Magnetic Resonance Imaging, Vol. 15, NO. 10, pp. 1193-1198, 1997 0 1997 Elsevier Science Inc. All rights reserved.

Printed in the USA. 0730-725x/97 $17.00 + 00

ELSEVIER

l Original Contribution

PI1 SO730-725X(97)00180-X

‘H MAGNETIC RESONANCE IMAGING AND 31P MAGNETIC RESONANCE SPECTROSCOPY IN EXPERIMENTAL FILARIASIS

AMITA SHUKLA-DAVE,* NIGAR FATMA,* RA.IA ROY,? S. SRIVASTAVA,$ R.K. CHATTEFUEE,*

V. GOVINDARAIU,§ A. KASI VISWANATHAN,~ AND P. RAGHUNATHAN$

Divisions of *Parasitology, tMedicinal Chemistry, and $Toxicology, Central Drug Research Institute, Lucknow, India and $Department of NMR, All India Institute of Medical Sciences, New Delhi, India

‘H Magnetic resonance imaging and 31P magnetic resonance spectroscopy (MRS) have been carried out in experi- mental rodent iilariasis, i.e., Acanthocheibema viteae infection in the rodent host, Mastomys coucha The T,-weighted image of the infected host shows fine hyperintense thread like structures of adult tilariid nests in the cervical region. “P MRS of normal and infected hosts, localized over the same region of interest, show seven major peaks corresponding to phosphomonoesters (mchuiing ghrcose-6-phosphate+ fructose-tkphosphate, fructose~l-6-diphosphate, phosphorylchotie, and adenhte monophospbate or AMP), morganic phosphate, glycerophosphorylcholine, phos- phoenolpyruvate, phosphocreatine and nucleoside di- and tri-phosphates. Concentrations of phosphomonoesters (PMEs) are higher in the normal rodent compared with the infected ones. In vivo 31P MRS provides a non-invasive assessment of tissue hioenergetics and phosphohpid metabolism. 0 1997 Elsevier Science Inc.

Keywords: ‘H Magnetic resonance imaging; 31P Magnetic resonance spectroscopy; Phosphorus metabolism; Acanthocheilonema viteae; Mastomys coucha; tilariasis.

INTRODUCTION

Lymphatic filariasis is a major public health problem and inflicts a considerable social and economic burden on many tropical and subtropical countries. It is currently estimated that 1.1 billion people are at risk of being infected, with 120 million people aheady infected in 73 endemic countries.’ The disease itself is not life threatening but causes chronic suffering, including episodic fever, lymphangitis, lymphad- en&is, epididymo-orchitis, etc., which may lead to massive swollen limbs, enlarged scrotum, breasts and clitoris. The conditions commonly known as elephantiasis are directly responsible for a huge loss of man-hours, productivity and economic well-being. Although insufficient to explain the clinical spectrum of filarial diseases, the pathophysiology of filarial infections is generally attributed to the presence of parasites, repeated attacks of lymphangitis, and/or the host’s immune response to the parasite. Recent strategies for fi- laria control are first being focused on the parasite and its detection, then on the pathogenesis of filarial diseases, and finally on both individual and mass chemotherapy? Mag-

netic resonance imaging (MIU) has considerable potential for detecting live filarial parasites in situ and observing lymphatic alteration after drug therapy in filariasis; this potential has been exploited by only a few investigators. Case et al.3 have visualized lymphatic abnormalities in ferrets infected with Brugiu malayi, while Tanoura et al? examined the potential utility of MR to monitor response of hyperplastic lymph nodes to AMl-227 in ferrets infected with filariasis. Eberhard et al5 used this conventional tech- nique MIU for imaging microfilaria-positive chimpanzees and mangabey monkeys to detect adult Onchocercu volvu- Zus worms or nodules. The aim of the present study is to employ MRI and magnetic resonance spectroscopy (MRS) to delineate the morphological and metabolic characteristics of filariasis (Acanthocheilonemu viteae) in the rat (Musto- mys coucha).

