13 (c), 1.6 × 10−13 (d), 3.9 × 10−13 (e) 7.8 × 10−13 (f), 1.2 × 10−12 (g), and 1.6 × 10−12 (h).) (B) SERS intensity of the 999-cm−1 peak as a function of Cspore : ISERS vs. Cspore (a) and 1/ISERS vs. 1/Cspore (b).
CaDPA samples are further compared in both high and low concentrations. While the general spectral characteristics are similar, a major distinction is that the peak width for the CaDPA extracted from spore samples is narrower than that observed for the
as-synthesized CaDPA samples. This difference is believed to reflect the existence of surface competitive adsorptions of chemical species on the Au NP based substrate between the spore solution after acid extraction of CaDPA and the solution of the as-prepared
Fig. 4. SERS spectra in 500–1800 cm−1 region for CaDPA of different concentrations in the presence of 0.02 M HNO3 with an AuNPs(60 nm)/PVP/Au substrate (A). (CCaDPA [M] (from (a) to (i)) = 3.0 × 10−7 (a), 4.0 × 10−7 (b), 6.0 × 10−7 (c), 1.0 × 10−6 (d), 5.0 × 10−6 (e), 1.0 × 10−5 (f), 1.0 × 10−4 (g), 1.0 × 10−3 (h), 1.0 × 10−2 (i).) A magnified view of SERS spectra in the 900–1100 cm−1 region is shown in B. (Unit for the scale bars: cps.)
160
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Fig. 5. SERS intensity of the 1001-cm−1 peak as a function of CCaDPA : ISERS vs. CCaDPA (A) and 1/ISERS vs. 1/CCaDPA (B). (Linear regression without the two low concentration points (B), R2 = 0.99, slope = 8.3 × 10−10 , int. = 1.1 × 10−6 .)
CaDPA. We believe it is unique for our Au NP based substrates, which is possibly originated from differences in the adsorption modes for the biomarker molecules on the SERS substrates. In our previous study of DPA [7], we showed different linearity characteristics for I vs. C depending on the concentration regions. By plotting the previous data [7] in terms of 1/ISERS vs. 1/CDPA , nonlinearity was observed for CDPA < 0.01 ppm (i.e., 6.0 × 10−8 M) and CDPA < 1 ppm (i.e., 6.0 × 10−6 M) whereas a linearity was observed for CDPA > 1 ppm. It is apparent that in the high concentration range (CDPA > 1 ppm) the data display a 1/ISERS vs. 1/CDPA linear relationship. The slope and intercept (1.5 × 10−8 and 1.8 × 10−5 ) translate to Imax = 5.5 × 104 , and KDPA × Imax = 6.6 × 107 , yielding a value of 1.2 × 103 M−1 for KDPA . To ensure the reproducibility, the concentration dependence of DPA was also examined for the SERS substrate prepared in the current study. The observed main peaks at 1000 and 1030 cm−1 are in agreement with our
previous reports [7,8]. From the linear relationship of I vs. C, a value of 1.9 × 103 M−1 was obtained for KDPA . From the 1/I vs. 1/CDPA linear plot, the slope and the intercept translate to Imax = 5.5 × 104 , and KDPA = 1.2 × 103 M−1 , practically the same as 1.3 × 103 M−1 for KCaDPA obtained in the current work under the similar concentration range, but different from that reported for AgFON substrate [3,4]. Another possible factor is the difference in adsorption mode of CaDPA or DPA between Au NP and Ag NP based substrates. As we described previously [7], the 1000-cm−1 peak position for the ringbreathing mode of DPA was found to be somewhat lower than those observed for SERS substrates with Ag colloids, e.g., 1020 cm−1 [3–6] and 1010 cm−1 [9,10]. This difference is believed to reflect the difference of the surface adsorption sites and the surface orientation of the adsorbed DPA molecules between Au and Ag particles. The surface orientation of DPA adsorbed on the Au NP was believed to
Fig. 6. Comparison of SERS spectra (in 500–1800 cm−1 region) between acid-extracted spore samples (a) and as-synthesized CaDPA samples (b) under two different concentrations: (A) spore sample ((a) 1.6 × 10−13 M, extracted with 0.02 M HNO3 ) and CaDPA ((b) 6.0 × 10−7 M in the presence of 0.02 M HNO3 ); (B) spore sample ((a) 1.6 × 10−12 M extracted with 0.02 M HNO3 ) and CaDPA ((b) 1.0 × 10−5 M in the presence of 0.02 M HNO3 ).
