Nanoplastic incorporation into an organismal skeleton

Studies on the effects of global marine plastic pollution have largely focused on physiological responses of few organism groups (e.g., corals, fishes). Here, we report the first observation of polymer nanoparticles being incorporated into the calcite skeleton of a large benthic foraminifera (LBF), a significant contributor to global carbonate production. While previous work on LBF has documented selectivity in feeding behaviour and a high degree of specialization regarding skeletal formation, in this study, abundant cases of nanoplastic encrustation into the calcite tests were observed. Nanoplastic incorporation was associated with formation of new chambers, in conjunction with rapid nanoplastic ingestion and subsequent incomplete egestion. Microalgae presence in nanoplastic treatments significantly increased the initial feeding response after 1 day, but regardless of microalgae presence, nanoplastic ingestion was similar after 6 weeks of chronic exposure. While ~ 40% of ingesting LBF expelled all nanoplastics from their cytoplasm, nanoplastics were still attached to the test surface and subsequently encrusted by calcite. These findings highlight the need for further investigation regarding plastic pollution impacts on calcifying organisms, e.g., the function of LBF as potential plastic sinks and alterations in structural integrity of LBF tests that will likely have larger ecosystem-level impacts on sediment production.

www.nature.com/scientificreports/ mineralization. Formation of new test chambers in these organisms is initiated through a cytoplasmic extrusion of the LBF, providing the shape for the newly forming chamber and a thin primary organic layer as the base for CaCO 3 precipitation 28,29 . During a secondary calcification stage, the chamber wall thickening, the majority of CaCO 3 is precipitated 30 . Like other photosymbiotic LBF, Amphistegina spp. (referred to as Amphistegina hereafter) exhibit a facultative heterotrophic mode, acquiring nutrients through both symbiont autotrophy and heterotrophic feeding [31][32][33][34] . As heterotrophic feeding is vital for the growth of these organisms 31 , organismal ingestion and egestion mechanisms of nanoplastics were assessed in food choices of: (1) nanoplastic particles only (fluorescent functionalized polystyrene; 'nanoplastic-only treatment'), (2) a mixture of Nannochloropsis microalgae and nanoplastic particles ('mixed treatment'), and (3) microalgae only ('control treatment'). The size of nanoplastic particles and microalgae (~ 1 μm) was chosen to correspond to the size of microalgae naturally occurring in the marine environment, to minimize potential bias to size.

Nanoplastic incorporation into skeletal structure of LBF
This study documents the first known instance of nanoplastic incorporation into the skeleton of a calcifying organism (Fig. 1). Ingestion responses were documented with fluorescence microscopy after 1-day and 6-week exposure to nanoplastic and egestion after an additional 2-week recovery (Fig. 2). Subsequently, skeletal incorporation was investigated using Scanning Electron Microscopy (SEM). At the test surface of newly formed chambers, nanoplastic particles appear to be encrusted by calcite, primarily observed close to ornamental spikes at the aperture (Fig. 1a). Encrusted nanoplastic particles formed dome-like structures often in accumulations of larger aggregates (> 5 particles) observed in various degrees of encrustation ( Fig. 1c-e). As the LBF aperture is an area of active food uptake, the incorporation of nanoplastic into the test likely occurred in conjunction with ingestion and incomplete egestion processes. While previous studies of foraminifera 22,23,25 have observed the uptake of nanoplastic particles, egestion of foreign particles is largely unknown. Following the 2-week recovery period, the percentage of LBF containing nanoplastic significantly decreased compared to 6-week exposure (F 2,95 = 22.73, p < 0.001; Supplementary Table 2a). However, nearly half of the total LBF still contained nanoplastic particles www.nature.com/scientificreports/ (47.0 ± 5.5% LBF, n = 160). While LBF seemed to efficiently egest those particles from their inner cytoplasm, they appear to have stuck to the outer primary organic layer present at the sites of calcification 29,35,36 and were thus passively incorporated in the calcite test. The calcite overgrowth on nanoplastic particles appears as calcite crusts, likely precipitated in the process of chamber wall thickening or new chamber formation 35,37 (Fig. 1e). Previous studies document Amphistegina calcification mechanisms through the initial formation of calcite spheres that subsequently are bound by an organic matrix (seen as organic filaments) to form a primary wall structure 35 . Although these calcite spheres are slightly larger 35 (~ 2-3 μm) than the nanoplastic spheres used in this study, similar organic filaments were observed spanning across multiple nanoplastic particles. This indicates that nanoplastic spheres could be perceived by the LBF as potential 'building-material' for a newly forming chamber or during localized repair processes. The www.nature.com/scientificreports/ process of biomineralization has been previously described in detail for the rotaliid foraminifera Ammonia beccarii 38 . The calcification stages of A. beccarii, in particular the calcification between organic layers 38 , resembles structures seen at nanoplastic incorporation sites investigated here (Supplementary Figs. 3 and 4). Nanoplastic spheres could even present a possible nucleation site for crystal formation 39 , as previously seen with bacteria and diatoms 40,41 , thereby acting as catalysts in LBF chamber formation or localized repair.

