Polyaniline–polystyrene membrane for simple and efficient retrieval of double-stranded DNA from aqueous media

Abstract

Extraction of nucleic acids from biological samples is a necessary step in almost all biotechnological procedures. However, while DNA-containing samples are usually very distinct in origin, no unique common methodology for performing these tasks is currently available and no single material that could satisfy the necessary requirements to carry them out exists. Consequently, the development of new systems or methods devoted to the improvement of DNA extraction protocols in terms of yield, purity, time, and cost continues to be of great interest. In this work, we describe the preparation of a novel type of polyaniline/polystyrene (PANI/PS) membrane that not only exhibits high levels of capture of DNA strands dissolved in aqueous solutions but also allows for their simple and efficient posterior release. The composite membrane consists of electrospun PS fibres that were modified by the incorporation of PANI nanograins obtained through an in situ chemical polymerization. We have used sodium salt of Salmon Sperm DNA aqueous solutions as a model system to examine how large would be the adsorption capacity of these membranes towards nucleic acid molecules. In batch-mode experiments, we have found a maximum adsorption capacity of 153.8 mg of DNA per gram of PANI, with the saturation limit attained within only 9 min. In addition, we have demonstrated that it is possible to achieve the complete release of the adsorbed DNA by simple changes on the pH of the aqueous medium. We also performed DNA adsorption experiments by adapting the composite membrane to a spin column. The corresponding results indicate that in this experimental setup the PANI/PS composite membranes appear as a promising material for the purification of nucleic acids and other valuable biomolecules.

---

1. Introduction

DNA molecules play a fundamental role in the control of all sustaining life processes of a given organism and in the hereditary instruction encoding for the next generation. In the last few decades, the ability to manipulate and analyse DNA sequences has gained huge scientific and social relevance in a wide range of disciplines and applications, such as genetics,¹ pharmacology,² diagnosis³ and forensic medicine.⁴ In all of them, a cornerstone is the search for improved methods for DNA extraction and purification.

Nowadays, one can usually separate the DNA fraction from a given biological sample by adopting two different methodologies based either on liquid–liquid or solid-phase extractions.⁵ Examples of the liquid–liquid methodologies, which imply the consecutive use of organic solvents and a series of precipitation and centrifugation steps to extract the DNA, are the widely adopted phenol–chloroform extraction method and the CTAB procedure.^6,7 Meanwhile, solid-phase extraction methods involve the use of beads, columns, and monoliths composed by silica, diatoms, cellulose, or resins to trap the DNA.^8–10 In spite of the prevalence of those methodologies in the field work, they are not free of some drawbacks, such as the need of a strict adherence to complicated, time-consuming, and labour-intensive protocols, the risk of cross contamination, and (in special, for the case of liquid–liquid methods) use of toxic chemicals.¹¹ More recently, nanostructured systems with improved affinities towards DNA strands have been developed. These nanomaterials could enhance the yield and purity of the retrieved target, while simultaneously reducing the need of using toxic substances and lowering the cost of the solid-phase extraction methods. Use of magnetic nanocomposites (MNCs) is an important example of the application of these materials for DNA extraction and purification. Typically, the corresponding procedure involves the completion of several steps, in which first the MNCs are placed in a heterogeneous medium containing the target DNA, whose physical-chemical conditions (pH, ionic strength) are then adjusted to foster the interaction between the MNC and the dissolved DNA chains. Subsequently, one can spatially confine the freshly formed MNC–DNA complex by applying an external magnetic field, before submitted them to successive purification steps. Finally, the DNA strands are desorbed from the MNC and later analysed.¹² Usually, those magnetic adsorbents are produced in two separated steps, in which first the magnetic nanoparticles are synthesized and then their surface is coated by a material with high DNA affinity (such as silica or an organic polymer),^11,13 thus forming a core–shell structure. Another approach for DNA extraction corresponds to the use of membrane-based solid matrices. Some examples of these materials are the spin-column (Sigma-Prep™, Wizard miniprep, among others) and FTA cards technologies already available in the market.

In this work, we propose the use of fibrous membranes functionalized with DNA affinity nanostructures as a convenient material for use in novel protocols, which appear as reliable advantageous alternatives to the current DNA extraction methods based on solid matrices. We employed the electrospinning (ES) technique to produce the membranes. ES is a flexible, low-cost and versatile electrorheological procedure that allows the preparation of micro and nanostructured fibres from the most diverse sets of materials.¹⁴ Large surface areas and high porosity degree are intrinsic characteristics of ES membranes. When used as adsorbent agents for the extraction of biomolecules, these materials exhibit superior target flowrates and increased target mass transfer relative to either porous membranes prepared by use of casting methods,¹⁵ or monoliths (porous silica rods¹⁶) commonly adopted for the same purposes.¹⁷ Nonetheless, extraction approaches based on the use of ES membranes have been sparingly explored either for direct DNA extraction¹⁸ or for using DNA adsorption as an intermediate step for gene delivery procedures.^19,20 In this study, we developed a hierarchical material based on an in situ chemical polymerization of aniline atop ES membranes of PS. This configuration of the composite elements favours the DNA adsorption by a synergetic effect of the membrane porosity and the chemical activity of the nanostructured PANI.