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
161
be possibly an edge-on mode via its aromatic ring interacting with the surface, which may account for the significant enhancement for the ring-breathing mode. Since DPA is adsorbed on the surface via carboxylate groups, the interaction between carboxylate groups and alumina or silver surfaces of the AgFON substrate [3–5] could be stronger than the interaction between DPA and the Au NP substrate because of the ionic character of the bonding between the negatively charged carboxylates and the positively charged Al(III) or Ag(I) surface species. In the case of DPA on Au surface, the adsorption occurs likely through a coordination interaction via its pyridine ring with Au surface [8], which should be weaker than the electrostatic interaction. Such a difference in the adsorption mode is likely responsible for the difference between KCaDPA (1.3 × 103 M−1 ) determined in this work and KCaDPA (9.0 × 103 M−1 ) reported for AgFON in the previous work [3]. 3.3. Assessment of extraction efficiency and detection limit Table 1 summarizes the data determined from the above linear relationships, including Kspore and KCaDPA . In these experiments spore samples were diluted from a stock solution with a concentration of 1.6 × 10−12 M and spore density of 7.9 × 1010 spores g−1 . Using Kspore = 4.5 × 1011 M−1 and KCaDPA = 1.3 × 103 M−1 , moles of CaDPA extracted from the spores can be obtained from the ratio of Kspore /KCaDPA , which yields 3.5 × 108 moles of CaDPA extracted from the spores. Since the mass of CaDPA (WCaDPA = 3.5 × 108 × (167 + 40)) is 7.2 × 1010 g and the mass of spores (Wspore = 6.02 × 1023 /(7.92 × 1010 spores g−1 )) is 7.6 × 1012 g, an extracted percentage (7.2 × 1010 g/7.6 × 1012 g) of 1.0% is obtained, which gives an extraction efficiency (1.0%/8.9%) of 11%. A slightly different value of 15% was also obtained for the extraction efficiency from another set of data (Table 1), indicating some degree of variability of the determined extraction efficiency value. In the early report [3] for spore samples with a spore concentration of 3.1 × 10−13 M and spore density of 5.6 × 1010 spores g−1 , the calculation using Kspore (1.7 × 1013 M−1 ) and KCaDPA (9.0 × 103 M−1 ) gave a value of 1.9 × 109 M CaDPA. Using WCaDPA (1.9 × 109 × (167 + 40)) = 3.9 × 1011 g and Wspore 23 (6.02 × 10 /(5.6 × 1010 spores g−1 )) = 1.1 × 1013 g, the calculated extracted percentage (3.9 × 1011 g/1.08 × 1013 g) and the extraction efficiency (3.0%/8.9%) give 3.0% and 34%, respectively. Note that in a later report by the same authors [4], Kspore = 9.0 × 1013 M−1 and KCaDPA = 4.9 × 104 M−1 were obtained. In this case, Kspore /KCaDPA ratio is 1.8 × 109 , which is similar to that reported in their early work [3]. Apparently, the extraction efficiency (34%) is somewhat larger than what was observed in our experiment (11%), probably due to difference in experimental condition for the preparation of spore samples. The limit of detection (LOD) was also estimated from the SERS data for the lowest concentration by considering a minimum signal being above three times of the noise level. Based on the data corresponding to the lowest spore concentration (3.1 × 10−14 M) (not shown), the estimate yielded a value of 2.5 × 10−14 