The influence of climate change on the fate and behavior of different carbon nanotubes materials and implication to estuarine invertebrates

The widespread use of Carbon nanotubes (CNTs) has been increasing exponentially, leading to a significant potential release into the environment. Nevertheless, the toxic effects of CNTs in natural aquatic systems are related to their ability to interact with abiotic compounds. Considering that salinity variations are one of the main challenges in the environment and thus may influence the behavior and toxicity of CNTs, a laboratory experiment was performed exposing the tube-building polychaete Diopatra neapolitana (Delle Chiaje 1841) for 28 days to pristine multi-walled carbon nanotube (MWCNTs) and carboxylated MWCNTs, maintained at control salinity 28 and low salinity 21. An innovative approach based on thermogravimetric analysis (TGA) was adopted for the first time to assess the presence of MWCNTs aggregates in the organisms. Both CNTs generated toxic impacts in terms of regenerative capacity, energy reserves and metabolic capacity as well as oxidative and neuro status, however greater toxic impacts were observed in polychaetes exposed to carboxylated MWCNTs. Moreover, both CNTs maintained under control salinity (28) generated higher toxic impacts in the polychaetes compared to individuals maintained under low salinity (21), indicating that exposed polychaetes tend to be more sensitive to the alteration induced by salinity variations on the chemical behavior of both MWCNTs in comparison to salt stress.


INTRODUCTION
Carbon nanotubes (CNTs) are artificial carbon-structured substances formed by one to more than one hundred graphene layers rolled into cylindrical tubes (Zhao and Liu 2012). The current typical industrial CNTs types with one graphene layer are called single-walled carbon nanotubes (SWCNTs), while tubes containing more than one layer are multi-walled carbon nanotubes (MWCNTs) (Jackson et al. 2013). As a result of their unique chemical, physical and electronic properties, CNTs became potentially useful in a wide variety of applications, such as water treatment, medical applications, optics, electronic engineering, photovoltaic devices, automotive industry, sports equipment and cosmetics (Zhao and Liu 2012;Freixa et al. 2018). In the last decade, the widespread use of CNTs increased exponentially leading to a significant potential release into the environment (Potočnik 2011;Sanchez et al. 2012). In aquatic ecosystems, CNTs can be accumulated in river sediments, or remain suspended in the water column and transported to the marine systems (Freixa et al. 2018). The predicted environmental concentrations (PECs) of CNTs in aquatic environments have been estimated, by modeling studies, to range between 0.001 to 1000 µg/L (Zhang et al. 2017). Nevertheless, the environmental behavior and effects of CNTs in natural aquatic systems are related to their ability to interact with abiotic and biotic compounds and aggregate, creating clusters that exhibit a colloidal behavior (Khosravi-katuli et al. 2017). Generally, CNTs, due to their nearly insoluble in any solvent, significantly restrict their application (Mwangi et al. 2012). For this reason, scientists have explored different methods to increase the solubility of CNTs such as for both practical use and toxicological research. As example, it has been already demonstrated that A C C E P T E D M A N U S C R I P T 5 by TGA, it is possible to detect the degradation temperature of the material subjected to analysis. Indeed, TGA may represent a valuable tool for the quantitative determination of the composition of composite materials, proving that the degradation temperatures of single components are widely separated from each other (Renneckar et al. 2004;Zhang et al. 2015).
To the best of our knowledge TGA has never been used for the detection of MWCNTs in living organisms, but it could potentially emerge as a successful technique for the identification and quantification of MWCNTs in biological matrices since their degradation temperatures generally lie in-between those of organic and inorganic materials (Zhang et al. 2015;Wang et al. 2008).
A C C E P T E D M A N U S C R I P T 6

Model species
The tube-building polychaete Diopatra neapolitana was selected as model organism for this study as it plays important ecological roles by acting as a food source for other marine biota and increasing the physical complexity and biodiversity of habitats (Arias et al. 2016). Also, it has been demonstrated that D. neapolitana is a good sentinel species of metal contamination (Freitas et al. 2012) organic matter enrichment (Carregosa et al. 2014), pharmaceutical drugs (Freitas et al. 2015), salinity shifts and pH decrease (Pires et al. 2015) and NMs (De Marchi et al. 2017c;.

Experimental set up
Polychaete specimens were collected in the Mira channel, the southern shallow arm of the Ria de Aveiro lagoon (Portugal). In the laboratory, organisms were pushed out from their tubes, and placed in different aquaria (20 l each) for laboratory acclimation period (20 days).
The aquaria were filled with a mixture of fine and medium sediment from the sampling area (see sediment details in De Marchi et al. (2018a)). Artificial seawater with the salinity 28 was used by the addition of artificial sea salt (Tropic Marin® Sea Salt) to deionized water, one day prior to utilization. Temperature was kept to 18 ± 1 ºC, photoperiod of 12 h light: 12 h dark, pH 8.0 ± 1 and constant aeration. During this period every two-three days the specimens were fed ad libitum with small fragments of frozen cockles or mussels (Pires et al. 2012). To assess the impact of different CNTs on the regenerative capacity of D. neapolitana, immediately before the exposure assay individuals were removed from their new tubes, anaesthetized with a 4% MgCl 2 .6H 2 O solution, and amputated at the 60 th chaetiger under a stereomicroscope (Pires et al. 2012).