2. Experimental methods

2.1. Materials

PS (PM = 280 kDa), ammonium persulfate (APS), sodium dodecyl sulfate (SDS) and Salmon Sperm DNA were purchased from Sigma-Aldrich (USA). Dimethylformamide (DMF) and sodium hydroxide (NaOH) were obtained from Dinâmica (Brazil). Hydrochloric acid (HCl) and aniline (C₆H₅NH₂) were acquired from the Brazilian companies Química Moderna and Nuclear, respectively. Glycine (C₂H₅NO₂) and SYBR Green dye nucleic acid dye were obtained from Promega (USA) and Thermo Fisher Scientific (USA). We used all reagents as received, with exception of the aniline monomer, which was distilled under reduced pressure prior to use. High purity water (Millipore, USA) was used in all experiments.

2.2. Preparation of the composite membranes

PS membranes were obtained by the ES method, and exposed to an air plasma treatment during 3 min using a PDC-002 plasma cleaner (Harrick, USA), as previously described.¹⁴ The PS membranes were then modified through the in situ polymerization of aniline in aqueous media. For this, we initially placed the membranes in a 150 mL beaker containing 37.5 mL of HCl 1 M and 0.52 mmol of aniline. After 15 min, we added 12.5 mL of HCl 1 M containing 0.33 mmol of APS. The polymerization process occurred at 5 °C under constant stirring for 12 h. After this, we repeatedly washed the membranes with deionized water and dried them at room temperature. Finally, the PANI/PS membranes were weighed in order to determine the amount of deposited PANI.

2.3. Characterization methods

The fibres were morphologically characterized and their diameters determined by use of a Quanta 450 FEG scanning electron microscope (FEI, USA). The chemical composition of the membranes was analyzed by FTIR spectroscopy using an IRTracer-100 equipment (Shimadzu, Japan). The wetting properties of the ES membranes were examined by employing a CAM 100 contact angle meter instrument (KSV, Finland). The DNA concentration in solution was determined using a NanoDrop 2000c spectrophotometer (ThermoFisher Scientific, USA). DNA adsorption and desorption were monitored by using a SYBR Green dye (ThermoFisher Scientific, USA) and an Axio Imager 2 microscope (Zeiss, Germany).

2.4. Adsorption experiments

The DNA adsorption capacity of the PANI/PS membranes was evaluated as a function of the interaction time and the initial DNA concentration in the aqueous solution. A stock solution was prepared by dissolving 40 mg of Salmon Sperm DNA in 100 mL of deionized water. In all experiments, we used 2 mL Eppendorf tubes and PANI/PS membranes with dimensions of 0.75 cm × 3.00 cm.

To implement the adsorption experiments we first allowed a glycine/HCl buffer solution (pH 2.8) to permeate a membrane enclosed in an Eppendorf tube, which was subsequently vortexed before the addition of the DNA solution. We adjusted the relative volumes of the glycine/HCl and DNA solutions accordingly to the desired final DNA concentration. A fixed 2 mL total volume of solution was adopted in all cases. The DNA concentrations evaluated in the adsorption experiments were (5, 11, 22, 49, 79, 111, 168) mg L⁻¹. We carried out all the experiments at room temperature using an orbital shaker operating at 250 rpm.

We estimated the amount of adsorbed DNA onto the PANI/PS membrane as

where C₀ and C_f are the initial and final (mg L⁻¹) DNA concentrations in the solution, respectively.

The membrane DNA adsorption capacity at time t, i.e., the amount (in mg) of DNA adsorbed per mass unit (g) of membrane, was denoted as q_t and evaluated as

where V is the total volume of the solution (mL), C₀ is the initial concentration of DNA (mg L⁻¹) in solution, C_t is the corresponding DNA concentration (mg L⁻¹) in the solution at that instant and m is the corresponding mass of PANI deposited in the PS fibres. After the process reaches equilibrium, the final membrane DNA adsorption capacity, q_e, corresponds to the total amount (in mg) of DNA adsorbed per mass unit (g) of membrane, i.e.,where C_e is the equilibrium concentration of DNA (mg L⁻¹) in solution. The experimental data were fitted to the Langmuir and Freundlich isotherm and kinetic models.²¹

2.5. Desorption experiments

In order to study the nucleic acid desorption, we have saturated a PANI/PS membrane with the target DNA, which was subsequently washed with 1 mL of deionized water. Then, we placed the membrane in an Eppendorf tube and added 1 mL of an elution aqueous solution (3.96 μmol of SDS, 8068 μM of NaOH, pH 11.7). Finally, the tube was stirred in an orbital shaker operating at 250 rpm for 10 min and the supernatant collected and analysed by UV-Vis.

