X-ray microanalysis of chlorine and phosphorus content in biguanide-treated Acanthamoeba castellanii


Professor A.D. Russell, Welsh School of Pharmacy, Cardiff University, Cardiff CF1 3XF, UK (e-mail: russellD2@cardiff.ac.uk).


Energy dispersive analysis of X-rays (EDAX) was used to study the effects of chlorhexidine diacetate (CHA) and polyhexamethylene biguanide (PHMB) on Acanthamoeba castellanii. A high variation of elements occurred in untreated individual cells and only two elements, Cl (a biocide marker) and P, were investigated. X-ray dot mapping of untreated trophozoites and cysts revealed that Cl in cells was uniformly distributed throughout the cytoplasm, whereas P was less dense in the vacuoles. X-ray dots of Cl in biocide-treated trophozoites and cysts appeared denser and evenly distributed within the cells as the biguanide concentration increased. Quantitative analysis of either CHA or PHMB within the cells using Cl as an elemental marker was unsatisfactory because of the high Cl levels in untreated cells. The apparent increases of P in some experiments with treated cells might be associated with reduced permeability, protein coagulation or aggregation of phospholipids.

Energy dispersive analysis of X-rays (EDAX) is a method for elemental analysis at the ultrastructural level. A major advantage of EDAX is its ability to correlate morphological appearance with chemical composition ( Hayat 1980). The principle of X-ray microanalysis is based on the fact that, when the electrons from external sources strike the atoms in the material, energy in the form of an X-ray photon is emitted, thus giving characteristic X-rays of the element ( Russ 1978; Hall & Gupta 1982; Murr 1982; Hobbs et al. 1986 ; Hiom 1993).

In morphological studies, chemical fixation is normally used to preserve the cell structure. However, during chemical fixation and dehydration procedures, diffusion and extraction of various cell constituents, including ions, normally occur ( Hayat 1989).

To minimize the loss of diffusible ions, cryofixation is an alternative method used to prepare a biological sample for quantitative X-ray microanalysis. Freeze substitution is one of several fast-freezing methods available. This involves quenching the cells in a coolant such as liquid nitrogen; the frozen specimens are then embedded in a low temperature resin ( Hayat 1989).

A cryoprotectant is required for the rapid-freezing step to prevent ice crystal formation. The cryoprotectant used in this method is polyvinyl pyrrolidone (PVP). It is less toxic and does not interfere with cell function but delays extracellular ice nucleation, allowing supercooling of the cytoplasm ( Mackenzie 1977).

Recently, EDAX has been used in this laboratory to locate and quantify chlorhexidine diacetate (CHA) within treated Saccharomyces cerevisiae cells ( Hiom et al. 1995 ) and Pseudomonas aeruginosa bacteriophage F116 ( Maillard et al. 1995 ).

The aim of this study was to investigate the chlorine- and phosphorus-containing compounds within biguanide-treated cells of Acanthamoeba castellanii and to attempt to relate the findings to those obtained previously ( Khunkitti et al. 1996 , 1997a, b).

Materials and methods

Sample preparation

Trophozoites and cysts of the test organism, Acanthamoeba castellanii (Neff strain), prepared as described previously ( Khunkitti et al. 1996 , 1997a, b) were washed twice with saline solution and adjusted to give suspensions of 2 and 4 mg dry wt ml−1, respectively. Equal volumes (5·0 ml) of cell suspension and test biocide solution at 20 °C were mixed to give the final required concentration and held at 20 °C for 1 h. The suspension was immediately washed twice with 25% (w/v) polyvinyl pyrrolidone and the treated cells were harvested at 500 g for 10 min

Test biocides consisted of chlorhexidine diacetate (CHA) and polyhexamethylene biguanide (PHMB). Details of the procedure by which A. castellanii was exposed to these agents were described previously ( Khunkitti et al. 1996 , 1997a, b).

