Electron and light microscopy studies on particulate EPR spin probes lithium phthalocyanine, fusinite and synthetic chars

Electron and light microscopy studies onparticulate EPR spin probes lithium phthalocyanine,fusinite and synthetic chars

S. W. NORBY,* H. M. SWARTZ† & R. B. CLARKSON‡*College of Medicine, and the Department of Veterinary Clinical Medicine, University of Illinois,Urbana, Illinois, U.S.A.†Dartmouth Medical School, Department of Radiology (H.M.S.), Hanover, New Hampshire, U.S.A.‡College of Medicine, the Department of Veterinary Clinical Medicine, and the Illinois EPRResearch Center, University of Illinois, Urbana, Illinois, U.S.A.

Key words. EPR spin probes, fusinite, lithium phthalocyanine, microscopy,synthetic chars.

Summary

This is the first detailed study on the morphology of thewidely used particulate electron paramagnetic resonance(EPR) probes lithium phthalocyanine (LiPc), methyl-LiPc,methoxy-LiPc, fusinite and synthetic carbon-based chars, bymeans of both light and electronmicroscopy (LM and EM).The importance of these EPR probes for the measurement ofO2/NO has been reported previously. Under LM, LiPc with itsdistinct crystalline structure differs significantly from thenoncrystalline black flakes of both LiPc derivatives, methyl-LiPc and methoxy-LiPc. Unlike the shiny carbon-basedsynthetic chars, which have no characteristic morphology,fusinite, a fraction of fossilized coal maceral, displaysdistinctive fine, parallel channels. SEM studies reveal astriated surface and interlocking multilayered structure ofLiPc that is markedly different from either the multilayeredstacked methyl-LiPc or botryoidal methoxy-LiPc. Theregularly spaced pores and channels of fusinite, a reflectionof its plant origin, contrast sharply with the randomlydistributed pores of all sizes of the various synthetic chars.Furthermore, the combined results of both LM and EMstudies strongly suggest the nonperturbing nature of fusiniteand LiPc when they were used as EPR probes to measureoxygen in cells or tissues. We hope that this study, inconjunction with the numerous EPR functional studies ofprobes reported earlier, provides valuable information for thefuture development of new particulate EPR spin probes.

Introduction

The application of the electron paramagnetic resonance(EPR) technique to measure [O2] has been well documented

(Hyde & Subczynski, 1989). Basically, EPR spectra of manyparamagnetic substances are affected by the presence ofmolecular oxygen because it is paramagnetic. The effects ofoxygen on EPR spectra are proportional to [O2], andtherefore can be used for quantitative studies wherever EPRspectra can be generated. The parameter that is often usedfor oxygen measurements in the EPR spectrum is the peak-to-peak linewidth (DBpp). DBpp increases reproducibly andreversibly when a specific spin probe is exposed to increased[O2]. The underlying physical mechanism responsible forthis phenomenon is exchange broadening, and is dueprimarily to the magnetic exchange coupling of the radicalspins with the triplet spin of O2 (Turek et al., 1989b).Because the effects of oxygen on EPR spectra are due tophysical interactions at the molecular level, oxygen is notconsumed in the process. The paramagnetic materials usedin this process are called spin probes or spin labels. Therecent developments of a variety of particulate EPR spinprobes, such as lithium phthalocyanine (LiPc), fusinite andsynthetic chars, have generated numerous EPR studies onthe in vitro and in vivo measurements of extra- andintracellular oxygen (O2) as well as nitric oxide (NO) inbiological systems (Swartz et al., 1991, 1992; Duret et al.,1992; Smirnov et al., 1993a,b, 1994; Vahidi et al., 1994;Clarkson et al., 1995). The importance of measuring theconcentration of oxygen has been especially emphasized inthese studies. The chemical and physical properties of thesespin probes have been characterized and well documentedin the literature (Sing, 1955; Sing et al., 1985; Turek &Andre, 1987a,b; Belarbi et al., 1989; Turek et al., 1989a;Boyer & Clarkson, 1994; Byrne & Marsh, 1995). However,the morphologies of spin probes have not been studiedextensively. In view of the correlation between structure

Journal of Microscopy, Vol. 192, Pt 2, November 1998, pp. 172–185.Received 16 March 1997; accepted 16 June 1998

172

q 1998 The Royal Microscopical Society

Correspondence to: Shong Wan Norby. Tel: 1 217 333 8716; fax: 1 217 333

8868; e-mail: [email protected]

and function, it is important to study these spin probes,whenever possible, with special reference to their surfacetexture, total surface area and the number and sizes of poresper unit mass. In this study, fusinite is the only spin probethat has not been modified chemically in the laboratory toinduce changes in its physical structure. For the other spinprobes, once we understand the relationship betweenstructure (chemical and physical) and function, steps maybe taken to alter the structure chemically and/or physicallyif necessary. By the proper application of this information,the design and synthesis of new and improved particulatespin probes that have increased sensitivity to pO2 and/orpNO may be feasible.

