Kinetoplast DNA Signatures of Trypanosoma cruzi Strains ... - NCBI

American Journal of Pathology, Vol. 149, No. 6, December 1996 Copyright C) American Society for Investigative Pathology

Kinetoplast DNA Signatures of Trypanosoma cruzi Strains Obtained Directly from Infected Tissues

Annamaria R. Vago,* Andrea M. Macedo,* Riva P. Oliveira,* Luciana 0. Andrade,* Egler Chiari,t Lucia M. C. Galvdo,t Debora d'Avila Reis,$ Maria Elizabeth S. Pereira,f Andrew J. G. Simpson,§ Sebastiao Tostes,11 and S6rgio D. J. Pena* From the Departamento de Bioquimica e Imunologia* and Departamento de Parasitologia,t Universidade Federal de Minas Gerais, Belo Horizonte, the Centro de Pesquisas "Rene Rachot ' , Funda!ao Oswaldo Cruz, Belo Horizonte, the Ludwig Institute for Cancer Researchb, Sao Paulo, and the Faculdade de Medicina do TrIangulo Mineiro,1 Uberaha, Brazil

We report here a polymerase chain reaction (PCR)-based DNA profiling technique that permits Trypanosoma cruzi strain characterization by direct study of infected tissues. This is based on application of a recently developed method of DNA fragment identification, called low-stringency single specifc primer PCR (LSSP-PCR), to the study of the variable region of kinetoplast DNA (kDNA) minicircles from T. cruzi Thus, we can translate the intraspecific polymorphism in the nucleotide sequence of kDNA minicircles into a specifc and highly reproducible kDNA signature. Comparison with the phenogram obtained by DNA fingerprinting analysis of a set of T. cruzi strains showed good qualitative correlation between the degree of divergence of the LSSP-PCR profiles and the genetic distance between the strains. kDNA signatures of heart tissue from acutely or chronically infected animals revealed perfect concordance with the patterns obtained from cultured parasites for the CL and Colombiana strains but not for the Y strain, which is known to be multiclonal. However, the match was perfect for studies with two clones of the Y strain. We take this as evidence that in some multiclonal strains there is heterogeneity among

the clones in the degree of tropismfor the heart tissue. Finaly, we showed that it is possible to obtain a T. cruzi kDNA signature from the heart of a human patient with chronic Chagasic myocardiopathy. kDNA signatures obtained by LSSPPCR of sequences amplifiedfrom infected tissues constitute a new tool to study the molecular epidemiology of Chagas' disease. (Am J Pathol

1996, 149.2153-2159)

Chagas' disease, caused by the protozoon Trypanosoma cruzi, has a variable clinical course, ranging from symptomless infection to severe chronic disease with cardiovascular or gastrointestinal involvement or even overwhelming acute episodes. The factors influencing this clinical variability have not been elucidated, but most likely genetic variation of both the host and parasite is important. Thus, research into intraspecific genetic polymorphisms of T. cruzi has the potential of leading to a better understanding of the molecular epidemiology of the disease. The first breakthrough in the high resolution of genetic variability in T. cruzi was the discovery of kinetoplast DNA (kDNA) restriction fragment length polymorphisms.1 It was demonstrated that the nucleotide sequence of the 330-bp variable portion of the kDNA minicircle molecule evolves rapidly enough to produce differences betweenisolates but not so rapidly as to preclude a stable genetic identity of the strain. The name schizodeme was given to the characteristic restriction profile of a strain. Further characterization of strain genetic identity, this time exSupported by grants-in-aid from the Conselho Nacional de Pesquisas (CNPq), Financiadora de Estudos e Projetos (FINEP), and Funda9ao de Amparo A Pesquisa do Estado de Minas Gerais (FAPEMIG). Accepted for publication July 29, 1996. Address reprint requests to Dr. Sbrgio D. J. Pena, Departamento de Bioquimica e Imunologia, Instituto de Ciencias Biologicas UFMG, 30161-970 Belo Horizonte, Brazil.

