Physico-chemical properties and toxicological effects on plant and algal models of carbon nanosheets from a nettle fibre clone

Carbon nanosheets are two-dimensional nanostructured materials that have applications as energy storage devices, electrochemical sensors, sample supports, filtration membranes, thanks to their high porosity and surface area. Here, for the first time, carbon nanosheets have been prepared from the stems and leaves of a nettle fibre clone, by using a cheap and straight-forward procedure that can be easily scaled up. The nanomaterial shows interesting physical parameters, namely interconnectivity of pores, graphitization, surface area and pore width. These characteristics are similar to those described for the nanomaterials obtained from other fibre crops. However, the advantage of nettle over other plants is its fast growth and easy propagation of homogeneous material using stem cuttings. This last aspect guarantees homogeneity of the starting raw material, a feature that is sought-after to get a nanomaterial with homogeneous and reproducible properties. To evaluate the potential toxic effects if released in the environment, an assessment of the impact on plant reproduction performance and microalgal growth has been carried out by using tobacco pollen cells and the green microalga Pseudokirchneriella subcapitata. No inhibitory effects on pollen germination are recorded, while algal growth inhibition is observed at higher concentrations of leaf carbon nanosheets with lower graphitization degree.


Results and discussion
Surface morphology and porosity of CNS. Plant biomass is chiefly composed of cellulose. This biopolymer is composed for approximately 44% of C; however, the C content can increase to around 80% with pyrolysis at high temperatures and reach 95% using various activating agents 23,24 . The preparation of activated C involves two steps: (i) the pyrolysis of biomass in an inert atmosphere and (ii) the activation of carbonized C with activating agents. The former step produces C from raw materials and the latter is used to enlarge the diameters of fine pores and develop new pores after the carbonization step 25,26 . The morphology, porous nature and specific surface area of C nanomaterials derived from plant biomass mainly depend on the activating agents, e.g. ZnCl 2 , KOH, NaOH, H 3 PO 4 and carbonates 2,27,28 . Among them, sodium bicarbonate (NaHCO 3 ) is a well-known activating agent used to generate pores in C nanomaterials and increase their specific surface area 2,28 . Pyrolysis at 650 °C and NaHCO 3 were used to prepare CNS from nettle biomass. The surface morphology and porous nature of the nettle-derived CNS were investigated through FESEM measurements.
The FESEM micrographs of the CNS obtained with 1:1 and 1:2 nettle powder: NaHCO 3 ratio show an interconnected nanosheet-like structure containing a network of pores ( Fig. 1). Figure 1(panels a-c) shows the FESEM images at different magnification of CNS1:1 prepared from nettle stems, while Fig. 1(d-f and g-i) those relative to the nettle leaf-derived CNS1:1 and CNS1:2, respectively. The chemical activation of CNS using NaHCO 3 treatment at a high temperature forms pores in the C skeleton, thus resulting in the formation of interconnected highly porous CNS. The ImageJ software (ImageJ bundled with 64-bit Java 1.8.0_172, https:// imagej. nih. gov/ ij/ downl oad. html) was used to measure the thickness of the prepared CNS: the average thicknesses were 36, 32 and 27 nm for the nettle stem-derived CNS1:1 (Fig. 1b) and leaf-derived CNS1:1 (Fig. 1e) and CNS1:2 (Fig. 1h), respectively. It was observed that changing the ratio of activating agent and nettle powder did not alter the morphology of the CNS. The above results clearly indicate that NaHCO 3 activation can increase the active surface area of the prepared C by producing pores. The porous structure of the C is useful for several applications, such as supercapacitor, electrochemical sensing 29 (fast diffusion of the electrolyte 2 ) and to increase the surface area for biological applications 30 .
Elemental analysis of CNS. EDS was used to detect the presence of elements in the prepared C samples.
Both qualitative and quantitative EDS results are shown in Fig. 2. These results indicate the presence of very high C content (> 90% atomic weight) as a major element and low O content (between 1.29 and 6.03% atomic weight) as a minor element in all the prepared CNS. Traces of Ca, Si and Cl are also present, which generally occur in plant tissues 31 and, consequently, in plant-derived C nanomaterials 32 . Additionally, nettle is known as a silicifier, with its trichomes typically containing silica at the tip 33 .