Clinical application of these spin probes is our final goal;therefore, it is absolutely vital that we make spin probeswith the following characteristics: nontoxic (with no/minimal interference with the metabolism and viability ofthe cells/tissues); small dimension, which facilitates injec-tion and subsequent delivery to specific target site(s) via thecirculatory system; chemically stable for long-term mon-itoring of the [O2] or [NO]; sufficiently inert so that they willnot elicit immune responses detrimental to health; and alsononcarcinogenic. We studied the particulate EPR spinprobes LiPc and derivatives, methyl LiPc and methoxyLiPc, fusinite and several synthetic chars with LM, trans-mission electron microscopy (TEM) and scanning electronmicroscopy (SEM). Furthermore, we conducted microscopystudies on both cultured cells that have endocytosed andtissues that have been injected with one of the spin probes.The results from this morphological study, in conjunctionwith previous documented functional studies using some ofthe same particulate EPR oxygen and nitric oxide spinprobes, help us better understand the relationship betweenstructure and function. We believe that the analysis of ourdata will provide the foundation for designing/synthesizingnew and more sensitive particulate EPR spin probes that areespecially suitable for clinical application.

Materials and methods

LiPc, methyl- and methoxy-LiPc were generously providedby Dr M. Moussavi (LETI, Grenoble, France). Fusinite fromsouthern Illinois coal (Coal Characterization Facility, South-ern Illinois University at Carbondale) was ground, acidwashed, rinsed and graded according to size. Fusinite of lessthan 10 mm was used for biological studies. Synthetic chars,products of charring sucrose with activating agents (ZnCl2

or KOH) under an inert nitrogen atmosphere and at hightemperature (620 8C), were kindly provided by Steve Boyer(University of Berkeley, California, U.S.A.). SEM (ISI-DS130microscopy at 7–10 kV) was done on LiPc, and derivativesmethoxy-LiPc [LiPc(CH3O)8], methyl-LiPc, fusinite andsynthetic carbon-based chars. Samples for SEM studieswere air dried prior to mounting on specimen stubs with

Araldite resin. The samples (with the exception of fusiniteand chars) were coated in an SPI sputter coater with gold/palladium and stored in a desiccator. For the SEM study oftissue injected with spin probe, the tissue was first fixed with2% glutaraldehyde in phosphate buffer as well as dehy-drated with an ascending series of concentrations of ethanolprior to critical point drying and the completion of theremaining preparative steps. All scanning images wererecorded on Polaroid PN-55 film. Cultured Chinese hamsterovary (CHO) cells and mouse macrophages (RAW cells)were prepared for TEM study as follows: samples of culturedCHO and RAW cells (fusinite particles of <5 mm were addedto the cultures) with endocytosed fusinite particles wereharvested, washed (3 ×) free of serum and extracellularfusinite, pelleted by centrifugation and embedded in 2%agar. The solidified agar blocks of cells were trimmed to1-mm3 blocks. To continue processing the agar blocks ofcells, the standard TEM protocol for tissue sample prepara-tion was followed. It included fixation with 4% glutaralde-hyde in 0·1 M phosphate buffer, washing (3 ×) withphosphate buffer; post-fixation with 1% osmium tetroxidein 0·1 M phosphate buffer, rinsing with phosphate buffer;dehydration with ascending series of ethanol ¹ 25%, 50%,75%, 95% (1 × each) and 100% (3 ×), 5 min each; continueddehydration with 100% propylene oxide (for 15 min);infiltration with 1:1 (propylene oxide: Epon 812) for 1 h,and 1:3 (propylene oxide: Epon 812) overnight at roomtemperature. The samples were embedded in pure embed-ding mixture (Epon 812) in the final castings, and allowedto polymerize at 60 8C for 72 h. The cured cell blocks weretrimmed, sectioned, stained with both lead citrate anduranyl acetate, and studied with a Hitachi H-300 transmis-sion electron microscope at an operating voltage of 100 kVand images recorded with Kodak electron microscope film4489 (3 1/400 × 400). LM studies were carried out on fusiniteand LiPc that had been separately injected into thegastrocnemius muscle of mice. Tissue blocks (< 5 mm3 insize) containing the specific fusinite or LiPc spin probe wereremoved from the experimental animals. Tissue blocks alsowere removed from identical sites from normal controlanimals. All tissue blocks were fixed in Bouins. The JB-4embedding kit (Polysciences, Inc., PA, U.S.A.) was used forsample preparation, and a standard protocol of LM samplepreparation was followed. Cured tissue blocks weretrimmed, and sections (about 8 mm thick) were cut with amicrotome (Reichert OMU2). Tissue sections were mountedon slides and stained with haematoxylin. Finally, stainedsections were mounted with Permount (Fisher Scientific, PA,U.S.A.), protected with a coverglass, studied with the lightmicroscope (Nikon DIAPHOT 300) and recorded on KodakT-Max 100 film. Preliminary in vivo oximetry studies of thetissue injected with individual fusinite or LiPc spin probewere carried out with an L-band EPR spectrometer prior tothe LM study.