A C C E P T E D M A N U S C R I P T
A C C E P T E D M A N U S C R I P T 8 mechanical strength (up to 15 to 20 times stronger than steel and 5 times lighter) and their thermal conductivity (same as that of diamond and more than five times that of copper). The combination of these properties enables a whole new variety of useful applications (Li et al., 2011). In details, f-MWCNTs are used as additives in polymers, catalysts electron field emitters for cathode ray lighting elements, flat panel display, gas-discharge tubes in telecom networks, electromagnetic-wave absorption and shielding, energy conversion, lithium-battery anodes, nanotube composites, nanoprobes, nanolithography, nanoelectrodes, drug delivery sensors reinforcements in composites and supercapacitor (MWCNTs-COOH: TNMC1 series, http://www.timesnano.com). The Nf-MWCNTs are used in different markets such as transportation (automotive, aeronautic, boats), electronics (electronic packaging, EMI-shielding, sensors), energy (Lithium-ion), industrial applications (Oil&Gas, dynamic rubber parts, coatings, heating elements) and sport goods (http://www.nanocyl.com/product/nc7000/). Moreover, for the choice of these nanoparticles, it was also considered the Organization for Economic Cooperation and Development (OECD)'s Working Party on Manufactured Nanomaterials (WPMN), which launched the Sponsorship Programme for the Testing of Manufactured Nanomaterials (OECD, 2010). This programme promotes international co-operation on the human health and environmental safety of manufactured nanomaterials, and involves the safety testing and risk assessment of engineering nanoparticles. The OECD WPMN has published a list of ENPs selected considering their commercial use, production volume of the materials, availability of such materials for testing and the existing information that would readily be available on the materials and this list comprised the CNTs.
MWCNTs in bivalves can generated lysosomal damage (Miller et al. 2015), accumulation of these materials in intestinal lumen and digestive gland and histopathological changes in the epithelium and swelling of the connective tissue (Anisimova et al. 2015); oxidative stress and alteration of energy-related responses and metabolism (Andrade et al. 2018;Freitas et al. 2018;De Marchi et al., 2017a; and decreased the cellular integrity (Sekar et al. 2016). Studies testing the impacts of CNTs in polychaetes are less, but also confirmed cellular damage and alteration in organisms energy reserves and metabolism (De Marchi et al. 2017c;2018a;2019).
For each condition, inside each aquarium, individuals used for the regenerative capacity analysis (n=9) were separated from the ones used for the biochemical analysis (n=9) using a plastic barrier in each aquarium.

Materials description
Both pristine and carboxylated MWCNTs were produced via the Catalytic Chemical Vapor Deposition (CCVD) process and characterized using Scanning Electron Microscopy Both CNTs were weighed (stock solution of 50 mg/L) and suspended in seawater (depending on the test condition, control-28 and low-21). Both materials were sonicated using a Hz ultrasound bath (IKA Labortechnik IKASONIC U50): 1 h for Nf-MWCNTs, while the f-

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T few minutes. Both Nf-MWCNT and f-MWCNT concentrations were re-established weekly after complete water renewals to ensure the same exposure concentrations during the experiment.
The added MWCNTs (f and Nf) were maintained homogenously dispersed in the seawater using one submersible circulation pump per aquarium, increasing CNTs mass suspended in the water column (Vonk et al. 2009).

Characterization analysis in the water media
The average size distribution measured by dynamic light scattering (DLS) and the polydispersity index (PDI) of both MWCNTs suspensions at different salinity levels were analyzed. As reported in the literature, DLS measurements represent a well-established method for the determination of the average diameter of carbon nanotubes in aqueous dispersions (Liu et al. 2011). In the present work, DLS measurements were carried out to obtain data regarding the tendency to aggregate and the settling behavior of suspended CNT materials in aqueous media. Measurements were performed on 1000 µL of suspension in four samples per condition, and five analyses per sample performed by DLS using a DelsaTM NanoC Particle Size Analyzer (Beckman Coulter). Each analysis was carried out by performing 120 acquisitions. Due to the inherent heterogeneity and colloidal instability of the analyzed samples, DLS analyses were repeated several times to ensure reproducible results. Intensity distributions were obtained by analyzing the autocorrelation functions through the Contin algorithm which is particularly appropriate for polydisperse and multimodal systems (Varenne et al. 2016). The cumulant method was used to obtain information on the particle's average hydrodynamic radii and on the PDI (Tardani and Mesa 2015).