Some trace amounts of Cu and S can also be seen in all the prepared samples: the presence of Cu is due to the substrate, because of the drop drying of CNS on the surface of Cu tape used as conductive support. S is a fake peak around 2.3 eV and derives from the combination of O and Si; more specifically, it is a sum of O Kα and Si Kα peaks. www.nature.com/scientificreports/ Elemental analysis was also carried via X-ray photoelectron spectroscopy (XPS) on a representative nettle CNS sample (nettle leaves 1:1), which confirmed the existence of C as major element with O and N as minor elements and Si as a trace element in the prepared CNS (Fig. 3).
In conclusion, the EDS and XPS analyses reveal that C is the major element in the nettle-derived CNS. The presence of O is also confirmed, which is common in biomass-derived C 34 . These analyses further confirm the absence of any metal impurities in the nettle-derived CNS.
TGA of CNS. The thermogravimetric analysis (TGA) analysis was performed in air to investigate the thermal decomposition of the prepared CNS. The TGA curves shown in Fig. 4 demonstrate that the mass loss occurred in two steps. In the first step, the mass loss occurred in the temperature range of 40 °C to 100 °C, which is mainly due to the evaporation of the moisture and residual water molecules. In the second step, the major mass loss was observed between 400 °C and 700 °C in all the three samples, showing the start of thermal decomposition at around 400 °C and stabilizing at around 700 °C. Another aspect worth being highlighted is the final mass: above 700 °C, all the prepared CNS resulted in complete decomposition (i.e. zero mass) which indicates the preparation of metal-free CNS by this process 35 . Diffraction pattern of CNS. XRD was used to investigate the diffraction nature and phase formation of the prepared CNS. Figure 5 shows the XRD patterns of nettle-derived CNS prepared with different activation processes, i.e. (a) CNS1:1 (nettle stems: NaHCO 3 ), (b) CNS1:1 (nettle leaves: NaHCO 3 ) and (c) CNS1:2 (nettle leaves: NaHCO 3 ). All the samples show identical diffraction patterns and all the peaks are well consistent with the pure graphitic C. The XRD spectra of the activated CNS show two peaks: the first characteristic broad peak at around 2θ = 24° is related to C(002) diffraction peak, which corresponds to the reflection of graphitic C and a second less intense, broader peak at around 2θ = 42° is assigned to C(100) diffraction peak, which accounts for amorphous C (JCPDF: 41-1487) 36,37 . No additional peak from the minor trace elements could be observed in the XRD spectra, due to their relatively low concentration, suggesting the formation of pure CNS.     www.nature.com/scientificreports/ Raman Spectroscopy. Raman spectroscopy was used to investigate various disordered domains and the graphitization degree of the prepared nettle-derived CNS. The Raman spectra show two peaks, i.e. the G-band and D-band (Fig. 6), with the latter accounting for disorder and the former for perfect graphite crystals 38 . Typically, the G band centered at 1570 cm −1 corresponds to the two dimensional in-plane motion of strongly bonded sp 2 C atoms, highlighting a characteristic of graphitic C and the D band centered at 1340 cm −1 relates to the defect sites or disordered tetrahedral sp 3 -hybridized C atoms 39,40 . The ratio of integrated intensities of D band and G band (I D /I G ) represents the number of defects in the structure 38 . The ratios of areas under D-band and G-band was used to calculate I D /I G

41
. The calculated I D /I G ratios are 1.2, 1.1 and 1.3 for nettle stem-derived CNS1:1, nettle leaf-derived CNS1:1 and CNS 1:2, respectively. This high intensity ratios entail the existence of disordered C in the prepared porous CNS 2 . These results suggest that the I D /I G ratio of nettle leaf-derived CNS1:1 is lowest, i.e. this samples possess the lowest number of defects 42 . This means that the highest degree of graphitization is found in NaHCO 3 -activated nettle leaf-derived CNS1:1.
