Graphene Oxide − Upconversion Nanoparticle Based Portable Sensors for Assessing Nutritional De ﬁ ciencies in Crops

: The development of innovative technologies to rapidly detect biomarkers associated with nutritional de ﬁ ciencies in crops is highly relevant to agriculture and thus could impact the future of food security. Zinc (Zn) is an important micronutrient in plants, and de ﬁ ciency leads to poor health, quality, and yield of crops. We have developed portable sensors, based on graphene oxide and upconversion nanoparticles, which could be used in the early detection of Zn de ﬁ ciency in crops, sensing mRNAs encoding members of the ZIP-transporter family in crops. ZIPs are membrane transport proteins, some of which are up-regulated at the early stages of Zn de ﬁ ciency, and they are part of the biological mechanism by which crops respond to nutritional de ﬁ ciency. The principle of these sensors is based on the intensity of the optical output resulting from the interaction of oligonucleotide-coated upconversion nanoparticles and graphene oxide in the absence or presence of a speci ﬁ c oligonucleotide target. The sensors can reliably detect mRNAs in RNA extracts from plants using a smartphone camera. Our work introduces the development of accurate and highly sensitive sensors for use in the ﬁ eld to determine crop nutrient status and ultimately facilitate economically important nutrient management decisions.

M icronutrient malnutrition is a global problem and its alleviation is an important challenge to tackling food security concerns. Zinc is a mineral micronutrient and deficiency in humans is widespread, resulting in a whole range of heath defects. 1 Cereal crops are a staple source of food across the world but are low in mineral micronutrients in their edible parts. 2 In addition to low Zn in grain, growth in soils where there is low Zn availability leads to reduced agricultural yield, and so there are consequences for quality and quantity of food. Fertilizer applications to ensure adequate yields is not only expensive but has serious environmental consequences; therefore developing technology to use these more efficiently would be an important advance for sustainable agriculture. Accumulation of mineral micronutrients in plant cells depends on the presence of membrane transporter proteins, and members of several different families have been implicated in Zn transport in crops including metal tolerance proteins (MTPs), 3−6 heavy metal ATPases 7 and ZRT, IRT-related proteins (ZIPs). 8,9 Some of these transporters are upregulated under nutrient stress conditions to ensure more efficient uptake and use of the available nutrients. The sophisticated sensing and response mechanisms responsible for homeostatic control of Zn are starting to be elucidated in cereal crops, and particular members of the ZIP family seem to play a key role as early indicators of Zn-deficiency stress. 8−10 The ability to monitor or predict plant nutrient status would be a clear advance to the agricultural sector. The development of technology employing state-of-the-art approaches to monitor biomarkers is important for basic research, but could also be applied to monitor crop performance directly in the field and thereby take actions to avoid loss in yield. Most techniques to detect biomarkers rely on complex laboratory procedures that are often time-consuming and require skilled personnel. For example, the real time polymerase chain reaction (RT-PCR), usually employed for the detection of altered gene expression in various organisms, involves a procedure for the conversion of extracted RNAs into complementary DNAs, requiring a sufficient level of purification of extracts, and then amplification and detection with expensive equipment. 11 Therefore, low-cost and rapid methods for direct detection of biomarkers, without having to implement additional purification and amplification steps are of great interest for a range of applications. In this context, research has been focused on the development of DNA sensors. 12−15 Most types of sensors usually involve the attachment of oligonucleotide sense strands on a functional surface and the output is recorded optically, electrically, or electrochemically upon binding of a complementary target. Among these, optical sensors involving the controlled quenching of fluorescence molecules have been broadly utilized. The most common design involves the incorporation of an energy transfer pair where the fluorescence of a donor organic dye is quenched by an acceptor, typically a surface or another molecule. However, despite their enormous success as labels in a variety of platforms, most commonly used organic dyes exhibit phenomena such as photobleaching and photoblinking, which can strongly hinder their potential. In addition, photoexcitation in the UV−visible region typically introduces nonspecific backgrounds related to autofluorescence of biological environments. 