Characterization and bioaccumulation in the organisms
Thermogravimetric Analysis (TGA) was performed on different entire tissue samples (10 mg) by using a TGA Q500 instrument (TA Instruments, Italy) in the temperature range 30 -900 ºC, at heating rate of 10ºC/min and under air flow of 60 mL/min. TGA, which records the weight loss of the materials as a function of temperature in a chosen atmosphere, was used for the first time as innovative methods to detect the presence of MWCNT materials in D. neapolitana organisms exposed to aqueous media at different salinity levels containing different concentrations of the selected contaminants. A preliminary study was carried out to investigate the feasibility of using TGA by analysing the organisms exposed to the maximum and minimum values of the concentration range of the MWCNT materials used (0.001 and 0.01 mg/L): (NF_S21: organisms exposed to Nf-MWCNTs in an aqueous medium with salinity 21; NF_S28: organisms exposed to Nf-MWCNTs in an aqueous medium with salinity 28; F_S21: organisms exposed to f-MWCNTs in an aqueous medium with salinity 21; F_S28: organisms exposed to f-MWCNTs in an aqueous medium with salinity 28) and non-contaminated organisms (CRTL S21_1, CRTL S21_2, CRTL S21_3: representing organisms immersed in an aqueous medium with salinity 21 and not exposed to carbon nanotubes; CRTL S28_1, CRTL S28_2, CRTL S28_3: organisms immersed in an aqueous medium with salinity 28 and not exposed to carbon nanotubes). The derivative of the TGA curves (DTG curves) were reported to better highlight the temperatures at which thermal degradation of the samples occurred, which could help to detect the presence of MWCNTs (f and Nf) by comparison with their DTG profile.
A C C E P T E D M A N U S C R I P T 12 width of the regenerated body part was measured, and the number of new segments counted.
Percentage of regenerated body width was calculated by comparing the width of the new segments with the width of the old segments (Pires et al. 2012). New segments were identified by the lighter colour and/or the narrower width compared to the old body segments (Pires et al. 2012).

Biochemical parameters: energy reserves and metabolic capacity, indicators of oxidative stress and neurotoxicity
At the end of the experimental period (28 days), the whole body of frozen organisms (3 per aquarium, 9 per condition) was individually pulverized with liquid nitrogen, divided in 0.2 g aliquots, and used for biochemical analyses. Extractions were performed with specific buffers to determine: a) energy reserves and metabolic capacity (protein (PROT) content, glycogen (GLY) content, electron transport system (ETS) activity); b) indicators of oxidative stress (lipid peroxidation (LPO) levels, reduced (GSH) and oxidized (GSSG) glutathione ratio, superoxide dismutase (SOD) activity; catalase (CAT) activity, Glutathione S-transferases (GSTs) activity); c) and neurotoxicity (Acetylcholinesterase (ATChI-ChE) activity). The methodologies used to perform each specific biomarker are described in detail in De Marchi et al. (2018a;.

Characterization analysis in water media
As reported in the literature DLS measurements represent a well-established method for the determination of the average diameter of CNTs in aqueous dispersions (Liuet al. 2011).
Here DLS analyses were repeated several times to ensure reproducible results due to the inherent heterogeneity and colloidal instability of the analyzed samples. Intensity distributions were obtained by analyzing the autocorrelation functions through the Contin algorithm which is particularly appropriate for polydisperse and multimodal systems (Varenne et al. 2016). The cumulant method was used to obtain information on the particle's average hydrodynamic radii, and on the PDI (Tardani and Mesa 2015). The mean size (nm) and PDI of functionalized MWCNTs (f) and pristine MWCNTs (Nf) suspended particles measured in artificial seawater at different salinity levels (21 and 28) are reported in Table 1. Under salinity 21, the mean size of Nf-MWCNTs were largely higher than that of f-MWCNTs. The reliability of the mean diameter values recorded at time 28 days was compromised by the presence of microaggregates of unknown origin as evidenced by the control. Both Nf and f-MWCNTs suspended at 0.01 mg/L were found to agglomerate and remain dispersed in the medium until 28 days. The concentration of 0.001 mg/L was found to lie below the limit of instrumental detection in the adopted experimental conditions. At salinity 28, the reliability of the mean diameter values obtained at time 21 and 28 days was compromised by the presence of microaggregates of unknown origin as evidenced by the controls. Up to 14 days Nf and f MWCNTs displayed a different behavior: f-MWCNTs were found to agglomerate and remain dispersed in the medium while Nf-MWCNTs particles were not detectable by DLS analysis due to settlement and/or

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 15 uptake by marine organisms. At 0.001 mg/L concentration was found to lie below the limit of instrumental detection under the adopted conditions.

Characterization and bioaccumulation in the organisms
Preliminary experiments employing TGA analysis did not allow to identify the presence of MWCNTs (f and Nf) in the exposed organisms due to the overlapping of the degradation peaks recorded between 500°C and 700°C in the DTG curves of non-contaminated D.
neapolitana and CNTs used as reference. The second degradation peak registered in the organisms not exposed to the contaminants strongly suggests the need for an optimization of the sample preparation for TGA analysis. To this purpose sample pre-treatment aiming to remove the material responsible for the degradation peaks overlapping that of MWCNTs in the DTG curves could represent an effective approach to raise TGA as useful technique for the detection of MWCNTs in marine organisms.