The high graphitization degree of the nettle stem-derived CNS is due to the structure of the tissues: just like hemp 3 , the stem of nettle also contains bast fibres ( Supplementary Fig. 1), that are long elements with a thick cellulosic wall (called gelatinous, or G-layer). Bast fibres are indeed characterized by crystalline cellulose and are hypolignified 43 . There are however differences between hemp and nettle bast fibres. The latter do not show labelling when the LM10 antibody recognizing the hemicellulose xylan is used and display a different structure of the G-layer 18 compared to hemp 44 . At the transmission electron microscope (TEM), the G-layer of nettle bast fibres appears indeed less compact than that of hemp, with the formation of local loose regions where the cellulosic layer seems to be flaking off ( Supplementary Fig. 1c) 18 . Such a loose structure may facilitate the process of CNS preparation by favouring disassembly of the layers and penetration of the activating agent.
Nitrogen adsorption-desorption isotherms and corresponding BJH pore-size distributions. The pore-network structure and Brunauer-Emmett-Teller (BET) specific surface area for the prepared nettle-derived CNS were investigated based on the physical nitrogen adsorption-desorption analysis. High surface area and combined pores containing C samples are very important for improved biological applications. NaHCO 3 was used as an activating agent to improve the porous structure and enhance the BET specific surface area of the prepared CNS. The nitrogen adsorption isotherms of the prepared samples are shown in Fig. 7 and described according to the IUPAC report 45 . It can be clearly seen that all the three samples show similar type-IV isothermal sorption curves, due to the presence of hysteresis loops [45][46][47] . The initial isotherm curvatures of the prepared C samples are assigned to micro-and mesopores, while the ascending curvature of the plateau (type H3 loop) in the high relative pressure range (P/P 0 = 0.85 − 1.00) is inherent to macropores [45][46][47] . The wider and higher ascending curvature in the CNS prepared from leaves indicates the presence of more macropores than the CNS prepared from the stems. This more pronounced adsorption property in the isotherms of CNS from leaves is assigned to an enhanced N 2 uptake associated with micropore filling.
The corresponding Barrett-Joyner-Halenda (BJH) pore size distribution of the prepared samples is shown in Fig. 8(a-c), while the structure properties of the prepared CNS using different ratios of the nettle powder and NaHCO 3 are summarized in Table 1. The corresponding cumulative pore size distribution based on the BJH method is shown in Fig. 8d.
The two samples of CNS prepared using nettle leaves have a relatively similar BET specific surface area, while the CNS prepared using nettle stems show significantly higher value than the leaf samples ( Table 1). The average pore volume and pore diameter of the CNS prepared from nettle leaves are larger than the CNS prepared from stems. This observation can be explained by the presence of a high number of macropores in the CNS prepared from nettle leaves, while the additional micropores formation in the CNS prepared from stems contribute to the improvement of the BET specific surface area 48 . As shown in Table 1, CNS1:1 prepared from nettle stems show the highest BET surface area of 718 m 2 ⁄g, while CNS1:1 and CNS1:2 have BET surface areas of 482 m 2 /g and 478 m 2 /g, respectively. The BET specific surface area is in the same range as the one previously calculated for cornstalks-derived C nanomaterial (between 320-790 m 2 ⁄g) 4 , while the thickness is in agreement with the values reported for CNS prepared from hemp (10-30 nm) 3 , tal palm (22 nm) 2 , waste coffee ground (ca. 20 nm) 1 . Hence, nettle leaves and stems are suitable renewable resources for the preparation of CNS with the advantage Figure 6. Raman spectra of CNS prepared at 650 °C from nettle leaves (a,c) and (b) stems using NaHCO 3 as activating agent. The ratio of nettle powder and NaHCO 3 is 1:1 (a,b) and 1:2 (c). Effects of nettle-derived CNS on the germination of tobacco pollen. Evaluating the environmental impact of either the production or utilization of a new material (such as CNS) implies a complex system of analysis requiring the contribution of different, but complementary, scientific disciplines. In plants, C nanomaterials negatively impact germination in both monocots and dicots. For example, in rice, graphene at 50 µg/mL delayed by three days the germination of the seeds compared to the control group and also caused a decrease in moisture content, probably due to the physical blocking of the seed coat pores and consequent water uptake impairment 49 . Recently, a study focused on an economically-important crop, i.e. tomato, showed that while C nanotubes (single-walled and multi-walled) in soil delayed early growth and flowering, they did not affect later developmental stages 50 . Single-walled nanotubes, however, caused an increase in salicylic acid, a finding denoting the presence of a stress response in the plants which is not present in the case of multi-walled tubes 50 . The germination of tomato seeds primed with C nanomaterials was not affected; however, parameters such as root length and hypocotyl biomass were impaired, while chlorophyll content, as well as vitamin C, β-carotene, phenols and flavonoids increased 51 . These results show that C nanomaterials can be used as biostimulants, but their stimulating effects should be carefully determined beforehand to establish the right concentration for the crop examined.