Ingestion patterns have no effect on skeletal incorporation
LBF generally utilize reticulopodia (a temporary extension of cytoplasm) to migrate food particles to aperture openings, subsequently engulfing particles via vacuolization 42 . LBF that were incubated in nanoplastic-only treatments exhibited a regulatory ability to avoid nanoplastic ingestion after 1 day of nanoplastic exposure (35.0 ± 5.1% LBF ingestion, n = 80, mean ± SE; Fig. 3). However, this potential regulatory ability against nanoplastic uptake was muted during a subsequent 6-week nanoplastic exposure, as seen in the high nanoplastic ingestion occurrences (71.0 ± 6.2% LBF ingestion; n = 80). In nanoplastic-only treatments, LBF specimens showed reduced growth (28.3 ± 8.9 μm growth in 8 weeks) compared to the mixed treatment (80.7 ± 12.9 μm growth) and the control treatment (96.6 ± 10.8 μm growth; F 2,95 = 4.32, p = 0.02, Supplementary Table 1). As LBF growth rates and calcification partly depend on heterotrophic nutrient uptake of LBF 31 , the significantly decreased growth of LBF is likely due to providing nanoplastic without natural food sources, as nanoplastic presumably has no nutritional value. The frequency of initial nanoplastic ingestion was significantly influenced by the presence of a natural food source. The presence of microalgae in the mixed treatment stimulated and significantly increased the initial feeding response during 1-day exposure (52.5 ± 5.8% LBF vs. 35.0 ± 5.1% in nanoplastic-only treatment; n = 80) compared to the nanoplastic-only treatment, and continued to increase during 6-weeks exposure (81.0 ± 6.7% LBF ingestion vs 71.0 ± 6.2% LBF ingestion; n = 80). This supports the importance of mimicking natural conditions in feeding experiments, showing that nanoplastic uptake in the presence of naturally occurring food sources is more common 24 . Although the ingestion of nanoplastic particles was initially increased in the presence of the natural food source, these ingestion patterns had no effect on egestion and skeletal incorporation, as the difference in egestion between the different treatments was insignificant (~ 40% egestion in both, nanoplastic-only and mixed treatment). Passive ingestion of nanoplastic through the pores was clearly distinguishable from active ingestion through the aperture and occurred exclusively in specimens that bleached during the experiment. Thus, we conclude that ingestion of nanoplastics is the result of active feeding. Bleaching of LBF holobionts remained low in all treatments (< 10%; Supplementary Table 1) and was not significantly altered by nanoplastic presence during the exposure period studied here (F 2,95 = 1.11, p = 0.33, Supplementary Table 3). This contrasts previous studies that have shown that diatoms can be negatively impacted by microplastic exposure 43,44 .