16 In recent years, the development of upconversion lanthanide-doped nanoparticles (UCNPs) has offered an alternative to conventional fluorophores because of their photochemical stability, absence of blinking and photobleaching and, most important, the upconversion mechanism itself that results in a low autofluorescence background when using near-infrared excitation. 17,18 Because of such properties, UCNPs have been successfully employed in a wide range of biomedical applications. 19 Our group has recently reported the use of hexagonal phase NaYF 4 :Er 3+ ,Yb 3+ upconversion nanoparticles, functionalized with single-stranded DNAs (UCNPs-ssDNA) for the detection of oligonucleotides with high sensitivity. This assay was first designed to combine the emissive optical properties of UCNPs and the quenching ability of graphene oxide (GO) for the detection of poly-A sequences in the picomolar range. 20 Subsequently, sensors to detect specific mRNA biomarkers present in Alzheimer disease and prostate cancer and optimized for use in blood plasma and cell lysate were developed. 21 GO has been widely reported to act as an effective quencher in various fluorescence-based assays. 22−24 The quenching activity of GO depends on the interactions occurring between the sp 2 hybridized carbon of GO and the aromatic structures and polar groups present in single-stranded oligonucleotides. The working principle of our assay relies on UCNPs-ssDNA interactions with GO in the absence of a complementary oligonucleotide target, through the formation of hydrogen and π−π stacking interactions between the nucleobases and the aromatic system on GO. 25,26 In the presence of a specific target sequence, the ssDNA on UCNPs is hybridized with its complementary strand, and the UCNPs no longer adsorb to the GO. The intensity of the upconversion fluorescence emitted in the presence or absence of a specific complementary target confirms the presence or absence of the complementary target. In our previous work, 20,21 the output detection of the GO/ UCNPs sensors was achieved using laboratory-based grating spectrometers and single-photon counting systems, which precluded the use of these sensors in the field.
Here we describe the development of portable GO/UCNPs sensors for the detection of biomarkers related to the nutrient deficiency in crops. Specifically, we chose to focus on Zn nutrition, detecting the induction of ZIP mRNAs as indicators for Zn deficiency. The portable GO/UCNPs sensors include the use of a smartphone camera as an optical detector and they selectively detect specific mRNAs in a mixture of RNAs. A schematic illustration of the principle of the sensors is shown in Scheme 1. RNA is isolated from crops and mixed with upconversion nanoparticles, which are coated with oligonucleotides designed to hybridize with the target mRNAs. Graphene oxide is added to the solution, and the sample is irradiated with a portable 980 nm laser source. A smartphone camera detects the fluorescent signature of the sample, which confirms the presence or absence of the mRNA target. The advantage of these sensors is not only the portability but also the ability to apply them directly on RNA extracts, eliminating the additional steps for reverse transcription or DNA amplification required by RT-PCR.

RESULTS AND DISCUSSION
Synthesis and Oligonucleotide Functionalization of Upconversion Nanoparticles. Hexagonal phase NaYF 4 :Yb 3+ (25%), Er 3+ (3%) upconversion nanoparticles (core UCNPs) with an average size of 32.1 ± 2.8 nm (Figure 1a,b) were synthesized following a well-established solvothermal method. 27 The core particles of the UCNPs were coated with a NaYF 4 shell following a modified method previously reported. 28 The formation of a shell has been shown to result in an increase at the upconversion emission, which is related to a decrease in surface defects of the lattice and concomitantly reduced energy dissipation, as well as to an improved light harvesting, promoted by the shell formation. 29−31 The use of NaYF 4 :Yb 3+ (25%), Er 3+ (3%) @NaYF 4 (core−shell UCNPs) allows an increase of the upconversion emission and, thus, improves the performance of the sensor. 29 The synthesized core−shell UCNPs exhibited a larger size than the core ones (40.8 ± 3.4 nm, Figure 1c,d), as well as an enhanced fluorescence emission compared to the core UCNPs ( Figure  1e), while as expected the upconversion nanoparticles maintained their hexagonal phase after the core−shell Scheme 1. Schematic Illustration of the Detection of mRNA Biomarkers Related to the Nutrient Deficiency in Crops a a RNA extracts from crops are mixed with oligonucleotide-coated upconversion nanoparticles for a few minutes to allow hybridization of the target sequences. Then, graphene oxide is added, and the sample is irradiated with a portable 980 nm laser source. The fluorescence output is detected by the camera of a smartphone confirming the presence or absence of the target mRNA sequence.

ACS Nano
Article formation as confirmed by XRD measurements (Supporting Information Figure S1).