As a first step towards the study of the effects of C nanomaterials on plants, it is possible to use a simplified system consisting in cell models that respond quickly to environmental stressors 52,53 . Obviously, such analyses do not allow an overall assessment of the environmental impact of a material or technology, but they can provide an initial indication of their toxic potential. Many model cells can be used in such an assessment, including the pollen tube, because it is a critical structure during plant reproduction 54,55 . One of the most important reasons for selecting the pollen tube as a model cell is that the evaluation of toxic effects can be quantified in a relatively simple way. In fact, the pollen tube is a cell that grows linearly and in which many cytological events follow one another in a precise spatial-temporal order. These characteristics make the pollen tube a model cell system suitable for providing measurable data 56 .
Pollen is very sensitive to environmental pollutants compared to other plant cells and this allows to assess the impact of a wide range of chemicals on plant metabolism 57,58 . Therefore, it is used as an indicator of air pollution because the performance of pollen (described in terms of germination and elongation of pollen tubes) is affected by airborne gas pollutants. Atmospheric particulate matter also affects pollen performance, for example Ag and Pd nanoparticles impair in vitro pollen germination and pollen elongation in kiwi 59,60 . Graphene oxide nanoparticles can also have a negative effect on the germination and growth of tobacco and hazel pollen tubes 19 .
In Supplementary Fig. 2, images from inverted microscopy and SEM are shown. CNS tend to form aggregates, especially at the highest concentration ( Supplementary Fig. 2a-c), rather than evenly distribute in the   Fig. 2d-f).
In the presence of pollen grains, the CNS clusters tend to adhere to the pollen grains ( Supplementary Fig. 3). It was previously suggested that the pollenkitt of N. tabacum pollen grains could act as a matrix favouring the adsorption of CNS 19 . The presence of CNS aggregates sometimes made it difficult to correctly visualize the pollen and therefore to evaluate the germination percentage.
As shown in Fig. 9, the percentage of germinated pollen increases slightly from 1 to 3 h of growth. From personal experience, a value above 50% is normally considered as acceptable in the case of frozen pollen. Germination data are thus consistent with what is usually observed. Data in Fig. 9 also show that the presence of stem-and leaf-derived CNS has no effect on pollen germination at any of the three concentrations tested for a total germination time of 3 h. Therefore, even if CNS form aggregates around pollen grains, this has no impact on pollen germination. As can be seen in Fig. 9, CNS do not affect pollen growth, even at the maximum concentration. Therefore, nettle-derived CNS have no inhibitory effects on tobacco pollen germination. After all, the germination of pollen grains is only one of the processes that take place during the reproduction of seed plants. Therefore, this result is not surprising because the possible negative effects of CNS could affect other steps, such as pollen activation, or pollen tube growth 19 , or specific molecular events 22 . Future studies, focused on the further development of pollen tubes, will clarify the possible effects of CNS during other processes related to pollen tubes  Table 1. BET surface area, average pore volume and average pore diameter of the prepared nettle-derived CNS at 650 °C, using different NaHCO 3 : tissue powder ratios.

Samples
Biomass sources BET surface area (m 2 /g) Average pore volume (cm 3 /g) Average pore diameter (Å) www.nature.com/scientificreports/ such as, for example, the transport of sperm cells or a possible inhibitory effect of nanocomposites in planta i.e. on the adhesion, activation and growth of the pollen tube on stigma and style.
