Electronic microscopy evidence for mitochondria as targets for Cd/Se/Te‐based quantum dot 705 toxicity in vivo

Abstract

The safety of quantum dots (QDs) 705 was evaluated in this study. Mice were treated with QD705 (intravenous) at a single dose of (40 pmol) for 4, 12, 16, and 24 weeks. Effects of QD705 on kidneys were examined. While there was a lack of histopathology, reduction in renal functions was detected at 16 weeks. Electron microscopic examination revealed alterations in proximal convoluted tubule (PCT) cell mitochondria at even much earlier time, including disorientation and reduction of mitochondrial number (early change), mitochondrial swelling, and later compensatory mitochondrial hypertrophy (enlargement mitochondria: giant mitochondria with hyperplastic inner cristae) as well as mitochondrial hyperplasia (increase in mitochondrial biogenesis and numbers) were observed. Such changes probably represent compensatory attempts of the mitochondria for functional loss or reduction of mitochondria in QD705 treated animals. Moreover, degeneration of mitochondria (myelin‐figure and cytoplasmic membranous body formation) and degradation of cytoplasmic materials (isolated cytoplasmic pockets of degenerated materials and focal cytoplasmic degradation) also occurred in later time points (16–24 weeks). Such mitochondrial changes were not identical with those induced by pure cadmium. Taken together, we suggest that mitochondria appeared to be the target of QD705 toxicity and specific mitochondrial markers may be useful parameters for toxicity assessments of QDs or other metal‐based nanomaterials.

Introduction

Quantum dots (QDs) 705 is a cadmium/selenium/tellurium‐based QDs with unique fluorescent properties and was found to be highly suitable for biomedical applications [1]. Because the core of QD705 consists of toxic metals, cadmium and tellurium, the “safety” of QD705 for biomedical utilization has remained a serious concern. While there is no lack of investigation on the safety or toxicity of QD705 [[2], [3], [4], [5]], there was a concern that many of these studies were not conducted by toxicologists or evaluated by qualified pathologists, and, thus, the conclusions in some of these studies were deemed questionable [2]. Furthermore, most pathology or cytotoxicity assessments in the past were dependent on relatively nonsensitive and nonspecific histopathologic methods and these traditional approaches were unlikely to detect subtle changes (subcellular changes) in cells and tissues. More sensitive or specific methods are therefore needed to provide more reliable and accountable assessments for the evaluation of QD705 toxicity.

A review further pointed out that most negative reports in the past were generated from short termed in vivo studies [2]. Indeed, our previous studies have demonstrated that QDs have an extremely long retention time in the biologic system [[6], [7]]. It also has preferential distributions in certain organs and tissues in vivo [[8], [9]]. Therefore, overly early evaluations or searching for pathologic changes in wrongly selected tissues may yield misleading results and conclusions. Our recent studies further demonstrated that kidney was the target organ for QD705 accumulation and deposition [6]. Based on these findings, kidney should be considered as one of the “target organs” for potential QD‐toxicity.

Changes in mitochondria by QDs have been suggested by several investigators in the past [[10], [11], [12]]. As early as 2005, Lovric and colleagues [13] had already projected that mitochondria were “early targets of QD‐stress and are severely damaged by QDs.” However, actual demonstrations on mitochondrial alterations and damages by QDs are still lacking. In our present study, we intend to examine subcellular, especially mitochondrial, changes in QD705‐treated renal tubular epithelial cells by electron microscopy. Information obtained in this study, we believe, not only provide the needed evidence on mitochondrial changes induced by QDs, but will also provide an important model for future assessments on the safety or toxicity of QDs.

Materials and methods

Quantum dots used in the study

The nanoparticles, QD705, used in our experiments were purchased from Invitrogen, Inc. (Hayward, CA, USA) as Qtracker 705 nontargeted quantum dots. This QD705 consisted of a Cd/Se/Te core covered with a thin ZnS shell and methoxy‐polyethylene glycol (PEG)‐5000 coating. QD705 analyses in our laboratory revealed that our QD705 consisted of 56.16 ± 8.54% Cd, 5.80 ± 3.60% Se, and 0.92 ± 0.20% Te and had a diameter of approximately 20 nm.