Implications for LBF and their sedimentological and ecological importance
In this study, we did not observe any deleterious effects of nanoplastics on the LBF calcareous tests (i.e., no test dissolution, breakages, major deformations). No effect on growth was observed when microalgal food sources were provided (i.e., in the mixed treatment; Supplementary Table 3). The frequent ingestion of nanoplastic particles in this study suggests an alternative mode of plastic invasion into marine food webs via unicellular organisms 45,46 , exacerbating bioaccumulation risks. Although egestion pathways reduce the residence time of microplastic inside the cytoplasm, egestion efforts without the benefit of nutrient uptake will potentially have negative impacts on foraminiferal energy budgets and consequentially on calcification and reproduction 31,33 .
Despite the partial egestion of ingested nanoplastic particles from the foraminiferal cytoplasm, particles were found to be abundantly incorporated into the test as a result of calcification. Skeletal incorporation of any type of foreign grains has previously not been documented for calcifying foraminifera, or any other calcite www.nature.com/scientificreports/ precipitating organisms. Further research employing sectioning methods like FIB-SEM for exploring the spatial relations of the skeletal carbonate and nanoplastics will allow for deeper insights into the encrustation. Additional work is also needed to ascertain further impacts of nanoplastics on LBF physiology and biomineralization. This is especially the case for longer-term or permanent nanoplastic exposure, as constant nanoplastic incorporation might promote skeletal malformations to a higher degree than relatively short-term exposure. One could hypothesize that the test structure and properties of hyaline foraminifera might change through the incorporation of nanoplastics, with potentially negative effects for the light conditions for symbiont photosynthesis, or test stability. Adhesion, ingestion and skeletal incorporation of nanoplastics could also become a potential sink for nanoplastic pollution, as previously hypothesized in the case of scleratinian corals 13,47,48 . Since LBF are essential components of tropical coral reef communities, the large-scale incorporation of nanoplastic into LBF tests as well as the potential consequences (e.g., test instability, toxicity) could influence ecosystem functions, such as carbonate production and coastline stability.

Methods
Collection and culture maintenance. We used specimens of Amphistegina gibbosa and Amphistegina lobifera from established long-term cultures in the marine experimental facility MAREE at the Leibniz Centre for Tropical Marine Research (ZMT) in Bremen, Germany. The collection of these LBF was described by Stuhr et al. 34,49 . The cultures have been maintained at the ZMT for 4-6 years at the time this experiment was conducted (summer 2021).
In culture, the LBF were kept in 500-ml containers filled with freshly made artificial seawater (made with Red Sea Salt) at a temperature of ~ 24 °C and light intensities of < 30 µmol m −2 s −1 (using a JBL Solar Ultra MARIN Day 15000K fluorescent light). It is assumed that the LBF used in this study are clonal progeny from the original cohort. During the culture period prior to the start of this experiment, the LBF used in this study were fed once a month with diluted Nannochloropsis algae concentrate (12 × 10 9 cells ml −1 , BlueBioTech GmbH, Germany). Culture maintenance is described further by Schmidt et al. 50 .