2.6. Fluorescence microscopy

The DNA adsorption–desorption process was monitored through fluorescence microscopy using the SYBR Green dye and an Axio Imager 2 microscope equipped with 450–490 nm excitation and 515–565 nm emission windows. For this, 2 μL of dye were added to 2 mL of a 50 mg L⁻¹ glycine/DNA solution. Subsequently, we performed the adsorption–desorption process shown in Sections 2.4 and 2.5.

2.7. DNA extraction by centrifugation

As shown in Scheme 1, we have prepared a device similar to a commercial spin column, after stacking in an Eppendorf tube 4 PS membranes and 4 PANI/PS membranes cut as 7 mm diameter disks. We have evaluated the performance of this setup for the retrieval of Salmon Sperm DNA from an aqueous solution by following a fixed protocol. Initially, 500 μL of glycine/HCl buffer solution (pH 2.8) were added to the tube and centrifuged. After that, the filtered was discarded and we added 500 μL of the 50 mg L⁻¹ DNA/glycine/HCl solution. Then, the tube was once again centrifuged and we measured the UV-Vis absorption to determine the DNA concentration remaining in the solution. Afterwards, we performed a washing step by adding 500 μL of water and centrifuging the solution, and then discarding the filtered solution. Desorption of the DNA retained in the membrane was achieved by adding 500 μL of an elution aqueous solution (3.96 μmol of SDS, 8068 μM de NaOH, pH 11.7) to the tube. Finally, we once again transferred the tube to the centrifuge and the eluted DNA was collected from the bottom of the device. All absorption measurements were performed at λ = 260 nm. At each step, the centrifuge operated at 4000 rpm for 20 s. All DNA adsorption–desorption processes were conducted in duplicate.

Scheme 1 [1] Devised experimental setup for the spin column using PANI/PS membranes. [2] DNA extraction process, (a) add of DNA/glycine/HCl solution, (b) washing process and (c) add of elution aqueous solution.

3. Results and discussion

3.1. Characterization of polyaniline/polystyrene membranes

PANI is well-known as an intrinsically conducting polymer that has been widely applied due to its peculiar properties, such as good electrical conductivity, electrochromic changes, and redox activity, among others.^21,22 However, its characteristic non-solubility and brittleness make relatively hard the further processing of the material into more convenient forms, such as films or fibres.¹⁴ As a manner of overcoming this limitation and exploiting some of desirable properties of the polymer, we have employed the PS membranes as a scaffold to deposit nanostructured PANI chains. The electrospinning technique can provide a simple and effective way to prepare polymeric films that could offer the required mechanical support and several examples of PANI based hybrid polymeric systems have been already reported.²³ For instance, use of PANI/PS blends as electrodialysis membranes for treating wastewater has been suggested before,^24,25 and superhydrophobic PANI/PS films have been proposed as anti-corrosive material for surface protection.²⁶

In Fig. 1a, one can observe that the PS fibres are homogenous, with an average diameter of (1.6 ± 0.4) μm, and no evidence of defects. In the inset of Fig. 1a we show the corresponding measurement of the water contact angle, whose value (125 ± 3)° indicates a high hydrophobicity, as a result from the combination of a low surface energy with the microstructured morphology of the PS fibres.

Fig. 1 SEM micrograph and contact angle (inset) of (a) PS membranes and (b) PANI/PS membranes. (c) Magnification of an individual PANI/PS fibre. (d) FTIR spectra of (a) PS and (b) PANI/PS membranes.

In the in situ deposition of the conducting polymer on the membrane, one has to choose the relative amounts of the reagents in a proper manner: while a high enough amount of aniline could lead to the solubilisation of polystyrene¹⁴ (hence compromising the integrity of the fibres), a low monomer concentration could result in a partial covering of the mat. Hence, we adopted the use of dilute concentrations and a monomer to oxidant ratio that favours both the preferential deposition of the PANI chains on the membranes and the integrity of the membrane. To determine this, we prepared hybrid membranes with varying relative PANI/PS amounts, by using different concentrations of the aniline monomer (0.05, 0.52, 5.00 and 20.00 mmol). In this manner, we found that the PS–PANI membranes obtained by using a 0.52 mmol concentration of the monomer were those to present the better performance for the DNA adsorption (please see Fig. SI.1 in the ESI for more details†).

