AEROXIDE^®TiO₂, a mixture of rutile and anatase crystals with a nominal diameter of 25 nm (Evonik, Germany), was utilized in the present study. The stock suspension (1000 mg/L) of TiO₂ NPs was prepared using AEROXIDE^®TiO₂ powder in Milli-Q water and sonicated using an NR-50M ultrasonic homogenizer (Microtec, Japan) at 50 W for 60 min with 2 s on/off cycles placed in ice to prevent aggregation. Stock suspensions were shielded from light and stored at -20°C. For each experiment, the thawed stock suspension was similarly sonicated for approximately 60 min to disperse the TiO₂ particles, which tend to clump due to their cohesive nature, bringing them to a particle size of 25 nm. The resulting dispersion was diluted with artificial sea water (ASW), which was filtered through a 0.2µm mesh membrane filter (ADVANTEC^®, Japan) of 35ppt Coral Pro Salt (Red Sea, USA) in Milli-Q water, according to the experimental conditions.

The study employed specimens derived from the culture strain A. veneta (Ishitani et al., 2023), which has been maintained in ASW at 23°C under a 12:12 hour light/dark cycle with weekly feedings of fresh Dunaliella salina (NIES-2257). This species forms a calcium carbonate test, which is sequentially increased by the maturity phase prior to reproduction. Adult (ca. 10 chambers) specimens of A. veneta showing reticulopodia extension and activity were randomly selected from subcultures, and three individuals were transferred to each petri dish (Aznor Petri Dish φ55 × 17 mm Azwan, Japan) containing 10 mL of either TiO₂ NPs seawater medium or ASW (control). For each medium condition, three replicated petri dishes were prepared and incubated at 23°C under a 12:12 h light/dark cycle for five weeks. All cultures per experiment were run in parallel.

Four different experimental conditions were set for TiO₂ NPs concentrations, as well as the frequency of medium changes, feeding frequency, and amount of food ( Table 1 ). In both culture experiments 1 and 2, TiO₂ NPs seawater media were set up at 1, 10, and 50 ppm, in addition to the control. The frequency of medium changes differed between culture experiments 1 and 2: once a week in Experiment 1 and twice a week in Experiment 2. At each time, 80 µL of food was added to each petri dish. In culture experiment 3, TiO₂ NPs seawater media were set up at 1, 5, and 10 ppm, in addition to the control, with weekly medium changes and with feeding 40 µL of food and increased feeding to a total of 80 µL per week through addition of 20 µL food every few days. Culture experiment 4 involved TiO₂ NPs seawater media at 5 and 10 ppm, in addition to the control, with weekly medium changes with 80 µL of food and increased feeding to a total of 160 µL per week through addition of 40 µL food every few days. In experiments 2 and 4, the number of feedings and the amount of food were increased to keep double the amount of fresh D. salina in the petri dishes per week ( Table 1 ).

At the beginning of culture experiments, we photographed each individual using an inverted research microscope (IX73 Olympus, Japan) with a microscope digital camera (DS-Ri2 Nikon Japan) and assigned an individual identification number. We photographed each of all studied individual weekly to check for signs of life or death, regarding the observations: extension and movement of reticulopodia and the movement of the individual to escape from light emission of the inverted microscope for a few minutes. Specimens, which lacked these three observation points, were categorized as dead. Individual growth was evaluated by the number of chamber increases from the previous week.

We firstly tested normality and variances for total numbers of increased chambers for five weeks at each culture experimental condition with Kolmogorov-Smirnov and F test in R v.4.0.2 (R Core Team, 2021), respectively. According to the normality and variances, we conducted the Student-t and Man-Whitney-U tests in R v.4.0.2. The p values of multiple comparisons were adjusted with the Bonferroni correction.

We independently conducted a culture experiment of D. salina, a food source for foraminifers, to examine cell divisions in TiO₂ NPs seawater medium. We prepared seven replicates for each of the 10 mL culture media with 1, 5, and 10 ppm TiO₂ NPs and ASW (control) in petri dishes and added 80 µL of D. salina to each of the 28 dishes. During the 6-day culture experiments under the same conditions as foraminiferal experiments, we examined the cell density of D. salina in each of the four different media daily. All 10 mL medium was transferred into a 15 mL centrifuge tube, and the petri dish was rinsed with 3 mL ASW to recover all remaining cells. After centrifugation at 3,300 rpm for 10 min at 23°C using a micro high-speed centrifuge (CF 16RX II Hitachi, Japan), the pellet of D. salina and TiO₂ NPs was suspended with 30 µL of ASW and transferred to a 1.5 ml tube. We immediately added 0.5% xanthan gum solution to a final concentration of 0.01% to inhibit the motility of D. salina. Three replicates were made, each of which was mounted with 10 µL of cell suspension to a cell counter (AS ONE, Japan). Following the manufacturer’s protocol of the cell counter, we counted the number of cells derived from the control and 1 ppm samples using a stereomicroscope (SMZ18 Nikon, Japan) equipped with an inclined-angle lens tube. As TiO₂ NPs aggregations in the 5 and 10 ppm overlapped on D. salina and avoided observation, we captured the autofluorescence of D. salina with a digital camera (DS-Ri2 Nikon, Japan) using fluorescence observation (U-HGLGPS Olympus Japan, Excitation filter: BP530-550, Barrier filter: BA575IF) equipped with an inverted research microscope (IX73 Olympus, Japan) and counted the cell numbers.