Animal treatments

Six‐week‐old male Imprinting Control Region (ICR) mice (BioLASCO, Taipei, Taiwan) were used in our study. All animals were acclimated for 2 weeks in the animal facilities at the National Health Research Institutes (NHRI) before treatment. All animal treatments and experimental protocol for this study were subjected to prior review and approval of the Animal Care Committee at NHRI before usage. Furthermore, animal handling was made in accordance to standard animal husbandry practices and regulations. Animals were also treated humanely with regard for alleviation of suffering throughout the study. All mice in this study were housed under a 12‐hour light/dark cycle and at 23 ± 1°C, 39–43% relative humidity. Water and food were available ad libitum. Ten mice were randomly selected for per time point of study. Each mouse was intravenously injected (via tail vein), with 40 pmol of QD705 in saline (20 μl of 2 μM solution). The injection volume was 100 μl per mouse. (The intravenous route of exposure was used to mimic potential human exposures for medical imaging applications.) Sacrifices (under pentobarbital anesthesia) were performed at 4, 12, 16 and 24 weeks after dosing. Since kidneys were found to be the “major organ” for QD705 accumulations in mice, these organs were used in our present evaluations. Kidneys were carefully removed, weighed, and prepared for studies in this investigation.

Histopathology and tissue preparations

Tissue specimens from the kidneys were fixed in buffered 10% formalin for 48 hours prior to tissue processing (dehydration and paraffin embedding). Sections were cut at 3 μm thick with rotary microtome. Hematoxylin and eosin (H&E) were used for general histology staining in accordance to standard H&E staining procedures.

Clinical biochemistry analysis

The concentrations of blood urea nitrogen (BUN) and creatinine in serum were determined for assessing renal functions. Blood was collected from heart after anesthetization. Blood samples were centrifuged at 1500 × g for 15 minutes to remove the clot. A total of 400 μl serum was used to determine BUN and creatinine concentrations with VITROS 350 Chemistry System (Ortho Clinical Diagnostics, Johnson & Johnson, Kowloon, Hong Kong) at the Veterinary Hospital, National Taiwan University, Taipei, Taiwan.

Electron microscopy findings

Renal cortex from the kidney was sampled, diced into 1 cubic meter sizes, and fixed with 1% glutadehyde for 48 hours. All tissue samples were further fixed and stained with 0.5% osmium tetroxide over night before embedding in epoxy resin. All sections were cut at 1‐μm thick and examined with a transmission electron microscope (H‐7650; Hitachi, Tokyo, Japan).

Statistical analysis

Comparison of data between groups was performed by Student t test. Differences between data were considered significant at p < 0.05.

Results

Histopathologic examinations

Tissue sections were stained with standard H&E stain and were examined with a research light microscope. No specific or significant histopathology was detected in the kidney tissues via H&E staining at any time point (4, 12, 16, and 24 weeks) of the study.

Measurements for renal function

Renal functions were assessed by measuring BUN and creatinine concentrations in serum at 12 and 16 weeks. Both concentrations were not elevated at 12 weeks and then significantly elevated at 16 weeks following injection of 40 pmol of QD705 (Table 1). These results showed that renal function, determined by clinical biochemistry methods, was reduced at 16 weeks after 40 pmol of QD705 injection.

Table 1.

Effects of QD705 on renal function.

Time (wk)	Treatment	BUN (mg/dL)	Creatinine (mg/dL)
12	Control	21.42 ± 4.70	0.26 ± 0.05
	QD705	24.08 ± 2.61	0.28 ± 0.07
16	Control	22.33 ± 2.58	0.27 ± 0.05
	QD705	27.83 ± 3.82*	0.32 ± 0.04*

*p < 0.05, as compared with control group.

BUN = blood urea nitrogen.

