ARTICLE

Quantitative Computer Image Analysis of Chondroitin Sulfate A Expression in Placentas Infected with Plasmodium Falciparum

Correspondence to: Dominique Gaillard, Unité INSERM U314, Université de Reims Champagne-Ardenne, CHU Maison Blanche, 45 rue Cognacq-Jay, 51092 Reims Cedex, France.

	Summary

Top Summary Introduction Materials and Methods Results Discussion Literature Cited

Most pathological conditions resulting from infection with the human malaria parasite Plasmodium falciparum occur as a consequence of the sequestration by several adhesion molecules of parasite-infected red blood cells (IRBCs). Recent reports have provided evidence that placental vascular endothelial ligands for IRBCs were mostly restricted to chondroitin sulfate A (CSA). The expression of CSA in malaria-infected placentas was investigated in a prospective case-control study in a hypoendemic area (Dakar, Senegal). The tissue distribution of CSA was measured in the terminal villi by immunostaining combined with image processing in 20 infected and 20 noninfected frozen sections of placenta. The villous surface immunostained by anti-CSA antibody was higher in infected than in noninfected placentas (p<0.03), in placentas with active infection than in those with past chronic infection (p<0.05), and in infected placentas with positive imprints than in those with negative imprints (not significant; p=0.06). Labeling was found in the extracellular matrix and in endothelial and stromal cells of all the placentas. Syncytiotrophoblast immunostaining was detected in all placentas associated with active or active chronic infection (n=7) but in only 4/13 placentas with past chronic infection (p<0.01). The presence of P. falciparum in the imprint was significantly correlated with immunostaining of CSA in syncytiotrophoblasts (p=0.003). These results suggest that CSA can play an important role in the sequestration of P. falciparum in human placentas during the acute phase of infection. (J Histochem Cytochem 47:751–756, 1999)

Key Words: image analysis, chondroitin sulfate A, placenta, Plasmodium falciparum, infection

	Introduction

Top Summary Introduction Materials and Methods Results Discussion Literature Cited

During pregnancy, Plasmodium falciparum malaria is associated with a number of maternofetal complications such as pre-eclampsia (Sartelet et al. 1996 ). There is some evidence that maternal circulation in the placenta particularly retains P. falciparum-infected red blood cells (IRBCs), but this alone cannot explain why parasitemia is usually much higher in the placenta than in the peripheral blood (Desowitz and Buchbinder 1992 ). The sequestration of IRBCs in vessels of various deep organs is believed to play a central role in the pathogenesis of severe P. falciparum infection (Aikawa et al. 1990 ) and in placenta infected with P. falciparum (Bray and Sinden 1979 ).

The availability of in vitro parasite cultures and assays to detect and measure cytoadherence of IRBCs on target cells has enabled the identification of several host proteins that are putative receptors to which IRBCs bind (Chulay and Ockenhouse 1990 ). Several receptors have been identified, such as CD36 (Oquendo et al. 1989 ), ICAM1 (Berendt et al. 1989 ) and, more recently, chondroitin sulfate A (CSA) (Robert et al. 1995 ; Rogerson et al. 1995 ). CSA is a glycosaminoglycan that is a natural component of the extracellular matrix. However, CSA was contained in different proteoglycans, among them thrombomodulin, which is located on the plasma membrane of endothelial cells and placental syncytiotrophoblasts. Recently, Rogerson et al. 1997b have demonstrated that thrombomodulin supports binding of selected lines of IRBCs and that adhesion is dependent on the presence of the CSA chain of thrombomodulin. In addition, a recent report has provided evidence that placental vascular endothelial ligands for IRBCs were mostly restricted to CSA (Fried and Duffy 1996 ); the authors proposed that placental infection with P. falciparum and maternal malaria were associated with distinct P. falciparum subpopulation, migrating from the peripheral circulation of pregnant women. These parasite subpopulations, which rarely emerge in nonpregnant hosts, appear to be preferentially sequestered and to multiply in placenta.