Physiological parameter: Regenerative capacity
The mean values for the percentage (%) of regenerated body width and the number (#) of new chaetigers in D. neapolitana after 11 th , 18 th and 28 th days of amputation are presented in Table 2 and illustrated in Figure 2. All the results were discussed considering: i) the effects of exposure concentrations of both MWCNT materials maintained under both salinity levels (significant differences (p ≤ 0.05) among exposure concentrations were represented with different letters: uppercase and regular letters for Nf-MWCNT at salinity 28; lowercase and regular letters for Nf-MWCNTs at salinity 21; uppercase and bold letters for f-MWCNT at salinity 28; lowercase and bold letters for f-MWCNT at salinity 21); ii) the effects of the carboxylation/functionalization of the surface of MWCNTs in organisms maintained under both salinity levels for each exposure concentration (significant differences (p ≤ 0.05) between f-

A C C E P T E D M A N U S C R I P T
16 MWCNT and Nf-MWCNTs within each salinity at each exposure concentration were represented with bold hashes (#)); iii) the effects of salinity shifts in organisms exposed to both MWCNT materials in each exposure concentration (significant differences (p ≤ 0.05) between the two salinities for each MWCNTs and exposure concentration were represented with bold asterisks (*)).

11 th day
After amputation all individuals were healing the cut region, however no significant differences were observed in terms of % of regenerated body width as well as # of new chaetigers between individuals non-exposed (0.00 mg/L) and exposed to both MWCNT materials in all tested concentrations (0.001 and 0.01 mg/L) under both salinity levels (control-28 and low-21).

18 th day
i) Considering the effects of exposure concentrations (0.001 vs 0.01 mg/L), for the same salinity (21 or 28) and MWCNTs (Nf or f) the results of % of regenerated body width and # of new chaetigers showed that for f-MWCNT submitted to both salinities no significant differences were observed between concentrations, while significantly lower values was detected only in individuals exposed to 0.01 mg/L Nf-MWCNTs under salinity 28 in comparison to remaining conditions. ii) Considering the effects of MWCNTs (Nf vs f), for each concentration (0.001 or 0.01 mg/L) and each salinity (28 or 21), no significant differences were observed between organisms exposed to different MWCNTs in terms of % of regenerated body width, while regarding the # of new chaetigers, significant differences between materials were observed only in polychaetes

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 17 exposed to 0.01 mg/L under salinity 28 showing a lower # of chaetigers for individuals contaminated with Nf-MWCNTs.
iii) Considering the effects of salinity (21 vs 28), for each MWCNTs (Nf or f) at each exposure concentration (0.001 or 0.01 mg/L), differences between salinities (28 and 21) were only observed at 0.01 mg/L Nf-MWCNTs with lower % of regenerated body width in individuals maintained under salinity 28 in comparison to individuals maintained under salinity 21, while no significant differences were found in terms of # of new chaetigers for both MWCNT materials (f and Nf) and for both salinities.

28 th day
i) Considering the effects of exposure concentrations, the results of both % of regenerated body width and # of new chaetigers showed only significantly lower values between individuals exposed to 0.01 mg/L of Nf-MWCNTs under control salinity in comparison to the remaining concentrations, while no significant differences were observed in individuals under salinity 21 exposed to both CNTs.
ii) When comparing organisms exposed to the same salinity and exposure concentration, the % of regenerated body width as well as the # of new chaetigers did not show significant differences between individuals exposed to different MWCNT materials.
iii) For each MWCNTs (f and Nf) at each exposure concentration, differences between salinities (28 and 21) were not observed both for % of regenerated body width and # of new chaetigers.

Biochemical parameters
As for the physiological parameter, all the results of biochemical parameters were discussed considering: i) the effects of exposure concentrations of both MWCNT materials

A C C E P T E D M A N U S C R I P T
18 maintained under both salinity levels (significant differences (p ≤ 0.05) among exposure concentrations were represented with different letters in the graphics: uppercase and regular letters for Nf-MWCNT at salinity 28; lowercase and regular letters for Nf-MWCNTs at salinity 21; uppercase and bold letters for f-MWCNT at salinity 28; lowercase and bold letters for f-MWCNT at salinity 21) ii) the effects of the carboxylation/functionalization of the surface of MWCNTs in organisms maintained under both salinity levels for each exposure concentration (significant differences (p ≤ 0.05) between f-MWCNT and Nf-MWCNTs within each salinity at each exposure concentration were represented with bold asterisks (*) in the table 3) ; iii) the effects of salinity shifts in organisms exposed to both MWCNT materials in each exposure concentration (significant differences (p ≤ 0.05) between the two salinities for each MWCNTs and exposure concentration were represented with bold asterisks (*) in the graphics).

Protein (PROT) content
i) Considering the effects of exposure concentrations, results of PROT content in D.
neapolitana exposed to 0.01 mg/L of Nf-MWCNT under salinity 28 showed significantly higher PROT content in comparison to the remaining concentrations, while in individuals exposed to f-MWCNTs under both salinities no significant differences were observed among exposure concentrations ( Figure 3A).
ii) When comparing D. neapolitana exposed to different MWCNTs at the same salinity and exposure concentration, significant differences between materials were observed only in polychaetes exposed to 0.01 mg/L under salinity 28 showing an increase of the content for individuals contaminated with Nf-MWCNTs (Table 3).