The PANI-modified PS membranes are shown in Fig. 1b, where it is possible to note that their intrinsically porosity is not lost. This is probably due to our choice of the relative concentrations of reagents and to the use of surface plasma treatment, since both conditions foster the PANI growth along the fibre surfaces. In Fig. 1c, we present a magnification of the surface of an individual PANI/PS fibre, where it is possible to observe PANI nanograins deposited along the fibre axis. Hence, the resulting morphology is a consequence of the presence of carboxylic groups generated during the air plasma treatment and of the use of dilute concentrations.^27–29 The water contact angle of the PANI/PS membranes was determined as 0°, in a confirmation of their strong hydrophilic character.

We have used FTIR analysis to assess the chemical composition of the membranes (Fig. 1d). The PS membrane spectrum is showed in curve (a), where one can observe peaks at 3059 cm⁻¹ and 3026 cm⁻¹, corresponding to the C–H vibration of the aromatic ring,³⁰ and the vibrations at 2920 cm⁻¹ and 2850 cm⁻¹, which are related to the C–H bond stretching. Meanwhile, the peak at 1600 cm⁻¹ is attributed to the C=C ring stretching,³¹ and the peaks at 1492 cm⁻¹ and 1448 cm⁻¹ to the bending vibrations of C–H.³⁰ In the corresponding spectrum of the PANI/PS membranes (curve b), it is possible to observe the characteristic bands of the PANI salt emeraldine state, where the peaks at 1597 cm⁻¹ and 1490 cm⁻¹ are related to the C=C stretching vibration from the quinoid and benzenoid rings, respectively. In 1298 cm⁻¹, one can find the characteristic C–N stretching vibration from the benzenoid ring. Finally, while the band at 1141 cm⁻¹ is attributed to the aromatic C–H in plane bending, the peak at 813 cm⁻¹ represents the 1,4 substituted benzenoid ring.¹⁴

3.2. Adsorption experiments

In its emeraldine salt form, PANI is a polycation capable of interacting with anions and polyanions such as DNA macromolecules. By taking advantage of this property, in a previous work we reported the development of a PANI/γ-Fe₂O₃ nanocomposite and its use as a solid-phase adsorbent that allowed the magnetic retrieval of DNA from aqueous solutions.¹² In the present work, we advance further along this line by investigating the adsorption properties of a new DNA retrieval system based on the use of PANI/PS membranes.

Effect of the interaction time and of the initial DNA concentration. The time allotted for the interaction to occur and the initial DNA concentration are two important parameters that determine not only how fast the saturation of the composite membrane would happen, but also its maximum adsorption capacity. To study those parameters, we first assured that the composite presented a positive charge, by wetting the PANI/PS membranes with the glycine/HCl buffer. After that, the DNA solution was added to the composite to allow the formation of the DNA–PANI/PS complex. For all DNA concentrations shown in Fig. 2, one can observe that the degree of DNA adsorbed per unit time is higher in the first minutes and tends to saturate at long enough interaction time, so that 9 min or less were enough for the system to reach equilibrium. We attribute this behaviour to the existence of a larger number of empty sites on the surface of the membrane during the first minutes of interaction. In Fig. 3, we show the adsorption capacity and adsorption percentage as functions of the initial DNA concentration. In these plots one can appreciate in more detail that higher amounts of DNA were retrieved when more concentrated solutions are used, a fact that can be explained by the increase in the driving force required to permit the mass transfer between the adsorbate and the adsorbent.³²

Fig. 2 Effect of interaction time on the degree of DNA adsorbed by the PANI/PS membranes for (a) 5 mg L⁻¹, (b) 11 mg L⁻¹, (c) 22 mg L⁻¹, (d) 49 mg L⁻¹, (e) 79 mg L⁻¹, (f) 111 mg L⁻¹ and (g) 168 mg L⁻¹ DNA concentration.

Fig. 3 Effect of the initial DNA concentration (■) on the adsorption degree and (●) the q_e of the PANI/PS membranes.