Effects of the nettle-derived CNS on the growth of P. subcapitata. The evaluation of the ecotoxicological effects of nettle CNS was complemented by tests on the green microalgae P. subcapitata. This is a phytoplanktonic freshwater aquatic model for ecotoxicological investigations and is widely used for studies addressing the interaction of nanoparticles with microalgae cells in aquatic environments 61,62 . In the literature, several studies have investigated the effects of C nanomaterials on microalgae and shown a higher toxicity in the case of oxidized nanomaterials. Oxidation occurs as a spontaneous aging/weathering process after the nanomaterials are released in the environment. Chlorella pyrenoidosa increased the antioxidant response as a result of exposure to oxidized multi-walled C nanotubes, with enhanced pentose phosphate pathway, cell division and formation of polyphosphate bodies 63 . However, when these protective responses were overwhelmed, the algae suffered from toxicity caused by membrane damage and denaturation of macromolecules. Oxidized C nanotubes adsorbed to the cell surface and penetrated via puncture and subsequent endocytosis 63 . Different concentrations (3.4-6.25-12.5 and 25 µg/mL) were tested and the number of algal cells was counted after 72 h of growth. As it can be observed in Fig. 10, at higher concentrations, the nettle CNS cause a decrease in the number of P. subcapitata cells for the stem-and leaf-derived CNS prepared using a ratio NaHCO 3 : nettle powder equal to 1:2. The nettle CNS tend to form aggregates (Supplementary Fig. 2) and the accumulation of C nanomaterials around algal cells was reported to cause shading and, subsequently, to reduce light availability thus resulting in growth impairment 64,65 .
The toxicity of C nanomaterials is due to different physical parameters, i.e. size, surface charge and area 66 . A higher surface area could adsorb more transition metals and ultimately trigger the formation of more reactive oxygen species (ROS) via Haber-Weiss and Fenton-like reactions 67 . The BET surface area for the leaf CNS is, however, the same, irrespective of the ratio of activating agent used ( Fig. 7 and Table 1).
The toxicity of C nanomaterials in algae can also be due to the presence of heavy metal impurities due to low-efficiency manufacturing process (which, for the lowest manufacturing processes, can constitute up to 99.9% of the total production output 68 ). A recent study has shown that metal impurities in C nanotubes affect the esterase activity of the algae Attheya ussuriensis, Chaetoceros muelleri, Heterosigma akashiwo and Porphyridium purpureum (with the latter showing the lowest changes) 69 . It should however be noted that, in the present study, no metal impurities were detected (Figs. 2 and 3).
The higher inhibitory impact on growth by the nettle leaf-derived CNS1:2 may also be ascribed to its high defective nature (higher I D /I G , lower graphitization) and smaller average sheet thickness with a higher adsorption average pore width. Such properties of leaf-derived CNS1:2 facilitate the easy absorption at higher concentrations and thus cause a decrease in the number of P. subcapitata cells. Although the EC 25 value of leaf CNS1:2 is higher, it should however be noted that all the EC 25 values reported for the nettle CNS lie within the same www.nature.com/scientificreports/ order of magnitude; thereby, no statistically significant differences are visible between the CNS. The values in Table 2 report the various essential parameters that may contribute to the toxicity of nettle C nanomaterials on microalgal cells.

Conclusions
The preparation of highly porous activated CNS from nettle tissues with interesting properties in terms of porosity and active surface area was demonstrated in the present study by using pyrolysis at 650 °C. The stems and leaves of U. dioica clone 13 were used as raw material and NaHCO 3 as activating agent. Various ratios of the nettle powder and NaHCO 3 were mixed and carbonized together under N 2 environment to get the porous CNS. The C obtained from this process showed high porosity and BET specific surface area. The highest BET specific surface area (718 m 2 /g) was obtained from the nettle stem-derived CNS, which is in the range of commercially available activated C. The potential ecotoxicological effects of nettle CNS were evaluated on tobacco pollen and the green microalga P. subcapitata. Tobacco pollen germinated both under control conditions and in the presence of progressively increasing concentrations of CNS; under the experimental conditions used, CNS formed dark aggregates that  www.nature.com/scientificreports/ were easily observable as their concentration increased. The presence of such dense aggregates forming a sheath around the algal cells and reducing light availability is probably the reason for the observed growth inhibition. The growth inhibition for algae was however associated to high concentrations of the CNS. The strongest inhibitory effects on the growth of P. subcapitata were observed for leaf CNS1:2, a finding suggesting that the low degree of graphitization has a negative effect on algal growth. By using a type of biomass that can be propagated rapidly by stem cuttings and in large amounts, highly porous CNS can be produced at a large scale and low cost for useful industrial applications. Overall, the work here presented is novel because it shows an additional way to valorize a neglected species considered a weed. The advantage of nettle over other plants is its fast growth and propagation using stem cuttings, which ensures homogeneous starting raw material.