ACCEPTED MANUSCRIPT
iii) Significant differences between salinities (28 and 21) were observed in PROT content when organisms were exposed to 0.01 mg/L of Nf-MWCNTs, showing higher content in individuals maintained at salinity control in comparison to low salinity ( Figure 3A).

Glycogen (GLY) content
i) Along the increasing Nf-MWCNTs exposure concentrations, all the exposed polychaetes maintained at salinity 21 decreased significantly their GLY content in comparison to non-exposed ones ( Figure 2B), while under salinity 28 no significant differences were observed between contaminated and non-contaminated organisms. Opposite results were observed in organisms submitted to f-MWCNTs, showing a significantly decrease on the GLY content in exposed individuals under salinity 28 in comparison to the control, while no significant differences between different concentrations and control were observed when D. neapolitana were submitted to low salinity 21 ( Figure 3B).
ii) When comparing D. neapolitana exposed to the same salinity and exposure concentration, significant differences between MWCNT materials (f and Nf) were observed in organisms exposed to both concentrations at salinity 28, showing decrease of the content in individuals contaminated with f-MWCNTs (Table 3).
iii) For each of the MWCNT (f and Nf) and exposure concentration, differences between salinities (28 and 21) were observed only at lowest exposure concentration (0.001 mg/L) for specimens under Nf-MWCNTs, with significantly higher GLY content in organisms maintained to salinity 28 in comparison to the specimens under salinity 21 ( Figure 3B).

Electron transport system (ETS) activity
i) At salinity 28 D. neapolitana presented a significant decrease of ETS activity only at 0.01 mg/L Nf-MWCNTs in comparison to the remaining conditions, while at salinity 21, no

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 20 significant differences were observed among concentrations. Regarding individuals exposed to f-MWCNTs, significantly higher ETS activity was observed in contaminated individuals under both salinity levels compared to control ( Figure 3C).
ii) When comparing specimens exposed to different MWCNTs at the same salinity and exposure concentration, significant differences between materials were observed only in polychaetes exposed to 0.01 mg/L under salinity 28 showing lower activity for individuals contaminated with Nf-MWCNTs in comparison to individuals exposed to f-MWCNTs (Table 3).
iii) For each MWCNT (f and Nf) and exposure concentration, differences between salinities (28 and 21) were only observed at 0.01 mg/L Nf-MWCNTs, with lower ETS activity in individuals maintained at salinity 28 in comparison to organisms under salinity 21 ( Figure 3C).

Lipid peroxidation (LPO) level
i) Under salinity 28 the level of LPO in polychaetes exposed to Nf-MWCNTs increased at the highest exposure concentration (0.01 mg/L) with significant differences in comparison to the other treatments, while in organisms under salinity 21 the LPO at 0.001 and 0.01 mg/L was significantly higher than levels observed in non-exposed organisms, and no significant differences were observed between individuals exposed to these two concentrations ( Figure   4A). Considering the organisms exposed to f-MWCNTs and maintained at salinity control, the level of LPO increased with the increasing of exposure concentrations with significant differences among all concentrations, while under low salinity significantly lower LPO level was observed in all contaminated D. neapolitana in comparison to control organisms.
ii) When comparing individuals exposed to different MWCNTs at the same salinity and exposure concentration, significant differences between materials were observed only in polychaetes exposed to 0.01 mg/L under salinity 28 with higher values in individuals

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 21 contaminated with Nf-MWCNTs in comparison to organisms contaminated with f-MWCNTs (Table 3).
iii) For each of the MWCNT (f and Nf) and exposure concentration, differences between salinities (28 and 21) were observed only at the highest exposure concentration for specimens under Nf-MWCNTs, with significantly higher LPO level in organisms maintained to control salinity 28 in comparison to polychaetes under salinity 21 ( Figure 4A).

Reduced (GSH) and oxidized (GSSG) glutathione ratio
i) Significantly lower ratio of GSH and GSSG was observed in organisms contaminated with 0.01 mg/L Nf-MWCNTs under salinity 28 in comparison to the remaining concentrations, while no significant differences were observed in specimens maintained at salinity 21. Under salinity 28, GSH/GSSG content in D. neapolitana exposed to f-MWCNTs decreased with the increasing of exposure concentrations with significant differences among all treatments. A similar trend was also observed for individuals submitted to salinity 21, however no significant differences were found between contaminated organisms ( Figure 4B).
ii) When comparing organisms exposed to the same salinity and exposure concentration, significant differences between MWCNT materials (f and Nf) were observed only in D. neapolitana exposed to 0.01 mg/L at salinity 28, showing the lowest ratio in individuals contaminated with Nf-MWCNTs (Table 3).
iii) For each of the MWCNT (f and Nf) and exposure concentration, differences between salinities (28 and 21) were observed in polychaetes exposed to the highest concentration of MWCNTs (both f and Nf), with lower ratio in individuals maintained at control salinity 28 compared to individuals under salinity 21 ( Figure 4B). neapolitana showed that for Nf-MWCNTs under both salinities (28 and 21), no significant differences were observed among all conditions. In polychaetes exposed to f-MWCNTs the SOD activity significantly increased with the increasing exposure concentrations under salinity 28, while under salinity 21 the activity of this enzyme significantly increased at the highest exposure concentration (0.01 mg/L) in comparison to the other treatments ( Figure 5A).
ii) Comparing organisms under the same salinity and exposure concentration, significantly higher SOD activity at 0.001 mg/L was observed in polychaetes exposed to f-MWCNTs compared to Nf-MWCNTs under salinity control, while at 0.01 mg/L significant differences between materials were found in organisms maintained under both salinities (21 and 28), showing higher activity in individuals contaminated with f-MWCNTs (Table 3).
iii) For each of the MWCNT (f and Nf) and exposure concentration, differences between salinities (28 and 21) were observed only at the lowest exposure concentration (0.001 mg/L) for specimens under f-MWCNTs, with significantly higher SOD activity in organisms maintained to control salinity 28 in comparison to organisms under salinity 21 ( Figure 5A).