The core−shell UCNPs were coated with poly(acrylic acid) (PAA) following a ligand exchange reaction as previously reported. 32 The PAA is hydrophilic, which facilitates the transfer of nanoparticles in water. Moreover, it contains carboxylic groups, which can be functionalized with aminemodified molecules. In our experiments amine-modified sense oligonucleotides strands were coupled to PAA coated core− shell UCNPs using EDC coupling and the successful coupling was confirmed by Z-potential and UV−vis spectroscopy ( Figure S2) as well as FT-IR ( Figure S3). The sense oligonucleotide sequences were designed to be complementary to short oligonucleotide sequences for specific MTP and ZIP mRNAs (Table S1). Members of these families that are implicated in Zn transport were chosen.
Detection of OsMTP1 mRNA in Transgenic Arabidopsis Using a Portable GO/UCNPs Sensor. To test the ability of a portable set up to detect the fluorescence output of our sensors, we first chose to detect the mRNA for the rice Zn transporter protein OsMTP1. For this purpose, we used RNA extracts from three types of plants: Arabidopsis wild-type plants (wild-type); Arabidopsis mtp1-1 mutant that lacks expression of AtMTP1 (mtp1-1) and finally mtp1-1 mutant lines genetically modified to express the rice gene OsMTP1 under the 35S promoter (mtp1-1 + OsMTP1). Only the latter should show a signal for OsMTP1 mRNA. 4 For the purpose of our experiment, the upconversion nanoparticles were functionalized with sense strands designed to detect OsMTP1 mRNA, and they were mixed with the RNA extracts and then graphene oxide was added. As a control experiment, a second batch of nanoparticles was functionalized with sense strands for the detection of AtACT2 mRNA, a housekeeping gene. The fluorescence output of the solutions was monitored using the portable configuration shown in Figure 2a. The set up consists of a compact 980 nm wavelength laser producing 200 mW of optical power, a cuvette, and a detection system consisting of a collection of lenses, a 900 nm absorptive short-pass filter, and a smartphone with a camera. Figure 2b shows images of the fluorescence output, which were used to obtain the total integrated fluorescence intensity over the entire interaction length in the cuvette. Figure 2c shows the total camera intensity of fluorescence detected for the sensor assays. While for all three types of plants similar levels of AtACT2 mRNA were detected, high levels of OsMTP1 mRNA were only detected in the transgenic mtp1-1 mutant plants transformed with OsMTP1.
Detection of ZIP mRNAs in Barley Using Portable GO/ UCNPs Sensors. Detecting biomarkers which are early signs for Zn deficiency is useful in crop management. Having demonstrated the successful detection of the OsMTP1 mRNA in transgenic Arabidopsis, we constructed sensors to evaluate the presence or absence of ZIP mRNAs in barley following exposure to Zn deficiency. First, barley was grown in hydroponics for 7 days in the presence of Zn, and then half

ACS Nano
Article of the plants were grown for a further 9 days in the presence of Zn while the other half were grown in the absence of Zn. This was a slightly different time course to that previously reported to induce Zn deficiency, 10 but a similar reduction in shoot biomass was observed under Zn deficiency, as illustrated in Figure 3a,b. RNA was extracted from the roots and shoots of these plants, and it was used to detect the mRNAs for HvZIP5, HvZIP7, and HvZIP13, which are regulated by the presence or absence of Zn in barley. 8,10 To confirm that the relevant mRNAs were present in the extracts, we used RT-PCR to detect the associated complementary DNAs (cDNAs). Figure 3 panels c−e show the data obtained for all three ZIP cDNAs normalized to the cDNA for the housekeeping gene HvACT1. Under the time-course used here, we observed an induction of HvZIP5, HvZIP7, HvZIP13 for the plants exposed to Zn deficiency; thus indicating they were good candidates to investigate further using the relevant GO/UCNPs sensors. For this purpose, three batches of core−shell UCNPs coated with sense oligonucleotide strands for the detection of HvZIP5, HvZIP7, and HvZIP13 mRNAs were synthesized and mixed directly with the RNA extracts of the different plants and then graphene oxide was added. The fluorescent signature of all the different samples was recorded by a smartphone camera using the portable laser configuration shown in Figure 2a. The signals were obtained as integrated camera intensities and were subsequently normalized to the signals of the corresponding HvACT1 housekeeping gene for each sample, in order to obtain a normalized relative quantification (NRQ) similar to RT-PCR. Again, the ZIPs were seen to be upregulated under Zn deficiency in both roots and shoots (Figure 3f−h). Therefore, the portable sensors constructed here are able to directly detect the barley genes that are up-regulated under Zn-deficiency in a cocktail of RNA extracts, thereby showing promise for future use in the field.