M for LOD. By linear extrapolation in this region with the consideration of 3 times of noise level, this value corresponds to LOD for the detection of the spores, which translates to 3.0 × 103 spores per 0.2 L sample solution in comparison with the LOD value reported for the spore detection using AgFON substrates (1.1–2.1 × 10−14 M or 1.4–2.6 × 103 spores per 0.2 L) [3,4]. For the as-synthesized CaDPA samples corresponding to the lowest CaDPA concentration (3.0 × 10−7 M) (not shown), the linear extrapolation in this region with the consideration of 3 times of noise level yielded a value of 1.3 × 10−7 M for the LOD. This value is one order of magnitude lower than the LOD reported for the CaDPA detection using AgFON substrates (3.1 × 10−6 M in 0.2 L [3] or 1.9 × 10−6 M in 0.2 L [4]). In Table 2, the LOD data from
Fig. 7. Comparison of SERS spectra for spore samples aged for different lengths of time without using reagent for extraction: (a) blank (0.02 M HNO3 ), (b) 2 months, (c) 5 months (supernatant after centrifugation), (d) 5 months (after stirring), (e) 1 year (50% dilution) and (f) fresh spore suspension extracted by 0.02 M HNO3 . The concentration of stock solution of spores: 1.6 × 10−12 M for (b), (c), and (d), 2.6 × 10−12 M for (e), and 1.6 × 10−12 M for (f).
our SERS detection using Au NP based substrates are summarized, in comparison with the LOD values reported for SERS detection of spores and CaDPA using AgFON substrates [3,4]. In view of the relatively large difference of adsorption equilibrium constants between our Au NP substrates and the previous AgFON based substrates, it is remarkable that the difference in LOD in terms of spore concentration is relatively small, with a value of 3.0 × 103 spores per 0.2 L for Au NP substrates, and 1.4–2.6 × 103 spores per 0.2 L for AgFON substrates [3,4]. In comparison, the difference in LOD in terms of as-prepared CaDPA concentration is much greater, with a value of 1.3 × 10−7 M for Au NP substrates, and 1.9–3.1 × 10−6 M for AgFON substrates [3,4]. 3.4. Detection of spores samples without extraction acid In addition to extraction of CaDPA from spore samples using 0.02 M HNO3 , the possible self-release of CaDPA from the spore suspension was also examined as a function of storage time without adding HNO3 or sonication. CaDPA was found to be released slowly from spore suspension. In Fig. 7, a representative set of SERS spectra for several samples of spores are compared, including spore samples after storage under 4 ◦ C for different time lengths. While there are small differences of the detected bands among the samples aged for different time lengths, these samples showed the diagnostic band at ∼1000 cm−1 characteristic of CaDPA. The peak width is similar to that observed for the spore samples after acid extraction. Since there is no intentional means being applied for the extraction, the detected CaDPA was termed as “self-released” CaDPA from the spores. The estimated concentrations of the equivalent spores corresponding to the naturally released CaDPA after ageing for different time lengths are shown in Table 3.
162
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163
Table 1 A summary of Kspore and KCaDPA , and extraction efficiency for CaDPA from spore samples. K (M−1 )
Expt.