Adsorption isotherms. To allow for a better understanding of the process involving the DNA adsorption onto the PANI/PS membrane surface, we examined how well we could fit our experimental data by use of the Langmuir and Freundlich isotherm models. The Langmuir isotherm model assumes that the adsorption process involves a finite number of identical and energetically equivalent sites, resulting in the formation of an adsorbed monolayer. Instead, the Freundlich isotherm model considers the existence of active sites of different natures and that the adsorption occurs in multilayers.¹²

In their linearized forms, one can express these models as

andwhere C_e (mg L⁻¹) is the concentration of DNA at equilibrium, q_e (mg g⁻¹) is the DNA adsorption capacity, b (L mg⁻¹) is the Langmuir constant, q_m (mg g⁻¹) is the maximum adsorption capacity of DNA, and K_F and n are Freundlich constants related with the adsorption capacity (mg g⁻¹) and the intensity of the PANI/PS membrane, respectively. The values of the constants of each isotherm were determined from the slope and intercept of the plots C_e/q_e vs. C_e and log q_e vs. log C_e, respectively (see Fig. SI.2†). The corresponding results, which are summarized in Table 1, allowed us to conclude that the Langmuir model best fits our data (with a correlation coefficient of 0.99), a result that agrees with the idea that the PANI amine groups are the sole active sites that adsorb the DNA strands. We then estimated the maximum adsorption capacity as equal to 153.8 mg of DNA per gram of PANI. In Fig. 4, we show the fitting of our experimental data by the Langmuir and Freundlich isotherms. In the ESI,† we present the corresponding linear fittings.

Table 1 Langmuir and Freundlich isotherms parameters for DNA adsorption onto PANI/PS NW membranes

Langmuir			Freundlich
qm (mg g^−1)	b (L mg^−1)	R^2	KF (mg g^−1)	1/n	R^2
153.8	0.112	0.99	39.8	0.26	0.95

---

Fig. 4 DNA adsorption isotherms of the experimental data for (■) PANI/PS membranes and ([thick line, graph caption]) Langmuir and ([dash dash, graph caption]) Freundlich fittings.

As a manner of evaluating the relative performance of the PANI/PS membrane, in Table 2 we list the adsorption and desorption capacities of some materials previously employed for removing Salmon Sperm DNA. One can easily see that the membranes here discussed exhibit an excellent compromise between the values of the adsorption/desorption capacities and the total time required to complete these corresponding tasks.

Table 2 Comparison of the PANI/PS NW membrane performance with those of other DNA adsorbents^a

Adsorbent	Adsorption capacity (mg g^−1)	Adsorption time (min)	% desorption	Desorption time (min)	Ref.
a MSP: mesoporous; MNC: magnetic nanocomposite.
Polymeric ionic liquid microspheres	190.7	1	80.7	—	34
MSP ZrBF	238.6	480	89.0	240	35
MSP Fe3O4/SiO2	46.0	—	—	—	36
MSP Fe3O4/SiO2	121	1200	89.5	60	13
Fe3O4/SiO2	57.8	—	—	—	37
MSP Fe3O4/SiO2	375	—	—	—	38
Allophane spherules	28	—	—	—	39
PANI/γ-Fe2O3 MNC	75.2	10	94	2	12
PANI/PS NW membrane	153.8	9	100	10	This work

---

Adsorption kinetics. The study of the adsorption kinetics is an important manner of investigating the nature of the mechanism involved in the adsorbate–adsorbent interaction. We have fitted our experimental data according to three different kinetic models (pseudo-first order, pseudo-second order and Morris–Weber – see Fig. SI.3 in the ESI†). The first and second of these models are associated with physical and chemical processes and can be expressed in their linearized form as²¹

and

t/q_t = 1/(k₂q_e²) +(1/q_e)t, (8)

respectively. In the above expressions, q_e and q_t are the adsorption capacities (mg g⁻¹) at equilibrium and at time t, k₁ (min⁻¹), k₂ (g mg⁻¹ min⁻¹) are the pseudo-first order, pseudo-second order adsorption kinetic rate constants, respectively, and t is the time (min). As for the Morris–Weber model, it describes diffusive processes and can be written as³³

q_t = k_id(t)^0.5 + C, (9)

where k_id (min⁻¹) and C are the intraparticle diffusion rate constant.

In Table 3, we present the results obtained by fitting our data to each one of these kinetics models. From the corresponding R² values, one can see that the adsorption kinetics of the DNA molecules onto the PANI/PS membrane is best described by the pseudo-second order model (please see the corresponding fitting in the ESI†). Therefore, we can say that the DNA strands follow chemical processes when adsorbing into the PANI/PS membrane. The fitting of the kinetics data to the pseudo-second order model is shown in Fig. 5.

Table 3 Kinetic parameters for DNA adsorption onto PANI/PS NW membrane

qe,exp (mg g^−1)	First-order			Second-order			Morris–Weber
	K1 (min^−1)	qe,calc (mg g^−1)	R^2	k2 (g mg^−1 min^−1)	qe,calc (mg g^−1)	R^2	kid (min^−1)	R^2
153.8	0.20	78.4	0.80	0.014	140.8	0.99	37.3	0.79

---

Fig. 5 (■) Effect of interaction time on the DNA adsorption capacity for the PANI/PS membranes and ([thick line, graph caption]) pseudo-second order fitting.