CONCLUSIONS
There is great potential in the application of nanotechnology in agriculture with major goals including better water management, optimizing nutrient application and use for increased yields, reducing release of damaging chemicals in the environment, and more efficient use of plant protection products. Here, we showed the construction of sensors, which are able to detect mRNAs related to membrane transport proteins up-regulated during Zn deficiency in plants. Most important, these sensors are portable, which means that they could be employed in the field. The principle of the sensors relies on the hybridization of mRNA target sequences with sense oligonucleotide strands attached to the surface of fluorescent upconversion nanoparticles. If the mRNA targets are bound to the upconversion nanoparticles then the particles' fluorescence is not quenched by graphene oxide indicating the presence of the target. The sensors were made portable using a detection system including a portable laser at 980 nm, a set of lenses and a smartphone camera. The sensors constructed here were successfully used to detect OsMTP1 mRNA in RNA extracts from OsMTP1-transformed Arabidopsis and the HvZIP5, HvZIP7, HvZIP13, mRNAs in RNA extracts from barley grown under Zn deficiency conditions. Our study suggests strategies to create sensors for use in the field, which will benefit agriculture, informing decisions related to the nutrient status of crops.
Methods. Transmission electron microscopy (TEM) samples were prepared by placing a drop of a diluted nanoparticle solution on a TEM copper grid coated with a Formvar film and left to dry in air. TEM images were obtained with a Hitachi HT7700 transmission electron microscope operating at a voltage of 75 kV. The nanoparticle size distribution was analyzed with ImageJ (National Institutes of Health, USA) and Soft Imaging Viewer software (Olympus, Japan), measuring over 300 nanoparticles for the statistical analysis. The upconversion fluorescence measurements were performed using a portable diode laser (Roithner Laser Technik GmbH) with continuous wave (980 nm) at 200 mW. The emitted fluorescence was collected perpendicular to the excitation beam using a 35 mm focal length lens, and was imaged onto a smartphone camera (Samsung S4). A short pass IR-blocking filter (Schott KG3) was used to suppress scattered excitation light and select only the fluorescence emission. Camera control software with manual focus, exposure, and ISO control was used to obtain images at constant exposure settings. For each measurement, 30 camera images were taken to average out variations, and the total camera intensity in the area of interest was analyzed using Matlab. A cuvette with the corresponding solvent was measured under illumination with the 980 nm laser beam and set as the blank for each measurement. The Z-potential measurements were performed using a Zetasizer Nano ZS instrument (Malvern Instruments, U.K.), and the accumulation time was determined automatically for each sample. The acquired data was processed using the software provided by Malvern (Zetasizer software v7.03). The UV−vis data for confirmatory experiments were acquired using a UV-2700 spectrophotometer (Shimadzu, UK). All statistics were done using ANOVA one-way with GraphPad Prism version 6.07 for Windows, GraphPad Software, La Jolla California USA, www.graphpad.com. Quantification of the emitted fluorescence for the biosensor obtained from the pictures acquired by the camera on the smartphone was performed by using Matlab on triplicates of at least 10 pictures for samples. Quantified fluorescence thresholds were set up by using the emitted fluorescence from house-keeping genes Actin as internal standard. The powder Xray diffractions were collected using a Bruker D2 Phaser equipment with a Cu Kα (1.54 Å) X-ray source between 5°< 2θ < 40°.
Statistical Analysis. Two-way ANOVA was used for statistical analysis, conducted using Minitab and Prism software, with the threshold for statistical significance difference taken at a 95% confidence interval. Bonferroni, Tukey and Dunnet's posthoc tests were used as indicated to determine significant differences. Experimental measurements were taken in triplicate with at least 10 measurements each for fluorescence and photon counts. RT-PCR analyses were taken as triplicated values of technical values. Statistical analyses were conducted on SEM ± error analysis of the mean average n = 3.