C
Spore – (a) Spore – (b) CaDPA In Ref. [3]
7.9 × 10 spore g 7.9 × 1010 spore g−1 3.0 × 10−7 –1.0 × 10−2 M CD 5.6 × 1010 spore g−1 3.1 × 10−6 –1.3 × 10−3 M CDc
−1
10
6.5 × 10 4.5 × 1011 M−1 1.3 × 103 M−1 1.7 × 1013 M−1 9.0 × 103 M−1 11
E.% a
E.E.% b
1.4% 1.0%
15% (a) 11% (b)
3.0%
34%
Note: Spore conc. = 1.6 × 10−12 M (9.4 × 108 spores mL−1 ); spore density = 7.9 × 1010 spores/g; laser power = 0.1 mW. K refers to Kspore or KCaDPA . a E% represents % of extraction. b E.E.% refers to extraction efficiency. c CD refers to CaDPA. Table 2 A summary of LODs from our SERS detection using Au NP based substrates in comparison with those reported using AgFON substrates. LODCaDPA a
SERS substrate
LODspore a
AgFON [3] AgFON [4] AuNPs/Au
2.6 × 10 spore (2.1 × 10 M) 1.4 × 103 spore (1.1 × 10−14 M) 3 −14 3.0 × 10 spore (2.5 × 10 M) 3
−14
−6
3.1 × 10 M 1.9 × 10−6 M 1.3 × 10−7 M
Plaser (mW) 50 50 0.1
Note: a 0.2 L 0.02 M HNO3 .
For a fresh solution of spore samples, the self-released CaDPA (e.g., after 2–5 months) corresponded to one third or fourth of the equivalent spore concentration achieved by the acid extraction. For an aged spore solution, the self-released CaDPA (e.g., after 1 year) corresponded to a spore concentration which is almost the same as the equivalent spore concentration achieved by the acid extraction. To understand the mechanism responsible for the self-release of CaDPA from spores and its implication for the SERS detection, we considered two possible pathways. The first involved considering the possibility of laser-induced release. However, several control experiments with different time lengths of exposures by the laser power used for the SERS measurement did not reveal significant release of measurable signals. This ruled out the possibility of laserinduced release. The second pathway involved considering some possible conditions for germination, denaturation and inactivation. It is known that the spore inner membrane could become permeable during germination in which ion fluxes and CaDPAs are released from inside the spore as a result of spore re-hydration. In a recent report [44], some non-nutrient agents and nutrient germinants bound to specific receptors located in the spore’s inner membrane were shown to trigger the release of ∼10% spore (dry weight) of DPA which is located exclusively in the spore’s central region or core as a 1:1 chelate with divalent cations (CaDPA). CaDPA release triggers hydrolysis of the spore’s peptidoglycan (PG) cortex, which then allows swelling of the spore core and further water uptake, resulting in a core water content similar to that in growing cells. In another recent report [45], it is pointed out that spores can rapidly return to active growth through germination followed by outgrowth although spores of bacteria of the Bacillus species
Table 3 Estimated concentrations of the equivalent spores corresponding to the naturally released CaDPA after ageing for different time lengths (based on linear relationship in Fig. 3B(b)). Spore samples fresh stock suspension (b) 2 months (c) 5 month (cent’d) (d) 5 month (stirred) (e) 1 year (50%)b
SERS peak & Ia −1
999 cm , 7962 999 cm−1 , 3306 1000 cm−1 , 4397 1000 cm−1 , 4856 999 cm−1 , 7461
Conc. of spore (M) 1.6 × 10−12 4.4 × 10−13 6.2 × 10−13 7.1 × 10−13 1.3 × 10−12
Note: The stock solution was not subjected to acid extraction, and was not homogenized in the bottle. a I refers to integrated area. b Curve (e) in Fig. 7 is 50% of the stock solution (2.6 × 10−12 M) after ∼1 year.