Specimen preparation

The cell pellet was quenched in liquid nitrogen and transferred to Lowicryl HM20 embedding medium at – 20 °C for 2 d. After resin infiltration, the cells were embedded in fresh Lowicryl HM20 at 4 °C under u.v. irradiation for 24 h until the resin became polymerized. The samples were then transferred to a u.v. light chamber at 22 °C for another 24 h before sectioning. Sample blocks were cut into 500 nm thick sections on a Reichert ultracut OMU4 microtome (Reichert-Jung, Austria). The sections were collected on carbon-coated copper grids.

X-ray microanalysis

The sections of the sample, mounted on carbon-coated copper grids, were held flat on a low background specimen holder and analysed using an energy dispersive analyser of X-rays (EDAX 9100/60) operated in conjunction with a Phillips (Eindhoven, The Netherlands) 400T electron microscope. The electron beam was focused on to the whole area of an individual cell. The following instrumental conditions were maintained constant throughout the analysis: accelerating voltage (80 kV); beam current (10 mA); sample tilt (12° C); beam spot size (100 nm). X-ray spectra obtained during 50 s intervals were either stored in a computer floppy disk for calculation, or photocopied via the video hard-copy printer (Tektronix 2200; Beaverton, OR, USA). Experiments were carried out in quintuplicate.

Quantitative analysis

Computerized analysis of the samples was performed using a computer thin section software, EDAX version 2·2, which allows the use of Hall’s Model for quantification and X-ray dot mapping ( Maegdlin & Zutkis 1982). The characteristic X-ray peak emission at the Kα line of the elements (Cl and P) was detected at 2·62 and 2·01 keV, respectively. The intensity of the corresponding radiation was determined by the window analysis method. The background X-ray emission was selected between 2·76 and 2·92 keV and its corresponding radiation intensity recorded for quantitative analysis. The relative concentrations of the element concerned were determined using the Mass Fraction Method of Hall described by Maegdlin & Zutkis (1982). The standards used for quantification included NaCl and K2HPO4, and were prepared according to the method described by Sobota et al. (1984) .

X-ray dot mapping

The distribution of elements (Cl and P) in the cells was studied by X-ray dot mapping using the window analysis programme of the EDAX thin section software, version 2·2. The X-ray dot maps were obtained via the scanning transmission STEM attachment (Phillips EM 400T). The characteristic electron dots of the elements were recorded on Ilford 120 roll film using a Wistar camera ( Hiom et al. 1995 ; Maillard et al. 1995 ).

Results and discussion

Localization of elements within A. castellanii

Typical X-ray spectra from quantitative X-ray microanalysis of untreated A. castellanii trophozoites and cysts showed that cells contained elements such as phosphorus (P), sulphur (S), chlorine (Cl) and calcium (Ca). As X-ray microanalysis was carried out by measuring elements in individual cells, a high variation in quantity of each element was observed. This was also reported by Sobota et al. (1984) . Due to the high variation of the elements in individual cells, only two major elements, Cl, a biocide marker, and P, the apparent element peak, were investigated.

X-ray dot mapping of untreated trophozoites ( Fig. 1a) and untreated cysts ( Fig. 2a) revealed that Cl in the cells was uniformly distributed throughout the cytoplasm whereas P was less dense in the vacuoles. Figure 1(b-d) and 2 (b-d) illustrate the X-ray dot mapping of Cl and P within the treated trophozoites and cysts, respectively. X-ray dots of Cl in biocide-treated trophozoites and cysts appeared denser and evenly distributed within the cells as concentrations increased. Similar findings were also found in CHA-treated yeast cells ( Hiom et al. 1995 ). This suggests that interaction of the biguanide with the cells may have non-specific target sites. The distribution of P in treated trophozoites was less dense than in vacuoles. This was not as apparent in treated cysts. A fluctuation of P in treated cells was observed.

Figure 1.

Electron micrographs of PHMB-treated trophozoites showing localization of chlorine (Cl) and phosphorus (P), respectively. Arrow indicates vacuole (V). Concentrations (μg ml−1): (a) untreated trophozoite (bar: 1 μm, × 6000); (b) 5 (bar: 5 μm, × 10 000); (c) (bar: 5 μm, × 16 000); (d) 500 (bar: 5 μm, × 8600)

Figure 2.