Synthesis of Core UCNPs. The synthesis of upconversion nanocrystals was performed following a previously reported protocol with some modifications. 33 Briefly the rare earth salts, YCl 3 ·6H 2 O (245 mg, 0.81 mmol), YbCl 3 ·6H 2 O (107 mg, 0.28 mmol), and ErCl 3 ·6H 2 O (15 mg, 0.04 mmol) were introduced in a 100 mL round-bottom flask. Then, oleic acid (6 mL, 19 nmol) and 1-octadecene (15 mL, 2.5 mmol) were added and the solution was heated up at 150°C under Ar. After 1 h, the mixture was cooled down to room temperature. To this solution, a mixture of NaOH (100 mg, 2.5 mmol) and NH 4 F (150 mg, 4 mmol) dissolved in 10 mL of dry MeOH were added dropwise. Then, the solution was stirred for other 30 min at room temperature and it was gradually heated up to 130°C under Ar. After 20 min at 130°C under Ar, it was then stirred for other 20 min at 130°C under vacuum to ensure the complete evaporation of the MeOH. Finally, the temperature was raised to 305°C at 15°C/min rate under Ar and the mixture was stirred for 1 h 10 min to form the nanoparticles. After completion of the reaction, the upconversion nanoparticles were left to cool down to room temperature. EtOH (20 mL) was then added to the nanoparticles solution and the mixture was centrifuged (8000 rpm, 10 min). This process was repeated three times to purify the nanoparticles. The nanoparticles' white pellet was left to dry at 80°C for several hours before being utilized in further experiments.
Synthesis of Core−Shell UCNPs. Core−shell nanocrystals were prepared following a previously published protocol. 28 Briefly, YCl 3 · 6H 2 O (150 mg, 0.8 mmol) was dissolved in a solution of 1-octadecene (15 mL) and oleic acid (6 mL), and stirred for 1 h under Ar at 150°C. The solution was cooled down to 40°C and core UCNPs (150 mg) in CHCl 3 (10 mL) were added dropwise. After 20 min at 40°C, the solution was gradually heated up to 100°C under Ar and stirred for 45 min. Once the mixture was cooled down to room temperature under the Ar flux, a solution of NaOH (100 mg, 2.5 mmol) and NH 4 F(150 mg, 4 mmol) dissolved in dry MeOH (10 mL) was added dropwise. Then, the solution was stirred for another 30 min at room temperature, and it was gradually heated up to 130°C under Ar. After 20 min at 130°C under Ar, it was stirred for another 20 min at 130°C under vacuum to ensure the complete evaporation of MeOH. Finally, the temperature was raised to 305°C at 15°C/min rate under Ar, and the mixture was stirred for 1 h 20 min to form the core−shell UCNPs. After completion of the reaction, the nanoparticles were left to cool down to room temperature. Then, EtOH (20 mL) was added to the nanoparticles solution, and the mixture was centrifuged (8000 rpm, 10 min). This process was repeated three times to purify the nanoparticles. The core−shell UCNPs were collected as white powder and stored in CHCl 3 .
Ligand Exchange on Core−Shell UCNPs. To bring the core− shell UCNPs in water, a ligand exchange protocol was followed to coat the nanoparticle surface with PAA. 34 In a typical reaction, PAA (0.25 g, MW ≈ 1.8 kDa) in THF (3 mL) was added to the core−shell UCNPs coated with oleic acid (21 mg in 7 mL THF). The solution was stirred for 48 h at room temperature and then centrifuged at 8000 rpm for 15 min. The particles were resuspended in THF and centrifuged again for another two times. Then the particles were suspended in EtOH (20 mL) and isolated via centrifugation (8000 rpm, 15 min) as a pellet. After drying at 80°C for several hours, the particles were suspended in sterile DNase/RNase free Milli-Q water and stored at 4°C.
Synthesis and Characterizations of ssDNA PAA Coated Core−Shell UCNPs. Core−shell UCNPs coated with PAA were coupled to amino-modified oligonucleotides using EDC coupling chemistry. In a typical reaction, PAA-coated UCNPs (0.5 mg mL −1 ) were suspended in borate buffer (pH 8. Afterward, the particles were purified by three centrifugation steps (16400 rpm, 4°C, 10 min). The UCNPs-ssDNA were resuspended in sterile Milli-Q and stored at −20°C for further use.