q 1998 The Royal Microscopical Society, Journal of Microscopy, 192, 172–185

LM AND EM OF EPR SPIN PROBES 173

Fig. 1. Scanning electron micrographs of (a) lateral view of single LiPc crystals (top crystal is intact with typical cylindrical crystal structureand the bottom crystal is broken at places showing the structure beneath the relatively smooth surface); (b) the lower crystal shown in (a) ata higher magnification, revealing the porous ultrastructure; (c) the porous surface shown in (b) at a higher magnification, hence the shapeand size of the pores could be easily observed; (d) end view of a single crystal which has been truncated at one end with a razor blade; (e)compound LiPc crystal and (f) portion of the compound crystal shown in (e) enlarged, showing the rough and striated surface. Scalebars ¼ 100 mm, 10 mm, 1 mm, 50 mm, 100 mm and 10 mm in a, b, c, d, e and f, respectively.

174 S. W. NORBY ET AL .


Results and discussion

SEM has proven to be a useful tool for studying themorphology of these particulate EPR spin probes. LiPc existseither as a single (Figs. 1a, d, 2a) or compound (Fig. 1e)crystal. At moderately high magnifications, both the porous(2700 ×, Fig. 1b) and striated nature of the surface (1430 ×,Fig. 1f) of LiPc are visible. At higher magnification(10 000 ×) the pores (Fig. 1c) are noticeably irregular inshape, and also vary in size. Large pores measure as muchas 4 mm long and 1 mm across, while small pores are about0·1–0·2 mm in diameter. The lamellar composition of someof the LiPc crystals are clearly shown in Fig. 2(a,c). Thesame samples when viewed at higher magnificationrevealed a moderately rough surface texture, with smallpits and interlocking layers (Fig. 2b,d).

Unlike LiPc, methyl-LiPc and methoxy-LiPc appear asblack flakes without any well-defined crystal structurewhen examined under the light microscope. SEM studiesshow that the morphology of methyl-LiPc (Fig. 3a–d)differs from that of methoxy-LiPc (Fig. 3e,f), and the

morphology of both differs significantly from that of LiPc(Figs. 1 and 2). Micrographs of methyl-LiPc taken at lowmagnification (50 ×) illustrate the very rough but rela-tively uniform surface texture (Fig. 3a). At highermagnifications of 1500 × (Fig. 3c) 3000 × (Fig. 3d) and5000 × (Fig. 3b), the irregular multilayered ultrastructureis very apparent. Unlike the lamellar structure seen in LiPc,the layers of methyl-LiPc are not structurally aligned inone specific direction, thus giving it a random multilayeredappearance (Fig. 3b,c). The rough and pitted surfaces areshown in Fig. 3(c,d). In comparison to methyl-LiPc,methoxy-LiPc has a definite botryoidal structure (Fig. 3e)and a slightly rough surface texture at low (100 ×)magnification. At a magnification of 2100 × (Fig. 3f), thedistinct surface texture consists of moderately rounded,subequal grains that cluster and/or fuse together in ageneral botryoidal pattern. Both methyl- and methoxy-LiPc (Fig. 3) show a significant increase in their surfaceareas when compared to the surface area of the sameamount of LiPc (Figs. 1 and 2).