The aim of this study was to analyze and quantify expression of CSA in P. falciparum-infected or noninfected human placenta by immunohistochemistry and image analysis, to determine if (a) CSA is more strongly expressed in the villi of infected placentas and (b) CSA is found in syncytiotrophoblasts.

	Materials and Methods

Top Summary Introduction Materials and Methods Results Discussion Literature Cited

A prospective case-control study was carried out in two maternity centers in Dakar (Senegal) where malaria is hypoendemic with a low and seasonal transmission: the Hôpital Principal de Dakar and the Public Health Center in Ouakam. Placentas were collected with informed consent of the parturients during a 5-month period (July to November, the rainy season), when malarial transmission was maximal (Trape et al. 1992 ). Because it had previously been reported that placental pathology in malarial infection is most marked in the medial placental zone (Walter et al. 1982 ), samples for histological examination were processed from this area. These samples were fixed in Bouin's solution rather than in formalin (to eliminate formalin pigment formation) or were frozen in isopentan. Imprints of placental samples were stained with Giemsa's dye and examined for the presence of malaria parasites. Venous blood of parturients and newborn cord blood samples were examined for the presence of malaria parasites by the QBC test (Levine et al. 1989 ). Twenty infected and 20 noninfected samples were chosen randomly for immunohistochemical study.

Pathological Examination
All fixed samples were embedded in paraffin. Paraffin sections 5 µm thick were stained with hematoxylin, eosin, and saffron (HES). Sections were examined under light microscopy and polarized light to assess the deposition of malarial pigment. To confirm the presence of malarial pigment and to eliminate the artifact due to formalin pigment, cryostat sections of positive samples were taken and stained with HES.

Malarial infection of the placenta was classified into four groups, using the histological classification of Bulmer et al. 1993 :

• Category N, noninfected

• Category 1, active infection: parasites in maternal erythrocytes in the intervillous space; pigment in erythrocytes and circulating monocytes in the intervillous space, but neither pigment nor cells within fibrin.

• Category 2, active chronic infection: parasites in maternal erythrocytes in the intervillous space; pigment in erythrocytes and circulating monocytes within the intervillous space, and pigment in fibrin or cells within fibrin and/or chorionic villous syncytiotrophoblasts or stroma.

• Category 3, past chronic infection: parasites not present; pigment confined to fibrin or cells within fibrin.

Placentas were deemed infected when they showed histological lesions of P. falciparum infection and belonged to Category 1, 2, or 3, irrespective of the positivity or the negativity of the imprint. Placentas without such histological lesions were considered noninfected.

Immunohistochemical Study
For each placental sample, 5-µm cryostat sections were cut. The slides were rinsed for 10 min with PBS at room temperature (RT). Blocking nonimmune serum was applied for 10 min and the tissue section was incubated for 1 hr at RT with a mouse IgG₁ primary MAb against CSA (1:200; Chemicon International-Euromedex, Souffelweyersheim, France). Secondary incubation was performed with an avidin–biotin immunoperoxidase kit (Universal, LSAB Kit; Dako, Glostrup, Denmark) and the sections were stained at RT for 5 min with diaminobenzidine (DAB; Immunotech, Marseille, France). The sections were counterstained with ethyl green (CAS; Elmhurst, NY) which actually gives a nuclear stain (Figure 1A and Figure 2).