Catalase (CAT) activity
i) Considering the effects of exposure concentrations, the results of CAT activity in organisms exposed to both MWCNT materials and under both salinity levels did not evidence any significant differences between concentrations ( Figure 5B).
ii) When comparing organisms exposed to the same salinity and exposure concentration, significant differences between MWCNT materials (f and Nf) were observed only in D. neapolitana exposed to 0.01 mg/L at salinity 28, showing higher activity in individuals contaminated with f-MWCNTs (Table 3).

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 23 iii) For each of the MWCNTs (f and Nf) and exposure concentration, no significant differences in terms of CAT activity were observed between salinities (28 and 21) ( Figure 5B).

Glutathione S-transferases (GSTs) activity
i) Polychaetes maintained at salinity 28 decreased significantly the activity of GSTs when exposed to 0.001 mg/L Nf-MWCNTs, but at the highest exposure concentration (0.01 mg/L) the enzyme activity significantly increased to values higher than control levels. Under salinity 21 significantly lower GSTs activity was observed in contaminated organisms with 0.001 mg/L in comparison to control individuals, however at 0.01 mg/L the activity reached again the same value of non-contaminated individual without significant differences. When organisms were exposed to f-MWCNTs under salinity 28 the activity significantly decreased only at the highest exposure concentration (0.01 mg/L) in comparison to the other treatments, while under salinity 21 the GSTs activity decreased in all contaminated organisms in comparison to noncontaminated ones ( Figure 5C).
ii) When comparing organisms exposed to the same salinity and exposure concentration, significant differences between polychaetes exposed to different MWCNTs were observed at 0.001 mg/L, with lower activity in D. neapolitana exposed to Nf-MWCNTs under salinity 28 compared to individuals exposed to f-MWCNTs. Significant differences between materials were also observed in individuals exposed to 0.01 mg/L maintained under salinity 28, showing an opposite response with higher activity in individuals contaminated with Nf-MWCNTs in comparison to f-MWCNTs (Table 3).
iii) For each of the MWCNTs (f and Nf) and exposure concentration, slight differences between salinities (28 and 21) were observed between organisms exposed to 0.001 mg/L f-MWCNTs, showing higher GSTs activity under salinity 28 compared to individuals under salinity 21. Significant differences between salinities were also recorded in individuals submitted to 0.01

A C C E P T E D M A N U S C R I P T
24 mg/L Nf-MWCNTs, with higher enzyme activity in individuals under salinity 28 in comparison to low salinity ( Figure 5C).

Neurotoxicity
Acetylcholinesterase (ATChI-ChE) activity i) At salinity 28 D. neapolitana presented a significant increase of ATChI-ChE activity only at 0.01 mg/L Nf-MWCNTs, while under salinity 21 no significant differences were observed between concentrations. Polychaetes exposed to f-MWCNTs under both salinities showed no significant differences in the ATChI-ChE activity between different treatments ( Figure 6).
ii) When comparing organisms exposed to the same salinity and exposure concentration but different MWCNTs significantly higher neuro-enzyme activity was only recorded in organisms contaminated with 0.01 mg/L Nf-MWCNTs at salinity 28 in comparison to f-MWCNTs (Table 3).
iii) For each of the MWCNTs (f and Nf) and exposure concentration, differences between salinities (28 and 21) were observed only between organisms exposed to 0.01 mg/L Nf-MWCNTs, showing higher ATChI-ChE activity under salinity 28 compared to individuals under salinity 21 ( Figure 6). . This hypothesis was confirmed in the present results due to a greater antioxidant enzymes activity such as SOD in organisms exposed to f-MWCNTs compared to Nf-MWCNTs.