3.3. Desorption experiments

The DNA strands adsorb onto the PANI covered PS fibres through electrostatic interactions between the positively charged imine groups of the conducting polymer and the negatively charged phosphate groups that compose the external ridge of the nucleotide strands. Hence, we have used a NaOH aqueous solution (pH 11.7) to deprotonate the PANI chains and so foster their desorption from the fibres. However, only 67% of the captured DNA was desorbed when we adopted this procedure. For that reason, and with the expectation that the SDS sulfate anions could help to improve the electrostatic repulsion between the PANI and the phosphates groups of the DNA, we decided to include SDS in the previous alkaline solution; with this, the release of the DNA from the surface of the membrane would be easier. In Fig. 6, we show the UV-Vis spectra collected at every step of this experiment, where the curve C_i [C_f] corresponds to the absorption spectrum of the DNA solution before [after] the adsorption takes place (corresponding to a concentration of 53.9 [41.4] mg L⁻¹). The degree of DNA adsorption onto the fibres is illustrated by the considerable decrease in the absorbance peak at λ = 260 nm. Later, when performing the desorption process (curve C_Des), it was possible to release all DNA absorbed by the membrane. (The concentration of the solution was 25 mg L⁻¹, but the volume used for desorb the DNA was only 1 mL, i.e., half of the volume used in the adsorption process.) For purposes of comparison, we note that when we used an aqueous SDS solution (3.96 μmol, pH 6.0) only 40.6% of the adsorbed DNA was released.

Fig. 6 DNA solution UV-Vis spectra before (C_i) and after (C_f) interaction with the PANI/PS membranes and DNA solution spectrum after the desorption process (C_Des).

3.4. Fluorescence microscopy

We followed the processes of how the DNA adsorbs and desorbs from the membrane surface by employing a fluorescent dye that strongly interacts with both single and double strand nucleic acids, but present as a special characteristic the fact that its fluorescence is remarkably larger in the latter case. In Fig. 7, we present the fluorescence images of PANI/PS membrane before and after of the DNA adsorption and desorption process when in presence of SYBR Green. As one can observe from Fig. 7a and b, respectively, neither the PANI/PS membrane is fluorescent by itself, nor it interacts with (or physically retains) the fluorescent dye. However, as shown in Fig. 7c, a high density of bright green dots can be observed after we exposed the fibres to the DNA-containing solution, confirming that in fact the membrane has captured DNA strands. That assumption was corroborated by the observation that the number of bright dots remaining at the membrane surface decreases after the DNA desorption step (Fig. 7d). This is exactly the result that one would expect if the adsorbed DNA had been successfully eluted in a large degree.

Fig. 7 Fluorescence images of the PANI/PS membrane (a) before interaction with DNA or dye, (b) dyed without DNA adsorption, (c) dyed after DNA adsorption, and (d) dyed after DNA desorption.

3.5. DNA extraction by centrifugation

After confirming that the DNA is indeed captured by the PANI/PS membranes immersed in the aqueous solution (i.e., in the “batch” mode), we tested the performance of these materials as a nucleic acid absorbent after they were incorporated in a spin column setup. For this, we designed the device represented in Scheme 1, where a set of four consecutive non-plasma treated PS disks was disposed in tandem with a set of four PANI/PS disks. Due to their high hydrophobicity (see the inset of Fig. 1a), the pure PS membranes act as a physical barrier that favours the retention of the solution, allowing a better interaction between the DNA and PANI/PS disks. In Fig. 8, we show the UV-Vis absorption spectra of the DNA solution at every step of the experiments performed using the spin column setup. In curve C_i, the band at 260 nm corresponds to a DNA concentration of 50.9 mg L⁻¹. After passing the DNA solution through the spin column, the corresponding intensity (curve C_f) experiments a large attenuation, and in fact, we have estimated the final DNA concentration as corresponding to only 11.2 mg L⁻¹. Hence, 78% of the DNA present in the original solution has been successfully retrieved. After the application of the elution solution, the nucleic acid strands would desorb from the PANI/PS membranes, as confirmed by the data shown in curve C_Des, where the 260 nm band intensity corresponds to a concentration of 28.8 mg L⁻¹, revealing that 72.5% of the total amount of DNA adsorbed has been recuperated into the solution.

Fig. 8 DNA solution UV-Vis spectra before (C_i) and after (C_f) interaction with PANI/PS membranes using a spin column and DNA solution spectrum after the desorption process (C_Des).

If the adsorption results of the batch mode are compared to those of the spin column setup, the latter appears as remarkably more efficient. When using the spin column, we observed that not only the percentage of adsorption was three times higher, but also that the whole procedure was faster than the batch mode. We attribute this fact to an increased degree of diffusion of the aqueous DNA solution onto the membrane. One can take this experiment as a proof of the concept that PANI/PS membranes can be easily adapted to a spin column configuration.