Electron microscopy findings

Electron microscopic appearance of a normal proximal convoluted tubule (PCT) epithelial cell is illustrated in Fig. 1A. Note the consistency in sizes and shape of the mitochondria that are aligned within the confines of the basal infoldings in the PCT epithelial cell in a well‐oriented fashion. The normal mitochondria also have a uniform matrical density with well‐developed inner cristae (1, 2). With electron microscopy, a series of morphologic alterations and pathologic changes could be identified in the renal PCT epithelial cells from QD705‐treated mice. At 4 weeks after 40 pmol of QD705 injection, many PCT epithelial cells were found to have lost their basal infoldings. Mitochondria of various sizes were also found to be scattered in the cell cytoplasm without alignment and orientation (Fig. 1B). Closer examinations revealed that many mitochondria were edematous and swollen (enlarged). These abnormal swollen mitochondria acquired rounded configuration and with “watery” matrices. With such edematous condition of the mitochondria, many inner cristae of the mitochondria became shortened or broken (Fig. 2B). In addition, another type of mitochondrial enlargement was also observed at this time of exposure (4 weeks). Unlike the swollen mitochondria, these giant or hypertrophied mitochondria could become extremely enormous in size. These giant mitochondria maintained a floccular matrical density and prominent inner cristae (Fig. 2C). Such abnormal giant mitochondria could still be found at 24 weeks after QD705 exposure (Fig. 3). At this late time, some of these hypertrophied giant mitochondria also developed hyperplastic cristae (Fig. 3A) give rise to long and complex cristae networks resembling fingerprint patterns within these mitochondria (Fig. 3B).

Figure 1.

(A) Normal renal proximal convoluted tubule (PCT) epithelial cells, mouse, normal. Note the uniformity in sizes of the mitochondria and the well orientation (alignment) of the mitochondria within the confines of the basal infoldings. (×20,000); (B) PCT epithelial cell, mouse, QD705‐treated (4 weeks). Note the great variations in mitochondrial size, absence of basal infoldings, and total disorientation of the mitochondria within the cell cytoplasm (×10,000).

Figure 2.

(A) Normal mitochondria from mouse renal PCT epithelial cell. Note the intactness of the inner cristae within the mitochondrial matrix of uniform density (×30,000); (B) swollen mitochondria from PCT epithelial cells, mouse, QD705‐treated (4 weeks). Note the roundness and enlargement of the mitochondrial shape and size. Many of these edematous mitochondria had matrices with watery appearance and broken inner cristae (×20,000); (C) giant or mega mitochondria from a PCT epithelial cell, mice, QD705‐treated (4 weeks). Note the tremendously enlarged size of the mitochondria was also observed. Despite the enlargement, there was maintenance of the mitochondrial matrix density as well as intactness of the inner cristae (×40,000).

Figure 3.

(A) APCT epithelial cell from a mouse treated with QD705 (24 weeks) showing many hypertrophied mitochondria (giant or mega mitochondria) with long and distinct inner cristae (×15,000); (B) mitochondria in a PCT epithelial cell, mouse, QD705‐treated (24 weeks). Some of the hypertrophied mitochondria also developed highly hyperplastic inner cristae. These long and complex cristae networks of inner cristae give rise to an appearance resembling finger print pattern (×20,000).

By contrast to “enlargements” of mitochondria (either swollen or hypertrophied), some PCT epithelial cells displayed large numbers of mitochondria within extremely small sizes (micro‐mitochondria). A comparison of normal, giant, and micro‐mitochondria is demonstrated in Fig. 4A. Careful examinations revealed that many mitochondria were engaged in budding activity (mitochondrial division) leading to many micro‐mitochondria in these cells (Fig. 4B). These variants in mitochondrial sizes (giant and micro‐mitochondria) could be observed in PCT cells throughout our study period (4–24 weeks). At late period (around 16–24 weeks), some PCT epithelial cells displayed cytoplasm with extremely high density of mitochondria (Fig. 5A). Such significant increase in mitochondrial number was probably the result of the increased mitochondrial budding activity in these cells at an earlier time (4–16 weeks).

Figure 4.

(A) PCT epithelial cell, QD705‐treated (4 weeks). Asides from development of giant or mega mitochondria (Mg), many extremely small mitochondria (micro‐mitochondria) (indicated by arrows) as compared with those of normal size (N) were also found (×10,000); (B) budding mitochondria in a PCT epithelial cell, QD705‐treated (4 weeks). Between 4–16 weeks, mitochondrial budding (division) was frequently found in PCT epithelial cells. Such rapid and increased budding of mitochondria (biogenesis of mitochondria) may be responsible to the many small (micro‐) mitochondria seen during this time period (×50,000).

