Quantitative Visualization of Lanthanum Accumulation in Lanthanum Carbonate-Administered Human Stomach Tissues Using Mass Spectrometry Imaging

Platinum, a transition metal that is widely used in anti-cancer agents, also results in the development of nephropathy due to severe adverse reactions caused by platinum-induced nephrotoxicity. Reports on imaging with metals other than platinum remain are limited, even in preclinical studies. Furthermore, most of these are case reports, and the relationship between the distribution of the metal and clinical observations in human samples is not well understood. Here we report on visualizing lanthanum (139La), a component of Fosrenol, which is usually used for the treatment of hyperphosphatemia. Gastric inflammation, also known as hemorrhagic gastritis, is the main adverse event caused by Fosrenol. To conduct this study, 139La was visualized in gastric biopsy samples obtained from a patient using quantitative laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS). We also compared the distribution of 139La in tissue and histochemical results. The areas where 139La accumulated corresponded to the macrophage-positive areas observed in immunohistochemistry studies using an anti-CD68 antibody. In contrast, we observed a debris-like crystal morphology in hematoxylin and eosin staining tissues. The debris was also associated with 139La accumulation. The abnormal accumulation of 139La crystals caused the observed inflammation. This phenomenon was previously characterized, but this is the first report in which 139La distribution and histochemical results are compared using LA-ICP-MS.


INTRODUCTION
Metal imaging using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) has recently become a popular method that is used in many research elds, including basic biology and toxicology. An important feature of LA-ICP-MS imaging is its ability to provide quantitative data, including a wide dynamic range because the metals and e ciently ionized, even on the tissue surfaces. erefore, many reports concerning metal imaging using LA-ICP-MS have appeared, including those on the endogenous distribution of metals in animal and human tissues, [1][2][3] toxicological studies related to exogenous cadmium in rat fetuses, 4) e ects of dibutyltin on placental and fetal toxicity in rats, 5) and lead poisoning via accumulation in the bones of water birds. 6) Additionally, toxicological studies of pharmaceuticals that contain metals are also important applications of LA-ICP-MS imaging. [7][8][9][10][11] Pharmaceutically relevant metals include alkali, alkaline earth, and transition metals and are used widely in drug development. Table 1 provides a summary of pharmaceuticals and diseases that have been investigated using LA-ICP-MS. As shown in Table 1, there are many reports on anti-cancer agents containing platinum, which inhibit DNA synthesis. A major adverse event of these platinum-based compounds includes nephrotoxicity, and almost all reports involved the visualization of the distribution of platinum inside kidney tissues. 11) As can be DOI: 10.5702/massspectrometry.A0086 Page 1 of 5 (page number not for citation purpose) seen in Table 1, animal model studies predominate, and only a few reports on metal visualization using human clinical samples are available.
In this study, we visualized lanthanum ( 139 La) using human gastric biopsy tissues using a quantitative imaging method. 12) Lanthanum, a component of the drug Fosrenol, is present in the form of lanthanum carbonate hydrate and is used in the treatment of hyperphosphatemia. e main adverse event of this drug is hemorrhagic gastritis. In this study, we compared the visualization of 139 La based on ion images and histological observations in an attempt to elucidate hemorrhagic gastritis. Our ndings revealed that a signi cant accumulation of 139 La corresponded to macrophage-positive areas, as visualized using anti-CD68 antibodies.

Specimen collection
e patient (a 66-year-old male) with hemorrhagic gastritis was subjected to gastric biopsies at the Oka hospital (Honjo, Saitama, Japan), which were collected for inspection. Written informed consent was obtained from the patient as a comprehensive agreement for future studies. e obtained tissues were stored as formalin-xed, para n-embedded (FFPE) tissues. e samples were then anonymized during their analysis. For the LA-ICP-MS analysis, this study was approved by the Institutional Review Board of the University of Tokyo (17-29; Visualization of lanthanum distribution included in Fosrenol at in ammatory sites). Sectioning e FFPE sample was sliced into 8 µm sections using a rotary microtome (RM2145, Leica, Wetzlar, Germany) and mounted on glass slides in a water bath at 40°C. e tissue slides were dried overnight at 37°C on a attering plate (Sakura Finetek Japan, Tokyo, Japan). Serial sections were prepared and used for LA-ICP-MS imaging, immunohistochemistry, and hematoxylin and eosin (H&E) staining.

Depara nization
Depara nization was performed by three cycles of washing with xylene at a temperature of approximately 25°C for 5 min. e sample for LA-ICP-MS analysis was directly dried to avoid the delocalization of the analytes. Samples for immunohistochemistry or H&E staining were further washed with 100% ethanol (10 min, twice), 95% ethanol (10 min, twice), and water (5 min, twice). A er washing, the staining procedure was performed.

Hematoxylin and eosin staining
e H&E staining protocol a er depara nization was as follows: tissue slides were stained with Mayer's hematoxylin solution for 1 min and washed with water (1 min, twice). e slides were then stained with 1% Eosin Y solution for 1 min and washed with water (1 min, twice). A er staining, the samples were dehydrated in ethanol (1 min, three times), and permeation was performed with xylene (1 min, twice). e resulting stained tissues were then mounted with a coverslip using a mounting medium and dried overnight.

Immunohistochemistry using an anti-CD68 antibody
CD68 is a transmembrane glycoprotein that is highly expressed on monocytes and macrophages. Herein, immu-

LA-ICP-MS imaging experiment
Laser ablation was performed using an in-house laser ablation system equipped with an IFRIT femtosecond laser (Cyber Laser, Tokyo, Japan). e laser ablation system was connected to iCAP-Qc ( ermo Fisher Scienti c, Bremen, Germany). e operational settings of the LA-ICP-MS instrument were optimized to obtain the maximum intensity for 139 La using a NIST SRM 610 glass standard material. e instrumentation and analytical conditions are summarized in Table 2.