It has been postulated (Boyer & Clarkson, 1994) that the


Fig. 2. Scanning electron micrographs of (a) a single and a compound LiPc crystal; (b) the compound crystal shown in (a) enlarged, showingoverlapping layers; (c) another compound LiPc crystal at higher magnification; and (d) an enlargement of (c) showing the interlocking layersat places. Scale bars ¼ 50 mm, 5 mm, 50 mm and 5 mm in a, b, c and d, respectively.


surface area of a solid spin probe may play an importantrole in (1) the response time and (2) the range of sensitivityof the solid spin probe in terms of oxygen-dependent peak-to-peak linewidth change (DBpp) when exposed to various

[O2]. Our data (Fig. 4a,b) suggest that the greater thesurface area of the solid spin probe, the shorter the responsetime, and also the greater the sensitivity to [O2]. Althoughthe most satisfactory range of sensitivity to gaseous oxygen

Fig. 3. Scanning electron micrographs of methyl-LiPc at increasing magnification showing the rough texture in (a) and much increased sur-face in (b), and (c), due to random multiple stacking, and the distribution of pits in (d) over the surface. Micrographs of methoxy-LiPc showthe characteristic botryoidal morphology at both low and high magnification in e and f, respectively. This kind of structural arrangement atboth macro and micro levels serves to increase the total surface area of the sample significantly. Scale bars ¼ 500 mm, 5 mm, 10 mm, 10 mm,100 mm and 10 mm in a, b, c, d, e and f, respectively.



was obtained with methoxy-LiPc (Fig. 4b), especially in thephysiological range of oxygen, LiPc (Fig. 4a) remains theprime choice as a solid EPR spin probe for oxygenmeasurements. The reason is that the oxygen sensitivityrange of LiPc dropped less significantly than that ofmethoxy-LiPc when both were suspended in aqueousmedia. At room temperature, both LiPc and methoxy-LiPcexhibit a single-line EPR spectrum. In comparison, the EPRspectrum of methyl-LiPc consists of two superimposed lines(i.e. a broad and a narrow component). In spite of the muchincreased surface area in methyl-LiPc, we found that thechemically added methyl groups resulted in a significantreduction in its sensitivity to change in pO2. This reductionin its oxygen response coupled with the two-componentspectrum renders it the least satisfactory of the three lithiumspin probes described thus far.

A survey of the literature indicates that numerousoximetry and/or nitric oxide studies have been done withLiPc (Swartz et al., 1991, 1992; Liu et al., 1993; Smirnovet al., 1993a,b, 1994); however, none has been reported foreither methyl- or methoxy-LiPc. When LiPc was used forour in vivo and in vitro oximetry studies (Swartz et al., 1991,1992; Liu et al., 1993; Smirnov et al., 1993a,b, 1994), areduction in sensitivity to oxygen was often observed.Apparently, the surface characteristics of all these spinprobes were adversely effected in the presence of water orother solutions, sometimes affecting the quality of these EPRspin probes not only in their increased response time tochanges in pO2/pNO, but also in a decrease in EPR spectralsensitivity (see Fig. 4c).

We have successfully introduced individual LiPc crystalsinto the gastrocnemius muscles of a group of mice tomonitor [O2] in the muscle tissue over a 2-month period.At the end of the experiment, we proceeded to process themuscle with the spin probes for both LM and SEM studies.From the LM study, the LiPc crystal in the gastrocnemiusmuscle has a tendency to shatter during the sectioningprocess (Fig. 5c). From an examination of numerousconsecutive haematoxylin-stained tissue sections, we con-clude that the presence of LiPc, although foreign to theanimal, has elicited only a minor immune response asdemonstrated by the recruitment of some leucocytes to thesite where the LiPc crystal (Fig. 5a,b,d) is located. If severeinflammation of the tissues had occurred, more drasticsymptomatic histological changes, including the encapsu-lation of the crystal, would have been observed. Prior tokilling the animal for the LM histological study, anexamination of each animal showed no visible signs ofintense redness or gross tumour formation at the regionwhere the crystal was introduced. Furthermore, none ofthe mice injected with LiPc crystals died during the courseof the experiment. This may serve as an indication of thenontoxic and inert nature of the LiPc crystals. Thesecharacteristics are vital requisites in the design and


Fig. 4. (a) Calibration of oxygen-dependent peak-to-peak linewidth(DBpp) of (a) LiPc and (b) methoxy-LiPc as a function of [O2],respectively; (c) calibration of DBpp of fusinite (<5 mm) in wateras a function of [O2]; insets for each calibration show response ofthe same spin probe at the range of physiological [O2].