Figure 5.

(A) PCT epithelial cells, QD705‐treated (24 weeks). These PCT cells showed very high density of mitochondria population in their cytoplasm (×6000). This hyperplastic (increase in number) condition of mitochondria was probably resulted from an increased biogenesis (budding) of mitochondria; (B) PCT epithelial cell, QD705‐treated (12 weeks). At earlier time periods (4–12 weeks), many PCT epithelial cells actually showed a reduction in mitochondrial number as compared to similar cells from control animals (×10,000).

While some PCT epithelial cells showed increases in mitochondrial size (mitochondrial hypertrophy) and number, many PCT cells actually also displayed a drastic reduction in mitochondrial numbers. This finding of reduction in mitochondrial number in the PCT cells was especially prominent at an earlier time period (4–12 weeks). This is shown in Fig. 5B. Therefore, we suggest that increases in mitochondrial size and numbers probably represent morphologic alterations of the mitochondria as a compensational attempt for the reductions in mitochondrial number and function induced by QD705. A comparative illustration of the great variants in mitochondrial morphology (size, shape, numbers, and orientations) in the kidney of QD705 ‐treated animals compared with the normal controls; this is demonstrated in Fig. 6.

Figure 6.

QD705 (12 weeks). A comparative illustration of PCT epithelial cells demonstrating great variances in mitochondrial size, shape, and number. The cell in upper left represents appearance of a normal PCT with normal mitochondria. Note the uniformity of size and shape as well as numbers and one‐directional orientation (alignment) of the mitochondria. The other two cells (upper right and lower half) displayed tremendous reduction in mitochondria number and great variances in size, shape, and orientations of the mitochondria (×6000).

Aside from such compensatory modifications and alterations of mitochondria, degenerations of mitochondria in the PCT cells of QD705‐treated animals were also observed. Mitochondrial breakdowns to form membranous bodies (myelin figures) could be detected as early as 4 weeks (Fig. 7A and B) leading to many multiple concentric membranous bodies (CMBs) in many PCT cells (Fig. 7C). By 12 weeks, isolated “cytoplasmic pockets” (CPs) containing degenerating cytoplasmic materials, including mitochondria, could be found in many PCT cells (Fig. 8A). At 16–24 weeks, multiple and dense focal cytoplasmic degenerative (FCD) areas could be detected in many PCT epithelial cells (Fig. 8B). Various time‐related mitochondrial alterations are represented and summarized in Fig. 9.

Figure 7.

Mitochondrial degeneration in PCT cells, mouse, QD705‐treated. (A) 4 weeks, break down of inner membranes of mitochondria (×30,000); (B) 12 weeks, myelin formation in cytoplasm near mitochondrial population. These may represent membranous degeneration of mitochondria (×20,000); (C) 24 weeks, large concentric membranous bodies in the cell cytoplasm could be demonstrated. These large concentric membranous bodies may represent aggregates of myelin bodies in cells (×15,000).

Figure 8.

Cytoplasmic degradation in PCT epithelial cells, mouse, QD705. At late time period (12–24 weeks), areas of cytoplasmic degradation in some PCT cells were also observed. (A) 12 weeks, isolated cytoplasmic pockets with confined degenerative cytoplasmic materials, including mitochondria, were demonstrated (×10,000); (B) 24 weeks, many focal cytoplasmic degeneration areas consisted of degraded proteins and membranous materials could be found in some PCT cells from animals treated with QD705. These focal cytoplasmic degradation areas represent generalized cytoplasmic degeneration independent or secondary to mitochondrial degeneration (×6000).

Figure 9.

Diagrammatic summary on time‐correlated mitochondrial alterations in PCT epithelial cells from mouse treated with QD705. Alteration of mitochondrial structure, numbers and degeneration by QD705.