View larger version (95K): [in this window] [in a new window]   Figure 1. (A) Frozen section of P. falciparum-infected placenta (Category 3, exhibiting cytoplasmic expression of CSA in stromal and endothelial villous cells (S, stroma of terminal villi; C, fetal capillaries) and in syncytiotrophoblast (small arrows) with membranous localization. CSA. Bar = 70 µm. (B) In the same field as in A, Surface 1 delineates the counterstaining ethyl green, the brown staining of DAB, and the very light background staining in the cytoplasm and the extracellular matrix detectable by image analysis. Bar = 140 µm. (C) In the same field as in A Surface 2 delineates the DAB labeling. Bar = 70 µm. Figure 2. Representative frozen section of noninfected placenta, showing positive cytoplasmic stromal cells (S, stroma of terminal villi; C, fetal capillaries) and extracellular matrix immunostaining for CSA but not in syncytiotrophoblasts (small arrows). CSA. Bar = 70 µm. Figure 3. Cell imprints of P. falciparum-infected placental tissue, showing red blood cells infected with P. falciparum (mostly schizonts at all maturation stages and, to a lesser extent, trophozoites; no gametocytes were detected). Some macrophages contained malarial pigment. Giemsa. Bar = 25 µm. Figure 4. Bouin-fixed, paraffin-embedded placenta infected with P. falciparum. Tissue section shows the presence of many infected erythrocytes in the intervillous spaces (small arrow); and hemozoin pigment in erythrocytes, macrophages (arrowhead), and fibrin (large arrow) (Category 2 in the Bulmer classification). HES. Bar = 70 µm.

Negative control slides were processed after incubation with normal nonimmune mouse serum instead of primary antibody. To test the specificity of reactivity of the antibody used, control slides were treated with chondroitinase AC. Slides of infected and uninfected placentas were incubated with 0.1 U of chondroitinase AC (Chondroitinase AC; Sigma-Aldrich, St Quentin Falavier, France) for 30 min at 37C, after which the slides were rinsed several times with PBS and the normal immunohistochemical technique was performed with anti-CSA antibody as primary antibody. Positive control slides were performed from normal frozen skin biopsy specimens with anti-KL-1 antibody (1:50, Immunotech) as primary antibody.

Quantitative Image Analysis
The CSA staining was quantified by image cytometry using a multiparameter image analysis research station (Discovery; Becton-Dickinson, Richmond, CA) equipped with three CDD cameras, one black-and-white camera for low magnification, one for high magnification, and one color CDD camera. For each slide, the fields to be measured were selected by microscopic examination under low magnification using a x10 objective lens, and included terminal villi.

A first surface (Surface 1) (Figure 1B) was delineated under low magnification (x125) by illuminating specimens with monochromatic light (620 nm) suitable to analyze the counterstaining ethyl green, the brown staining of DAB, and the very light background staining in the cytoplasm and the extracellular matrix detectable by image analysis. In the same field, a second surface (Surface 2) (Figure 1C) was delineated with monochromatic light (500 nm), corresponding to the absorption peak of the DAB wavelength (Richmond et al. 1988 ). For both surfaces, a 50-pixel threshold, 10 µm in diameter, was defined in order to discard from the analysis red blood cells or inflammatory cells present in or outside the villi. This process was automatically repeated under high magnification (x500) for 30–40 fields/slide for completion of the analysis. The total number of pixels in the positively immunostained area (Surface 2) was divided by the total number of pixels in the positively counterstained area (Surface 1) and multiplied by 100.

To ensure the reproducibility of the immunoperoxidase staining, the samples were processed twice with 10 slides of infected placentas and 10 slides of noninfected placentas, using identical parameters. To eliminate nonspecific or background immunostaining, a measurement was done in terminal villi. The computer of the multiparameter image analysis research station evaluated the intensity of labeling in this entire area and established a histogram of this intensity. In this histogram, the threshold of intensity was adjusted in Surface 1 so that the light brown stain present across the tissue was eliminated, and in Surface 2 so that intensity of labeling was similar to that of the epidermal area in our control slides.

Statistical Analysis
The relationship between the expression of CSA and the other parameters (infection of the placenta, presence of P. falciparum in imprints, histologic category) was evaluated with the Fisher exact test or the chi-square test. p values less than 0.05 were considered statistically significant. These analyses were performed using the SAS statistical software package (SAS Institute; Carey, NC).