2153

2154

Vago et al

4/P December 1996, Vol. 149, No. 6

ploring nuclear DNA variability, came from studies of DNA fingerprinting with multilocal core probes directed to families of minisatellites.2 DNA fingerprinting afforded approximately the same high level of strain discrimination as schizodeme analysis, while having the advantages of higher stability and greater operational simplicity.2 Other molecular approaches to the study of genetic polymorphism in T. cruzi, eg., isoenzyme analysis3 or random amplified polymorphic DNA,4 were not capable of producing the same resolution as schizodeme analysis and DNA fingerprinting. Both schizodeme analysis and DNA fingerprinting require large amounts of parasites for study, which means that trypanosomes have to be initially cultivated from the blood of patients and later multiplied in laboratory animals or axenic cultures. In this respect, an important confounding factor is that some T. cruzi strains are known to be multiclonal populations.1'2 As culture involves adaptation to new environmental conditions and remarkable population expansion, there is ample opportunity for clonal selection. Consequently, the parasite strain obtained from a patient might be distinct from the population that was actually circulating in the blood. On the other hand, the circulating parasite population is not necessarily identical to the set of clone(s) that actually infected the heart or other specific tissues. Among the multiclonal population infecting an individual (from one or multiple encounters with infective reduviid bugs), we may have clones with specific tropism to different tissues, and it is exactly the distribution of these clones that may influence or determine the clinical course of that given patient. Thus, it is not surprising that studies in our laboratory did not show any significant correlation between DNA fingerprinting patterns of the T. cruzi strains cultured from patients and the clinical course of the disease (R. P. Oliveira, A. M. Macedo, and S. D. J. Pena, unpublished observations). To unravel the molecular epidemiology of Chagas' disease at a fine level, we need to be able to study parasite variability directly in clinical tissues. We report here our success in developing a polymerase chain reaction (PCR)-based DNA profiling technique that permits strain characterization in T. cruzi by direct study of infected tissues. To achieve this, we applied a sensitive technique for DNA fragment identification recently developed by us and called low-stringency single specific primer PCR (LSSP-PCR)5,6 to the study of the variable region of kDNA from T. cruzi. With this technique we could translate the intraspecific polymorphism in the nucle-

otide sequence of kDNA minicircles into a specific and highly reproducible kDNA signature.

Materials and Methods Strains and Animals Most of the T. cruzi strains used in this study were obtained from patients, animal reservoirs, or triatomid vectors in an endemic region of Chagas' disease in the state of Minas Gerais, Brazil. Parasites were obtained and total DNA was isolated as previously described.2 To obtain infected heart samples, outbred Swiss mice were inoculated with 5 x 104 trypomastigotes of the CL,7 Y,8 or the Colombiana9 strain and with 5 x 103 trypomastigotes of the Ystrain clones (YP2 and YP3).10 Animals were inoculated and then sacrificed after 14 days (acute phase) or 3 to 6 months (chronic phase). Heart tissue sections of approximately 0.5 x 5 x 5 mm were taken and exhaustively washed with isotonic saline. The tissue was then minced, subjected to alkaline lysis with 50 mmol/L NaOH, neutralized with 130 mmol/L Tris/HCI (pH 7.0), and used directly in the PCR reaction after 10-fold dilution in water. Extraction of DNA from mouse heart tissue fixed in formaldehyde and embedded in paraffin was done in exactly the same way as described below for the human sample.

Patients We analyzed a paraffin-embedded heart tissue sample from a 61-year-old male from Uberaba, Minas Gerais, Brazil, who died in 1993 with chronic Chagasic cardiopathy. The necropsy findings were cardiac enlargement (weight, 470 g) with chronic productive epicarditis. The histochemical analysis of the heart tissue showed intense inflammatory foci and fibrosis but no intracytoplasmic amastigote forms of the parasite. For extraction of DNA from the paraffin block, we used a technique that precludes the use of xylene. One heart tissue section of approximately 0.5 x 4 x 4 mm was obtained, and excess paraffin was trimmed off with a blade. The section was minced and directly subjected to digestion with proteinase K (1 j,g/,ul) in 100 g] of lysis buffer (100 mmol/L Tris/HCI pH 8.0, 40 mmol/L EDTA, pH 8.0, 0.5% sodium dodecyl sulfate), at 37°C for 48 hours, followed by heating at 1000C for 10 minutes to inactivate the enzyme. The lysate was used directly in the PCR reaction after 10-fold dilution in water.