Electron micrographs of PHMB-treated cysts showing localization of chlorine (Cl) and phosphorus (P), respectively. Concentrations (μg ml−1): (a) untreated cyst (bar: 1 μm, × 6000); (b) 10 (bar: 5 μm, × 9000); (c) 200 (bar: 5 μm, ×9000); (d) 500 (bar: 5 μm, × 7700)

The levels of Cl and P in untreated cysts were greater than in untreated trophozoites ( Table 1). This could be due to the presence of MgCl2 in PGY medium causing physiological changes during encystment ( Bowen et al. 1969 ; Bower & Korn 1969). Sobota et al. (1984) demonstrated that trophozoites grown in medium 10 times higher in ionic content than the normal growth medium contain higher levels of elements. The content of P and Cl in trophozoites was comparatively high in comparison with the study of Sobota et al. (1984) . This could be due to difference in growth medium, age of culture and method of specimen preparation.

Table 1.  Chlorine- and phosphorus-containing compounds in trophozoites after treatment with biguanides for 1 h at 20 °C (n = 5)
BiocideConcentration (μg ml−1) Cl ± s. d. (mmol l−1 kg−1 dry wt) P ± s. d. (mmol l−1 kg−1 dry wt)
  1. s. d., standard deviation.

CHA0191·5 ± 44·6 556·9 ± 79·5
5191·4 ± 88·2 763·7 ± 283·9
10420·5 ± 213·4 932·6 ± 194·8
75446·9 ± 169·4 567·8 ± 320·6
500477·3 ± 162·9 609·6 ± 399·0
PHMB0191·5 ± 44·6 556·9 ± 79·5
5212·7 ± 92·41392·3 ± 168·0
10294·9 ± 107·5 735·4 ± 213·4
200465·4 ± 196·41311·0 ± 250·9
500805·6 ± 151·7 818·1 ± 239·4

Uptake of biguanides

The levels of Cl in biguanide-treated trophozoites and cysts increased as concentration increased. The amount of Cl in PHMB-treated trophozoites was relatively greater than in CHA-treated cells. By contrast, the level of Cl was greater in CHA-treated cysts than in PHMB-treated cysts.

In biguanide uptake studies ( Khunkitti et al. 1997a ), the amount of biguanide uptake by cysts was less than by trophozoites. Those results did not agree with this study, possibly due to measuring different parameters. In the present study, EDAX determined the Cl level in cells, whereas biocides remaining in the supernatant fluid formed the basis of the uptake investigation.

In this investigation, Cl in the cells comes from the cell itself and from biguanide molecules. In general, Cl (a basic cellular anion) exists in the form of electrolytes such as potassium chloride and sodium chloride in the intracellular constituents. These elements are often highly diffusible ( Morgan et al. 1978 ). The Cl content of untreated trophozoites (191·5 mmol l−1 kg−1 dry wt) and cysts (386·2 mmol l−1 kg−1 dry wt) is relatively high. Therefore, the increase in Cl after treatment with biguanides indicates biocide uptake but may not represent the amount of biocide within the cells.

As demonstrated in Table 1, the levels of Cl in CHA- and PHMB-treated cysts were higher than in trophozoites ( Khunkitti et al. 1996 ). The amount of pentose leakage from biguanide-treated cysts was much less than that from trophozoites, and cysts contain an extremely dense cytoplasm ( Bower & Korn 1969). As a result, the treated cysts may retain the diffusible Cl and Cl from biguanides which interact with the cellular substances. In addition, the number of Cl atoms in CHA and PHMB may also affect the levels of Cl in treated cells.