Experimental setup
Food choices. To understand the impact of nanoplastic presence on physiological performance, feeding behavior and calcification of LBF, two food sources were provided in this study. The first was sterilized Nannochloropsis algae (12 × 10 9 cells ml −1 , BlueBioTech GmbH, Germany), a natural food source the LBF regularly encounter. The algae stock solution was created freshly prior to each feeding session, by diluting 0.015 ml algae concentrate in 100 ml seawater (concentration of stock solution: 3.60 × 10 8 cells ml −1 ). The stock solution was autoclaved to avoid algal blooms inside the experimental treatments. The second food source were nanoplastic particles (Polystyrene (PS) grains; Fluoresbrite ® Polychromatic Red Microspheres, 1.0 μm, 4.55 × 10 10 particles ml −1 ). A stock solution of the nanoplastic concentrate was created freshly before feeding, by diluting 0.1 ml concentrate in 5 ml seawater (concentration of stock solution: 4.55 × 10 9 particles ml −1 ). Both food particles, algae and nanoplastic, were approximately in the same size range (~ 1 μm). Nanoplastic concentration in nanoplasticonly and mixed treatments was 1.8 × 10 7 particles ml −1 . Nannochloropsis concentration in mixed and control treatments was 1.8 × 10 4 cells ml −1 .
Experimental exposure. For the duration of the experimental exposure to different food choices, LBF were kept in 12-well polystyrene plates (5 LBF per well; CELLSTAR ). Three experimental treatments were set up: nanoplastic particles as the only food source ('nanoplastic-only'; 9 × 10 7 particles per 5-ml well), nanoplastic particles and algae ('mixed'; 9 × 10 7 nanoplastic particles and 9 × 10 4 Nannochloropsis cells per 5-ml well), Nannochloropsis algae only ('control'; 9 × 10 4 cells per 5-ml well). Each treatment consisted of 80 LBF with 16 replicates. The concentration of Nannochloropsis cells inside the wells was relatively low to remain close to the culturing conditions. There was no direct comparison between the amount of algae uptake and nanoplastic uptake. Prior to the experiment start, the vitality of all LBF used in this experiment was assessed visually through symbiont coloring and symbiont fluorescence intensity, to assure the health of all LBF. The chosen LBF were then fed weekly with the respective food choices over a course of 6 weeks. For an additional 2 weeks of recovery period, nanoplastic was not added to allow for detection of egestion, however, microalgae were still provided.
Evaluation of ingestion, vitality, and size. Every week, all LBF were rinsed individually with fresh seawater to avoid excessive accumulation of nanoplastic and algae inside the wells and on the LBF itself. This also assured that all nanoplastic that was covering only the outside test of the LBF would not be included in the ingestion identification. Pictures of the LBF were taken after 1 day, 6 weeks and 8 weeks (6 weeks exposure + 2 weeks recovery) to assess feeding response (ingestion of nanoplastic) and vitality (i.e., symbiont coloring and presence of symbiont fluorescence as an indicator for bleaching). Ingestion was assessed by counting the LBF individuals that had nanoplastic particles solely inside their first (newest) chamber and LBF that showed nanoplastic particles further inside their test. Sizes of all LBF individuals were monitored by measuring the diameter of the specimens prior to the experiment start, after 1 day, 6 weeks and 8 weeks on optical micrographs taken with the Keyence Digital Microscope VHX-2000. Fluorescence imaging was done with an additional fluorescence adapter (Filter set Cyan, Excitation 490-515 nm, NIGHTSEA, USA).

Documentation of nanoplastic incorporation.
Sections of individual LBF were prepared to detect the exact position of nanoplastic inside the LBF and account for potential incorporation of nanoplastic particles into the calcite test. High resolution fluorescence microscopy was done using a Leica DM6000B microscope. Potential nanoplastic incorporation sites were further analyzed with scanning electron microscopy (SEM; Teneo Thermo www.nature.com/scientificreports/ Fisher Scientific). Iridium coating was used to obtain SEM images at multiple magnifications up to 80.000×. For the highest magnification of 80.000×, 2 kV high voltage and a working distance of 2 mm was used. All other images were taken at 5 kV high voltage and working distances of 10 mm. As aggressive removal of organics might have damaged the nanoplastic beads, the methods chosen here leave the possibility that remains of extracellular polymeric substances (EPS) have stayed in the sample and might have influenced the surface. EDX analysis was conducted to confirm the presence of CaCO 3 encrustations ( Supplementary Fig. 5). For the EDX an energy of 10 keV was applied, resulting in a penetration depth which is larger than particle size. Thus, the interpretation of the EDX data is limited as the foraminiferal shell material underlying the nanoplastics particle could have been excited. Nevertheless, the sharp peaks argue against a pure substrate signal. Evidence of nanoplastic incorporation is mainly based on morphological information acquired from SEM. Data analysis. The number of LBF that ingested nanoplastic was counted within treatments and converted to percentages (% LBF per treatment that ingested nanoplastic). The same was done for the number of bleached LBF (as an indicator for vitality). A two-way ANOVA was conducted to detect significant differences in feeding responses (total ingestion of nanoplastic and nanoplastic inside the first chamber only), vitality and organism size between treatments. Treatment (nanoplastic-only, mixed, control) and duration (1 day = T1, 6 weeks = T2, 8 weeks = T3) were assigned as fixed factors. Prior to the ANOVA analyses, all data were tested for normality (Shapiro-Wilk) and homogeneity of variance (Levene's test). Normality and homogeneity assumptions (p > 0.05) were met in size and bleaching measurements, however, the nanoplastic feeding observations were not normally distributed. We still proceeded with the analysis due to the robust nature of ANOVA tests. All ANOVA analyses were performed in R Version 4.0.2 using the R stats package 51 and vegan package 52 .

Data availability
The data generated or analyzed during this study are included in the supplementary information files of this published article.