Discussion

Although QDs have tremendous potential for various applications, their biosafety is still very controversial. Some publications claimed either safety or toxicity of this nanomaterial [[2], [3], [4], [5]]. A recent study coupling with histopathology and blood chemistry (organ functional markers) claimed that QD705 was totally “safe” with 80 days' toxicity evaluation [3]. In our present study, alterations in renal functions were not observed until at 16 weeks (112 days) in animals treated with QD705 indicating that QD705 may have a slow or delayed toxicity and may not be noticeable at early time periods. Furthermore, our own histopathology studies via H&E staining and light microscopic examination, either in the present or past [9], also failed to detect significant histopathologic damages in the kidneys of QD705‐treated animals even at late periods of the studies. These failures to detect reduction in renal functions at earlier time points and the absence of renal histopathology raised the question: Are general histopathology (H&E staining) and blood chemistry sensitive and specific enough to detect subtle and specific injuries in cells and tissues induced by QDs or other nanomaterials? The answer is obviously no. These methods are very general and nonspecific techniques in detecting more overt cell and tissue damages with significant cell degenerations and deaths (necrosis or apoptosis). Indeed, by means of more specific MitoTracker Red staining method, aggregations of swollen or enlarged mitochondria in QD705‐treated cells have been reported before [[13], [14]], suggesting that QDs might induce more subcellular changes without overt cell degenerations or death. These findings further suggested a functional loss or reduction in the mitochondrial function by QDs [[13], [15]]. These investigators further indicated that such changes occurred prior to cell deaths (apoptosis) suggesting that mitochondria may actually be early targets of QDs toxicity. In our present study, we have observed a reduction in the number of mitochondria in the PCT cells at earlier time points of QD exposure (4–12 weeks). Despite there was demonstration on reduction of mitochondrial enzymes and functions upon QD exposures [[13], [15]], actual mitochondrial changes and damages during the periods of QD705 injection still need to be demonstrated and verified. Since mitochondria are subcellular organelles that are difficult to be visualized at light microscopic level, mitochondrial alterations and damages must be demonstrated with electron microscopy. In our present study, as early as 4 weeks after exposure to QD705, swelling of mitochondria in the renal PCT epithelial cells could be readily demonstrated by electron microscopy. Alternatively, this type of mitochondrial change represents only non‐specific mitochondrial injury, which can be induced by many renal toxicants, including cadmium chloride (CdCl₂) [[16], [17], [18]]. Since cadmium is a major component of QD705, the influence by cadmium released from degraded QD705 [6] cannot be totally excluded.

While the mitochondria in normal PCT are usually fairly consistent in size and are always well oriented (positioned) within the confines of the basal infoldings of the PCT cells, many PCT epithelial cells in QD705‐treated animals showed an absence of basal infoldings and totally disoriented mitochondria. This ultrastructural finding probably represents the disruption of “tubular mitochondrial network” described by Lovric and colleagues [13] via light microscopic examination of QD‐treated Michigan Cancer Foundation (MCF)‐7 cells. However, with electron microscopy, a great variation in mitochondrial sizes could also be demonstrated in this study. Most notably, many mitochondria became extremely large (giant‐mitochondria). Unlike swollen mitochondria, these giant mitochondria maintained good matrical density and intact inner cristae. Abnormal proliferation of the inner membrane (cristae), giving rise to “finger‐print” configuration of the enlarged mitochondria was also demonstrated at later time of the study (16–24 weeks). We suggest that the enlargement of mitochondria (hypertrophied mitochondria) with hyperplastic cristae represents a compensatory response to the reduction of mitochondrial numbers or mitochondrial functions (loss of cytochrome C oxidase activity) reported by other investigators [[13], [15]]. Similar compensatory effects on defective of mitochondria have been observed in other health conditions such as in heart muscles of certain cardiomyopathy patients [19] and in skeletal muscles of patients with congenital muscular dystrophy [20]. The changes in mitochondria that we have observed in this study are striking and exciting. Because except in some chronic human diseases [[19], [20], [21], [22], [23]], no ordinary chemical is known to induce such dynamic structural changes in mitochondria. Electron microscopy is a much more sensitive tool to detect “pathology” than traditional histopathology. To our knowledge, we are the first to demonstrate actual structural and pathologic alterations and degeneration of mitochondria induced by QD705 and eluted to its health risk.