Fig. 5. Light micrographs of histological sections of mouse gastrocnemius muscle injected with LiPc (a, b, c, d). Vacuolation at the site of eachLiPc crystal observed in all micrographs is due to the shattering of the spin probes during the sectioning process; it is especially evident in (c)where large numbers of shattered fragments are seen lying on top of the characteristic striated skeletal muscle cells. Minimal immuneresponse to the presence of LiPc is clearly evident in (b) and (d) where only a few small lymphocytes are seen to have been recruited tothe site of the crystal. Scale bar ¼ 100 mm.

Fig. 6. Scanning electron micrographs of mouse gastrocnemius muscle injected with LiPc (a) at low magnification and (b) at high magnifica-tion. The absence of tumour formation or encapsulation of the crystal is especially noted in (b) where the outline of the crystal is highlyvisible. Scale bars ¼ 50 mm and 10 mm in a and b, respectively.



synthesis of future particulate spin probes suitable forclinical application. The SEM studies (Fig. 6a,b) alsoindicate a lack of major immune response at the site ofthe LiPc crystal. In addition, the absence of tumour was

evident, when the outline of the LiPc crystal injected ineach of the gastrocnemius samples is clearly visible athigher magnification (Fig. 6b).

Fusinite, a coal maceral derived from pyrolysed woody


Fig. 7. Scanning electron micrographs of cross-sections of fossilized carbon, fusinite, in various sizes (a, b, c and f) and at various magnifica-tions where prominent, regularly spaced large and small pores resembling xylem and phloem of modern-day plants are observed. Longitu-dinal (d) and oblique (e) sections of fusinite display numerous long, parallel, closely spaced channels that provide ample surface for gasinteractions. Scale bars ¼ 50 mm, 10 mm, 50 mm, 100 mm, 100 mm and 20 mm in a, b, c, d, e and f, respectively.


Fig. 8. LM histological studies. Longitudinal sections (a, b, c) and cross-sections (d, e) of mouse gastrocnemius muscles injected with fusinite.Various cell types present could be easily identified in the haematoxylin-stained sections. Skeletal muscle cells are readily recognized by theircomparatively large dimension and their distinctive striations and multiple peripheral nuclei. Recruited and interspersed among the promi-nent opaque black specks of fusinite are fair numbers of polygonal-shaped small lymphocytes with their typical thin cytoplasm and darklystained nuclei; other cells recruited to the same area include the comparatively large macrophage and neutrophils – both can be seen tocontain phagocytic fusinite particles if examined very closely. Adjacent to the muscle and fusinite particles is the distinctive connective tissuesconsisting of collagen fibres with numerous spindle-shaped fibroblasts. Disc-shaped red blood cells within small blood vessels are also visible.Normal control sample of longitudinal section of gastrocnemius muscle is shown in (f). Scale bar represents 100 mm.




Fig. 9. Transmission electron micrographs of two adjacent RAW cells with endocytosed fusinite. (a) A very large fusinite (labelled f) and a fewsmall fusinite particles are found in the cytoplasm of one and only a few small fusinite particles in the other. All fusinite particles have astriated appearance. The cell membrane remains intact, as does the nucleus (labelled n) with its prominent nucleolus. That the endocytosedfusinite is enclosed by membrane is clearly shown in RAW cell (b), where intact mitochondria, ribosomes, endoplasmic reticulum and lyso-somes are also observed. Scale bar represents 1 mm.


Fig. 10. Transmission electron micrographs of CHO cell (a) with endocytosed fusinite (labelled f) showing similar striated ultrastructure.Located within the cytoplasm are a large nucleus, numerous rough and smooth endoplasmic reticulum, comparable to the subcellular orga-nelles found in normal CHO cells (b) that have not been exposed to fusinite particles. Scale bar represents 1 mm.



tissue, was studied by SEM, LM and TEM. The SEM studies,performed at various magnifications, provide information toform a complete picture of the structure of fusinite. At lowmagnification of a relatively large piece of fusinite (about50 mm), the original woody plant structure is visible (Fig.7a,b,c,f), with regularly spaced large and small pores whichare the cross-sectional views of elongated channels. Theselarge and small channels are very similar to the xylem andphloem of modern-day plants. These channels are parallel toeach other (Fig. 7d,e) and arranged in a distinct pattern. Thenumerous channels serve to increase the surface area offusinite extensively, hence facilitating gas/surface interactions.