Materials and methods
Growth of plants and chemicals. The whole stems and leaves of stinging nettle (U. dioica clone 13) were collected from 1.5 months-old plants grown under controlled conditions 17 . The clone used produces a higher yield of bast fibres (16%) compared to wild nettle (ca. 4-5%) and it was bred by G. Bredemann between 1927Bredemann between -1950 at the Institute of Applied Botany in Hamburg 70 . Sodium bicarbonate (NaHCO 3 ) and hydrochloric acid (HCl) were obtained from Sigma-Aldrich. The nitrogen (N 2 ) gas with a purity of 99.99% was purchased from SCG gas supply center, Jubail, Saudi Arabia. De-ionized (DI) water was obtained from a water purification system (Barnstead Nanopure, Thermo Scientific, USA).
Preparation of CNS. CNS were prepared by pyrolysis of the nettle stems and leaves using NaHCO 3 as an activating agent. The stems and leaves were collected, washed with DI water and then dried in an electric oven at 100 °C for 24 h. The dried stems and leaves were separately crushed by a high-precision grinding machine to obtain a very fine powder. Powder (particle size ≤ 100 µm) was collected after passing it through a 100 µm mesh. Three different samples were prepared by mixing the obtained powder with different ratios of NaHCO 3 . The ratio of nettle stem powder and NaHCO 3 was 1:1 (referred to as CNS1:1), while for the leaves and NaHCO 3 the mass ratios were 1:1 and 1:2 (referred to as CNS1:1 and CNS1:2, respectively). The final uniform mixture was then placed in an alumina crucible and carbonized at 650 °C for 5 h at a heating and cooling rate of 10 °C/min and 5 °C/min, respectively, in a tube furnace under the protection of nitrogen flow. For the last step, the samples were placed in glass beakers containing 0.5 M HCl and ultrasonicated for 2 h and then washed two times with DI water. Following this step, the obtained samples were then filtered through a filter paper. Finally, the filtered samples were dried in an electric oven at 60 °C for 24 h to obtain the nettle-derived CNS.
Characterization of CNS. The morphology of the prepared samples was analyzed using FESEM (TESCAN LYRA 3, Czech Republic). The FESEM was operated at 20 kV. EDS data were recorded with an Oxford Instruments Xmass detector, equipped with the FESEM and analyzed with the LINK INCA program system.
The TGA of CNS were recorded by the thermal analyzer (TGA1 STAR e SYSTEM), having a temperature range from room temperature up to 1000 °C. An X-ray monochromator (ESCALAB 250Xi XPS Microprobe, Thermo Scientific, USA) was utilized to record the XPS of the prepared CNS. An X-ray diffractometer, Rigaku Ultima (high-resolution) equipped with Cu-K (alpha) radiation was used to obtain the XRD patterns of the prepared CNS. The graphite structure of the synthesized CNS was detected and studied by using a Raman spectrometer, iHR320 with CCD detector (HORIBA), equipped with green laser (300 mW) having an excitation wavelength λ o = 532 nm. For the measurement of the surface area and pore size of the prepared CNS, BET and BJH analyses were performed by using Micromeritics ChemiSorb 2750. A high-precision grinding machine was used to make a fine powder and a Power Sonic 603 ultrasonic cleaner was used for sonication.
Optical and transmission electron microscopy (TEM) of nettle stem tissues. The preparation of nettle stem tissue sections and Toluidine Blue O staining were performed as previously described 17,71 . Nettle bottom internodes (5 mm thickness, localized at the stem base) were fixed in glutaraldehyde/paraformaldehyde/caffeine (1%/2%/1% v/v in Milli-Q water) under vacuum for 15 min and o/n at 4 °C, dehydrated in an ethanol series (70-95-100%), impregnated in resin containing PEG 400 (2% v/v) and dimethacrylate ethylene glycol (0.4% w/v) and finally included in Technovit 7100 resin (Technovit 7100, Heraeus Kulzer GmbH, Hanau, Germany). Cross sections of 10 μm thickness were prepared using a microtome and stained with 0.05% w/v Toluidine Blue O.