Phosphorus in biguanide-treated cells

The X-ray microanalysis spectra of CHA-treated cells were examined (data not shown). At low concentration (5 μg ml−1), the levels of P in trophozoites treated with CHA increased by a factor of 1·4 and with PHMB, by the factor of 2·5 of P in untreated cells ( Table 1). The maximal leakage of P occurred at 75 μg ml−1 CHA and increased thereafter. PHMB-treated cells also showed a similar pattern, with the maximal P leakage at 10 μg ml−1.

Cysts treated with CHA had a higher P content than untreated cysts by a factor of 1·25, whereas PHMB-treated cysts contained a greater amount of P by a factor of 1·12 ( Table 2). A similar leakage pattern of P to trophozoites was observed. P levels were minimal at 100 μg ml−1 CHA and 200 μg ml−1 PHMB, respectively.

Table 2.  Chlorine- and phosphorus-containing compounds in cysts after treatment with biguanides for 1 h at 20 °C (n = 5)
BiocideConcentration (μg ml−1) Cl ± s. d. (mmol l−1 kg−1 dry wt) P ± s. d. (mmol l−1 kg−1 dry wt)
  1. s. d., standard deviation.

CHA0 386·2 ± 51·0752·7 ± 126·4
10 384·1 ± 98·4941·0 ± 274·4
100 973·8 ± 128·8455·6 ± 79·4
3001388·5 ± 649·1534·0 ± 228·1
5001576·5 ± 782·1834·9 ± 321·0
PHMB0 386·2 ± 51·0752·7 ± 126·4
10 419·5 ± 77·4845·0 ± 279·8
100 548·6 ± 33·2504·8 ± 190·1
200 712·4 ± 45·3360·5 ± 50·7
500 992·8 ± 136·9406·8 ± 111·3

Rye & Wiseman (1964) demonstrated that 32P release from bacteria increased as CHA concentrations increased and became constant at higher concentrations. However, with EDAX studies, the pattern of P within the biguanide-treated cells appeared higher at low concentrations in trophozoites and cysts and the diphasic leakage occurred as concentration increased.

Biocide-induced P leakage can be assessed by detecting the presence of inorganic P in the supernatant fluid ( Denyer & Hugo 1991) whereas the amount of P within the cells using EDAX represents all P-containing compounds. In general, P occurs in the cells as the element associated with lipids, proteins and polysaccharides and is thus considered as a relatively non-exchangeable element ( Morgan et al. 1978 ) and as one of the anions to balance all cations in Acanthamoeba ( Sobota et al. 1984 ). Therefore, the initial P release from treated cells could be inorganic P followed by P-containing macromolecules.

The increased levels of P in treated cysts and trophozoites are puzzling. Different sections of the same samples were analysed to ensure that no systematic errors took place during the measurement; the results were all the same. Levels of P were much higher than those of Cl in all the untreated cells. High density of P electron dots was often confined to cell walls, phospholipids and nuclei.

Possible, albeit currently speculative, reasons for the increased P levels in treated cells are as follows. (i) Reduced membrane permeability; high concentrations of CHA cause clustering small vesicles along the plasma membrane ( Khunkitti et al. 1998 ). These may act as a physical barrier in limiting the P release. (ii) Coagulation of proteins; electron dense precipitates were found within the cell walls of PHMBtreated cells ( Khunkitti et al. 1998 ). These materials were not seen in the untreated samples. This observation suggests that the precipitates were formed as a result of the reaction of cell wall proteins to biguanide treatment. Heavy deposition of these materials in the plasma membrane may function as a major barrier in reducing the leakage of P elements. (iii) Aggregation of phospholipids; it is possible that CHA and PHMB alter the structural arrangement and nature of phospholipids. Crystallization of these lipids may take place in the presence of biguanides.


EDAX may be used to study ion localization qualitatively and quantitatively within the treated cells. High variations of elements in treated cells indicate the response of individual cells to the biocides. Furthermore, quantitative analysis of either CHA of PHMB within the cells using Cl as an elemental marker does not apply for Acanthamoeba because of the high Cl levels in untreated cells. Adaptability of Acanthamoeba by producing P-containing compounds within the cells may possibly be one of the defence mechanisms of Acanthamoeba to biocides.