	Results

Top Summary Introduction Materials and Methods Results Discussion Literature Cited

Six (30%) positive imprints (Figure 3) were detected out of the 20 infected placentas (Figure 4). Among these 20 parturients, two (10%) had a positive QBC test but no positive QBC test was recorded for cord blood samples.

Quantitative Image Analysis
Surface 1 assessed the total surface of the villi, including cells and the extracellular matrix except for the capillary lumen (Figure 1A and Figure 1B). Surface 2 corresponds to a cellular or an extracellular area stained by the chromogen (Figure 1A and Figure 1C). The percentage of the surface immunostained for CSA was significantly higher in infected placentas than in noninfected placentas (Fisher test p<0.03) (Table 1; Figure 5). The percentage of the villous area immunostained for CSA was higher in placentas with positive imprints than in those with negative imprints, although nonsignificantly (Fisher test p=0.06). However, this percentage was significantly higher in Categories 1 and 2 than in Category N (Fisher test p<0.05). No significant difference could be found between Category N and Category 3. Considering Categories 1 and 2 as a single entity because both refer to stages of acute infection, it appears that placentas in this 1+2 category have larger CSA surface immunostaining than placentas belonging to Categories N or 3 (Fisher test p<0.05).

View larger version (17K): [in this window] [in a new window]   Figure 5. Average of percentage of surface of terminal villi immunostained by CSA evaluated by quantitative computer image analysis. Category 1, active infection (n = 3); Category 2, active chronic infection (n = 4); Category 3, past chronic infection (n = 13); Category N, noninfected (n = 20) [histological classification from Bulmer et al. 1993 ].

Immunohistochemical Localization of CSA
CSA was detected in the extracellular matrix, in a few endothelial cells, and in stromal cells of the terminal villi in all placentas, whether or not they were infected (Figure 1A and Figure 2). In 11 infected placentas, immunohistochemical examination showed cytoplasmic expression of CSA in syncytiotrophoblasts, and sometimes this expression was membranous (Figure 1A). There was a significant correlation between the positivity of the imprints and the presence of CSA in syncytiotrophoblasts (chi-square test p=0.003). Placentas with active infection (Category 1) and with active chronic infection (Category 2) showed syncytiotrophoblast immunostaining of CSA more frequently than placentas with past chronic infection (Category 3) (Fisher test p<0.02) (Table 1).

The two types of negative control slides, after incubation with normal nonimmune mouse serum instead of primary antibody and after incubation with chondroitinase AC, were both negative.

	Discussion

Top Summary Introduction Materials and Methods Results Discussion Literature Cited

Histological lesions of P. falciparum-induced placentitis are well known (Walter et al. 1982 ). The Bulmer classification (Bulmer et al. 1993 ) captures the progression of infection with P. falciparum. Furthermore, it allows a delineation between active and chronic inflammation, which do not exhibit similar immunologic markers. In active inflammation, the number of parasites in the placenta is usually higher than in the peripheral blood (Sartelet et al. 1997 ). Sequestration of the IRBCs in placenta by syncytiotrophoblasts would therefore play a major role in placental infection with P. falciparum (Bray and Sinden 1979 ), and CSA appears to be a very important element in this mechanism of sequestration (Fried and Duffy 1996 ).

If the amount of the immunoreactivity (i.e., the "antigen"–antibody immune complex formation) in a tissue section is proportional to the amount of antigen in the section, immunohistochemistry is therefore applicable to assess the antigen content in this section (Raivich et al. 1993 ; Watanabe et al. 1994 ). It is noteworthy that the average of staining intensity depends primarily on the number of positive cells, especially in sections that include very few positively stained cells (Matsuo et al. 1995 ; Watanabe et al. 1996 ). In the work reported here, we studied the surface area of the terminal villi (a number of pixels), rather than the intensity (a value for pixels), of the immunostaining. This surface corresponds to two components of the terminal villi: (a) stromal and epithelial cells and (b) the extracellular matrix. The extracellular matrix, however, has been reported to decrease significantly in infected placentas compared to noninfected placentas (Leopardi et al. 1996 ).