We used six specimens: two for control and four for the 10 ppm TiO₂ NPs treatment. Two out of four 10 ppm TiO₂ NPs treated specimens formed a new chamber during the culture experiment, while the other two showed no growth. Each test was cleaned with a brush in water and dried in air. These tests were placed on carbon tape and glued to a stage where carbon deposition was performed with the carbon coater (JEE-420 JEOL, Japan). The surface structure and element analysis of foraminiferal tests were conducted with a field emission SEM equipped with an energy dispersive X-ray spectrometer (EF-SEM-EDS) (JSM-6500F JEOL, Japan).

Ammonia veneta specimens, cultured in the 5 ppm TiO₂ NPs medium, were fixed with 2.5% glutaraldehyde in filtered artificial sea water (FASW) for longer than 24 hours at 4°C. The fixed specimens were embedded in 1% aqueous agarose and then decalcified with 0.2% EGTA in 0.81 mol/L aqueous sucrose solution (pH 7.0) for several days. The specimens were rinsed with FASW, then postfixed with 2% osmium tetroxide in FASW for 2 hours at 4°C, dehydrated in a graded ethanol series, and embedded in epoxy resin (Quetol 651). Ultrathin sections (100 nm for observation and 200 nm for analysis) were cut using a diamond knife on an Ultracut S ultramicrotome (Leica Reichert, Germany), stained with 2% aqueous uranyl acetate and lead staining solution (0.3% lead nitrate and 0.3% lead acetate), and examined using TEM equipped with an energy dispersive X-ray spectrometer (TEM-EDS) (FEI Tecnai G2 20, USA) operated at 200 kV. Elemental mapping was performed in scanning TEM (STEM) mode.

Living specimens of A. veneta are filled with cytoplasm inside the test and extend fine reticulopodia radially from the aperture and pores of the test to capture food particles in their surroundings ( Figure 1 ). In 1 ppm TiO₂ NPs seawater medium, TiO₂ NPs aggregate up to ~2 µm within 24 hours and accumulate with other wastes on the bottom (Ishitani et al., 2023). TiO₂ NPs is aggregating faster in high TiO₂ NPs concentration: over 1 µm in hydrodynamic diameter within 1 hour in 5 and 10 ppm TiO₂ NPs seawater and over 15 µm in 50 ppm TiO₂ NPs seawater (Morelli et al., 2018; Reyes et al., 2021; Palmeira-Pinto et al., 2023). These aggregates visibly settled at the bottom of the petri dish as brownish substances during the culture experiment with TiO₂ NPs seawater media ( Figure 1 ). However, these depositions were not observed in the surrounding areas of living A. veneta, particularly around the reticulopodia extensions ( Figure 1 ). At the onset of the culture experiment, all studied specimens extended fine reticulopodia. In the control and 1 ppm TiO₂ NPs seawater medium, specimens showed reticulopodia activity and also added new chambers and grew cytoplasm, which typically filled the interior of the test. Under TiO₂ NPs seawater media concentrations greater than 10 ppm, reticulopodia were retracted, and cytoplasm continuously shrank ( Figures 2A–D ). After reduction of cytoplasm, any reticulopodia activity and movement of specimens were not observed. In these conditions, aggregations containing TiO₂ NPs were observed around foraminiferal specimens.

Holes appeared on the surface of the tests in some specimens exposed to TiO₂ NPs seawater media concentrations higher than 10 ppm ( Figure 3 ). After a 5-week experiment, additional holes were observed on the umbilical side of the test.

In Experiment 1, control specimens increased by an average of ~2 chambers every week, reaching a total of ~9 chambers over five weeks ( Figure 4A ; Supplementary Table S1 ). One out of nine control specimens reproduced in the third week. By the end of the 5-week experiment, three specimens had reproduced clones, forming an average of ~20 chambers during their entire lifespan. In the 1 ppm TiO₂ treatment, eight out of nine specimens increased by ~1 chamber per week, although five specimens died after 3 and 5 weeks, respectively. In the 10 ppm TiO₂ treatment, six out of nine specimens increased by ~1 chamber per week, but all specimens ceased growth by the third week and died by the fourth week. In the 50 ppm TiO₂ treatment, only one specimen increased by 1 chamber in the first week, but all specimens died by the fourth week. The total number of increased chambers over the 5-week experiment showed a significant difference between the control specimens and the 10 and 50 ppm TiO₂ treated specimens ( Figure 4B ).