Sensor Calibration. The sensors were calibrated in order to find the concentration of GO needed to achieve maximum quenching of core−shell UCNPs fluorescence. These experiments were carried out by adding increasing concentrations of GO (0.1−0.7 mg mL −1 ) to a solution of a fixed concentration of ssDNA PAA-coated core−shell UCNPs (0.5 mg mL −1 ) while measuring the fluorescence signature of the sample. It was found that 0.5 mg mL −1 of GO to 0.5 mg mL −1 of nanoparticles was the optimal ratio corresponding to the maximum quenching of the UCNPs' fluorescence (see Figure S4).
Plants Cultivation and Harvesting. Arabidopsis thaliana plants were grown as previously described 7,35 in soil containing equal proportions of vermiculite, Levington M2, and John Innes No. 2

ACS Nano
Article compost (Fargro) in 8 cm pots; 0.28 g l-1 Intercept insecticide (Bayer, Canada) was present, and the soil was sterilized by autoclaving at 121°C for 15 min at 1 bar pressure. Wild-type (ecotype Wassilewskija), mtp1-1 mutant, and transgenic lines expressing OsMTP1 were grown in a controlled-environment growth room with a day/night cycle (23°C 16 h light, 120 μmol m −2 s −1 ; 18°C 8 h dark). The T-DNA knockout mtp1-1 mutant 36 was kindly provided by Professor Masayoshi Maeshima (Nagoya University, Japan). The OsMTP1 gene induction in Arabidopsis was carried out following previous reported procedures. 2 To grow hydroponically, Hordeum vulgare L. cv Golden Promise seed was sterilized in 1% bleach for 15 min, rinsed in sterile water, germinated on wet tissue paper for 5 days, and then individual seedlings were grown in aerated 1 L hydroponic culture pots (Thermo Fischer Scientific, UK) in a controlled environment room (21°C, 16 h light, 220 μmol m −2 s −1 , 55% humidity; 16°C, 8 h dark, 65% humidity). Pots were filled with a nutrient solution (as previously described), 10  . For the Zndeficient treatment, Zn was omitted from the media. The nutrient solution was replaced every 3 days, and roots and shoots were harvested separately either for fresh weight determinations or for freezing in liquid N 2 for subsequent RNA extraction.
Isolation of RNA Extracts. Total RNA was extracted from Arabidopsis using Trizol Reagent according to the manufacturer's instructions (Invitrogen Life Technologies RNA, UK). RNA from hydroponically grown barley was also isolated using the Trizol method. First-strand cDNA synthesis using 1 μg of total RNA was carried out using the ImProm-IITM reverse transcriptase kit (Promega, USA) with an oligo(dT) primer according to manufacturer's instructions.
PCR Analysis in Arabidopsis thaliana. Standard 10 μL PCR reactions were performed using BioMix Red (Bioline), forward (F) and reverse (R) primers (Table S3) with either genomic DNA or cDNA. Reactions were performed in a peqSTAR 96 Universal PCR machine (Peqlab, Germany). For Arabidopsis, multiple plants were pooled from a particular genotype to generate 1 μg/μL of RNA for each biological replicate, which was then used to generate cDNA for gene expression comparisons. 4 PCR and RT-PCR Analysis in Hordeum vulgare. Standard 10 μL PCR reactions were performed using BioMix Red (Bioline), forward (F) and reverse (R) primers (Table S3) with cDNA. Reactions were performed in a peqSTAR 96 Universal PCR machine (Peqlab, Germany). Real-time PCR was performed as described previously 35 with specific forward and reverse primers. For barley, tissues from three plants were pooled for each biological replicate. Gene expression levels were calculated based on previous methods, 37 standardized by normalizing to HvACT1 for barley 2 and analyzed using Opticon software. They were expressed relative to levels of gene in Barley in the 7 days samples and within the same sample house-keeping gene. The primers sequences used for monitoring expression of a range of genes are reported in Table S2. All the forward primers were the same sequences used for PCR, Real Time PCR, and sensing detection based on the UCNPs.

* S Supporting Information
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.8b03261. The raw data is available at DOI: http://doi.org/10.5258/ SOTON/D0512 Experimental procedures for the fabrication and function of the sensors, oligonucleotide sequences, further characterization of materials (PDF)