Additional information on fusinite applications inbiological studies has been gathered. In our earlier study(Swartz et al., 1991), when a fusinite suspension wasinjected into the gastrocnemius muscle of a mouse, itmaintained its function as an EPR oxygen spin probe at thatsite over a long period of time (over 6 months). Groups ofmice in that experiment (Swartz et al., 1991) weremonitored with an L-band EPR spectrometer to measurethe [O2] of the muscle tissue where the fusinite suspensionhad been injected. The sensitivity of fusinite as an oxygenspin probe was tested throughout the duration of theexperiment by the application of a tourniquet around thehind leg proximal to the gastrocnemius muscle andmeasuring the EPR signals with the EPR spectrometer. Atthe end of experiment, the mice were killed and thegastrocnemius muscle removed and processed for histologi-cal study with the light microscope.

Unlike LiPc, one could easily prepare a sample of fusiniteof 5 mm or less in size by repeated grinding and filtration insuccession. The exhaustive screening of haematoxylin-stained sections prepared from all the experimental miceshows that the sectioning process caused less shattering ofthe fusinite when compared with sections of mouse muscleinjected with LiPc. Similar to the observation made on

muscle tissue injected with LiPc, the introduction of fusiniteinto the muscle tissue did not elicit a major immuneresponse. Evidently, some leucocytes were recruited to thesite of fusinite particles, but there were also no signs ofencapsulation or tumour formation (Fig. 8a–e). Further-more, there were no visible signs of migration of the fusiniteparticles away from the site of injection. Again, repeatedexaminations of the mice at various time intervals prior tokilling the animals showed no indication of the presence ofinduced tumours. Furthermore, all the mice in thisexperiment survived, with the exception of one which diedfrom improper anaesthesia. Our data suggest that fusinitemay be used for in vivo oximetry studies owing to itsapparent nontoxic and inert nature, but most of all, becauseit maintained its functional capacity as an EPR spin probe tomonitor [O2] continuously over a few months.

For in vitro studies (Swartz et al., 1991; Vahidi et al.,1994; Clarkson et al., 1995), fusinite had been used tomeasure the intracellular [O2] of both CHO and RAW cellsin culture by EPR, whereas soluble nitroxide (unable tocross the cell membrane owing to its charge) was used asthe spin probe to measure extracellular [O2]. At thecompletion of these experiments, cell samples wereprocessed for TEM study. Micrographs from this studyindicate that fusinite particles were in the cytoplasm (Figs.9 and 10a). Apparently, the presence of fusinite within thecells did not interfere with the ultrastructure of the cellsbecause both CHO and RAW cells show no signs ofsubcellular organelle disruption or damage when com-pared with a normal cell (Fig. 10b). The physiologicalfunction of the cells with endocytosed fusinite particle(s)was not compromised, as both experimental and controlcultures continued to grow and divide in a similar fashion.Therefore, we conclude that a nonexcessive amount offusinite may be safely used as a particulate spin probe formeasuring intracellular oxygen in vitro.


Table 1. The preparative conditions and the BET surface areas (calculated from nitrogen adsorption data) of the synthetic chars (Boyer 1992,personal communication).

Carbonaceous Heat treatment BET Surface areaSample no. precursor Activating agent:* sucrose temperature (8C) (m2 g–1)

1 sucrose 2:1 620 11382 sucrose 1:1 620 10553 sucrose 1:2 620 6724 sucrose 1:3 620 6705 sucrose 1:4 620 11006 sucrose 2:1 620 433

*ZnCl2 was the chemical agent used for preparing samples 1–5, and KOH was the chemical agent used for preparing sample 6. For the defi-nition for BET surface area see Boyer & Clarkson (1994).


In our laboratory, a preliminary study on synthetic charswas conducted. Our initial goal is to succeed in synthesizing achar that will surpass the performance of fusinite as an EPRoxygen spin probe. The synthetic char samples used in thisstudy are listed in Table 1. The precursor material used was