In Experiments 2, all specimens, both in the control and the 1 ppm TiO₂ treatment, increased by 6–8 chambers over 5 weeks ( Figure 5A ; Supplementary Table S1 ). During the experiment, four and five specimens reproduced clones in the control and the 1 ppm TiO₂ treatment, respectively. However, in both the 10 and 50 ppm TiO₂ treatments, all specimens showed no growth and died by the fourth week. The total number of increased chambers over the 5-week experiment was the same between the control and the 1 ppm TiO₂ treated specimens, but there were significant differences between the controls and the 10 and 50 ppm TiO₂ treatments ( Figure 5B ).

In Experiment 3, all specimens, both in the control and the 1 ppm TiO₂ treatment, increased by ~8 chambers over 5 weeks ( Figure 6A ; Supplementary Table S1 ). During this experiment, one specimen reproduced clones in each of the control and the 1 ppm TiO₂ treatments, but some specimens died. In the 5 ppm TiO₂ treatment, three out of nine specimens increased by ~1 chamber per week, but all had stopped growing and died by the third week. In the 10 ppm TiO₂ treatment, all individuals showed no growth and died by the third week. The total number of increased chambers over the 5-week experiment was the same between the control and the 1 ppm TiO₂ treated specimens, but there were significant differences between the controls and the 5 and 10 ppm TiO₂ treatments ( Figure 6B ).

In Experiment 4, eight out of nine control specimens increased by ~8 chambers over 5 weeks ( Figure 7A ; Supplementary Table S1 ). At the end of the experiment, three specimens reproduced clones. In the 5 ppm TiO₂ treatment, four out of nine specimens increased by ~1 chamber per week, but they stopped growing by the third week and all died by the fifth week. In the 10 ppm TiO₂ treatment, only two specimens increased by one chamber in the first week, but all individuals died by the third week. The total number of increased chambers over the 5-week experiment showed significant differences between the controls and the 5 and 10 ppm TiO₂ treatments, as well as those of Experiment 3 ( Figure 7B ).

Morphological differences between control and 10 ppm TiO₂ treated specimens were observed on the surface of the tests, but pores were clearly opened on both specimens ( Figures 8A–D ; Supplementary Figure S1 ). The test surface was smooth on control specimens but slightly rough on 10 ppm TiO₂ treated specimens ( Figures 8B, D ). Moreover, some adhesive materials were observed on the test surface of TiO₂ treated specimens, even when cleaned with brushes. These adhesive materials were often observed from the 10 ppm TiO₂ treated specimens, which showed no growth during the culture experiment ( Supplementary Figure S1 ). The SEM-EDS mapping showed a trace amount of titanium from adhesive materials and on the surface of the test but not covering the aperture and pores of the test ( Figures 8E, F ).

Time-series changes in D. salina cell density were recorded for each culture condition, including control and 1, 5, and 10 ppm TiO₂ seawater media, over a 6-day period. In the control, D. salina cell density continuously increased for four days, reaching a stationary phase thereafter ( Figure 9 ). Over the 6-day experiment, the cell density was approximately 10 times higher than the initial density. In the 1 ppm TiO₂ seawater medium, cell density started a slow increase by the second day, significantly rising afterward and reaching a level comparable to the control by the fifth day. However, in the 5 and 10 ppm TiO₂ seawater media, cell densities slightly increased but remained at 2000 cells/mL throughout the experiment. By the end of the experiment, cell densities in the 5 and 10 ppm TiO₂ seawater media were less than half of those in the control and the 1 ppm seawater medium.

Angular-shaped grains with distinct contours were observed in vesicles distributed in the latest three chambers by TEM observations ( Figures 10A–C ; Supplementary Figure S2 ). These grains, several tens of nanometers in size, exhibited a darker contrast than the surrounding organics and resin, indicating a heavier and more electronically dense element. STEM-EDS analysis detected titanium X-ray peaks from these angular grains ( Figures 10D–F ), but no titanium peaks were detected from the background ( Figures 10G, H ). Additionally, STEM-EDS mapping identified abundant titanium-bearing grains in the vesicles, each tens of nanometers in size ( Figures 11A, B ).

Aggregates of angular grains covered with light-contrast materials were found along the cell, relevant to the surface of the decalcified test ( Figure 12A ). STEM-EDS analysis detected titanium X-ray peaks from these grains, as well as those in vesicles ( Figure 12B ), while no titanium peaks were detected from the background ( Figure 12C ).