are very resistant to a variety of harsh conditions and can survive for many years. Wet-heat treatment in spore killing is associated with protein denaturation and enzyme inactivation, which results in the release of the spore core’s CaDPA through rupture of the spore’s inner membrane. It is therefore believed that a combination of non-nutrient/nutrient induced germination and wet-heat induced denaturation could have played an important role in the release of CaDPA. 4. Conclusions In conclusion, the Au NP based SERS substrates have been shown to function as an effective probe for the highly sensitive detection of biomarker CaDPA released from B. subtilis spores. The LOD (1.5 × 109 spores L−1 ) for our Au NP based SERS substrates has been found to be quite comparable with that reported for Ag NP based SERS substrates. This finding is remarkable in view of the fact that the surface adsorption equilibrium constant for the Au NP based SERS substrates is smaller by a factor of 1–2 order magnitude than that for Ag NP based SERS substrates. The results have also revealed the viability of SERS detection of CaDPA released from the spores under ambient conditions without extraction using any reagents. This is an intriguing finding in view of the detection of CaDPA released from the spores under the ambient condition without any extraction procedures. One important implication of this finding is the viability of the Au NP based SERS substrates as a biocompatible probe with high resolution and sensitivity for rapid and sensitive detection of bacteria spores with minimized interferences from possible co-existent species in biological samples. Acknowledgements This work is financially supported by the National Natural Science Foundation of China (Grant Numbers: 21035001, 20775023, 20605007, and 20675005) and “973” National Key Basic Research Program of China (2007CB310500, and 2011CB911000). References [1] S.J. Park, T.A. Taton, C.A. Mirkin, Science 295 (2002) 1503–1506. [2] C.L. Haynes, A.D. McFarland, R.P. Van Duyne, Anal. Chem. 77 (2005) 338A–346A. [3] X.Y. Zhang, M.A. Young, O. Lyandres, R.P. Van Duyne, J. Am. Chem. Soc. 127 (2005) 4484–4489. [4] X.Y. Zhang, J. Zhao, A.V. Whitney, J.W. Elam, R.P. Van Duyne, J. Am. Chem. Soc. 128 (2006) 10304–10309.
H.-W. Cheng et al. / Analytica Chimica Acta 707 (2011) 155–163 [5] X.Y. Zhang, N.C. Shah, R.P. Van Duyne, Vib. Spectrosc. 42 (2006) 2–8. [6] C.R. Yonzon, X.Y. Zhang, J. Zhao, R.P. Van Duyne, Spectroscopy 22 (2007) 42–56. [7] H.W. Cheng, S.Y. Huan, H.L. Wu, G.L. Shen, R.Q. Yu, Anal. Chem. 81 (2009) 9902–9912. [8] H.W. Cheng, W.Q. Luo, G.L. Wen, S.Y. Huan, G.L. Shen, R.Q. Yu, Analyst 135 (2010) 2993–3001. [9] S.E.J. Bell, J.N. Mackle, N.M.S. Sirimuthu, Analyst 130 (2005) 545–549. [10] J.W. Lee, H.B. Lee, K. Kim, Anal. Bioanal. Chem. 397 (2010) 557–562. [11] J.K. Daniels, G. Chumanov, J. Phys. Chem. B 109 (2005) 17936–17942. [12] J.K. Daniels, T.P. Caldwell, K.A. Christensen, G. Chumanov, Anal. Chem. 78 (2006) 1724–1729. [13] D.D. Evanoff Jr.J., T.P. Heckel, K.A. Caldwell, G. Christensen, Chumanov, J. Am. Chem. Soc. 128 (2006) 12618–12619. [14] J. He, X.F. Luo, S.W. Chen, L.L. Cao, M. Sun, Z.N. Yu, J. Chromatogr. A 994 (2003) 207–212. [15] L.H. Harrison, J.W. Ezzell, T.G. Abshire, S. Kidd, A.F. Kaufmann, J. Infect. Dis. 160 (1989) 706–710. [16] L.J. Kricka, P. Fortina, Clin. Chem. 47 (2001) 1479–1482. [17] S.E.J. Bell, N.M.S. Sirimuthu, Chem. Soc. Rev. 37 (2008) 1012–1024. [18] K. Kneipp, H. Kneipp, J. Kneipp, Acc. Chem. Res. 39 (2006) 443. [19] J. Kneipp, H. Kneipp, K. Kneipp, Chem. Soc. Rev. 37 (2008) 1052–1060. [20] M.D. Porter, R.J. Lipert, L.M. Siperko, G. Wang, R. Narayanana, Chem. Soc. Rev. 37 (2008) 1001–1011. [21] S. Nie, S. Emory, Science 275 (1997) 1102. [22] J.D. Driskell, R.J. Lipert, M.D. Porter, J. Phys. Chem. B 110 (2006) 17444–17451. [23] P.N. Njoki, S. Lim, D. Mott, H.Y. Park, B. Khan, S. Mishra, R. Sujakumar, J. Luo, C.J. Zhong, J. Phys. Chem. C 111 (2007) 14664–14669. [24] S. Lim, P.N. Njok, H.Y. Park, X. Wang, L. Wang, D. Mott, C.J. Zhong, Nanotechnology 19 (2008) 305102. [25] A. Gultekin, A. Ersoz, D. Hur, N.Y. Sariozlu, A. Denizli, R. Say, Appl. Surf. Sci. 256 (2009) 142–148.