We reported two important results which can be helpful in understanding the action of CSA in placentas infected with P. falciparum. First, the expression of CSA in the terminal villi was higher in infected placentas than in noninfected placentas, and in placentas with acute infection than in those with past chronic infection. Two hypotheses can be formulated on the basis of these results: (a) P. falciparum alone can induce the expression of CSA in vivo, as Fried and Duffy 1996 demonstrated in vitro; and (b) the major molecules associated with inflammation, such as TNF, can directly induce the expression of CSA and then increase the expression of CSA in placentas from Categories 1 and 2. However, in vitro studies of syncytiotrophoblasts revealed that TNF did not modify the expression of CSA (Maubert et al. 1997 ).

The second major result of the present study is that CSA was detected in the cytoplasm and in plasma membrane of syncytiotrophoblasts in most infected placentas, and especially in placentas with active inflammation. This suggests that the CSA adhesion molecule detected on the surface of syncytiotrophoblasts, may play a major role in the sequestration of IRBCs in placenta. In fact, the increased surface immunostained for CSA in infected placentas corresponded to an increase in the number of immunostained cells, particularly syncytiotrophoblasts, and was dependent on the intensity and the acuteness of the inflammation, as well as on the presence of IRBCs in the intervillous space.

CSA appears to be directly involved in both the sequestration of IRBCs in the placenta and, particularly, in the acute phase of malarial infection during pregnancy. This research strongly suggests that prevention of maternal malaria can be achieved by means of inhibitors of CSA/IRBC interactions, which could be potentially mediated by either antibody-based therapy or subunit vaccines, as has already been suggested (Rogerson and Brown 1997a ).

	Acknowledgments

We would like to thank the midwives of the Public Health Center of Ouakam and of the Hôpital Principal de Dakar, and particularly Ms Coumba Diop and Ms Carole Todote. We thank Dr G. Michel for support in this work and J.M. Zahm for technical assistance. We also thank Effie Gournis (Yale University; New Haven, CT) for helpful comments and revision of the manuscript.

Received for publication December 1, 1998; accepted January 19, 1999.

	Literature Cited

Top Summary Introduction Materials and Methods Results Discussion Literature Cited

Aikawa M, Iseki M, Barnwell JW, Taylor D, Oo MM, Howard RJ (1990) The pathology of the human cerebral malaria. Am J Trop Med Hyg 43:30-37

Berendt AR, Simmons DL, Tansey J, Newbold CI, Marsh K (1989) Intercellular adhesion molecule-1 is an endothelial cell adhesion receptor for Plasmodium falciparum. Nature 341:57-59[Medline]

Bulmer JN, Rasheed FN, Morrison L, Francis N, Greenwood BM (1993) Placental malaria. I: Pathological classification. Histopathology 22:211-218[Medline]

Bray RS, Sinden RE (1979) The sequestration of Plasmodium infected erythrocytes in the placenta. Trans R Soc Trop Med Hyg 73:716-718[Medline]

Chulay JD, Ockenhouse CF (1990) Host receptors for malaria-infected erythrocytes. Am J Trop Mod Hyg 43(suppl):6-14

Desowitz RS, Buchbinder G (1992) The absence of Plasmodium falciparum gametocytes in the placental blood of a woman with peripheral parasitaemia. Ann Trop Med Parasitol 86:191-192[Medline]

Fried M, Duffy PE (1996) Adherence of Plasmodium falciparum to chondroitin sulfate A in human placenta. Science 272:1502-1504[Abstract]

Leopardi O, Naughten W, Salvia L, Colecchia M, Matteeli A, Zucchi A, Shein A, Muchi JA, Carosi G, Ghione M (1996) Malaric placentas: a quantitative study and clinico-pathological correlation. Pathol Res Pract 192:892-898[Medline]

Levine RA, Wardlaw SC, Patton CL (1989) Detection of haematoparasites using Quantitative Buffy Coat analysis tubes. Parasitol Today 4:132-134