MATERIALS AND METHODS

Host-parasite model. A. viteae filarial larvae were transmitted to Mustomys coucha by the tick vector, Or- nithodorus moubata6”

RECEIVED 5/21/96; ACCEFTED 7117197. Drug Research Institute, Chattar Manzil, P.O. Box 173, Luc- Address correspondence to: R.K. Chatterjee, Ph.D., Central know 226 001, India. E-mail: root@csdri.ren.nic.in

1193

1194 Magnetic Resonance Imaging 0 Volume 1.5, Number 10, 1997

Ex-vivo study. Adult A. viteae parasites were recov- ered from the subcutaneous tissues after sacrificing the animals for ex-vivo experiments.

In-vivo study. Animals were anaesthetised by inject- ing thiopentone intraperitoneally (Intraval sodium, May and Baker Pharmaceuticals, U.K.) at a dose of 50 mg/kg prior to MRI and MRS examination.

NMR ANALYSIS

Ex-vivo NMR. The 3’P nuclear magnetic resonance (NMR) spectra were performed by fixing the adult par- asites in low gelling agarose containing Ringer log so- lution. The experiments were carried out at 121 MHz on a Bruker Avance DRX 300 NMR spectrometer equipped with a 5-mm multinuclear probe head. The experimental conditions were as follows: transients 4000, pulse width 90”, delay time 3 s, and the FID (Free Induction Decay) was line broadened by 20 Hz prior to Fourier transfor- mation (IT).

Spectra of perchloric acid extracts were analyzed to assign resonances in the in vivo 31P NMR spectra by the method of Shukla et al.*

In-vivo NMR. 4.7T horizontal bore MRILMRS re- search system (BRUKER BIOSPEC 47140) was used’*” in combination with a volume resonator for ‘H NIvW31P MRS studies. A single turn double-tuned (‘H and 31P) surface coil of 5.0-cm diameter was used for 31P spec- troscopic studies.

All rats were imaged in a supine position after they had been anaesthetized with thiopentone at a single dose of 50 mgikg prior to the experiment. Each rat was then secured atraumatically with tape and placed inside the bore of the magnet, taking care to align the region of interest (ROI), namely the cervical region (comprising of skin and subcutaneous tissues), with the magnet iso- center. The coil was tuned to the desired nucleus (31P or ‘H). Prior to each MRILMRS scan, coarse shimming of the magnet was performed manually to obtain maximum signal intensity of the observed FID. Final adjustments of the magnetic field homogeneity were then made using the automated “compushim” procedure provided by BRUKER. This operation routinely produced a line width of 80-90 Hz, as measured by full width at half height of an acquired water spectrum.g”O

Control and infected rats were studied using different MRI and MRS methods. ‘H NMR was performed within a 15-cm-diameter gradient tube (maximum gradient strength being 225 mT/m) using a volume resonator of 7.2-cm inner diameter. Tl- and T,-weighted images were obtained using a quadrature transmitter/receiver coil op- erating at a proton resonance frequency of 200 MHz. A multiple slice, multiple echo (MSME) imaging proto-

col” was used with proper adjustment of 180” and 90” pulse attenuation to yield a strong echo. A 2-mm-thick transverse slice was imaged initially through the ROI. Using this as a “pilot” image, coronal slices of l-mm thickness were then imaged for a clearer visualization of the parasite-infected region. Figure 1B shows slices start- ing from the surface of the infected animal; it was, therefore, necessary to employ T2 contrast. The habitat of A. viteae filariids is the subcutaneous tissue layer of the host. Since the cervical region was clearly seen (surgi- cally) to contain the maximum number of parasites (Fig. ZB), this was chosen as the ROI in the present study for both the MRI as well as MRS studies.

T,-weighted images of the infected host were ob- tained with a data matrix of 256 X 256 covering a field of view (FOV) of 6 X 6 cm. The top five coronal slices of thickness (1 mm) were taken, and the third slice is illustrated in Fig. lA, which clearly depicts the soft tissue anatomy of the rat covering the ROI. T,-weighted images were obtained using an echo time of 80.0 ms and a repetition time of 2000 ms. Repetition time values longer than 2000 ms in T,-weighted images did not improve the quality of the image obtained and, further- more, required longer experimental time. The number of averages (n) was kept at n = 4 for T,-weighted images and n = 8 for T,-weighted images to obtain MRI images with a satisfactory signal-to-noise ratio.