A C C E P T E D M A N U S C R I P T
Interactions of CNTs with organisms can be external, as attachment of the NMs onto the skin or exoskeleton, or internal, via food intake, or both (Mesarič et al. 2015). All of these interactions can cause different physiological disturbances, as also demonstrated in the present study, and the generation of oxidative stress, which leads to toxicity with direct damage of the lipid membranes, due to the high affinity of CNTs for lipid membranes and a consequent activation of antioxidant enzymes (Mesarič et al. 2015). These responses were already demonstrated when polychaetes were exposed to different carbon NMs (De Marchi et al. 2017c;Monserrat et al. 2017). However, successful CNTs uptake in the exposed organisms are important prerequisites for bioaccumulation in the body and consequent cellular damage which are directly related to the characteristics of the CNTs such as heterogeneous purity, length, type of functionalization (Costa et al. 2016). Looking to the results of the present study, while in the organisms exposed to Nf-MWCNTs under salinity control the LPO increased only at the highest exposure concentration, in polychaetes exposed to f-MWCNTs the damage of the lipid membranes was also observed at the lowest exposure concentration, assuming that these different responses were directly related to the availability of the CNT materials. While Nf-MWCNTs, due to their poor suspendability as demonstrated by DLS analysis, could be less available for the organisms, f-MWCNTs were more dispersible in the water column probably increasing their mobility and thus may intensify the risk of exposure and toxicity and possible uptake (Jackson et al. 2013).
The generation of LPO is known to be responsible for the activation/inactivation of the antioxidant defense system (Lapresta-Fernández et al. 2012). Reduced glutathione (GSH) /

A C C E P T E D M A N U S C R I P T
28 oxidized glutathione (GSSG), considered as a regulatory molecule and sensor of the redox state of cells (Mocan et al. 2010), and antioxidant enzymes play an important role in organisms defense system against oxidative damage (Ighodaro and Akinloye 2017). In the present study, the polychaetes exposed to f-MWCNTs under salinity control showed dose-dependent increased of the LPO with a consequence dose-dependent decrease of GSH/GSSG as well as increase of SOD activity. This result suggested a compensatory response of cellular defense systems against cellular damage, while in organisms exposed to Nf-MWCNTs under control salinity, the GSH/GSSG decreased at the highest exposure concentration but the SOD activity did not increase under this condition. This behavior may be due to an excessive ROS production, especially under the highest exposure concentration, leading to oxidative damage and a loss of compensatory mechanisms as a consequence of insufficient antioxidant mechanisms (Fukai and Ushio-Fukai 2011;Walters et al. 2016) which may contribute to higher LPO levels recorded at this condition.
As multicomponent enzymes involved in the detoxification of different xenobiotics, glutathione S-transferases (GSTs) play important roles in protecting tissues from oxidative stress (Fournier et al. 1992) and they have been already used as biomarkers of cellular damage as these enzymes exhibit many of the required characteristics, i.e. specific localization, high cytosolic concentration and relatively short half-life (Pérez et al. 2004). In previously published studies GSTs showed different mechanisms of action when exposed to different NMs, assuming that GSTs activity may be either increased or decreased due to production of lipid Nf-MWCNTs (insoluble) and increased activities in organisms exposed to f-MWCNTs (soluble), both under control salinity.
Neurotoxicity was also investigated in the present study, evaluating the activity of acetylcholinesterase, AChE, which are specific esterases that mainly hydrolyse choline-based esters, several of which are used as neurotransmitters (Mennillo et al. 2017;Augustinsson, 1971). Although in recent years the number of studies that investigated the interactions between ChE and NMs have been increasing, generally demonstrating an inhibition of their activity in invertebrates as a consequence of CNTs exposure (De Marchi et al. 2017a;Monserrat et al. 2017;De Marchi et al. 2018a;, opposite results were observed in the present study, showing no inhibition of AChE activity in exposed organisms under both materials and both salinities. Such result may be related to the fact that organisms try to reduce neurotransmitter excess in the synaptic clefts, which was already showed in the bivalve Perna indica exposed to arsenic (As) (Rajkumar, 2013). II) Alteration induced by salinity shifts modifying the sensitivity of the polychaetes and the toxicity of the CNTs was also observed in the present study. It was already demonstrated that organisms exposed to salinity stress must increase their energy expenditure to successfully acclimate to the stressor and ensure cellular protection (Rivera-ingraham 2017). When organisms are exposed to low salinity level initiate a series of mechanisms (energetically costly) that allow them to hyper-regulate (i.e. to maintain their extracellular fluids at a higher osmolality than that of their surrounding medium) and this osmoregulation is considered to be an energetically costly process (Rivera-ingraham 2017). This hypothesis supported our results, showing that when the polychaetes were exposed to salinity 21 under both MWCNT materials there was an increase of the energy expenditure (showed by decrease of the GLY) and an increase of metabolic activity (expressed by an increase of ETS activity) demonstrating that the alteration induced by salinity shifts modifying the sensitivity of the polychaetes to the CNTs, but