Matsuo T, Mori H, Nishimura Y, Maeda T, Nakagawa JL, Obasi A (1995) Quantification of immunohistochemistry using an image analyser: correlation with hormone concentrations in pituitary adenomas. Histochem J 27:989-996[Medline]

Maubert B, Guilbert LJ, Deloron P (1997) Cytoadherence of Plasmodium falciparum to intercellular adhesion molecule-1 and chondroitin-4 sulfate expressed by the syncytiotrophoblast in the human. Infect Immun 65:1251-1257[Abstract]

Oquendo P, Hundt E, Lawler J, Seed B (1989) CD36 directly mediates cytoadherence of Plasmodium falciparum parasited erythrocytes. Cell 58:95-101[Medline]

Raivich G, Gehrmann J, Graeber MB, Kreutzberg GW (1993) Quantitative immunohistochemistry in the rat facial nucleus with [¹²⁵I]-iodinated secondary antibodies and in situ autoradiography: non-linear binding characteristics of primary monoclonal and polyclonal antibodies. J Histochem Cytochem 41:579-584[Abstract]

Richmond RE, Peirreira MA, Carter HW, Long RE (1988) Quantitative and qualitative immunohistochemical detection of myc and src oncogene proteins in normal, nodular and neoplastic rat liver. J Histochem Cytochem 36:179-184[Abstract]

Robert C, Pouvelle B, Meyer P, Muanza K, Fujitoka H, Aikawa M, Scherf A, Gysin J (1995) Chondroitin-4-sulfate (proteoglycan), a receptor for Plasmodium falciparum-infected erythrocyte adherence on brain microvascular endothelial cells. Res Immunol 146:383-393[Medline]

Rogerson SJ, Brown GV (1997a) Chondroitin sulphate A as an adherence receptor for Plasmodium falsiparum-infected erythrocytes. Parasitol Today 13:70-75[Medline]

Rogerson SJ, Chaiyaroj SC, Ng K, Reeder JC, Brown GV (1995) Chondroitin sulfate A is a cell surface receptor for Plasmodium falciparum infected erythrocytes. J Exp Med 182:15-20[Abstract/Free Full Text]

Rogerson SJ, Novakovic S, Cooke BM, Brown GV (1997b) Plasmodium falciparum-infected erythrocytes adhere to the proteoglycan thrombomodulin in static and flow-base systems. Exp Parasitol 86:8-18[Medline]

Sartelet H, Milko–Sartelet I, Garraud O, Picot S (1997) Plasmodium falciparum persists in the placenta after 3 days' treatment with quinine. Trans R Soc Trop Med Hyg 91:431[Medline]

Sartelet H, Rogier C, Milko–Sartelet I, Michel G, Angel G (1996) Malaria associated pre-eclampsia in Senegal. Lancet 347:1121[Medline]

Trape JF, Zante E, Legros F (1992) Vector density gradients and epidemiology of urban malaria in Dakar, Senegal. Am J Trop Med Hyg 47:181-189

Walter PR, Garin Y, Blot P (1982) Placental pathologic changes in malaria. Am J Pathol 109:330-342[Abstract]

Watanabe J, Asaka Y, Kanamura S (1996) Relationship between immunostaining intensity and antigen content in sections. J Histochem Cytochem 44:1451-1458[Abstract]

Watanabe J, Asaka Y, Tanaka T, Kanamura S (1994) Measurement of NADPH-cytochrome P-450 reductase content in rat liver sections by quantitative immunohistochemistry with a video image processor. J Histochem Cytochem 42:1161-1167[Abstract]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Sartelet, H. Articles by Gaillard, D. Search for Related Content PubMed PubMed Citation Articles by Sartelet, H. Articles by Gaillard, D. Social Bookmarking                      What's this?

Guidelines | Subscriptions | About | exPRESS - Current - Archive | Business Information | Contact