163
[26] A. Dhawan, Y. Du, F. Yan, M.D. Gerhold, V. Misra, T. Vo-Dinh, IEEE Sens. J. 10 (2010) 608–616. [27] G.C. Zhu, Y.J. Hu, J. Gao, L. Zhong, Anal. Chim. Acta 697 (2011) 61–66. [28] M.A. Woo, S.M. Lee, G.S. Kim, J.H. Baek, M.S. Noh, J.E. Kim, S.J. Park, A.M. Tehrani, S.C. Park, Y.T. Seo, Y.K. Kim, Y.S. Lee, D.H. Jeong, M.H. Cho, Anal. Chem. 81 (2009) 1008–1015. [29] J. Kneipp, H. Kneipp, A. Rajadurai, R.W. Redmond, K. Kneipp, J. Raman Spectrosc. 40 (2009) 1–5. [30] J.A. Lindsay, T.C. Beaman, P. Gerhardt, J. Bacteriol. 163 (1985) 735–737. [31] P.S. Margolis, A. Driks, R. Losick, J. Bacteriol. 174 (1993) 528–540. [32] R.E. Marquis, G.R. Bender, J. Bacteriol. 161 (1985) 789–791. [33] W.L. Nicholson, P. Setlow, Sporulation, germination and outgrowth, in: C.R. Harwood, S.M. Cutting (Eds.), Molecular biological methods for Bacillus., John Wiley and Sons, Chichester, England, 1990, pp. 391–450. [34] M. Paidhungat, P. Setlow, J. Bacteriol. 182 (2000) 2513–2519. [35] D.L. Popham, S. Sengupta, P. Setlow, Appl. Environ. Microbiol. 61 (1995) 3633–3638. [36] Y. Rotman, M.L. Fields, Anal. Biochem. 22 (1967) 168. [37] R. Tennen, B. Setlow, K.L. Davis, C.A. Loshon, P. Setlow, J. Appl. Microbiol. 89 (2000) 1–10. [38] G.F. Bailey, S. Karp, L.E. Sacks, J. Bacteriol. 89 (1965) 984–987. [39] K.R. Brown, M.J. Natan, Langmuir 14 (1998) 726–728. [40] Q. Zhou, Q. Fan, Y. Zhuang, Y. Li, G. Zhao, J.W. Zheng, J. Phys. Chem. B 110 (2006) 12029–12033. [41] X.J. Liu, S.G. Huan, Y. Bu, G.L. Shen, R.Q. Yu, Talanta 75 (2008) 797–803. [42] G.K. Liu, J. Hu, P.C. Zheng, G.L. Shen, J.H. Jiang, R.Q. Yu, Y. Cui, B. Ren, J. Phys. Chem. C 112 (2008) 6499–6508. [43] A.A. Kolomenskii, H.A. Schuessler, Spectrochim. Acta A 61 (2005) 647–651. [44] L.X. Peng, D. Chen, Y.Q. Li, P. Setlow, Anal. Chem. 81 (2009) 4035–4042. [45] P.F. Zhang, L.B. Kong, P. Setlow, Y.Q. Li, Appl. Environ. Microbiol. 76 (2010) 1796–1805.