As a test of the stability of the PANI/PS membranes, in a series of experiments we subjected them to several consecutive cycles adsorption–desorption of DNA molecules. In Fig. SI.4,† one can observe that the membrane's ability to adsorb DNA rapidly decreases along the consecutive cycles of use. Even so, the PANI/PS mats preserved their physical integrity during the process (Fig. SI.5†), and could be easily manipulated afterwards. However, to minimize the risks of cross-contamination, as a general rule biotechnology protocols do not recommend the recycling of materials used for the separation or purification of molecules of interest. That was the main reason why we decided not to pursue further the optimization of the membranes characteristics with regard to the improvement of their DNA binding capabilities for use in consecutive adsorption/desorption cycles.

4. Conclusion

In this work, we reported the successful production of PANI/PS membranes that result from the preparation of PS membranes through ES and their subsequent modification by an in situ chemical polymerization of aniline. By following this approach, we could obtain flexible and porous membranes that exhibited physical–chemical properties corresponding to an actual PANI coverage of the nanostructured PS fibres. With this, we were able to overcome at once the tendency of intrinsically conducting polymers to form agglomerates and the brittleness problems normally associated to their use. We tested the DNA adsorption capacity of the PANI/PS membranes as a function of several parameters, such as the interaction time and the initial DNA concentration. In those studies, we have found that when used in batch mode the PANI/PS membranes possess a high capacity of DNA adsorption (153.8 mg g⁻¹) after an interaction time of only 9 min. The degree of desorption was another relevant parameter that we examined, and we have established that 100% of the amount of adsorbed DNA could be released in only 10 min. The relevance of those values arises when they are compared with the capacities of other materials previously reported in the literature: relative to them, the PANI/PS membrane present the best combination of favourable properties such as high adsorption capacity, elevated degree of DNA desorbed by elution and short adsorption–desorption times. We have exploited the fact that one can be easily tailor the PANI/PS membranes into several alternate forms (i.e., bent, cut or compacted) without losing their integrity to adapt them for use in a spin column experimental setup. When we used the mats in this configuration, the corresponding results indicate that not only the overall process was faster, but also that the amount of adsorbed DNA was higher than that obtained in the batch mode.

Therefore, we suggest the integration of these PANI/PS membranes in active components of DNA extraction kits that could become viable alternatives to those currently available. We note that even though one should not expect that in their pristine form the PANI/PS membranes would be selective towards nucleic acid chains, standard protocols for DNA extraction from actual samples usually involve a series of successive steps where enzymes, chaotropic agents and buffer solutions are used for the previous removal of other types of biomolecules. In addition to be cost-effective, these PANI/PS membranes could also be employed in different biomedical applications, such as DNA scaffolds for genetic transfection, biosensors and agents for extraction and purification of other high valuable biomolecules from several media.

Acknowledgements

This work was supported by the Brazilian agencies CNPq, FACEPE and ELINOR Nanobiotechnology network. W. B., J. M., A. C. and J. A. are grateful for FACEPE and Brazilian Health Ministry scholarships. The authors would like to thank the Analytic Center of the Universidade Federal do Ceará for the SEM micrographs.