A C C E P T E D M A N U S C R I P T
30 did not modified the toxicity of the NMs, as for both exposure conditions (f and Nf) it was possible to observe the same trend in terms of energy reserves consumption and metabolic activity in the exposed organisms.
Mitochondria, as the main energy producers in eukaryotic cells, play a central role in acclimation processes. However, they also represent the main source of reactive oxygen and nitrogen species (ROS/RNS), although the relationship between mitochondrial respiration and ROS/RNS formation is not fully understood. ROS/RNS can potentially lead to the LPO (such as those composing cellular membranes), as well as damaging other cellular molecules; ROS/RNS potentially have negative consequences for acclimation to hyper-and hypo-osmotic conditions. However, the lipid electrophiles resulting from such processes can have, along with ROS/RNS themselves, a role in the activation of cellular defences (Rivera-ingraham 2017; Sokolova 2018).
In the present study, organisms exposed at both f-MWCNTs and Nf-MWCNTs under low salinity presented an increase of LPO and activation of antioxidant enzymes in terms of increase of SOD activity as well as decrease of GSH/GSSG and decrease of GSTs especially at the highest exposure concentration, demonstrating that the alteration induced by salinity shifts on the sensitivity of the polychaetes to low salinity caused major toxicity in comparison to the chemical behaviour of both MWCNTs under this condition. In fact, despite estuarine invertebrates are often exposed to short-term (tidal) and long-term (rain periods) changes in salinity, the increased stress may lead to physiological and morphological abnormalities when exposed to low salinity (Verdelhos et al. 2015).
Nevertheless, in general, in the present study, both Nf-MWCNTs and f-MWCNTs under salinity 28 generated greater alterations on energy reserves and metabolic activity, oxidative stress biomarker responses and antioxidant enzymes activities compared to individuals maintained under salinity 21, assuming that exposed polychaetes tend to be more sensitive to the alteration induced by salinity variations on the chemical behavior of both MWCNTs in  Ward and Kach (2009) revealed that the larger aggregates can considerably increase the uptake of polystyrene NMs by suspension filter-feeding bivalves (the mussels M. edulis and oysters Crassostrea virginica), which were either dispersed or embedded within aggregates, showing that both these species more efficiently captured and ingested NMs that were incorporated into aggregates compared to those freely suspended.

A C C E P T E D
These findings are in agreement with our results, showing major toxic impacts in organisms exposed to higher salinity 28.

CONCLUSIONS
The results of the present study demonstrated clearly that both CNTs generated toxic impacts in terms of energy reserves and metabolic capacity as well as oxidative and neuro status. However, when comparing f-MWCNTs and Nf-MWCNTs, greater toxic impacts in the polychaetes D. neapolitana were observed by functionalized CNTs due to availability of the CNT materials. While Nf-MWCNTs, due to their insolubility, may be less available for the organisms, f-MWCNTs were more dispersible in the water column probably increasing their mobility and thus increasing the risk of exposure, possible uptake and toxicity, leading to a much higher cellular damage concluding that nanomaterial toxicity can be attributed to core structure and surface functionalization. Moreover, when evaluating if the alteration induced by salinity shifts could modify the sensitivity of the polychaetes and/or the toxicity of the CNTs, the present findings demonstrated that Nf-MWCNTs and f-MWCNTs under salinity 28 generated greater toxic impacts in the polychaetes compared to individuals maintained under salinity 21, assuming

A C C E P T E D M A N U S C R I P T
32 that the alteration induced by salinity shifts on the chemical behaviour of both MWCNTs and consequent fate in exposed polychaetes caused major toxicity in comparison to the sensitivity of the organisms to low salinity. Thus, nanomaterials toxicity was not only attributed to core structure and surface functionalization, but also to the physico-chemical parameters of the media which alter the behavior of the CNTs and consequently the toxicity in the exposed organisms. Considering that CNTs are one of the most promising classes of new materials to emerge from nanotechnology to date, the results of these studies provided essential information about the behavior of commercially important CNTs in complex physiological environment giving a scientifically grounded knowledge base for the risk assessment of these materials. Moreover, the complex interactions between climate change and pollutants may be particularly problematic for species living at the edge of their physiological tolerance range where acclimation capacity may be limited. On this topic, understanding how pollutants behave once reaching the environment, and how different climate related factors (e.g salinity shifts) may influence their fate, transport and toxicity, will be of major relevance to predict interactions between climate change and contaminant exposures.
Based on the results here presented, data obtained highlight the need to develop standard protocols for CNTs toxicological testing to characterize the behaviour and fate of these materials in different compartments of the aquatic environment, exposure conditions following environmental relevant concentrations and point out the importance to use a broad range of biomarkers to evaluate the possible toxic effects of these new emerging pollutants. Moreover, this study improved the understanding of biological responses of polychaetes exposed to combined CNTs and predicted climate change scenarios. Preliminary investigation on nanohorn toxicity in marine mussels and polychaetes.

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 47

Conflict of Interest
The Authors whose names are listed immediately below certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affi liations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.

ACCEPTED MANUSCRIPT
A C C E P T E D M A N U S C R I P T 53 Highlights  An innovative approach was adopted for the first time to assess the presence of CNTs aggregates in the organisms  Physiological and biochemical alterations induced in Diopatra neapolitana exposed to different CNT materials  Major toxicity caused by salinity 28 on the chemical behavior of CNTs and in exposed polychaetes in comparison to salinity 21  Greater toxic impacts induced in exposed organisms by f-MWCNTs compared to Nf-MWCNTs ACCEPTED MANUSCRIPT