kDNA Signatures of T. cruzi in Infected Tissues

2155

AJP December 1996, Vol. 149, No. 6

kDNA Signatures The production of kDNA signatures by LSSP-PCR is a two-step procedure. The first step consists in the specific PCR amplification of the 330-bp fragment representing the four variable regions of the T. cruzi minicircle DNA. The products are then gel purified and subjected to LSSP-PCR in a second step.56 The first-step (specific) PCR was carried out using the primers S35 (5'-AAATAATGTACGGGGGAGATGCATGA-3') and S36 (5'-GGGTTCGATTGGGGTTGGTGT-3').11 The DNA template consisted of 2.5 ng of the total DNA of parasites grown in culture or 2.5 ,ul of the lysate obtained from human and mouse tissue samples, diluted 10 times in double-distilled water. Twenty-five amplification cycles were carried out with annealing at 600C for 1 minute, extension at 720C for 2 minutes, and denaturation at 940C for 1 minute, preceded by an initial denaturation at 940C for 5 minutes. The PCR products were run in 2% agarose gel (1/3 low melting point agarose, Sigma Chemical Co., St. Louis, MO) and stained by ethidium bromide. The bands corresponding to the 330-bp fragment were visualized by long-wave ultraviolet and excised from the gel. These fragments were diluted 1:10 in double-distilled water, and 1 ,xl of the dilution (approximately 15 ng of DNA) was used as template for the LSSP-PCR reaction, which was performed exactly as described5'6 using primer S35 as driver. A 5-,ul volume of the amplification reaction products was loaded in each lane in 6% polyacrylamide gel and silver stained as described previously. 12

DNA Fingerprinting DNA fingerprinting was obtained exactly as described previously2 using the multilocal probe 33.1513 labeled by the incorporation of biotin.14 The distance metric D = 1 S (where S is the proportion of shared bands in the DNA fingerprint) was used to build a phenetic tree by the unweighted pair group method with arithmetic mean (UPGMA) on the MEGA program, version 1.0.15 -

Results kDNA Signatures of Culture Parasites To evaluate its usefulness in the genetic discrimination of T. cruzi strains, the LSSP-PCR analysis was performed on the four 330-bp fragments that constitute the variable region of the T. cruzi kDNA mi-

nicircle. This was the same target segment utilized for the schizodeme classification based on restriction analysis.1 We initially studied 16 different T. cruzi strains grown in culture. LSSP-PCR produced a multiband pattern that constituted the kDNA signatures and was variable among some of the strains although very similar in others (Figure 1 B). When we compare the kDNA signatures with the phenogram obtained by DNA fingerprinting analysis of these strains with a multilocal biotinylated 33.15 probe (Figure 1A), it is evident that the former afford a lesser genetic resolution, but there is a good qualitative correlation between the degree of divergence of the LSSP-PCR profiles and the genetic distance between the strains as measured by DNA fingerprinting band sharing. For instance, the group made up of strains 181, 226, 222, 231, and 115, all isolated from Chagas' disease patients from the same endemic geographic area in Minas Gerais state, were clustered in the phenogram and showed identical kDNA signatures. The same was observed for strains CL, 182, and 167, which had very similar LSSP-PCR profiles to the 1030 and 1023 strains, which were isolated from humans or T. infestans vector. These 5 strains also clustered in the DNA fingerprinting phenogram. On the other hand, these two groups of 5 strains differed considerably from each other in their kDNA signatures, as they differed from several of the other isolates analyzed, eg, 229, DE, LA, and the 1017/1001 cluster.