Phosphorus-31 spectra were measured with a surface coil placed directly over the ROI, which covered a length of 5 cm and was of 2.5 cm depth.12-14 The 31P spectra were acquired using a single radiofrequency pulse. A total of 4096 transients were acquired covering a spectral width of 5 KHz with an acquisition size of 4K. For 3 s, repetition time, the flip angle was adjusted to optimise signal intensity.

The data generated for MRI and MRS studies have been repeated thrice. Three animals for each set of ex- periments were employed in the present study.

HISTOPATHOLOGY

Animals were anesthetized with an overdose of an- aesthesia and then dissected from the dorsal side, recov- ering parasites from the subcutaneous tissue of the neck region. Tissue samples from the area in the neck ROI showing the presence of the parasites in different tissue planes were taken. For confirmation, a small group of parasites recovered from another infected animal was also examined. These samples were preserved in 10% formalin and processed for paraffin embedding and sec- tioning by standard methods. Sections of 4-6 micron thickness were stained by haematoxylin and eosin and examined by light microscopy to locate the parasites in specific tissue compartments.

NMR in experimental filariasis 0 SHUKLA-DAVE ET AL. 119.5

E

Fig. 1. A, T,-weighted coronal image of the infected host. B, An infected rat dissected from the dorsal view. Arrow heads showing fine threadlike structures in the neck region of the rat depicting adult A. viteae parasites. C, Histological section of ROI of infected rat showing parasites (P) embedded in fat cells (FC). D, Histological section of ROI of infected rat showing parasites (P) embedded in intermuscular connective tissue (CT). E, Histological section of adult A. viteae parasite showing larval stages (LS) in the uterus of the female worm.

RESULTS AND DISCUSSION The T,-weighted image of the normal host was

similar to those of corresponding images from the infected animal, and it was therefore necessary to proceed with T, contrast. A T,-weighted image of the

infected host clearly shows parasite-mediated pathol- ogy. Fine, hyperintense, threadlike structures were identified as nests of adult A. viteae filariids in the ROI (Fig. 2, arrows). To confirm our findings, ex-vivo experiments were carried out. In the dissected infected

1196 Magnetic Resonance Imaging 0 Volume 15, Number 10, 1997

Fig. 2. T,-weighted coronal image showing hyperintense, threadlike structures (arrows) as nests of adult A. viteae in the cervical region (ROI) of the host.

rat (Fig. 1B) the arrow heads indicate the fine thread like structure of A. viteae adult parasites embedded in the subcutaneous tissue of the ROI. Recovered para- sites with a wet weight of 120 mg were washed with saline and kept inside a 20-mm NMR tube containing low melting point (37°C) agarose in Ringer log solu- tion. The T,-weighted images showed a hyperintense bundle of adult parasites surrounded by the buffer medium (Fig. 3A). The long T2 values of the parasites were a clear demonstration of their fluidity due to water content.

Histological sections of the ROI showed filarial par- asites (mainly the gravid female worms) lying in be- tween the group of fat cells (Fig. 1C) or in the intermus- cular connective tissue (Fig. ID). The section of parasites alone showed larval stages in the female uterus (Fig. 1E).

An ex-vivo 31P spectrum of the above A. viteae parasites in Ringer log solution enriched agarose gel was acquired. The spectra were composed of 15 peaks (Fig. 3C), which were assigned on the basis of chem- ical shift analysis. The shifts for peaks 9-10, 11-12 and 15 were consistent with the YNTP+PNDP, aNTP + CXNDP and the /3NTP phosphorus resonances bound to Mg. Peaks 13 and 14 contained NADP and the phosphorus resonance of UDPGlc. Peaks 7 and 8 were composed of glycerophosphorylcholine and phosphoenolpyruvate, whereas peaks l-6 gave a broad envelope of resonances originating from sugar

odm i i 4 -io 45 -io

C

Fig. 3. A, T,-weighted image showing hyperintense bundle of adult A. viteae (arrows) surrounded by the buffer medium (see text). B, 31P-NMR spectra of PCA extract of adult A. viteae parasites. C, 3’P-NMR spectra of adult A. viteae parasites in low gelling agarose enriched with buffer medium.