References

1. S. Branco, P. Gladieux, C. E. Ellison, A. Kuo, K. LaButti, A. Lipzen, I. V. Grigoriev, H.-L. Liao, R. Vilgalys, K. G. Peay, J. W. Taylor and T. D. Bruns, Mol. Ecol., 2015, 24, 2747–2758.
2. B. L. Kieffer, K. Befort, C. Gaveriaux-Ruff and C. G. HiRTH, Proc. Natl. Acad. Sci. U. S. A., 1992, 89, 12048–12052.
3. D. Koontz, K. Baecher, M. Amin, S. Nikolova, M. Gallagher and S. Dollard, J. Clin. Virol., 2015, 66, 95–99.
4. M. Kasu and K. Shires, Leg. Med., 2015, 17, 232–238.
5. S. C. Tan and B. C. Yiap, J. Biomed. Biotechnol., 2009, 2009, 1–10.
6. M. Murray and W. F. Thompson, Nucleic Acids Res., 1980, 8, 4321–4326.
7. G. C. Allen, M. A. Flores-Vergara, S. Krasynanski, S. Kumar and W. F. Thompson, Nat. Protoc., 2006, 1, 2320–2325.
8. W. Jiang, P. Liang, B. Wang, J. Fang, J. Lang, G. Tian, J. Jiang and T. F. Zhu, Nat. Protoc., 2015, 10, 768–779.
9. M. C. Little, Process for the purification of DNA on diatomaceous earth, US Pat., US5,075,430, 1991.
10. E. L. Kendall, E. Wienhold and D. L. DeVoe, Biomicrofluidics, 2014, 8, 044109.
11. J. L. Corchero and A. Villaverde, Trends Biotechnol., 2009, 27, 468–476.
12. J. C. Medina-Llamas, A. E. Chávez-Guajardo, C. A. S. Andrade, K. G. B. Alves and C. P. de Melo, J. Colloid Interface Sci., 2014, 434, 167–174.
13. X. Li, J. Zhang and H. Gu, Langmuir, 2011, 27, 6099–6106.
14. J. J. Alcaraz-Espinoza, A. E. Chávez-Guajardo, J. C. Medina-Llamas, C. A. S. Andrade and C. P. de Melo, ACS Appl. Mater. Interfaces, 2015, 7, 7231–7240.
15. R. Wang, Y. Liu, B. Li, B. S. Hsiao and B. Chu, J. Membr. Sci., 2012, 392–393, 167–174.
16. H. Minakuchi, K. Nakanishi, N. Soga, N. Ishizuka and N. Tanaka, Anal. Chem., 1996, 68, 3498–3501.
17. O. Hardick, S. Dods, B. Stevens and D. G. Bracewell, J. Biotechnol., 2015, 213, 74–82.
18. S. Demirci, A. Celebioglu and T. Uyar, Carbohydr. Polym., 2014, 113, 200–207.
19. S. Sakai, Y. Yamada, T. Yamaguchi, T. Ciach and K. Kawakami, J. Biomed. Mater. Res., Part A, 2009, 88, 281–287.
20. J. Zhang, Y. Duan, D. Wei, L. Wang, H. Wang, Z. Gu and D. Kong, J. Biomed. Mater. Res., Part A, 2011, 96, 212–220.
21. J. Lin and L. Wang, Front. Environ. Sci. Eng. China, 2009, 3, 320–324.
22. G. Ćirić-Marjanović, Synth. Met., 2013, 177, 1–47.
23. F. Miao, C. Shao, X. Li, N. Lu, K. Wang, X. Zhang and Y. Liu, Electrochim. Acta, 2015, 176, 293–300.
24. F. D. R. Amado, E. Gondran, J. Z. Ferreira, M. A. S. Rodrigues and C. A. Ferreira, J. Membr. Sci., 2004, 234, 139–145.
25. F. D. R. Amado, M. A. S. Rodrigues, F. D. P. Morisso, A. M. Bernardes, J. Z. Ferreira and C. A. Ferreira, J. Colloid Interface Sci., 2008, 320, 52–61.
26. Y. Zhu, J. Zhang, Y. Zheng, Z. Huang, L. Feng and L. Jiang, Adv. Funct. Mater., 2006, 16, 568–574.
27. L. Zhang, Y. Long, Z. Chen and M. Wan, Adv. Funct. Mater., 2004, 14, 693–698.
28. H. Gao, J. Zhang, F. Liu, Z. Ao, S. Liu, S. Zhu, D. Han and B. Yang, J. Mater. Chem. B, 2014, 2, 7192–7200.
29. N.-R. Chiou, C. Lu, J. Guan, L. J. Lee and A. J. Epstein, Nat. Nanotechnol., 2007, 2, 354–357.
30. X. Lu, J. Zhou, Y. Zhao, Y. Qiu and J. Li, Chem. Mater., 2008, 20, 3420–3424.
31. S. Nair, E. Hsiao and S. H. Kim, Chem. Mater., 2009, 21, 115–121.
32. H. Javadian, P. Vahedian and M. Toosi, Appl. Surf. Sci., 2013, 284, 13–22.
33. H. Javadian, F. Ghorbani, H.-a. Tayebi and S. H. Asl, Arabian J. Chem., 2015, 8, 837–849.
34. X. Wang, L. Xing, Y. Shu, X. Chen and J. Wang, Anal. Chim. Acta, 2014, 837, 64–69.
35. Y. Tang, Y. Ren and X. Shi, Inorg. Chem., 2013, 52, 1388–1397.
36. Z. Teng, J. Li, F. Yan, R. Zhao and W. Yang, J. Mater. Chem., 2009, 19, 1811–1815.
37. X. Li, J. Zhang and H. Gu, Langmuir, 2012, 28, 2827–2834.
38. J. Zhang, X. Li, J. M. Rosenholm and H.-c. Gu, J. Colloid Interface Sci., 2011, 361, 16–24.
39. Y.-T. Huang, D. J. Lowe, G. J. Churchman, L. A. Schipper, R. Cursons, H. Zhang, T.-Y. Chen and A. Cooper, Appl. Clay Sci., 2016, 120, 40–50.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra17353e

---