kDNA Signatures of Infected Hearts The next step in our study was to evaluate whether LSSP-PCR could be reliably used to identify the T. cruzi strain present in tissues of an infected animal. We chose to start our studies with animals in the acute phase of infection, because of a higher abundance of parasites. Thus, animals were infected with trypomastigotes of the CL and the Y strains and a comparison was made of the kDNA signature of the cultured strains and of the parasites present in the myocardium of infected mice. In the case of the CL strain, the two kDNA signatures were identical (Figure 2A), which would have allowed identification of the infecting strain by comparison with known patterns. Indeed, the CL strain kDNA signature would have been recognized among those of the strains shown in Figure 1B. On the other hand, the two signatures obtained with the Y strain consistently differed (Figure 2B, lanes 2 and 3). We suspected that this was caused by selection of clones on the basis of differing tissue tropism, as the Y strain is

2156

Vago et al


r-

I-m I1

--LI

7

r-

r..

I,

r__. 0-"

rLi+

[LI

181 226 222 231 115 229 DE LA 1017 1001 0301023 JM CL 182 167

M

-_

Figure 1. Correlation between the LSSP-PCR profiles of 16 strains of T. cruLzi and the phenogram obtained by DNA finigerprinting. A: Phenogram obtained with 16 differenit T. cruzi strains analyzed by DNA fingerprinting using biotinylated multilocal 33 15 probe. The tree was obtainied by the unweighted pair grou.p with arithmetic mean method using the percentage of zunshared bands as the measuire fo- genetic distance. B: LSSP-PCR profiles obtainedfromn 330-bpfragments of minicircles of the same 16T. crLzi strains. The S35primer was usedfor the second round qf amplffication. Five mzicroliters of the LSSP-PCR reaction products were loaded in each lane in a 6% polvacrylamide gel and silver stained. Migration of the markers of the 1-kb ladder (Life Technologies, Gaithersburg, MD) is showtn in the lane M.

known to be multiclonal,j but despite the excellent reproducibility of LSSP-PCR (see below) we could not rule out a priori the possibility that the different profiles were due to technical artifacts. To test between the two alternatives, we infected mice with two

clones of the Y strain (YP2 and YP3) and the results this time revealed that the clones, although differing widely from each other, had identical kDNA signatures in the cultures and in the myocardia (Figure 2B, lanes 5 to 8).


2157


)1

2

3

1

2

3

4

5

6

7

8

,Rft-OVI

Figure 2. Analysis of the kDNA signatures obtainedfrom 330-bp minicirclefragment ofT. cruzi amplifiedfrom culturedparasites orfrom the heart of acutely infected mice. A: kDNA signatures obtained with the CL strain from culture (lane 2) and heart tissue (lane 3). The migration of the markers of the 1-kb ladder (Life Technologies) is shown in lane 1. Five microliters of the LSSP-PCR reaction products were loaded in each lane in a 6% polyacrylamide gel and silver stained. B: kDNA signatures obtained with the Ystrain and three of its clones. Lane 2, Ystrain, culture; lane 3, Ystrain, heart tissue; lane 4, clone YP1, culture; lane 5, clone YP2, culture; lane 6, clone YP2, heart tissue; lane 7, clone YP3, culture, lane 8, clone YP3, heart tissue. Lane 1 shows the migration of the markers of the 1-kb ladder (Life Technologies).

kDNA Signatures of Chronically Infected Hearts We next tested whether strain identification would be possible in the chronic phase of the disease. We chose to study this with the Colombiana strain, which is known to exhibit persistence of parasites in heart tissue during chronic infection.9 Again, the kDNA signatures were identical in acute and chronic infections (Figure 3, lanes 1 to 3). Finally, we decided to ascertain whether it would be possible to obtain kDNA signatures from human subjects with Chagas' disease. We studied postmortem heart tissue from a patient with chronic Chagasic myocardiopathy and obtained good PCR amplification of kDNA, which permitted the production of a clear LSSP-PCR pat-

tern that did not match any of the known strains

(Figure 3, lane 4).

Reproducibility of the kDNA Signatures With LSSP-PCR, identification of a clone or a strain present in an infected tissue is achieved by comparison with reference signatures obtained on different days and kept on file. Thus, day-to-day and operator reproducibility are critical. We have observed that LSSP-PCR patterns were very reproducible when obtained on separate days, using distinct thermocyclers and electrophoretic runs, even when undertaken by different workers in our laboratory. For instance, the kDNA signatures of the CL strain in

Vago et al

2158


AbAwd0k IFIW P.