While the formation of giant mitochondria is an attempt of “compensation” for reduced mitochondrial number or function, we believe that “decompensation” of these enlarged mitochondria would eventually occur leading to the degenerative CMB found in the PCT epithelial cells. At later time point (24 weeks), multiple areas of CP of degenerations and FCD were formed as cytoplasmic attempts to confine and degrade the degenerative cellular debris. It is interesting to note that despite all these mitochondrial changes and cytoplasmic degradation, the affected PCT cells were still “alive.” This may explain that cell death or “remarkable histopathology” was not readily observed at light microscopic level, especially via non‐specific H&E staining.

Aside from “enlargement” in size, mitochondria can also compensate their functional loss by increase in number. This type of “abnormal mitochondrial proliferation” has been observed and described in certain drug treatments or under highly stressful conditions [[22], [23]]. Abnormal mitochondrial biogenesis (proliferation) is one of the prominent histological hallmarks in muscle from patients with mitochondrial encephalomyopathy [21]. This phenomenon is known as compensatory mitochondrial biogenesis which takes place as an attempt to counteract energy deficits [23]. Indeed, careful examination of the QD‐treated kidneys also revealed an increase in extremely small mitochondria (micro‐mitochondria) in many PCT cells. Frequent mitochondrial budding (division), which is usually seldom captured in normal cells, was also evident. This active mitochondrial budding denoted active mitochondrial proliferation (biogenesis), which was observed at earlier time point of the study (4–12 weeks). Overcompensation may have occurred in some cells. We believe that the high density or population of mitochondria in some PCT cells at later study time points represented such overcompensation.

In our previous study, we have demonstrated release of free cadmium from QD705 in the kidney tissues [6]. Because cadmium is highly toxic, many assumptions were made that suggested the toxicity of cadmium‐based QDs was probably induced by the cadmium released [[1], [2], [24], [25]]. If it is so, then the mitochondrial changes observed in our present study should be identical to those reported in cadmium toxicity [[16], [17], [18]]. However, aside from mitochondrial swelling, most of the mitochondrial alterations observed in our study, such as compensatory changes in size and numbers of mitochondria and decompensation of giant mitochondria to form CMBs and FCDs, were not commonly seen or described in cadmium nephropathy [[16], [17], [18], [26], [27]]. Furthermore, renal tubular necrosis and glomerular changes, which are frequently observed in cadmium nephropathy [[27], [28]], are not found in the QD‐treated animals. Thus, the QD‐toxicity observed in this study should not be equated with those of pure cadmium toxicity. Furthermore, tellurium, another metallic component in the QD705 core, is also known to be a renal toxin [[29], [30], [31]]. We cannot exclude the possibility that the specific mitochondrial changes that we have observed may represent combined effects of cadmium and tellurium or other factors.

Our previous studies have identified the kidneys to be the target organ for QD705 accumulation [[6], [8], [9]]. Now we further demonstrated that mitochondria in PCT cells are specific targets for QD705 toxicity. It should be pointed out that mitochondria are actually favorite targets for toxicity of many metals, such as Hg, Cd, Cu, Fe, As, Al, and Pb [[16], [17], [18], [32], [33], [34], [35], [36]], which are known to cause mitochondrial dysfunction, oxidative stress, and degeneration. Furthermore, mitochondrial permeability transition, a pore‐mediated impairment of mitochondrial function, has been indicated to participate in cytotoxic action of metal ions [37]. Thus, we propose that mitochondrial markers should be developed as specific parameters for sensitive toxicity evaluation, not just for QDs, but also for many other metal‐based nanomaterials. It must be pointed out that cell death alone, either in vitro or in vivo, is not a good or sensitive parameter to define toxicity. The most worrisome toxic endpoint, in fact, should be subtle cell alterations (including cancer) without cell death (apoptosis). Thus, traditional histopathology or other general cytotoxicity methods, which may be useful in detecting overt cell deaths or tissue damages, are too insensitive and nonspecific for the detection of early and subtle cellular changes induced by some nanomaterials. Nanotechnology is one of the most revolutionary developments in the 21st century, influencing almost every aspect of human life. Sensitive and innovative approaches for biosafety assessment are needed to meet the new challenges of the 21st century. Hopefully our study will inspire developments of new and specific parameters, such as mitochondrial markers, for future biosafety evaluations in the future.

Acknowledgment

This work was supported by grant NM‐100‐PP08 from the Center for Nanomedicine Research, National Health Research Institutes, Taiwan.