NMR in experimental filariasis l SHUKLA-DAVE ET AL. 1197

-I 2b I I 1 I ppm 15 10 5 0 -5 -10 -15 -20 -25 -30

Fig. 4. 31P-NMR spectra from localized normal host ROI.

phosphates, phosphomonoesters and inorganic phos- phate. Six control spectra were recorded.

Spectra of extract (Fig. 3B) based on direct chemical analysis indicated that resonances in the downfield monophosphate region of the spectrum arise from pri- marily glucose-6-phosphate (peak l), adenine monphos- phate (AMP) (peak 3), fructose-6-phosphate (peak 2), phosphocholine and fructose Id-diphosphate (peak 4) and inorganic phosphate (peak 5). A resonance 0.55 + 0.01 ppm downfield of glycerophosphorylcholine (peak 7) in the 3’P NMR spectrum was assigned to glycero- phosphorylethanolamine (peak 6).

On the basis of its chemical shift, the most prominent resoance of the spectrum arose from phosphoenolpymvate (PEP), (peak 8). Peak 9 is a doublet originating from the Y-phosphate of NTP, and peak 10 is a doublet originating from the p-phosphates of NDP; a doublet of doublets orig- inating from the a-phosphates of NTP and NDP (peaks 11 and 12) was observed. NASDP + (II) and uridine phos- phate of UDPGlc were detected (peak 13). Peak 14 of glucose phosphate of UDPGlc and a triplet of P-phosphate of NTP (peak 15) were also detected. The confirmation of the resonance assignments of PEP (peak 8) was done by adding t&odium salt of PEP in the perchloric acid extract after the normal 31P NMR experiment.

Fully resolved spectra of the extract confirmed the assignments of the ex-vivo 3’P NMR spectra of adult parasites embedded in gel.

Characterization and identification of various phos- phorus containing metabolites of normal and infected hosts were carried out by 31P MRS, the ROI being once again the cervical region.

The 31P MR spectra of the normal host cervical region

consisted of seven major peaks. Based on our chemical shift analysis, the peak assignments shown in Fig. 4 have been made. The major peaks originate from sugar phosphates and phosphomonoesters (peak l), Pi (peak 2), PCr (peak 3), peak 4 unassigned, YNTP+pNDP (peak 5), a-phosphates of NTPSNDP (peak 6) and PNTP (peak 7).

The 31P NMR spectra of infected hosts (A. viteae infection in Mustom~s) were also composed of the above-mentioned seven major peaks (Fig. 5). However, the major difference between the spectra obtained from normal and infected hosts was in the concentration of sugar phosphates and phosphomonoester (PME) (peak 1). The concentration of metabolites giving rise to peak 1 was higher in normal animals in comparison to the infected ones. The PME peaks originate from multiple components. Phosphoethanolamine and phosphocholine are cell membrane precursors and contribute to the PME peak with the signal from glycolytic intermediates and AMP. The relative contributions of these compounds to the PME and sugar phosphate (SP) peaks may change with disease15z’6 as observed in our study.

Therefore, the detection of nematodes in vivo from our MRI experiments, combined with 31P MRS using a surface coil of appropriate diameter localized over the same region, provides a non-invasive assessment of nor- mal and infected tissue bioenergetics and phospholipid metabolism. Such information is thought to be very useful in developing appropriate filaricidal drugs and monitoring their curative efficiency in infected animals.

Acknowledgmenfs-Financial assistance in the form of a fellowship to

A.S.D. and N.F. from the Council of Scientific & Industrial Research, New Delhi, India is gratefully acknowledged.