sucrose, and the basic differences between the samples were:(1) the type of chemical activating agent, one of which wasZnCl2 and the other being KOH, and (2) the ratio ofactivating agent to sucrose used during preparation. Theobjective of the activation process was to enhance the volumeand to enlarge the diameters of the pores during the charringprocess. A survey of the scanning electron micrographs of allthe different chars shows that the general micromorphologyof the chars differs markedly from that of fusinite describedearlier. No distinctive regular structural patterns are visible.Obviously, owing to the nature of the precursor, sucrose, it isnot surprising that the typical plant structure frequentlyobserved in the fossilized fusinite samples was completelyabsent. However, subtle differences in the pore sizes as well asin the distribution of the pores can be seen when a carefulcomparison is made between scanning electron micrographs(Fig. 11) of samples 1, 2 and 5 (see Table 1). The ratios ofactivating agent (ZnCl2) to sucrose are 2:1, 1:1 and 1:4,respectively. According to Boyer & Clarkson (1994), the BETsurface areas calculated from nitrogen adsorption data of thechar samples are dependent on the ratio of activating agentpresent during carbonization of sucrose (Table 1). They haveshown in their paper that ZnCl2-activated chars (samples 1–5) displayed a high degree of linearity and a moderate EPRlinewidth response over a broad range of oxygen pressures,while the KOH-activated char (sample 6) displayed linearityover a narrow range. Furthermore, they suggested that atrend was apparent when the EPR linewidth response wascorrelated with the adsorption pore volume distributions forthe two differently activated chars. Their conclusion was thata high degree of microporosity yielded a char with a linearlinewidth response to oxygen, while increasing nonlinearbehaviour was displayed as the mesoporous characterincreased. In this study, the International Union of Pureand Applied Chemistry classification of porosities wasfollowed: micropores, width <2 nm; mesopores, widthbetween 2 and 50 nm; and macropores, width >50 nm.Boyer & Clarkson also made an important observation thatincreasing the BET surface of the char sample increases itsresponse to oxygen. In order to fully understand thecorrelation between the porosity, particle size and geometryand the magnitude of oxygen response, the linearity of thisresponse and the kinetics of this response, reference to themanuscript of Boyer & Clarkson (1994) is recommended.

Conclusion

For the first time, by utilizing LM, SEM and TEM, wecompleted a detailed study on the micromorphology andstructure of the particulate EPR spin probes LiPc, methyl-LiPc, methoxy-LiPc, fusinite and synthetic chars, whichhave been used in various in vivo and in vitro oximetrystudies (Swartz et al., 1991, 1992; Duret et al., 1992; Liuet al., 1993; Smirnov et al., 1993a,b, 1994; Clarkson et al.,

Fig. 11. Scanning electron micrographs of synthetic chars fromcarbonizing sucrose with activating agent ZnCl2. ZnCl2: sucroseis 2:1 in (a), 1:1 in (b) and 1:4 in (c). A comparison of these micro-graphs demonstrates the variation in both size and distribution ofpores as well as the overall general morphology. Scalebars ¼ 100 mm, 10 mm and 50 mm in a, b and c, respectively.



1995; Vahidi et al., 1994). The micrographs clearly showthat each spin probe exhibits its unique morphology. Whencrystalline LiPc was chemically altered to add a methyl ormethoxy group, only methoxy-LiPc in a nonaqueousenvironment showed increased response and sensitivity tooxygen over LiPc. Therefore, the increased surface area inboth derivatives observed in the microscopy study was onlyone of the many factors that contributed to the finalfunctional characteristic of the EPR oxygen spin probe. Themicromorphology and structure of synthetic chars could bealtered, as shown by microscopy and adsorption analysis, bychanging certain chemical and/or physical parametersduring synthesis. Therefore, synthetic chars are excellentcandidates for developing new oxygen-sensitive EPR spinprobes. Their small size (# 8 mm) and their inertness andstability allows them to be delivered to the target area, andalso to remain as a functional reporter of the [O2] in thebiological environment over a significant time period. Thesehave been confirmed by our microscopy study on these spinprobes in cells and tissues. We strongly believe that ourdetailed and complete microscopy study of these EPR spinprobes contributes new and useful information towardsfuture development of new and improved particulate EPRoxygen probes, especially useful for clinical application.

Acknowledgments

This research was supported by NIH grant PO1-GM51630(to H.M.S. and R.B.C.) and RO1-GM42208 (to R.B.C.) Thisstudy used resources of the IERC (NIH P41-RR01811).

References

Belarbi, Z., Sirlin, C., Simon, J. & Andre, J.-J. (1989) Electrical andmagnetic properties of liquid crystalline molecular materials:lithium and lutetium phthalocyanine derivatives. J. Phys. Chem.93, 8105–8110.

Boyer, S.J. & Clarkson, R.B. (1994) Electron paramagneticresonance studies of an active carbon: the influence ofpreparation procedure on the oxygen response of the linewidth.Colloids Surfaces A: Physicochem. Eng. Aspects, 82, 217–224.

Byrne, J.F. & Marsh, H. (1995) Introductory overview. Porosity inCarbons: Characterization and Applications (ed. by J. W. Patrick),pp. 1–48. Halsted Press, New York.