4

3

2

1

5

.;

-,

.1,

I

I:

U.-pow

00'

Figure 3. kDNA signatures of the 330-bp minicircle fragment of T. cruzi amplifiedfrom chronically infected hearts. Lanes 1 to 3, heart tissues from mice itnfected with the Colombiana strain (lane 1, aculte phase; lane 2, chronic phase, 3 months; lane 3, chronic phase, 6 months). Lane 4 shous the kDNA signature obtained with the heart tissue obtained at the autopsy of a human subject with chronic Chagas' disease. Five mnicroliters of the LSSP-PCR reaction products u'ere loaded in each lane in a 6%polyacrylamidegel and silverstained. The migrationi of the molecular size standards of the 1-kb ladder (Life Technologies) is shoun in lane 5.

Figure 1 and those in lanes 1 and 2 of Figure 2 are virtually identical. Likewise, we have also determined that parasites present in specimens subjected to formalin fixation and paraffin inclusion have kDNA signatures identical to those in fresh tissues.

Discussion LSSP-PCR is a simple and rapid PCR-based technique for detecting DNA sequence variation that is both highly sensitive and informative.5'6 It consists of submitting a purified DNA fragment to multiple cycles of PCR amplification in the presence of a single oligonucleotide primer (driver), specific for one of the extremities of the fragment, under conditions of very

low stringency. The driver hybridizes with high specificity to its complementary extremity and with low specificity to multiple sites within the fragment in a sequence-dependent manner. The reaction thus yields a large number of products that can be resolved by polyacrylamide gel electrophoresis to give rise to a multiband DNA fragment signature that reflects the underlying sequence. Changes as small as single base mutations can drastically alter the multiband pattern, producing new signatures that are diagnostic of the specific alterations. LSSP-PCR is generally applicable to the detection of single or multiple mutations in any gene-sized DNA fragment and has been used for detection of single base changes in human genetic diseases,5 as an identity test in humans based on polymorphisms of mitochondrial DNA,16 and for the genetic typing of papillomavirus.17 The application of LSSP-PCR to the characterization of variation of the 330-bp variable portion of the kDNA minicircle molecule permitted the recognition of complex banding patterns (kDNA signatures) that allow recognition of clones and strains with good, although not absolute, discrimination. Using LSSP-PCR, we attempted to identify strains in infected tissues. In the case of the CL strain and of clones of the Y strain, recognition from the study of acutely infected myocardia was straightforward, because the kDNA signatures were identical to the culture counterparts. However, with the Y strain, the pattern from heart tissue was consistently different from culture. We take this as a suggestion that in multiclonal strains, such as Y, there is heterogeneity among the clones in the degree of tropism for the heart. We also could obtain reliable and reproducible kDNA signatures from myocardial tissue from chronically infected mice (with the Colombiana strain) and, in the single case studied, from a human heart. We have previously demonstrated the excellent stability and reproducibility of LSSP-PCR profiles in different applications.5'16 Likewise, kDNA signatures from the T. cruzi strains present in infected tissues were found to be reproducible. Identical kDNA signatures were obtained with LSSP-PCR reactions from several different mice inoculated with the same strain, and were reproducible on different days, using different thermocyclers and when performed by different people. LSSP-PCR is a two-step procedure that obviously depends on the success of the initial amplification of the target fragments. The presence of inhibitors in the analyzed samples can decrease the efficiency of the first-step PCR amplification, but provided that careful attention is paid to the purifica-


2159


tion of the amplicon, artifacts of the kDNA signatures are not expected. Indeed, identical kDNA signatures were observed from the same tissue, whether fresh or formalin fixed and paraffin embedded. The fact that kDNA is a naturally amplified target facilitates LSSP-PCR even in suboptimal specimens, such as archival material. In conclusion, our results indicate that the kDNA signatures obtained by LSSP-PCR of T. cruzi sequences amplified from infected tissues constitute a new tool to study the molecular epidemiology of Chagas' disease. The next obvious step is the study of more human cases. If we succeed in observing recurring kDNA signatures that show any correlation with the clinical form of the disease, we will have opened new vistas for understanding the pathogenesis of this important parasitic disease.