Sample processing for immunoTEM and detection of crystalline cellulose with CBM3a were as previously reported 18 . Briefly, the bottom internode was sampled by using a clean razor blade and thereafter fixed for 2 h at room temperature (RT) and o/n at 4 °C in 2% v/v glutaraldehyde and 1.6% v/v paraformaldehyde in 0.1 M phosphate buffer pH 6.9. Samples were rinsed with phosphate buffer two times for 10 min each, then dehydrated in a graded series of ethanol. Samples were infiltrated with LR-White resin at 1:1 ratio with ethanol to pure resin. After polymerization for 48 h at 40 °C, ultrafine sections were obtained with a diamond knife and the ultramicrotome LKB NOVA. The sections were collected on gold grids and blocked for 20 min with normal goat serum diluted 1:30 in dilution buffer (0.05 M Tris-HCl pH 7.6, 0.9% w/v NaCl and 0.2% w/v BSA). Sections were incubated with the CBM3a protein 72 at 5 μg/mL at RT for 1.5 h in conjunction with the mouse monoclonal anti-His antibody diluted 1:100 in dilution buffer + 0.1% v/v Tween 20 for 1.5 h. Samples were washed and incubated with the anti-mouse antibody conjugated with 10 nm gold particles for 45 min at RT. Sections were visualized with the Philips MORGAGNI 268 80 kV transmission electron microscope, equipped with MEGAview II camera and elaborated with the Analysis software. www.nature.com/scientificreports/ Pollen grains' germination assay and microscopy. Pollen grains of tobacco were obtained from plants grown at the Botanical Garden of the University of Siena (Italy). For analysis, the pollen was used after having been stored at -20 °C. The pollen was germinated in BK germination medium with 12% sucrose (referred to as BK suc ) 73 . Germination time was usually 2-3 h. In the case of CNS treatment, the nanosheets were added to the growth medium at different concentrations (10,50 or 100 µg/mL) and the pollen tube growth was monitored exactly as for the control. CNS were initially sonicated to facilitate dispersion, after that they were added to the growth medium. Germination tests were carried out in 24-well plates with pollen at the concentration of 1 mg/ mL. At different time points, pollen tubes were observed with an optical microscope (Zeiss AxioPhot, 10X objective, MRm videocamera). The germination of pollen grains was evaluated by observing 150 pollen grains for each condition. For a statistical evaluation of the effects of CNS at different concentrations on pollen germination, a two-way ANOVA was applied after checking normality with the Shapiro-Wilk test and homogeneity with the Levene's test (SPSS Statistics v19 IBM SPSS, Chicago, IL, USA).
P. subcapitata growth in the presence of nettle CNS. The growth inhibition test was performed with the green algal species P. subcapitata (CCAP 278/4, Scotland, United Kingdom). The test is a modification of the OECD 201 standard 74 in a 12-well plate format. A pre-culture of algae was started 3 days before the beginning of the experiment and algae in exponential growth phase were used. The OECD medium 74 (FUJIFILM Wako Pure Chemical Corporation) was used in order to prepare the test concentrations, which were inoculated with exponentially growing algae in a concentration of 10 5 cells/mL. One mL of algae with or without the different nettle CNS at increasing concentrations (3.4-25 mg/L) were placed in the wells of 12-well plates and were continuously shaken at 50 rpm (Gerhardt Analytical Systems, Germany) under a cool white fluorescent light (46.2 mmol m -2 s -1 ) and temperature 20 °C ± 1 °C. Algal suspension with the highest concentration of the carrier solvent (2.5% v/v DMSO) was used as negative control. Three replicates per concentration and for the control were used. After 72 h, the cell concentration was determined manually in a Bürker-Türk haemocytometer. Growth rates were calculated as indicated in the guideline and the results were expressed as % inhibition of growth rate relative to the untreated control.

Data availability
All data generated or analyzed during this study are included in this published article and the supplementary material.