Clarkson, R.B., Norby, S.W., Smirnov, A., Boyer, S., Vahidi, N.,Nims, R.W. & Wink, D.A. (1995) Direct measurement of theaccumulation and mitochondrial conversion of nitric oxidewithin Chinese hamster ovary cells using an intracellularelectron paramagnetic resonance technique. Biochim. Biophys.Acta, 1243, 496–502.

Duret, D., Moussavi, M., Jeandey, C. & Beranger, M. (1992) Oxygenconcentration measurements using the ESR line modification ofPcLi molecules. Sensors Actuators B, 6, 266–269.

Hyde, J.S. & Subczynski, W.K. (1989) Spin-label oximetry. BiologicalMagnetic Resonance (ed. by L. J. Berliner and J. Reuben), Vol. 8,pp. 399–422. Plenum Press, New York.

Liu, K.J., Gast, P., Moussavi, M., Norby, S.W., Vahidi, N., Walczak,T., Wu, M. & Swartz, H.M. (1993) Lithium phthalocyanine: aprobe for electron paramagnetic resonance oximetry in viablebiological systems. Proc. Natl. Acad. Sci. USA, 90, 5438–5442.

Sing, K.S.W. (1955) Physisorption of gases by porous carbons.Porosity in Carbons: Characterization and Applications (ed. by J. W.Patrick), pp. 49–66. Halsted Press, New York.

Sing, K.S.W., Everett, D.H., Haul, R.A.W., Moscou, L., Pierotti, R.A.,Rouquerol, J. & Siemieniewska, T. (1985) Reporting physisorp-tion data for gas/solid systems – with special reference to thedetermination of surface area and porosity. Pure Appl. Chem. 57,603–619.

Smirnov, A.I., Norby, S.W., Clarkson, R.B., Walczak, T. & Swartz,H.M. (1993a) Simultaneous multi-site EPR spectroscopy in vivo.Magn. Res. Med. 30, 213–220.

Smirnov, A.I., Norby, S.W., Walczak, T., Liu, K.J. & Swartz, H.M.(1994) Physical and instrumental considerations in the use oflithium phthalocyanine of measurements of the concentration ofthe oxygen. J. Mag. Res. B, 103, 95–102.

Smirnov, A.I., Norby, S.W., Weyhenmeyer, J.A. & Clarkson, R.B.(1993b) The effect of temperature on the respiration of culturedneural cells as studied by a novel electron paramagneticresonance technique. Biochim. Biophys. Acta, 1200, 205–214.

Swartz, H.M., Boyer, S., Brown, D., Chang, K., Gast, P., Glockner,J.F., Hu, H., Liu, K.J., Moussavi, M., Nilges, M., Norby, S.W.,Smirnov, A., Vahidi, N., Walczak, T., Wu, M. & Clarkson, R.B.(1992) The use of EPR for the measurement of the concentrationof oxygen in vivo in tissues under physiologically pertinentconditions and concentrations. Oxygen Transport to Tissue XIV(ed. by W. Erdmann and D. F. Bruley), pp. 221–228. PlenumPress, New York.

Swartz, H.M., Boyer, S., Gast, P., Glockner, J.F., Hu, H., Liu, K.J.,Moussavi, M., Norby, S.W., Vahidi, N., Walczak, T., Wu, M. &Clarkson, R.B. (1991) Measurements of pertinent concentra-tions of oxygen in vivo. Magn. Res. Med. 20, 333–339.

Turek, P. & Andre, J.-J. (1987a) Extreme spin exchange narrowingin a neutral phthalocyanine radical: the lithium phthalocyanine.Solid State Commun. 63, 741–744.

Turek, P. & Andre, J.-J. (1987b) Preparation and study of a lithiumphthalocyanine radical: optical and magnetic properties. Chem.Phys. Lett. 134, 471–476.

Turek, P., Andre, J.-J., Moussavi, M. & Fillion, G. (1989a) Septetspin state in the lithium phthalocyanine p-radical compound.Role of dioxygen. Mol. Cryst. Liq. Cryst. 176, 535–546.

Turek, P., Moussavi, M. & Andre, J.-J. (1989b) Magnetic propertiesof the lithium phthalocyanine p radical. Role of dioxygen.Europhys. Lett. 8, 275–280.

Vahidi, N., Clarkson, R.B., Liu, K.J., Norby, S.W., Wu, M. & Swartz,H.M. (1994) In vivo and in vitro EPR oximetry with fusinite: anew coal-derived, particulate EPR probe. Magn. Res. Med. 31,139–146.



Documents

Electron and light microscopy studies on particulate EPR spin probes lithium phthalocyanine, fusinite and synthetic chars