198 Magnetic Resonance Imaging 0 Volume 15, Number 10, 1997

I I 1 I I I r I 1 I ppm 20 15 10 5 0 -5 -10 -15 -20 -25 -30

Fig. 5. 31P-NMR spectra from localized infected host ROI.

REFERENCES

1. Ottesen, E.A.; Ramachandran, C.P. Lymphatic filariasis infection and disease: Control strategies. Parasitol. Today 11:129-131; 1995.

2. Ottesen, E.A. The human filariasis: New understanding, new therapeutic strategies. Current Opinion in Infections Diseases 7550-558; 1994.

3. Case, T.C.; Unger, E.; Bemas, M.J.; Witte, M.H.; Witte, C.L.; McNeill, G.; Crandall, C.; Crandall, R. Lymphatic imaging in experimental filariasis using magnetic reso- nance. Invest. Radiol. 27:293-297; 1992.

4.

5.

6.

7.

8.

Tanoura, T.; Bemas, M.; Darkazanli, A.; Elam, E.; Unger, E.; Witte, M.H.; Green, A. MR lymphography with iron oxide compound AMI-227. Studies in ferrets with filaria- sis. Am. J. Roentgenol. 159:875-881; 1992. Eberhard, M.L.; Dickerson, J.W.; Tsang, V.C.W.; Walker, E.M.; Ottesen, E.A.; Chandrashekar, R.; Weil, G.J.; Trpis, M.; Strobert, E.; Constantinidis, I.; Swen- son, R.B. Onchocerca volvulus: Parasitologic and sero- logic responses in experimentally infected chimpanzees and mangabey monkeys. Exptl. Parasitol. 80:454-462; 1995. Worms, M.J.; Terry, R.J.; Terry, A. Dipetalonema viteae, a filarial parasite of the jird Meriones unguiculatus. J. Parasitol. 47:963-970; 1961. Singh, D.P.; Misra, S.; Chatterjee, R.K. Response of Di- petalonema viteae in Mastomys natalensis to known anti- filarials. Indian J. Exptl. Biol. 26:48-52; 1988. Shukla, A.; Roy, R.; Bhaduri, A.P.; Chatterjee, R.K. A new approach in study of bioenergetics of macrofilariae using

9.

10.

11.

12.

13.

14.

15.

16.

nuclear magnetic resonance (NMR) spectroscopy. Indian J. Pharmacol. 27: 106-l 10; 1995. John, J.; Govindraju, V.; Raghunathan, P.; Mohankumar, V. Magnetic resonance imaging of temporal changes of neurotoxic lesion in the rat. Brain Research Bulletin 40: 273-277; 1996. Mohankumar, V.; John, J.; Govindraju, V.; Khan, N.A.; Raghunathan, P. Magnetic resonance imaging of NMDA- induced lesion of the medial preoptic area and changes in sleep, temperature and sex behaviour. Neuroscience Re- search 24:207-214; 1996. Morris, D.G. Nuclear Magnetic Resonance Imaging In Medicine and Biology. Oxford: Clarendon Press; 1986. Ackerman, J.H.; Grove, T.H.; Wong, G.G.; Gadian, D.G.; Radda, G.K. Mapping of metabolites in whole animals by 31P-NMR using surface coils. Nature 283: 167-170; 1980. Bendall, M.R.; Gordon, R.E. Depth and refocussing pulses designed for multipulse NMR with surface coils. J. Magn. Reson. 53:365-385; 1983. Ernst, R.R.; Bodenhausen, G.; Wokaun, A. One-dimen- sional Fourier Spectroscopy. In: Principles of Nuclear Magnetic Resonance In One and Two Dimensions. New York: Oxford Press, 1987: pp. 124-125. Radda, G.K. The use of NMR spectroscopy for the under- standing of disease. Science 223640-645; 1986. Taylor-Robinson, SD.; Marcus, C.D. Tissue behaviour measurements using phosphorus 3 1NMR. In: D.M. Grant, R.K. Harris, (Eds). Encyclopaedia of Nuclear Magnetic Resonance. New York: John Wiley & Sons, 1996; pp. 4765-477 1.