7.

8.

9.

10.

11.

References 1. Morel C, Chiari E, Plessmann Camargo E, Mattei DM, Romanha AJ, Simpson L: Strains and clones of Trypanosoma cruzi can be characterized by pattern of restriction endonuclease products of kinetoplast DNA minicircles. Proc Natl Acad Sci USA 1980, 77:68106814 2. Macedo AM, Martins MS, Chiari E, Pena SDJ: DNA fingerprinting of Trypanosoma cruzi: a new tool for characterization of strains and clones. Mol Biochem Parasitol 1992, 55:147-154 3. Carneiro M, Romanha AJ, Chiari E: Biological characterization of Trypanosoma cruzi strains from different zymodemes and schizodemes. Mem Inst Oswaldo

12.

13.

14.

15.

Cruz 1991, 86(4):387-393 4. Steindel M, Dias Neto E, Menezes CLP, Romanha AJ, Simpson AJG: Random amplified DNA analysis of T. cruzi strains. Mol Biochem Parasitol 1993, 60:71-80 5. Pena SDJ, Barreto G, Vago AR, De Marco L, Reinach FC, Dias Neto E, Simpson AJG: Sequence specific DNA fragment by PCR amplification with single specific primers at low stringency (LSSP-PCR). Proc Natl Acad Sci USA 1994, 91:1916-1919 6. Pena SDJ, Simpson AJG: LSSP-PCR: multiplex mutation detection using sequence-specific gene signa-

16.

17.

tures. Laboratory Protocols for Mutation Detection. Edited by U. Landegren. Oxford, Oxford University Press, 1996, pp 42-47 Brener Z, Chiari E: Aspects of early growth of different Trypanosoma cruzi strains in culture medium. J Parasitol 1965, 51:922-926 Silva LHP, Nussenzweig V: Sobre uma cepa de Trypanosoma cruzi altamente virulenta para o camundongo branco. Folia Clin Biol 1953, 20:191-207 Federici EE, Abelmann WH, Neva FA: Chronic and progressive myocarditis and myositys in C3H mice infected with Trypanosoma cruzi. Am J Trop Med Hyg 1964, 13:272-280 Marques de Araujo S, Chiari E: Caracterizaao biol6gica de clones das cepas Y, CL e MR de Trypanosoma cruzi em camundongos. Mem Inst Oswaldo Cruz 1988, 83:175-181 Avila H, Goncalves AM, Nehme NS, Morel CM, Simpson L: Schizodeme analysis of Trypanosoma cruzi stocks from South and Central America by analysis of PCR amplified minicircle variable region sequences. Mol Biochem Parasitol 1990, 42:175-187 Santos FR, Pena SDJ, Epplen JT: Genetic and population study of a Y-linked tetranucleotide repeat DNA polymorphism with a simple non-isotopic technique. Hum Genet 1993, 90:655-656 Jeffreys AJ, Wilson V, Thein SL: Hypervariable "minisatellite" regions in human DNA. Nature 1985, 316: 67-73 Macedo AM, Medeiros AC, Pena SDJ: A general method for efficient non-isotopic labeling of DNA probes cloned in M13 vectors: application to DNA fingerprinting. Nucleic Acids Res 1989, 17:4414 Kumar S, Koikiro T, Masatoshi N: MEGA: Molecular Evolutionary Genetics Analysis, version 1.0. University Park, PA, The Pennsylvania State University Barreto G, Vago AR, Ginther C, Simpson AJG, Pena SDJ: Mitochondrial D-loop "signatures" produced by low-stringency single specific primer PCR (LSSP-PCR) constitute a simple comparative human identity test. Am J Hum Genet 1996, 58:609-616 Villa LL, Caballero OL, Levi JO, Pena SDJ, Simpson AJG: An approach to human papillomavirus identification using low-stringency single specific primer PCR. Mol Cell Probes 1995, 9:45-48