Image reconstruction and quanti cation
Image reconstruction was performed using the iQuant2 so ware. 13) For quantitative imaging, a previously published method calibrated by NIST SRM 610 was used. 12)

Quanti cation using NIST SRM610
To convert the raw ion count into concentration in the human tissue sample, 139 La detection was performed using the NIST SRM610 standard under the same conditions as were used for the tissues. e resulting ion count is shown in Fig. 1a. According to the SRM610 measurement, a plateau in ion intensity was observed on the glass, at a 139 La count of 3.2×10 6 cps. However, since the ablated volumes would di er between the tissue section and glass standard, a correction for the ablated volume di erence was made. e trace of the laser ablation was examined by optical microscopy to estimate the ablated volume of the glass standard for comparison to the ablated volumes of the tissue sections with glass. e obtained diameter and depth values were found to be 12 and 4 µm, respectively, and an intensity value of 3.2×10 6 cps was obtained from this spot. Using the measured volumes of the standard glass and tissue section, the signal intensity was corrected using the volume ratio based on the equation below, and quantitative values were obtained from the imaging results.
where C sam and C std are the concentrations of the sample and standard, respectively; I sam and I std are the ion counts of the sample and standard, respectively; V sam and V std are the volumes of laser ablation; and ρ std and ρ sam are the densities of the sample and standard, respectively. Based on a study by Pearce et al., the overall average concentration of La in NIST SRM610 was determined to be 509.2±138.0 µg g −1 (ppm). us, ρ std was reported as 2.6 g/cm 3 , and ρ sam was de ned as 1.0 g/cm 3 .

Quantitative lanthanum imaging
Quantitative La imaging was performed using the conversion method described above. Two biopsies (Figs. 2a and 2b) from the same patient were used in this experiment. As shown in Fig. 2a, a region in which La had accumulated was observed in the rightwards section. To visualize the tissue morphology, we also mapped the distribution of phosphorus and provide overlay images of 139 La (magenta) and 31 P (green). Comparing the anti-CD68 staining results and the La distribution, the positive regions corresponded to the La accumulated regions. e highest section in Fig.  2a was approximately 10 −2 ppm/pixel, whereas a hot spot was observed at the bottom le side at 10 −3 ppm/pixel. Interestingly, some of the anti-CD68 positive sections did not show correspondingly high 139 La concentrations, especially those shown in Fig. 2b. Based on this result, we conclude that anti-CD68 positive sites were observed even in areas where the concentration of 139 La was not high because the in ammatory site had spread mainly from the core of the in ammation where large amounts of 139 La had accumulated.
La imaging is typically performed using optical microscopy, scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDS). [14][15][16] However, these studies only report on the qualitative distribution of 139 La based on macroscopic and microscopic observations. Quantitative data were reported by Namie et al. using ICP-MS, 17) but the authors did not report on the spatial distribution of 139 La. In addition, the 139 La distribution results were not directly compared with histocytes that were positive for CD68 using  18) However, the signal intensity and spatial resolution were insu cient to permit the level of accumulation of 139 La to be determined. To date, no reports regarding quantitative 139 La distribution coupled with CD68 immunohistochemistry have appeared. erefore, our results using LA-ICP-MS represent the rst report of the quantitative determination of the distribution of 139 La coupled with excellent agreement for histocytes. In addition, as shown in Fig. 2, the areas of the highest accumulation of 31 P were consistent with the highest accumulation of 139 La. In previous studies, these distributions were presumed to indicate the presence of lanthanum phosphate. 19) We conclude, however, that the observed colocalization is due to the binding of 139 La with 31 P which is located on the inner wall of the stomach, with the formation of insoluble lanthanum phosphate.

High-magni cation observations in H&E stained tissues
Serial sections were prepared for LA-ICP-MS, anti-CD68 staining, and H&E staining and a representative region with a high accumulation is shown in Figs. 2a and 2b. For this tissue, the same area was observed by H&E staining at high magni cation (×40 and ×100; Figs. 3a and 3b). Interestingly, a large amount of crystal-like debris was found in the cytoplasm around the in ammatory regions. In contrast, these morphological materials were not observed near areas of low or zero 139 La concentrations (Fig.  3c). erefore, it can be concluded that this debris was derived from 139 La crystals. e debris was in the form of elongated crystals, similar to those reported by Murakami et al. and Makino et al. 18,19) In ammation is considered to be induced by these crystal-shaped La formations. In addition to these crystals, leukocytes (mainly lymphocytes) were observed around the debris, supporting our interpretation (Fig. 3b).

CONCLUSION
We report herein on the quantitative visualization of La using LA-ICP-MS in gastric biopsy samples obtained from a Fosrenol-administrated patient. An in ammation response occurred when La was accumulated at a concentration of a few hundred ppm. is accumulation was in good agreement with the in ammatory response, and a signi cant amount of debris was observed in these areas, as evidenced by H&E staining. e debris was likely derived from crystals of La. To date, investigations of the distribution of La in human clinical samples using LA-ICP-MS has not been reported. is study provides the rst clear correlation between the in ammatory response and the accumulation of La with a high, quantitative spatial resolution.  In both sections, 139 La was mainly accumulated at highly CD 68 positive sites. From the quantitation using trace elements in the glass, the maximum concentrations were 1.9×10 4 and 9.3 ×10 3 ppm, respectively. Phosphorous images are also provided to con rm tissue morphology (1 µg/g=1 ppm). Scale bars: 500 µm.